{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.5, "eval_steps": 500, "global_step": 40000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.4154530763626099, "learning_rate": 2.25e-07, "loss": 1.3087, "step": 10 }, { "grad_norm": 1.516983985900879, "learning_rate": 4.75e-07, "loss": 1.3101, "step": 20 }, { "grad_norm": 1.3894928693771362, "learning_rate": 7.25e-07, "loss": 1.3043, "step": 30 }, { "grad_norm": 1.1667842864990234, "learning_rate": 9.75e-07, "loss": 1.3054, "step": 40 }, { "grad_norm": 1.039514422416687, "learning_rate": 1.2250000000000001e-06, "loss": 1.2882, "step": 50 }, { "grad_norm": 0.6983387470245361, "learning_rate": 1.475e-06, "loss": 1.2683, "step": 60 }, { "grad_norm": 0.8124768733978271, "learning_rate": 1.7250000000000002e-06, "loss": 1.249, "step": 70 }, { "grad_norm": 0.3786017894744873, "learning_rate": 1.975e-06, "loss": 1.2317, "step": 80 }, { "grad_norm": 0.46728870272636414, "learning_rate": 2.225e-06, "loss": 1.2225, "step": 90 }, { "grad_norm": 0.3138982355594635, "learning_rate": 2.4750000000000004e-06, "loss": 1.2344, "step": 100 }, { "grad_norm": 0.43287360668182373, "learning_rate": 2.725e-06, "loss": 1.217, "step": 110 }, { "grad_norm": 0.5593885183334351, "learning_rate": 2.975e-06, "loss": 1.2105, "step": 120 }, { "grad_norm": 0.6069411039352417, "learning_rate": 3.225e-06, "loss": 1.167, "step": 130 }, { "grad_norm": 0.595902681350708, "learning_rate": 3.4750000000000006e-06, "loss": 1.1386, "step": 140 }, { "grad_norm": 0.40489286184310913, "learning_rate": 3.725e-06, "loss": 1.1098, "step": 150 }, { "grad_norm": 0.4490557312965393, "learning_rate": 3.975e-06, "loss": 1.0922, "step": 160 }, { "grad_norm": 0.6013497114181519, "learning_rate": 4.225e-06, "loss": 1.087, "step": 170 }, { "grad_norm": 0.5094527006149292, "learning_rate": 4.475e-06, "loss": 1.0827, "step": 180 }, { "grad_norm": 0.5680581331253052, "learning_rate": 4.7250000000000005e-06, "loss": 1.0756, "step": 190 }, { "grad_norm": 0.7737853527069092, "learning_rate": 4.975000000000001e-06, "loss": 1.0698, "step": 200 }, { "grad_norm": 0.6116843819618225, "learning_rate": 5.225e-06, "loss": 1.0614, "step": 210 }, { "grad_norm": 0.49338728189468384, "learning_rate": 5.475e-06, "loss": 1.0606, "step": 220 }, { "grad_norm": 0.5798864364624023, "learning_rate": 5.725e-06, "loss": 1.048, "step": 230 }, { "grad_norm": 0.4625064432621002, "learning_rate": 5.975e-06, "loss": 1.0468, "step": 240 }, { "grad_norm": 0.5644728541374207, "learning_rate": 6.2250000000000005e-06, "loss": 1.042, "step": 250 }, { "grad_norm": 0.5585410594940186, "learning_rate": 6.475000000000001e-06, "loss": 1.049, "step": 260 }, { "grad_norm": 0.5381026268005371, "learning_rate": 6.725000000000001e-06, "loss": 1.0436, "step": 270 }, { "grad_norm": 0.48326992988586426, "learning_rate": 6.975000000000001e-06, "loss": 1.0427, "step": 280 }, { "grad_norm": 0.5189052224159241, "learning_rate": 7.2249999999999994e-06, "loss": 1.0347, "step": 290 }, { "grad_norm": 0.49118053913116455, "learning_rate": 7.4750000000000004e-06, "loss": 1.0365, "step": 300 }, { "grad_norm": 0.5370896458625793, "learning_rate": 7.725e-06, "loss": 1.0345, "step": 310 }, { "grad_norm": 0.4393181800842285, "learning_rate": 7.975e-06, "loss": 1.0382, "step": 320 }, { "grad_norm": 0.47321298718452454, "learning_rate": 8.225e-06, "loss": 1.033, "step": 330 }, { "grad_norm": 0.42894235253334045, "learning_rate": 8.475000000000001e-06, "loss": 1.0255, "step": 340 }, { "grad_norm": 0.5298031568527222, "learning_rate": 8.725e-06, "loss": 1.0256, "step": 350 }, { "grad_norm": 1.1084742546081543, "learning_rate": 8.975e-06, "loss": 1.0038, "step": 360 }, { "grad_norm": 1.0837687253952026, "learning_rate": 9.225e-06, "loss": 0.9783, "step": 370 }, { "grad_norm": 1.1430495977401733, "learning_rate": 9.475e-06, "loss": 0.9609, "step": 380 }, { "grad_norm": 1.2151986360549927, "learning_rate": 9.725000000000001e-06, "loss": 0.9435, "step": 390 }, { "grad_norm": 1.6320743560791016, "learning_rate": 9.975e-06, "loss": 0.9096, "step": 400 }, { "grad_norm": 1.1177517175674438, "learning_rate": 1.0225e-05, "loss": 0.8769, "step": 410 }, { "grad_norm": 1.6174978017807007, "learning_rate": 1.0475e-05, "loss": 0.8349, "step": 420 }, { "grad_norm": 1.4267441034317017, "learning_rate": 1.0725e-05, "loss": 0.8267, "step": 430 }, { "grad_norm": 1.6050726175308228, "learning_rate": 1.0975e-05, "loss": 0.7912, "step": 440 }, { "grad_norm": 1.4691901206970215, "learning_rate": 1.1225e-05, "loss": 0.765, "step": 450 }, { "grad_norm": 1.6230883598327637, "learning_rate": 1.1475000000000001e-05, "loss": 0.7344, "step": 460 }, { "grad_norm": 1.442430853843689, "learning_rate": 1.1725e-05, "loss": 0.7121, "step": 470 }, { "grad_norm": 1.8194056749343872, "learning_rate": 1.1975e-05, "loss": 0.6792, "step": 480 }, { "grad_norm": 1.4651265144348145, "learning_rate": 1.2225e-05, "loss": 0.6453, "step": 490 }, { "grad_norm": 1.9267743825912476, "learning_rate": 1.2475e-05, "loss": 0.6129, "step": 500 }, { "grad_norm": 2.049866199493408, "learning_rate": 1.2725000000000001e-05, "loss": 0.5753, "step": 510 }, { "grad_norm": 1.836199402809143, "learning_rate": 1.2975e-05, "loss": 0.5422, "step": 520 }, { "grad_norm": 2.2650907039642334, "learning_rate": 1.3225000000000001e-05, "loss": 0.5148, "step": 530 }, { "grad_norm": 2.037248373031616, "learning_rate": 1.3475000000000002e-05, "loss": 0.4985, "step": 540 }, { "grad_norm": 1.4967652559280396, "learning_rate": 1.3725000000000002e-05, "loss": 0.4831, "step": 550 }, { "grad_norm": 1.734890103340149, "learning_rate": 1.3975000000000003e-05, "loss": 0.4773, "step": 560 }, { "grad_norm": 1.8290252685546875, "learning_rate": 1.4225e-05, "loss": 0.4694, "step": 570 }, { "grad_norm": 2.0998547077178955, "learning_rate": 1.4475e-05, "loss": 0.445, "step": 580 }, { "grad_norm": 1.9377562999725342, "learning_rate": 1.4725e-05, "loss": 0.4233, "step": 590 }, { "grad_norm": 1.791024923324585, "learning_rate": 1.4975e-05, "loss": 0.4118, "step": 600 }, { "grad_norm": 1.9979017972946167, "learning_rate": 1.5225e-05, "loss": 0.3786, "step": 610 }, { "grad_norm": 1.9694970846176147, "learning_rate": 1.5475e-05, "loss": 0.3618, "step": 620 }, { "grad_norm": 2.30204176902771, "learning_rate": 1.5725e-05, "loss": 0.33, "step": 630 }, { "grad_norm": 1.9860674142837524, "learning_rate": 1.5975000000000002e-05, "loss": 0.3007, "step": 640 }, { "grad_norm": 2.062673807144165, "learning_rate": 1.6225e-05, "loss": 0.2911, "step": 650 }, { "grad_norm": 1.8429611921310425, "learning_rate": 1.6475e-05, "loss": 0.2804, "step": 660 }, { "grad_norm": 1.5752352476119995, "learning_rate": 1.6725000000000003e-05, "loss": 0.2546, "step": 670 }, { "grad_norm": 2.2148196697235107, "learning_rate": 1.6975000000000003e-05, "loss": 0.2333, "step": 680 }, { "grad_norm": 2.381422281265259, "learning_rate": 1.7225e-05, "loss": 0.2293, "step": 690 }, { "grad_norm": 1.586865782737732, "learning_rate": 1.7475e-05, "loss": 0.2156, "step": 700 }, { "grad_norm": 2.8625845909118652, "learning_rate": 1.7725e-05, "loss": 0.2191, "step": 710 }, { "grad_norm": 1.9748612642288208, "learning_rate": 1.7975e-05, "loss": 0.1885, "step": 720 }, { "grad_norm": 2.229268789291382, "learning_rate": 1.8225e-05, "loss": 0.1551, "step": 730 }, { "grad_norm": 1.7495851516723633, "learning_rate": 1.8475000000000002e-05, "loss": 0.163, "step": 740 }, { "grad_norm": 2.3251953125, "learning_rate": 1.8725e-05, "loss": 0.1622, "step": 750 }, { "grad_norm": 1.8665645122528076, "learning_rate": 1.8975e-05, "loss": 0.1606, "step": 760 }, { "grad_norm": 1.2769687175750732, "learning_rate": 1.9225e-05, "loss": 0.1458, "step": 770 }, { "grad_norm": 2.189958095550537, "learning_rate": 1.9475000000000002e-05, "loss": 0.1445, "step": 780 }, { "grad_norm": 1.8894394636154175, "learning_rate": 1.9725000000000002e-05, "loss": 0.1412, "step": 790 }, { "grad_norm": 1.895442008972168, "learning_rate": 1.9975e-05, "loss": 0.1408, "step": 800 }, { "grad_norm": 1.458067536354065, "learning_rate": 2.0225000000000004e-05, "loss": 0.1367, "step": 810 }, { "grad_norm": 1.8932210206985474, "learning_rate": 2.0475e-05, "loss": 0.1365, "step": 820 }, { "grad_norm": 1.8285574913024902, "learning_rate": 2.0725e-05, "loss": 0.1315, "step": 830 }, { "grad_norm": 1.4206339120864868, "learning_rate": 2.0975e-05, "loss": 0.1241, "step": 840 }, { "grad_norm": 1.371285080909729, "learning_rate": 2.1225e-05, "loss": 0.1277, "step": 850 }, { "grad_norm": 1.50456702709198, "learning_rate": 2.1475e-05, "loss": 0.1319, "step": 860 }, { "grad_norm": 2.002007007598877, "learning_rate": 2.1725e-05, "loss": 0.1226, "step": 870 }, { "grad_norm": 1.886924386024475, "learning_rate": 2.1975000000000002e-05, "loss": 0.1273, "step": 880 }, { "grad_norm": 1.5372599363327026, "learning_rate": 2.2225e-05, "loss": 0.1174, "step": 890 }, { "grad_norm": 2.1205878257751465, "learning_rate": 2.2475e-05, "loss": 0.1074, "step": 900 }, { "grad_norm": 1.3410552740097046, "learning_rate": 2.2725000000000003e-05, "loss": 0.1045, "step": 910 }, { "grad_norm": 1.5549544095993042, "learning_rate": 2.2975000000000003e-05, "loss": 0.1085, "step": 920 }, { "grad_norm": 1.4378851652145386, "learning_rate": 2.3225000000000002e-05, "loss": 0.1104, "step": 930 }, { "grad_norm": 1.5319527387619019, "learning_rate": 2.3475e-05, "loss": 0.1047, "step": 940 }, { "grad_norm": 1.5329798460006714, "learning_rate": 2.3725e-05, "loss": 0.1079, "step": 950 }, { "grad_norm": 1.4354236125946045, "learning_rate": 2.3975e-05, "loss": 0.1078, "step": 960 }, { "grad_norm": 1.8273038864135742, "learning_rate": 2.4225e-05, "loss": 0.112, "step": 970 }, { "grad_norm": 1.710262656211853, "learning_rate": 2.4475000000000002e-05, "loss": 0.1068, "step": 980 }, { "grad_norm": 1.5799199342727661, "learning_rate": 2.4725e-05, "loss": 0.109, "step": 990 }, { "grad_norm": 1.6639446020126343, "learning_rate": 2.4975e-05, "loss": 0.113, "step": 1000 }, { "grad_norm": 1.4148406982421875, "learning_rate": 2.5225e-05, "loss": 0.1037, "step": 1010 }, { "grad_norm": 1.0834399461746216, "learning_rate": 2.5475e-05, "loss": 0.106, "step": 1020 }, { "grad_norm": 1.2255778312683105, "learning_rate": 2.5725e-05, "loss": 0.1119, "step": 1030 }, { "grad_norm": 1.2928969860076904, "learning_rate": 2.5974999999999998e-05, "loss": 0.1025, "step": 1040 }, { "grad_norm": 1.4941201210021973, "learning_rate": 2.6225e-05, "loss": 0.095, "step": 1050 }, { "grad_norm": 1.6863523721694946, "learning_rate": 2.6475e-05, "loss": 0.0944, "step": 1060 }, { "grad_norm": 1.1537200212478638, "learning_rate": 2.6725e-05, "loss": 0.0989, "step": 1070 }, { "grad_norm": 1.475977897644043, "learning_rate": 2.6975000000000002e-05, "loss": 0.0898, "step": 1080 }, { "grad_norm": 1.4176764488220215, "learning_rate": 2.7225e-05, "loss": 0.0931, "step": 1090 }, { "grad_norm": 1.6688342094421387, "learning_rate": 2.7475e-05, "loss": 0.102, "step": 1100 }, { "grad_norm": 1.1329275369644165, "learning_rate": 2.7725e-05, "loss": 0.0844, "step": 1110 }, { "grad_norm": 1.6303564310073853, "learning_rate": 2.7975000000000002e-05, "loss": 0.1, "step": 1120 }, { "grad_norm": 1.2959305047988892, "learning_rate": 2.8225e-05, "loss": 0.0981, "step": 1130 }, { "grad_norm": 1.7076715230941772, "learning_rate": 2.8475e-05, "loss": 0.0953, "step": 1140 }, { "grad_norm": 1.6095266342163086, "learning_rate": 2.8725e-05, "loss": 0.1014, "step": 1150 }, { "grad_norm": 1.2872703075408936, "learning_rate": 2.8975000000000003e-05, "loss": 0.0917, "step": 1160 }, { "grad_norm": 1.5120097398757935, "learning_rate": 2.9225000000000002e-05, "loss": 0.0946, "step": 1170 }, { "grad_norm": 1.444676160812378, "learning_rate": 2.9475e-05, "loss": 0.0946, "step": 1180 }, { "grad_norm": 1.4531761407852173, "learning_rate": 2.9725000000000004e-05, "loss": 0.0898, "step": 1190 }, { "grad_norm": 1.0419379472732544, "learning_rate": 2.9975000000000004e-05, "loss": 0.0922, "step": 1200 }, { "grad_norm": 1.229286789894104, "learning_rate": 3.0225000000000003e-05, "loss": 0.0906, "step": 1210 }, { "grad_norm": 0.9782979488372803, "learning_rate": 3.0475000000000002e-05, "loss": 0.0806, "step": 1220 }, { "grad_norm": 1.2256121635437012, "learning_rate": 3.0725e-05, "loss": 0.0835, "step": 1230 }, { "grad_norm": 1.2048929929733276, "learning_rate": 3.0975e-05, "loss": 0.0931, "step": 1240 }, { "grad_norm": 1.3468683958053589, "learning_rate": 3.122500000000001e-05, "loss": 0.0915, "step": 1250 }, { "grad_norm": 1.3145710229873657, "learning_rate": 3.1475e-05, "loss": 0.0745, "step": 1260 }, { "grad_norm": 1.2712877988815308, "learning_rate": 3.1725e-05, "loss": 0.0964, "step": 1270 }, { "grad_norm": 1.6587239503860474, "learning_rate": 3.1975e-05, "loss": 0.0875, "step": 1280 }, { "grad_norm": 1.4921205043792725, "learning_rate": 3.2225e-05, "loss": 0.0881, "step": 1290 }, { "grad_norm": 1.1982781887054443, "learning_rate": 3.2474999999999997e-05, "loss": 0.0877, "step": 1300 }, { "grad_norm": 1.243086338043213, "learning_rate": 3.2725e-05, "loss": 0.0809, "step": 1310 }, { "grad_norm": 1.320361852645874, "learning_rate": 3.2975e-05, "loss": 0.0977, "step": 1320 }, { "grad_norm": 1.2380154132843018, "learning_rate": 3.3225e-05, "loss": 0.0792, "step": 1330 }, { "grad_norm": 1.3226662874221802, "learning_rate": 3.3475e-05, "loss": 0.0738, "step": 1340 }, { "grad_norm": 1.6450095176696777, "learning_rate": 3.3725e-05, "loss": 0.0879, "step": 1350 }, { "grad_norm": 1.2922295331954956, "learning_rate": 3.3975e-05, "loss": 0.0896, "step": 1360 }, { "grad_norm": 1.1081315279006958, "learning_rate": 3.4225e-05, "loss": 0.0795, "step": 1370 }, { "grad_norm": 1.2442125082015991, "learning_rate": 3.4475000000000005e-05, "loss": 0.0802, "step": 1380 }, { "grad_norm": 0.9677568674087524, "learning_rate": 3.4725000000000004e-05, "loss": 0.0718, "step": 1390 }, { "grad_norm": 1.0887181758880615, "learning_rate": 3.4975e-05, "loss": 0.075, "step": 1400 }, { "grad_norm": 1.3916051387786865, "learning_rate": 3.5225e-05, "loss": 0.0758, "step": 1410 }, { "grad_norm": 0.9190086722373962, "learning_rate": 3.5475e-05, "loss": 0.0754, "step": 1420 }, { "grad_norm": 1.3555165529251099, "learning_rate": 3.5725e-05, "loss": 0.0888, "step": 1430 }, { "grad_norm": 1.1770647764205933, "learning_rate": 3.5975e-05, "loss": 0.0786, "step": 1440 }, { "grad_norm": 0.6849499940872192, "learning_rate": 3.6225000000000006e-05, "loss": 0.073, "step": 1450 }, { "grad_norm": 1.071390151977539, "learning_rate": 3.6475000000000006e-05, "loss": 0.0807, "step": 1460 }, { "grad_norm": 1.3181883096694946, "learning_rate": 3.6725000000000005e-05, "loss": 0.0718, "step": 1470 }, { "grad_norm": 1.2026164531707764, "learning_rate": 3.6975000000000004e-05, "loss": 0.0726, "step": 1480 }, { "grad_norm": 1.3269777297973633, "learning_rate": 3.7225000000000004e-05, "loss": 0.0811, "step": 1490 }, { "grad_norm": 1.2252689599990845, "learning_rate": 3.7475e-05, "loss": 0.0737, "step": 1500 }, { "grad_norm": 1.0553340911865234, "learning_rate": 3.7725e-05, "loss": 0.0751, "step": 1510 }, { "grad_norm": 1.1153234243392944, "learning_rate": 3.7975e-05, "loss": 0.0768, "step": 1520 }, { "grad_norm": 0.9873115420341492, "learning_rate": 3.8225e-05, "loss": 0.0719, "step": 1530 }, { "grad_norm": 1.1770172119140625, "learning_rate": 3.8475e-05, "loss": 0.0757, "step": 1540 }, { "grad_norm": 1.0055499076843262, "learning_rate": 3.8725e-05, "loss": 0.0758, "step": 1550 }, { "grad_norm": 0.9166172742843628, "learning_rate": 3.8975e-05, "loss": 0.0769, "step": 1560 }, { "grad_norm": 1.3790513277053833, "learning_rate": 3.9225e-05, "loss": 0.0784, "step": 1570 }, { "grad_norm": 0.9946162104606628, "learning_rate": 3.9475000000000004e-05, "loss": 0.0701, "step": 1580 }, { "grad_norm": 1.1103980541229248, "learning_rate": 3.9725e-05, "loss": 0.0714, "step": 1590 }, { "grad_norm": 0.8524609804153442, "learning_rate": 3.9975e-05, "loss": 0.0756, "step": 1600 }, { "grad_norm": 0.8936760425567627, "learning_rate": 4.0225e-05, "loss": 0.0692, "step": 1610 }, { "grad_norm": 0.9490578174591064, "learning_rate": 4.0475e-05, "loss": 0.0693, "step": 1620 }, { "grad_norm": 1.1713149547576904, "learning_rate": 4.0725e-05, "loss": 0.077, "step": 1630 }, { "grad_norm": 1.270061731338501, "learning_rate": 4.0975e-05, "loss": 0.0761, "step": 1640 }, { "grad_norm": 1.045929193496704, "learning_rate": 4.1225e-05, "loss": 0.0685, "step": 1650 }, { "grad_norm": 1.0888994932174683, "learning_rate": 4.1475000000000005e-05, "loss": 0.0675, "step": 1660 }, { "grad_norm": 0.9357777833938599, "learning_rate": 4.1725000000000005e-05, "loss": 0.0752, "step": 1670 }, { "grad_norm": 1.1290347576141357, "learning_rate": 4.1975000000000004e-05, "loss": 0.068, "step": 1680 }, { "grad_norm": 1.1104294061660767, "learning_rate": 4.2225e-05, "loss": 0.0691, "step": 1690 }, { "grad_norm": 0.961200475692749, "learning_rate": 4.2475e-05, "loss": 0.0789, "step": 1700 }, { "grad_norm": 0.9760602116584778, "learning_rate": 4.2725e-05, "loss": 0.0677, "step": 1710 }, { "grad_norm": 0.9489273428916931, "learning_rate": 4.2975e-05, "loss": 0.0704, "step": 1720 }, { "grad_norm": 0.7732637524604797, "learning_rate": 4.322500000000001e-05, "loss": 0.0709, "step": 1730 }, { "grad_norm": 0.7987691760063171, "learning_rate": 4.3475000000000006e-05, "loss": 0.0665, "step": 1740 }, { "grad_norm": 1.0278087854385376, "learning_rate": 4.3725000000000006e-05, "loss": 0.0658, "step": 1750 }, { "grad_norm": 1.0088249444961548, "learning_rate": 4.3975e-05, "loss": 0.0632, "step": 1760 }, { "grad_norm": 1.015378475189209, "learning_rate": 4.4225e-05, "loss": 0.08, "step": 1770 }, { "grad_norm": 0.8454896211624146, "learning_rate": 4.4475e-05, "loss": 0.0704, "step": 1780 }, { "grad_norm": 0.8381644487380981, "learning_rate": 4.4725e-05, "loss": 0.0661, "step": 1790 }, { "grad_norm": 1.060675024986267, "learning_rate": 4.4975e-05, "loss": 0.0736, "step": 1800 }, { "grad_norm": 0.9768335819244385, "learning_rate": 4.5225e-05, "loss": 0.0679, "step": 1810 }, { "grad_norm": 1.0793521404266357, "learning_rate": 4.5475e-05, "loss": 0.0704, "step": 1820 }, { "grad_norm": 0.856987476348877, "learning_rate": 4.5725e-05, "loss": 0.0667, "step": 1830 }, { "grad_norm": 0.7979984283447266, "learning_rate": 4.5975e-05, "loss": 0.062, "step": 1840 }, { "grad_norm": 0.7637870907783508, "learning_rate": 4.6225e-05, "loss": 0.063, "step": 1850 }, { "grad_norm": 0.9637273550033569, "learning_rate": 4.6475000000000005e-05, "loss": 0.0628, "step": 1860 }, { "grad_norm": 0.8197232484817505, "learning_rate": 4.6725000000000004e-05, "loss": 0.0637, "step": 1870 }, { "grad_norm": 0.8717010021209717, "learning_rate": 4.6975000000000003e-05, "loss": 0.0606, "step": 1880 }, { "grad_norm": 0.8143038749694824, "learning_rate": 4.7225e-05, "loss": 0.0615, "step": 1890 }, { "grad_norm": 0.8906691074371338, "learning_rate": 4.7475e-05, "loss": 0.0663, "step": 1900 }, { "grad_norm": 0.8654858469963074, "learning_rate": 4.7725e-05, "loss": 0.0651, "step": 1910 }, { "grad_norm": 0.8163949847221375, "learning_rate": 4.7975e-05, "loss": 0.059, "step": 1920 }, { "grad_norm": 0.8157826066017151, "learning_rate": 4.822500000000001e-05, "loss": 0.0611, "step": 1930 }, { "grad_norm": 0.7930881977081299, "learning_rate": 4.8475000000000006e-05, "loss": 0.0752, "step": 1940 }, { "grad_norm": 1.3163224458694458, "learning_rate": 4.8725000000000005e-05, "loss": 0.0623, "step": 1950 }, { "grad_norm": 0.854286253452301, "learning_rate": 4.8975000000000005e-05, "loss": 0.0628, "step": 1960 }, { "grad_norm": 0.8835042715072632, "learning_rate": 4.9225000000000004e-05, "loss": 0.0631, "step": 1970 }, { "grad_norm": 0.9762257933616638, "learning_rate": 4.9475e-05, "loss": 0.0687, "step": 1980 }, { "grad_norm": 0.5900201201438904, "learning_rate": 4.9725e-05, "loss": 0.0631, "step": 1990 }, { "grad_norm": 0.8055667877197266, "learning_rate": 4.9975e-05, "loss": 0.0614, "step": 2000 }, { "grad_norm": 0.8474117517471313, "learning_rate": 5.0225e-05, "loss": 0.0641, "step": 2010 }, { "grad_norm": 1.122346043586731, "learning_rate": 5.047500000000001e-05, "loss": 0.0627, "step": 2020 }, { "grad_norm": 0.6619302034378052, "learning_rate": 5.0725e-05, "loss": 0.0592, "step": 2030 }, { "grad_norm": 0.9135761260986328, "learning_rate": 5.0975000000000006e-05, "loss": 0.064, "step": 2040 }, { "grad_norm": 0.8559839725494385, "learning_rate": 5.1225e-05, "loss": 0.0646, "step": 2050 }, { "grad_norm": 0.8072445392608643, "learning_rate": 5.1475000000000004e-05, "loss": 0.0643, "step": 2060 }, { "grad_norm": 0.9622520804405212, "learning_rate": 5.1725000000000004e-05, "loss": 0.064, "step": 2070 }, { "grad_norm": 0.7415013313293457, "learning_rate": 5.197500000000001e-05, "loss": 0.0671, "step": 2080 }, { "grad_norm": 0.7717751860618591, "learning_rate": 5.2225e-05, "loss": 0.0658, "step": 2090 }, { "grad_norm": 0.843674898147583, "learning_rate": 5.247500000000001e-05, "loss": 0.0596, "step": 2100 }, { "grad_norm": 1.0089781284332275, "learning_rate": 5.2725e-05, "loss": 0.0615, "step": 2110 }, { "grad_norm": 0.8794126510620117, "learning_rate": 5.297500000000001e-05, "loss": 0.0607, "step": 2120 }, { "grad_norm": 1.0187033414840698, "learning_rate": 5.3225e-05, "loss": 0.0648, "step": 2130 }, { "grad_norm": 0.8640639185905457, "learning_rate": 5.3475e-05, "loss": 0.0636, "step": 2140 }, { "grad_norm": 0.9254222512245178, "learning_rate": 5.3725000000000005e-05, "loss": 0.0628, "step": 2150 }, { "grad_norm": 0.90449458360672, "learning_rate": 5.3975e-05, "loss": 0.0521, "step": 2160 }, { "grad_norm": 0.777798593044281, "learning_rate": 5.4225000000000003e-05, "loss": 0.0633, "step": 2170 }, { "grad_norm": 0.9453645348548889, "learning_rate": 5.4474999999999996e-05, "loss": 0.0618, "step": 2180 }, { "grad_norm": 0.753568172454834, "learning_rate": 5.4725e-05, "loss": 0.0605, "step": 2190 }, { "grad_norm": 0.7776861786842346, "learning_rate": 5.4975e-05, "loss": 0.0553, "step": 2200 }, { "grad_norm": 0.7204363346099854, "learning_rate": 5.522500000000001e-05, "loss": 0.0529, "step": 2210 }, { "grad_norm": 0.9355034828186035, "learning_rate": 5.5475e-05, "loss": 0.0567, "step": 2220 }, { "grad_norm": 0.724220335483551, "learning_rate": 5.5725000000000006e-05, "loss": 0.0581, "step": 2230 }, { "grad_norm": 0.7577541470527649, "learning_rate": 5.5975e-05, "loss": 0.0601, "step": 2240 }, { "grad_norm": 0.6884009838104248, "learning_rate": 5.6225000000000005e-05, "loss": 0.0681, "step": 2250 }, { "grad_norm": 0.87007075548172, "learning_rate": 5.6475e-05, "loss": 0.0669, "step": 2260 }, { "grad_norm": 0.8507005572319031, "learning_rate": 5.6725e-05, "loss": 0.0691, "step": 2270 }, { "grad_norm": 0.6103546619415283, "learning_rate": 5.6975e-05, "loss": 0.0604, "step": 2280 }, { "grad_norm": 0.7729164958000183, "learning_rate": 5.722500000000001e-05, "loss": 0.0565, "step": 2290 }, { "grad_norm": 0.7087283730506897, "learning_rate": 5.7475e-05, "loss": 0.0588, "step": 2300 }, { "grad_norm": 0.7450974583625793, "learning_rate": 5.772500000000001e-05, "loss": 0.0593, "step": 2310 }, { "grad_norm": 0.7838680744171143, "learning_rate": 5.7975e-05, "loss": 0.0604, "step": 2320 }, { "grad_norm": 0.886369526386261, "learning_rate": 5.8225000000000006e-05, "loss": 0.0649, "step": 2330 }, { "grad_norm": 0.6738531589508057, "learning_rate": 5.8475000000000005e-05, "loss": 0.0617, "step": 2340 }, { "grad_norm": 0.8939045071601868, "learning_rate": 5.8725000000000004e-05, "loss": 0.0564, "step": 2350 }, { "grad_norm": 0.9854339957237244, "learning_rate": 5.8975000000000004e-05, "loss": 0.0643, "step": 2360 }, { "grad_norm": 0.7381295561790466, "learning_rate": 5.922500000000001e-05, "loss": 0.06, "step": 2370 }, { "grad_norm": 0.7221106290817261, "learning_rate": 5.9475e-05, "loss": 0.0611, "step": 2380 }, { "grad_norm": 0.7176597118377686, "learning_rate": 5.9724999999999995e-05, "loss": 0.0571, "step": 2390 }, { "grad_norm": 0.5730379819869995, "learning_rate": 5.9975e-05, "loss": 0.0587, "step": 2400 }, { "grad_norm": 0.7387498617172241, "learning_rate": 6.0225e-05, "loss": 0.0526, "step": 2410 }, { "grad_norm": 0.6399484872817993, "learning_rate": 6.0475000000000006e-05, "loss": 0.0569, "step": 2420 }, { "grad_norm": 0.6622423529624939, "learning_rate": 6.0725e-05, "loss": 0.0555, "step": 2430 }, { "grad_norm": 0.8700543642044067, "learning_rate": 6.0975000000000005e-05, "loss": 0.0563, "step": 2440 }, { "grad_norm": 0.7052351236343384, "learning_rate": 6.1225e-05, "loss": 0.0594, "step": 2450 }, { "grad_norm": 0.663169801235199, "learning_rate": 6.1475e-05, "loss": 0.0632, "step": 2460 }, { "grad_norm": 0.4682319164276123, "learning_rate": 6.1725e-05, "loss": 0.0527, "step": 2470 }, { "grad_norm": 0.7458738684654236, "learning_rate": 6.1975e-05, "loss": 0.0615, "step": 2480 }, { "grad_norm": 0.6027103662490845, "learning_rate": 6.2225e-05, "loss": 0.0543, "step": 2490 }, { "grad_norm": 0.8132599592208862, "learning_rate": 6.2475e-05, "loss": 0.0622, "step": 2500 }, { "grad_norm": 0.6589983105659485, "learning_rate": 6.2725e-05, "loss": 0.0604, "step": 2510 }, { "grad_norm": 0.6678787469863892, "learning_rate": 6.297500000000001e-05, "loss": 0.0566, "step": 2520 }, { "grad_norm": 0.8375526666641235, "learning_rate": 6.3225e-05, "loss": 0.0561, "step": 2530 }, { "grad_norm": 0.5772702097892761, "learning_rate": 6.347500000000001e-05, "loss": 0.0539, "step": 2540 }, { "grad_norm": 0.8374665975570679, "learning_rate": 6.3725e-05, "loss": 0.0603, "step": 2550 }, { "grad_norm": 0.633335292339325, "learning_rate": 6.397500000000001e-05, "loss": 0.0525, "step": 2560 }, { "grad_norm": 0.7409762740135193, "learning_rate": 6.4225e-05, "loss": 0.0538, "step": 2570 }, { "grad_norm": 0.7259706258773804, "learning_rate": 6.447500000000001e-05, "loss": 0.0601, "step": 2580 }, { "grad_norm": 0.7602354288101196, "learning_rate": 6.4725e-05, "loss": 0.0536, "step": 2590 }, { "grad_norm": 0.769178032875061, "learning_rate": 6.497500000000001e-05, "loss": 0.0583, "step": 2600 }, { "grad_norm": 0.8308601379394531, "learning_rate": 6.5225e-05, "loss": 0.0559, "step": 2610 }, { "grad_norm": 0.7161239981651306, "learning_rate": 6.5475e-05, "loss": 0.0521, "step": 2620 }, { "grad_norm": 0.8504297137260437, "learning_rate": 6.5725e-05, "loss": 0.0579, "step": 2630 }, { "grad_norm": 0.486929327249527, "learning_rate": 6.5975e-05, "loss": 0.0558, "step": 2640 }, { "grad_norm": 0.7214643955230713, "learning_rate": 6.6225e-05, "loss": 0.0554, "step": 2650 }, { "grad_norm": 0.6431878209114075, "learning_rate": 6.6475e-05, "loss": 0.053, "step": 2660 }, { "grad_norm": 0.6826710104942322, "learning_rate": 6.672500000000001e-05, "loss": 0.0557, "step": 2670 }, { "grad_norm": 0.8162904381752014, "learning_rate": 6.6975e-05, "loss": 0.0515, "step": 2680 }, { "grad_norm": 0.6216828227043152, "learning_rate": 6.722500000000001e-05, "loss": 0.058, "step": 2690 }, { "grad_norm": 0.6386335492134094, "learning_rate": 6.7475e-05, "loss": 0.0573, "step": 2700 }, { "grad_norm": 0.8852646350860596, "learning_rate": 6.7725e-05, "loss": 0.0619, "step": 2710 }, { "grad_norm": 0.6463866233825684, "learning_rate": 6.7975e-05, "loss": 0.0537, "step": 2720 }, { "grad_norm": 0.8776310682296753, "learning_rate": 6.8225e-05, "loss": 0.0572, "step": 2730 }, { "grad_norm": 0.6673681735992432, "learning_rate": 6.8475e-05, "loss": 0.0582, "step": 2740 }, { "grad_norm": 0.6958160996437073, "learning_rate": 6.8725e-05, "loss": 0.0553, "step": 2750 }, { "grad_norm": 0.6443694829940796, "learning_rate": 6.8975e-05, "loss": 0.0582, "step": 2760 }, { "grad_norm": 0.6842501759529114, "learning_rate": 6.9225e-05, "loss": 0.0498, "step": 2770 }, { "grad_norm": 0.5554088354110718, "learning_rate": 6.9475e-05, "loss": 0.0519, "step": 2780 }, { "grad_norm": 0.5094532370567322, "learning_rate": 6.9725e-05, "loss": 0.0549, "step": 2790 }, { "grad_norm": 0.729341983795166, "learning_rate": 6.997500000000001e-05, "loss": 0.05, "step": 2800 }, { "grad_norm": 0.6997498869895935, "learning_rate": 7.022500000000001e-05, "loss": 0.0485, "step": 2810 }, { "grad_norm": 0.6703678965568542, "learning_rate": 7.0475e-05, "loss": 0.0461, "step": 2820 }, { "grad_norm": 0.6599840521812439, "learning_rate": 7.072500000000001e-05, "loss": 0.0554, "step": 2830 }, { "grad_norm": 0.5110449194908142, "learning_rate": 7.0975e-05, "loss": 0.0539, "step": 2840 }, { "grad_norm": 0.4814050495624542, "learning_rate": 7.122500000000001e-05, "loss": 0.0496, "step": 2850 }, { "grad_norm": 0.732448399066925, "learning_rate": 7.1475e-05, "loss": 0.0505, "step": 2860 }, { "grad_norm": 0.49642619490623474, "learning_rate": 7.172500000000001e-05, "loss": 0.0495, "step": 2870 }, { "grad_norm": 0.6120973825454712, "learning_rate": 7.1975e-05, "loss": 0.0505, "step": 2880 }, { "grad_norm": 0.474577933549881, "learning_rate": 7.2225e-05, "loss": 0.048, "step": 2890 }, { "grad_norm": 0.6242506504058838, "learning_rate": 7.2475e-05, "loss": 0.0497, "step": 2900 }, { "grad_norm": 0.6373989582061768, "learning_rate": 7.272499999999999e-05, "loss": 0.051, "step": 2910 }, { "grad_norm": 0.8021920919418335, "learning_rate": 7.2975e-05, "loss": 0.0495, "step": 2920 }, { "grad_norm": 0.6013539433479309, "learning_rate": 7.3225e-05, "loss": 0.0548, "step": 2930 }, { "grad_norm": 0.6601709127426147, "learning_rate": 7.347500000000001e-05, "loss": 0.0572, "step": 2940 }, { "grad_norm": 0.5770432353019714, "learning_rate": 7.3725e-05, "loss": 0.0538, "step": 2950 }, { "grad_norm": 0.557041347026825, "learning_rate": 7.397500000000001e-05, "loss": 0.0583, "step": 2960 }, { "grad_norm": 0.6183140277862549, "learning_rate": 7.4225e-05, "loss": 0.0499, "step": 2970 }, { "grad_norm": 0.8379376530647278, "learning_rate": 7.447500000000001e-05, "loss": 0.0487, "step": 2980 }, { "grad_norm": 0.6987119317054749, "learning_rate": 7.4725e-05, "loss": 0.0485, "step": 2990 }, { "grad_norm": 0.5286829471588135, "learning_rate": 7.4975e-05, "loss": 0.0521, "step": 3000 }, { "grad_norm": 0.6669318079948425, "learning_rate": 7.5225e-05, "loss": 0.0569, "step": 3010 }, { "grad_norm": 0.5450798273086548, "learning_rate": 7.5475e-05, "loss": 0.052, "step": 3020 }, { "grad_norm": 0.7608873844146729, "learning_rate": 7.5725e-05, "loss": 0.0509, "step": 3030 }, { "grad_norm": 0.7558789849281311, "learning_rate": 7.5975e-05, "loss": 0.0476, "step": 3040 }, { "grad_norm": 0.578813374042511, "learning_rate": 7.6225e-05, "loss": 0.0502, "step": 3050 }, { "grad_norm": 0.5277841091156006, "learning_rate": 7.6475e-05, "loss": 0.0502, "step": 3060 }, { "grad_norm": 0.5524429082870483, "learning_rate": 7.672500000000001e-05, "loss": 0.0518, "step": 3070 }, { "grad_norm": 0.6779712438583374, "learning_rate": 7.697500000000001e-05, "loss": 0.0499, "step": 3080 }, { "grad_norm": 0.5634255409240723, "learning_rate": 7.722500000000001e-05, "loss": 0.0506, "step": 3090 }, { "grad_norm": 0.6444821357727051, "learning_rate": 7.747500000000001e-05, "loss": 0.0532, "step": 3100 }, { "grad_norm": 0.8489841818809509, "learning_rate": 7.7725e-05, "loss": 0.0552, "step": 3110 }, { "grad_norm": 0.5425320863723755, "learning_rate": 7.797500000000001e-05, "loss": 0.0462, "step": 3120 }, { "grad_norm": 0.6298826932907104, "learning_rate": 7.8225e-05, "loss": 0.0606, "step": 3130 }, { "grad_norm": 0.6492961049079895, "learning_rate": 7.8475e-05, "loss": 0.0522, "step": 3140 }, { "grad_norm": 0.7294120788574219, "learning_rate": 7.8725e-05, "loss": 0.0564, "step": 3150 }, { "grad_norm": 0.5288587808609009, "learning_rate": 7.8975e-05, "loss": 0.0465, "step": 3160 }, { "grad_norm": 0.5039998292922974, "learning_rate": 7.9225e-05, "loss": 0.0494, "step": 3170 }, { "grad_norm": 0.5864385366439819, "learning_rate": 7.9475e-05, "loss": 0.0504, "step": 3180 }, { "grad_norm": 0.598970353603363, "learning_rate": 7.9725e-05, "loss": 0.0521, "step": 3190 }, { "grad_norm": 0.558262825012207, "learning_rate": 7.9975e-05, "loss": 0.0485, "step": 3200 }, { "grad_norm": 0.6093278527259827, "learning_rate": 8.022500000000001e-05, "loss": 0.0471, "step": 3210 }, { "grad_norm": 0.8396059274673462, "learning_rate": 8.0475e-05, "loss": 0.0536, "step": 3220 }, { "grad_norm": 0.6502605676651001, "learning_rate": 8.072500000000001e-05, "loss": 0.057, "step": 3230 }, { "grad_norm": 0.6641515493392944, "learning_rate": 8.0975e-05, "loss": 0.0513, "step": 3240 }, { "grad_norm": 0.7277132272720337, "learning_rate": 8.122500000000001e-05, "loss": 0.0509, "step": 3250 }, { "grad_norm": 0.6885650157928467, "learning_rate": 8.1475e-05, "loss": 0.0531, "step": 3260 }, { "grad_norm": 0.592635452747345, "learning_rate": 8.172500000000001e-05, "loss": 0.052, "step": 3270 }, { "grad_norm": 0.6484315395355225, "learning_rate": 8.1975e-05, "loss": 0.0472, "step": 3280 }, { "grad_norm": 0.44711288809776306, "learning_rate": 8.2225e-05, "loss": 0.0501, "step": 3290 }, { "grad_norm": 0.5349039435386658, "learning_rate": 8.2475e-05, "loss": 0.0519, "step": 3300 }, { "grad_norm": 0.5816120505332947, "learning_rate": 8.2725e-05, "loss": 0.0501, "step": 3310 }, { "grad_norm": 0.6468127965927124, "learning_rate": 8.2975e-05, "loss": 0.0533, "step": 3320 }, { "grad_norm": 0.5564659833908081, "learning_rate": 8.3225e-05, "loss": 0.0479, "step": 3330 }, { "grad_norm": 0.5348617434501648, "learning_rate": 8.347500000000001e-05, "loss": 0.0507, "step": 3340 }, { "grad_norm": 0.5908156633377075, "learning_rate": 8.3725e-05, "loss": 0.0527, "step": 3350 }, { "grad_norm": 0.6047680377960205, "learning_rate": 8.397500000000001e-05, "loss": 0.0491, "step": 3360 }, { "grad_norm": 0.41470372676849365, "learning_rate": 8.422500000000001e-05, "loss": 0.0487, "step": 3370 }, { "grad_norm": 0.6324091553688049, "learning_rate": 8.447500000000001e-05, "loss": 0.0458, "step": 3380 }, { "grad_norm": 0.6604275107383728, "learning_rate": 8.4725e-05, "loss": 0.0523, "step": 3390 }, { "grad_norm": 0.5711779594421387, "learning_rate": 8.4975e-05, "loss": 0.0535, "step": 3400 }, { "grad_norm": 0.6499602794647217, "learning_rate": 8.5225e-05, "loss": 0.0555, "step": 3410 }, { "grad_norm": 0.61171555519104, "learning_rate": 8.5475e-05, "loss": 0.051, "step": 3420 }, { "grad_norm": 0.4923974871635437, "learning_rate": 8.5725e-05, "loss": 0.0462, "step": 3430 }, { "grad_norm": 0.6211041212081909, "learning_rate": 8.5975e-05, "loss": 0.0541, "step": 3440 }, { "grad_norm": 0.6566500663757324, "learning_rate": 8.6225e-05, "loss": 0.0431, "step": 3450 }, { "grad_norm": 0.6915661096572876, "learning_rate": 8.6475e-05, "loss": 0.049, "step": 3460 }, { "grad_norm": 0.48505115509033203, "learning_rate": 8.672500000000001e-05, "loss": 0.0551, "step": 3470 }, { "grad_norm": 0.5928274393081665, "learning_rate": 8.6975e-05, "loss": 0.0483, "step": 3480 }, { "grad_norm": 0.5292720198631287, "learning_rate": 8.7225e-05, "loss": 0.0482, "step": 3490 }, { "grad_norm": 0.5630860924720764, "learning_rate": 8.747500000000001e-05, "loss": 0.0467, "step": 3500 }, { "grad_norm": 0.6130035519599915, "learning_rate": 8.7725e-05, "loss": 0.0544, "step": 3510 }, { "grad_norm": 0.7033230662345886, "learning_rate": 8.797500000000001e-05, "loss": 0.0506, "step": 3520 }, { "grad_norm": 0.5440663695335388, "learning_rate": 8.8225e-05, "loss": 0.05, "step": 3530 }, { "grad_norm": 0.5826355814933777, "learning_rate": 8.847500000000001e-05, "loss": 0.0518, "step": 3540 }, { "grad_norm": 0.5115913152694702, "learning_rate": 8.8725e-05, "loss": 0.0487, "step": 3550 }, { "grad_norm": 0.49898388981819153, "learning_rate": 8.897500000000001e-05, "loss": 0.0496, "step": 3560 }, { "grad_norm": 0.541337788105011, "learning_rate": 8.9225e-05, "loss": 0.0517, "step": 3570 }, { "grad_norm": 0.5494643449783325, "learning_rate": 8.9475e-05, "loss": 0.0514, "step": 3580 }, { "grad_norm": 0.4959094524383545, "learning_rate": 8.9725e-05, "loss": 0.0517, "step": 3590 }, { "grad_norm": 0.5134307742118835, "learning_rate": 8.9975e-05, "loss": 0.051, "step": 3600 }, { "grad_norm": 0.6030277013778687, "learning_rate": 9.0225e-05, "loss": 0.048, "step": 3610 }, { "grad_norm": 0.6210982799530029, "learning_rate": 9.0475e-05, "loss": 0.0473, "step": 3620 }, { "grad_norm": 0.6427984833717346, "learning_rate": 9.072500000000001e-05, "loss": 0.0508, "step": 3630 }, { "grad_norm": 0.5490468144416809, "learning_rate": 9.0975e-05, "loss": 0.0466, "step": 3640 }, { "grad_norm": 0.6036748886108398, "learning_rate": 9.122500000000001e-05, "loss": 0.0425, "step": 3650 }, { "grad_norm": 0.5942227840423584, "learning_rate": 9.1475e-05, "loss": 0.0455, "step": 3660 }, { "grad_norm": 0.40086278319358826, "learning_rate": 9.172500000000001e-05, "loss": 0.045, "step": 3670 }, { "grad_norm": 0.4066513180732727, "learning_rate": 9.1975e-05, "loss": 0.0454, "step": 3680 }, { "grad_norm": 0.5664783120155334, "learning_rate": 9.2225e-05, "loss": 0.0521, "step": 3690 }, { "grad_norm": 0.483974426984787, "learning_rate": 9.2475e-05, "loss": 0.0421, "step": 3700 }, { "grad_norm": 0.4465171694755554, "learning_rate": 9.2725e-05, "loss": 0.0444, "step": 3710 }, { "grad_norm": 0.44451186060905457, "learning_rate": 9.2975e-05, "loss": 0.0475, "step": 3720 }, { "grad_norm": 0.4905838966369629, "learning_rate": 9.3225e-05, "loss": 0.0522, "step": 3730 }, { "grad_norm": 0.5948286652565002, "learning_rate": 9.3475e-05, "loss": 0.048, "step": 3740 }, { "grad_norm": 0.5052937269210815, "learning_rate": 9.3725e-05, "loss": 0.0487, "step": 3750 }, { "grad_norm": 0.46981412172317505, "learning_rate": 9.397500000000001e-05, "loss": 0.0517, "step": 3760 }, { "grad_norm": 0.472815603017807, "learning_rate": 9.422500000000001e-05, "loss": 0.042, "step": 3770 }, { "grad_norm": 0.3996998965740204, "learning_rate": 9.4475e-05, "loss": 0.0501, "step": 3780 }, { "grad_norm": 0.43792060017585754, "learning_rate": 9.472500000000001e-05, "loss": 0.0461, "step": 3790 }, { "grad_norm": 0.3636253774166107, "learning_rate": 9.4975e-05, "loss": 0.0499, "step": 3800 }, { "grad_norm": 0.5306479334831238, "learning_rate": 9.522500000000001e-05, "loss": 0.0539, "step": 3810 }, { "grad_norm": 0.559519350528717, "learning_rate": 9.5475e-05, "loss": 0.0561, "step": 3820 }, { "grad_norm": 0.5238432884216309, "learning_rate": 9.572500000000001e-05, "loss": 0.0465, "step": 3830 }, { "grad_norm": 0.3780970573425293, "learning_rate": 9.5975e-05, "loss": 0.0449, "step": 3840 }, { "grad_norm": 0.5398630499839783, "learning_rate": 9.622500000000001e-05, "loss": 0.0465, "step": 3850 }, { "grad_norm": 0.41128361225128174, "learning_rate": 9.6475e-05, "loss": 0.0498, "step": 3860 }, { "grad_norm": 0.4804386496543884, "learning_rate": 9.6725e-05, "loss": 0.043, "step": 3870 }, { "grad_norm": 0.44075873494148254, "learning_rate": 9.6975e-05, "loss": 0.0434, "step": 3880 }, { "grad_norm": 0.5246647000312805, "learning_rate": 9.7225e-05, "loss": 0.0482, "step": 3890 }, { "grad_norm": 0.5387880206108093, "learning_rate": 9.747500000000001e-05, "loss": 0.0524, "step": 3900 }, { "grad_norm": 0.48840731382369995, "learning_rate": 9.7725e-05, "loss": 0.046, "step": 3910 }, { "grad_norm": 0.5133220553398132, "learning_rate": 9.797500000000001e-05, "loss": 0.0534, "step": 3920 }, { "grad_norm": 0.49491602182388306, "learning_rate": 9.8225e-05, "loss": 0.0447, "step": 3930 }, { "grad_norm": 0.42988649010658264, "learning_rate": 9.847500000000001e-05, "loss": 0.0473, "step": 3940 }, { "grad_norm": 0.47287631034851074, "learning_rate": 9.8725e-05, "loss": 0.0464, "step": 3950 }, { "grad_norm": 0.5049149394035339, "learning_rate": 9.897500000000001e-05, "loss": 0.0451, "step": 3960 }, { "grad_norm": 0.46181783080101013, "learning_rate": 9.9225e-05, "loss": 0.0428, "step": 3970 }, { "grad_norm": 0.4115452766418457, "learning_rate": 9.9475e-05, "loss": 0.0437, "step": 3980 }, { "grad_norm": 0.4008096158504486, "learning_rate": 9.9725e-05, "loss": 0.0451, "step": 3990 }, { "grad_norm": 0.5836281776428223, "learning_rate": 9.9975e-05, "loss": 0.0483, "step": 4000 }, { "grad_norm": 0.4279605448246002, "learning_rate": 9.999999653982884e-05, "loss": 0.043, "step": 4010 }, { "grad_norm": 0.4389435350894928, "learning_rate": 9.999998457874392e-05, "loss": 0.0427, "step": 4020 }, { "grad_norm": 0.5150651335716248, "learning_rate": 9.999996407402913e-05, "loss": 0.0498, "step": 4030 }, { "grad_norm": 0.48244622349739075, "learning_rate": 9.999993502568801e-05, "loss": 0.0447, "step": 4040 }, { "grad_norm": 0.41431787610054016, "learning_rate": 9.999989743372548e-05, "loss": 0.0461, "step": 4050 }, { "grad_norm": 0.4197765588760376, "learning_rate": 9.999985129814798e-05, "loss": 0.0451, "step": 4060 }, { "grad_norm": 0.3942737281322479, "learning_rate": 9.99997966189634e-05, "loss": 0.0399, "step": 4070 }, { "grad_norm": 0.43094155192375183, "learning_rate": 9.999973339618107e-05, "loss": 0.0453, "step": 4080 }, { "grad_norm": 0.4451828896999359, "learning_rate": 9.999966162981179e-05, "loss": 0.0445, "step": 4090 }, { "grad_norm": 0.3957744538784027, "learning_rate": 9.999958131986784e-05, "loss": 0.0406, "step": 4100 }, { "grad_norm": 0.44088831543922424, "learning_rate": 9.999949246636293e-05, "loss": 0.0503, "step": 4110 }, { "grad_norm": 0.4508195221424103, "learning_rate": 9.999939506931224e-05, "loss": 0.048, "step": 4120 }, { "grad_norm": 0.5352017283439636, "learning_rate": 9.999928912873243e-05, "loss": 0.0511, "step": 4130 }, { "grad_norm": 0.5133150219917297, "learning_rate": 9.999917464464159e-05, "loss": 0.0441, "step": 4140 }, { "grad_norm": 0.47296959161758423, "learning_rate": 9.999905161705929e-05, "loss": 0.0467, "step": 4150 }, { "grad_norm": 0.4317520558834076, "learning_rate": 9.999892004600653e-05, "loss": 0.0439, "step": 4160 }, { "grad_norm": 0.41307884454727173, "learning_rate": 9.999877993150581e-05, "loss": 0.0479, "step": 4170 }, { "grad_norm": 0.47972530126571655, "learning_rate": 9.999863127358108e-05, "loss": 0.0462, "step": 4180 }, { "grad_norm": 0.6004597544670105, "learning_rate": 9.999847407225773e-05, "loss": 0.0486, "step": 4190 }, { "grad_norm": 0.5376521944999695, "learning_rate": 9.999830832756262e-05, "loss": 0.0413, "step": 4200 }, { "grad_norm": 0.3597925007343292, "learning_rate": 9.999813403952407e-05, "loss": 0.0414, "step": 4210 }, { "grad_norm": 0.46408572793006897, "learning_rate": 9.999795120817187e-05, "loss": 0.0414, "step": 4220 }, { "grad_norm": 0.369221031665802, "learning_rate": 9.999775983353725e-05, "loss": 0.0411, "step": 4230 }, { "grad_norm": 0.5834969878196716, "learning_rate": 9.999755991565292e-05, "loss": 0.0418, "step": 4240 }, { "grad_norm": 0.3855322003364563, "learning_rate": 9.999735145455303e-05, "loss": 0.0438, "step": 4250 }, { "grad_norm": 0.44607189297676086, "learning_rate": 9.99971344502732e-05, "loss": 0.045, "step": 4260 }, { "grad_norm": 0.4182288944721222, "learning_rate": 9.999690890285053e-05, "loss": 0.0453, "step": 4270 }, { "grad_norm": 0.5636694431304932, "learning_rate": 9.999667481232356e-05, "loss": 0.0461, "step": 4280 }, { "grad_norm": 0.3594149351119995, "learning_rate": 9.999643217873225e-05, "loss": 0.0414, "step": 4290 }, { "grad_norm": 0.45584267377853394, "learning_rate": 9.999618100211809e-05, "loss": 0.0396, "step": 4300 }, { "grad_norm": 0.3696536719799042, "learning_rate": 9.999592128252402e-05, "loss": 0.0402, "step": 4310 }, { "grad_norm": 0.49256792664527893, "learning_rate": 9.999565301999437e-05, "loss": 0.0441, "step": 4320 }, { "grad_norm": 0.503288745880127, "learning_rate": 9.999537621457502e-05, "loss": 0.0404, "step": 4330 }, { "grad_norm": 0.5206612944602966, "learning_rate": 9.999509086631323e-05, "loss": 0.0465, "step": 4340 }, { "grad_norm": 0.4433129131793976, "learning_rate": 9.99947969752578e-05, "loss": 0.0511, "step": 4350 }, { "grad_norm": 0.4087696075439453, "learning_rate": 9.999449454145891e-05, "loss": 0.0423, "step": 4360 }, { "grad_norm": 0.41787442564964294, "learning_rate": 9.999418356496827e-05, "loss": 0.0447, "step": 4370 }, { "grad_norm": 0.5632086992263794, "learning_rate": 9.999386404583899e-05, "loss": 0.0462, "step": 4380 }, { "grad_norm": 0.3860396444797516, "learning_rate": 9.999353598412568e-05, "loss": 0.0495, "step": 4390 }, { "grad_norm": 0.3858436942100525, "learning_rate": 9.999319937988442e-05, "loss": 0.0417, "step": 4400 }, { "grad_norm": 0.48116278648376465, "learning_rate": 9.999285423317268e-05, "loss": 0.0428, "step": 4410 }, { "grad_norm": 0.4289935231208801, "learning_rate": 9.999250054404947e-05, "loss": 0.0433, "step": 4420 }, { "grad_norm": 0.4070263206958771, "learning_rate": 9.99921383125752e-05, "loss": 0.0406, "step": 4430 }, { "grad_norm": 0.4471568465232849, "learning_rate": 9.99917675388118e-05, "loss": 0.0412, "step": 4440 }, { "grad_norm": 0.3605797588825226, "learning_rate": 9.99913882228226e-05, "loss": 0.0438, "step": 4450 }, { "grad_norm": 0.4948781132698059, "learning_rate": 9.999100036467242e-05, "loss": 0.0406, "step": 4460 }, { "grad_norm": 0.3856710195541382, "learning_rate": 9.999060396442753e-05, "loss": 0.0391, "step": 4470 }, { "grad_norm": 0.3492943048477173, "learning_rate": 9.999019902215566e-05, "loss": 0.0426, "step": 4480 }, { "grad_norm": 0.5384465456008911, "learning_rate": 9.998978553792602e-05, "loss": 0.0484, "step": 4490 }, { "grad_norm": 0.33625274896621704, "learning_rate": 9.998936351180926e-05, "loss": 0.042, "step": 4500 }, { "grad_norm": 0.3698599636554718, "learning_rate": 9.998893294387747e-05, "loss": 0.0391, "step": 4510 }, { "grad_norm": 0.5760732889175415, "learning_rate": 9.998849383420426e-05, "loss": 0.0391, "step": 4520 }, { "grad_norm": 0.4065379500389099, "learning_rate": 9.998804618286465e-05, "loss": 0.0428, "step": 4530 }, { "grad_norm": 0.5309622883796692, "learning_rate": 9.99875899899351e-05, "loss": 0.0428, "step": 4540 }, { "grad_norm": 0.3387792408466339, "learning_rate": 9.99871252554936e-05, "loss": 0.0363, "step": 4550 }, { "grad_norm": 0.47070568799972534, "learning_rate": 9.998665197961955e-05, "loss": 0.0388, "step": 4560 }, { "grad_norm": 0.3736090064048767, "learning_rate": 9.998617016239379e-05, "loss": 0.0395, "step": 4570 }, { "grad_norm": 0.540499746799469, "learning_rate": 9.998567980389869e-05, "loss": 0.0449, "step": 4580 }, { "grad_norm": 0.42824262380599976, "learning_rate": 9.998518090421802e-05, "loss": 0.0461, "step": 4590 }, { "grad_norm": 0.4077673554420471, "learning_rate": 9.998467346343703e-05, "loss": 0.0419, "step": 4600 }, { "grad_norm": 0.4336542785167694, "learning_rate": 9.998415748164243e-05, "loss": 0.0439, "step": 4610 }, { "grad_norm": 0.33464235067367554, "learning_rate": 9.998363295892238e-05, "loss": 0.0404, "step": 4620 }, { "grad_norm": 0.36474913358688354, "learning_rate": 9.998309989536652e-05, "loss": 0.0393, "step": 4630 }, { "grad_norm": 0.39171960949897766, "learning_rate": 9.998255829106593e-05, "loss": 0.0434, "step": 4640 }, { "grad_norm": 0.31233838200569153, "learning_rate": 9.998200814611316e-05, "loss": 0.0395, "step": 4650 }, { "grad_norm": 0.3567310869693756, "learning_rate": 9.998144946060219e-05, "loss": 0.041, "step": 4660 }, { "grad_norm": 0.31406185030937195, "learning_rate": 9.998088223462852e-05, "loss": 0.0394, "step": 4670 }, { "grad_norm": 0.46675726771354675, "learning_rate": 9.998030646828905e-05, "loss": 0.0369, "step": 4680 }, { "grad_norm": 0.4438435733318329, "learning_rate": 9.997972216168217e-05, "loss": 0.0403, "step": 4690 }, { "grad_norm": 0.4167032241821289, "learning_rate": 9.997912931490771e-05, "loss": 0.0486, "step": 4700 }, { "grad_norm": 0.38077372312545776, "learning_rate": 9.9978527928067e-05, "loss": 0.0416, "step": 4710 }, { "grad_norm": 0.335506409406662, "learning_rate": 9.997791800126277e-05, "loss": 0.0365, "step": 4720 }, { "grad_norm": 0.38144057989120483, "learning_rate": 9.997729953459927e-05, "loss": 0.0387, "step": 4730 }, { "grad_norm": 0.39465269446372986, "learning_rate": 9.997667252818214e-05, "loss": 0.0402, "step": 4740 }, { "grad_norm": 0.3490985035896301, "learning_rate": 9.997603698211855e-05, "loss": 0.038, "step": 4750 }, { "grad_norm": 0.3253363370895386, "learning_rate": 9.99753928965171e-05, "loss": 0.041, "step": 4760 }, { "grad_norm": 0.45837974548339844, "learning_rate": 9.997474027148781e-05, "loss": 0.0461, "step": 4770 }, { "grad_norm": 0.4361877143383026, "learning_rate": 9.997407910714223e-05, "loss": 0.0407, "step": 4780 }, { "grad_norm": 0.3769112527370453, "learning_rate": 9.997340940359332e-05, "loss": 0.0442, "step": 4790 }, { "grad_norm": 0.38700175285339355, "learning_rate": 9.997273116095552e-05, "loss": 0.0404, "step": 4800 }, { "grad_norm": 0.49159565567970276, "learning_rate": 9.997204437934473e-05, "loss": 0.0374, "step": 4810 }, { "grad_norm": 0.3580368757247925, "learning_rate": 9.997134905887829e-05, "loss": 0.0396, "step": 4820 }, { "grad_norm": 0.3751048445701599, "learning_rate": 9.997064519967501e-05, "loss": 0.0364, "step": 4830 }, { "grad_norm": 0.36665958166122437, "learning_rate": 9.996993280185517e-05, "loss": 0.04, "step": 4840 }, { "grad_norm": 0.41432008147239685, "learning_rate": 9.99692118655405e-05, "loss": 0.0385, "step": 4850 }, { "grad_norm": 0.3556528389453888, "learning_rate": 9.996848239085417e-05, "loss": 0.0366, "step": 4860 }, { "grad_norm": 0.3248606324195862, "learning_rate": 9.996774437792085e-05, "loss": 0.0416, "step": 4870 }, { "grad_norm": 0.3990243077278137, "learning_rate": 9.996699782686664e-05, "loss": 0.0395, "step": 4880 }, { "grad_norm": 0.3853633403778076, "learning_rate": 9.996624273781909e-05, "loss": 0.0408, "step": 4890 }, { "grad_norm": 0.39898088574409485, "learning_rate": 9.996547911090725e-05, "loss": 0.0387, "step": 4900 }, { "grad_norm": 0.40821942687034607, "learning_rate": 9.996470694626157e-05, "loss": 0.0388, "step": 4910 }, { "grad_norm": 0.43259531259536743, "learning_rate": 9.996392624401403e-05, "loss": 0.0431, "step": 4920 }, { "grad_norm": 0.3773435950279236, "learning_rate": 9.996313700429801e-05, "loss": 0.0415, "step": 4930 }, { "grad_norm": 0.35636499524116516, "learning_rate": 9.996233922724836e-05, "loss": 0.0359, "step": 4940 }, { "grad_norm": 0.38820600509643555, "learning_rate": 9.996153291300141e-05, "loss": 0.041, "step": 4950 }, { "grad_norm": 0.3808906376361847, "learning_rate": 9.996071806169494e-05, "loss": 0.0431, "step": 4960 }, { "grad_norm": 0.33854877948760986, "learning_rate": 9.995989467346817e-05, "loss": 0.034, "step": 4970 }, { "grad_norm": 0.3747018873691559, "learning_rate": 9.995906274846183e-05, "loss": 0.044, "step": 4980 }, { "grad_norm": 0.3585336208343506, "learning_rate": 9.995822228681803e-05, "loss": 0.0408, "step": 4990 }, { "grad_norm": 0.3699122369289398, "learning_rate": 9.99573732886804e-05, "loss": 0.0386, "step": 5000 }, { "grad_norm": 0.3823486268520355, "learning_rate": 9.995651575419402e-05, "loss": 0.0394, "step": 5010 }, { "grad_norm": 0.4399341940879822, "learning_rate": 9.995564968350541e-05, "loss": 0.0394, "step": 5020 }, { "grad_norm": 0.38932934403419495, "learning_rate": 9.995477507676256e-05, "loss": 0.0351, "step": 5030 }, { "grad_norm": 0.464145690202713, "learning_rate": 9.995389193411493e-05, "loss": 0.0352, "step": 5040 }, { "grad_norm": 0.45307108759880066, "learning_rate": 9.995300025571339e-05, "loss": 0.0398, "step": 5050 }, { "grad_norm": 0.4135781228542328, "learning_rate": 9.995210004171034e-05, "loss": 0.0434, "step": 5060 }, { "grad_norm": 0.4142097234725952, "learning_rate": 9.995119129225956e-05, "loss": 0.0402, "step": 5070 }, { "grad_norm": 0.36967402696609497, "learning_rate": 9.995027400751637e-05, "loss": 0.0392, "step": 5080 }, { "grad_norm": 0.4670371413230896, "learning_rate": 9.994934818763751e-05, "loss": 0.0411, "step": 5090 }, { "grad_norm": 0.35396382212638855, "learning_rate": 9.994841383278115e-05, "loss": 0.0399, "step": 5100 }, { "grad_norm": 0.3430791199207306, "learning_rate": 9.994747094310695e-05, "loss": 0.0397, "step": 5110 }, { "grad_norm": 0.3813588619232178, "learning_rate": 9.994651951877604e-05, "loss": 0.0346, "step": 5120 }, { "grad_norm": 0.3367784321308136, "learning_rate": 9.994555955995099e-05, "loss": 0.0368, "step": 5130 }, { "grad_norm": 0.32991546392440796, "learning_rate": 9.994459106679581e-05, "loss": 0.0407, "step": 5140 }, { "grad_norm": 0.3053739070892334, "learning_rate": 9.994361403947603e-05, "loss": 0.0381, "step": 5150 }, { "grad_norm": 0.38162171840667725, "learning_rate": 9.994262847815854e-05, "loss": 0.0402, "step": 5160 }, { "grad_norm": 0.2757958769798279, "learning_rate": 9.99416343830118e-05, "loss": 0.0357, "step": 5170 }, { "grad_norm": 0.34541019797325134, "learning_rate": 9.994063175420565e-05, "loss": 0.0352, "step": 5180 }, { "grad_norm": 0.35190093517303467, "learning_rate": 9.99396205919114e-05, "loss": 0.0381, "step": 5190 }, { "grad_norm": 0.32873621582984924, "learning_rate": 9.993860089630185e-05, "loss": 0.0365, "step": 5200 }, { "grad_norm": 0.5136183500289917, "learning_rate": 9.993757266755123e-05, "loss": 0.0446, "step": 5210 }, { "grad_norm": 0.4805842638015747, "learning_rate": 9.993653590583522e-05, "loss": 0.0463, "step": 5220 }, { "grad_norm": 0.32354170083999634, "learning_rate": 9.993549061133102e-05, "loss": 0.0381, "step": 5230 }, { "grad_norm": 0.4423300623893738, "learning_rate": 9.993443678421719e-05, "loss": 0.0343, "step": 5240 }, { "grad_norm": 0.2841018736362457, "learning_rate": 9.993337442467384e-05, "loss": 0.0388, "step": 5250 }, { "grad_norm": 0.3258613049983978, "learning_rate": 9.993230353288248e-05, "loss": 0.0358, "step": 5260 }, { "grad_norm": 0.4958130419254303, "learning_rate": 9.993122410902608e-05, "loss": 0.0377, "step": 5270 }, { "grad_norm": 0.38689181208610535, "learning_rate": 9.993013615328912e-05, "loss": 0.0378, "step": 5280 }, { "grad_norm": 0.319393128156662, "learning_rate": 9.992903966585747e-05, "loss": 0.0383, "step": 5290 }, { "grad_norm": 0.3548632562160492, "learning_rate": 9.992793464691852e-05, "loss": 0.0446, "step": 5300 }, { "grad_norm": 0.4161016047000885, "learning_rate": 9.992682109666105e-05, "loss": 0.0446, "step": 5310 }, { "grad_norm": 0.47750401496887207, "learning_rate": 9.992569901527538e-05, "loss": 0.0436, "step": 5320 }, { "grad_norm": 0.3699932396411896, "learning_rate": 9.99245684029532e-05, "loss": 0.0371, "step": 5330 }, { "grad_norm": 0.3565710186958313, "learning_rate": 9.992342925988774e-05, "loss": 0.0404, "step": 5340 }, { "grad_norm": 0.3969672620296478, "learning_rate": 9.992228158627361e-05, "loss": 0.0397, "step": 5350 }, { "grad_norm": 0.3416878283023834, "learning_rate": 9.992112538230693e-05, "loss": 0.0353, "step": 5360 }, { "grad_norm": 0.3799988031387329, "learning_rate": 9.991996064818527e-05, "loss": 0.035, "step": 5370 }, { "grad_norm": 0.34365206956863403, "learning_rate": 9.991878738410768e-05, "loss": 0.0367, "step": 5380 }, { "grad_norm": 0.46991395950317383, "learning_rate": 9.991760559027457e-05, "loss": 0.0371, "step": 5390 }, { "grad_norm": 0.43776142597198486, "learning_rate": 9.991641526688793e-05, "loss": 0.0332, "step": 5400 }, { "grad_norm": 0.3916773200035095, "learning_rate": 9.991521641415113e-05, "loss": 0.0364, "step": 5410 }, { "grad_norm": 0.27511146664619446, "learning_rate": 9.991400903226904e-05, "loss": 0.0357, "step": 5420 }, { "grad_norm": 0.2834094166755676, "learning_rate": 9.991279312144794e-05, "loss": 0.0419, "step": 5430 }, { "grad_norm": 0.3195440173149109, "learning_rate": 9.991156868189564e-05, "loss": 0.037, "step": 5440 }, { "grad_norm": 0.43989297747612, "learning_rate": 9.991033571382131e-05, "loss": 0.0391, "step": 5450 }, { "grad_norm": 0.3676360249519348, "learning_rate": 9.990909421743569e-05, "loss": 0.037, "step": 5460 }, { "grad_norm": 0.4460879862308502, "learning_rate": 9.990784419295085e-05, "loss": 0.0369, "step": 5470 }, { "grad_norm": 0.49686020612716675, "learning_rate": 9.990658564058044e-05, "loss": 0.041, "step": 5480 }, { "grad_norm": 0.36375656723976135, "learning_rate": 9.990531856053948e-05, "loss": 0.0368, "step": 5490 }, { "grad_norm": 0.34415075182914734, "learning_rate": 9.99040429530445e-05, "loss": 0.0352, "step": 5500 }, { "grad_norm": 0.4916974604129791, "learning_rate": 9.990275881831346e-05, "loss": 0.0399, "step": 5510 }, { "grad_norm": 0.4549710750579834, "learning_rate": 9.990146615656577e-05, "loss": 0.0405, "step": 5520 }, { "grad_norm": 0.3515833914279938, "learning_rate": 9.990016496802233e-05, "loss": 0.0379, "step": 5530 }, { "grad_norm": 0.3527772128582001, "learning_rate": 9.989885525290548e-05, "loss": 0.0368, "step": 5540 }, { "grad_norm": 0.2951643764972687, "learning_rate": 9.989753701143897e-05, "loss": 0.0378, "step": 5550 }, { "grad_norm": 0.3731999397277832, "learning_rate": 9.989621024384812e-05, "loss": 0.0363, "step": 5560 }, { "grad_norm": 0.4068029820919037, "learning_rate": 9.989487495035959e-05, "loss": 0.0367, "step": 5570 }, { "grad_norm": 0.38421064615249634, "learning_rate": 9.989353113120156e-05, "loss": 0.0319, "step": 5580 }, { "grad_norm": 0.32749971747398376, "learning_rate": 9.989217878660366e-05, "loss": 0.0372, "step": 5590 }, { "grad_norm": 0.3439282774925232, "learning_rate": 9.989081791679695e-05, "loss": 0.0338, "step": 5600 }, { "grad_norm": 0.40047770738601685, "learning_rate": 9.988944852201397e-05, "loss": 0.0333, "step": 5610 }, { "grad_norm": 0.395082950592041, "learning_rate": 9.988807060248873e-05, "loss": 0.0353, "step": 5620 }, { "grad_norm": 0.3719947338104248, "learning_rate": 9.988668415845665e-05, "loss": 0.0348, "step": 5630 }, { "grad_norm": 0.32991740107536316, "learning_rate": 9.988528919015466e-05, "loss": 0.0333, "step": 5640 }, { "grad_norm": 0.2997605800628662, "learning_rate": 9.988388569782112e-05, "loss": 0.0331, "step": 5650 }, { "grad_norm": 0.4000813961029053, "learning_rate": 9.988247368169583e-05, "loss": 0.0421, "step": 5660 }, { "grad_norm": 0.3173542618751526, "learning_rate": 9.988105314202007e-05, "loss": 0.0356, "step": 5670 }, { "grad_norm": 0.39883512258529663, "learning_rate": 9.987962407903659e-05, "loss": 0.0364, "step": 5680 }, { "grad_norm": 0.2827059328556061, "learning_rate": 9.987818649298957e-05, "loss": 0.0321, "step": 5690 }, { "grad_norm": 0.39205679297447205, "learning_rate": 9.987674038412465e-05, "loss": 0.0368, "step": 5700 }, { "grad_norm": 0.4997949004173279, "learning_rate": 9.987528575268891e-05, "loss": 0.0343, "step": 5710 }, { "grad_norm": 0.3786291182041168, "learning_rate": 9.987382259893095e-05, "loss": 0.0356, "step": 5720 }, { "grad_norm": 0.3565458059310913, "learning_rate": 9.987235092310074e-05, "loss": 0.0351, "step": 5730 }, { "grad_norm": 0.3358219861984253, "learning_rate": 9.987087072544978e-05, "loss": 0.0376, "step": 5740 }, { "grad_norm": 0.4125051498413086, "learning_rate": 9.9869382006231e-05, "loss": 0.0365, "step": 5750 }, { "grad_norm": 0.39243578910827637, "learning_rate": 9.986788476569875e-05, "loss": 0.0334, "step": 5760 }, { "grad_norm": 0.2702019512653351, "learning_rate": 9.986637900410887e-05, "loss": 0.0375, "step": 5770 }, { "grad_norm": 0.4191756248474121, "learning_rate": 9.986486472171869e-05, "loss": 0.0351, "step": 5780 }, { "grad_norm": 0.346282035112381, "learning_rate": 9.986334191878692e-05, "loss": 0.0383, "step": 5790 }, { "grad_norm": 0.44664669036865234, "learning_rate": 9.986181059557378e-05, "loss": 0.0362, "step": 5800 }, { "grad_norm": 0.3473651707172394, "learning_rate": 9.986027075234094e-05, "loss": 0.0355, "step": 5810 }, { "grad_norm": 0.3083897531032562, "learning_rate": 9.985872238935152e-05, "loss": 0.0378, "step": 5820 }, { "grad_norm": 0.40029266476631165, "learning_rate": 9.985716550687008e-05, "loss": 0.0389, "step": 5830 }, { "grad_norm": 0.40106573700904846, "learning_rate": 9.985560010516264e-05, "loss": 0.0378, "step": 5840 }, { "grad_norm": 0.43405017256736755, "learning_rate": 9.985402618449668e-05, "loss": 0.0358, "step": 5850 }, { "grad_norm": 0.27713045477867126, "learning_rate": 9.985244374514118e-05, "loss": 0.0407, "step": 5860 }, { "grad_norm": 0.31817156076431274, "learning_rate": 9.985085278736651e-05, "loss": 0.0385, "step": 5870 }, { "grad_norm": 0.36309725046157837, "learning_rate": 9.984925331144452e-05, "loss": 0.0355, "step": 5880 }, { "grad_norm": 0.37270066142082214, "learning_rate": 9.984764531764851e-05, "loss": 0.0341, "step": 5890 }, { "grad_norm": 0.47628462314605713, "learning_rate": 9.984602880625326e-05, "loss": 0.036, "step": 5900 }, { "grad_norm": 0.3054281175136566, "learning_rate": 9.9844403777535e-05, "loss": 0.038, "step": 5910 }, { "grad_norm": 0.4386266767978668, "learning_rate": 9.984277023177135e-05, "loss": 0.0352, "step": 5920 }, { "grad_norm": 0.3834841847419739, "learning_rate": 9.984112816924148e-05, "loss": 0.0322, "step": 5930 }, { "grad_norm": 0.3154723644256592, "learning_rate": 9.983947759022596e-05, "loss": 0.0359, "step": 5940 }, { "grad_norm": 0.412314772605896, "learning_rate": 9.983781849500682e-05, "loss": 0.0327, "step": 5950 }, { "grad_norm": 0.4822341799736023, "learning_rate": 9.98361508838676e-05, "loss": 0.0339, "step": 5960 }, { "grad_norm": 0.37582430243492126, "learning_rate": 9.98344747570932e-05, "loss": 0.0366, "step": 5970 }, { "grad_norm": 0.39552026987075806, "learning_rate": 9.983279011497004e-05, "loss": 0.0352, "step": 5980 }, { "grad_norm": 0.32126152515411377, "learning_rate": 9.983109695778596e-05, "loss": 0.0371, "step": 5990 }, { "grad_norm": 0.30864349007606506, "learning_rate": 9.982939528583032e-05, "loss": 0.0379, "step": 6000 }, { "grad_norm": 0.2857007682323456, "learning_rate": 9.982768509939385e-05, "loss": 0.031, "step": 6010 }, { "grad_norm": 0.3535921275615692, "learning_rate": 9.982596639876879e-05, "loss": 0.0323, "step": 6020 }, { "grad_norm": 0.2967619001865387, "learning_rate": 9.982423918424881e-05, "loss": 0.031, "step": 6030 }, { "grad_norm": 0.307161420583725, "learning_rate": 9.982250345612908e-05, "loss": 0.0359, "step": 6040 }, { "grad_norm": 0.3253406584262848, "learning_rate": 9.982075921470611e-05, "loss": 0.0348, "step": 6050 }, { "grad_norm": 0.302277535200119, "learning_rate": 9.981900646027802e-05, "loss": 0.0326, "step": 6060 }, { "grad_norm": 0.29730919003486633, "learning_rate": 9.981724519314425e-05, "loss": 0.0368, "step": 6070 }, { "grad_norm": 0.29800665378570557, "learning_rate": 9.981547541360581e-05, "loss": 0.0382, "step": 6080 }, { "grad_norm": 0.30230239033699036, "learning_rate": 9.981369712196508e-05, "loss": 0.0314, "step": 6090 }, { "grad_norm": 0.3190537095069885, "learning_rate": 9.981191031852592e-05, "loss": 0.0388, "step": 6100 }, { "grad_norm": 0.3350618779659271, "learning_rate": 9.981011500359362e-05, "loss": 0.0339, "step": 6110 }, { "grad_norm": 0.3524513840675354, "learning_rate": 9.9808311177475e-05, "loss": 0.0393, "step": 6120 }, { "grad_norm": 0.34016522765159607, "learning_rate": 9.980649884047826e-05, "loss": 0.0334, "step": 6130 }, { "grad_norm": 0.3818851113319397, "learning_rate": 9.980467799291307e-05, "loss": 0.0339, "step": 6140 }, { "grad_norm": 0.4049653112888336, "learning_rate": 9.980284863509058e-05, "loss": 0.0359, "step": 6150 }, { "grad_norm": 0.3830108046531677, "learning_rate": 9.980101076732334e-05, "loss": 0.0325, "step": 6160 }, { "grad_norm": 0.3871265947818756, "learning_rate": 9.979916438992544e-05, "loss": 0.035, "step": 6170 }, { "grad_norm": 0.30365070700645447, "learning_rate": 9.979730950321237e-05, "loss": 0.0326, "step": 6180 }, { "grad_norm": 0.27912193536758423, "learning_rate": 9.979544610750104e-05, "loss": 0.0313, "step": 6190 }, { "grad_norm": 0.30921095609664917, "learning_rate": 9.97935742031099e-05, "loss": 0.0341, "step": 6200 }, { "grad_norm": 0.3795010447502136, "learning_rate": 9.979169379035878e-05, "loss": 0.0336, "step": 6210 }, { "grad_norm": 0.4343709349632263, "learning_rate": 9.978980486956899e-05, "loss": 0.0364, "step": 6220 }, { "grad_norm": 0.3518691658973694, "learning_rate": 9.978790744106332e-05, "loss": 0.0336, "step": 6230 }, { "grad_norm": 0.4451911449432373, "learning_rate": 9.978600150516594e-05, "loss": 0.0347, "step": 6240 }, { "grad_norm": 0.230643630027771, "learning_rate": 9.978408706220259e-05, "loss": 0.0383, "step": 6250 }, { "grad_norm": 0.30884402990341187, "learning_rate": 9.978216411250032e-05, "loss": 0.0392, "step": 6260 }, { "grad_norm": 0.3112923800945282, "learning_rate": 9.978023265638778e-05, "loss": 0.0312, "step": 6270 }, { "grad_norm": 0.24474281072616577, "learning_rate": 9.977829269419495e-05, "loss": 0.0344, "step": 6280 }, { "grad_norm": 0.28444135189056396, "learning_rate": 9.977634422625335e-05, "loss": 0.0325, "step": 6290 }, { "grad_norm": 0.32341915369033813, "learning_rate": 9.97743872528959e-05, "loss": 0.0414, "step": 6300 }, { "grad_norm": 0.32371658086776733, "learning_rate": 9.9772421774457e-05, "loss": 0.0361, "step": 6310 }, { "grad_norm": 0.404091477394104, "learning_rate": 9.977044779127252e-05, "loss": 0.0345, "step": 6320 }, { "grad_norm": 0.39526596665382385, "learning_rate": 9.976846530367971e-05, "loss": 0.0366, "step": 6330 }, { "grad_norm": 0.32155829668045044, "learning_rate": 9.976647431201735e-05, "loss": 0.0366, "step": 6340 }, { "grad_norm": 0.34719952940940857, "learning_rate": 9.976447481662568e-05, "loss": 0.0366, "step": 6350 }, { "grad_norm": 0.31656256318092346, "learning_rate": 9.976246681784629e-05, "loss": 0.0384, "step": 6360 }, { "grad_norm": 0.28702425956726074, "learning_rate": 9.976045031602234e-05, "loss": 0.0343, "step": 6370 }, { "grad_norm": 0.419617623090744, "learning_rate": 9.975842531149837e-05, "loss": 0.0377, "step": 6380 }, { "grad_norm": 0.3313044011592865, "learning_rate": 9.975639180462043e-05, "loss": 0.0358, "step": 6390 }, { "grad_norm": 0.3963448405265808, "learning_rate": 9.975434979573596e-05, "loss": 0.0336, "step": 6400 }, { "grad_norm": 0.31834280490875244, "learning_rate": 9.97522992851939e-05, "loss": 0.0401, "step": 6410 }, { "grad_norm": 0.2908938527107239, "learning_rate": 9.975024027334461e-05, "loss": 0.0381, "step": 6420 }, { "grad_norm": 0.2564719319343567, "learning_rate": 9.974817276053993e-05, "loss": 0.034, "step": 6430 }, { "grad_norm": 0.36664479970932007, "learning_rate": 9.974609674713315e-05, "loss": 0.0342, "step": 6440 }, { "grad_norm": 0.39386236667633057, "learning_rate": 9.9744012233479e-05, "loss": 0.0349, "step": 6450 }, { "grad_norm": 0.3791062831878662, "learning_rate": 9.974191921993366e-05, "loss": 0.0323, "step": 6460 }, { "grad_norm": 0.33615657687187195, "learning_rate": 9.973981770685474e-05, "loss": 0.034, "step": 6470 }, { "grad_norm": 0.3916633427143097, "learning_rate": 9.97377076946014e-05, "loss": 0.0362, "step": 6480 }, { "grad_norm": 0.3401600420475006, "learning_rate": 9.973558918353412e-05, "loss": 0.0358, "step": 6490 }, { "grad_norm": 0.36579206585884094, "learning_rate": 9.973346217401494e-05, "loss": 0.0317, "step": 6500 }, { "grad_norm": 0.37133264541625977, "learning_rate": 9.973132666640726e-05, "loss": 0.0328, "step": 6510 }, { "grad_norm": 0.3498920500278473, "learning_rate": 9.972918266107602e-05, "loss": 0.0315, "step": 6520 }, { "grad_norm": 0.31046780943870544, "learning_rate": 9.972703015838756e-05, "loss": 0.0333, "step": 6530 }, { "grad_norm": 0.30670058727264404, "learning_rate": 9.97248691587097e-05, "loss": 0.0322, "step": 6540 }, { "grad_norm": 0.27853623032569885, "learning_rate": 9.972269966241166e-05, "loss": 0.0329, "step": 6550 }, { "grad_norm": 0.385928213596344, "learning_rate": 9.972052166986417e-05, "loss": 0.0359, "step": 6560 }, { "grad_norm": 0.4275452494621277, "learning_rate": 9.971833518143938e-05, "loss": 0.0383, "step": 6570 }, { "grad_norm": 0.30175328254699707, "learning_rate": 9.971614019751093e-05, "loss": 0.0315, "step": 6580 }, { "grad_norm": 0.3535175919532776, "learning_rate": 9.971393671845383e-05, "loss": 0.0331, "step": 6590 }, { "grad_norm": 0.34291931986808777, "learning_rate": 9.971172474464464e-05, "loss": 0.037, "step": 6600 }, { "grad_norm": 0.26116055250167847, "learning_rate": 9.97095042764613e-05, "loss": 0.037, "step": 6610 }, { "grad_norm": 0.3246641457080841, "learning_rate": 9.970727531428324e-05, "loss": 0.0292, "step": 6620 }, { "grad_norm": 0.35776904225349426, "learning_rate": 9.970503785849132e-05, "loss": 0.0383, "step": 6630 }, { "grad_norm": 0.35751572251319885, "learning_rate": 9.970279190946788e-05, "loss": 0.0357, "step": 6640 }, { "grad_norm": 0.27430471777915955, "learning_rate": 9.970053746759667e-05, "loss": 0.0329, "step": 6650 }, { "grad_norm": 0.3194725215435028, "learning_rate": 9.969827453326292e-05, "loss": 0.0376, "step": 6660 }, { "grad_norm": 0.3701923191547394, "learning_rate": 9.969600310685332e-05, "loss": 0.0364, "step": 6670 }, { "grad_norm": 0.3049522340297699, "learning_rate": 9.969372318875596e-05, "loss": 0.0364, "step": 6680 }, { "grad_norm": 0.40526798367500305, "learning_rate": 9.969143477936043e-05, "loss": 0.0366, "step": 6690 }, { "grad_norm": 0.3317222595214844, "learning_rate": 9.968913787905775e-05, "loss": 0.0356, "step": 6700 }, { "grad_norm": 0.33450639247894287, "learning_rate": 9.968683248824045e-05, "loss": 0.0354, "step": 6710 }, { "grad_norm": 0.32762616872787476, "learning_rate": 9.968451860730238e-05, "loss": 0.0354, "step": 6720 }, { "grad_norm": 0.35022491216659546, "learning_rate": 9.968219623663896e-05, "loss": 0.0328, "step": 6730 }, { "grad_norm": 0.26442989706993103, "learning_rate": 9.967986537664702e-05, "loss": 0.0334, "step": 6740 }, { "grad_norm": 0.3442971706390381, "learning_rate": 9.967752602772483e-05, "loss": 0.0309, "step": 6750 }, { "grad_norm": 0.3364148437976837, "learning_rate": 9.967517819027212e-05, "loss": 0.0305, "step": 6760 }, { "grad_norm": 0.36402344703674316, "learning_rate": 9.967282186469009e-05, "loss": 0.0296, "step": 6770 }, { "grad_norm": 0.3396092653274536, "learning_rate": 9.967045705138135e-05, "loss": 0.0326, "step": 6780 }, { "grad_norm": 0.26374322175979614, "learning_rate": 9.966808375074998e-05, "loss": 0.0331, "step": 6790 }, { "grad_norm": 0.3070227801799774, "learning_rate": 9.966570196320154e-05, "loss": 0.0337, "step": 6800 }, { "grad_norm": 0.32372409105300903, "learning_rate": 9.966331168914299e-05, "loss": 0.034, "step": 6810 }, { "grad_norm": 0.3470175266265869, "learning_rate": 9.966091292898277e-05, "loss": 0.0341, "step": 6820 }, { "grad_norm": 0.47592228651046753, "learning_rate": 9.965850568313076e-05, "loss": 0.0311, "step": 6830 }, { "grad_norm": 0.2367602437734604, "learning_rate": 9.965608995199827e-05, "loss": 0.0324, "step": 6840 }, { "grad_norm": 0.3112695515155792, "learning_rate": 9.965366573599812e-05, "loss": 0.0352, "step": 6850 }, { "grad_norm": 0.3561722934246063, "learning_rate": 9.965123303554453e-05, "loss": 0.0388, "step": 6860 }, { "grad_norm": 0.2608005702495575, "learning_rate": 9.964879185105317e-05, "loss": 0.0311, "step": 6870 }, { "grad_norm": 0.2737438380718231, "learning_rate": 9.964634218294119e-05, "loss": 0.0379, "step": 6880 }, { "grad_norm": 0.264662504196167, "learning_rate": 9.964388403162714e-05, "loss": 0.0293, "step": 6890 }, { "grad_norm": 0.2972363531589508, "learning_rate": 9.96414173975311e-05, "loss": 0.0305, "step": 6900 }, { "grad_norm": 0.27286532521247864, "learning_rate": 9.963894228107451e-05, "loss": 0.0317, "step": 6910 }, { "grad_norm": 0.38336804509162903, "learning_rate": 9.963645868268032e-05, "loss": 0.0336, "step": 6920 }, { "grad_norm": 0.2651558220386505, "learning_rate": 9.963396660277289e-05, "loss": 0.0328, "step": 6930 }, { "grad_norm": 0.25609463453292847, "learning_rate": 9.963146604177807e-05, "loss": 0.0301, "step": 6940 }, { "grad_norm": 0.38214823603630066, "learning_rate": 9.962895700012311e-05, "loss": 0.0314, "step": 6950 }, { "grad_norm": 0.2872646450996399, "learning_rate": 9.962643947823677e-05, "loss": 0.0324, "step": 6960 }, { "grad_norm": 0.2868938148021698, "learning_rate": 9.962391347654921e-05, "loss": 0.0301, "step": 6970 }, { "grad_norm": 0.39823397994041443, "learning_rate": 9.962137899549204e-05, "loss": 0.0334, "step": 6980 }, { "grad_norm": 0.3423458933830261, "learning_rate": 9.961883603549835e-05, "loss": 0.0315, "step": 6990 }, { "grad_norm": 0.23993316292762756, "learning_rate": 9.961628459700267e-05, "loss": 0.0316, "step": 7000 }, { "grad_norm": 0.3461841940879822, "learning_rate": 9.961372468044095e-05, "loss": 0.0351, "step": 7010 }, { "grad_norm": 0.35355862975120544, "learning_rate": 9.961115628625062e-05, "loss": 0.0349, "step": 7020 }, { "grad_norm": 0.342862069606781, "learning_rate": 9.960857941487056e-05, "loss": 0.0286, "step": 7030 }, { "grad_norm": 0.33051595091819763, "learning_rate": 9.960599406674106e-05, "loss": 0.0319, "step": 7040 }, { "grad_norm": 0.29291823506355286, "learning_rate": 9.960340024230393e-05, "loss": 0.0318, "step": 7050 }, { "grad_norm": 0.3712957501411438, "learning_rate": 9.960079794200232e-05, "loss": 0.0331, "step": 7060 }, { "grad_norm": 0.3146769404411316, "learning_rate": 9.959818716628096e-05, "loss": 0.0313, "step": 7070 }, { "grad_norm": 0.2686498165130615, "learning_rate": 9.95955679155859e-05, "loss": 0.0308, "step": 7080 }, { "grad_norm": 0.3388281464576721, "learning_rate": 9.959294019036472e-05, "loss": 0.0327, "step": 7090 }, { "grad_norm": 0.3510092496871948, "learning_rate": 9.959030399106646e-05, "loss": 0.0276, "step": 7100 }, { "grad_norm": 0.3059179186820984, "learning_rate": 9.958765931814153e-05, "loss": 0.0327, "step": 7110 }, { "grad_norm": 0.27594587206840515, "learning_rate": 9.958500617204184e-05, "loss": 0.0319, "step": 7120 }, { "grad_norm": 0.34035325050354004, "learning_rate": 9.958234455322075e-05, "loss": 0.0286, "step": 7130 }, { "grad_norm": 0.40871119499206543, "learning_rate": 9.957967446213308e-05, "loss": 0.0332, "step": 7140 }, { "grad_norm": 0.27637606859207153, "learning_rate": 9.957699589923501e-05, "loss": 0.0288, "step": 7150 }, { "grad_norm": 0.3043307960033417, "learning_rate": 9.957430886498431e-05, "loss": 0.031, "step": 7160 }, { "grad_norm": 0.2526138126850128, "learning_rate": 9.957161335984008e-05, "loss": 0.0295, "step": 7170 }, { "grad_norm": 0.314525842666626, "learning_rate": 9.956890938426291e-05, "loss": 0.0312, "step": 7180 }, { "grad_norm": 0.30385854840278625, "learning_rate": 9.956619693871482e-05, "loss": 0.0289, "step": 7190 }, { "grad_norm": 0.2717941999435425, "learning_rate": 9.956347602365934e-05, "loss": 0.03, "step": 7200 }, { "grad_norm": 0.25288820266723633, "learning_rate": 9.956074663956135e-05, "loss": 0.0308, "step": 7210 }, { "grad_norm": 0.2835371494293213, "learning_rate": 9.955800878688726e-05, "loss": 0.0314, "step": 7220 }, { "grad_norm": 0.316801518201828, "learning_rate": 9.955526246610489e-05, "loss": 0.0315, "step": 7230 }, { "grad_norm": 0.35801637172698975, "learning_rate": 9.955250767768349e-05, "loss": 0.0277, "step": 7240 }, { "grad_norm": 0.41948917508125305, "learning_rate": 9.95497444220938e-05, "loss": 0.0299, "step": 7250 }, { "grad_norm": 0.31397855281829834, "learning_rate": 9.954697269980797e-05, "loss": 0.0296, "step": 7260 }, { "grad_norm": 0.4101783037185669, "learning_rate": 9.954419251129962e-05, "loss": 0.0316, "step": 7270 }, { "grad_norm": 0.2892381548881531, "learning_rate": 9.95414038570438e-05, "loss": 0.0288, "step": 7280 }, { "grad_norm": 0.19944261014461517, "learning_rate": 9.953860673751703e-05, "loss": 0.0319, "step": 7290 }, { "grad_norm": 0.24661116302013397, "learning_rate": 9.953580115319725e-05, "loss": 0.0291, "step": 7300 }, { "grad_norm": 0.3387734293937683, "learning_rate": 9.953298710456387e-05, "loss": 0.0293, "step": 7310 }, { "grad_norm": 0.3474023938179016, "learning_rate": 9.953016459209771e-05, "loss": 0.0306, "step": 7320 }, { "grad_norm": 0.30612170696258545, "learning_rate": 9.952733361628108e-05, "loss": 0.0348, "step": 7330 }, { "grad_norm": 0.3130338191986084, "learning_rate": 9.952449417759772e-05, "loss": 0.0322, "step": 7340 }, { "grad_norm": 0.4382362365722656, "learning_rate": 9.952164627653279e-05, "loss": 0.0348, "step": 7350 }, { "grad_norm": 0.31644898653030396, "learning_rate": 9.951878991357292e-05, "loss": 0.0296, "step": 7360 }, { "grad_norm": 0.3211268186569214, "learning_rate": 9.951592508920622e-05, "loss": 0.0295, "step": 7370 }, { "grad_norm": 0.23415659368038177, "learning_rate": 9.951305180392219e-05, "loss": 0.0298, "step": 7380 }, { "grad_norm": 0.2823311984539032, "learning_rate": 9.951017005821178e-05, "loss": 0.0302, "step": 7390 }, { "grad_norm": 0.2597612142562866, "learning_rate": 9.95072798525674e-05, "loss": 0.0303, "step": 7400 }, { "grad_norm": 0.3604186177253723, "learning_rate": 9.950438118748293e-05, "loss": 0.0345, "step": 7410 }, { "grad_norm": 0.25895994901657104, "learning_rate": 9.950147406345366e-05, "loss": 0.0307, "step": 7420 }, { "grad_norm": 0.31738442182540894, "learning_rate": 9.949855848097635e-05, "loss": 0.0368, "step": 7430 }, { "grad_norm": 0.34275129437446594, "learning_rate": 9.949563444054916e-05, "loss": 0.0345, "step": 7440 }, { "grad_norm": 0.38651153445243835, "learning_rate": 9.949270194267178e-05, "loss": 0.0339, "step": 7450 }, { "grad_norm": 0.3494931757450104, "learning_rate": 9.948976098784526e-05, "loss": 0.0324, "step": 7460 }, { "grad_norm": 0.28972023725509644, "learning_rate": 9.948681157657213e-05, "loss": 0.0375, "step": 7470 }, { "grad_norm": 0.2783524692058563, "learning_rate": 9.948385370935638e-05, "loss": 0.0322, "step": 7480 }, { "grad_norm": 0.3416198194026947, "learning_rate": 9.94808873867034e-05, "loss": 0.0333, "step": 7490 }, { "grad_norm": 0.33667856454849243, "learning_rate": 9.947791260912009e-05, "loss": 0.034, "step": 7500 }, { "grad_norm": 0.25375908613204956, "learning_rate": 9.947492937711474e-05, "loss": 0.0295, "step": 7510 }, { "grad_norm": 0.3600921034812927, "learning_rate": 9.947193769119707e-05, "loss": 0.0272, "step": 7520 }, { "grad_norm": 0.3062167167663574, "learning_rate": 9.946893755187834e-05, "loss": 0.031, "step": 7530 }, { "grad_norm": 0.3231102228164673, "learning_rate": 9.946592895967115e-05, "loss": 0.0313, "step": 7540 }, { "grad_norm": 0.3663909435272217, "learning_rate": 9.94629119150896e-05, "loss": 0.0306, "step": 7550 }, { "grad_norm": 0.30482977628707886, "learning_rate": 9.94598864186492e-05, "loss": 0.0341, "step": 7560 }, { "grad_norm": 0.25158149003982544, "learning_rate": 9.945685247086696e-05, "loss": 0.0291, "step": 7570 }, { "grad_norm": 0.25972598791122437, "learning_rate": 9.945381007226129e-05, "loss": 0.0335, "step": 7580 }, { "grad_norm": 0.2933083176612854, "learning_rate": 9.945075922335203e-05, "loss": 0.0262, "step": 7590 }, { "grad_norm": 0.27695053815841675, "learning_rate": 9.944769992466049e-05, "loss": 0.0281, "step": 7600 }, { "grad_norm": 0.3641439378261566, "learning_rate": 9.944463217670945e-05, "loss": 0.0311, "step": 7610 }, { "grad_norm": 0.3399907946586609, "learning_rate": 9.944155598002307e-05, "loss": 0.0306, "step": 7620 }, { "grad_norm": 0.3243508040904999, "learning_rate": 9.943847133512701e-05, "loss": 0.0291, "step": 7630 }, { "grad_norm": 0.3709075152873993, "learning_rate": 9.943537824254834e-05, "loss": 0.0368, "step": 7640 }, { "grad_norm": 0.2818834185600281, "learning_rate": 9.943227670281559e-05, "loss": 0.0331, "step": 7650 }, { "grad_norm": 0.2114933431148529, "learning_rate": 9.942916671645873e-05, "loss": 0.0264, "step": 7660 }, { "grad_norm": 0.29602864384651184, "learning_rate": 9.942604828400916e-05, "loss": 0.0303, "step": 7670 }, { "grad_norm": 0.3240678012371063, "learning_rate": 9.942292140599975e-05, "loss": 0.0316, "step": 7680 }, { "grad_norm": 0.3523012101650238, "learning_rate": 9.94197860829648e-05, "loss": 0.032, "step": 7690 }, { "grad_norm": 0.33539876341819763, "learning_rate": 9.941664231544004e-05, "loss": 0.0324, "step": 7700 }, { "grad_norm": 0.24272127449512482, "learning_rate": 9.941349010396264e-05, "loss": 0.0318, "step": 7710 }, { "grad_norm": 0.34551212191581726, "learning_rate": 9.941032944907125e-05, "loss": 0.0258, "step": 7720 }, { "grad_norm": 0.23443731665611267, "learning_rate": 9.940716035130596e-05, "loss": 0.0303, "step": 7730 }, { "grad_norm": 0.28460556268692017, "learning_rate": 9.940398281120821e-05, "loss": 0.0288, "step": 7740 }, { "grad_norm": 0.31294411420822144, "learning_rate": 9.940079682932102e-05, "loss": 0.0296, "step": 7750 }, { "grad_norm": 0.22121766209602356, "learning_rate": 9.939760240618877e-05, "loss": 0.0304, "step": 7760 }, { "grad_norm": 0.35882335901260376, "learning_rate": 9.939439954235729e-05, "loss": 0.0285, "step": 7770 }, { "grad_norm": 0.2631855905056, "learning_rate": 9.939118823837387e-05, "loss": 0.0292, "step": 7780 }, { "grad_norm": 0.30760452151298523, "learning_rate": 9.938796849478725e-05, "loss": 0.0272, "step": 7790 }, { "grad_norm": 0.2649596929550171, "learning_rate": 9.938474031214755e-05, "loss": 0.0311, "step": 7800 }, { "grad_norm": 0.2083115428686142, "learning_rate": 9.938150369100643e-05, "loss": 0.0327, "step": 7810 }, { "grad_norm": 0.29728788137435913, "learning_rate": 9.93782586319169e-05, "loss": 0.0289, "step": 7820 }, { "grad_norm": 0.2404378056526184, "learning_rate": 9.937500513543348e-05, "loss": 0.0296, "step": 7830 }, { "grad_norm": 0.35104918479919434, "learning_rate": 9.937174320211207e-05, "loss": 0.0306, "step": 7840 }, { "grad_norm": 0.27994590997695923, "learning_rate": 9.936847283251009e-05, "loss": 0.0258, "step": 7850 }, { "grad_norm": 0.2800155282020569, "learning_rate": 9.936519402718632e-05, "loss": 0.0284, "step": 7860 }, { "grad_norm": 0.3109774589538574, "learning_rate": 9.936190678670102e-05, "loss": 0.0276, "step": 7870 }, { "grad_norm": 0.2868209183216095, "learning_rate": 9.935861111161593e-05, "loss": 0.0331, "step": 7880 }, { "grad_norm": 0.2520010769367218, "learning_rate": 9.935530700249416e-05, "loss": 0.026, "step": 7890 }, { "grad_norm": 0.22236652672290802, "learning_rate": 9.935199445990028e-05, "loss": 0.0293, "step": 7900 }, { "grad_norm": 0.26979514956474304, "learning_rate": 9.934867348440033e-05, "loss": 0.0352, "step": 7910 }, { "grad_norm": 0.2832021117210388, "learning_rate": 9.934534407656176e-05, "loss": 0.0291, "step": 7920 }, { "grad_norm": 0.2936552166938782, "learning_rate": 9.93420062369535e-05, "loss": 0.0315, "step": 7930 }, { "grad_norm": 0.2743794023990631, "learning_rate": 9.933865996614589e-05, "loss": 0.0316, "step": 7940 }, { "grad_norm": 0.26523834466934204, "learning_rate": 9.933530526471068e-05, "loss": 0.0279, "step": 7950 }, { "grad_norm": 0.31827616691589355, "learning_rate": 9.933194213322114e-05, "loss": 0.0299, "step": 7960 }, { "grad_norm": 0.2624361217021942, "learning_rate": 9.932857057225192e-05, "loss": 0.0295, "step": 7970 }, { "grad_norm": 0.3376995623111725, "learning_rate": 9.932519058237912e-05, "loss": 0.0291, "step": 7980 }, { "grad_norm": 0.22986117005348206, "learning_rate": 9.932180216418032e-05, "loss": 0.0264, "step": 7990 }, { "grad_norm": 0.27322039008140564, "learning_rate": 9.931840531823446e-05, "loss": 0.0274, "step": 8000 }, { "grad_norm": 0.3751048743724823, "learning_rate": 9.9315000045122e-05, "loss": 0.0292, "step": 8010 }, { "grad_norm": 0.310342013835907, "learning_rate": 9.931158634542481e-05, "loss": 0.0323, "step": 8020 }, { "grad_norm": 0.2686484158039093, "learning_rate": 9.930816421972617e-05, "loss": 0.0266, "step": 8030 }, { "grad_norm": 0.30585983395576477, "learning_rate": 9.930473366861086e-05, "loss": 0.0328, "step": 8040 }, { "grad_norm": 0.31397712230682373, "learning_rate": 9.930129469266505e-05, "loss": 0.0274, "step": 8050 }, { "grad_norm": 0.2483936846256256, "learning_rate": 9.929784729247638e-05, "loss": 0.0263, "step": 8060 }, { "grad_norm": 0.33135929703712463, "learning_rate": 9.929439146863389e-05, "loss": 0.0295, "step": 8070 }, { "grad_norm": 0.35986220836639404, "learning_rate": 9.92909272217281e-05, "loss": 0.0328, "step": 8080 }, { "grad_norm": 0.3505386412143707, "learning_rate": 9.928745455235097e-05, "loss": 0.0281, "step": 8090 }, { "grad_norm": 0.24151688814163208, "learning_rate": 9.928397346109588e-05, "loss": 0.031, "step": 8100 }, { "grad_norm": 0.2716277539730072, "learning_rate": 9.928048394855762e-05, "loss": 0.0319, "step": 8110 }, { "grad_norm": 0.25290295481681824, "learning_rate": 9.92769860153325e-05, "loss": 0.0311, "step": 8120 }, { "grad_norm": 0.2370290458202362, "learning_rate": 9.927347966201819e-05, "loss": 0.03, "step": 8130 }, { "grad_norm": 0.34491756558418274, "learning_rate": 9.926996488921383e-05, "loss": 0.0312, "step": 8140 }, { "grad_norm": 0.23664215207099915, "learning_rate": 9.926644169752001e-05, "loss": 0.029, "step": 8150 }, { "grad_norm": 0.2824476361274719, "learning_rate": 9.926291008753875e-05, "loss": 0.0296, "step": 8160 }, { "grad_norm": 0.25934457778930664, "learning_rate": 9.92593700598735e-05, "loss": 0.0287, "step": 8170 }, { "grad_norm": 0.3207644522190094, "learning_rate": 9.925582161512915e-05, "loss": 0.0294, "step": 8180 }, { "grad_norm": 0.37441304326057434, "learning_rate": 9.925226475391205e-05, "loss": 0.0281, "step": 8190 }, { "grad_norm": 0.25758448243141174, "learning_rate": 9.924869947682993e-05, "loss": 0.0256, "step": 8200 }, { "grad_norm": 0.32591453194618225, "learning_rate": 9.924512578449204e-05, "loss": 0.0282, "step": 8210 }, { "grad_norm": 0.2768457531929016, "learning_rate": 9.924154367750901e-05, "loss": 0.0248, "step": 8220 }, { "grad_norm": 0.3185722827911377, "learning_rate": 9.923795315649293e-05, "loss": 0.0264, "step": 8230 }, { "grad_norm": 0.1953597366809845, "learning_rate": 9.92343542220573e-05, "loss": 0.0263, "step": 8240 }, { "grad_norm": 0.27554699778556824, "learning_rate": 9.92307468748171e-05, "loss": 0.0265, "step": 8250 }, { "grad_norm": 0.28353458642959595, "learning_rate": 9.922713111538873e-05, "loss": 0.0276, "step": 8260 }, { "grad_norm": 0.21622861921787262, "learning_rate": 9.922350694439003e-05, "loss": 0.0314, "step": 8270 }, { "grad_norm": 0.25771114230155945, "learning_rate": 9.921987436244024e-05, "loss": 0.0309, "step": 8280 }, { "grad_norm": 0.36892539262771606, "learning_rate": 9.921623337016008e-05, "loss": 0.0284, "step": 8290 }, { "grad_norm": 0.31864258646965027, "learning_rate": 9.921258396817172e-05, "loss": 0.0264, "step": 8300 }, { "grad_norm": 0.23525747656822205, "learning_rate": 9.920892615709874e-05, "loss": 0.0306, "step": 8310 }, { "grad_norm": 0.31325310468673706, "learning_rate": 9.920525993756612e-05, "loss": 0.0282, "step": 8320 }, { "grad_norm": 0.2787887752056122, "learning_rate": 9.920158531020036e-05, "loss": 0.0257, "step": 8330 }, { "grad_norm": 0.31143873929977417, "learning_rate": 9.919790227562933e-05, "loss": 0.0262, "step": 8340 }, { "grad_norm": 0.2586553394794464, "learning_rate": 9.919421083448237e-05, "loss": 0.0267, "step": 8350 }, { "grad_norm": 0.27026045322418213, "learning_rate": 9.919051098739022e-05, "loss": 0.0254, "step": 8360 }, { "grad_norm": 0.22813290357589722, "learning_rate": 9.918680273498514e-05, "loss": 0.027, "step": 8370 }, { "grad_norm": 0.33415159583091736, "learning_rate": 9.918308607790072e-05, "loss": 0.0284, "step": 8380 }, { "grad_norm": 0.2704885005950928, "learning_rate": 9.917936101677205e-05, "loss": 0.0287, "step": 8390 }, { "grad_norm": 0.3874081075191498, "learning_rate": 9.917562755223564e-05, "loss": 0.0269, "step": 8400 }, { "grad_norm": 0.35243770480155945, "learning_rate": 9.917188568492944e-05, "loss": 0.0289, "step": 8410 }, { "grad_norm": 0.3592638075351715, "learning_rate": 9.916813541549283e-05, "loss": 0.0291, "step": 8420 }, { "grad_norm": 0.3778051435947418, "learning_rate": 9.916437674456663e-05, "loss": 0.0267, "step": 8430 }, { "grad_norm": 0.27459993958473206, "learning_rate": 9.916060967279308e-05, "loss": 0.029, "step": 8440 }, { "grad_norm": 0.2640787363052368, "learning_rate": 9.91568342008159e-05, "loss": 0.0273, "step": 8450 }, { "grad_norm": 0.3583521246910095, "learning_rate": 9.915305032928019e-05, "loss": 0.0296, "step": 8460 }, { "grad_norm": 0.27528446912765503, "learning_rate": 9.914925805883253e-05, "loss": 0.0298, "step": 8470 }, { "grad_norm": 0.25999680161476135, "learning_rate": 9.914545739012088e-05, "loss": 0.0276, "step": 8480 }, { "grad_norm": 0.2584124803543091, "learning_rate": 9.91416483237947e-05, "loss": 0.0263, "step": 8490 }, { "grad_norm": 0.32234764099121094, "learning_rate": 9.913783086050485e-05, "loss": 0.0276, "step": 8500 }, { "grad_norm": 0.26476675271987915, "learning_rate": 9.913400500090364e-05, "loss": 0.0297, "step": 8510 }, { "grad_norm": 0.35436779260635376, "learning_rate": 9.913017074564479e-05, "loss": 0.0297, "step": 8520 }, { "grad_norm": 0.28617146611213684, "learning_rate": 9.912632809538348e-05, "loss": 0.0321, "step": 8530 }, { "grad_norm": 0.3194282352924347, "learning_rate": 9.912247705077629e-05, "loss": 0.0332, "step": 8540 }, { "grad_norm": 0.2676764726638794, "learning_rate": 9.911861761248127e-05, "loss": 0.0285, "step": 8550 }, { "grad_norm": 0.32962220907211304, "learning_rate": 9.91147497811579e-05, "loss": 0.0305, "step": 8560 }, { "grad_norm": 0.279344767332077, "learning_rate": 9.911087355746709e-05, "loss": 0.0261, "step": 8570 }, { "grad_norm": 0.25487497448921204, "learning_rate": 9.910698894207117e-05, "loss": 0.0283, "step": 8580 }, { "grad_norm": 0.3410746455192566, "learning_rate": 9.910309593563392e-05, "loss": 0.0282, "step": 8590 }, { "grad_norm": 0.23029069602489471, "learning_rate": 9.909919453882057e-05, "loss": 0.0263, "step": 8600 }, { "grad_norm": 0.2624494135379791, "learning_rate": 9.90952847522977e-05, "loss": 0.0261, "step": 8610 }, { "grad_norm": 0.2780649662017822, "learning_rate": 9.909136657673346e-05, "loss": 0.027, "step": 8620 }, { "grad_norm": 0.30223914980888367, "learning_rate": 9.908744001279731e-05, "loss": 0.0239, "step": 8630 }, { "grad_norm": 0.2568209171295166, "learning_rate": 9.90835050611602e-05, "loss": 0.0265, "step": 8640 }, { "grad_norm": 0.2870122492313385, "learning_rate": 9.90795617224945e-05, "loss": 0.0273, "step": 8650 }, { "grad_norm": 0.2963080406188965, "learning_rate": 9.907560999747405e-05, "loss": 0.0273, "step": 8660 }, { "grad_norm": 0.31362152099609375, "learning_rate": 9.907164988677408e-05, "loss": 0.0293, "step": 8670 }, { "grad_norm": 0.2340802550315857, "learning_rate": 9.906768139107124e-05, "loss": 0.0287, "step": 8680 }, { "grad_norm": 0.32710427045822144, "learning_rate": 9.906370451104367e-05, "loss": 0.0271, "step": 8690 }, { "grad_norm": 0.3108914792537689, "learning_rate": 9.905971924737088e-05, "loss": 0.0244, "step": 8700 }, { "grad_norm": 0.23276133835315704, "learning_rate": 9.905572560073387e-05, "loss": 0.0263, "step": 8710 }, { "grad_norm": 0.24805641174316406, "learning_rate": 9.905172357181501e-05, "loss": 0.023, "step": 8720 }, { "grad_norm": 0.3245159983634949, "learning_rate": 9.904771316129817e-05, "loss": 0.0297, "step": 8730 }, { "grad_norm": 0.24186715483665466, "learning_rate": 9.904369436986862e-05, "loss": 0.0252, "step": 8740 }, { "grad_norm": 0.2591281533241272, "learning_rate": 9.903966719821303e-05, "loss": 0.0299, "step": 8750 }, { "grad_norm": 0.2781001627445221, "learning_rate": 9.903563164701956e-05, "loss": 0.0268, "step": 8760 }, { "grad_norm": 0.2865314185619354, "learning_rate": 9.903158771697778e-05, "loss": 0.0233, "step": 8770 }, { "grad_norm": 0.27870237827301025, "learning_rate": 9.902753540877867e-05, "loss": 0.0242, "step": 8780 }, { "grad_norm": 0.3464711904525757, "learning_rate": 9.902347472311466e-05, "loss": 0.025, "step": 8790 }, { "grad_norm": 0.2405622899532318, "learning_rate": 9.901940566067962e-05, "loss": 0.0307, "step": 8800 }, { "grad_norm": 0.34293925762176514, "learning_rate": 9.901532822216883e-05, "loss": 0.0273, "step": 8810 }, { "grad_norm": 0.33228304982185364, "learning_rate": 9.901124240827904e-05, "loss": 0.0308, "step": 8820 }, { "grad_norm": 0.2495785653591156, "learning_rate": 9.900714821970835e-05, "loss": 0.0327, "step": 8830 }, { "grad_norm": 0.29274892807006836, "learning_rate": 9.900304565715641e-05, "loss": 0.0314, "step": 8840 }, { "grad_norm": 0.29178959131240845, "learning_rate": 9.899893472132419e-05, "loss": 0.0334, "step": 8850 }, { "grad_norm": 0.3137081265449524, "learning_rate": 9.899481541291415e-05, "loss": 0.029, "step": 8860 }, { "grad_norm": 0.3291626572608948, "learning_rate": 9.899068773263016e-05, "loss": 0.032, "step": 8870 }, { "grad_norm": 0.3424113094806671, "learning_rate": 9.898655168117754e-05, "loss": 0.0265, "step": 8880 }, { "grad_norm": 0.2880123555660248, "learning_rate": 9.898240725926302e-05, "loss": 0.0286, "step": 8890 }, { "grad_norm": 0.3278730511665344, "learning_rate": 9.897825446759478e-05, "loss": 0.0238, "step": 8900 }, { "grad_norm": 0.2500796616077423, "learning_rate": 9.897409330688241e-05, "loss": 0.0257, "step": 8910 }, { "grad_norm": 0.2618340849876404, "learning_rate": 9.896992377783692e-05, "loss": 0.0232, "step": 8920 }, { "grad_norm": 0.35392194986343384, "learning_rate": 9.89657458811708e-05, "loss": 0.0256, "step": 8930 }, { "grad_norm": 0.27675551176071167, "learning_rate": 9.896155961759792e-05, "loss": 0.0244, "step": 8940 }, { "grad_norm": 0.4425216019153595, "learning_rate": 9.895736498783361e-05, "loss": 0.0287, "step": 8950 }, { "grad_norm": 0.37865397334098816, "learning_rate": 9.895316199259462e-05, "loss": 0.0307, "step": 8960 }, { "grad_norm": 0.22828209400177002, "learning_rate": 9.894895063259909e-05, "loss": 0.027, "step": 8970 }, { "grad_norm": 0.3536059558391571, "learning_rate": 9.894473090856667e-05, "loss": 0.0303, "step": 8980 }, { "grad_norm": 0.25264859199523926, "learning_rate": 9.894050282121839e-05, "loss": 0.0286, "step": 8990 }, { "grad_norm": 0.18521511554718018, "learning_rate": 9.893626637127668e-05, "loss": 0.0324, "step": 9000 }, { "grad_norm": 0.3343427777290344, "learning_rate": 9.893202155946546e-05, "loss": 0.0295, "step": 9010 }, { "grad_norm": 0.2631494700908661, "learning_rate": 9.892776838651006e-05, "loss": 0.0235, "step": 9020 }, { "grad_norm": 0.30685293674468994, "learning_rate": 9.892350685313722e-05, "loss": 0.0286, "step": 9030 }, { "grad_norm": 0.30925923585891724, "learning_rate": 9.891923696007513e-05, "loss": 0.0288, "step": 9040 }, { "grad_norm": 0.21273507177829742, "learning_rate": 9.891495870805336e-05, "loss": 0.024, "step": 9050 }, { "grad_norm": 0.23723646998405457, "learning_rate": 9.891067209780298e-05, "loss": 0.0259, "step": 9060 }, { "grad_norm": 0.28669309616088867, "learning_rate": 9.890637713005646e-05, "loss": 0.0265, "step": 9070 }, { "grad_norm": 0.2769131362438202, "learning_rate": 9.890207380554767e-05, "loss": 0.0283, "step": 9080 }, { "grad_norm": 0.3294835686683655, "learning_rate": 9.889776212501196e-05, "loss": 0.0285, "step": 9090 }, { "grad_norm": 0.2620512843132019, "learning_rate": 9.889344208918605e-05, "loss": 0.0265, "step": 9100 }, { "grad_norm": 0.3117540180683136, "learning_rate": 9.888911369880812e-05, "loss": 0.0276, "step": 9110 }, { "grad_norm": 0.24844777584075928, "learning_rate": 9.888477695461777e-05, "loss": 0.0276, "step": 9120 }, { "grad_norm": 0.2620811462402344, "learning_rate": 9.888043185735607e-05, "loss": 0.026, "step": 9130 }, { "grad_norm": 0.24476051330566406, "learning_rate": 9.887607840776542e-05, "loss": 0.0273, "step": 9140 }, { "grad_norm": 0.21743415296077728, "learning_rate": 9.887171660658975e-05, "loss": 0.0227, "step": 9150 }, { "grad_norm": 0.3038538694381714, "learning_rate": 9.886734645457435e-05, "loss": 0.0265, "step": 9160 }, { "grad_norm": 0.2976208031177521, "learning_rate": 9.886296795246597e-05, "loss": 0.0283, "step": 9170 }, { "grad_norm": 0.313384473323822, "learning_rate": 9.885858110101276e-05, "loss": 0.0245, "step": 9180 }, { "grad_norm": 0.1951734572649002, "learning_rate": 9.885418590096434e-05, "loss": 0.0254, "step": 9190 }, { "grad_norm": 0.2712096869945526, "learning_rate": 9.88497823530717e-05, "loss": 0.0264, "step": 9200 }, { "grad_norm": 0.2941609025001526, "learning_rate": 9.884537045808732e-05, "loss": 0.0285, "step": 9210 }, { "grad_norm": 0.2617374062538147, "learning_rate": 9.884095021676502e-05, "loss": 0.0288, "step": 9220 }, { "grad_norm": 0.23736511170864105, "learning_rate": 9.883652162986017e-05, "loss": 0.025, "step": 9230 }, { "grad_norm": 0.28262442350387573, "learning_rate": 9.883208469812943e-05, "loss": 0.0274, "step": 9240 }, { "grad_norm": 0.2697758972644806, "learning_rate": 9.882763942233098e-05, "loss": 0.0229, "step": 9250 }, { "grad_norm": 0.30339959263801575, "learning_rate": 9.882318580322441e-05, "loss": 0.0246, "step": 9260 }, { "grad_norm": 0.2987322509288788, "learning_rate": 9.881872384157067e-05, "loss": 0.0251, "step": 9270 }, { "grad_norm": 0.327715128660202, "learning_rate": 9.881425353813225e-05, "loss": 0.0288, "step": 9280 }, { "grad_norm": 0.30393436551094055, "learning_rate": 9.880977489367296e-05, "loss": 0.0265, "step": 9290 }, { "grad_norm": 0.2847493886947632, "learning_rate": 9.88052879089581e-05, "loss": 0.0262, "step": 9300 }, { "grad_norm": 0.3307715654373169, "learning_rate": 9.880079258475434e-05, "loss": 0.023, "step": 9310 }, { "grad_norm": 0.2072727531194687, "learning_rate": 9.879628892182985e-05, "loss": 0.0234, "step": 9320 }, { "grad_norm": 0.28411397337913513, "learning_rate": 9.879177692095416e-05, "loss": 0.0232, "step": 9330 }, { "grad_norm": 0.29028478264808655, "learning_rate": 9.878725658289825e-05, "loss": 0.0233, "step": 9340 }, { "grad_norm": 0.23575714230537415, "learning_rate": 9.878272790843454e-05, "loss": 0.0248, "step": 9350 }, { "grad_norm": 0.29951202869415283, "learning_rate": 9.877819089833682e-05, "loss": 0.0248, "step": 9360 }, { "grad_norm": 0.2444654405117035, "learning_rate": 9.877364555338038e-05, "loss": 0.0223, "step": 9370 }, { "grad_norm": 0.32148370146751404, "learning_rate": 9.876909187434186e-05, "loss": 0.0261, "step": 9380 }, { "grad_norm": 0.21168909966945648, "learning_rate": 9.876452986199939e-05, "loss": 0.0224, "step": 9390 }, { "grad_norm": 0.3129773437976837, "learning_rate": 9.875995951713248e-05, "loss": 0.0252, "step": 9400 }, { "grad_norm": 0.31936511397361755, "learning_rate": 9.875538084052207e-05, "loss": 0.0207, "step": 9410 }, { "grad_norm": 0.33602139353752136, "learning_rate": 9.875079383295053e-05, "loss": 0.0247, "step": 9420 }, { "grad_norm": 0.33414819836616516, "learning_rate": 9.874619849520167e-05, "loss": 0.0284, "step": 9430 }, { "grad_norm": 0.2787718176841736, "learning_rate": 9.874159482806069e-05, "loss": 0.0253, "step": 9440 }, { "grad_norm": 0.22230473160743713, "learning_rate": 9.873698283231426e-05, "loss": 0.028, "step": 9450 }, { "grad_norm": 0.2091957926750183, "learning_rate": 9.87323625087504e-05, "loss": 0.0229, "step": 9460 }, { "grad_norm": 0.2997121512889862, "learning_rate": 9.872773385815863e-05, "loss": 0.0222, "step": 9470 }, { "grad_norm": 0.3393774926662445, "learning_rate": 9.872309688132986e-05, "loss": 0.026, "step": 9480 }, { "grad_norm": 0.21818907558918, "learning_rate": 9.871845157905639e-05, "loss": 0.0251, "step": 9490 }, { "grad_norm": 0.24156902730464935, "learning_rate": 9.871379795213201e-05, "loss": 0.0239, "step": 9500 }, { "grad_norm": 0.26524510979652405, "learning_rate": 9.87091360013519e-05, "loss": 0.0271, "step": 9510 }, { "grad_norm": 0.2618923485279083, "learning_rate": 9.870446572751262e-05, "loss": 0.0223, "step": 9520 }, { "grad_norm": 0.20701199769973755, "learning_rate": 9.869978713141224e-05, "loss": 0.0201, "step": 9530 }, { "grad_norm": 0.2960413694381714, "learning_rate": 9.869510021385016e-05, "loss": 0.0225, "step": 9540 }, { "grad_norm": 0.2374277561903, "learning_rate": 9.869040497562727e-05, "loss": 0.0293, "step": 9550 }, { "grad_norm": 0.26360049843788147, "learning_rate": 9.868570141754587e-05, "loss": 0.027, "step": 9560 }, { "grad_norm": 0.23041941225528717, "learning_rate": 9.868098954040965e-05, "loss": 0.0251, "step": 9570 }, { "grad_norm": 0.24978552758693695, "learning_rate": 9.867626934502374e-05, "loss": 0.0275, "step": 9580 }, { "grad_norm": 0.3058309257030487, "learning_rate": 9.86715408321947e-05, "loss": 0.0287, "step": 9590 }, { "grad_norm": 0.30726197361946106, "learning_rate": 9.86668040027305e-05, "loss": 0.0243, "step": 9600 }, { "grad_norm": 0.22430849075317383, "learning_rate": 9.866205885744053e-05, "loss": 0.0286, "step": 9610 }, { "grad_norm": 0.31294816732406616, "learning_rate": 9.865730539713563e-05, "loss": 0.0226, "step": 9620 }, { "grad_norm": 0.28636959195137024, "learning_rate": 9.8652543622628e-05, "loss": 0.0246, "step": 9630 }, { "grad_norm": 0.3756641745567322, "learning_rate": 9.864777353473132e-05, "loss": 0.03, "step": 9640 }, { "grad_norm": 0.2976136803627014, "learning_rate": 9.864299513426068e-05, "loss": 0.0307, "step": 9650 }, { "grad_norm": 0.2623140513896942, "learning_rate": 9.863820842203254e-05, "loss": 0.0282, "step": 9660 }, { "grad_norm": 0.3400615453720093, "learning_rate": 9.863341339886483e-05, "loss": 0.0231, "step": 9670 }, { "grad_norm": 0.2936703562736511, "learning_rate": 9.86286100655769e-05, "loss": 0.0229, "step": 9680 }, { "grad_norm": 0.24218615889549255, "learning_rate": 9.862379842298953e-05, "loss": 0.0267, "step": 9690 }, { "grad_norm": 0.21970446407794952, "learning_rate": 9.861897847192485e-05, "loss": 0.0252, "step": 9700 }, { "grad_norm": 0.27100232243537903, "learning_rate": 9.86141502132065e-05, "loss": 0.0254, "step": 9710 }, { "grad_norm": 0.2791595160961151, "learning_rate": 9.860931364765946e-05, "loss": 0.025, "step": 9720 }, { "grad_norm": 0.2739853858947754, "learning_rate": 9.860446877611021e-05, "loss": 0.0242, "step": 9730 }, { "grad_norm": 0.25219517946243286, "learning_rate": 9.859961559938655e-05, "loss": 0.0266, "step": 9740 }, { "grad_norm": 0.32021111249923706, "learning_rate": 9.85947541183178e-05, "loss": 0.0256, "step": 9750 }, { "grad_norm": 0.273194283246994, "learning_rate": 9.858988433373463e-05, "loss": 0.0248, "step": 9760 }, { "grad_norm": 0.253176212310791, "learning_rate": 9.858500624646918e-05, "loss": 0.0224, "step": 9770 }, { "grad_norm": 0.24501651525497437, "learning_rate": 9.858011985735497e-05, "loss": 0.0203, "step": 9780 }, { "grad_norm": 0.20256631076335907, "learning_rate": 9.857522516722693e-05, "loss": 0.0217, "step": 9790 }, { "grad_norm": 0.26777559518814087, "learning_rate": 9.857032217692145e-05, "loss": 0.0226, "step": 9800 }, { "grad_norm": 0.265764445066452, "learning_rate": 9.856541088727631e-05, "loss": 0.0306, "step": 9810 }, { "grad_norm": 0.24403178691864014, "learning_rate": 9.856049129913072e-05, "loss": 0.0263, "step": 9820 }, { "grad_norm": 0.2923603355884552, "learning_rate": 9.85555634133253e-05, "loss": 0.0273, "step": 9830 }, { "grad_norm": 0.24516703188419342, "learning_rate": 9.855062723070208e-05, "loss": 0.0273, "step": 9840 }, { "grad_norm": 0.23456716537475586, "learning_rate": 9.854568275210454e-05, "loss": 0.0272, "step": 9850 }, { "grad_norm": 0.34495100378990173, "learning_rate": 9.854072997837754e-05, "loss": 0.0254, "step": 9860 }, { "grad_norm": 0.2452862560749054, "learning_rate": 9.853576891036737e-05, "loss": 0.0203, "step": 9870 }, { "grad_norm": 0.291759729385376, "learning_rate": 9.853079954892177e-05, "loss": 0.0242, "step": 9880 }, { "grad_norm": 0.24348050355911255, "learning_rate": 9.852582189488983e-05, "loss": 0.0232, "step": 9890 }, { "grad_norm": 0.2400985211133957, "learning_rate": 9.852083594912212e-05, "loss": 0.0221, "step": 9900 }, { "grad_norm": 0.2817951738834381, "learning_rate": 9.851584171247058e-05, "loss": 0.0237, "step": 9910 }, { "grad_norm": 0.27593713998794556, "learning_rate": 9.851083918578863e-05, "loss": 0.0222, "step": 9920 }, { "grad_norm": 0.2727307379245758, "learning_rate": 9.850582836993103e-05, "loss": 0.0244, "step": 9930 }, { "grad_norm": 0.2495252639055252, "learning_rate": 9.850080926575397e-05, "loss": 0.0205, "step": 9940 }, { "grad_norm": 0.26649340987205505, "learning_rate": 9.849578187411515e-05, "loss": 0.0294, "step": 9950 }, { "grad_norm": 0.2989371716976166, "learning_rate": 9.849074619587354e-05, "loss": 0.0246, "step": 9960 }, { "grad_norm": 0.31876659393310547, "learning_rate": 9.848570223188964e-05, "loss": 0.0235, "step": 9970 }, { "grad_norm": 0.2575185000896454, "learning_rate": 9.848064998302531e-05, "loss": 0.0203, "step": 9980 }, { "grad_norm": 0.23824641108512878, "learning_rate": 9.847558945014386e-05, "loss": 0.0224, "step": 9990 }, { "grad_norm": 0.2901687026023865, "learning_rate": 9.847052063410996e-05, "loss": 0.023, "step": 10000 }, { "grad_norm": 0.2364916205406189, "learning_rate": 9.846544353578977e-05, "loss": 0.0224, "step": 10010 }, { "grad_norm": 0.2725428342819214, "learning_rate": 9.846035815605081e-05, "loss": 0.0224, "step": 10020 }, { "grad_norm": 0.28419724106788635, "learning_rate": 9.845526449576204e-05, "loss": 0.024, "step": 10030 }, { "grad_norm": 0.24462123215198517, "learning_rate": 9.845016255579383e-05, "loss": 0.0221, "step": 10040 }, { "grad_norm": 0.26677900552749634, "learning_rate": 9.844505233701794e-05, "loss": 0.0206, "step": 10050 }, { "grad_norm": 0.231193408370018, "learning_rate": 9.843993384030757e-05, "loss": 0.0214, "step": 10060 }, { "grad_norm": 0.28157225251197815, "learning_rate": 9.843480706653737e-05, "loss": 0.0218, "step": 10070 }, { "grad_norm": 0.27320972084999084, "learning_rate": 9.84296720165833e-05, "loss": 0.029, "step": 10080 }, { "grad_norm": 0.25023069977760315, "learning_rate": 9.842452869132286e-05, "loss": 0.0225, "step": 10090 }, { "grad_norm": 0.3048573136329651, "learning_rate": 9.841937709163489e-05, "loss": 0.0213, "step": 10100 }, { "grad_norm": 0.23919513821601868, "learning_rate": 9.841421721839962e-05, "loss": 0.02, "step": 10110 }, { "grad_norm": 0.3072323799133301, "learning_rate": 9.840904907249879e-05, "loss": 0.0235, "step": 10120 }, { "grad_norm": 0.2863621115684509, "learning_rate": 9.840387265481545e-05, "loss": 0.0246, "step": 10130 }, { "grad_norm": 0.2856663763523102, "learning_rate": 9.839868796623411e-05, "loss": 0.0214, "step": 10140 }, { "grad_norm": 0.26640549302101135, "learning_rate": 9.839349500764072e-05, "loss": 0.0232, "step": 10150 }, { "grad_norm": 0.34254148602485657, "learning_rate": 9.83882937799226e-05, "loss": 0.0252, "step": 10160 }, { "grad_norm": 0.2528793513774872, "learning_rate": 9.838308428396849e-05, "loss": 0.0246, "step": 10170 }, { "grad_norm": 0.23682279884815216, "learning_rate": 9.837786652066854e-05, "loss": 0.0249, "step": 10180 }, { "grad_norm": 0.297897070646286, "learning_rate": 9.837264049091437e-05, "loss": 0.0233, "step": 10190 }, { "grad_norm": 0.3181309401988983, "learning_rate": 9.836740619559893e-05, "loss": 0.023, "step": 10200 }, { "grad_norm": 0.33556243777275085, "learning_rate": 9.836216363561659e-05, "loss": 0.0273, "step": 10210 }, { "grad_norm": 0.3347345292568207, "learning_rate": 9.835691281186322e-05, "loss": 0.0274, "step": 10220 }, { "grad_norm": 0.30249112844467163, "learning_rate": 9.8351653725236e-05, "loss": 0.0236, "step": 10230 }, { "grad_norm": 0.2021283656358719, "learning_rate": 9.83463863766336e-05, "loss": 0.0215, "step": 10240 }, { "grad_norm": 0.2570987343788147, "learning_rate": 9.834111076695602e-05, "loss": 0.0205, "step": 10250 }, { "grad_norm": 0.24541670083999634, "learning_rate": 9.833582689710477e-05, "loss": 0.0254, "step": 10260 }, { "grad_norm": 0.2833375334739685, "learning_rate": 9.833053476798268e-05, "loss": 0.0261, "step": 10270 }, { "grad_norm": 0.2866086959838867, "learning_rate": 9.832523438049404e-05, "loss": 0.0248, "step": 10280 }, { "grad_norm": 0.2946564257144928, "learning_rate": 9.831992573554454e-05, "loss": 0.0206, "step": 10290 }, { "grad_norm": 0.2982623279094696, "learning_rate": 9.831460883404128e-05, "loss": 0.0238, "step": 10300 }, { "grad_norm": 0.2756432592868805, "learning_rate": 9.830928367689278e-05, "loss": 0.0228, "step": 10310 }, { "grad_norm": 0.256780207157135, "learning_rate": 9.830395026500896e-05, "loss": 0.0211, "step": 10320 }, { "grad_norm": 0.2476087063550949, "learning_rate": 9.829860859930115e-05, "loss": 0.0243, "step": 10330 }, { "grad_norm": 0.286111056804657, "learning_rate": 9.829325868068212e-05, "loss": 0.0214, "step": 10340 }, { "grad_norm": 0.37515103816986084, "learning_rate": 9.8287900510066e-05, "loss": 0.0274, "step": 10350 }, { "grad_norm": 0.3188357651233673, "learning_rate": 9.828253408836834e-05, "loss": 0.026, "step": 10360 }, { "grad_norm": 0.3494161367416382, "learning_rate": 9.827715941650615e-05, "loss": 0.0241, "step": 10370 }, { "grad_norm": 0.30381909012794495, "learning_rate": 9.82717764953978e-05, "loss": 0.0223, "step": 10380 }, { "grad_norm": 0.2660832107067108, "learning_rate": 9.826638532596308e-05, "loss": 0.0227, "step": 10390 }, { "grad_norm": 0.2528570890426636, "learning_rate": 9.82609859091232e-05, "loss": 0.0225, "step": 10400 }, { "grad_norm": 0.279561311006546, "learning_rate": 9.825557824580076e-05, "loss": 0.0224, "step": 10410 }, { "grad_norm": 0.2073567807674408, "learning_rate": 9.82501623369198e-05, "loss": 0.0179, "step": 10420 }, { "grad_norm": 0.21019811928272247, "learning_rate": 9.824473818340574e-05, "loss": 0.022, "step": 10430 }, { "grad_norm": 0.25286808609962463, "learning_rate": 9.823930578618541e-05, "loss": 0.0213, "step": 10440 }, { "grad_norm": 0.2732088565826416, "learning_rate": 9.823386514618709e-05, "loss": 0.0212, "step": 10450 }, { "grad_norm": 0.23915117979049683, "learning_rate": 9.82284162643404e-05, "loss": 0.0215, "step": 10460 }, { "grad_norm": 0.22454191744327545, "learning_rate": 9.822295914157642e-05, "loss": 0.0217, "step": 10470 }, { "grad_norm": 0.2789269685745239, "learning_rate": 9.821749377882763e-05, "loss": 0.0211, "step": 10480 }, { "grad_norm": 0.253140389919281, "learning_rate": 9.821202017702791e-05, "loss": 0.0212, "step": 10490 }, { "grad_norm": 0.34793907403945923, "learning_rate": 9.820653833711253e-05, "loss": 0.0213, "step": 10500 }, { "grad_norm": 0.24965886771678925, "learning_rate": 9.820104826001822e-05, "loss": 0.0227, "step": 10510 }, { "grad_norm": 0.33400869369506836, "learning_rate": 9.819554994668305e-05, "loss": 0.0254, "step": 10520 }, { "grad_norm": 0.2663857936859131, "learning_rate": 9.819004339804654e-05, "loss": 0.0237, "step": 10530 }, { "grad_norm": 0.2966526746749878, "learning_rate": 9.818452861504961e-05, "loss": 0.0245, "step": 10540 }, { "grad_norm": 0.3291342258453369, "learning_rate": 9.81790055986346e-05, "loss": 0.026, "step": 10550 }, { "grad_norm": 0.19958826899528503, "learning_rate": 9.817347434974523e-05, "loss": 0.0215, "step": 10560 }, { "grad_norm": 0.2816254794597626, "learning_rate": 9.816793486932664e-05, "loss": 0.0229, "step": 10570 }, { "grad_norm": 0.2153339684009552, "learning_rate": 9.816238715832538e-05, "loss": 0.0199, "step": 10580 }, { "grad_norm": 0.23130807280540466, "learning_rate": 9.815683121768939e-05, "loss": 0.025, "step": 10590 }, { "grad_norm": 0.23872573673725128, "learning_rate": 9.815126704836804e-05, "loss": 0.0233, "step": 10600 }, { "grad_norm": 0.20730245113372803, "learning_rate": 9.81456946513121e-05, "loss": 0.0257, "step": 10610 }, { "grad_norm": 0.23323945701122284, "learning_rate": 9.814011402747373e-05, "loss": 0.0265, "step": 10620 }, { "grad_norm": 0.2826184332370758, "learning_rate": 9.813452517780651e-05, "loss": 0.0225, "step": 10630 }, { "grad_norm": 0.37264353036880493, "learning_rate": 9.81289281032654e-05, "loss": 0.0266, "step": 10640 }, { "grad_norm": 0.2612908184528351, "learning_rate": 9.812332280480683e-05, "loss": 0.0226, "step": 10650 }, { "grad_norm": 0.2199002206325531, "learning_rate": 9.811770928338854e-05, "loss": 0.0229, "step": 10660 }, { "grad_norm": 0.17794561386108398, "learning_rate": 9.811208753996979e-05, "loss": 0.0202, "step": 10670 }, { "grad_norm": 0.2398097962141037, "learning_rate": 9.810645757551113e-05, "loss": 0.019, "step": 10680 }, { "grad_norm": 0.2774912714958191, "learning_rate": 9.810081939097459e-05, "loss": 0.0204, "step": 10690 }, { "grad_norm": 0.21221177279949188, "learning_rate": 9.809517298732356e-05, "loss": 0.0226, "step": 10700 }, { "grad_norm": 0.35394740104675293, "learning_rate": 9.80895183655229e-05, "loss": 0.0212, "step": 10710 }, { "grad_norm": 0.22063297033309937, "learning_rate": 9.808385552653877e-05, "loss": 0.024, "step": 10720 }, { "grad_norm": 0.1957472264766693, "learning_rate": 9.807818447133886e-05, "loss": 0.0226, "step": 10730 }, { "grad_norm": 0.3313548266887665, "learning_rate": 9.807250520089215e-05, "loss": 0.0196, "step": 10740 }, { "grad_norm": 0.2190708965063095, "learning_rate": 9.806681771616908e-05, "loss": 0.02, "step": 10750 }, { "grad_norm": 0.23964525759220123, "learning_rate": 9.80611220181415e-05, "loss": 0.0217, "step": 10760 }, { "grad_norm": 0.24396094679832458, "learning_rate": 9.805541810778264e-05, "loss": 0.0216, "step": 10770 }, { "grad_norm": 0.2300400733947754, "learning_rate": 9.804970598606716e-05, "loss": 0.0213, "step": 10780 }, { "grad_norm": 0.24982690811157227, "learning_rate": 9.804398565397106e-05, "loss": 0.0278, "step": 10790 }, { "grad_norm": 0.20676158368587494, "learning_rate": 9.803825711247183e-05, "loss": 0.0238, "step": 10800 }, { "grad_norm": 0.22856757044792175, "learning_rate": 9.803252036254831e-05, "loss": 0.0201, "step": 10810 }, { "grad_norm": 0.25168484449386597, "learning_rate": 9.802677540518076e-05, "loss": 0.0255, "step": 10820 }, { "grad_norm": 0.18558180332183838, "learning_rate": 9.802102224135081e-05, "loss": 0.0243, "step": 10830 }, { "grad_norm": 0.2661195695400238, "learning_rate": 9.801526087204155e-05, "loss": 0.0229, "step": 10840 }, { "grad_norm": 0.2133924663066864, "learning_rate": 9.800949129823743e-05, "loss": 0.0194, "step": 10850 }, { "grad_norm": 0.2410089075565338, "learning_rate": 9.80037135209243e-05, "loss": 0.0232, "step": 10860 }, { "grad_norm": 0.22677341103553772, "learning_rate": 9.799792754108946e-05, "loss": 0.0224, "step": 10870 }, { "grad_norm": 0.27355825901031494, "learning_rate": 9.799213335972152e-05, "loss": 0.0216, "step": 10880 }, { "grad_norm": 0.21076233685016632, "learning_rate": 9.798633097781058e-05, "loss": 0.0242, "step": 10890 }, { "grad_norm": 0.2513548731803894, "learning_rate": 9.79805203963481e-05, "loss": 0.02, "step": 10900 }, { "grad_norm": 0.19351020455360413, "learning_rate": 9.797470161632697e-05, "loss": 0.0225, "step": 10910 }, { "grad_norm": 0.2034849375486374, "learning_rate": 9.796887463874145e-05, "loss": 0.0223, "step": 10920 }, { "grad_norm": 0.27106401324272156, "learning_rate": 9.796303946458718e-05, "loss": 0.0207, "step": 10930 }, { "grad_norm": 0.24355590343475342, "learning_rate": 9.795719609486127e-05, "loss": 0.0218, "step": 10940 }, { "grad_norm": 0.20170798897743225, "learning_rate": 9.795134453056219e-05, "loss": 0.0187, "step": 10950 }, { "grad_norm": 0.2628694474697113, "learning_rate": 9.794548477268979e-05, "loss": 0.019, "step": 10960 }, { "grad_norm": 0.18226423859596252, "learning_rate": 9.793961682224537e-05, "loss": 0.0184, "step": 10970 }, { "grad_norm": 0.18573971092700958, "learning_rate": 9.793374068023156e-05, "loss": 0.0213, "step": 10980 }, { "grad_norm": 0.2814602851867676, "learning_rate": 9.792785634765247e-05, "loss": 0.0241, "step": 10990 }, { "grad_norm": 0.24594654142856598, "learning_rate": 9.792196382551357e-05, "loss": 0.0215, "step": 11000 }, { "grad_norm": 0.19445006549358368, "learning_rate": 9.791606311482171e-05, "loss": 0.0196, "step": 11010 }, { "grad_norm": 0.19854941964149475, "learning_rate": 9.791015421658518e-05, "loss": 0.0219, "step": 11020 }, { "grad_norm": 0.24307556450366974, "learning_rate": 9.790423713181362e-05, "loss": 0.0223, "step": 11030 }, { "grad_norm": 0.21011842787265778, "learning_rate": 9.789831186151814e-05, "loss": 0.0206, "step": 11040 }, { "grad_norm": 0.35026347637176514, "learning_rate": 9.789237840671118e-05, "loss": 0.0232, "step": 11050 }, { "grad_norm": 0.28576669096946716, "learning_rate": 9.78864367684066e-05, "loss": 0.0211, "step": 11060 }, { "grad_norm": 0.2300906777381897, "learning_rate": 9.788048694761968e-05, "loss": 0.0213, "step": 11070 }, { "grad_norm": 0.25583192706108093, "learning_rate": 9.787452894536709e-05, "loss": 0.0224, "step": 11080 }, { "grad_norm": 0.2542971670627594, "learning_rate": 9.786856276266685e-05, "loss": 0.0196, "step": 11090 }, { "grad_norm": 0.22022055089473724, "learning_rate": 9.786258840053845e-05, "loss": 0.0196, "step": 11100 }, { "grad_norm": 0.26894623041152954, "learning_rate": 9.785660586000273e-05, "loss": 0.0204, "step": 11110 }, { "grad_norm": 0.2761552929878235, "learning_rate": 9.785061514208196e-05, "loss": 0.0201, "step": 11120 }, { "grad_norm": 0.1929217278957367, "learning_rate": 9.784461624779977e-05, "loss": 0.022, "step": 11130 }, { "grad_norm": 0.24248257279396057, "learning_rate": 9.783860917818123e-05, "loss": 0.0217, "step": 11140 }, { "grad_norm": 0.2564442455768585, "learning_rate": 9.783259393425277e-05, "loss": 0.0237, "step": 11150 }, { "grad_norm": 0.21255779266357422, "learning_rate": 9.782657051704221e-05, "loss": 0.0177, "step": 11160 }, { "grad_norm": 0.24567356705665588, "learning_rate": 9.782053892757883e-05, "loss": 0.0202, "step": 11170 }, { "grad_norm": 0.18994764983654022, "learning_rate": 9.781449916689324e-05, "loss": 0.0228, "step": 11180 }, { "grad_norm": 0.2392175942659378, "learning_rate": 9.780845123601746e-05, "loss": 0.0195, "step": 11190 }, { "grad_norm": 0.1516280174255371, "learning_rate": 9.780239513598492e-05, "loss": 0.0197, "step": 11200 }, { "grad_norm": 0.24005086719989777, "learning_rate": 9.779633086783047e-05, "loss": 0.0242, "step": 11210 }, { "grad_norm": 0.26286813616752625, "learning_rate": 9.779025843259031e-05, "loss": 0.0194, "step": 11220 }, { "grad_norm": 0.23991088569164276, "learning_rate": 9.778417783130204e-05, "loss": 0.0203, "step": 11230 }, { "grad_norm": 0.2578740417957306, "learning_rate": 9.777808906500468e-05, "loss": 0.0186, "step": 11240 }, { "grad_norm": 0.22028635442256927, "learning_rate": 9.777199213473862e-05, "loss": 0.0206, "step": 11250 }, { "grad_norm": 0.2258797436952591, "learning_rate": 9.77658870415457e-05, "loss": 0.0249, "step": 11260 }, { "grad_norm": 0.36247143149375916, "learning_rate": 9.775977378646906e-05, "loss": 0.026, "step": 11270 }, { "grad_norm": 0.2010679990053177, "learning_rate": 9.775365237055331e-05, "loss": 0.0199, "step": 11280 }, { "grad_norm": 0.27493247389793396, "learning_rate": 9.774752279484445e-05, "loss": 0.0195, "step": 11290 }, { "grad_norm": 0.26263102889060974, "learning_rate": 9.774138506038984e-05, "loss": 0.0213, "step": 11300 }, { "grad_norm": 0.2703741490840912, "learning_rate": 9.773523916823826e-05, "loss": 0.0255, "step": 11310 }, { "grad_norm": 0.21234829723834991, "learning_rate": 9.772908511943986e-05, "loss": 0.0195, "step": 11320 }, { "grad_norm": 0.2516130805015564, "learning_rate": 9.77229229150462e-05, "loss": 0.0209, "step": 11330 }, { "grad_norm": 0.21666017174720764, "learning_rate": 9.771675255611024e-05, "loss": 0.023, "step": 11340 }, { "grad_norm": 0.2344212532043457, "learning_rate": 9.771057404368632e-05, "loss": 0.0229, "step": 11350 }, { "grad_norm": 0.2201387882232666, "learning_rate": 9.770438737883018e-05, "loss": 0.0193, "step": 11360 }, { "grad_norm": 0.3038453459739685, "learning_rate": 9.769819256259898e-05, "loss": 0.022, "step": 11370 }, { "grad_norm": 0.27618709206581116, "learning_rate": 9.769198959605119e-05, "loss": 0.0241, "step": 11380 }, { "grad_norm": 0.22515647113323212, "learning_rate": 9.768577848024678e-05, "loss": 0.0224, "step": 11390 }, { "grad_norm": 0.24324388802051544, "learning_rate": 9.767955921624702e-05, "loss": 0.0219, "step": 11400 }, { "grad_norm": 0.31333839893341064, "learning_rate": 9.767333180511465e-05, "loss": 0.0222, "step": 11410 }, { "grad_norm": 0.24642115831375122, "learning_rate": 9.766709624791373e-05, "loss": 0.0255, "step": 11420 }, { "grad_norm": 0.25071948766708374, "learning_rate": 9.766085254570975e-05, "loss": 0.025, "step": 11430 }, { "grad_norm": 0.3323870897293091, "learning_rate": 9.76546006995696e-05, "loss": 0.0242, "step": 11440 }, { "grad_norm": 0.2865903973579407, "learning_rate": 9.764834071056155e-05, "loss": 0.0199, "step": 11450 }, { "grad_norm": 0.1878250241279602, "learning_rate": 9.764207257975526e-05, "loss": 0.0215, "step": 11460 }, { "grad_norm": 0.19101636111736298, "learning_rate": 9.763579630822179e-05, "loss": 0.0223, "step": 11470 }, { "grad_norm": 0.21792732179164886, "learning_rate": 9.762951189703356e-05, "loss": 0.0218, "step": 11480 }, { "grad_norm": 0.21752913296222687, "learning_rate": 9.762321934726442e-05, "loss": 0.0195, "step": 11490 }, { "grad_norm": 0.26005473732948303, "learning_rate": 9.761691865998959e-05, "loss": 0.0247, "step": 11500 }, { "grad_norm": 0.24519501626491547, "learning_rate": 9.76106098362857e-05, "loss": 0.0212, "step": 11510 }, { "grad_norm": 0.22920238971710205, "learning_rate": 9.760429287723072e-05, "loss": 0.0229, "step": 11520 }, { "grad_norm": 0.2486458420753479, "learning_rate": 9.759796778390406e-05, "loss": 0.0217, "step": 11530 }, { "grad_norm": 0.2646237313747406, "learning_rate": 9.759163455738653e-05, "loss": 0.0232, "step": 11540 }, { "grad_norm": 0.20852546393871307, "learning_rate": 9.75852931987603e-05, "loss": 0.0172, "step": 11550 }, { "grad_norm": 0.19822125136852264, "learning_rate": 9.757894370910891e-05, "loss": 0.0246, "step": 11560 }, { "grad_norm": 0.24397386610507965, "learning_rate": 9.757258608951733e-05, "loss": 0.0237, "step": 11570 }, { "grad_norm": 0.17563265562057495, "learning_rate": 9.75662203410719e-05, "loss": 0.0202, "step": 11580 }, { "grad_norm": 0.28649991750717163, "learning_rate": 9.755984646486034e-05, "loss": 0.0182, "step": 11590 }, { "grad_norm": 0.21265198290348053, "learning_rate": 9.75534644619718e-05, "loss": 0.0168, "step": 11600 }, { "grad_norm": 0.20389725267887115, "learning_rate": 9.754707433349676e-05, "loss": 0.0197, "step": 11610 }, { "grad_norm": 0.2171027958393097, "learning_rate": 9.754067608052715e-05, "loss": 0.0228, "step": 11620 }, { "grad_norm": 0.2345639318227768, "learning_rate": 9.753426970415622e-05, "loss": 0.0183, "step": 11630 }, { "grad_norm": 0.22720754146575928, "learning_rate": 9.752785520547868e-05, "loss": 0.0197, "step": 11640 }, { "grad_norm": 0.23892919719219208, "learning_rate": 9.752143258559056e-05, "loss": 0.0198, "step": 11650 }, { "grad_norm": 0.22184735536575317, "learning_rate": 9.751500184558933e-05, "loss": 0.0206, "step": 11660 }, { "grad_norm": 0.2035127431154251, "learning_rate": 9.750856298657383e-05, "loss": 0.0156, "step": 11670 }, { "grad_norm": 0.17319782078266144, "learning_rate": 9.750211600964428e-05, "loss": 0.0177, "step": 11680 }, { "grad_norm": 0.22965875267982483, "learning_rate": 9.749566091590226e-05, "loss": 0.0243, "step": 11690 }, { "grad_norm": 0.2775254547595978, "learning_rate": 9.748919770645083e-05, "loss": 0.023, "step": 11700 }, { "grad_norm": 0.32182103395462036, "learning_rate": 9.748272638239432e-05, "loss": 0.021, "step": 11710 }, { "grad_norm": 0.2825569212436676, "learning_rate": 9.747624694483855e-05, "loss": 0.0236, "step": 11720 }, { "grad_norm": 0.23078487813472748, "learning_rate": 9.746975939489065e-05, "loss": 0.0183, "step": 11730 }, { "grad_norm": 0.1892591416835785, "learning_rate": 9.746326373365918e-05, "loss": 0.0179, "step": 11740 }, { "grad_norm": 0.22016140818595886, "learning_rate": 9.745675996225403e-05, "loss": 0.0198, "step": 11750 }, { "grad_norm": 0.23057502508163452, "learning_rate": 9.745024808178657e-05, "loss": 0.0204, "step": 11760 }, { "grad_norm": 0.2970694303512573, "learning_rate": 9.744372809336947e-05, "loss": 0.02, "step": 11770 }, { "grad_norm": 0.2607668936252594, "learning_rate": 9.743719999811682e-05, "loss": 0.025, "step": 11780 }, { "grad_norm": 0.23094968497753143, "learning_rate": 9.743066379714412e-05, "loss": 0.0195, "step": 11790 }, { "grad_norm": 0.24990642070770264, "learning_rate": 9.74241194915682e-05, "loss": 0.0256, "step": 11800 }, { "grad_norm": 0.2205488532781601, "learning_rate": 9.741756708250731e-05, "loss": 0.0198, "step": 11810 }, { "grad_norm": 0.235418438911438, "learning_rate": 9.741100657108109e-05, "loss": 0.0244, "step": 11820 }, { "grad_norm": 0.23818162083625793, "learning_rate": 9.740443795841054e-05, "loss": 0.0225, "step": 11830 }, { "grad_norm": 0.2300974279642105, "learning_rate": 9.739786124561805e-05, "loss": 0.0196, "step": 11840 }, { "grad_norm": 0.21710632741451263, "learning_rate": 9.73912764338274e-05, "loss": 0.0219, "step": 11850 }, { "grad_norm": 0.23707321286201477, "learning_rate": 9.738468352416377e-05, "loss": 0.022, "step": 11860 }, { "grad_norm": 0.21918252110481262, "learning_rate": 9.737808251775369e-05, "loss": 0.024, "step": 11870 }, { "grad_norm": 0.22120508551597595, "learning_rate": 9.737147341572512e-05, "loss": 0.0199, "step": 11880 }, { "grad_norm": 0.17044790089130402, "learning_rate": 9.736485621920735e-05, "loss": 0.0199, "step": 11890 }, { "grad_norm": 0.2633146047592163, "learning_rate": 9.735823092933108e-05, "loss": 0.0287, "step": 11900 }, { "grad_norm": 0.2000637948513031, "learning_rate": 9.735159754722838e-05, "loss": 0.0206, "step": 11910 }, { "grad_norm": 0.29666030406951904, "learning_rate": 9.734495607403275e-05, "loss": 0.0207, "step": 11920 }, { "grad_norm": 0.33612027764320374, "learning_rate": 9.733830651087901e-05, "loss": 0.0227, "step": 11930 }, { "grad_norm": 0.2422572374343872, "learning_rate": 9.733164885890338e-05, "loss": 0.0216, "step": 11940 }, { "grad_norm": 0.267545610666275, "learning_rate": 9.732498311924349e-05, "loss": 0.0252, "step": 11950 }, { "grad_norm": 0.23066267371177673, "learning_rate": 9.731830929303833e-05, "loss": 0.0233, "step": 11960 }, { "grad_norm": 0.2665710151195526, "learning_rate": 9.731162738142827e-05, "loss": 0.021, "step": 11970 }, { "grad_norm": 0.22269146144390106, "learning_rate": 9.730493738555506e-05, "loss": 0.0238, "step": 11980 }, { "grad_norm": 0.25476816296577454, "learning_rate": 9.729823930656186e-05, "loss": 0.0203, "step": 11990 }, { "grad_norm": 0.2711181342601776, "learning_rate": 9.729153314559316e-05, "loss": 0.0211, "step": 12000 }, { "grad_norm": 0.22684314846992493, "learning_rate": 9.728481890379486e-05, "loss": 0.0229, "step": 12010 }, { "grad_norm": 0.17920762300491333, "learning_rate": 9.727809658231428e-05, "loss": 0.0207, "step": 12020 }, { "grad_norm": 0.23924830555915833, "learning_rate": 9.727136618230003e-05, "loss": 0.0219, "step": 12030 }, { "grad_norm": 0.27787551283836365, "learning_rate": 9.726462770490219e-05, "loss": 0.024, "step": 12040 }, { "grad_norm": 0.29174670577049255, "learning_rate": 9.725788115127214e-05, "loss": 0.0192, "step": 12050 }, { "grad_norm": 0.23999947309494019, "learning_rate": 9.725112652256274e-05, "loss": 0.0217, "step": 12060 }, { "grad_norm": 0.21006347239017487, "learning_rate": 9.724436381992812e-05, "loss": 0.0213, "step": 12070 }, { "grad_norm": 0.288537859916687, "learning_rate": 9.723759304452387e-05, "loss": 0.0208, "step": 12080 }, { "grad_norm": 0.17738349735736847, "learning_rate": 9.72308141975069e-05, "loss": 0.0184, "step": 12090 }, { "grad_norm": 0.21498005092144012, "learning_rate": 9.722402728003557e-05, "loss": 0.0209, "step": 12100 }, { "grad_norm": 0.17372187972068787, "learning_rate": 9.721723229326953e-05, "loss": 0.0189, "step": 12110 }, { "grad_norm": 0.1978607475757599, "learning_rate": 9.721042923836992e-05, "loss": 0.0214, "step": 12120 }, { "grad_norm": 0.2697259783744812, "learning_rate": 9.720361811649914e-05, "loss": 0.0207, "step": 12130 }, { "grad_norm": 0.19612038135528564, "learning_rate": 9.719679892882106e-05, "loss": 0.0172, "step": 12140 }, { "grad_norm": 0.2236517071723938, "learning_rate": 9.718997167650085e-05, "loss": 0.0201, "step": 12150 }, { "grad_norm": 0.23421171307563782, "learning_rate": 9.718313636070515e-05, "loss": 0.0216, "step": 12160 }, { "grad_norm": 0.29893746972084045, "learning_rate": 9.717629298260192e-05, "loss": 0.0224, "step": 12170 }, { "grad_norm": 0.2393772006034851, "learning_rate": 9.716944154336047e-05, "loss": 0.0191, "step": 12180 }, { "grad_norm": 0.20764672756195068, "learning_rate": 9.716258204415157e-05, "loss": 0.0187, "step": 12190 }, { "grad_norm": 0.3080677390098572, "learning_rate": 9.715571448614728e-05, "loss": 0.0226, "step": 12200 }, { "grad_norm": 0.20813128352165222, "learning_rate": 9.71488388705211e-05, "loss": 0.019, "step": 12210 }, { "grad_norm": 0.2465483397245407, "learning_rate": 9.714195519844788e-05, "loss": 0.0203, "step": 12220 }, { "grad_norm": 0.23032209277153015, "learning_rate": 9.713506347110386e-05, "loss": 0.0198, "step": 12230 }, { "grad_norm": 0.24761758744716644, "learning_rate": 9.712816368966662e-05, "loss": 0.0188, "step": 12240 }, { "grad_norm": 0.22430187463760376, "learning_rate": 9.712125585531517e-05, "loss": 0.0207, "step": 12250 }, { "grad_norm": 0.210874542593956, "learning_rate": 9.711433996922988e-05, "loss": 0.019, "step": 12260 }, { "grad_norm": 0.2291673868894577, "learning_rate": 9.710741603259245e-05, "loss": 0.0239, "step": 12270 }, { "grad_norm": 0.20653052628040314, "learning_rate": 9.710048404658603e-05, "loss": 0.0211, "step": 12280 }, { "grad_norm": 0.22628863155841827, "learning_rate": 9.709354401239508e-05, "loss": 0.0197, "step": 12290 }, { "grad_norm": 0.25819310545921326, "learning_rate": 9.708659593120546e-05, "loss": 0.0206, "step": 12300 }, { "grad_norm": 0.24366986751556396, "learning_rate": 9.707963980420443e-05, "loss": 0.0215, "step": 12310 }, { "grad_norm": 0.22170352935791016, "learning_rate": 9.707267563258058e-05, "loss": 0.0178, "step": 12320 }, { "grad_norm": 0.24680040776729584, "learning_rate": 9.70657034175239e-05, "loss": 0.0196, "step": 12330 }, { "grad_norm": 0.2513481080532074, "learning_rate": 9.705872316022577e-05, "loss": 0.0238, "step": 12340 }, { "grad_norm": 0.20380881428718567, "learning_rate": 9.705173486187891e-05, "loss": 0.0196, "step": 12350 }, { "grad_norm": 0.23518119752407074, "learning_rate": 9.704473852367741e-05, "loss": 0.0263, "step": 12360 }, { "grad_norm": 0.17543476819992065, "learning_rate": 9.70377341468168e-05, "loss": 0.0207, "step": 12370 }, { "grad_norm": 0.2998526990413666, "learning_rate": 9.703072173249389e-05, "loss": 0.0208, "step": 12380 }, { "grad_norm": 0.22180643677711487, "learning_rate": 9.702370128190693e-05, "loss": 0.0193, "step": 12390 }, { "grad_norm": 0.31075674295425415, "learning_rate": 9.701667279625552e-05, "loss": 0.0225, "step": 12400 }, { "grad_norm": 0.24090136587619781, "learning_rate": 9.700963627674065e-05, "loss": 0.0197, "step": 12410 }, { "grad_norm": 0.21964602172374725, "learning_rate": 9.700259172456466e-05, "loss": 0.0195, "step": 12420 }, { "grad_norm": 0.17924994230270386, "learning_rate": 9.699553914093124e-05, "loss": 0.0175, "step": 12430 }, { "grad_norm": 0.21820694208145142, "learning_rate": 9.698847852704553e-05, "loss": 0.0197, "step": 12440 }, { "grad_norm": 0.22381623089313507, "learning_rate": 9.6981409884114e-05, "loss": 0.0286, "step": 12450 }, { "grad_norm": 0.22642408311367035, "learning_rate": 9.697433321334443e-05, "loss": 0.0196, "step": 12460 }, { "grad_norm": 0.27203109860420227, "learning_rate": 9.696724851594607e-05, "loss": 0.024, "step": 12470 }, { "grad_norm": 0.17331796884536743, "learning_rate": 9.696015579312952e-05, "loss": 0.0183, "step": 12480 }, { "grad_norm": 0.1835806965827942, "learning_rate": 9.695305504610668e-05, "loss": 0.0179, "step": 12490 }, { "grad_norm": 0.18036296963691711, "learning_rate": 9.694594627609092e-05, "loss": 0.0183, "step": 12500 }, { "grad_norm": 0.19551387429237366, "learning_rate": 9.693882948429691e-05, "loss": 0.0176, "step": 12510 }, { "grad_norm": 0.24262671172618866, "learning_rate": 9.693170467194071e-05, "loss": 0.0194, "step": 12520 }, { "grad_norm": 0.1960844099521637, "learning_rate": 9.692457184023977e-05, "loss": 0.0208, "step": 12530 }, { "grad_norm": 0.33481740951538086, "learning_rate": 9.691743099041291e-05, "loss": 0.0203, "step": 12540 }, { "grad_norm": 0.3375990390777588, "learning_rate": 9.691028212368027e-05, "loss": 0.0218, "step": 12550 }, { "grad_norm": 0.26271143555641174, "learning_rate": 9.690312524126342e-05, "loss": 0.0188, "step": 12560 }, { "grad_norm": 0.2151772528886795, "learning_rate": 9.689596034438527e-05, "loss": 0.0188, "step": 12570 }, { "grad_norm": 0.27696725726127625, "learning_rate": 9.688878743427012e-05, "loss": 0.0178, "step": 12580 }, { "grad_norm": 0.23556122183799744, "learning_rate": 9.688160651214359e-05, "loss": 0.0179, "step": 12590 }, { "grad_norm": 0.26734521985054016, "learning_rate": 9.687441757923273e-05, "loss": 0.0196, "step": 12600 }, { "grad_norm": 0.2490423172712326, "learning_rate": 9.68672206367659e-05, "loss": 0.0183, "step": 12610 }, { "grad_norm": 0.20568297803401947, "learning_rate": 9.686001568597291e-05, "loss": 0.0204, "step": 12620 }, { "grad_norm": 0.20002958178520203, "learning_rate": 9.685280272808486e-05, "loss": 0.018, "step": 12630 }, { "grad_norm": 0.280409038066864, "learning_rate": 9.684558176433424e-05, "loss": 0.0181, "step": 12640 }, { "grad_norm": 0.21689613163471222, "learning_rate": 9.683835279595495e-05, "loss": 0.0187, "step": 12650 }, { "grad_norm": 0.20565904676914215, "learning_rate": 9.683111582418216e-05, "loss": 0.0153, "step": 12660 }, { "grad_norm": 0.2739284336566925, "learning_rate": 9.682387085025254e-05, "loss": 0.0165, "step": 12670 }, { "grad_norm": 0.2573322355747223, "learning_rate": 9.681661787540401e-05, "loss": 0.0198, "step": 12680 }, { "grad_norm": 0.24332661926746368, "learning_rate": 9.680935690087593e-05, "loss": 0.0209, "step": 12690 }, { "grad_norm": 0.2680313289165497, "learning_rate": 9.680208792790901e-05, "loss": 0.0178, "step": 12700 }, { "grad_norm": 0.2771071791648865, "learning_rate": 9.679481095774529e-05, "loss": 0.0175, "step": 12710 }, { "grad_norm": 0.16593296825885773, "learning_rate": 9.678752599162822e-05, "loss": 0.0191, "step": 12720 }, { "grad_norm": 0.2125425785779953, "learning_rate": 9.678023303080259e-05, "loss": 0.0181, "step": 12730 }, { "grad_norm": 0.2484748810529709, "learning_rate": 9.677293207651459e-05, "loss": 0.0189, "step": 12740 }, { "grad_norm": 0.20254670083522797, "learning_rate": 9.676562313001173e-05, "loss": 0.0187, "step": 12750 }, { "grad_norm": 0.2732097804546356, "learning_rate": 9.675830619254293e-05, "loss": 0.0194, "step": 12760 }, { "grad_norm": 0.3516128659248352, "learning_rate": 9.675098126535843e-05, "loss": 0.0213, "step": 12770 }, { "grad_norm": 0.2795286774635315, "learning_rate": 9.674364834970988e-05, "loss": 0.0223, "step": 12780 }, { "grad_norm": 0.2603814899921417, "learning_rate": 9.673630744685028e-05, "loss": 0.0175, "step": 12790 }, { "grad_norm": 0.20894835889339447, "learning_rate": 9.672895855803397e-05, "loss": 0.0201, "step": 12800 }, { "grad_norm": 0.2150842249393463, "learning_rate": 9.672160168451667e-05, "loss": 0.0164, "step": 12810 }, { "grad_norm": 0.32513001561164856, "learning_rate": 9.671423682755549e-05, "loss": 0.0197, "step": 12820 }, { "grad_norm": 0.19205784797668457, "learning_rate": 9.670686398840888e-05, "loss": 0.0214, "step": 12830 }, { "grad_norm": 0.19088035821914673, "learning_rate": 9.669948316833664e-05, "loss": 0.0195, "step": 12840 }, { "grad_norm": 0.24088677763938904, "learning_rate": 9.669209436859997e-05, "loss": 0.019, "step": 12850 }, { "grad_norm": 0.2862408757209778, "learning_rate": 9.66846975904614e-05, "loss": 0.0232, "step": 12860 }, { "grad_norm": 0.23529212176799774, "learning_rate": 9.667729283518483e-05, "loss": 0.018, "step": 12870 }, { "grad_norm": 0.2128528207540512, "learning_rate": 9.666988010403557e-05, "loss": 0.0186, "step": 12880 }, { "grad_norm": 0.2630966603755951, "learning_rate": 9.66624593982802e-05, "loss": 0.0195, "step": 12890 }, { "grad_norm": 0.26447898149490356, "learning_rate": 9.665503071918675e-05, "loss": 0.0205, "step": 12900 }, { "grad_norm": 0.21905027329921722, "learning_rate": 9.664759406802456e-05, "loss": 0.0172, "step": 12910 }, { "grad_norm": 0.18972203135490417, "learning_rate": 9.664014944606437e-05, "loss": 0.0197, "step": 12920 }, { "grad_norm": 0.24976800382137299, "learning_rate": 9.663269685457822e-05, "loss": 0.0193, "step": 12930 }, { "grad_norm": 0.23582036793231964, "learning_rate": 9.662523629483962e-05, "loss": 0.0205, "step": 12940 }, { "grad_norm": 0.21916361153125763, "learning_rate": 9.661776776812333e-05, "loss": 0.0169, "step": 12950 }, { "grad_norm": 0.22607946395874023, "learning_rate": 9.661029127570553e-05, "loss": 0.0198, "step": 12960 }, { "grad_norm": 0.2463037371635437, "learning_rate": 9.660280681886373e-05, "loss": 0.0185, "step": 12970 }, { "grad_norm": 0.24223634600639343, "learning_rate": 9.659531439887685e-05, "loss": 0.0196, "step": 12980 }, { "grad_norm": 0.26892924308776855, "learning_rate": 9.658781401702511e-05, "loss": 0.0264, "step": 12990 }, { "grad_norm": 0.15391597151756287, "learning_rate": 9.658030567459015e-05, "loss": 0.0185, "step": 13000 }, { "grad_norm": 0.20606476068496704, "learning_rate": 9.65727893728549e-05, "loss": 0.0191, "step": 13010 }, { "grad_norm": 0.25185829401016235, "learning_rate": 9.656526511310375e-05, "loss": 0.0185, "step": 13020 }, { "grad_norm": 0.2240242213010788, "learning_rate": 9.655773289662233e-05, "loss": 0.0182, "step": 13030 }, { "grad_norm": 0.21432587504386902, "learning_rate": 9.655019272469772e-05, "loss": 0.0237, "step": 13040 }, { "grad_norm": 0.22224999964237213, "learning_rate": 9.654264459861832e-05, "loss": 0.0209, "step": 13050 }, { "grad_norm": 0.210889995098114, "learning_rate": 9.653508851967391e-05, "loss": 0.0175, "step": 13060 }, { "grad_norm": 0.2859139144420624, "learning_rate": 9.65275244891556e-05, "loss": 0.0163, "step": 13070 }, { "grad_norm": 0.1917818784713745, "learning_rate": 9.651995250835591e-05, "loss": 0.0161, "step": 13080 }, { "grad_norm": 0.28376421332359314, "learning_rate": 9.651237257856862e-05, "loss": 0.0189, "step": 13090 }, { "grad_norm": 0.24288560450077057, "learning_rate": 9.6504784701089e-05, "loss": 0.0155, "step": 13100 }, { "grad_norm": 0.25409460067749023, "learning_rate": 9.649718887721357e-05, "loss": 0.0205, "step": 13110 }, { "grad_norm": 0.20432789623737335, "learning_rate": 9.648958510824028e-05, "loss": 0.0173, "step": 13120 }, { "grad_norm": 0.21751879155635834, "learning_rate": 9.648197339546837e-05, "loss": 0.0188, "step": 13130 }, { "grad_norm": 0.23161198198795319, "learning_rate": 9.647435374019851e-05, "loss": 0.0197, "step": 13140 }, { "grad_norm": 0.19783392548561096, "learning_rate": 9.646672614373266e-05, "loss": 0.0207, "step": 13150 }, { "grad_norm": 0.2559168338775635, "learning_rate": 9.645909060737418e-05, "loss": 0.0185, "step": 13160 }, { "grad_norm": 0.2880434989929199, "learning_rate": 9.645144713242778e-05, "loss": 0.0215, "step": 13170 }, { "grad_norm": 0.3026072382926941, "learning_rate": 9.64437957201995e-05, "loss": 0.0191, "step": 13180 }, { "grad_norm": 0.2313661128282547, "learning_rate": 9.643613637199678e-05, "loss": 0.0207, "step": 13190 }, { "grad_norm": 0.21623563766479492, "learning_rate": 9.642846908912839e-05, "loss": 0.0197, "step": 13200 }, { "grad_norm": 0.13684117794036865, "learning_rate": 9.642079387290444e-05, "loss": 0.018, "step": 13210 }, { "grad_norm": 0.23206710815429688, "learning_rate": 9.641311072463644e-05, "loss": 0.0201, "step": 13220 }, { "grad_norm": 0.2381032556295395, "learning_rate": 9.640541964563722e-05, "loss": 0.0215, "step": 13230 }, { "grad_norm": 0.23971526324748993, "learning_rate": 9.639772063722096e-05, "loss": 0.022, "step": 13240 }, { "grad_norm": 0.17700313031673431, "learning_rate": 9.639001370070324e-05, "loss": 0.0185, "step": 13250 }, { "grad_norm": 0.24898169934749603, "learning_rate": 9.638229883740095e-05, "loss": 0.0189, "step": 13260 }, { "grad_norm": 0.20926599204540253, "learning_rate": 9.637457604863233e-05, "loss": 0.0162, "step": 13270 }, { "grad_norm": 0.16781547665596008, "learning_rate": 9.636684533571703e-05, "loss": 0.017, "step": 13280 }, { "grad_norm": 0.2254173904657364, "learning_rate": 9.635910669997599e-05, "loss": 0.0151, "step": 13290 }, { "grad_norm": 0.19177494943141937, "learning_rate": 9.635136014273154e-05, "loss": 0.019, "step": 13300 }, { "grad_norm": 0.25063973665237427, "learning_rate": 9.634360566530735e-05, "loss": 0.0196, "step": 13310 }, { "grad_norm": 0.3087405860424042, "learning_rate": 9.633584326902845e-05, "loss": 0.0195, "step": 13320 }, { "grad_norm": 0.24878033995628357, "learning_rate": 9.632807295522124e-05, "loss": 0.0188, "step": 13330 }, { "grad_norm": 0.1917106956243515, "learning_rate": 9.632029472521342e-05, "loss": 0.0186, "step": 13340 }, { "grad_norm": 0.2291882485151291, "learning_rate": 9.631250858033409e-05, "loss": 0.0179, "step": 13350 }, { "grad_norm": 0.19463373720645905, "learning_rate": 9.630471452191371e-05, "loss": 0.0188, "step": 13360 }, { "grad_norm": 0.20099224150180817, "learning_rate": 9.629691255128405e-05, "loss": 0.0186, "step": 13370 }, { "grad_norm": 0.20065951347351074, "learning_rate": 9.628910266977825e-05, "loss": 0.0172, "step": 13380 }, { "grad_norm": 0.2116747796535492, "learning_rate": 9.628128487873083e-05, "loss": 0.0163, "step": 13390 }, { "grad_norm": 0.24116162955760956, "learning_rate": 9.627345917947761e-05, "loss": 0.0186, "step": 13400 }, { "grad_norm": 0.23744642734527588, "learning_rate": 9.626562557335579e-05, "loss": 0.0164, "step": 13410 }, { "grad_norm": 0.1811514049768448, "learning_rate": 9.625778406170393e-05, "loss": 0.0164, "step": 13420 }, { "grad_norm": 0.23056335747241974, "learning_rate": 9.624993464586193e-05, "loss": 0.0188, "step": 13430 }, { "grad_norm": 0.3126709461212158, "learning_rate": 9.624207732717105e-05, "loss": 0.0203, "step": 13440 }, { "grad_norm": 0.17006397247314453, "learning_rate": 9.623421210697386e-05, "loss": 0.0184, "step": 13450 }, { "grad_norm": 0.2020941972732544, "learning_rate": 9.622633898661434e-05, "loss": 0.0199, "step": 13460 }, { "grad_norm": 0.20464475452899933, "learning_rate": 9.621845796743778e-05, "loss": 0.0218, "step": 13470 }, { "grad_norm": 0.24230563640594482, "learning_rate": 9.621056905079082e-05, "loss": 0.019, "step": 13480 }, { "grad_norm": 0.3060840964317322, "learning_rate": 9.620267223802149e-05, "loss": 0.0163, "step": 13490 }, { "grad_norm": 0.1918579339981079, "learning_rate": 9.619476753047911e-05, "loss": 0.0159, "step": 13500 }, { "grad_norm": 0.21348707377910614, "learning_rate": 9.618685492951438e-05, "loss": 0.0182, "step": 13510 }, { "grad_norm": 0.2091357558965683, "learning_rate": 9.617893443647938e-05, "loss": 0.0163, "step": 13520 }, { "grad_norm": 0.21435245871543884, "learning_rate": 9.617100605272746e-05, "loss": 0.0155, "step": 13530 }, { "grad_norm": 0.27174806594848633, "learning_rate": 9.616306977961338e-05, "loss": 0.0163, "step": 13540 }, { "grad_norm": 0.22894567251205444, "learning_rate": 9.615512561849326e-05, "loss": 0.0165, "step": 13550 }, { "grad_norm": 0.2323436141014099, "learning_rate": 9.61471735707245e-05, "loss": 0.023, "step": 13560 }, { "grad_norm": 0.20948708057403564, "learning_rate": 9.613921363766592e-05, "loss": 0.0174, "step": 13570 }, { "grad_norm": 0.1678377389907837, "learning_rate": 9.613124582067763e-05, "loss": 0.0161, "step": 13580 }, { "grad_norm": 0.20279757678508759, "learning_rate": 9.612327012112112e-05, "loss": 0.0183, "step": 13590 }, { "grad_norm": 0.19554418325424194, "learning_rate": 9.611528654035921e-05, "loss": 0.0155, "step": 13600 }, { "grad_norm": 0.23457832634449005, "learning_rate": 9.610729507975611e-05, "loss": 0.0181, "step": 13610 }, { "grad_norm": 0.2188928872346878, "learning_rate": 9.609929574067731e-05, "loss": 0.017, "step": 13620 }, { "grad_norm": 0.20065073668956757, "learning_rate": 9.609128852448967e-05, "loss": 0.0183, "step": 13630 }, { "grad_norm": 0.2069512903690338, "learning_rate": 9.608327343256143e-05, "loss": 0.0179, "step": 13640 }, { "grad_norm": 0.3740241229534149, "learning_rate": 9.607525046626216e-05, "loss": 0.0177, "step": 13650 }, { "grad_norm": 0.2606186866760254, "learning_rate": 9.606721962696272e-05, "loss": 0.0162, "step": 13660 }, { "grad_norm": 0.2417176365852356, "learning_rate": 9.60591809160354e-05, "loss": 0.0183, "step": 13670 }, { "grad_norm": 0.253496915102005, "learning_rate": 9.605113433485378e-05, "loss": 0.0173, "step": 13680 }, { "grad_norm": 0.2482839822769165, "learning_rate": 9.604307988479279e-05, "loss": 0.0187, "step": 13690 }, { "grad_norm": 0.22498297691345215, "learning_rate": 9.603501756722876e-05, "loss": 0.0172, "step": 13700 }, { "grad_norm": 0.22978350520133972, "learning_rate": 9.602694738353927e-05, "loss": 0.0196, "step": 13710 }, { "grad_norm": 0.2805382311344147, "learning_rate": 9.601886933510331e-05, "loss": 0.0206, "step": 13720 }, { "grad_norm": 0.25460708141326904, "learning_rate": 9.60107834233012e-05, "loss": 0.0174, "step": 13730 }, { "grad_norm": 0.2368362843990326, "learning_rate": 9.60026896495146e-05, "loss": 0.0194, "step": 13740 }, { "grad_norm": 0.2298370599746704, "learning_rate": 9.599458801512652e-05, "loss": 0.0153, "step": 13750 }, { "grad_norm": 0.2403576672077179, "learning_rate": 9.598647852152129e-05, "loss": 0.0151, "step": 13760 }, { "grad_norm": 0.1869267374277115, "learning_rate": 9.597836117008462e-05, "loss": 0.0185, "step": 13770 }, { "grad_norm": 0.2569224536418915, "learning_rate": 9.597023596220356e-05, "loss": 0.0205, "step": 13780 }, { "grad_norm": 0.22227337956428528, "learning_rate": 9.596210289926643e-05, "loss": 0.0172, "step": 13790 }, { "grad_norm": 0.23740506172180176, "learning_rate": 9.5953961982663e-05, "loss": 0.0181, "step": 13800 }, { "grad_norm": 0.21488867700099945, "learning_rate": 9.594581321378431e-05, "loss": 0.0192, "step": 13810 }, { "grad_norm": 0.24785883724689484, "learning_rate": 9.593765659402276e-05, "loss": 0.0189, "step": 13820 }, { "grad_norm": 0.16703125834465027, "learning_rate": 9.59294921247721e-05, "loss": 0.0168, "step": 13830 }, { "grad_norm": 0.20406243205070496, "learning_rate": 9.59213198074274e-05, "loss": 0.0159, "step": 13840 }, { "grad_norm": 0.258932501077652, "learning_rate": 9.59131396433851e-05, "loss": 0.0194, "step": 13850 }, { "grad_norm": 0.15906313061714172, "learning_rate": 9.590495163404297e-05, "loss": 0.0201, "step": 13860 }, { "grad_norm": 0.19853387773036957, "learning_rate": 9.589675578080009e-05, "loss": 0.018, "step": 13870 }, { "grad_norm": 0.21654126048088074, "learning_rate": 9.588855208505694e-05, "loss": 0.018, "step": 13880 }, { "grad_norm": 0.17610648274421692, "learning_rate": 9.588034054821529e-05, "loss": 0.0164, "step": 13890 }, { "grad_norm": 0.18230432271957397, "learning_rate": 9.587212117167826e-05, "loss": 0.0193, "step": 13900 }, { "grad_norm": 0.24894928932189941, "learning_rate": 9.586389395685033e-05, "loss": 0.0152, "step": 13910 }, { "grad_norm": 0.25254717469215393, "learning_rate": 9.585565890513733e-05, "loss": 0.0162, "step": 13920 }, { "grad_norm": 0.19976076483726501, "learning_rate": 9.584741601794636e-05, "loss": 0.0165, "step": 13930 }, { "grad_norm": 0.21246333420276642, "learning_rate": 9.58391652966859e-05, "loss": 0.0167, "step": 13940 }, { "grad_norm": 0.2040940672159195, "learning_rate": 9.583090674276583e-05, "loss": 0.0182, "step": 13950 }, { "grad_norm": 0.24562913179397583, "learning_rate": 9.582264035759726e-05, "loss": 0.016, "step": 13960 }, { "grad_norm": 0.25059306621551514, "learning_rate": 9.58143661425927e-05, "loss": 0.018, "step": 13970 }, { "grad_norm": 0.2343519777059555, "learning_rate": 9.580608409916601e-05, "loss": 0.0196, "step": 13980 }, { "grad_norm": 0.17232495546340942, "learning_rate": 9.579779422873233e-05, "loss": 0.0209, "step": 13990 }, { "grad_norm": 0.17937737703323364, "learning_rate": 9.578949653270819e-05, "loss": 0.0226, "step": 14000 }, { "grad_norm": 0.21802952885627747, "learning_rate": 9.578119101251144e-05, "loss": 0.0146, "step": 14010 }, { "grad_norm": 0.20713980495929718, "learning_rate": 9.577287766956127e-05, "loss": 0.0168, "step": 14020 }, { "grad_norm": 0.22563405334949493, "learning_rate": 9.57645565052782e-05, "loss": 0.0171, "step": 14030 }, { "grad_norm": 0.21527451276779175, "learning_rate": 9.575622752108407e-05, "loss": 0.0193, "step": 14040 }, { "grad_norm": 0.20858077704906464, "learning_rate": 9.57478907184021e-05, "loss": 0.0187, "step": 14050 }, { "grad_norm": 0.21709994971752167, "learning_rate": 9.573954609865681e-05, "loss": 0.0178, "step": 14060 }, { "grad_norm": 0.19478358328342438, "learning_rate": 9.573119366327408e-05, "loss": 0.0191, "step": 14070 }, { "grad_norm": 0.22116144001483917, "learning_rate": 9.57228334136811e-05, "loss": 0.0172, "step": 14080 }, { "grad_norm": 0.2022261768579483, "learning_rate": 9.571446535130641e-05, "loss": 0.0189, "step": 14090 }, { "grad_norm": 0.14604324102401733, "learning_rate": 9.570608947757988e-05, "loss": 0.0183, "step": 14100 }, { "grad_norm": 0.24647250771522522, "learning_rate": 9.569770579393274e-05, "loss": 0.0171, "step": 14110 }, { "grad_norm": 0.18320399522781372, "learning_rate": 9.56893143017975e-05, "loss": 0.018, "step": 14120 }, { "grad_norm": 0.25963085889816284, "learning_rate": 9.568091500260806e-05, "loss": 0.0171, "step": 14130 }, { "grad_norm": 0.1839406043291092, "learning_rate": 9.567250789779961e-05, "loss": 0.017, "step": 14140 }, { "grad_norm": 0.21880608797073364, "learning_rate": 9.566409298880872e-05, "loss": 0.0171, "step": 14150 }, { "grad_norm": 0.2456142008304596, "learning_rate": 9.565567027707326e-05, "loss": 0.0197, "step": 14160 }, { "grad_norm": 0.23191657662391663, "learning_rate": 9.56472397640324e-05, "loss": 0.0185, "step": 14170 }, { "grad_norm": 0.2445622980594635, "learning_rate": 9.563880145112675e-05, "loss": 0.0223, "step": 14180 }, { "grad_norm": 0.26126089692115784, "learning_rate": 9.563035533979814e-05, "loss": 0.0188, "step": 14190 }, { "grad_norm": 0.24051155149936676, "learning_rate": 9.562190143148981e-05, "loss": 0.0211, "step": 14200 }, { "grad_norm": 0.201939657330513, "learning_rate": 9.561343972764627e-05, "loss": 0.0138, "step": 14210 }, { "grad_norm": 0.24526570737361908, "learning_rate": 9.560497022971343e-05, "loss": 0.0169, "step": 14220 }, { "grad_norm": 0.22548547387123108, "learning_rate": 9.559649293913847e-05, "loss": 0.0169, "step": 14230 }, { "grad_norm": 0.3159067928791046, "learning_rate": 9.558800785736993e-05, "loss": 0.0201, "step": 14240 }, { "grad_norm": 0.2571752071380615, "learning_rate": 9.557951498585767e-05, "loss": 0.02, "step": 14250 }, { "grad_norm": 0.21058785915374756, "learning_rate": 9.557101432605293e-05, "loss": 0.0171, "step": 14260 }, { "grad_norm": 0.2405351996421814, "learning_rate": 9.556250587940818e-05, "loss": 0.0185, "step": 14270 }, { "grad_norm": 0.21665196120738983, "learning_rate": 9.555398964737734e-05, "loss": 0.018, "step": 14280 }, { "grad_norm": 0.18496540188789368, "learning_rate": 9.554546563141555e-05, "loss": 0.0185, "step": 14290 }, { "grad_norm": 0.2140117734670639, "learning_rate": 9.553693383297937e-05, "loss": 0.0215, "step": 14300 }, { "grad_norm": 0.2483241856098175, "learning_rate": 9.552839425352663e-05, "loss": 0.0168, "step": 14310 }, { "grad_norm": 0.28488874435424805, "learning_rate": 9.551984689451652e-05, "loss": 0.0183, "step": 14320 }, { "grad_norm": 0.23417390882968903, "learning_rate": 9.551129175740953e-05, "loss": 0.0178, "step": 14330 }, { "grad_norm": 0.1885925978422165, "learning_rate": 9.550272884366754e-05, "loss": 0.0167, "step": 14340 }, { "grad_norm": 0.2013378143310547, "learning_rate": 9.549415815475369e-05, "loss": 0.0182, "step": 14350 }, { "grad_norm": 0.1656717211008072, "learning_rate": 9.548557969213247e-05, "loss": 0.0173, "step": 14360 }, { "grad_norm": 0.22734372317790985, "learning_rate": 9.547699345726972e-05, "loss": 0.0153, "step": 14370 }, { "grad_norm": 0.191575825214386, "learning_rate": 9.546839945163257e-05, "loss": 0.016, "step": 14380 }, { "grad_norm": 0.21763455867767334, "learning_rate": 9.545979767668953e-05, "loss": 0.0187, "step": 14390 }, { "grad_norm": 0.16965730488300323, "learning_rate": 9.54511881339104e-05, "loss": 0.0181, "step": 14400 }, { "grad_norm": 0.18583247065544128, "learning_rate": 9.54425708247663e-05, "loss": 0.0189, "step": 14410 }, { "grad_norm": 0.15710724890232086, "learning_rate": 9.543394575072972e-05, "loss": 0.0161, "step": 14420 }, { "grad_norm": 0.24522897601127625, "learning_rate": 9.542531291327441e-05, "loss": 0.0158, "step": 14430 }, { "grad_norm": 0.24453626573085785, "learning_rate": 9.541667231387552e-05, "loss": 0.0158, "step": 14440 }, { "grad_norm": 0.2503807246685028, "learning_rate": 9.540802395400949e-05, "loss": 0.0174, "step": 14450 }, { "grad_norm": 0.1970362514257431, "learning_rate": 9.539936783515406e-05, "loss": 0.0172, "step": 14460 }, { "grad_norm": 0.2582394778728485, "learning_rate": 9.539070395878835e-05, "loss": 0.019, "step": 14470 }, { "grad_norm": 0.2365778535604477, "learning_rate": 9.538203232639277e-05, "loss": 0.0156, "step": 14480 }, { "grad_norm": 0.2153443545103073, "learning_rate": 9.537335293944907e-05, "loss": 0.0178, "step": 14490 }, { "grad_norm": 0.2510625422000885, "learning_rate": 9.536466579944032e-05, "loss": 0.0166, "step": 14500 }, { "grad_norm": 0.22122831642627716, "learning_rate": 9.535597090785091e-05, "loss": 0.0187, "step": 14510 }, { "grad_norm": 0.18174763023853302, "learning_rate": 9.534726826616656e-05, "loss": 0.016, "step": 14520 }, { "grad_norm": 0.20871084928512573, "learning_rate": 9.53385578758743e-05, "loss": 0.0158, "step": 14530 }, { "grad_norm": 0.20615345239639282, "learning_rate": 9.532983973846252e-05, "loss": 0.0195, "step": 14540 }, { "grad_norm": 0.15742690861225128, "learning_rate": 9.53211138554209e-05, "loss": 0.0151, "step": 14550 }, { "grad_norm": 0.22752836346626282, "learning_rate": 9.531238022824047e-05, "loss": 0.0165, "step": 14560 }, { "grad_norm": 0.2404741495847702, "learning_rate": 9.530363885841355e-05, "loss": 0.0196, "step": 14570 }, { "grad_norm": 0.2023254632949829, "learning_rate": 9.52948897474338e-05, "loss": 0.0179, "step": 14580 }, { "grad_norm": 0.2327239215373993, "learning_rate": 9.528613289679622e-05, "loss": 0.0164, "step": 14590 }, { "grad_norm": 0.1509394645690918, "learning_rate": 9.52773683079971e-05, "loss": 0.0161, "step": 14600 }, { "grad_norm": 0.2309248298406601, "learning_rate": 9.526859598253407e-05, "loss": 0.018, "step": 14610 }, { "grad_norm": 0.2104867398738861, "learning_rate": 9.525981592190609e-05, "loss": 0.0194, "step": 14620 }, { "grad_norm": 0.1986987292766571, "learning_rate": 9.525102812761342e-05, "loss": 0.0182, "step": 14630 }, { "grad_norm": 0.18607953190803528, "learning_rate": 9.524223260115768e-05, "loss": 0.0174, "step": 14640 }, { "grad_norm": 0.21233713626861572, "learning_rate": 9.523342934404175e-05, "loss": 0.0153, "step": 14650 }, { "grad_norm": 0.31393295526504517, "learning_rate": 9.522461835776989e-05, "loss": 0.0216, "step": 14660 }, { "grad_norm": 0.2654819190502167, "learning_rate": 9.521579964384764e-05, "loss": 0.0162, "step": 14670 }, { "grad_norm": 0.24641306698322296, "learning_rate": 9.52069732037819e-05, "loss": 0.0186, "step": 14680 }, { "grad_norm": 0.25661733746528625, "learning_rate": 9.519813903908083e-05, "loss": 0.0171, "step": 14690 }, { "grad_norm": 0.38516518473625183, "learning_rate": 9.5189297151254e-05, "loss": 0.0176, "step": 14700 }, { "grad_norm": 0.2200002521276474, "learning_rate": 9.518044754181218e-05, "loss": 0.0151, "step": 14710 }, { "grad_norm": 0.2029087245464325, "learning_rate": 9.51715902122676e-05, "loss": 0.0183, "step": 14720 }, { "grad_norm": 0.1754121035337448, "learning_rate": 9.516272516413368e-05, "loss": 0.0179, "step": 14730 }, { "grad_norm": 0.18747101724147797, "learning_rate": 9.515385239892525e-05, "loss": 0.0197, "step": 14740 }, { "grad_norm": 0.17998608946800232, "learning_rate": 9.514497191815839e-05, "loss": 0.0183, "step": 14750 }, { "grad_norm": 0.19577737152576447, "learning_rate": 9.513608372335055e-05, "loss": 0.016, "step": 14760 }, { "grad_norm": 0.15405043959617615, "learning_rate": 9.512718781602045e-05, "loss": 0.0178, "step": 14770 }, { "grad_norm": 0.2377178966999054, "learning_rate": 9.511828419768823e-05, "loss": 0.016, "step": 14780 }, { "grad_norm": 0.23860648274421692, "learning_rate": 9.510937286987521e-05, "loss": 0.0184, "step": 14790 }, { "grad_norm": 0.1798333078622818, "learning_rate": 9.510045383410408e-05, "loss": 0.017, "step": 14800 }, { "grad_norm": 0.23698052763938904, "learning_rate": 9.509152709189892e-05, "loss": 0.0169, "step": 14810 }, { "grad_norm": 0.22944682836532593, "learning_rate": 9.508259264478504e-05, "loss": 0.0193, "step": 14820 }, { "grad_norm": 0.26514366269111633, "learning_rate": 9.507365049428909e-05, "loss": 0.0185, "step": 14830 }, { "grad_norm": 0.2098889946937561, "learning_rate": 9.506470064193902e-05, "loss": 0.0193, "step": 14840 }, { "grad_norm": 0.21956516802310944, "learning_rate": 9.505574308926414e-05, "loss": 0.0166, "step": 14850 }, { "grad_norm": 0.19320078194141388, "learning_rate": 9.504677783779505e-05, "loss": 0.0138, "step": 14860 }, { "grad_norm": 0.19913630187511444, "learning_rate": 9.503780488906365e-05, "loss": 0.0146, "step": 14870 }, { "grad_norm": 0.20805475115776062, "learning_rate": 9.502882424460319e-05, "loss": 0.0175, "step": 14880 }, { "grad_norm": 0.18778297305107117, "learning_rate": 9.501983590594821e-05, "loss": 0.0188, "step": 14890 }, { "grad_norm": 0.1816745102405548, "learning_rate": 9.501083987463455e-05, "loss": 0.0164, "step": 14900 }, { "grad_norm": 0.18693998456001282, "learning_rate": 9.500183615219942e-05, "loss": 0.0157, "step": 14910 }, { "grad_norm": 0.20558983087539673, "learning_rate": 9.49928247401813e-05, "loss": 0.0199, "step": 14920 }, { "grad_norm": 0.23979400098323822, "learning_rate": 9.498380564011997e-05, "loss": 0.0187, "step": 14930 }, { "grad_norm": 0.26915064454078674, "learning_rate": 9.497477885355656e-05, "loss": 0.0197, "step": 14940 }, { "grad_norm": 0.23560000956058502, "learning_rate": 9.496574438203353e-05, "loss": 0.0157, "step": 14950 }, { "grad_norm": 0.2653893232345581, "learning_rate": 9.495670222709459e-05, "loss": 0.0153, "step": 14960 }, { "grad_norm": 0.16116546094417572, "learning_rate": 9.494765239028483e-05, "loss": 0.0204, "step": 14970 }, { "grad_norm": 0.20141169428825378, "learning_rate": 9.493859487315057e-05, "loss": 0.0168, "step": 14980 }, { "grad_norm": 0.24919697642326355, "learning_rate": 9.492952967723953e-05, "loss": 0.0147, "step": 14990 }, { "grad_norm": 0.29414990544319153, "learning_rate": 9.492045680410068e-05, "loss": 0.0177, "step": 15000 }, { "grad_norm": 0.15417231619358063, "learning_rate": 9.491137625528436e-05, "loss": 0.016, "step": 15010 }, { "grad_norm": 0.26690343022346497, "learning_rate": 9.490228803234215e-05, "loss": 0.0172, "step": 15020 }, { "grad_norm": 0.25168576836586, "learning_rate": 9.489319213682701e-05, "loss": 0.0151, "step": 15030 }, { "grad_norm": 0.2283739447593689, "learning_rate": 9.488408857029316e-05, "loss": 0.0133, "step": 15040 }, { "grad_norm": 0.2379574030637741, "learning_rate": 9.487497733429616e-05, "loss": 0.0155, "step": 15050 }, { "grad_norm": 0.2955033481121063, "learning_rate": 9.486585843039286e-05, "loss": 0.0146, "step": 15060 }, { "grad_norm": 0.1975097954273224, "learning_rate": 9.485673186014143e-05, "loss": 0.0147, "step": 15070 }, { "grad_norm": 0.23816418647766113, "learning_rate": 9.484759762510137e-05, "loss": 0.0162, "step": 15080 }, { "grad_norm": 0.1947396695613861, "learning_rate": 9.483845572683346e-05, "loss": 0.0161, "step": 15090 }, { "grad_norm": 0.19376258552074432, "learning_rate": 9.48293061668998e-05, "loss": 0.0154, "step": 15100 }, { "grad_norm": 0.19446057081222534, "learning_rate": 9.48201489468638e-05, "loss": 0.0175, "step": 15110 }, { "grad_norm": 0.21651515364646912, "learning_rate": 9.481098406829016e-05, "loss": 0.0166, "step": 15120 }, { "grad_norm": 0.22557683289051056, "learning_rate": 9.480181153274495e-05, "loss": 0.0174, "step": 15130 }, { "grad_norm": 0.22109054028987885, "learning_rate": 9.479263134179548e-05, "loss": 0.0207, "step": 15140 }, { "grad_norm": 0.1850217580795288, "learning_rate": 9.478344349701039e-05, "loss": 0.0217, "step": 15150 }, { "grad_norm": 0.24423737823963165, "learning_rate": 9.477424799995964e-05, "loss": 0.0184, "step": 15160 }, { "grad_norm": 0.222717747092247, "learning_rate": 9.476504485221448e-05, "loss": 0.0184, "step": 15170 }, { "grad_norm": 0.23475567996501923, "learning_rate": 9.475583405534748e-05, "loss": 0.0171, "step": 15180 }, { "grad_norm": 0.2309248447418213, "learning_rate": 9.474661561093251e-05, "loss": 0.02, "step": 15190 }, { "grad_norm": 0.2041691392660141, "learning_rate": 9.473738952054478e-05, "loss": 0.0176, "step": 15200 }, { "grad_norm": 0.1947823166847229, "learning_rate": 9.472815578576073e-05, "loss": 0.018, "step": 15210 }, { "grad_norm": 0.27836930751800537, "learning_rate": 9.471891440815817e-05, "loss": 0.0172, "step": 15220 }, { "grad_norm": 0.14768050611019135, "learning_rate": 9.470966538931621e-05, "loss": 0.0159, "step": 15230 }, { "grad_norm": 0.21622733771800995, "learning_rate": 9.470040873081525e-05, "loss": 0.0159, "step": 15240 }, { "grad_norm": 0.23631992936134338, "learning_rate": 9.469114443423698e-05, "loss": 0.0162, "step": 15250 }, { "grad_norm": 0.164640411734581, "learning_rate": 9.468187250116445e-05, "loss": 0.0169, "step": 15260 }, { "grad_norm": 0.1690230369567871, "learning_rate": 9.467259293318197e-05, "loss": 0.0153, "step": 15270 }, { "grad_norm": 0.22955118119716644, "learning_rate": 9.466330573187514e-05, "loss": 0.0176, "step": 15280 }, { "grad_norm": 0.2149248868227005, "learning_rate": 9.46540108988309e-05, "loss": 0.0156, "step": 15290 }, { "grad_norm": 0.27198362350463867, "learning_rate": 9.46447084356375e-05, "loss": 0.0155, "step": 15300 }, { "grad_norm": 0.22377045452594757, "learning_rate": 9.463539834388447e-05, "loss": 0.0152, "step": 15310 }, { "grad_norm": 0.21917124092578888, "learning_rate": 9.462608062516263e-05, "loss": 0.0176, "step": 15320 }, { "grad_norm": 0.21238091588020325, "learning_rate": 9.461675528106413e-05, "loss": 0.0151, "step": 15330 }, { "grad_norm": 0.34570056200027466, "learning_rate": 9.460742231318244e-05, "loss": 0.0186, "step": 15340 }, { "grad_norm": 0.23594196140766144, "learning_rate": 9.459808172311229e-05, "loss": 0.0186, "step": 15350 }, { "grad_norm": 0.22879670560359955, "learning_rate": 9.458873351244972e-05, "loss": 0.017, "step": 15360 }, { "grad_norm": 0.18693485856056213, "learning_rate": 9.457937768279211e-05, "loss": 0.0197, "step": 15370 }, { "grad_norm": 0.2433399111032486, "learning_rate": 9.45700142357381e-05, "loss": 0.0184, "step": 15380 }, { "grad_norm": 0.24638678133487701, "learning_rate": 9.456064317288765e-05, "loss": 0.0178, "step": 15390 }, { "grad_norm": 0.18670199811458588, "learning_rate": 9.455126449584201e-05, "loss": 0.0194, "step": 15400 }, { "grad_norm": 0.15768824517726898, "learning_rate": 9.454187820620375e-05, "loss": 0.0178, "step": 15410 }, { "grad_norm": 0.19697265326976776, "learning_rate": 9.453248430557673e-05, "loss": 0.0176, "step": 15420 }, { "grad_norm": 0.2408006191253662, "learning_rate": 9.452308279556611e-05, "loss": 0.019, "step": 15430 }, { "grad_norm": 0.2068658024072647, "learning_rate": 9.451367367777835e-05, "loss": 0.0202, "step": 15440 }, { "grad_norm": 0.2655520737171173, "learning_rate": 9.450425695382122e-05, "loss": 0.0167, "step": 15450 }, { "grad_norm": 0.18300262093544006, "learning_rate": 9.449483262530375e-05, "loss": 0.017, "step": 15460 }, { "grad_norm": 0.18935175240039825, "learning_rate": 9.448540069383633e-05, "loss": 0.0193, "step": 15470 }, { "grad_norm": 0.25726476311683655, "learning_rate": 9.447596116103061e-05, "loss": 0.019, "step": 15480 }, { "grad_norm": 0.22584331035614014, "learning_rate": 9.446651402849955e-05, "loss": 0.0241, "step": 15490 }, { "grad_norm": 0.2612643241882324, "learning_rate": 9.44570592978574e-05, "loss": 0.0206, "step": 15500 }, { "grad_norm": 0.24943320453166962, "learning_rate": 9.444759697071972e-05, "loss": 0.0187, "step": 15510 }, { "grad_norm": 0.2502191662788391, "learning_rate": 9.443812704870336e-05, "loss": 0.0195, "step": 15520 }, { "grad_norm": 0.19507087767124176, "learning_rate": 9.442864953342649e-05, "loss": 0.0175, "step": 15530 }, { "grad_norm": 0.21571223437786102, "learning_rate": 9.441916442650852e-05, "loss": 0.0162, "step": 15540 }, { "grad_norm": 0.21772539615631104, "learning_rate": 9.440967172957023e-05, "loss": 0.0175, "step": 15550 }, { "grad_norm": 0.2221076935529709, "learning_rate": 9.440017144423364e-05, "loss": 0.0156, "step": 15560 }, { "grad_norm": 0.21604934334754944, "learning_rate": 9.439066357212209e-05, "loss": 0.0156, "step": 15570 }, { "grad_norm": 0.2379390150308609, "learning_rate": 9.438114811486022e-05, "loss": 0.0159, "step": 15580 }, { "grad_norm": 0.18432138860225677, "learning_rate": 9.4371625074074e-05, "loss": 0.0191, "step": 15590 }, { "grad_norm": 0.2799726128578186, "learning_rate": 9.436209445139059e-05, "loss": 0.0167, "step": 15600 }, { "grad_norm": 0.19271481037139893, "learning_rate": 9.435255624843855e-05, "loss": 0.0175, "step": 15610 }, { "grad_norm": 0.16770713031291962, "learning_rate": 9.43430104668477e-05, "loss": 0.0151, "step": 15620 }, { "grad_norm": 0.18282826244831085, "learning_rate": 9.433345710824914e-05, "loss": 0.0166, "step": 15630 }, { "grad_norm": 0.23306399583816528, "learning_rate": 9.432389617427529e-05, "loss": 0.0182, "step": 15640 }, { "grad_norm": 0.18111756443977356, "learning_rate": 9.431432766655984e-05, "loss": 0.0169, "step": 15650 }, { "grad_norm": 0.31020814180374146, "learning_rate": 9.430475158673778e-05, "loss": 0.0183, "step": 15660 }, { "grad_norm": 0.18446139991283417, "learning_rate": 9.429516793644542e-05, "loss": 0.0177, "step": 15670 }, { "grad_norm": 0.2172493040561676, "learning_rate": 9.428557671732034e-05, "loss": 0.0185, "step": 15680 }, { "grad_norm": 0.18219824135303497, "learning_rate": 9.42759779310014e-05, "loss": 0.0152, "step": 15690 }, { "grad_norm": 0.21948067843914032, "learning_rate": 9.426637157912879e-05, "loss": 0.0154, "step": 15700 }, { "grad_norm": 0.3066687285900116, "learning_rate": 9.425675766334397e-05, "loss": 0.0172, "step": 15710 }, { "grad_norm": 0.2403554618358612, "learning_rate": 9.424713618528968e-05, "loss": 0.0172, "step": 15720 }, { "grad_norm": 0.2690398693084717, "learning_rate": 9.423750714661e-05, "loss": 0.0187, "step": 15730 }, { "grad_norm": 0.1920660436153412, "learning_rate": 9.422787054895022e-05, "loss": 0.0167, "step": 15740 }, { "grad_norm": 0.17829719185829163, "learning_rate": 9.4218226393957e-05, "loss": 0.0179, "step": 15750 }, { "grad_norm": 0.21343407034873962, "learning_rate": 9.420857468327828e-05, "loss": 0.0193, "step": 15760 }, { "grad_norm": 0.23882602155208588, "learning_rate": 9.419891541856323e-05, "loss": 0.0182, "step": 15770 }, { "grad_norm": 0.25232791900634766, "learning_rate": 9.41892486014624e-05, "loss": 0.0166, "step": 15780 }, { "grad_norm": 0.2422620803117752, "learning_rate": 9.417957423362756e-05, "loss": 0.0166, "step": 15790 }, { "grad_norm": 0.21191775798797607, "learning_rate": 9.416989231671178e-05, "loss": 0.0165, "step": 15800 }, { "grad_norm": 0.21113531291484833, "learning_rate": 9.416020285236946e-05, "loss": 0.0144, "step": 15810 }, { "grad_norm": 0.18866783380508423, "learning_rate": 9.415050584225626e-05, "loss": 0.0191, "step": 15820 }, { "grad_norm": 0.24091237783432007, "learning_rate": 9.414080128802914e-05, "loss": 0.0142, "step": 15830 }, { "grad_norm": 0.19407381117343903, "learning_rate": 9.413108919134632e-05, "loss": 0.0193, "step": 15840 }, { "grad_norm": 0.29919642210006714, "learning_rate": 9.412136955386734e-05, "loss": 0.0157, "step": 15850 }, { "grad_norm": 0.16948746144771576, "learning_rate": 9.411164237725303e-05, "loss": 0.0175, "step": 15860 }, { "grad_norm": 0.17628444731235504, "learning_rate": 9.41019076631655e-05, "loss": 0.0152, "step": 15870 }, { "grad_norm": 0.22076500952243805, "learning_rate": 9.409216541326815e-05, "loss": 0.0172, "step": 15880 }, { "grad_norm": 0.14973223209381104, "learning_rate": 9.408241562922564e-05, "loss": 0.0183, "step": 15890 }, { "grad_norm": 0.1934037059545517, "learning_rate": 9.407265831270395e-05, "loss": 0.02, "step": 15900 }, { "grad_norm": 0.23807282745838165, "learning_rate": 9.406289346537035e-05, "loss": 0.0194, "step": 15910 }, { "grad_norm": 0.24702884256839752, "learning_rate": 9.405312108889339e-05, "loss": 0.0227, "step": 15920 }, { "grad_norm": 0.23320326209068298, "learning_rate": 9.404334118494288e-05, "loss": 0.0159, "step": 15930 }, { "grad_norm": 0.2581568956375122, "learning_rate": 9.403355375518995e-05, "loss": 0.0166, "step": 15940 }, { "grad_norm": 0.17227257788181305, "learning_rate": 9.4023758801307e-05, "loss": 0.0186, "step": 15950 }, { "grad_norm": 0.15710128843784332, "learning_rate": 9.401395632496774e-05, "loss": 0.0166, "step": 15960 }, { "grad_norm": 0.22995789349079132, "learning_rate": 9.400414632784711e-05, "loss": 0.0161, "step": 15970 }, { "grad_norm": 0.22753983736038208, "learning_rate": 9.39943288116214e-05, "loss": 0.0166, "step": 15980 }, { "grad_norm": 0.1919258087873459, "learning_rate": 9.398450377796815e-05, "loss": 0.0166, "step": 15990 }, { "grad_norm": 0.21068371832370758, "learning_rate": 9.397467122856616e-05, "loss": 0.0195, "step": 16000 }, { "grad_norm": 0.22214944660663605, "learning_rate": 9.396483116509558e-05, "loss": 0.0177, "step": 16010 }, { "grad_norm": 0.16422733664512634, "learning_rate": 9.39549835892378e-05, "loss": 0.0179, "step": 16020 }, { "grad_norm": 0.22761031985282898, "learning_rate": 9.39451285026755e-05, "loss": 0.0156, "step": 16030 }, { "grad_norm": 0.2172476202249527, "learning_rate": 9.393526590709262e-05, "loss": 0.0158, "step": 16040 }, { "grad_norm": 0.17333579063415527, "learning_rate": 9.392539580417444e-05, "loss": 0.0159, "step": 16050 }, { "grad_norm": 0.2561842203140259, "learning_rate": 9.391551819560747e-05, "loss": 0.017, "step": 16060 }, { "grad_norm": 0.192958801984787, "learning_rate": 9.390563308307955e-05, "loss": 0.0137, "step": 16070 }, { "grad_norm": 0.19599869847297668, "learning_rate": 9.389574046827974e-05, "loss": 0.0161, "step": 16080 }, { "grad_norm": 0.2526998817920685, "learning_rate": 9.388584035289845e-05, "loss": 0.0163, "step": 16090 }, { "grad_norm": 0.25794917345046997, "learning_rate": 9.387593273862732e-05, "loss": 0.0159, "step": 16100 }, { "grad_norm": 0.24137233197689056, "learning_rate": 9.386601762715929e-05, "loss": 0.0152, "step": 16110 }, { "grad_norm": 0.24449370801448822, "learning_rate": 9.38560950201886e-05, "loss": 0.0155, "step": 16120 }, { "grad_norm": 0.18849647045135498, "learning_rate": 9.384616491941071e-05, "loss": 0.0185, "step": 16130 }, { "grad_norm": 0.23193302750587463, "learning_rate": 9.383622732652245e-05, "loss": 0.0176, "step": 16140 }, { "grad_norm": 0.2348398119211197, "learning_rate": 9.382628224322187e-05, "loss": 0.0166, "step": 16150 }, { "grad_norm": 0.24158364534378052, "learning_rate": 9.381632967120829e-05, "loss": 0.017, "step": 16160 }, { "grad_norm": 0.19573022425174713, "learning_rate": 9.380636961218235e-05, "loss": 0.0168, "step": 16170 }, { "grad_norm": 0.17404359579086304, "learning_rate": 9.379640206784597e-05, "loss": 0.0173, "step": 16180 }, { "grad_norm": 0.2127157598733902, "learning_rate": 9.378642703990229e-05, "loss": 0.0153, "step": 16190 }, { "grad_norm": 0.18619701266288757, "learning_rate": 9.37764445300558e-05, "loss": 0.0154, "step": 16200 }, { "grad_norm": 0.25574326515197754, "learning_rate": 9.376645454001222e-05, "loss": 0.0163, "step": 16210 }, { "grad_norm": 0.18335339426994324, "learning_rate": 9.375645707147858e-05, "loss": 0.0178, "step": 16220 }, { "grad_norm": 0.16021886467933655, "learning_rate": 9.374645212616316e-05, "loss": 0.0158, "step": 16230 }, { "grad_norm": 0.2502102851867676, "learning_rate": 9.373643970577555e-05, "loss": 0.0172, "step": 16240 }, { "grad_norm": 0.2165716290473938, "learning_rate": 9.372641981202659e-05, "loss": 0.0138, "step": 16250 }, { "grad_norm": 0.22344736754894257, "learning_rate": 9.37163924466284e-05, "loss": 0.019, "step": 16260 }, { "grad_norm": 0.16376079618930817, "learning_rate": 9.370635761129438e-05, "loss": 0.0162, "step": 16270 }, { "grad_norm": 0.21762481331825256, "learning_rate": 9.36963153077392e-05, "loss": 0.0199, "step": 16280 }, { "grad_norm": 0.22375187277793884, "learning_rate": 9.368626553767888e-05, "loss": 0.014, "step": 16290 }, { "grad_norm": 0.20727375149726868, "learning_rate": 9.367620830283057e-05, "loss": 0.0171, "step": 16300 }, { "grad_norm": 0.14850696921348572, "learning_rate": 9.366614360491281e-05, "loss": 0.0166, "step": 16310 }, { "grad_norm": 0.17698469758033752, "learning_rate": 9.365607144564539e-05, "loss": 0.0166, "step": 16320 }, { "grad_norm": 0.1756291538476944, "learning_rate": 9.364599182674934e-05, "loss": 0.0142, "step": 16330 }, { "grad_norm": 0.23946936428546906, "learning_rate": 9.3635904749947e-05, "loss": 0.0148, "step": 16340 }, { "grad_norm": 0.25852733850479126, "learning_rate": 9.362581021696202e-05, "loss": 0.0173, "step": 16350 }, { "grad_norm": 0.22865800559520721, "learning_rate": 9.361570822951921e-05, "loss": 0.0193, "step": 16360 }, { "grad_norm": 0.19270019233226776, "learning_rate": 9.360559878934476e-05, "loss": 0.0144, "step": 16370 }, { "grad_norm": 0.22924582660198212, "learning_rate": 9.359548189816611e-05, "loss": 0.0137, "step": 16380 }, { "grad_norm": 0.22820276021957397, "learning_rate": 9.358535755771193e-05, "loss": 0.0154, "step": 16390 }, { "grad_norm": 0.16451047360897064, "learning_rate": 9.357522576971221e-05, "loss": 0.0183, "step": 16400 }, { "grad_norm": 0.23957796394824982, "learning_rate": 9.356508653589819e-05, "loss": 0.0166, "step": 16410 }, { "grad_norm": 0.20600295066833496, "learning_rate": 9.355493985800237e-05, "loss": 0.0162, "step": 16420 }, { "grad_norm": 0.26462307572364807, "learning_rate": 9.354478573775857e-05, "loss": 0.0141, "step": 16430 }, { "grad_norm": 0.220207080245018, "learning_rate": 9.353462417690186e-05, "loss": 0.0147, "step": 16440 }, { "grad_norm": 0.17281238734722137, "learning_rate": 9.352445517716853e-05, "loss": 0.0178, "step": 16450 }, { "grad_norm": 0.23489247262477875, "learning_rate": 9.351427874029621e-05, "loss": 0.0181, "step": 16460 }, { "grad_norm": 0.17050617933273315, "learning_rate": 9.350409486802379e-05, "loss": 0.0152, "step": 16470 }, { "grad_norm": 0.22234457731246948, "learning_rate": 9.349390356209138e-05, "loss": 0.0163, "step": 16480 }, { "grad_norm": 0.1645386666059494, "learning_rate": 9.348370482424042e-05, "loss": 0.0142, "step": 16490 }, { "grad_norm": 0.18477633595466614, "learning_rate": 9.347349865621357e-05, "loss": 0.0141, "step": 16500 }, { "grad_norm": 0.21365520358085632, "learning_rate": 9.346328505975481e-05, "loss": 0.0127, "step": 16510 }, { "grad_norm": 0.19369599223136902, "learning_rate": 9.345306403660936e-05, "loss": 0.0162, "step": 16520 }, { "grad_norm": 0.21945422887802124, "learning_rate": 9.344283558852371e-05, "loss": 0.0168, "step": 16530 }, { "grad_norm": 0.2117282748222351, "learning_rate": 9.343259971724563e-05, "loss": 0.0146, "step": 16540 }, { "grad_norm": 0.17919808626174927, "learning_rate": 9.342235642452413e-05, "loss": 0.0129, "step": 16550 }, { "grad_norm": 0.17653129994869232, "learning_rate": 9.341210571210954e-05, "loss": 0.0159, "step": 16560 }, { "grad_norm": 0.1888950616121292, "learning_rate": 9.340184758175338e-05, "loss": 0.0167, "step": 16570 }, { "grad_norm": 0.1657741665840149, "learning_rate": 9.339158203520854e-05, "loss": 0.0144, "step": 16580 }, { "grad_norm": 0.1509900838136673, "learning_rate": 9.338130907422908e-05, "loss": 0.0146, "step": 16590 }, { "grad_norm": 0.1742399036884308, "learning_rate": 9.337102870057037e-05, "loss": 0.0147, "step": 16600 }, { "grad_norm": 0.18770025670528412, "learning_rate": 9.336074091598907e-05, "loss": 0.0139, "step": 16610 }, { "grad_norm": 0.21226978302001953, "learning_rate": 9.335044572224306e-05, "loss": 0.0147, "step": 16620 }, { "grad_norm": 0.16879944503307343, "learning_rate": 9.334014312109151e-05, "loss": 0.0182, "step": 16630 }, { "grad_norm": 0.1998162865638733, "learning_rate": 9.332983311429486e-05, "loss": 0.0158, "step": 16640 }, { "grad_norm": 0.18027056753635406, "learning_rate": 9.33195157036148e-05, "loss": 0.0208, "step": 16650 }, { "grad_norm": 0.17996352910995483, "learning_rate": 9.330919089081432e-05, "loss": 0.0153, "step": 16660 }, { "grad_norm": 0.23712168633937836, "learning_rate": 9.32988586776576e-05, "loss": 0.0197, "step": 16670 }, { "grad_norm": 0.16917669773101807, "learning_rate": 9.328851906591016e-05, "loss": 0.0153, "step": 16680 }, { "grad_norm": 0.1734936386346817, "learning_rate": 9.327817205733875e-05, "loss": 0.0128, "step": 16690 }, { "grad_norm": 0.11972777545452118, "learning_rate": 9.326781765371142e-05, "loss": 0.0138, "step": 16700 }, { "grad_norm": 0.12494630366563797, "learning_rate": 9.325745585679741e-05, "loss": 0.0168, "step": 16710 }, { "grad_norm": 0.15712566673755646, "learning_rate": 9.32470866683673e-05, "loss": 0.013, "step": 16720 }, { "grad_norm": 0.1973295360803604, "learning_rate": 9.323671009019288e-05, "loss": 0.0147, "step": 16730 }, { "grad_norm": 0.29092782735824585, "learning_rate": 9.322632612404725e-05, "loss": 0.019, "step": 16740 }, { "grad_norm": 0.20878049731254578, "learning_rate": 9.321593477170471e-05, "loss": 0.0165, "step": 16750 }, { "grad_norm": 0.18979725241661072, "learning_rate": 9.320553603494088e-05, "loss": 0.0172, "step": 16760 }, { "grad_norm": 0.21865522861480713, "learning_rate": 9.319512991553261e-05, "loss": 0.0177, "step": 16770 }, { "grad_norm": 0.3006628453731537, "learning_rate": 9.318471641525803e-05, "loss": 0.0176, "step": 16780 }, { "grad_norm": 0.2684422433376312, "learning_rate": 9.317429553589652e-05, "loss": 0.0149, "step": 16790 }, { "grad_norm": 0.1876288801431656, "learning_rate": 9.316386727922873e-05, "loss": 0.0157, "step": 16800 }, { "grad_norm": 0.21250469982624054, "learning_rate": 9.315343164703656e-05, "loss": 0.0117, "step": 16810 }, { "grad_norm": 0.24900281429290771, "learning_rate": 9.314298864110316e-05, "loss": 0.0146, "step": 16820 }, { "grad_norm": 0.1857692003250122, "learning_rate": 9.313253826321295e-05, "loss": 0.0155, "step": 16830 }, { "grad_norm": 0.21865788102149963, "learning_rate": 9.312208051515165e-05, "loss": 0.0138, "step": 16840 }, { "grad_norm": 0.17538058757781982, "learning_rate": 9.311161539870618e-05, "loss": 0.0167, "step": 16850 }, { "grad_norm": 0.18645866215229034, "learning_rate": 9.310114291566474e-05, "loss": 0.0139, "step": 16860 }, { "grad_norm": 0.20039254426956177, "learning_rate": 9.309066306781679e-05, "loss": 0.0152, "step": 16870 }, { "grad_norm": 0.17673605680465698, "learning_rate": 9.308017585695306e-05, "loss": 0.0146, "step": 16880 }, { "grad_norm": 0.19600452482700348, "learning_rate": 9.306968128486552e-05, "loss": 0.0157, "step": 16890 }, { "grad_norm": 0.22994369268417358, "learning_rate": 9.30591793533474e-05, "loss": 0.0172, "step": 16900 }, { "grad_norm": 0.2078111171722412, "learning_rate": 9.304867006419321e-05, "loss": 0.0147, "step": 16910 }, { "grad_norm": 0.23404042422771454, "learning_rate": 9.303815341919868e-05, "loss": 0.0189, "step": 16920 }, { "grad_norm": 0.1681479513645172, "learning_rate": 9.302762942016084e-05, "loss": 0.0131, "step": 16930 }, { "grad_norm": 0.23889461159706116, "learning_rate": 9.301709806887792e-05, "loss": 0.0146, "step": 16940 }, { "grad_norm": 0.19615747034549713, "learning_rate": 9.300655936714948e-05, "loss": 0.0178, "step": 16950 }, { "grad_norm": 0.19753314554691315, "learning_rate": 9.299601331677627e-05, "loss": 0.0195, "step": 16960 }, { "grad_norm": 0.1752304583787918, "learning_rate": 9.298545991956033e-05, "loss": 0.0185, "step": 16970 }, { "grad_norm": 0.2457280457019806, "learning_rate": 9.297489917730493e-05, "loss": 0.0154, "step": 16980 }, { "grad_norm": 0.239800363779068, "learning_rate": 9.296433109181464e-05, "loss": 0.015, "step": 16990 }, { "grad_norm": 0.18307335674762726, "learning_rate": 9.295375566489523e-05, "loss": 0.0166, "step": 17000 }, { "grad_norm": 0.2576836049556732, "learning_rate": 9.294317289835379e-05, "loss": 0.0215, "step": 17010 }, { "grad_norm": 0.21626725792884827, "learning_rate": 9.293258279399859e-05, "loss": 0.0179, "step": 17020 }, { "grad_norm": 0.21495628356933594, "learning_rate": 9.292198535363919e-05, "loss": 0.0157, "step": 17030 }, { "grad_norm": 0.18812932074069977, "learning_rate": 9.291138057908641e-05, "loss": 0.0165, "step": 17040 }, { "grad_norm": 0.34788456559181213, "learning_rate": 9.290076847215234e-05, "loss": 0.0194, "step": 17050 }, { "grad_norm": 0.18641312420368195, "learning_rate": 9.289014903465025e-05, "loss": 0.0172, "step": 17060 }, { "grad_norm": 0.21807558834552765, "learning_rate": 9.287952226839475e-05, "loss": 0.0172, "step": 17070 }, { "grad_norm": 0.21221035718917847, "learning_rate": 9.286888817520164e-05, "loss": 0.0139, "step": 17080 }, { "grad_norm": 0.2611849009990692, "learning_rate": 9.285824675688803e-05, "loss": 0.0185, "step": 17090 }, { "grad_norm": 0.16270023584365845, "learning_rate": 9.28475980152722e-05, "loss": 0.0144, "step": 17100 }, { "grad_norm": 0.26045817136764526, "learning_rate": 9.283694195217379e-05, "loss": 0.0145, "step": 17110 }, { "grad_norm": 0.19767536222934723, "learning_rate": 9.282627856941356e-05, "loss": 0.0157, "step": 17120 }, { "grad_norm": 0.17585183680057526, "learning_rate": 9.281560786881363e-05, "loss": 0.0158, "step": 17130 }, { "grad_norm": 0.25512099266052246, "learning_rate": 9.280492985219733e-05, "loss": 0.0183, "step": 17140 }, { "grad_norm": 0.1710432916879654, "learning_rate": 9.279424452138924e-05, "loss": 0.0162, "step": 17150 }, { "grad_norm": 0.17905260622501373, "learning_rate": 9.278355187821517e-05, "loss": 0.0163, "step": 17160 }, { "grad_norm": 0.16385629773139954, "learning_rate": 9.277285192450224e-05, "loss": 0.0132, "step": 17170 }, { "grad_norm": 0.16452378034591675, "learning_rate": 9.276214466207875e-05, "loss": 0.0139, "step": 17180 }, { "grad_norm": 0.17742861807346344, "learning_rate": 9.275143009277427e-05, "loss": 0.018, "step": 17190 }, { "grad_norm": 0.20725807547569275, "learning_rate": 9.274070821841964e-05, "loss": 0.014, "step": 17200 }, { "grad_norm": 0.21451465785503387, "learning_rate": 9.272997904084696e-05, "loss": 0.013, "step": 17210 }, { "grad_norm": 0.22913554310798645, "learning_rate": 9.271924256188951e-05, "loss": 0.0159, "step": 17220 }, { "grad_norm": 0.27058514952659607, "learning_rate": 9.270849878338189e-05, "loss": 0.0199, "step": 17230 }, { "grad_norm": 0.23310431838035583, "learning_rate": 9.269774770715991e-05, "loss": 0.0187, "step": 17240 }, { "grad_norm": 0.22391007840633392, "learning_rate": 9.268698933506061e-05, "loss": 0.0153, "step": 17250 }, { "grad_norm": 0.20207615196704865, "learning_rate": 9.267622366892235e-05, "loss": 0.0176, "step": 17260 }, { "grad_norm": 0.22535714507102966, "learning_rate": 9.266545071058465e-05, "loss": 0.0155, "step": 17270 }, { "grad_norm": 0.202767014503479, "learning_rate": 9.265467046188833e-05, "loss": 0.0155, "step": 17280 }, { "grad_norm": 0.18402281403541565, "learning_rate": 9.264388292467543e-05, "loss": 0.0156, "step": 17290 }, { "grad_norm": 0.15002749860286713, "learning_rate": 9.263308810078926e-05, "loss": 0.0168, "step": 17300 }, { "grad_norm": 0.15851089358329773, "learning_rate": 9.262228599207434e-05, "loss": 0.0148, "step": 17310 }, { "grad_norm": 0.18077293038368225, "learning_rate": 9.261147660037647e-05, "loss": 0.0153, "step": 17320 }, { "grad_norm": 0.16594336926937103, "learning_rate": 9.26006599275427e-05, "loss": 0.0183, "step": 17330 }, { "grad_norm": 0.18172140419483185, "learning_rate": 9.258983597542124e-05, "loss": 0.0161, "step": 17340 }, { "grad_norm": 0.20665878057479858, "learning_rate": 9.257900474586167e-05, "loss": 0.0179, "step": 17350 }, { "grad_norm": 0.12166878581047058, "learning_rate": 9.256816624071471e-05, "loss": 0.0144, "step": 17360 }, { "grad_norm": 0.19705678522586823, "learning_rate": 9.25573204618324e-05, "loss": 0.0123, "step": 17370 }, { "grad_norm": 0.20576637983322144, "learning_rate": 9.254646741106796e-05, "loss": 0.0145, "step": 17380 }, { "grad_norm": 0.23388388752937317, "learning_rate": 9.253560709027589e-05, "loss": 0.0158, "step": 17390 }, { "grad_norm": 0.17234647274017334, "learning_rate": 9.252473950131192e-05, "loss": 0.0176, "step": 17400 }, { "grad_norm": 0.1457904726266861, "learning_rate": 9.251386464603302e-05, "loss": 0.0144, "step": 17410 }, { "grad_norm": 0.25219258666038513, "learning_rate": 9.250298252629741e-05, "loss": 0.0141, "step": 17420 }, { "grad_norm": 0.18759150803089142, "learning_rate": 9.249209314396454e-05, "loss": 0.0143, "step": 17430 }, { "grad_norm": 0.22946606576442719, "learning_rate": 9.248119650089513e-05, "loss": 0.0153, "step": 17440 }, { "grad_norm": 0.24728648364543915, "learning_rate": 9.247029259895108e-05, "loss": 0.0173, "step": 17450 }, { "grad_norm": 0.15805895626544952, "learning_rate": 9.24593814399956e-05, "loss": 0.0135, "step": 17460 }, { "grad_norm": 0.17012950778007507, "learning_rate": 9.244846302589309e-05, "loss": 0.0169, "step": 17470 }, { "grad_norm": 0.1890903264284134, "learning_rate": 9.243753735850923e-05, "loss": 0.0157, "step": 17480 }, { "grad_norm": 0.20732243359088898, "learning_rate": 9.24266044397109e-05, "loss": 0.017, "step": 17490 }, { "grad_norm": 0.19487081468105316, "learning_rate": 9.241566427136624e-05, "loss": 0.0155, "step": 17500 }, { "grad_norm": 0.2906295955181122, "learning_rate": 9.240471685534463e-05, "loss": 0.0162, "step": 17510 }, { "grad_norm": 0.2379431575536728, "learning_rate": 9.239376219351667e-05, "loss": 0.0132, "step": 17520 }, { "grad_norm": 0.19793689250946045, "learning_rate": 9.238280028775425e-05, "loss": 0.0155, "step": 17530 }, { "grad_norm": 0.14790894091129303, "learning_rate": 9.237183113993041e-05, "loss": 0.0151, "step": 17540 }, { "grad_norm": 0.1922716349363327, "learning_rate": 9.236085475191952e-05, "loss": 0.0138, "step": 17550 }, { "grad_norm": 0.172505721449852, "learning_rate": 9.234987112559709e-05, "loss": 0.0153, "step": 17560 }, { "grad_norm": 0.2114626169204712, "learning_rate": 9.233888026283999e-05, "loss": 0.0163, "step": 17570 }, { "grad_norm": 0.27163374423980713, "learning_rate": 9.232788216552619e-05, "loss": 0.0166, "step": 17580 }, { "grad_norm": 0.16292493045330048, "learning_rate": 9.231687683553502e-05, "loss": 0.0161, "step": 17590 }, { "grad_norm": 0.2110169380903244, "learning_rate": 9.230586427474698e-05, "loss": 0.0136, "step": 17600 }, { "grad_norm": 0.16032956540584564, "learning_rate": 9.229484448504379e-05, "loss": 0.0156, "step": 17610 }, { "grad_norm": 0.21298927068710327, "learning_rate": 9.228381746830843e-05, "loss": 0.0143, "step": 17620 }, { "grad_norm": 0.17557291686534882, "learning_rate": 9.227278322642514e-05, "loss": 0.0146, "step": 17630 }, { "grad_norm": 0.16314220428466797, "learning_rate": 9.226174176127937e-05, "loss": 0.0163, "step": 17640 }, { "grad_norm": 0.19644732773303986, "learning_rate": 9.22506930747578e-05, "loss": 0.0119, "step": 17650 }, { "grad_norm": 0.20777320861816406, "learning_rate": 9.223963716874831e-05, "loss": 0.0122, "step": 17660 }, { "grad_norm": 0.1903933435678482, "learning_rate": 9.222857404514012e-05, "loss": 0.016, "step": 17670 }, { "grad_norm": 0.20678211748600006, "learning_rate": 9.221750370582355e-05, "loss": 0.0201, "step": 17680 }, { "grad_norm": 0.25635504722595215, "learning_rate": 9.220642615269028e-05, "loss": 0.0153, "step": 17690 }, { "grad_norm": 0.21239998936653137, "learning_rate": 9.219534138763311e-05, "loss": 0.0151, "step": 17700 }, { "grad_norm": 0.2569771111011505, "learning_rate": 9.218424941254613e-05, "loss": 0.0138, "step": 17710 }, { "grad_norm": 0.14362169802188873, "learning_rate": 9.217315022932468e-05, "loss": 0.0181, "step": 17720 }, { "grad_norm": 0.18235982954502106, "learning_rate": 9.216204383986528e-05, "loss": 0.0151, "step": 17730 }, { "grad_norm": 0.15979978442192078, "learning_rate": 9.215093024606574e-05, "loss": 0.0184, "step": 17740 }, { "grad_norm": 0.21010367572307587, "learning_rate": 9.213980944982506e-05, "loss": 0.0138, "step": 17750 }, { "grad_norm": 0.20114266872406006, "learning_rate": 9.212868145304346e-05, "loss": 0.021, "step": 17760 }, { "grad_norm": 0.21386680006980896, "learning_rate": 9.211754625762241e-05, "loss": 0.016, "step": 17770 }, { "grad_norm": 0.17126014828681946, "learning_rate": 9.210640386546463e-05, "loss": 0.0153, "step": 17780 }, { "grad_norm": 0.14575059711933136, "learning_rate": 9.209525427847405e-05, "loss": 0.016, "step": 17790 }, { "grad_norm": 0.19602926075458527, "learning_rate": 9.208409749855583e-05, "loss": 0.0148, "step": 17800 }, { "grad_norm": 0.20025832951068878, "learning_rate": 9.207293352761633e-05, "loss": 0.0189, "step": 17810 }, { "grad_norm": 0.1960688978433609, "learning_rate": 9.206176236756319e-05, "loss": 0.0169, "step": 17820 }, { "grad_norm": 0.21855135262012482, "learning_rate": 9.205058402030525e-05, "loss": 0.0136, "step": 17830 }, { "grad_norm": 0.1656602919101715, "learning_rate": 9.203939848775259e-05, "loss": 0.0155, "step": 17840 }, { "grad_norm": 0.25461432337760925, "learning_rate": 9.202820577181652e-05, "loss": 0.0151, "step": 17850 }, { "grad_norm": 0.2882099449634552, "learning_rate": 9.201700587440953e-05, "loss": 0.0175, "step": 17860 }, { "grad_norm": 0.341415137052536, "learning_rate": 9.200579879744544e-05, "loss": 0.02, "step": 17870 }, { "grad_norm": 0.21048197150230408, "learning_rate": 9.199458454283918e-05, "loss": 0.0144, "step": 17880 }, { "grad_norm": 0.18323716521263123, "learning_rate": 9.198336311250697e-05, "loss": 0.0182, "step": 17890 }, { "grad_norm": 0.2175145149230957, "learning_rate": 9.197213450836626e-05, "loss": 0.0153, "step": 17900 }, { "grad_norm": 0.25669944286346436, "learning_rate": 9.19608987323357e-05, "loss": 0.0172, "step": 17910 }, { "grad_norm": 0.18096506595611572, "learning_rate": 9.194965578633517e-05, "loss": 0.0186, "step": 17920 }, { "grad_norm": 0.1741887629032135, "learning_rate": 9.193840567228582e-05, "loss": 0.0135, "step": 17930 }, { "grad_norm": 0.22216752171516418, "learning_rate": 9.192714839210994e-05, "loss": 0.0162, "step": 17940 }, { "grad_norm": 0.20719623565673828, "learning_rate": 9.19158839477311e-05, "loss": 0.0127, "step": 17950 }, { "grad_norm": 0.2384558469057083, "learning_rate": 9.190461234107411e-05, "loss": 0.018, "step": 17960 }, { "grad_norm": 0.23778226971626282, "learning_rate": 9.189333357406496e-05, "loss": 0.016, "step": 17970 }, { "grad_norm": 0.21642230451107025, "learning_rate": 9.188204764863089e-05, "loss": 0.0148, "step": 17980 }, { "grad_norm": 0.17561082541942596, "learning_rate": 9.187075456670033e-05, "loss": 0.016, "step": 17990 }, { "grad_norm": 0.19204701483249664, "learning_rate": 9.1859454330203e-05, "loss": 0.0138, "step": 18000 }, { "grad_norm": 0.204100102186203, "learning_rate": 9.18481469410698e-05, "loss": 0.0161, "step": 18010 }, { "grad_norm": 0.1744370460510254, "learning_rate": 9.183683240123281e-05, "loss": 0.0143, "step": 18020 }, { "grad_norm": 0.21148833632469177, "learning_rate": 9.182551071262541e-05, "loss": 0.014, "step": 18030 }, { "grad_norm": 0.17550867795944214, "learning_rate": 9.181418187718218e-05, "loss": 0.0149, "step": 18040 }, { "grad_norm": 0.2360689640045166, "learning_rate": 9.180284589683888e-05, "loss": 0.016, "step": 18050 }, { "grad_norm": 0.29082155227661133, "learning_rate": 9.17915027735325e-05, "loss": 0.0179, "step": 18060 }, { "grad_norm": 0.19825154542922974, "learning_rate": 9.178015250920133e-05, "loss": 0.0183, "step": 18070 }, { "grad_norm": 0.194020614027977, "learning_rate": 9.176879510578477e-05, "loss": 0.0147, "step": 18080 }, { "grad_norm": 0.17289502918720245, "learning_rate": 9.17574305652235e-05, "loss": 0.0175, "step": 18090 }, { "grad_norm": 0.2091646045446396, "learning_rate": 9.174605888945942e-05, "loss": 0.0162, "step": 18100 }, { "grad_norm": 0.17336700856685638, "learning_rate": 9.173468008043564e-05, "loss": 0.015, "step": 18110 }, { "grad_norm": 0.25515079498291016, "learning_rate": 9.172329414009648e-05, "loss": 0.0209, "step": 18120 }, { "grad_norm": 0.18252073228359222, "learning_rate": 9.171190107038747e-05, "loss": 0.0167, "step": 18130 }, { "grad_norm": 0.1836196929216385, "learning_rate": 9.170050087325541e-05, "loss": 0.0128, "step": 18140 }, { "grad_norm": 0.16180972754955292, "learning_rate": 9.168909355064824e-05, "loss": 0.0114, "step": 18150 }, { "grad_norm": 0.19751319289207458, "learning_rate": 9.167767910451519e-05, "loss": 0.0141, "step": 18160 }, { "grad_norm": 0.21143637597560883, "learning_rate": 9.166625753680669e-05, "loss": 0.0149, "step": 18170 }, { "grad_norm": 0.22256405651569366, "learning_rate": 9.165482884947431e-05, "loss": 0.0154, "step": 18180 }, { "grad_norm": 0.17026254534721375, "learning_rate": 9.164339304447098e-05, "loss": 0.0179, "step": 18190 }, { "grad_norm": 0.21586935222148895, "learning_rate": 9.163195012375072e-05, "loss": 0.017, "step": 18200 }, { "grad_norm": 0.19957469403743744, "learning_rate": 9.16205000892688e-05, "loss": 0.0147, "step": 18210 }, { "grad_norm": 0.16990841925144196, "learning_rate": 9.160904294298175e-05, "loss": 0.0137, "step": 18220 }, { "grad_norm": 0.1742437332868576, "learning_rate": 9.159757868684727e-05, "loss": 0.0166, "step": 18230 }, { "grad_norm": 0.14150413870811462, "learning_rate": 9.15861073228243e-05, "loss": 0.0171, "step": 18240 }, { "grad_norm": 0.1749410331249237, "learning_rate": 9.157462885287296e-05, "loss": 0.0152, "step": 18250 }, { "grad_norm": 0.2257196456193924, "learning_rate": 9.156314327895461e-05, "loss": 0.0153, "step": 18260 }, { "grad_norm": 0.2144749015569687, "learning_rate": 9.155165060303185e-05, "loss": 0.0171, "step": 18270 }, { "grad_norm": 0.22404006123542786, "learning_rate": 9.154015082706841e-05, "loss": 0.0168, "step": 18280 }, { "grad_norm": 0.20610439777374268, "learning_rate": 9.152864395302936e-05, "loss": 0.0141, "step": 18290 }, { "grad_norm": 0.16320574283599854, "learning_rate": 9.151712998288085e-05, "loss": 0.013, "step": 18300 }, { "grad_norm": 0.19051234424114227, "learning_rate": 9.150560891859031e-05, "loss": 0.0195, "step": 18310 }, { "grad_norm": 0.1958145946264267, "learning_rate": 9.14940807621264e-05, "loss": 0.0139, "step": 18320 }, { "grad_norm": 0.22539806365966797, "learning_rate": 9.148254551545894e-05, "loss": 0.0178, "step": 18330 }, { "grad_norm": 0.20832835137844086, "learning_rate": 9.147100318055901e-05, "loss": 0.016, "step": 18340 }, { "grad_norm": 0.2667052447795868, "learning_rate": 9.145945375939888e-05, "loss": 0.0171, "step": 18350 }, { "grad_norm": 0.2237963080406189, "learning_rate": 9.144789725395203e-05, "loss": 0.0149, "step": 18360 }, { "grad_norm": 0.16010117530822754, "learning_rate": 9.14363336661931e-05, "loss": 0.0141, "step": 18370 }, { "grad_norm": 0.19362559914588928, "learning_rate": 9.142476299809806e-05, "loss": 0.0156, "step": 18380 }, { "grad_norm": 0.18736977875232697, "learning_rate": 9.1413185251644e-05, "loss": 0.0117, "step": 18390 }, { "grad_norm": 0.15934845805168152, "learning_rate": 9.140160042880923e-05, "loss": 0.0148, "step": 18400 }, { "grad_norm": 0.18727493286132812, "learning_rate": 9.139000853157327e-05, "loss": 0.0142, "step": 18410 }, { "grad_norm": 0.2303777039051056, "learning_rate": 9.137840956191688e-05, "loss": 0.0146, "step": 18420 }, { "grad_norm": 0.19079740345478058, "learning_rate": 9.136680352182199e-05, "loss": 0.0166, "step": 18430 }, { "grad_norm": 0.19323007762432098, "learning_rate": 9.135519041327177e-05, "loss": 0.0169, "step": 18440 }, { "grad_norm": 0.18977022171020508, "learning_rate": 9.134357023825058e-05, "loss": 0.0168, "step": 18450 }, { "grad_norm": 0.24107855558395386, "learning_rate": 9.133194299874398e-05, "loss": 0.0133, "step": 18460 }, { "grad_norm": 0.12496183067560196, "learning_rate": 9.132030869673876e-05, "loss": 0.0131, "step": 18470 }, { "grad_norm": 0.2088342159986496, "learning_rate": 9.130866733422288e-05, "loss": 0.0129, "step": 18480 }, { "grad_norm": 0.2534388303756714, "learning_rate": 9.129701891318556e-05, "loss": 0.0183, "step": 18490 }, { "grad_norm": 0.22327151894569397, "learning_rate": 9.128536343561718e-05, "loss": 0.0157, "step": 18500 }, { "grad_norm": 0.24228952825069427, "learning_rate": 9.127370090350934e-05, "loss": 0.016, "step": 18510 }, { "grad_norm": 0.2328600138425827, "learning_rate": 9.126203131885487e-05, "loss": 0.0156, "step": 18520 }, { "grad_norm": 0.20090967416763306, "learning_rate": 9.125035468364775e-05, "loss": 0.0144, "step": 18530 }, { "grad_norm": 0.21118541061878204, "learning_rate": 9.123867099988322e-05, "loss": 0.0161, "step": 18540 }, { "grad_norm": 0.16707338392734528, "learning_rate": 9.122698026955769e-05, "loss": 0.0166, "step": 18550 }, { "grad_norm": 0.15865390002727509, "learning_rate": 9.12152824946688e-05, "loss": 0.0138, "step": 18560 }, { "grad_norm": 0.1846635937690735, "learning_rate": 9.120357767721538e-05, "loss": 0.0134, "step": 18570 }, { "grad_norm": 0.16113655269145966, "learning_rate": 9.119186581919745e-05, "loss": 0.0123, "step": 18580 }, { "grad_norm": 0.22157487273216248, "learning_rate": 9.118014692261624e-05, "loss": 0.0148, "step": 18590 }, { "grad_norm": 0.2198432981967926, "learning_rate": 9.116842098947422e-05, "loss": 0.0139, "step": 18600 }, { "grad_norm": 0.23755665123462677, "learning_rate": 9.115668802177499e-05, "loss": 0.0165, "step": 18610 }, { "grad_norm": 0.26722753047943115, "learning_rate": 9.114494802152342e-05, "loss": 0.0174, "step": 18620 }, { "grad_norm": 0.1835045963525772, "learning_rate": 9.113320099072555e-05, "loss": 0.0164, "step": 18630 }, { "grad_norm": 0.173446923494339, "learning_rate": 9.112144693138864e-05, "loss": 0.0127, "step": 18640 }, { "grad_norm": 0.12361949682235718, "learning_rate": 9.110968584552111e-05, "loss": 0.0129, "step": 18650 }, { "grad_norm": 0.14354373514652252, "learning_rate": 9.109791773513264e-05, "loss": 0.0134, "step": 18660 }, { "grad_norm": 0.20337015390396118, "learning_rate": 9.108614260223403e-05, "loss": 0.013, "step": 18670 }, { "grad_norm": 0.19487524032592773, "learning_rate": 9.107436044883738e-05, "loss": 0.0162, "step": 18680 }, { "grad_norm": 0.17937670648097992, "learning_rate": 9.10625712769559e-05, "loss": 0.0151, "step": 18690 }, { "grad_norm": 0.13788801431655884, "learning_rate": 9.105077508860406e-05, "loss": 0.0135, "step": 18700 }, { "grad_norm": 0.16669730842113495, "learning_rate": 9.103897188579751e-05, "loss": 0.0125, "step": 18710 }, { "grad_norm": 0.14051353931427002, "learning_rate": 9.102716167055308e-05, "loss": 0.0126, "step": 18720 }, { "grad_norm": 0.17896658182144165, "learning_rate": 9.10153444448888e-05, "loss": 0.0159, "step": 18730 }, { "grad_norm": 0.18165169656276703, "learning_rate": 9.100352021082393e-05, "loss": 0.0148, "step": 18740 }, { "grad_norm": 0.23769527673721313, "learning_rate": 9.099168897037891e-05, "loss": 0.0145, "step": 18750 }, { "grad_norm": 0.12679032981395721, "learning_rate": 9.097985072557538e-05, "loss": 0.0117, "step": 18760 }, { "grad_norm": 0.16772393882274628, "learning_rate": 9.096800547843615e-05, "loss": 0.0135, "step": 18770 }, { "grad_norm": 0.22464506328105927, "learning_rate": 9.095615323098526e-05, "loss": 0.0133, "step": 18780 }, { "grad_norm": 0.19364666938781738, "learning_rate": 9.094429398524795e-05, "loss": 0.0118, "step": 18790 }, { "grad_norm": 0.24793414771556854, "learning_rate": 9.093242774325061e-05, "loss": 0.0154, "step": 18800 }, { "grad_norm": 0.15649910271167755, "learning_rate": 9.092055450702088e-05, "loss": 0.0142, "step": 18810 }, { "grad_norm": 0.1915372908115387, "learning_rate": 9.090867427858756e-05, "loss": 0.018, "step": 18820 }, { "grad_norm": 0.20356839895248413, "learning_rate": 9.089678705998066e-05, "loss": 0.0152, "step": 18830 }, { "grad_norm": 0.24186597764492035, "learning_rate": 9.088489285323139e-05, "loss": 0.0141, "step": 18840 }, { "grad_norm": 0.23540888726711273, "learning_rate": 9.087299166037212e-05, "loss": 0.0142, "step": 18850 }, { "grad_norm": 0.2597535252571106, "learning_rate": 9.086108348343647e-05, "loss": 0.0136, "step": 18860 }, { "grad_norm": 0.14900387823581696, "learning_rate": 9.08491683244592e-05, "loss": 0.0129, "step": 18870 }, { "grad_norm": 0.21057410538196564, "learning_rate": 9.08372461854763e-05, "loss": 0.0181, "step": 18880 }, { "grad_norm": 0.1830531656742096, "learning_rate": 9.082531706852492e-05, "loss": 0.0184, "step": 18890 }, { "grad_norm": 0.18707984685897827, "learning_rate": 9.081338097564342e-05, "loss": 0.0162, "step": 18900 }, { "grad_norm": 0.17318391799926758, "learning_rate": 9.080143790887137e-05, "loss": 0.0163, "step": 18910 }, { "grad_norm": 0.21545055508613586, "learning_rate": 9.07894878702495e-05, "loss": 0.0161, "step": 18920 }, { "grad_norm": 0.1851704716682434, "learning_rate": 9.077753086181974e-05, "loss": 0.0158, "step": 18930 }, { "grad_norm": 0.19039234519004822, "learning_rate": 9.076556688562524e-05, "loss": 0.016, "step": 18940 }, { "grad_norm": 0.22295747697353363, "learning_rate": 9.075359594371029e-05, "loss": 0.0155, "step": 18950 }, { "grad_norm": 0.19876070320606232, "learning_rate": 9.07416180381204e-05, "loss": 0.0169, "step": 18960 }, { "grad_norm": 0.20903059840202332, "learning_rate": 9.072963317090228e-05, "loss": 0.0148, "step": 18970 }, { "grad_norm": 0.23441289365291595, "learning_rate": 9.071764134410382e-05, "loss": 0.0168, "step": 18980 }, { "grad_norm": 0.18961374461650848, "learning_rate": 9.070564255977407e-05, "loss": 0.0173, "step": 18990 }, { "grad_norm": 0.22104206681251526, "learning_rate": 9.06936368199633e-05, "loss": 0.0177, "step": 19000 }, { "grad_norm": 0.1988217979669571, "learning_rate": 9.0681624126723e-05, "loss": 0.0158, "step": 19010 }, { "grad_norm": 0.19446052610874176, "learning_rate": 9.066960448210576e-05, "loss": 0.0154, "step": 19020 }, { "grad_norm": 0.22661161422729492, "learning_rate": 9.065757788816543e-05, "loss": 0.0175, "step": 19030 }, { "grad_norm": 0.1802348792552948, "learning_rate": 9.064554434695705e-05, "loss": 0.0135, "step": 19040 }, { "grad_norm": 0.23344352841377258, "learning_rate": 9.063350386053677e-05, "loss": 0.0163, "step": 19050 }, { "grad_norm": 0.16027042269706726, "learning_rate": 9.062145643096202e-05, "loss": 0.0138, "step": 19060 }, { "grad_norm": 0.15248720347881317, "learning_rate": 9.060940206029136e-05, "loss": 0.016, "step": 19070 }, { "grad_norm": 0.20324234664440155, "learning_rate": 9.059734075058457e-05, "loss": 0.0195, "step": 19080 }, { "grad_norm": 0.21586498618125916, "learning_rate": 9.058527250390257e-05, "loss": 0.0161, "step": 19090 }, { "grad_norm": 0.18039296567440033, "learning_rate": 9.057319732230752e-05, "loss": 0.0144, "step": 19100 }, { "grad_norm": 0.16866351664066315, "learning_rate": 9.056111520786273e-05, "loss": 0.0133, "step": 19110 }, { "grad_norm": 0.1801162213087082, "learning_rate": 9.054902616263268e-05, "loss": 0.0156, "step": 19120 }, { "grad_norm": 0.17856216430664062, "learning_rate": 9.05369301886831e-05, "loss": 0.0138, "step": 19130 }, { "grad_norm": 0.14521759748458862, "learning_rate": 9.052482728808083e-05, "loss": 0.0134, "step": 19140 }, { "grad_norm": 0.18124017119407654, "learning_rate": 9.051271746289391e-05, "loss": 0.0145, "step": 19150 }, { "grad_norm": 0.1606343388557434, "learning_rate": 9.050060071519162e-05, "loss": 0.0122, "step": 19160 }, { "grad_norm": 0.1990022361278534, "learning_rate": 9.048847704704437e-05, "loss": 0.0129, "step": 19170 }, { "grad_norm": 0.201959490776062, "learning_rate": 9.047634646052376e-05, "loss": 0.0162, "step": 19180 }, { "grad_norm": 0.14377468824386597, "learning_rate": 9.046420895770256e-05, "loss": 0.0152, "step": 19190 }, { "grad_norm": 0.15919743478298187, "learning_rate": 9.045206454065473e-05, "loss": 0.0144, "step": 19200 }, { "grad_norm": 0.1367906779050827, "learning_rate": 9.043991321145546e-05, "loss": 0.0122, "step": 19210 }, { "grad_norm": 0.1554226577281952, "learning_rate": 9.042775497218105e-05, "loss": 0.0113, "step": 19220 }, { "grad_norm": 0.24721449613571167, "learning_rate": 9.041558982490901e-05, "loss": 0.0144, "step": 19230 }, { "grad_norm": 0.19738972187042236, "learning_rate": 9.040341777171805e-05, "loss": 0.0134, "step": 19240 }, { "grad_norm": 0.17164137959480286, "learning_rate": 9.039123881468802e-05, "loss": 0.017, "step": 19250 }, { "grad_norm": 0.15104235708713531, "learning_rate": 9.037905295589998e-05, "loss": 0.012, "step": 19260 }, { "grad_norm": 0.1716039627790451, "learning_rate": 9.036686019743617e-05, "loss": 0.0135, "step": 19270 }, { "grad_norm": 0.13624060153961182, "learning_rate": 9.035466054137997e-05, "loss": 0.0161, "step": 19280 }, { "grad_norm": 0.1791962832212448, "learning_rate": 9.0342453989816e-05, "loss": 0.0145, "step": 19290 }, { "grad_norm": 0.1775127351284027, "learning_rate": 9.033024054483e-05, "loss": 0.0117, "step": 19300 }, { "grad_norm": 0.1743019074201584, "learning_rate": 9.031802020850894e-05, "loss": 0.0111, "step": 19310 }, { "grad_norm": 0.18125182390213013, "learning_rate": 9.030579298294092e-05, "loss": 0.013, "step": 19320 }, { "grad_norm": 0.13328762352466583, "learning_rate": 9.029355887021524e-05, "loss": 0.016, "step": 19330 }, { "grad_norm": 0.14886756241321564, "learning_rate": 9.028131787242238e-05, "loss": 0.0152, "step": 19340 }, { "grad_norm": 0.1361846923828125, "learning_rate": 9.026906999165399e-05, "loss": 0.0133, "step": 19350 }, { "grad_norm": 0.24462001025676727, "learning_rate": 9.025681523000291e-05, "loss": 0.0143, "step": 19360 }, { "grad_norm": 0.1449802815914154, "learning_rate": 9.024455358956315e-05, "loss": 0.0137, "step": 19370 }, { "grad_norm": 0.20838385820388794, "learning_rate": 9.023228507242984e-05, "loss": 0.0144, "step": 19380 }, { "grad_norm": 0.26524728536605835, "learning_rate": 9.022000968069937e-05, "loss": 0.0175, "step": 19390 }, { "grad_norm": 0.22918367385864258, "learning_rate": 9.020772741646928e-05, "loss": 0.0155, "step": 19400 }, { "grad_norm": 0.23580186069011688, "learning_rate": 9.019543828183826e-05, "loss": 0.0175, "step": 19410 }, { "grad_norm": 0.1402084231376648, "learning_rate": 9.018314227890616e-05, "loss": 0.0133, "step": 19420 }, { "grad_norm": 0.15188981592655182, "learning_rate": 9.017083940977408e-05, "loss": 0.0144, "step": 19430 }, { "grad_norm": 0.2090679109096527, "learning_rate": 9.015852967654422e-05, "loss": 0.0139, "step": 19440 }, { "grad_norm": 0.2120629996061325, "learning_rate": 9.014621308131996e-05, "loss": 0.0148, "step": 19450 }, { "grad_norm": 0.22583357989788055, "learning_rate": 9.01338896262059e-05, "loss": 0.0141, "step": 19460 }, { "grad_norm": 0.20906202495098114, "learning_rate": 9.012155931330777e-05, "loss": 0.0162, "step": 19470 }, { "grad_norm": 0.203958660364151, "learning_rate": 9.010922214473246e-05, "loss": 0.0159, "step": 19480 }, { "grad_norm": 0.17120401561260223, "learning_rate": 9.009687812258808e-05, "loss": 0.0136, "step": 19490 }, { "grad_norm": 0.19811135530471802, "learning_rate": 9.00845272489839e-05, "loss": 0.0121, "step": 19500 }, { "grad_norm": 0.13485828042030334, "learning_rate": 9.007216952603031e-05, "loss": 0.0173, "step": 19510 }, { "grad_norm": 0.1830756962299347, "learning_rate": 9.005980495583894e-05, "loss": 0.0125, "step": 19520 }, { "grad_norm": 0.27303022146224976, "learning_rate": 9.004743354052252e-05, "loss": 0.0143, "step": 19530 }, { "grad_norm": 0.1960548609495163, "learning_rate": 9.003505528219503e-05, "loss": 0.0142, "step": 19540 }, { "grad_norm": 0.24445226788520813, "learning_rate": 9.002267018297154e-05, "loss": 0.0166, "step": 19550 }, { "grad_norm": 0.1938912719488144, "learning_rate": 9.001027824496834e-05, "loss": 0.0138, "step": 19560 }, { "grad_norm": 0.18362317979335785, "learning_rate": 8.999787947030287e-05, "loss": 0.0148, "step": 19570 }, { "grad_norm": 0.17016476392745972, "learning_rate": 8.998547386109376e-05, "loss": 0.0117, "step": 19580 }, { "grad_norm": 0.29697486758232117, "learning_rate": 8.997306141946073e-05, "loss": 0.015, "step": 19590 }, { "grad_norm": 0.21898868680000305, "learning_rate": 8.996064214752481e-05, "loss": 0.0155, "step": 19600 }, { "grad_norm": 0.1620524525642395, "learning_rate": 8.994821604740806e-05, "loss": 0.0125, "step": 19610 }, { "grad_norm": 0.2004169225692749, "learning_rate": 8.993578312123377e-05, "loss": 0.0155, "step": 19620 }, { "grad_norm": 0.188628688454628, "learning_rate": 8.992334337112639e-05, "loss": 0.0125, "step": 19630 }, { "grad_norm": 0.219661682844162, "learning_rate": 8.991089679921154e-05, "loss": 0.0163, "step": 19640 }, { "grad_norm": 0.19429916143417358, "learning_rate": 8.989844340761599e-05, "loss": 0.0139, "step": 19650 }, { "grad_norm": 0.16111105680465698, "learning_rate": 8.988598319846768e-05, "loss": 0.0124, "step": 19660 }, { "grad_norm": 0.14441807568073273, "learning_rate": 8.987351617389574e-05, "loss": 0.0138, "step": 19670 }, { "grad_norm": 0.15946167707443237, "learning_rate": 8.98610423360304e-05, "loss": 0.0162, "step": 19680 }, { "grad_norm": 0.16994856297969818, "learning_rate": 8.984856168700317e-05, "loss": 0.0137, "step": 19690 }, { "grad_norm": 0.15943659842014313, "learning_rate": 8.983607422894658e-05, "loss": 0.0121, "step": 19700 }, { "grad_norm": 0.14706739783287048, "learning_rate": 8.982357996399442e-05, "loss": 0.0159, "step": 19710 }, { "grad_norm": 0.19158321619033813, "learning_rate": 8.981107889428164e-05, "loss": 0.0129, "step": 19720 }, { "grad_norm": 0.2205214649438858, "learning_rate": 8.979857102194428e-05, "loss": 0.0161, "step": 19730 }, { "grad_norm": 0.16667021811008453, "learning_rate": 8.978605634911968e-05, "loss": 0.0135, "step": 19740 }, { "grad_norm": 0.21227900683879852, "learning_rate": 8.977353487794616e-05, "loss": 0.0132, "step": 19750 }, { "grad_norm": 0.1634090095758438, "learning_rate": 8.976100661056334e-05, "loss": 0.0117, "step": 19760 }, { "grad_norm": 0.2387167364358902, "learning_rate": 8.974847154911197e-05, "loss": 0.0167, "step": 19770 }, { "grad_norm": 0.20993822813034058, "learning_rate": 8.973592969573393e-05, "loss": 0.0134, "step": 19780 }, { "grad_norm": 0.15226568281650543, "learning_rate": 8.972338105257228e-05, "loss": 0.0137, "step": 19790 }, { "grad_norm": 0.10034973919391632, "learning_rate": 8.971082562177125e-05, "loss": 0.0112, "step": 19800 }, { "grad_norm": 0.16783136129379272, "learning_rate": 8.96982634054762e-05, "loss": 0.0124, "step": 19810 }, { "grad_norm": 0.1986105889081955, "learning_rate": 8.96856944058337e-05, "loss": 0.0136, "step": 19820 }, { "grad_norm": 0.1709800511598587, "learning_rate": 8.967311862499144e-05, "loss": 0.0149, "step": 19830 }, { "grad_norm": 0.15480056405067444, "learning_rate": 8.966053606509825e-05, "loss": 0.0146, "step": 19840 }, { "grad_norm": 0.14363107085227966, "learning_rate": 8.964794672830417e-05, "loss": 0.0123, "step": 19850 }, { "grad_norm": 0.31112977862358093, "learning_rate": 8.963535061676038e-05, "loss": 0.0164, "step": 19860 }, { "grad_norm": 0.167728990316391, "learning_rate": 8.962274773261918e-05, "loss": 0.0173, "step": 19870 }, { "grad_norm": 0.24918238818645477, "learning_rate": 8.961013807803409e-05, "loss": 0.0177, "step": 19880 }, { "grad_norm": 0.19289280474185944, "learning_rate": 8.959752165515973e-05, "loss": 0.0154, "step": 19890 }, { "grad_norm": 0.21943894028663635, "learning_rate": 8.958489846615193e-05, "loss": 0.0153, "step": 19900 }, { "grad_norm": 0.23696449398994446, "learning_rate": 8.957226851316762e-05, "loss": 0.0183, "step": 19910 }, { "grad_norm": 0.16602608561515808, "learning_rate": 8.955963179836493e-05, "loss": 0.0132, "step": 19920 }, { "grad_norm": 0.1788434088230133, "learning_rate": 8.954698832390312e-05, "loss": 0.0133, "step": 19930 }, { "grad_norm": 0.18610110878944397, "learning_rate": 8.953433809194263e-05, "loss": 0.0137, "step": 19940 }, { "grad_norm": 0.1312485635280609, "learning_rate": 8.9521681104645e-05, "loss": 0.0121, "step": 19950 }, { "grad_norm": 0.18096153438091278, "learning_rate": 8.9509017364173e-05, "loss": 0.015, "step": 19960 }, { "grad_norm": 0.16604511439800262, "learning_rate": 8.949634687269052e-05, "loss": 0.0156, "step": 19970 }, { "grad_norm": 0.19101987779140472, "learning_rate": 8.948366963236259e-05, "loss": 0.0116, "step": 19980 }, { "grad_norm": 0.15930043160915375, "learning_rate": 8.947098564535538e-05, "loss": 0.0123, "step": 19990 }, { "grad_norm": 0.1877911388874054, "learning_rate": 8.945829491383627e-05, "loss": 0.0145, "step": 20000 }, { "grad_norm": 0.23551002144813538, "learning_rate": 8.944559743997374e-05, "loss": 0.0165, "step": 20010 }, { "grad_norm": 0.16486108303070068, "learning_rate": 8.943289322593746e-05, "loss": 0.0145, "step": 20020 }, { "grad_norm": 0.1956585943698883, "learning_rate": 8.942018227389821e-05, "loss": 0.0147, "step": 20030 }, { "grad_norm": 0.16870133578777313, "learning_rate": 8.940746458602795e-05, "loss": 0.0127, "step": 20040 }, { "grad_norm": 0.19059798121452332, "learning_rate": 8.939474016449979e-05, "loss": 0.0149, "step": 20050 }, { "grad_norm": 0.1918209344148636, "learning_rate": 8.938200901148799e-05, "loss": 0.0143, "step": 20060 }, { "grad_norm": 0.15293142199516296, "learning_rate": 8.936927112916795e-05, "loss": 0.0153, "step": 20070 }, { "grad_norm": 0.17285791039466858, "learning_rate": 8.935652651971622e-05, "loss": 0.0132, "step": 20080 }, { "grad_norm": 0.24250538647174835, "learning_rate": 8.934377518531052e-05, "loss": 0.0148, "step": 20090 }, { "grad_norm": 0.2000247687101364, "learning_rate": 8.933101712812967e-05, "loss": 0.0181, "step": 20100 }, { "grad_norm": 0.18851235508918762, "learning_rate": 8.931825235035374e-05, "loss": 0.0167, "step": 20110 }, { "grad_norm": 0.2679937481880188, "learning_rate": 8.930548085416382e-05, "loss": 0.0169, "step": 20120 }, { "grad_norm": 0.26009276509284973, "learning_rate": 8.92927026417422e-05, "loss": 0.0154, "step": 20130 }, { "grad_norm": 0.21014812588691711, "learning_rate": 8.92799177152724e-05, "loss": 0.014, "step": 20140 }, { "grad_norm": 0.22462187707424164, "learning_rate": 8.926712607693895e-05, "loss": 0.0167, "step": 20150 }, { "grad_norm": 0.20635633170604706, "learning_rate": 8.925432772892762e-05, "loss": 0.0149, "step": 20160 }, { "grad_norm": 0.14449088275432587, "learning_rate": 8.924152267342529e-05, "loss": 0.0118, "step": 20170 }, { "grad_norm": 0.13717393577098846, "learning_rate": 8.922871091261998e-05, "loss": 0.0128, "step": 20180 }, { "grad_norm": 0.19667914509773254, "learning_rate": 8.92158924487009e-05, "loss": 0.0142, "step": 20190 }, { "grad_norm": 0.14727506041526794, "learning_rate": 8.920306728385834e-05, "loss": 0.0126, "step": 20200 }, { "grad_norm": 0.1849784553050995, "learning_rate": 8.919023542028379e-05, "loss": 0.0135, "step": 20210 }, { "grad_norm": 0.1957162618637085, "learning_rate": 8.917739686016988e-05, "loss": 0.0168, "step": 20220 }, { "grad_norm": 0.15358717739582062, "learning_rate": 8.916455160571033e-05, "loss": 0.0133, "step": 20230 }, { "grad_norm": 0.22936122119426727, "learning_rate": 8.915169965910008e-05, "loss": 0.0143, "step": 20240 }, { "grad_norm": 0.2225513607263565, "learning_rate": 8.913884102253514e-05, "loss": 0.0161, "step": 20250 }, { "grad_norm": 0.19936934113502502, "learning_rate": 8.912597569821273e-05, "loss": 0.0165, "step": 20260 }, { "grad_norm": 0.25727808475494385, "learning_rate": 8.911310368833118e-05, "loss": 0.0127, "step": 20270 }, { "grad_norm": 0.20081132650375366, "learning_rate": 8.910022499508994e-05, "loss": 0.0152, "step": 20280 }, { "grad_norm": 0.20668824017047882, "learning_rate": 8.908733962068965e-05, "loss": 0.0144, "step": 20290 }, { "grad_norm": 0.1735188215970993, "learning_rate": 8.907444756733207e-05, "loss": 0.0128, "step": 20300 }, { "grad_norm": 0.1584666222333908, "learning_rate": 8.906154883722006e-05, "loss": 0.0171, "step": 20310 }, { "grad_norm": 0.20050817728042603, "learning_rate": 8.904864343255773e-05, "loss": 0.0136, "step": 20320 }, { "grad_norm": 0.16416358947753906, "learning_rate": 8.90357313555502e-05, "loss": 0.0138, "step": 20330 }, { "grad_norm": 0.21169880032539368, "learning_rate": 8.90228126084038e-05, "loss": 0.0138, "step": 20340 }, { "grad_norm": 0.2007894366979599, "learning_rate": 8.900988719332601e-05, "loss": 0.0148, "step": 20350 }, { "grad_norm": 0.20269738137722015, "learning_rate": 8.899695511252542e-05, "loss": 0.0139, "step": 20360 }, { "grad_norm": 0.2067916840314865, "learning_rate": 8.898401636821176e-05, "loss": 0.0153, "step": 20370 }, { "grad_norm": 0.20476116240024567, "learning_rate": 8.897107096259593e-05, "loss": 0.0168, "step": 20380 }, { "grad_norm": 0.17714093625545502, "learning_rate": 8.895811889788994e-05, "loss": 0.017, "step": 20390 }, { "grad_norm": 0.13612239062786102, "learning_rate": 8.894516017630692e-05, "loss": 0.0151, "step": 20400 }, { "grad_norm": 0.16428551077842712, "learning_rate": 8.893219480006118e-05, "loss": 0.0136, "step": 20410 }, { "grad_norm": 0.19970853626728058, "learning_rate": 8.891922277136817e-05, "loss": 0.0104, "step": 20420 }, { "grad_norm": 0.17300252616405487, "learning_rate": 8.890624409244441e-05, "loss": 0.0114, "step": 20430 }, { "grad_norm": 0.15357810258865356, "learning_rate": 8.889325876550763e-05, "loss": 0.0147, "step": 20440 }, { "grad_norm": 0.22307996451854706, "learning_rate": 8.888026679277666e-05, "loss": 0.0129, "step": 20450 }, { "grad_norm": 0.19527994096279144, "learning_rate": 8.886726817647147e-05, "loss": 0.0144, "step": 20460 }, { "grad_norm": 0.167945995926857, "learning_rate": 8.885426291881319e-05, "loss": 0.0141, "step": 20470 }, { "grad_norm": 0.1610669046640396, "learning_rate": 8.884125102202401e-05, "loss": 0.0154, "step": 20480 }, { "grad_norm": 0.18235866725444794, "learning_rate": 8.882823248832736e-05, "loss": 0.0166, "step": 20490 }, { "grad_norm": 0.14463543891906738, "learning_rate": 8.881520731994772e-05, "loss": 0.014, "step": 20500 }, { "grad_norm": 0.1956208348274231, "learning_rate": 8.880217551911077e-05, "loss": 0.0121, "step": 20510 }, { "grad_norm": 0.2567012310028076, "learning_rate": 8.878913708804323e-05, "loss": 0.0167, "step": 20520 }, { "grad_norm": 0.1460028737783432, "learning_rate": 8.877609202897308e-05, "loss": 0.0161, "step": 20530 }, { "grad_norm": 0.14275281131267548, "learning_rate": 8.876304034412933e-05, "loss": 0.0107, "step": 20540 }, { "grad_norm": 0.2011106312274933, "learning_rate": 8.874998203574214e-05, "loss": 0.0139, "step": 20550 }, { "grad_norm": 0.24639460444450378, "learning_rate": 8.873691710604284e-05, "loss": 0.0163, "step": 20560 }, { "grad_norm": 0.19705763459205627, "learning_rate": 8.872384555726387e-05, "loss": 0.0135, "step": 20570 }, { "grad_norm": 0.20936641097068787, "learning_rate": 8.871076739163878e-05, "loss": 0.015, "step": 20580 }, { "grad_norm": 0.20264239609241486, "learning_rate": 8.86976826114023e-05, "loss": 0.0146, "step": 20590 }, { "grad_norm": 0.1541741043329239, "learning_rate": 8.868459121879023e-05, "loss": 0.0125, "step": 20600 }, { "grad_norm": 0.20354047417640686, "learning_rate": 8.867149321603956e-05, "loss": 0.013, "step": 20610 }, { "grad_norm": 0.13500240445137024, "learning_rate": 8.865838860538835e-05, "loss": 0.0113, "step": 20620 }, { "grad_norm": 0.18945254385471344, "learning_rate": 8.864527738907585e-05, "loss": 0.0132, "step": 20630 }, { "grad_norm": 0.17377211153507233, "learning_rate": 8.863215956934239e-05, "loss": 0.013, "step": 20640 }, { "grad_norm": 0.21109245717525482, "learning_rate": 8.861903514842947e-05, "loss": 0.0143, "step": 20650 }, { "grad_norm": 0.1415761113166809, "learning_rate": 8.860590412857966e-05, "loss": 0.0163, "step": 20660 }, { "grad_norm": 0.20306454598903656, "learning_rate": 8.85927665120367e-05, "loss": 0.0145, "step": 20670 }, { "grad_norm": 0.1723753660917282, "learning_rate": 8.857962230104546e-05, "loss": 0.014, "step": 20680 }, { "grad_norm": 0.19185440242290497, "learning_rate": 8.856647149785193e-05, "loss": 0.014, "step": 20690 }, { "grad_norm": 0.1971760243177414, "learning_rate": 8.855331410470322e-05, "loss": 0.0136, "step": 20700 }, { "grad_norm": 0.16893786191940308, "learning_rate": 8.854015012384756e-05, "loss": 0.0134, "step": 20710 }, { "grad_norm": 0.15642224252223969, "learning_rate": 8.852697955753433e-05, "loss": 0.014, "step": 20720 }, { "grad_norm": 0.22082823514938354, "learning_rate": 8.851380240801399e-05, "loss": 0.016, "step": 20730 }, { "grad_norm": 0.22899819910526276, "learning_rate": 8.850061867753818e-05, "loss": 0.0171, "step": 20740 }, { "grad_norm": 0.19500166177749634, "learning_rate": 8.848742836835963e-05, "loss": 0.0142, "step": 20750 }, { "grad_norm": 0.22286668419837952, "learning_rate": 8.847423148273221e-05, "loss": 0.0151, "step": 20760 }, { "grad_norm": 0.19439052045345306, "learning_rate": 8.846102802291092e-05, "loss": 0.0175, "step": 20770 }, { "grad_norm": 0.15441979467868805, "learning_rate": 8.844781799115183e-05, "loss": 0.0107, "step": 20780 }, { "grad_norm": 0.2064681500196457, "learning_rate": 8.84346013897122e-05, "loss": 0.0127, "step": 20790 }, { "grad_norm": 0.20197845995426178, "learning_rate": 8.842137822085038e-05, "loss": 0.0128, "step": 20800 }, { "grad_norm": 0.2082214206457138, "learning_rate": 8.840814848682585e-05, "loss": 0.0113, "step": 20810 }, { "grad_norm": 0.16382725536823273, "learning_rate": 8.83949121898992e-05, "loss": 0.0131, "step": 20820 }, { "grad_norm": 0.14237327873706818, "learning_rate": 8.838166933233217e-05, "loss": 0.0137, "step": 20830 }, { "grad_norm": 0.14401336014270782, "learning_rate": 8.83684199163876e-05, "loss": 0.0116, "step": 20840 }, { "grad_norm": 0.1864144504070282, "learning_rate": 8.835516394432943e-05, "loss": 0.0134, "step": 20850 }, { "grad_norm": 0.18912188708782196, "learning_rate": 8.834190141842276e-05, "loss": 0.015, "step": 20860 }, { "grad_norm": 0.20153836905956268, "learning_rate": 8.83286323409338e-05, "loss": 0.0138, "step": 20870 }, { "grad_norm": 0.20829899609088898, "learning_rate": 8.831535671412986e-05, "loss": 0.0134, "step": 20880 }, { "grad_norm": 0.13502764701843262, "learning_rate": 8.830207454027938e-05, "loss": 0.0132, "step": 20890 }, { "grad_norm": 0.172804594039917, "learning_rate": 8.828878582165192e-05, "loss": 0.0109, "step": 20900 }, { "grad_norm": 0.18064400553703308, "learning_rate": 8.827549056051818e-05, "loss": 0.0133, "step": 20910 }, { "grad_norm": 0.20235341787338257, "learning_rate": 8.826218875914993e-05, "loss": 0.0127, "step": 20920 }, { "grad_norm": 0.18187105655670166, "learning_rate": 8.82488804198201e-05, "loss": 0.0116, "step": 20930 }, { "grad_norm": 0.2131505310535431, "learning_rate": 8.82355655448027e-05, "loss": 0.0163, "step": 20940 }, { "grad_norm": 0.17596104741096497, "learning_rate": 8.822224413637293e-05, "loss": 0.0125, "step": 20950 }, { "grad_norm": 0.18207398056983948, "learning_rate": 8.820891619680697e-05, "loss": 0.0128, "step": 20960 }, { "grad_norm": 0.1597854048013687, "learning_rate": 8.819558172838227e-05, "loss": 0.0122, "step": 20970 }, { "grad_norm": 0.17906001210212708, "learning_rate": 8.818224073337731e-05, "loss": 0.013, "step": 20980 }, { "grad_norm": 0.20311617851257324, "learning_rate": 8.816889321407169e-05, "loss": 0.0117, "step": 20990 }, { "grad_norm": 0.15109920501708984, "learning_rate": 8.815553917274615e-05, "loss": 0.0158, "step": 21000 }, { "grad_norm": 0.2529934346675873, "learning_rate": 8.81421786116825e-05, "loss": 0.013, "step": 21010 }, { "grad_norm": 0.2219238430261612, "learning_rate": 8.812881153316373e-05, "loss": 0.0153, "step": 21020 }, { "grad_norm": 0.1686869114637375, "learning_rate": 8.81154379394739e-05, "loss": 0.0136, "step": 21030 }, { "grad_norm": 0.20717480778694153, "learning_rate": 8.810205783289818e-05, "loss": 0.0144, "step": 21040 }, { "grad_norm": 0.21501699090003967, "learning_rate": 8.808867121572286e-05, "loss": 0.017, "step": 21050 }, { "grad_norm": 0.23042654991149902, "learning_rate": 8.807527809023537e-05, "loss": 0.0125, "step": 21060 }, { "grad_norm": 0.20134757459163666, "learning_rate": 8.80618784587242e-05, "loss": 0.0166, "step": 21070 }, { "grad_norm": 0.19258520007133484, "learning_rate": 8.804847232347902e-05, "loss": 0.0147, "step": 21080 }, { "grad_norm": 0.17575304210186005, "learning_rate": 8.803505968679054e-05, "loss": 0.0139, "step": 21090 }, { "grad_norm": 0.21794597804546356, "learning_rate": 8.802164055095061e-05, "loss": 0.0139, "step": 21100 }, { "grad_norm": 0.15198978781700134, "learning_rate": 8.80082149182522e-05, "loss": 0.0162, "step": 21110 }, { "grad_norm": 0.20629525184631348, "learning_rate": 8.79947827909894e-05, "loss": 0.0124, "step": 21120 }, { "grad_norm": 0.21376711130142212, "learning_rate": 8.798134417145738e-05, "loss": 0.0148, "step": 21130 }, { "grad_norm": 0.2016480416059494, "learning_rate": 8.796789906195243e-05, "loss": 0.015, "step": 21140 }, { "grad_norm": 0.16365109384059906, "learning_rate": 8.795444746477195e-05, "loss": 0.0157, "step": 21150 }, { "grad_norm": 0.17769980430603027, "learning_rate": 8.794098938221446e-05, "loss": 0.014, "step": 21160 }, { "grad_norm": 0.12367560714483261, "learning_rate": 8.792752481657957e-05, "loss": 0.0135, "step": 21170 }, { "grad_norm": 0.2009112387895584, "learning_rate": 8.791405377016802e-05, "loss": 0.0139, "step": 21180 }, { "grad_norm": 0.19796133041381836, "learning_rate": 8.790057624528163e-05, "loss": 0.0158, "step": 21190 }, { "grad_norm": 0.2036760151386261, "learning_rate": 8.788709224422333e-05, "loss": 0.0147, "step": 21200 }, { "grad_norm": 0.17943094670772552, "learning_rate": 8.787360176929717e-05, "loss": 0.0123, "step": 21210 }, { "grad_norm": 0.21004058420658112, "learning_rate": 8.786010482280834e-05, "loss": 0.0136, "step": 21220 }, { "grad_norm": 0.17432600259780884, "learning_rate": 8.784660140706306e-05, "loss": 0.0152, "step": 21230 }, { "grad_norm": 0.160610631108284, "learning_rate": 8.783309152436872e-05, "loss": 0.0132, "step": 21240 }, { "grad_norm": 0.18314044177532196, "learning_rate": 8.781957517703375e-05, "loss": 0.0139, "step": 21250 }, { "grad_norm": 0.1745738536119461, "learning_rate": 8.780605236736776e-05, "loss": 0.013, "step": 21260 }, { "grad_norm": 0.2272973507642746, "learning_rate": 8.779252309768142e-05, "loss": 0.0161, "step": 21270 }, { "grad_norm": 0.2032705545425415, "learning_rate": 8.777898737028652e-05, "loss": 0.0136, "step": 21280 }, { "grad_norm": 0.21564361453056335, "learning_rate": 8.776544518749591e-05, "loss": 0.015, "step": 21290 }, { "grad_norm": 0.2488187849521637, "learning_rate": 8.775189655162364e-05, "loss": 0.0132, "step": 21300 }, { "grad_norm": 0.2255420982837677, "learning_rate": 8.773834146498474e-05, "loss": 0.0175, "step": 21310 }, { "grad_norm": 0.18087494373321533, "learning_rate": 8.772477992989545e-05, "loss": 0.0149, "step": 21320 }, { "grad_norm": 0.1391696035861969, "learning_rate": 8.771121194867304e-05, "loss": 0.0101, "step": 21330 }, { "grad_norm": 0.17494361102581024, "learning_rate": 8.769763752363589e-05, "loss": 0.0145, "step": 21340 }, { "grad_norm": 0.19801610708236694, "learning_rate": 8.768405665710352e-05, "loss": 0.0139, "step": 21350 }, { "grad_norm": 0.1937042474746704, "learning_rate": 8.767046935139655e-05, "loss": 0.0132, "step": 21360 }, { "grad_norm": 0.16834422945976257, "learning_rate": 8.765687560883666e-05, "loss": 0.0123, "step": 21370 }, { "grad_norm": 0.19903616607189178, "learning_rate": 8.764327543174664e-05, "loss": 0.0113, "step": 21380 }, { "grad_norm": 0.17760290205478668, "learning_rate": 8.762966882245038e-05, "loss": 0.0141, "step": 21390 }, { "grad_norm": 0.18874740600585938, "learning_rate": 8.761605578327291e-05, "loss": 0.0124, "step": 21400 }, { "grad_norm": 0.17223365604877472, "learning_rate": 8.76024363165403e-05, "loss": 0.0136, "step": 21410 }, { "grad_norm": 0.16979412734508514, "learning_rate": 8.758881042457976e-05, "loss": 0.014, "step": 21420 }, { "grad_norm": 0.18732033669948578, "learning_rate": 8.757517810971957e-05, "loss": 0.0124, "step": 21430 }, { "grad_norm": 0.2244427651166916, "learning_rate": 8.756153937428913e-05, "loss": 0.0142, "step": 21440 }, { "grad_norm": 0.17683863639831543, "learning_rate": 8.754789422061889e-05, "loss": 0.0168, "step": 21450 }, { "grad_norm": 0.15133114159107208, "learning_rate": 8.753424265104052e-05, "loss": 0.0126, "step": 21460 }, { "grad_norm": 0.15442988276481628, "learning_rate": 8.752058466788659e-05, "loss": 0.0115, "step": 21470 }, { "grad_norm": 0.1998477727174759, "learning_rate": 8.750692027349097e-05, "loss": 0.0121, "step": 21480 }, { "grad_norm": 0.1720661073923111, "learning_rate": 8.749324947018847e-05, "loss": 0.0133, "step": 21490 }, { "grad_norm": 0.16796745359897614, "learning_rate": 8.747957226031507e-05, "loss": 0.0144, "step": 21500 }, { "grad_norm": 0.11605538427829742, "learning_rate": 8.746588864620787e-05, "loss": 0.0106, "step": 21510 }, { "grad_norm": 0.18860796093940735, "learning_rate": 8.745219863020498e-05, "loss": 0.0132, "step": 21520 }, { "grad_norm": 0.12659330666065216, "learning_rate": 8.743850221464564e-05, "loss": 0.0129, "step": 21530 }, { "grad_norm": 0.21053585410118103, "learning_rate": 8.742479940187026e-05, "loss": 0.0138, "step": 21540 }, { "grad_norm": 0.1711157262325287, "learning_rate": 8.74110901942202e-05, "loss": 0.013, "step": 21550 }, { "grad_norm": 0.1723143607378006, "learning_rate": 8.739737459403803e-05, "loss": 0.013, "step": 21560 }, { "grad_norm": 0.1812308430671692, "learning_rate": 8.738365260366737e-05, "loss": 0.011, "step": 21570 }, { "grad_norm": 0.19529083371162415, "learning_rate": 8.736992422545292e-05, "loss": 0.0155, "step": 21580 }, { "grad_norm": 0.20311647653579712, "learning_rate": 8.73561894617405e-05, "loss": 0.0144, "step": 21590 }, { "grad_norm": 0.14721928536891937, "learning_rate": 8.734244831487697e-05, "loss": 0.016, "step": 21600 }, { "grad_norm": 0.18780724704265594, "learning_rate": 8.732870078721035e-05, "loss": 0.0138, "step": 21610 }, { "grad_norm": 0.1484299749135971, "learning_rate": 8.731494688108972e-05, "loss": 0.0139, "step": 21620 }, { "grad_norm": 0.13511373102664948, "learning_rate": 8.730118659886523e-05, "loss": 0.0137, "step": 21630 }, { "grad_norm": 0.15422889590263367, "learning_rate": 8.728741994288814e-05, "loss": 0.0135, "step": 21640 }, { "grad_norm": 0.18232791125774384, "learning_rate": 8.727364691551079e-05, "loss": 0.012, "step": 21650 }, { "grad_norm": 0.20752371847629547, "learning_rate": 8.725986751908661e-05, "loss": 0.014, "step": 21660 }, { "grad_norm": 0.20618990063667297, "learning_rate": 8.724608175597016e-05, "loss": 0.0139, "step": 21670 }, { "grad_norm": 0.22931991517543793, "learning_rate": 8.723228962851699e-05, "loss": 0.0145, "step": 21680 }, { "grad_norm": 0.20164726674556732, "learning_rate": 8.721849113908385e-05, "loss": 0.0132, "step": 21690 }, { "grad_norm": 0.16540850698947906, "learning_rate": 8.720468629002848e-05, "loss": 0.0147, "step": 21700 }, { "grad_norm": 0.18725700676441193, "learning_rate": 8.719087508370978e-05, "loss": 0.0143, "step": 21710 }, { "grad_norm": 0.18815235793590546, "learning_rate": 8.717705752248772e-05, "loss": 0.0197, "step": 21720 }, { "grad_norm": 0.1288663148880005, "learning_rate": 8.71632336087233e-05, "loss": 0.0155, "step": 21730 }, { "grad_norm": 0.22917477786540985, "learning_rate": 8.71494033447787e-05, "loss": 0.0123, "step": 21740 }, { "grad_norm": 0.12078762799501419, "learning_rate": 8.713556673301708e-05, "loss": 0.0141, "step": 21750 }, { "grad_norm": 0.14764097332954407, "learning_rate": 8.712172377580278e-05, "loss": 0.0142, "step": 21760 }, { "grad_norm": 0.14192430675029755, "learning_rate": 8.710787447550114e-05, "loss": 0.0166, "step": 21770 }, { "grad_norm": 0.2429364025592804, "learning_rate": 8.70940188344787e-05, "loss": 0.0143, "step": 21780 }, { "grad_norm": 0.16676735877990723, "learning_rate": 8.708015685510293e-05, "loss": 0.0125, "step": 21790 }, { "grad_norm": 0.19703519344329834, "learning_rate": 8.706628853974252e-05, "loss": 0.0133, "step": 21800 }, { "grad_norm": 0.1417253315448761, "learning_rate": 8.705241389076715e-05, "loss": 0.0138, "step": 21810 }, { "grad_norm": 0.19405460357666016, "learning_rate": 8.703853291054764e-05, "loss": 0.0116, "step": 21820 }, { "grad_norm": 0.2373807430267334, "learning_rate": 8.702464560145587e-05, "loss": 0.0129, "step": 21830 }, { "grad_norm": 0.15047091245651245, "learning_rate": 8.701075196586476e-05, "loss": 0.0133, "step": 21840 }, { "grad_norm": 0.221115842461586, "learning_rate": 8.699685200614842e-05, "loss": 0.0163, "step": 21850 }, { "grad_norm": 0.15043854713439941, "learning_rate": 8.698294572468193e-05, "loss": 0.0138, "step": 21860 }, { "grad_norm": 0.19126668572425842, "learning_rate": 8.696903312384148e-05, "loss": 0.0153, "step": 21870 }, { "grad_norm": 0.16443312168121338, "learning_rate": 8.695511420600439e-05, "loss": 0.0128, "step": 21880 }, { "grad_norm": 0.19428576529026031, "learning_rate": 8.694118897354901e-05, "loss": 0.0136, "step": 21890 }, { "grad_norm": 0.17762432992458344, "learning_rate": 8.692725742885478e-05, "loss": 0.0144, "step": 21900 }, { "grad_norm": 0.14237737655639648, "learning_rate": 8.691331957430221e-05, "loss": 0.0137, "step": 21910 }, { "grad_norm": 0.2585836350917816, "learning_rate": 8.68993754122729e-05, "loss": 0.0176, "step": 21920 }, { "grad_norm": 0.2334396243095398, "learning_rate": 8.688542494514955e-05, "loss": 0.0171, "step": 21930 }, { "grad_norm": 0.1267423778772354, "learning_rate": 8.68714681753159e-05, "loss": 0.0161, "step": 21940 }, { "grad_norm": 0.18813049793243408, "learning_rate": 8.685750510515676e-05, "loss": 0.0122, "step": 21950 }, { "grad_norm": 0.17035603523254395, "learning_rate": 8.684353573705805e-05, "loss": 0.0134, "step": 21960 }, { "grad_norm": 0.1561555415391922, "learning_rate": 8.682956007340677e-05, "loss": 0.0158, "step": 21970 }, { "grad_norm": 0.21205958724021912, "learning_rate": 8.681557811659095e-05, "loss": 0.017, "step": 21980 }, { "grad_norm": 0.183713898062706, "learning_rate": 8.680158986899974e-05, "loss": 0.0148, "step": 21990 }, { "grad_norm": 0.24328270554542542, "learning_rate": 8.678759533302335e-05, "loss": 0.0114, "step": 22000 }, { "grad_norm": 0.17863787710666656, "learning_rate": 8.677359451105308e-05, "loss": 0.0136, "step": 22010 }, { "grad_norm": 0.21141280233860016, "learning_rate": 8.675958740548123e-05, "loss": 0.0146, "step": 22020 }, { "grad_norm": 0.20107771456241608, "learning_rate": 8.674557401870129e-05, "loss": 0.0133, "step": 22030 }, { "grad_norm": 0.1866605430841446, "learning_rate": 8.673155435310775e-05, "loss": 0.014, "step": 22040 }, { "grad_norm": 0.1920078545808792, "learning_rate": 8.671752841109617e-05, "loss": 0.013, "step": 22050 }, { "grad_norm": 0.1801520437002182, "learning_rate": 8.670349619506321e-05, "loss": 0.0127, "step": 22060 }, { "grad_norm": 0.235622838139534, "learning_rate": 8.66894577074066e-05, "loss": 0.0151, "step": 22070 }, { "grad_norm": 0.19780336320400238, "learning_rate": 8.667541295052513e-05, "loss": 0.0136, "step": 22080 }, { "grad_norm": 0.2725517749786377, "learning_rate": 8.666136192681865e-05, "loss": 0.0146, "step": 22090 }, { "grad_norm": 0.15944363176822662, "learning_rate": 8.664730463868811e-05, "loss": 0.017, "step": 22100 }, { "grad_norm": 0.18049658834934235, "learning_rate": 8.663324108853552e-05, "loss": 0.0134, "step": 22110 }, { "grad_norm": 0.15724274516105652, "learning_rate": 8.661917127876395e-05, "loss": 0.0142, "step": 22120 }, { "grad_norm": 0.19228105247020721, "learning_rate": 8.660509521177754e-05, "loss": 0.0171, "step": 22130 }, { "grad_norm": 0.17795680463314056, "learning_rate": 8.65910128899815e-05, "loss": 0.0133, "step": 22140 }, { "grad_norm": 0.1508927196264267, "learning_rate": 8.657692431578214e-05, "loss": 0.0129, "step": 22150 }, { "grad_norm": 0.1917557716369629, "learning_rate": 8.656282949158679e-05, "loss": 0.0113, "step": 22160 }, { "grad_norm": 0.2288837432861328, "learning_rate": 8.654872841980388e-05, "loss": 0.0129, "step": 22170 }, { "grad_norm": 0.17607107758522034, "learning_rate": 8.653462110284289e-05, "loss": 0.0139, "step": 22180 }, { "grad_norm": 0.19141936302185059, "learning_rate": 8.652050754311437e-05, "loss": 0.0129, "step": 22190 }, { "grad_norm": 0.16392646729946136, "learning_rate": 8.650638774302995e-05, "loss": 0.0145, "step": 22200 }, { "grad_norm": 0.1330307424068451, "learning_rate": 8.649226170500233e-05, "loss": 0.0135, "step": 22210 }, { "grad_norm": 0.15660201013088226, "learning_rate": 8.647812943144524e-05, "loss": 0.0153, "step": 22220 }, { "grad_norm": 0.2243049442768097, "learning_rate": 8.646399092477351e-05, "loss": 0.0138, "step": 22230 }, { "grad_norm": 0.16792233288288116, "learning_rate": 8.644984618740301e-05, "loss": 0.015, "step": 22240 }, { "grad_norm": 0.1804889589548111, "learning_rate": 8.643569522175073e-05, "loss": 0.0182, "step": 22250 }, { "grad_norm": 0.1979459673166275, "learning_rate": 8.642153803023463e-05, "loss": 0.0131, "step": 22260 }, { "grad_norm": 0.1398697942495346, "learning_rate": 8.640737461527383e-05, "loss": 0.0115, "step": 22270 }, { "grad_norm": 0.16490058600902557, "learning_rate": 8.639320497928845e-05, "loss": 0.0125, "step": 22280 }, { "grad_norm": 0.18482857942581177, "learning_rate": 8.637902912469969e-05, "loss": 0.0137, "step": 22290 }, { "grad_norm": 0.1692296266555786, "learning_rate": 8.636484705392982e-05, "loss": 0.0114, "step": 22300 }, { "grad_norm": 0.13989490270614624, "learning_rate": 8.635065876940216e-05, "loss": 0.0133, "step": 22310 }, { "grad_norm": 0.16160671412944794, "learning_rate": 8.633646427354112e-05, "loss": 0.0114, "step": 22320 }, { "grad_norm": 0.13742682337760925, "learning_rate": 8.632226356877213e-05, "loss": 0.0153, "step": 22330 }, { "grad_norm": 0.22607450187206268, "learning_rate": 8.630805665752173e-05, "loss": 0.0127, "step": 22340 }, { "grad_norm": 0.19273777306079865, "learning_rate": 8.629384354221748e-05, "loss": 0.0121, "step": 22350 }, { "grad_norm": 0.1542445868253708, "learning_rate": 8.627962422528797e-05, "loss": 0.012, "step": 22360 }, { "grad_norm": 0.11744873225688934, "learning_rate": 8.626539870916296e-05, "loss": 0.0129, "step": 22370 }, { "grad_norm": 0.1614408940076828, "learning_rate": 8.625116699627317e-05, "loss": 0.0116, "step": 22380 }, { "grad_norm": 0.19692577421665192, "learning_rate": 8.623692908905041e-05, "loss": 0.0143, "step": 22390 }, { "grad_norm": 0.14395707845687866, "learning_rate": 8.622268498992755e-05, "loss": 0.0132, "step": 22400 }, { "grad_norm": 0.1678454577922821, "learning_rate": 8.620843470133851e-05, "loss": 0.0124, "step": 22410 }, { "grad_norm": 0.15484729409217834, "learning_rate": 8.619417822571829e-05, "loss": 0.0152, "step": 22420 }, { "grad_norm": 0.2009819597005844, "learning_rate": 8.617991556550292e-05, "loss": 0.0143, "step": 22430 }, { "grad_norm": 0.17382150888442993, "learning_rate": 8.616564672312952e-05, "loss": 0.0147, "step": 22440 }, { "grad_norm": 0.17096637189388275, "learning_rate": 8.61513717010362e-05, "loss": 0.0122, "step": 22450 }, { "grad_norm": 0.19060079753398895, "learning_rate": 8.613709050166221e-05, "loss": 0.0118, "step": 22460 }, { "grad_norm": 0.13884443044662476, "learning_rate": 8.61228031274478e-05, "loss": 0.0116, "step": 22470 }, { "grad_norm": 0.23602065443992615, "learning_rate": 8.610850958083431e-05, "loss": 0.011, "step": 22480 }, { "grad_norm": 0.13861413300037384, "learning_rate": 8.609420986426409e-05, "loss": 0.0116, "step": 22490 }, { "grad_norm": 0.1886831372976303, "learning_rate": 8.60799039801806e-05, "loss": 0.0129, "step": 22500 }, { "grad_norm": 0.16404502093791962, "learning_rate": 8.606559193102828e-05, "loss": 0.0132, "step": 22510 }, { "grad_norm": 0.13799221813678741, "learning_rate": 8.605127371925273e-05, "loss": 0.012, "step": 22520 }, { "grad_norm": 0.19461846351623535, "learning_rate": 8.603694934730047e-05, "loss": 0.0126, "step": 22530 }, { "grad_norm": 0.1618703156709671, "learning_rate": 8.602261881761919e-05, "loss": 0.0136, "step": 22540 }, { "grad_norm": 0.238754540681839, "learning_rate": 8.600828213265759e-05, "loss": 0.0122, "step": 22550 }, { "grad_norm": 0.2668251693248749, "learning_rate": 8.599393929486539e-05, "loss": 0.0134, "step": 22560 }, { "grad_norm": 0.15778347849845886, "learning_rate": 8.59795903066934e-05, "loss": 0.0119, "step": 22570 }, { "grad_norm": 0.17004314064979553, "learning_rate": 8.596523517059347e-05, "loss": 0.0105, "step": 22580 }, { "grad_norm": 0.1733941286802292, "learning_rate": 8.59508738890185e-05, "loss": 0.0114, "step": 22590 }, { "grad_norm": 0.18217557668685913, "learning_rate": 8.593650646442246e-05, "loss": 0.0124, "step": 22600 }, { "grad_norm": 0.15720269083976746, "learning_rate": 8.59221328992603e-05, "loss": 0.0122, "step": 22610 }, { "grad_norm": 0.16254684329032898, "learning_rate": 8.590775319598813e-05, "loss": 0.0104, "step": 22620 }, { "grad_norm": 0.140626460313797, "learning_rate": 8.589336735706301e-05, "loss": 0.0115, "step": 22630 }, { "grad_norm": 0.24633006751537323, "learning_rate": 8.587897538494307e-05, "loss": 0.0122, "step": 22640 }, { "grad_norm": 0.22266073524951935, "learning_rate": 8.586457728208756e-05, "loss": 0.0148, "step": 22650 }, { "grad_norm": 0.13156758248806, "learning_rate": 8.585017305095667e-05, "loss": 0.0126, "step": 22660 }, { "grad_norm": 0.20877358317375183, "learning_rate": 8.583576269401173e-05, "loss": 0.0168, "step": 22670 }, { "grad_norm": 0.18092380464076996, "learning_rate": 8.582134621371504e-05, "loss": 0.0134, "step": 22680 }, { "grad_norm": 0.1592414528131485, "learning_rate": 8.580692361253e-05, "loss": 0.0131, "step": 22690 }, { "grad_norm": 0.15331824123859406, "learning_rate": 8.579249489292104e-05, "loss": 0.0125, "step": 22700 }, { "grad_norm": 0.1732124239206314, "learning_rate": 8.577806005735363e-05, "loss": 0.0138, "step": 22710 }, { "grad_norm": 0.1796896904706955, "learning_rate": 8.576361910829429e-05, "loss": 0.0123, "step": 22720 }, { "grad_norm": 0.1305357664823532, "learning_rate": 8.574917204821057e-05, "loss": 0.0118, "step": 22730 }, { "grad_norm": 0.1933080404996872, "learning_rate": 8.57347188795711e-05, "loss": 0.0132, "step": 22740 }, { "grad_norm": 0.1719045341014862, "learning_rate": 8.572025960484551e-05, "loss": 0.012, "step": 22750 }, { "grad_norm": 0.15759499371051788, "learning_rate": 8.57057942265045e-05, "loss": 0.0117, "step": 22760 }, { "grad_norm": 0.20346443355083466, "learning_rate": 8.569132274701984e-05, "loss": 0.0141, "step": 22770 }, { "grad_norm": 0.19167502224445343, "learning_rate": 8.567684516886427e-05, "loss": 0.0123, "step": 22780 }, { "grad_norm": 0.1538420170545578, "learning_rate": 8.56623614945116e-05, "loss": 0.0118, "step": 22790 }, { "grad_norm": 0.1552775502204895, "learning_rate": 8.564787172643675e-05, "loss": 0.0124, "step": 22800 }, { "grad_norm": 0.1873333901166916, "learning_rate": 8.563337586711559e-05, "loss": 0.0173, "step": 22810 }, { "grad_norm": 0.19834226369857788, "learning_rate": 8.561887391902506e-05, "loss": 0.0103, "step": 22820 }, { "grad_norm": 0.16541431844234467, "learning_rate": 8.560436588464316e-05, "loss": 0.0132, "step": 22830 }, { "grad_norm": 0.17999888956546783, "learning_rate": 8.55898517664489e-05, "loss": 0.0121, "step": 22840 }, { "grad_norm": 0.24224193394184113, "learning_rate": 8.557533156692236e-05, "loss": 0.016, "step": 22850 }, { "grad_norm": 0.15775614976882935, "learning_rate": 8.556080528854467e-05, "loss": 0.0137, "step": 22860 }, { "grad_norm": 0.24523364007472992, "learning_rate": 8.554627293379791e-05, "loss": 0.0127, "step": 22870 }, { "grad_norm": 0.15855202078819275, "learning_rate": 8.553173450516531e-05, "loss": 0.0126, "step": 22880 }, { "grad_norm": 0.21019212901592255, "learning_rate": 8.551719000513108e-05, "loss": 0.0127, "step": 22890 }, { "grad_norm": 0.1846770942211151, "learning_rate": 8.550263943618049e-05, "loss": 0.0137, "step": 22900 }, { "grad_norm": 0.16992206871509552, "learning_rate": 8.54880828007998e-05, "loss": 0.0127, "step": 22910 }, { "grad_norm": 0.1812451183795929, "learning_rate": 8.547352010147637e-05, "loss": 0.0127, "step": 22920 }, { "grad_norm": 0.23415586352348328, "learning_rate": 8.545895134069855e-05, "loss": 0.0126, "step": 22930 }, { "grad_norm": 0.15500082075595856, "learning_rate": 8.544437652095576e-05, "loss": 0.0137, "step": 22940 }, { "grad_norm": 0.18724003434181213, "learning_rate": 8.542979564473843e-05, "loss": 0.0148, "step": 22950 }, { "grad_norm": 0.2570995092391968, "learning_rate": 8.541520871453802e-05, "loss": 0.0147, "step": 22960 }, { "grad_norm": 0.1844482272863388, "learning_rate": 8.540061573284705e-05, "loss": 0.0159, "step": 22970 }, { "grad_norm": 0.10335902869701385, "learning_rate": 8.538601670215906e-05, "loss": 0.0119, "step": 22980 }, { "grad_norm": 0.1808565855026245, "learning_rate": 8.537141162496864e-05, "loss": 0.011, "step": 22990 }, { "grad_norm": 0.14999745786190033, "learning_rate": 8.535680050377137e-05, "loss": 0.0132, "step": 23000 }, { "grad_norm": 0.17087168991565704, "learning_rate": 8.534218334106391e-05, "loss": 0.0127, "step": 23010 }, { "grad_norm": 0.17226862907409668, "learning_rate": 8.532756013934393e-05, "loss": 0.0123, "step": 23020 }, { "grad_norm": 0.17891356348991394, "learning_rate": 8.531293090111012e-05, "loss": 0.011, "step": 23030 }, { "grad_norm": 0.1755797266960144, "learning_rate": 8.529829562886225e-05, "loss": 0.0133, "step": 23040 }, { "grad_norm": 0.16180431842803955, "learning_rate": 8.528365432510105e-05, "loss": 0.0121, "step": 23050 }, { "grad_norm": 0.19889011979103088, "learning_rate": 8.526900699232833e-05, "loss": 0.0149, "step": 23060 }, { "grad_norm": 0.19107794761657715, "learning_rate": 8.525435363304695e-05, "loss": 0.014, "step": 23070 }, { "grad_norm": 0.1713416576385498, "learning_rate": 8.523969424976072e-05, "loss": 0.0149, "step": 23080 }, { "grad_norm": 0.18875494599342346, "learning_rate": 8.522502884497457e-05, "loss": 0.0135, "step": 23090 }, { "grad_norm": 0.16719073057174683, "learning_rate": 8.521035742119437e-05, "loss": 0.0136, "step": 23100 }, { "grad_norm": 0.17530861496925354, "learning_rate": 8.519567998092712e-05, "loss": 0.0128, "step": 23110 }, { "grad_norm": 0.18370987474918365, "learning_rate": 8.518099652668075e-05, "loss": 0.012, "step": 23120 }, { "grad_norm": 0.16376522183418274, "learning_rate": 8.516630706096429e-05, "loss": 0.0128, "step": 23130 }, { "grad_norm": 0.16895002126693726, "learning_rate": 8.515161158628773e-05, "loss": 0.0124, "step": 23140 }, { "grad_norm": 0.17145726084709167, "learning_rate": 8.513691010516216e-05, "loss": 0.0146, "step": 23150 }, { "grad_norm": 0.12499267607927322, "learning_rate": 8.512220262009966e-05, "loss": 0.0128, "step": 23160 }, { "grad_norm": 0.14156265556812286, "learning_rate": 8.510748913361332e-05, "loss": 0.0123, "step": 23170 }, { "grad_norm": 0.1360696256160736, "learning_rate": 8.509276964821726e-05, "loss": 0.0123, "step": 23180 }, { "grad_norm": 0.2456054985523224, "learning_rate": 8.507804416642669e-05, "loss": 0.0149, "step": 23190 }, { "grad_norm": 0.15525484085083008, "learning_rate": 8.506331269075774e-05, "loss": 0.0164, "step": 23200 }, { "grad_norm": 0.15374812483787537, "learning_rate": 8.504857522372765e-05, "loss": 0.0106, "step": 23210 }, { "grad_norm": 0.15176038444042206, "learning_rate": 8.503383176785461e-05, "loss": 0.0107, "step": 23220 }, { "grad_norm": 0.18518508970737457, "learning_rate": 8.501908232565792e-05, "loss": 0.0129, "step": 23230 }, { "grad_norm": 0.16407209634780884, "learning_rate": 8.50043268996578e-05, "loss": 0.0119, "step": 23240 }, { "grad_norm": 0.15514127910137177, "learning_rate": 8.498956549237562e-05, "loss": 0.0096, "step": 23250 }, { "grad_norm": 0.12237125635147095, "learning_rate": 8.497479810633366e-05, "loss": 0.0099, "step": 23260 }, { "grad_norm": 0.17084437608718872, "learning_rate": 8.496002474405525e-05, "loss": 0.0124, "step": 23270 }, { "grad_norm": 0.16478270292282104, "learning_rate": 8.494524540806478e-05, "loss": 0.0133, "step": 23280 }, { "grad_norm": 0.16181842982769012, "learning_rate": 8.493046010088761e-05, "loss": 0.0125, "step": 23290 }, { "grad_norm": 0.1665535420179367, "learning_rate": 8.491566882505018e-05, "loss": 0.0119, "step": 23300 }, { "grad_norm": 0.1466463953256607, "learning_rate": 8.490087158307988e-05, "loss": 0.012, "step": 23310 }, { "grad_norm": 0.2123657912015915, "learning_rate": 8.488606837750518e-05, "loss": 0.0135, "step": 23320 }, { "grad_norm": 0.18916819989681244, "learning_rate": 8.487125921085552e-05, "loss": 0.0118, "step": 23330 }, { "grad_norm": 0.1662645787000656, "learning_rate": 8.485644408566141e-05, "loss": 0.0106, "step": 23340 }, { "grad_norm": 0.20076234638690948, "learning_rate": 8.484162300445431e-05, "loss": 0.0131, "step": 23350 }, { "grad_norm": 0.22850646078586578, "learning_rate": 8.482679596976676e-05, "loss": 0.014, "step": 23360 }, { "grad_norm": 0.11255618929862976, "learning_rate": 8.48119629841323e-05, "loss": 0.0137, "step": 23370 }, { "grad_norm": 0.14670130610466003, "learning_rate": 8.479712405008547e-05, "loss": 0.0154, "step": 23380 }, { "grad_norm": 0.17994233965873718, "learning_rate": 8.478227917016184e-05, "loss": 0.0121, "step": 23390 }, { "grad_norm": 0.21387289464473724, "learning_rate": 8.476742834689801e-05, "loss": 0.0124, "step": 23400 }, { "grad_norm": 0.15611164271831512, "learning_rate": 8.475257158283157e-05, "loss": 0.0143, "step": 23410 }, { "grad_norm": 0.18121813237667084, "learning_rate": 8.473770888050112e-05, "loss": 0.0154, "step": 23420 }, { "grad_norm": 0.21772606670856476, "learning_rate": 8.47228402424463e-05, "loss": 0.0149, "step": 23430 }, { "grad_norm": 0.15309935808181763, "learning_rate": 8.470796567120775e-05, "loss": 0.0127, "step": 23440 }, { "grad_norm": 0.17733319103717804, "learning_rate": 8.469308516932714e-05, "loss": 0.0137, "step": 23450 }, { "grad_norm": 0.17195455729961395, "learning_rate": 8.467819873934714e-05, "loss": 0.0129, "step": 23460 }, { "grad_norm": 0.17707020044326782, "learning_rate": 8.466330638381143e-05, "loss": 0.0128, "step": 23470 }, { "grad_norm": 0.1526450365781784, "learning_rate": 8.464840810526469e-05, "loss": 0.0119, "step": 23480 }, { "grad_norm": 0.14659301936626434, "learning_rate": 8.463350390625264e-05, "loss": 0.0101, "step": 23490 }, { "grad_norm": 0.15443380177021027, "learning_rate": 8.4618593789322e-05, "loss": 0.0128, "step": 23500 }, { "grad_norm": 0.2082311362028122, "learning_rate": 8.46036777570205e-05, "loss": 0.014, "step": 23510 }, { "grad_norm": 0.21829472482204437, "learning_rate": 8.458875581189688e-05, "loss": 0.0137, "step": 23520 }, { "grad_norm": 0.19062119722366333, "learning_rate": 8.457382795650092e-05, "loss": 0.015, "step": 23530 }, { "grad_norm": 0.17215359210968018, "learning_rate": 8.455889419338335e-05, "loss": 0.0143, "step": 23540 }, { "grad_norm": 0.12800684571266174, "learning_rate": 8.454395452509593e-05, "loss": 0.0118, "step": 23550 }, { "grad_norm": 0.121414914727211, "learning_rate": 8.452900895419146e-05, "loss": 0.0095, "step": 23560 }, { "grad_norm": 0.11814800649881363, "learning_rate": 8.451405748322376e-05, "loss": 0.0109, "step": 23570 }, { "grad_norm": 0.20665563642978668, "learning_rate": 8.449910011474759e-05, "loss": 0.0159, "step": 23580 }, { "grad_norm": 0.15503761172294617, "learning_rate": 8.448413685131876e-05, "loss": 0.0124, "step": 23590 }, { "grad_norm": 0.2183321714401245, "learning_rate": 8.446916769549407e-05, "loss": 0.0116, "step": 23600 }, { "grad_norm": 0.1506820172071457, "learning_rate": 8.445419264983136e-05, "loss": 0.0109, "step": 23610 }, { "grad_norm": 0.1587565690279007, "learning_rate": 8.443921171688947e-05, "loss": 0.0102, "step": 23620 }, { "grad_norm": 0.18281061947345734, "learning_rate": 8.442422489922819e-05, "loss": 0.0122, "step": 23630 }, { "grad_norm": 0.2355986386537552, "learning_rate": 8.440923219940838e-05, "loss": 0.0126, "step": 23640 }, { "grad_norm": 0.16725817322731018, "learning_rate": 8.439423361999189e-05, "loss": 0.011, "step": 23650 }, { "grad_norm": 0.15941768884658813, "learning_rate": 8.437922916354155e-05, "loss": 0.0119, "step": 23660 }, { "grad_norm": 0.1914660930633545, "learning_rate": 8.436421883262123e-05, "loss": 0.0131, "step": 23670 }, { "grad_norm": 0.14239682257175446, "learning_rate": 8.434920262979577e-05, "loss": 0.0103, "step": 23680 }, { "grad_norm": 0.21630612015724182, "learning_rate": 8.433418055763104e-05, "loss": 0.012, "step": 23690 }, { "grad_norm": 0.19829000532627106, "learning_rate": 8.431915261869389e-05, "loss": 0.0119, "step": 23700 }, { "grad_norm": 0.1715444177389145, "learning_rate": 8.43041188155522e-05, "loss": 0.0112, "step": 23710 }, { "grad_norm": 0.1840580701828003, "learning_rate": 8.428907915077481e-05, "loss": 0.0135, "step": 23720 }, { "grad_norm": 0.1788136214017868, "learning_rate": 8.42740336269316e-05, "loss": 0.0118, "step": 23730 }, { "grad_norm": 0.24670477211475372, "learning_rate": 8.425898224659345e-05, "loss": 0.0137, "step": 23740 }, { "grad_norm": 0.13469715416431427, "learning_rate": 8.42439250123322e-05, "loss": 0.012, "step": 23750 }, { "grad_norm": 0.16600817441940308, "learning_rate": 8.422886192672076e-05, "loss": 0.0123, "step": 23760 }, { "grad_norm": 0.17609041929244995, "learning_rate": 8.421379299233297e-05, "loss": 0.0118, "step": 23770 }, { "grad_norm": 0.18216747045516968, "learning_rate": 8.419871821174371e-05, "loss": 0.0126, "step": 23780 }, { "grad_norm": 0.1691918522119522, "learning_rate": 8.418363758752884e-05, "loss": 0.0128, "step": 23790 }, { "grad_norm": 0.23155562579631805, "learning_rate": 8.416855112226523e-05, "loss": 0.0126, "step": 23800 }, { "grad_norm": 0.19439052045345306, "learning_rate": 8.415345881853075e-05, "loss": 0.0126, "step": 23810 }, { "grad_norm": 0.1965436488389969, "learning_rate": 8.413836067890426e-05, "loss": 0.0139, "step": 23820 }, { "grad_norm": 0.17093969881534576, "learning_rate": 8.41232567059656e-05, "loss": 0.0104, "step": 23830 }, { "grad_norm": 0.1537761390209198, "learning_rate": 8.410814690229565e-05, "loss": 0.0121, "step": 23840 }, { "grad_norm": 0.18512602150440216, "learning_rate": 8.409303127047626e-05, "loss": 0.0126, "step": 23850 }, { "grad_norm": 0.18862473964691162, "learning_rate": 8.407790981309028e-05, "loss": 0.0114, "step": 23860 }, { "grad_norm": 0.14118431508541107, "learning_rate": 8.406278253272153e-05, "loss": 0.0121, "step": 23870 }, { "grad_norm": 0.1663968861103058, "learning_rate": 8.404764943195487e-05, "loss": 0.0114, "step": 23880 }, { "grad_norm": 0.22616241872310638, "learning_rate": 8.403251051337613e-05, "loss": 0.0131, "step": 23890 }, { "grad_norm": 0.13350993394851685, "learning_rate": 8.401736577957214e-05, "loss": 0.0128, "step": 23900 }, { "grad_norm": 0.15627801418304443, "learning_rate": 8.40022152331307e-05, "loss": 0.0112, "step": 23910 }, { "grad_norm": 0.15854787826538086, "learning_rate": 8.398705887664064e-05, "loss": 0.0106, "step": 23920 }, { "grad_norm": 0.25100451707839966, "learning_rate": 8.397189671269177e-05, "loss": 0.0104, "step": 23930 }, { "grad_norm": 0.21275846660137177, "learning_rate": 8.395672874387488e-05, "loss": 0.0113, "step": 23940 }, { "grad_norm": 0.22453393042087555, "learning_rate": 8.394155497278177e-05, "loss": 0.0125, "step": 23950 }, { "grad_norm": 0.22481930255889893, "learning_rate": 8.392637540200523e-05, "loss": 0.0142, "step": 23960 }, { "grad_norm": 0.1393231898546219, "learning_rate": 8.391119003413902e-05, "loss": 0.0098, "step": 23970 }, { "grad_norm": 0.2189590483903885, "learning_rate": 8.38959988717779e-05, "loss": 0.0098, "step": 23980 }, { "grad_norm": 0.18416845798492432, "learning_rate": 8.388080191751764e-05, "loss": 0.0113, "step": 23990 }, { "grad_norm": 0.17294074594974518, "learning_rate": 8.386559917395496e-05, "loss": 0.0129, "step": 24000 }, { "grad_norm": 0.15423744916915894, "learning_rate": 8.385039064368761e-05, "loss": 0.0111, "step": 24010 }, { "grad_norm": 0.14612828195095062, "learning_rate": 8.383517632931431e-05, "loss": 0.0131, "step": 24020 }, { "grad_norm": 0.18599684536457062, "learning_rate": 8.381995623343477e-05, "loss": 0.0113, "step": 24030 }, { "grad_norm": 0.20397649705410004, "learning_rate": 8.380473035864968e-05, "loss": 0.0136, "step": 24040 }, { "grad_norm": 0.16004933416843414, "learning_rate": 8.378949870756076e-05, "loss": 0.0132, "step": 24050 }, { "grad_norm": 0.17264245450496674, "learning_rate": 8.377426128277063e-05, "loss": 0.0161, "step": 24060 }, { "grad_norm": 0.14444050192832947, "learning_rate": 8.375901808688298e-05, "loss": 0.0109, "step": 24070 }, { "grad_norm": 0.12326563894748688, "learning_rate": 8.374376912250246e-05, "loss": 0.0094, "step": 24080 }, { "grad_norm": 0.15168257057666779, "learning_rate": 8.372851439223468e-05, "loss": 0.013, "step": 24090 }, { "grad_norm": 0.16446900367736816, "learning_rate": 8.371325389868627e-05, "loss": 0.0111, "step": 24100 }, { "grad_norm": 0.16637343168258667, "learning_rate": 8.369798764446482e-05, "loss": 0.0115, "step": 24110 }, { "grad_norm": 0.24710746109485626, "learning_rate": 8.368271563217893e-05, "loss": 0.0142, "step": 24120 }, { "grad_norm": 0.16299305856227875, "learning_rate": 8.366743786443817e-05, "loss": 0.0119, "step": 24130 }, { "grad_norm": 0.1488874852657318, "learning_rate": 8.365215434385309e-05, "loss": 0.0126, "step": 24140 }, { "grad_norm": 0.15575049817562103, "learning_rate": 8.36368650730352e-05, "loss": 0.0118, "step": 24150 }, { "grad_norm": 0.1408136934041977, "learning_rate": 8.362157005459705e-05, "loss": 0.0123, "step": 24160 }, { "grad_norm": 0.19978530704975128, "learning_rate": 8.360626929115213e-05, "loss": 0.0105, "step": 24170 }, { "grad_norm": 0.171030193567276, "learning_rate": 8.359096278531492e-05, "loss": 0.0153, "step": 24180 }, { "grad_norm": 0.2021617889404297, "learning_rate": 8.357565053970088e-05, "loss": 0.0157, "step": 24190 }, { "grad_norm": 0.1758628487586975, "learning_rate": 8.356033255692647e-05, "loss": 0.011, "step": 24200 }, { "grad_norm": 0.21214376389980316, "learning_rate": 8.354500883960911e-05, "loss": 0.012, "step": 24210 }, { "grad_norm": 0.1655486822128296, "learning_rate": 8.352967939036717e-05, "loss": 0.0146, "step": 24220 }, { "grad_norm": 0.1738501638174057, "learning_rate": 8.35143442118201e-05, "loss": 0.0137, "step": 24230 }, { "grad_norm": 0.16838665306568146, "learning_rate": 8.349900330658819e-05, "loss": 0.0116, "step": 24240 }, { "grad_norm": 0.1324981302022934, "learning_rate": 8.348365667729284e-05, "loss": 0.0122, "step": 24250 }, { "grad_norm": 0.14095664024353027, "learning_rate": 8.346830432655633e-05, "loss": 0.011, "step": 24260 }, { "grad_norm": 0.1394716501235962, "learning_rate": 8.345294625700195e-05, "loss": 0.0102, "step": 24270 }, { "grad_norm": 0.15683844685554504, "learning_rate": 8.343758247125402e-05, "loss": 0.0116, "step": 24280 }, { "grad_norm": 0.19871588051319122, "learning_rate": 8.342221297193776e-05, "loss": 0.0155, "step": 24290 }, { "grad_norm": 0.1449330747127533, "learning_rate": 8.34068377616794e-05, "loss": 0.0128, "step": 24300 }, { "grad_norm": 0.1945897489786148, "learning_rate": 8.339145684310615e-05, "loss": 0.0138, "step": 24310 }, { "grad_norm": 0.14013852179050446, "learning_rate": 8.337607021884618e-05, "loss": 0.0116, "step": 24320 }, { "grad_norm": 0.20387892425060272, "learning_rate": 8.336067789152867e-05, "loss": 0.0128, "step": 24330 }, { "grad_norm": 0.19834353029727936, "learning_rate": 8.334527986378369e-05, "loss": 0.0137, "step": 24340 }, { "grad_norm": 0.1816319227218628, "learning_rate": 8.332987613824239e-05, "loss": 0.0119, "step": 24350 }, { "grad_norm": 0.15236547589302063, "learning_rate": 8.331446671753685e-05, "loss": 0.0132, "step": 24360 }, { "grad_norm": 0.17002497613430023, "learning_rate": 8.329905160430007e-05, "loss": 0.0136, "step": 24370 }, { "grad_norm": 0.2691374123096466, "learning_rate": 8.328363080116611e-05, "loss": 0.0123, "step": 24380 }, { "grad_norm": 0.17470771074295044, "learning_rate": 8.326820431076997e-05, "loss": 0.0123, "step": 24390 }, { "grad_norm": 0.13383741676807404, "learning_rate": 8.325277213574759e-05, "loss": 0.0123, "step": 24400 }, { "grad_norm": 0.17219342291355133, "learning_rate": 8.32373342787359e-05, "loss": 0.0137, "step": 24410 }, { "grad_norm": 0.17522567510604858, "learning_rate": 8.322189074237285e-05, "loss": 0.0131, "step": 24420 }, { "grad_norm": 0.18884727358818054, "learning_rate": 8.32064415292973e-05, "loss": 0.0145, "step": 24430 }, { "grad_norm": 0.17828179895877838, "learning_rate": 8.319098664214907e-05, "loss": 0.0152, "step": 24440 }, { "grad_norm": 0.14625419676303864, "learning_rate": 8.3175526083569e-05, "loss": 0.0134, "step": 24450 }, { "grad_norm": 0.1840135008096695, "learning_rate": 8.316005985619889e-05, "loss": 0.0118, "step": 24460 }, { "grad_norm": 0.1950893998146057, "learning_rate": 8.314458796268147e-05, "loss": 0.0116, "step": 24470 }, { "grad_norm": 0.16481232643127441, "learning_rate": 8.312911040566047e-05, "loss": 0.012, "step": 24480 }, { "grad_norm": 0.21828225255012512, "learning_rate": 8.31136271877806e-05, "loss": 0.0143, "step": 24490 }, { "grad_norm": 0.20057864487171173, "learning_rate": 8.309813831168748e-05, "loss": 0.0147, "step": 24500 }, { "grad_norm": 0.15399424731731415, "learning_rate": 8.308264378002777e-05, "loss": 0.0114, "step": 24510 }, { "grad_norm": 0.16494616866111755, "learning_rate": 8.306714359544906e-05, "loss": 0.0114, "step": 24520 }, { "grad_norm": 0.1822587102651596, "learning_rate": 8.30516377605999e-05, "loss": 0.012, "step": 24530 }, { "grad_norm": 0.20364944636821747, "learning_rate": 8.30361262781298e-05, "loss": 0.0151, "step": 24540 }, { "grad_norm": 0.11769521981477737, "learning_rate": 8.302060915068924e-05, "loss": 0.0118, "step": 24550 }, { "grad_norm": 0.18690063059329987, "learning_rate": 8.300508638092972e-05, "loss": 0.0127, "step": 24560 }, { "grad_norm": 0.11806757003068924, "learning_rate": 8.298955797150361e-05, "loss": 0.0144, "step": 24570 }, { "grad_norm": 0.18195120990276337, "learning_rate": 8.297402392506433e-05, "loss": 0.0102, "step": 24580 }, { "grad_norm": 0.1289147287607193, "learning_rate": 8.295848424426617e-05, "loss": 0.0095, "step": 24590 }, { "grad_norm": 0.1633138805627823, "learning_rate": 8.29429389317645e-05, "loss": 0.0129, "step": 24600 }, { "grad_norm": 0.19083194434642792, "learning_rate": 8.292738799021556e-05, "loss": 0.0127, "step": 24610 }, { "grad_norm": 0.1782887727022171, "learning_rate": 8.291183142227656e-05, "loss": 0.0148, "step": 24620 }, { "grad_norm": 0.19839996099472046, "learning_rate": 8.289626923060572e-05, "loss": 0.0115, "step": 24630 }, { "grad_norm": 0.20495718717575073, "learning_rate": 8.288070141786218e-05, "loss": 0.0115, "step": 24640 }, { "grad_norm": 0.13151519000530243, "learning_rate": 8.286512798670605e-05, "loss": 0.0109, "step": 24650 }, { "grad_norm": 0.15621036291122437, "learning_rate": 8.284954893979842e-05, "loss": 0.0115, "step": 24660 }, { "grad_norm": 0.1693412959575653, "learning_rate": 8.283396427980131e-05, "loss": 0.0118, "step": 24670 }, { "grad_norm": 0.17424964904785156, "learning_rate": 8.281837400937771e-05, "loss": 0.0137, "step": 24680 }, { "grad_norm": 0.18432065844535828, "learning_rate": 8.28027781311916e-05, "loss": 0.0143, "step": 24690 }, { "grad_norm": 0.15674914419651031, "learning_rate": 8.278717664790785e-05, "loss": 0.0121, "step": 24700 }, { "grad_norm": 0.19725684821605682, "learning_rate": 8.277156956219234e-05, "loss": 0.0109, "step": 24710 }, { "grad_norm": 0.16883836686611176, "learning_rate": 8.275595687671189e-05, "loss": 0.0153, "step": 24720 }, { "grad_norm": 0.20366086065769196, "learning_rate": 8.27403385941343e-05, "loss": 0.0114, "step": 24730 }, { "grad_norm": 0.18172453343868256, "learning_rate": 8.272471471712828e-05, "loss": 0.0119, "step": 24740 }, { "grad_norm": 0.20533967018127441, "learning_rate": 8.270908524836355e-05, "loss": 0.0124, "step": 24750 }, { "grad_norm": 0.2553025186061859, "learning_rate": 8.269345019051074e-05, "loss": 0.015, "step": 24760 }, { "grad_norm": 0.1985418051481247, "learning_rate": 8.267780954624147e-05, "loss": 0.0113, "step": 24770 }, { "grad_norm": 0.1679641604423523, "learning_rate": 8.266216331822827e-05, "loss": 0.0118, "step": 24780 }, { "grad_norm": 0.2048112004995346, "learning_rate": 8.264651150914469e-05, "loss": 0.0139, "step": 24790 }, { "grad_norm": 0.19158095121383667, "learning_rate": 8.263085412166517e-05, "loss": 0.0118, "step": 24800 }, { "grad_norm": 0.22639845311641693, "learning_rate": 8.261519115846514e-05, "loss": 0.014, "step": 24810 }, { "grad_norm": 0.19986532628536224, "learning_rate": 8.259952262222096e-05, "loss": 0.0113, "step": 24820 }, { "grad_norm": 0.19772285223007202, "learning_rate": 8.258384851560997e-05, "loss": 0.0107, "step": 24830 }, { "grad_norm": 0.20022477209568024, "learning_rate": 8.256816884131044e-05, "loss": 0.011, "step": 24840 }, { "grad_norm": 0.17044521868228912, "learning_rate": 8.255248360200159e-05, "loss": 0.0124, "step": 24850 }, { "grad_norm": 0.18773221969604492, "learning_rate": 8.253679280036359e-05, "loss": 0.0101, "step": 24860 }, { "grad_norm": 0.1649075150489807, "learning_rate": 8.252109643907762e-05, "loss": 0.0115, "step": 24870 }, { "grad_norm": 0.15764443576335907, "learning_rate": 8.250539452082569e-05, "loss": 0.0112, "step": 24880 }, { "grad_norm": 0.1560094952583313, "learning_rate": 8.248968704829087e-05, "loss": 0.0092, "step": 24890 }, { "grad_norm": 0.15999509394168854, "learning_rate": 8.247397402415714e-05, "loss": 0.0109, "step": 24900 }, { "grad_norm": 0.15973320603370667, "learning_rate": 8.24582554511094e-05, "loss": 0.0116, "step": 24910 }, { "grad_norm": 0.16617341339588165, "learning_rate": 8.244253133183355e-05, "loss": 0.0151, "step": 24920 }, { "grad_norm": 0.19706247746944427, "learning_rate": 8.24268016690164e-05, "loss": 0.012, "step": 24930 }, { "grad_norm": 0.16547076404094696, "learning_rate": 8.241106646534571e-05, "loss": 0.0103, "step": 24940 }, { "grad_norm": 0.1278744786977768, "learning_rate": 8.23953257235102e-05, "loss": 0.0114, "step": 24950 }, { "grad_norm": 0.1421400010585785, "learning_rate": 8.237957944619956e-05, "loss": 0.0122, "step": 24960 }, { "grad_norm": 0.11784877628087997, "learning_rate": 8.236382763610437e-05, "loss": 0.0099, "step": 24970 }, { "grad_norm": 0.1845155954360962, "learning_rate": 8.234807029591619e-05, "loss": 0.0141, "step": 24980 }, { "grad_norm": 0.21461373567581177, "learning_rate": 8.233230742832752e-05, "loss": 0.0122, "step": 24990 }, { "grad_norm": 0.13403284549713135, "learning_rate": 8.231653903603178e-05, "loss": 0.0122, "step": 25000 }, { "grad_norm": 0.16524222493171692, "learning_rate": 8.23007651217234e-05, "loss": 0.01, "step": 25010 }, { "grad_norm": 0.1550874263048172, "learning_rate": 8.228498568809769e-05, "loss": 0.0117, "step": 25020 }, { "grad_norm": 0.17560186982154846, "learning_rate": 8.22692007378509e-05, "loss": 0.0125, "step": 25030 }, { "grad_norm": 0.1312147080898285, "learning_rate": 8.225341027368028e-05, "loss": 0.0107, "step": 25040 }, { "grad_norm": 0.15940386056900024, "learning_rate": 8.223761429828399e-05, "loss": 0.0122, "step": 25050 }, { "grad_norm": 0.13454821705818176, "learning_rate": 8.22218128143611e-05, "loss": 0.0116, "step": 25060 }, { "grad_norm": 0.17665515840053558, "learning_rate": 8.220600582461166e-05, "loss": 0.0137, "step": 25070 }, { "grad_norm": 0.22094590961933136, "learning_rate": 8.219019333173668e-05, "loss": 0.011, "step": 25080 }, { "grad_norm": 0.16325858235359192, "learning_rate": 8.217437533843805e-05, "loss": 0.0108, "step": 25090 }, { "grad_norm": 0.18719075620174408, "learning_rate": 8.215855184741867e-05, "loss": 0.0127, "step": 25100 }, { "grad_norm": 0.1435537040233612, "learning_rate": 8.21427228613823e-05, "loss": 0.0121, "step": 25110 }, { "grad_norm": 0.15852570533752441, "learning_rate": 8.21268883830337e-05, "loss": 0.0137, "step": 25120 }, { "grad_norm": 0.17820055782794952, "learning_rate": 8.211104841507855e-05, "loss": 0.0126, "step": 25130 }, { "grad_norm": 0.1482018083333969, "learning_rate": 8.209520296022346e-05, "loss": 0.0132, "step": 25140 }, { "grad_norm": 0.16792099177837372, "learning_rate": 8.207935202117599e-05, "loss": 0.0112, "step": 25150 }, { "grad_norm": 0.1742662787437439, "learning_rate": 8.206349560064463e-05, "loss": 0.0098, "step": 25160 }, { "grad_norm": 0.16617320477962494, "learning_rate": 8.204763370133881e-05, "loss": 0.0119, "step": 25170 }, { "grad_norm": 0.17158794403076172, "learning_rate": 8.203176632596892e-05, "loss": 0.0123, "step": 25180 }, { "grad_norm": 0.20583461225032806, "learning_rate": 8.20158934772462e-05, "loss": 0.0117, "step": 25190 }, { "grad_norm": 0.16622993350028992, "learning_rate": 8.200001515788294e-05, "loss": 0.0095, "step": 25200 }, { "grad_norm": 0.14978590607643127, "learning_rate": 8.198413137059228e-05, "loss": 0.0095, "step": 25210 }, { "grad_norm": 0.1576579511165619, "learning_rate": 8.196824211808835e-05, "loss": 0.0133, "step": 25220 }, { "grad_norm": 0.18895533680915833, "learning_rate": 8.195234740308617e-05, "loss": 0.0137, "step": 25230 }, { "grad_norm": 0.19108036160469055, "learning_rate": 8.193644722830171e-05, "loss": 0.0116, "step": 25240 }, { "grad_norm": 0.14911936223506927, "learning_rate": 8.19205415964519e-05, "loss": 0.0089, "step": 25250 }, { "grad_norm": 0.1449291855096817, "learning_rate": 8.190463051025456e-05, "loss": 0.0106, "step": 25260 }, { "grad_norm": 0.1524459421634674, "learning_rate": 8.188871397242843e-05, "loss": 0.0125, "step": 25270 }, { "grad_norm": 0.15716810524463654, "learning_rate": 8.187279198569326e-05, "loss": 0.0109, "step": 25280 }, { "grad_norm": 0.2025960087776184, "learning_rate": 8.185686455276966e-05, "loss": 0.0139, "step": 25290 }, { "grad_norm": 0.14523887634277344, "learning_rate": 8.184093167637921e-05, "loss": 0.0127, "step": 25300 }, { "grad_norm": 0.2117985635995865, "learning_rate": 8.182499335924437e-05, "loss": 0.0139, "step": 25310 }, { "grad_norm": 0.15180183947086334, "learning_rate": 8.18090496040886e-05, "loss": 0.0139, "step": 25320 }, { "grad_norm": 0.13181710243225098, "learning_rate": 8.179310041363621e-05, "loss": 0.0105, "step": 25330 }, { "grad_norm": 0.16371065378189087, "learning_rate": 8.17771457906125e-05, "loss": 0.0115, "step": 25340 }, { "grad_norm": 0.1476866900920868, "learning_rate": 8.176118573774371e-05, "loss": 0.0101, "step": 25350 }, { "grad_norm": 0.16373442113399506, "learning_rate": 8.174522025775692e-05, "loss": 0.011, "step": 25360 }, { "grad_norm": 0.16592414677143097, "learning_rate": 8.172924935338022e-05, "loss": 0.0125, "step": 25370 }, { "grad_norm": 0.1562318354845047, "learning_rate": 8.171327302734262e-05, "loss": 0.0131, "step": 25380 }, { "grad_norm": 0.13501478731632233, "learning_rate": 8.169729128237401e-05, "loss": 0.0127, "step": 25390 }, { "grad_norm": 0.14213936030864716, "learning_rate": 8.168130412120525e-05, "loss": 0.0123, "step": 25400 }, { "grad_norm": 0.171024352312088, "learning_rate": 8.16653115465681e-05, "loss": 0.0149, "step": 25410 }, { "grad_norm": 0.14456279575824738, "learning_rate": 8.164931356119526e-05, "loss": 0.0128, "step": 25420 }, { "grad_norm": 0.1519649177789688, "learning_rate": 8.163331016782032e-05, "loss": 0.0108, "step": 25430 }, { "grad_norm": 0.12916482985019684, "learning_rate": 8.161730136917785e-05, "loss": 0.0137, "step": 25440 }, { "grad_norm": 0.11928904056549072, "learning_rate": 8.160128716800333e-05, "loss": 0.0096, "step": 25450 }, { "grad_norm": 0.1320548802614212, "learning_rate": 8.158526756703313e-05, "loss": 0.0133, "step": 25460 }, { "grad_norm": 0.13902615010738373, "learning_rate": 8.156924256900455e-05, "loss": 0.0177, "step": 25470 }, { "grad_norm": 0.18167784810066223, "learning_rate": 8.155321217665584e-05, "loss": 0.0146, "step": 25480 }, { "grad_norm": 0.1686467081308365, "learning_rate": 8.153717639272614e-05, "loss": 0.0136, "step": 25490 }, { "grad_norm": 0.1872495710849762, "learning_rate": 8.152113521995555e-05, "loss": 0.0124, "step": 25500 }, { "grad_norm": 0.14896979928016663, "learning_rate": 8.150508866108505e-05, "loss": 0.0124, "step": 25510 }, { "grad_norm": 0.21315361559391022, "learning_rate": 8.148903671885657e-05, "loss": 0.0154, "step": 25520 }, { "grad_norm": 0.2238854020833969, "learning_rate": 8.147297939601292e-05, "loss": 0.0145, "step": 25530 }, { "grad_norm": 0.17180606722831726, "learning_rate": 8.145691669529792e-05, "loss": 0.014, "step": 25540 }, { "grad_norm": 0.13382813334465027, "learning_rate": 8.144084861945618e-05, "loss": 0.0093, "step": 25550 }, { "grad_norm": 0.17526951432228088, "learning_rate": 8.142477517123333e-05, "loss": 0.0106, "step": 25560 }, { "grad_norm": 0.1778167337179184, "learning_rate": 8.140869635337586e-05, "loss": 0.0123, "step": 25570 }, { "grad_norm": 0.14897117018699646, "learning_rate": 8.139261216863123e-05, "loss": 0.0137, "step": 25580 }, { "grad_norm": 0.19740450382232666, "learning_rate": 8.137652261974776e-05, "loss": 0.0141, "step": 25590 }, { "grad_norm": 0.12784329056739807, "learning_rate": 8.136042770947472e-05, "loss": 0.0124, "step": 25600 }, { "grad_norm": 0.12651023268699646, "learning_rate": 8.134432744056228e-05, "loss": 0.011, "step": 25610 }, { "grad_norm": 0.15097057819366455, "learning_rate": 8.132822181576158e-05, "loss": 0.0118, "step": 25620 }, { "grad_norm": 0.17803333699703217, "learning_rate": 8.131211083782459e-05, "loss": 0.0127, "step": 25630 }, { "grad_norm": 0.18659864366054535, "learning_rate": 8.129599450950424e-05, "loss": 0.0107, "step": 25640 }, { "grad_norm": 0.1354772299528122, "learning_rate": 8.127987283355438e-05, "loss": 0.011, "step": 25650 }, { "grad_norm": 0.20952492952346802, "learning_rate": 8.126374581272976e-05, "loss": 0.0119, "step": 25660 }, { "grad_norm": 0.19828274846076965, "learning_rate": 8.124761344978605e-05, "loss": 0.0167, "step": 25670 }, { "grad_norm": 0.1317204236984253, "learning_rate": 8.12314757474798e-05, "loss": 0.0132, "step": 25680 }, { "grad_norm": 0.16458728909492493, "learning_rate": 8.121533270856856e-05, "loss": 0.0091, "step": 25690 }, { "grad_norm": 0.189958855509758, "learning_rate": 8.119918433581069e-05, "loss": 0.0137, "step": 25700 }, { "grad_norm": 0.13879287242889404, "learning_rate": 8.118303063196551e-05, "loss": 0.0133, "step": 25710 }, { "grad_norm": 0.10606580972671509, "learning_rate": 8.116687159979326e-05, "loss": 0.0108, "step": 25720 }, { "grad_norm": 0.19452974200248718, "learning_rate": 8.115070724205508e-05, "loss": 0.0128, "step": 25730 }, { "grad_norm": 0.17362748086452484, "learning_rate": 8.113453756151296e-05, "loss": 0.0142, "step": 25740 }, { "grad_norm": 0.19260528683662415, "learning_rate": 8.111836256092995e-05, "loss": 0.0141, "step": 25750 }, { "grad_norm": 0.21844016015529633, "learning_rate": 8.110218224306985e-05, "loss": 0.0126, "step": 25760 }, { "grad_norm": 0.12712886929512024, "learning_rate": 8.108599661069745e-05, "loss": 0.0105, "step": 25770 }, { "grad_norm": 0.1838972568511963, "learning_rate": 8.106980566657845e-05, "loss": 0.0121, "step": 25780 }, { "grad_norm": 0.17138609290122986, "learning_rate": 8.10536094134794e-05, "loss": 0.0126, "step": 25790 }, { "grad_norm": 0.11505641043186188, "learning_rate": 8.103740785416783e-05, "loss": 0.01, "step": 25800 }, { "grad_norm": 0.1535670906305313, "learning_rate": 8.102120099141212e-05, "loss": 0.0111, "step": 25810 }, { "grad_norm": 0.14999733865261078, "learning_rate": 8.100498882798163e-05, "loss": 0.0121, "step": 25820 }, { "grad_norm": 0.12289073318243027, "learning_rate": 8.09887713666465e-05, "loss": 0.0143, "step": 25830 }, { "grad_norm": 0.28923434019088745, "learning_rate": 8.09725486101779e-05, "loss": 0.013, "step": 25840 }, { "grad_norm": 0.11285758763551712, "learning_rate": 8.095632056134784e-05, "loss": 0.0127, "step": 25850 }, { "grad_norm": 0.14812469482421875, "learning_rate": 8.094008722292925e-05, "loss": 0.016, "step": 25860 }, { "grad_norm": 0.12799230217933655, "learning_rate": 8.092384859769598e-05, "loss": 0.0101, "step": 25870 }, { "grad_norm": 0.1519114375114441, "learning_rate": 8.090760468842275e-05, "loss": 0.0164, "step": 25880 }, { "grad_norm": 0.1999591588973999, "learning_rate": 8.089135549788521e-05, "loss": 0.0116, "step": 25890 }, { "grad_norm": 0.1318873018026352, "learning_rate": 8.087510102885987e-05, "loss": 0.0111, "step": 25900 }, { "grad_norm": 0.2347613424062729, "learning_rate": 8.085884128412422e-05, "loss": 0.0154, "step": 25910 }, { "grad_norm": 0.27094969153404236, "learning_rate": 8.084257626645659e-05, "loss": 0.0133, "step": 25920 }, { "grad_norm": 0.13948829472064972, "learning_rate": 8.08263059786362e-05, "loss": 0.0106, "step": 25930 }, { "grad_norm": 0.1574452519416809, "learning_rate": 8.081003042344325e-05, "loss": 0.0102, "step": 25940 }, { "grad_norm": 0.15969012677669525, "learning_rate": 8.079374960365872e-05, "loss": 0.011, "step": 25950 }, { "grad_norm": 0.10254243016242981, "learning_rate": 8.077746352206463e-05, "loss": 0.0092, "step": 25960 }, { "grad_norm": 0.16004987061023712, "learning_rate": 8.076117218144377e-05, "loss": 0.0084, "step": 25970 }, { "grad_norm": 0.17225003242492676, "learning_rate": 8.074487558457991e-05, "loss": 0.0099, "step": 25980 }, { "grad_norm": 0.15075376629829407, "learning_rate": 8.072857373425768e-05, "loss": 0.01, "step": 25990 }, { "grad_norm": 0.13065491616725922, "learning_rate": 8.071226663326264e-05, "loss": 0.0095, "step": 26000 }, { "grad_norm": 0.1418628841638565, "learning_rate": 8.069595428438121e-05, "loss": 0.0146, "step": 26010 }, { "grad_norm": 0.18277676403522491, "learning_rate": 8.067963669040072e-05, "loss": 0.0104, "step": 26020 }, { "grad_norm": 0.18744057416915894, "learning_rate": 8.066331385410942e-05, "loss": 0.0141, "step": 26030 }, { "grad_norm": 0.18224860727787018, "learning_rate": 8.064698577829641e-05, "loss": 0.0108, "step": 26040 }, { "grad_norm": 0.14439161121845245, "learning_rate": 8.063065246575175e-05, "loss": 0.0097, "step": 26050 }, { "grad_norm": 0.16921041905879974, "learning_rate": 8.061431391926631e-05, "loss": 0.0112, "step": 26060 }, { "grad_norm": 0.16854725778102875, "learning_rate": 8.059797014163195e-05, "loss": 0.0117, "step": 26070 }, { "grad_norm": 0.17508524656295776, "learning_rate": 8.058162113564133e-05, "loss": 0.0121, "step": 26080 }, { "grad_norm": 0.1760987490415573, "learning_rate": 8.056526690408806e-05, "loss": 0.0133, "step": 26090 }, { "grad_norm": 0.15702641010284424, "learning_rate": 8.054890744976666e-05, "loss": 0.0113, "step": 26100 }, { "grad_norm": 0.11899559199810028, "learning_rate": 8.053254277547248e-05, "loss": 0.01, "step": 26110 }, { "grad_norm": 0.13759683072566986, "learning_rate": 8.051617288400182e-05, "loss": 0.0095, "step": 26120 }, { "grad_norm": 0.1863390952348709, "learning_rate": 8.049979777815182e-05, "loss": 0.0101, "step": 26130 }, { "grad_norm": 0.1856270730495453, "learning_rate": 8.048341746072054e-05, "loss": 0.0116, "step": 26140 }, { "grad_norm": 0.1309521645307541, "learning_rate": 8.046703193450696e-05, "loss": 0.0128, "step": 26150 }, { "grad_norm": 0.19463084638118744, "learning_rate": 8.04506412023109e-05, "loss": 0.0104, "step": 26160 }, { "grad_norm": 0.18253836035728455, "learning_rate": 8.043424526693306e-05, "loss": 0.0137, "step": 26170 }, { "grad_norm": 0.1547091156244278, "learning_rate": 8.04178441311751e-05, "loss": 0.0096, "step": 26180 }, { "grad_norm": 0.1179729700088501, "learning_rate": 8.04014377978395e-05, "loss": 0.01, "step": 26190 }, { "grad_norm": 0.20564791560173035, "learning_rate": 8.038502626972967e-05, "loss": 0.0136, "step": 26200 }, { "grad_norm": 0.1529185175895691, "learning_rate": 8.036860954964989e-05, "loss": 0.0137, "step": 26210 }, { "grad_norm": 0.13193698227405548, "learning_rate": 8.035218764040531e-05, "loss": 0.013, "step": 26220 }, { "grad_norm": 0.17470581829547882, "learning_rate": 8.033576054480199e-05, "loss": 0.0143, "step": 26230 }, { "grad_norm": 0.14730222523212433, "learning_rate": 8.031932826564688e-05, "loss": 0.0114, "step": 26240 }, { "grad_norm": 0.13431039452552795, "learning_rate": 8.030289080574782e-05, "loss": 0.0092, "step": 26250 }, { "grad_norm": 0.13313192129135132, "learning_rate": 8.028644816791349e-05, "loss": 0.0111, "step": 26260 }, { "grad_norm": 0.3267718553543091, "learning_rate": 8.027000035495351e-05, "loss": 0.0097, "step": 26270 }, { "grad_norm": 0.12405971437692642, "learning_rate": 8.025354736967836e-05, "loss": 0.0088, "step": 26280 }, { "grad_norm": 0.14032886922359467, "learning_rate": 8.023708921489941e-05, "loss": 0.0098, "step": 26290 }, { "grad_norm": 0.12812529504299164, "learning_rate": 8.022062589342887e-05, "loss": 0.0115, "step": 26300 }, { "grad_norm": 0.15344274044036865, "learning_rate": 8.020415740807993e-05, "loss": 0.0105, "step": 26310 }, { "grad_norm": 0.14711278676986694, "learning_rate": 8.018768376166656e-05, "loss": 0.0123, "step": 26320 }, { "grad_norm": 0.14825165271759033, "learning_rate": 8.017120495700368e-05, "loss": 0.0126, "step": 26330 }, { "grad_norm": 0.1470337063074112, "learning_rate": 8.015472099690704e-05, "loss": 0.0114, "step": 26340 }, { "grad_norm": 0.13658811151981354, "learning_rate": 8.013823188419332e-05, "loss": 0.0133, "step": 26350 }, { "grad_norm": 0.19031402468681335, "learning_rate": 8.012173762168006e-05, "loss": 0.011, "step": 26360 }, { "grad_norm": 0.15039251744747162, "learning_rate": 8.010523821218567e-05, "loss": 0.0097, "step": 26370 }, { "grad_norm": 0.21736909449100494, "learning_rate": 8.008873365852945e-05, "loss": 0.0124, "step": 26380 }, { "grad_norm": 0.13771986961364746, "learning_rate": 8.007222396353157e-05, "loss": 0.0129, "step": 26390 }, { "grad_norm": 0.17267313599586487, "learning_rate": 8.00557091300131e-05, "loss": 0.0116, "step": 26400 }, { "grad_norm": 0.14247366786003113, "learning_rate": 8.003918916079597e-05, "loss": 0.0117, "step": 26410 }, { "grad_norm": 0.1698984056711197, "learning_rate": 8.002266405870298e-05, "loss": 0.012, "step": 26420 }, { "grad_norm": 0.10331499576568604, "learning_rate": 8.000613382655782e-05, "loss": 0.0104, "step": 26430 }, { "grad_norm": 0.12870991230010986, "learning_rate": 7.998959846718505e-05, "loss": 0.0126, "step": 26440 }, { "grad_norm": 0.19351212680339813, "learning_rate": 7.997305798341012e-05, "loss": 0.0112, "step": 26450 }, { "grad_norm": 0.17923620343208313, "learning_rate": 7.995651237805937e-05, "loss": 0.012, "step": 26460 }, { "grad_norm": 0.15403172373771667, "learning_rate": 7.993996165395996e-05, "loss": 0.0121, "step": 26470 }, { "grad_norm": 0.13503101468086243, "learning_rate": 7.992340581393996e-05, "loss": 0.0113, "step": 26480 }, { "grad_norm": 0.15720467269420624, "learning_rate": 7.990684486082831e-05, "loss": 0.0125, "step": 26490 }, { "grad_norm": 0.13635165989398956, "learning_rate": 7.989027879745482e-05, "loss": 0.0106, "step": 26500 }, { "grad_norm": 0.18589016795158386, "learning_rate": 7.98737076266502e-05, "loss": 0.0121, "step": 26510 }, { "grad_norm": 0.21872729063034058, "learning_rate": 7.985713135124598e-05, "loss": 0.0113, "step": 26520 }, { "grad_norm": 0.2429024875164032, "learning_rate": 7.98405499740746e-05, "loss": 0.0127, "step": 26530 }, { "grad_norm": 0.15583884716033936, "learning_rate": 7.98239634979694e-05, "loss": 0.0153, "step": 26540 }, { "grad_norm": 0.16351935267448425, "learning_rate": 7.98073719257645e-05, "loss": 0.0126, "step": 26550 }, { "grad_norm": 0.143492192029953, "learning_rate": 7.979077526029499e-05, "loss": 0.0122, "step": 26560 }, { "grad_norm": 0.15169307589530945, "learning_rate": 7.977417350439675e-05, "loss": 0.0151, "step": 26570 }, { "grad_norm": 0.16451995074748993, "learning_rate": 7.97575666609066e-05, "loss": 0.0142, "step": 26580 }, { "grad_norm": 0.13521181046962738, "learning_rate": 7.974095473266216e-05, "loss": 0.0109, "step": 26590 }, { "grad_norm": 0.10825876891613007, "learning_rate": 7.972433772250198e-05, "loss": 0.0112, "step": 26600 }, { "grad_norm": 0.2076387256383896, "learning_rate": 7.970771563326544e-05, "loss": 0.014, "step": 26610 }, { "grad_norm": 0.14315077662467957, "learning_rate": 7.96910884677928e-05, "loss": 0.0138, "step": 26620 }, { "grad_norm": 0.21298183500766754, "learning_rate": 7.967445622892523e-05, "loss": 0.0118, "step": 26630 }, { "grad_norm": 0.1454935073852539, "learning_rate": 7.965781891950465e-05, "loss": 0.0102, "step": 26640 }, { "grad_norm": 0.1063150018453598, "learning_rate": 7.964117654237397e-05, "loss": 0.0102, "step": 26650 }, { "grad_norm": 0.14325018227100372, "learning_rate": 7.962452910037692e-05, "loss": 0.0111, "step": 26660 }, { "grad_norm": 0.15584105253219604, "learning_rate": 7.96078765963581e-05, "loss": 0.0108, "step": 26670 }, { "grad_norm": 0.1907406747341156, "learning_rate": 7.95912190331629e-05, "loss": 0.0105, "step": 26680 }, { "grad_norm": 0.1248830109834671, "learning_rate": 7.957455641363772e-05, "loss": 0.0104, "step": 26690 }, { "grad_norm": 0.1767905205488205, "learning_rate": 7.955788874062968e-05, "loss": 0.011, "step": 26700 }, { "grad_norm": 0.17209218442440033, "learning_rate": 7.95412160169869e-05, "loss": 0.0119, "step": 26710 }, { "grad_norm": 0.1910022497177124, "learning_rate": 7.952453824555824e-05, "loss": 0.0111, "step": 26720 }, { "grad_norm": 0.18355056643486023, "learning_rate": 7.95078554291935e-05, "loss": 0.0104, "step": 26730 }, { "grad_norm": 0.17193223536014557, "learning_rate": 7.94911675707433e-05, "loss": 0.0093, "step": 26740 }, { "grad_norm": 0.14593608677387238, "learning_rate": 7.947447467305915e-05, "loss": 0.0112, "step": 26750 }, { "grad_norm": 0.1497984677553177, "learning_rate": 7.94577767389934e-05, "loss": 0.0131, "step": 26760 }, { "grad_norm": 0.148638054728508, "learning_rate": 7.944107377139928e-05, "loss": 0.0102, "step": 26770 }, { "grad_norm": 0.17661331593990326, "learning_rate": 7.942436577313088e-05, "loss": 0.013, "step": 26780 }, { "grad_norm": 0.13317608833312988, "learning_rate": 7.940765274704312e-05, "loss": 0.0081, "step": 26790 }, { "grad_norm": 0.14482656121253967, "learning_rate": 7.939093469599181e-05, "loss": 0.0108, "step": 26800 }, { "grad_norm": 0.14432941377162933, "learning_rate": 7.93742116228336e-05, "loss": 0.0112, "step": 26810 }, { "grad_norm": 0.20567645132541656, "learning_rate": 7.935748353042602e-05, "loss": 0.0116, "step": 26820 }, { "grad_norm": 0.17889541387557983, "learning_rate": 7.934075042162744e-05, "loss": 0.0124, "step": 26830 }, { "grad_norm": 0.16109271347522736, "learning_rate": 7.932401229929705e-05, "loss": 0.0116, "step": 26840 }, { "grad_norm": 0.1391121745109558, "learning_rate": 7.9307269166295e-05, "loss": 0.0103, "step": 26850 }, { "grad_norm": 0.17370782792568207, "learning_rate": 7.92905210254822e-05, "loss": 0.0092, "step": 26860 }, { "grad_norm": 0.19839128851890564, "learning_rate": 7.927376787972045e-05, "loss": 0.0127, "step": 26870 }, { "grad_norm": 0.20770280063152313, "learning_rate": 7.92570097318724e-05, "loss": 0.0118, "step": 26880 }, { "grad_norm": 0.1534220576286316, "learning_rate": 7.924024658480158e-05, "loss": 0.0133, "step": 26890 }, { "grad_norm": 0.17059281468391418, "learning_rate": 7.922347844137233e-05, "loss": 0.0116, "step": 26900 }, { "grad_norm": 0.19448187947273254, "learning_rate": 7.92067053044499e-05, "loss": 0.012, "step": 26910 }, { "grad_norm": 0.16469967365264893, "learning_rate": 7.918992717690031e-05, "loss": 0.0109, "step": 26920 }, { "grad_norm": 0.19100050628185272, "learning_rate": 7.917314406159053e-05, "loss": 0.0123, "step": 26930 }, { "grad_norm": 0.11395150423049927, "learning_rate": 7.915635596138832e-05, "loss": 0.0116, "step": 26940 }, { "grad_norm": 0.22609581053256989, "learning_rate": 7.913956287916228e-05, "loss": 0.0152, "step": 26950 }, { "grad_norm": 0.13924583792686462, "learning_rate": 7.912276481778193e-05, "loss": 0.0119, "step": 26960 }, { "grad_norm": 0.1834266632795334, "learning_rate": 7.910596178011759e-05, "loss": 0.0103, "step": 26970 }, { "grad_norm": 0.19562646746635437, "learning_rate": 7.908915376904043e-05, "loss": 0.0112, "step": 26980 }, { "grad_norm": 0.22787751257419586, "learning_rate": 7.907234078742247e-05, "loss": 0.012, "step": 26990 }, { "grad_norm": 0.1465006172657013, "learning_rate": 7.90555228381366e-05, "loss": 0.0104, "step": 27000 }, { "grad_norm": 0.11589900404214859, "learning_rate": 7.903869992405656e-05, "loss": 0.0114, "step": 27010 }, { "grad_norm": 0.1398123800754547, "learning_rate": 7.902187204805691e-05, "loss": 0.0114, "step": 27020 }, { "grad_norm": 0.140925794839859, "learning_rate": 7.900503921301308e-05, "loss": 0.011, "step": 27030 }, { "grad_norm": 0.14628705382347107, "learning_rate": 7.898820142180133e-05, "loss": 0.0099, "step": 27040 }, { "grad_norm": 0.12192625552415848, "learning_rate": 7.897135867729879e-05, "loss": 0.0096, "step": 27050 }, { "grad_norm": 0.11406855285167694, "learning_rate": 7.89545109823834e-05, "loss": 0.0137, "step": 27060 }, { "grad_norm": 0.12831607460975647, "learning_rate": 7.8937658339934e-05, "loss": 0.0107, "step": 27070 }, { "grad_norm": 0.19393327832221985, "learning_rate": 7.892080075283026e-05, "loss": 0.0118, "step": 27080 }, { "grad_norm": 0.1396946758031845, "learning_rate": 7.890393822395263e-05, "loss": 0.0098, "step": 27090 }, { "grad_norm": 0.13688552379608154, "learning_rate": 7.88870707561825e-05, "loss": 0.0104, "step": 27100 }, { "grad_norm": 0.15246273577213287, "learning_rate": 7.887019835240203e-05, "loss": 0.0097, "step": 27110 }, { "grad_norm": 0.12128090113401413, "learning_rate": 7.885332101549427e-05, "loss": 0.0102, "step": 27120 }, { "grad_norm": 0.14470313489437103, "learning_rate": 7.883643874834308e-05, "loss": 0.0104, "step": 27130 }, { "grad_norm": 0.28309696912765503, "learning_rate": 7.881955155383321e-05, "loss": 0.0097, "step": 27140 }, { "grad_norm": 0.14371544122695923, "learning_rate": 7.880265943485017e-05, "loss": 0.0122, "step": 27150 }, { "grad_norm": 0.15018244087696075, "learning_rate": 7.878576239428038e-05, "loss": 0.0136, "step": 27160 }, { "grad_norm": 0.18080084025859833, "learning_rate": 7.87688604350111e-05, "loss": 0.0125, "step": 27170 }, { "grad_norm": 0.09706562757492065, "learning_rate": 7.875195355993042e-05, "loss": 0.0111, "step": 27180 }, { "grad_norm": 0.17506441473960876, "learning_rate": 7.873504177192724e-05, "loss": 0.0141, "step": 27190 }, { "grad_norm": 0.1403864175081253, "learning_rate": 7.87181250738913e-05, "loss": 0.0107, "step": 27200 }, { "grad_norm": 0.13364501297473907, "learning_rate": 7.870120346871324e-05, "loss": 0.0109, "step": 27210 }, { "grad_norm": 0.1426720917224884, "learning_rate": 7.86842769592845e-05, "loss": 0.0092, "step": 27220 }, { "grad_norm": 0.1509006768465042, "learning_rate": 7.866734554849732e-05, "loss": 0.0093, "step": 27230 }, { "grad_norm": 0.14622433483600616, "learning_rate": 7.865040923924486e-05, "loss": 0.0103, "step": 27240 }, { "grad_norm": 0.14605113863945007, "learning_rate": 7.863346803442104e-05, "loss": 0.0116, "step": 27250 }, { "grad_norm": 0.17430303990840912, "learning_rate": 7.861652193692067e-05, "loss": 0.0117, "step": 27260 }, { "grad_norm": 0.180787593126297, "learning_rate": 7.859957094963937e-05, "loss": 0.0106, "step": 27270 }, { "grad_norm": 0.09976784884929657, "learning_rate": 7.858261507547357e-05, "loss": 0.0108, "step": 27280 }, { "grad_norm": 0.1679612249135971, "learning_rate": 7.856565431732061e-05, "loss": 0.0119, "step": 27290 }, { "grad_norm": 0.12842300534248352, "learning_rate": 7.854868867807859e-05, "loss": 0.0142, "step": 27300 }, { "grad_norm": 0.11916430294513702, "learning_rate": 7.85317181606465e-05, "loss": 0.011, "step": 27310 }, { "grad_norm": 0.20397041738033295, "learning_rate": 7.85147427679241e-05, "loss": 0.0111, "step": 27320 }, { "grad_norm": 0.14046868681907654, "learning_rate": 7.849776250281205e-05, "loss": 0.0099, "step": 27330 }, { "grad_norm": 0.1593417227268219, "learning_rate": 7.84807773682118e-05, "loss": 0.0109, "step": 27340 }, { "grad_norm": 0.17351612448692322, "learning_rate": 7.846378736702565e-05, "loss": 0.0097, "step": 27350 }, { "grad_norm": 0.1387547254562378, "learning_rate": 7.844679250215671e-05, "loss": 0.0093, "step": 27360 }, { "grad_norm": 0.16731156408786774, "learning_rate": 7.842979277650898e-05, "loss": 0.0116, "step": 27370 }, { "grad_norm": 0.14196012914180756, "learning_rate": 7.84127881929872e-05, "loss": 0.012, "step": 27380 }, { "grad_norm": 0.12950557470321655, "learning_rate": 7.839577875449704e-05, "loss": 0.0117, "step": 27390 }, { "grad_norm": 0.17961491644382477, "learning_rate": 7.837876446394489e-05, "loss": 0.0107, "step": 27400 }, { "grad_norm": 0.182003915309906, "learning_rate": 7.836174532423805e-05, "loss": 0.0109, "step": 27410 }, { "grad_norm": 0.16572783887386322, "learning_rate": 7.834472133828466e-05, "loss": 0.0122, "step": 27420 }, { "grad_norm": 0.17369955778121948, "learning_rate": 7.832769250899359e-05, "loss": 0.0123, "step": 27430 }, { "grad_norm": 0.1935008317232132, "learning_rate": 7.831065883927464e-05, "loss": 0.0118, "step": 27440 }, { "grad_norm": 0.15727683901786804, "learning_rate": 7.829362033203841e-05, "loss": 0.0121, "step": 27450 }, { "grad_norm": 0.14460916817188263, "learning_rate": 7.827657699019628e-05, "loss": 0.0099, "step": 27460 }, { "grad_norm": 0.14350317418575287, "learning_rate": 7.825952881666052e-05, "loss": 0.0128, "step": 27470 }, { "grad_norm": 0.12770414352416992, "learning_rate": 7.824247581434418e-05, "loss": 0.0106, "step": 27480 }, { "grad_norm": 0.15980753302574158, "learning_rate": 7.822541798616116e-05, "loss": 0.0105, "step": 27490 }, { "grad_norm": 0.12670442461967468, "learning_rate": 7.820835533502617e-05, "loss": 0.0111, "step": 27500 }, { "grad_norm": 0.12963247299194336, "learning_rate": 7.819128786385475e-05, "loss": 0.0098, "step": 27510 }, { "grad_norm": 0.11105962842702866, "learning_rate": 7.817421557556329e-05, "loss": 0.0102, "step": 27520 }, { "grad_norm": 0.13944076001644135, "learning_rate": 7.815713847306893e-05, "loss": 0.0085, "step": 27530 }, { "grad_norm": 0.17424635589122772, "learning_rate": 7.81400565592897e-05, "loss": 0.0113, "step": 27540 }, { "grad_norm": 0.11604277789592743, "learning_rate": 7.812296983714444e-05, "loss": 0.0113, "step": 27550 }, { "grad_norm": 0.10539546608924866, "learning_rate": 7.810587830955281e-05, "loss": 0.0093, "step": 27560 }, { "grad_norm": 0.14693346619606018, "learning_rate": 7.808878197943528e-05, "loss": 0.0118, "step": 27570 }, { "grad_norm": 0.1732034534215927, "learning_rate": 7.807168084971312e-05, "loss": 0.0108, "step": 27580 }, { "grad_norm": 0.1383925974369049, "learning_rate": 7.805457492330849e-05, "loss": 0.0127, "step": 27590 }, { "grad_norm": 0.1447259485721588, "learning_rate": 7.803746420314428e-05, "loss": 0.0104, "step": 27600 }, { "grad_norm": 0.16340653598308563, "learning_rate": 7.802034869214428e-05, "loss": 0.0121, "step": 27610 }, { "grad_norm": 0.15778006613254547, "learning_rate": 7.800322839323303e-05, "loss": 0.0157, "step": 27620 }, { "grad_norm": 0.14787964522838593, "learning_rate": 7.798610330933593e-05, "loss": 0.0128, "step": 27630 }, { "grad_norm": 0.19680528342723846, "learning_rate": 7.796897344337922e-05, "loss": 0.0111, "step": 27640 }, { "grad_norm": 0.11528158187866211, "learning_rate": 7.795183879828989e-05, "loss": 0.0109, "step": 27650 }, { "grad_norm": 0.13964903354644775, "learning_rate": 7.793469937699579e-05, "loss": 0.0098, "step": 27660 }, { "grad_norm": 0.18471510708332062, "learning_rate": 7.791755518242558e-05, "loss": 0.0092, "step": 27670 }, { "grad_norm": 0.15009832382202148, "learning_rate": 7.790040621750876e-05, "loss": 0.0087, "step": 27680 }, { "grad_norm": 0.17278006672859192, "learning_rate": 7.788325248517558e-05, "loss": 0.0115, "step": 27690 }, { "grad_norm": 0.1448446363210678, "learning_rate": 7.786609398835715e-05, "loss": 0.0123, "step": 27700 }, { "grad_norm": 0.1663646548986435, "learning_rate": 7.784893072998541e-05, "loss": 0.011, "step": 27710 }, { "grad_norm": 0.2942851483821869, "learning_rate": 7.783176271299306e-05, "loss": 0.0096, "step": 27720 }, { "grad_norm": 0.14006859064102173, "learning_rate": 7.781458994031368e-05, "loss": 0.0106, "step": 27730 }, { "grad_norm": 0.16052715480327606, "learning_rate": 7.779741241488161e-05, "loss": 0.01, "step": 27740 }, { "grad_norm": 0.196994811296463, "learning_rate": 7.7780230139632e-05, "loss": 0.011, "step": 27750 }, { "grad_norm": 0.16509072482585907, "learning_rate": 7.776304311750087e-05, "loss": 0.0101, "step": 27760 }, { "grad_norm": 0.19066493213176727, "learning_rate": 7.7745851351425e-05, "loss": 0.0124, "step": 27770 }, { "grad_norm": 0.12591087818145752, "learning_rate": 7.772865484434197e-05, "loss": 0.0087, "step": 27780 }, { "grad_norm": 0.17968131601810455, "learning_rate": 7.77114535991902e-05, "loss": 0.0118, "step": 27790 }, { "grad_norm": 0.17929868400096893, "learning_rate": 7.769424761890893e-05, "loss": 0.0114, "step": 27800 }, { "grad_norm": 0.20515292882919312, "learning_rate": 7.767703690643817e-05, "loss": 0.0103, "step": 27810 }, { "grad_norm": 0.22845187783241272, "learning_rate": 7.76598214647188e-05, "loss": 0.0128, "step": 27820 }, { "grad_norm": 0.13215439021587372, "learning_rate": 7.764260129669241e-05, "loss": 0.0098, "step": 27830 }, { "grad_norm": 0.15461066365242004, "learning_rate": 7.76253764053015e-05, "loss": 0.0098, "step": 27840 }, { "grad_norm": 0.1606319099664688, "learning_rate": 7.760814679348932e-05, "loss": 0.0115, "step": 27850 }, { "grad_norm": 0.15966956317424774, "learning_rate": 7.759091246419992e-05, "loss": 0.0119, "step": 27860 }, { "grad_norm": 0.20836423337459564, "learning_rate": 7.75736734203782e-05, "loss": 0.0135, "step": 27870 }, { "grad_norm": 0.1689414232969284, "learning_rate": 7.755642966496985e-05, "loss": 0.0135, "step": 27880 }, { "grad_norm": 0.17974494397640228, "learning_rate": 7.753918120092132e-05, "loss": 0.014, "step": 27890 }, { "grad_norm": 0.18531939387321472, "learning_rate": 7.752192803117993e-05, "loss": 0.0104, "step": 27900 }, { "grad_norm": 0.15784020721912384, "learning_rate": 7.750467015869377e-05, "loss": 0.0135, "step": 27910 }, { "grad_norm": 0.18074847757816315, "learning_rate": 7.748740758641174e-05, "loss": 0.0102, "step": 27920 }, { "grad_norm": 0.19401228427886963, "learning_rate": 7.74701403172835e-05, "loss": 0.0124, "step": 27930 }, { "grad_norm": 0.18861380219459534, "learning_rate": 7.745286835425962e-05, "loss": 0.0093, "step": 27940 }, { "grad_norm": 0.20894791185855865, "learning_rate": 7.743559170029138e-05, "loss": 0.014, "step": 27950 }, { "grad_norm": 0.14832447469234467, "learning_rate": 7.741831035833087e-05, "loss": 0.0096, "step": 27960 }, { "grad_norm": 0.12301675230264664, "learning_rate": 7.740102433133102e-05, "loss": 0.0118, "step": 27970 }, { "grad_norm": 0.16569484770298004, "learning_rate": 7.738373362224553e-05, "loss": 0.0111, "step": 27980 }, { "grad_norm": 0.1490863710641861, "learning_rate": 7.73664382340289e-05, "loss": 0.0118, "step": 27990 }, { "grad_norm": 0.15703508257865906, "learning_rate": 7.734913816963647e-05, "loss": 0.0105, "step": 28000 }, { "grad_norm": 0.13458693027496338, "learning_rate": 7.73318334320243e-05, "loss": 0.0119, "step": 28010 }, { "grad_norm": 0.16071785986423492, "learning_rate": 7.731452402414934e-05, "loss": 0.01, "step": 28020 }, { "grad_norm": 0.11119303107261658, "learning_rate": 7.729720994896928e-05, "loss": 0.0106, "step": 28030 }, { "grad_norm": 0.13656987249851227, "learning_rate": 7.727989120944262e-05, "loss": 0.0105, "step": 28040 }, { "grad_norm": 0.22364383935928345, "learning_rate": 7.726256780852865e-05, "loss": 0.0131, "step": 28050 }, { "grad_norm": 0.20061752200126648, "learning_rate": 7.724523974918749e-05, "loss": 0.0103, "step": 28060 }, { "grad_norm": 0.20771253108978271, "learning_rate": 7.722790703438002e-05, "loss": 0.0098, "step": 28070 }, { "grad_norm": 0.15623918175697327, "learning_rate": 7.72105696670679e-05, "loss": 0.0102, "step": 28080 }, { "grad_norm": 0.15692897140979767, "learning_rate": 7.719322765021364e-05, "loss": 0.0079, "step": 28090 }, { "grad_norm": 0.14692972600460052, "learning_rate": 7.717588098678051e-05, "loss": 0.0135, "step": 28100 }, { "grad_norm": 0.2040403038263321, "learning_rate": 7.715852967973258e-05, "loss": 0.0087, "step": 28110 }, { "grad_norm": 0.20581883192062378, "learning_rate": 7.714117373203474e-05, "loss": 0.0102, "step": 28120 }, { "grad_norm": 0.14115287363529205, "learning_rate": 7.712381314665259e-05, "loss": 0.0094, "step": 28130 }, { "grad_norm": 0.16781611740589142, "learning_rate": 7.710644792655261e-05, "loss": 0.0104, "step": 28140 }, { "grad_norm": 0.10558803379535675, "learning_rate": 7.708907807470207e-05, "loss": 0.0083, "step": 28150 }, { "grad_norm": 0.12311101704835892, "learning_rate": 7.707170359406896e-05, "loss": 0.0104, "step": 28160 }, { "grad_norm": 0.1807747334241867, "learning_rate": 7.705432448762213e-05, "loss": 0.0123, "step": 28170 }, { "grad_norm": 0.16767765581607819, "learning_rate": 7.703694075833117e-05, "loss": 0.0105, "step": 28180 }, { "grad_norm": 0.13411863148212433, "learning_rate": 7.70195524091665e-05, "loss": 0.0133, "step": 28190 }, { "grad_norm": 0.13478484749794006, "learning_rate": 7.70021594430993e-05, "loss": 0.0104, "step": 28200 }, { "grad_norm": 0.12012676149606705, "learning_rate": 7.698476186310157e-05, "loss": 0.0109, "step": 28210 }, { "grad_norm": 0.1772623360157013, "learning_rate": 7.696735967214608e-05, "loss": 0.0116, "step": 28220 }, { "grad_norm": 0.1572984904050827, "learning_rate": 7.694995287320636e-05, "loss": 0.0108, "step": 28230 }, { "grad_norm": 0.1857835352420807, "learning_rate": 7.693254146925679e-05, "loss": 0.0117, "step": 28240 }, { "grad_norm": 0.18012382090091705, "learning_rate": 7.691512546327251e-05, "loss": 0.0116, "step": 28250 }, { "grad_norm": 0.15399475395679474, "learning_rate": 7.689770485822939e-05, "loss": 0.0124, "step": 28260 }, { "grad_norm": 0.21064592897891998, "learning_rate": 7.688027965710416e-05, "loss": 0.0097, "step": 28270 }, { "grad_norm": 0.12868797779083252, "learning_rate": 7.686284986287433e-05, "loss": 0.012, "step": 28280 }, { "grad_norm": 0.1550067961215973, "learning_rate": 7.684541547851817e-05, "loss": 0.0092, "step": 28290 }, { "grad_norm": 0.16386769711971283, "learning_rate": 7.68279765070147e-05, "loss": 0.0106, "step": 28300 }, { "grad_norm": 0.10087305307388306, "learning_rate": 7.68105329513438e-05, "loss": 0.011, "step": 28310 }, { "grad_norm": 0.13682547211647034, "learning_rate": 7.67930848144861e-05, "loss": 0.0117, "step": 28320 }, { "grad_norm": 0.1076432392001152, "learning_rate": 7.6775632099423e-05, "loss": 0.0086, "step": 28330 }, { "grad_norm": 0.12149666994810104, "learning_rate": 7.675817480913667e-05, "loss": 0.0087, "step": 28340 }, { "grad_norm": 0.17629270255565643, "learning_rate": 7.674071294661011e-05, "loss": 0.0092, "step": 28350 }, { "grad_norm": 0.132378488779068, "learning_rate": 7.672324651482707e-05, "loss": 0.0089, "step": 28360 }, { "grad_norm": 0.19200728833675385, "learning_rate": 7.670577551677209e-05, "loss": 0.0101, "step": 28370 }, { "grad_norm": 0.13827544450759888, "learning_rate": 7.668829995543047e-05, "loss": 0.0103, "step": 28380 }, { "grad_norm": 0.18231667578220367, "learning_rate": 7.667081983378832e-05, "loss": 0.0118, "step": 28390 }, { "grad_norm": 0.15584121644496918, "learning_rate": 7.66533351548325e-05, "loss": 0.009, "step": 28400 }, { "grad_norm": 0.1446003019809723, "learning_rate": 7.663584592155069e-05, "loss": 0.0145, "step": 28410 }, { "grad_norm": 0.14008557796478271, "learning_rate": 7.661835213693129e-05, "loss": 0.0101, "step": 28420 }, { "grad_norm": 0.14763212203979492, "learning_rate": 7.660085380396353e-05, "loss": 0.0111, "step": 28430 }, { "grad_norm": 0.1731070727109909, "learning_rate": 7.658335092563738e-05, "loss": 0.0113, "step": 28440 }, { "grad_norm": 0.17895396053791046, "learning_rate": 7.656584350494362e-05, "loss": 0.0107, "step": 28450 }, { "grad_norm": 0.17286106944084167, "learning_rate": 7.654833154487378e-05, "loss": 0.0104, "step": 28460 }, { "grad_norm": 0.10348524153232574, "learning_rate": 7.653081504842017e-05, "loss": 0.012, "step": 28470 }, { "grad_norm": 0.19879624247550964, "learning_rate": 7.65132940185759e-05, "loss": 0.0124, "step": 28480 }, { "grad_norm": 0.1621146947145462, "learning_rate": 7.649576845833481e-05, "loss": 0.0103, "step": 28490 }, { "grad_norm": 0.14918746054172516, "learning_rate": 7.647823837069156e-05, "loss": 0.0111, "step": 28500 }, { "grad_norm": 0.138729065656662, "learning_rate": 7.646070375864156e-05, "loss": 0.0085, "step": 28510 }, { "grad_norm": 0.12417460978031158, "learning_rate": 7.644316462518099e-05, "loss": 0.0098, "step": 28520 }, { "grad_norm": 0.16006244719028473, "learning_rate": 7.642562097330679e-05, "loss": 0.0104, "step": 28530 }, { "grad_norm": 0.156032994389534, "learning_rate": 7.640807280601671e-05, "loss": 0.01, "step": 28540 }, { "grad_norm": 0.15512673556804657, "learning_rate": 7.639052012630927e-05, "loss": 0.0108, "step": 28550 }, { "grad_norm": 0.143533855676651, "learning_rate": 7.63729629371837e-05, "loss": 0.0113, "step": 28560 }, { "grad_norm": 0.17378732562065125, "learning_rate": 7.635540124164009e-05, "loss": 0.0111, "step": 28570 }, { "grad_norm": 0.11868693679571152, "learning_rate": 7.633783504267922e-05, "loss": 0.0112, "step": 28580 }, { "grad_norm": 0.22056350111961365, "learning_rate": 7.632026434330269e-05, "loss": 0.0108, "step": 28590 }, { "grad_norm": 0.14367446303367615, "learning_rate": 7.630268914651282e-05, "loss": 0.0094, "step": 28600 }, { "grad_norm": 0.2262362539768219, "learning_rate": 7.628510945531278e-05, "loss": 0.0096, "step": 28610 }, { "grad_norm": 0.1373184621334076, "learning_rate": 7.626752527270641e-05, "loss": 0.0107, "step": 28620 }, { "grad_norm": 0.1674773246049881, "learning_rate": 7.62499366016984e-05, "loss": 0.0131, "step": 28630 }, { "grad_norm": 0.14671029150485992, "learning_rate": 7.623234344529416e-05, "loss": 0.0099, "step": 28640 }, { "grad_norm": 0.18634270131587982, "learning_rate": 7.62147458064999e-05, "loss": 0.0122, "step": 28650 }, { "grad_norm": 0.14494507014751434, "learning_rate": 7.619714368832254e-05, "loss": 0.0088, "step": 28660 }, { "grad_norm": 0.13823345303535461, "learning_rate": 7.61795370937698e-05, "loss": 0.0094, "step": 28670 }, { "grad_norm": 0.18616624176502228, "learning_rate": 7.61619260258502e-05, "loss": 0.0113, "step": 28680 }, { "grad_norm": 0.15809223055839539, "learning_rate": 7.614431048757298e-05, "loss": 0.0101, "step": 28690 }, { "grad_norm": 0.15431280434131622, "learning_rate": 7.612669048194814e-05, "loss": 0.0113, "step": 28700 }, { "grad_norm": 0.1619095504283905, "learning_rate": 7.610906601198646e-05, "loss": 0.011, "step": 28710 }, { "grad_norm": 0.11224380135536194, "learning_rate": 7.60914370806995e-05, "loss": 0.0088, "step": 28720 }, { "grad_norm": 0.18722876906394958, "learning_rate": 7.607380369109953e-05, "loss": 0.0108, "step": 28730 }, { "grad_norm": 0.1327032446861267, "learning_rate": 7.605616584619961e-05, "loss": 0.0098, "step": 28740 }, { "grad_norm": 0.17418408393859863, "learning_rate": 7.603852354901362e-05, "loss": 0.0107, "step": 28750 }, { "grad_norm": 0.18249854445457458, "learning_rate": 7.602087680255609e-05, "loss": 0.0103, "step": 28760 }, { "grad_norm": 0.15371859073638916, "learning_rate": 7.600322560984238e-05, "loss": 0.0116, "step": 28770 }, { "grad_norm": 0.16254805028438568, "learning_rate": 7.598556997388863e-05, "loss": 0.0114, "step": 28780 }, { "grad_norm": 0.16526010632514954, "learning_rate": 7.596790989771166e-05, "loss": 0.0098, "step": 28790 }, { "grad_norm": 0.157432422041893, "learning_rate": 7.595024538432914e-05, "loss": 0.0105, "step": 28800 }, { "grad_norm": 0.1458846628665924, "learning_rate": 7.59325764367594e-05, "loss": 0.0095, "step": 28810 }, { "grad_norm": 0.13108940422534943, "learning_rate": 7.59149030580216e-05, "loss": 0.0124, "step": 28820 }, { "grad_norm": 0.1135430857539177, "learning_rate": 7.589722525113562e-05, "loss": 0.0093, "step": 28830 }, { "grad_norm": 0.13523881137371063, "learning_rate": 7.587954301912216e-05, "loss": 0.0102, "step": 28840 }, { "grad_norm": 0.12403765320777893, "learning_rate": 7.586185636500263e-05, "loss": 0.0093, "step": 28850 }, { "grad_norm": 0.20183028280735016, "learning_rate": 7.584416529179914e-05, "loss": 0.0095, "step": 28860 }, { "grad_norm": 0.14890794456005096, "learning_rate": 7.582646980253465e-05, "loss": 0.0097, "step": 28870 }, { "grad_norm": 0.11258271336555481, "learning_rate": 7.580876990023282e-05, "loss": 0.009, "step": 28880 }, { "grad_norm": 0.14600394666194916, "learning_rate": 7.579106558791809e-05, "loss": 0.0087, "step": 28890 }, { "grad_norm": 0.1992729753255844, "learning_rate": 7.577335686861565e-05, "loss": 0.0119, "step": 28900 }, { "grad_norm": 0.18519796431064606, "learning_rate": 7.575564374535141e-05, "loss": 0.01, "step": 28910 }, { "grad_norm": 0.13870970904827118, "learning_rate": 7.573792622115207e-05, "loss": 0.0083, "step": 28920 }, { "grad_norm": 0.16469953954219818, "learning_rate": 7.572020429904507e-05, "loss": 0.0121, "step": 28930 }, { "grad_norm": 0.18179719150066376, "learning_rate": 7.570247798205861e-05, "loss": 0.0093, "step": 28940 }, { "grad_norm": 0.17396894097328186, "learning_rate": 7.568474727322164e-05, "loss": 0.0107, "step": 28950 }, { "grad_norm": 0.13722015917301178, "learning_rate": 7.566701217556384e-05, "loss": 0.0126, "step": 28960 }, { "grad_norm": 0.1455458253622055, "learning_rate": 7.564927269211564e-05, "loss": 0.0109, "step": 28970 }, { "grad_norm": 0.17771358788013458, "learning_rate": 7.563152882590824e-05, "loss": 0.0112, "step": 28980 }, { "grad_norm": 0.17010411620140076, "learning_rate": 7.56137805799736e-05, "loss": 0.0122, "step": 28990 }, { "grad_norm": 0.1524798572063446, "learning_rate": 7.559602795734439e-05, "loss": 0.0114, "step": 29000 }, { "grad_norm": 0.12427827715873718, "learning_rate": 7.557827096105408e-05, "loss": 0.0103, "step": 29010 }, { "grad_norm": 0.12648090720176697, "learning_rate": 7.55605095941368e-05, "loss": 0.0102, "step": 29020 }, { "grad_norm": 0.14770524203777313, "learning_rate": 7.554274385962752e-05, "loss": 0.0102, "step": 29030 }, { "grad_norm": 0.1398836225271225, "learning_rate": 7.552497376056191e-05, "loss": 0.0141, "step": 29040 }, { "grad_norm": 0.12923122942447662, "learning_rate": 7.550719929997639e-05, "loss": 0.0075, "step": 29050 }, { "grad_norm": 0.09787118434906006, "learning_rate": 7.548942048090813e-05, "loss": 0.0092, "step": 29060 }, { "grad_norm": 0.17485545575618744, "learning_rate": 7.547163730639506e-05, "loss": 0.012, "step": 29070 }, { "grad_norm": 0.13530413806438446, "learning_rate": 7.545384977947583e-05, "loss": 0.0105, "step": 29080 }, { "grad_norm": 0.16219554841518402, "learning_rate": 7.543605790318981e-05, "loss": 0.0107, "step": 29090 }, { "grad_norm": 0.1380588859319687, "learning_rate": 7.54182616805772e-05, "loss": 0.0102, "step": 29100 }, { "grad_norm": 0.18490087985992432, "learning_rate": 7.540046111467885e-05, "loss": 0.0084, "step": 29110 }, { "grad_norm": 0.19226980209350586, "learning_rate": 7.53826562085364e-05, "loss": 0.0101, "step": 29120 }, { "grad_norm": 0.13954167068004608, "learning_rate": 7.536484696519221e-05, "loss": 0.0111, "step": 29130 }, { "grad_norm": 0.1606956422328949, "learning_rate": 7.534703338768942e-05, "loss": 0.0128, "step": 29140 }, { "grad_norm": 0.20725975930690765, "learning_rate": 7.532921547907185e-05, "loss": 0.0115, "step": 29150 }, { "grad_norm": 0.24172258377075195, "learning_rate": 7.531139324238412e-05, "loss": 0.0125, "step": 29160 }, { "grad_norm": 0.18848691880702972, "learning_rate": 7.529356668067157e-05, "loss": 0.0113, "step": 29170 }, { "grad_norm": 0.15267413854599, "learning_rate": 7.527573579698023e-05, "loss": 0.0105, "step": 29180 }, { "grad_norm": 0.15675950050354004, "learning_rate": 7.525790059435693e-05, "loss": 0.0101, "step": 29190 }, { "grad_norm": 0.14229263365268707, "learning_rate": 7.524006107584926e-05, "loss": 0.0088, "step": 29200 }, { "grad_norm": 0.15308599174022675, "learning_rate": 7.522221724450544e-05, "loss": 0.0094, "step": 29210 }, { "grad_norm": 0.18155133724212646, "learning_rate": 7.520436910337451e-05, "loss": 0.0106, "step": 29220 }, { "grad_norm": 0.20511403679847717, "learning_rate": 7.518651665550627e-05, "loss": 0.0103, "step": 29230 }, { "grad_norm": 0.15947291254997253, "learning_rate": 7.516865990395117e-05, "loss": 0.0132, "step": 29240 }, { "grad_norm": 0.15080484747886658, "learning_rate": 7.515079885176047e-05, "loss": 0.0123, "step": 29250 }, { "grad_norm": 0.1447294056415558, "learning_rate": 7.513293350198612e-05, "loss": 0.009, "step": 29260 }, { "grad_norm": 0.12350407242774963, "learning_rate": 7.511506385768081e-05, "loss": 0.0091, "step": 29270 }, { "grad_norm": 0.12591268122196198, "learning_rate": 7.509718992189801e-05, "loss": 0.0099, "step": 29280 }, { "grad_norm": 0.15377360582351685, "learning_rate": 7.507931169769182e-05, "loss": 0.0118, "step": 29290 }, { "grad_norm": 0.1882753074169159, "learning_rate": 7.506142918811722e-05, "loss": 0.0115, "step": 29300 }, { "grad_norm": 0.13128386437892914, "learning_rate": 7.504354239622978e-05, "loss": 0.0106, "step": 29310 }, { "grad_norm": 0.14801619946956635, "learning_rate": 7.50256513250859e-05, "loss": 0.0103, "step": 29320 }, { "grad_norm": 0.1457812488079071, "learning_rate": 7.500775597774265e-05, "loss": 0.0109, "step": 29330 }, { "grad_norm": 0.10726290941238403, "learning_rate": 7.498985635725788e-05, "loss": 0.0087, "step": 29340 }, { "grad_norm": 0.13041557371616364, "learning_rate": 7.497195246669012e-05, "loss": 0.0108, "step": 29350 }, { "grad_norm": 0.16221880912780762, "learning_rate": 7.495404430909868e-05, "loss": 0.011, "step": 29360 }, { "grad_norm": 0.1712666004896164, "learning_rate": 7.493613188754356e-05, "loss": 0.0105, "step": 29370 }, { "grad_norm": 0.14696356654167175, "learning_rate": 7.49182152050855e-05, "loss": 0.01, "step": 29380 }, { "grad_norm": 0.1213688924908638, "learning_rate": 7.490029426478598e-05, "loss": 0.0093, "step": 29390 }, { "grad_norm": 0.20059040188789368, "learning_rate": 7.488236906970719e-05, "loss": 0.0096, "step": 29400 }, { "grad_norm": 0.18712754547595978, "learning_rate": 7.486443962291207e-05, "loss": 0.0085, "step": 29410 }, { "grad_norm": 0.11932803690433502, "learning_rate": 7.484650592746424e-05, "loss": 0.0094, "step": 29420 }, { "grad_norm": 0.1298755258321762, "learning_rate": 7.482856798642811e-05, "loss": 0.0106, "step": 29430 }, { "grad_norm": 0.16203705966472626, "learning_rate": 7.481062580286878e-05, "loss": 0.0102, "step": 29440 }, { "grad_norm": 0.13097994029521942, "learning_rate": 7.479267937985208e-05, "loss": 0.0096, "step": 29450 }, { "grad_norm": 0.17624531686306, "learning_rate": 7.477472872044456e-05, "loss": 0.0112, "step": 29460 }, { "grad_norm": 0.15426355600357056, "learning_rate": 7.475677382771347e-05, "loss": 0.0089, "step": 29470 }, { "grad_norm": 0.15122206509113312, "learning_rate": 7.473881470472683e-05, "loss": 0.0114, "step": 29480 }, { "grad_norm": 0.17568020522594452, "learning_rate": 7.472085135455337e-05, "loss": 0.0085, "step": 29490 }, { "grad_norm": 0.18514789640903473, "learning_rate": 7.470288378026256e-05, "loss": 0.0117, "step": 29500 }, { "grad_norm": 0.10680675506591797, "learning_rate": 7.468491198492451e-05, "loss": 0.0118, "step": 29510 }, { "grad_norm": 0.2207629382610321, "learning_rate": 7.466693597161013e-05, "loss": 0.0111, "step": 29520 }, { "grad_norm": 0.18693026900291443, "learning_rate": 7.464895574339104e-05, "loss": 0.0091, "step": 29530 }, { "grad_norm": 0.18029704689979553, "learning_rate": 7.463097130333958e-05, "loss": 0.0132, "step": 29540 }, { "grad_norm": 0.11827763170003891, "learning_rate": 7.461298265452876e-05, "loss": 0.0097, "step": 29550 }, { "grad_norm": 0.182710200548172, "learning_rate": 7.459498980003239e-05, "loss": 0.0108, "step": 29560 }, { "grad_norm": 0.15703251957893372, "learning_rate": 7.457699274292493e-05, "loss": 0.0094, "step": 29570 }, { "grad_norm": 0.10094356536865234, "learning_rate": 7.455899148628159e-05, "loss": 0.0086, "step": 29580 }, { "grad_norm": 0.12442620098590851, "learning_rate": 7.45409860331783e-05, "loss": 0.0094, "step": 29590 }, { "grad_norm": 0.18664568662643433, "learning_rate": 7.452297638669169e-05, "loss": 0.0113, "step": 29600 }, { "grad_norm": 0.19996654987335205, "learning_rate": 7.450496254989911e-05, "loss": 0.0123, "step": 29610 }, { "grad_norm": 0.13017259538173676, "learning_rate": 7.448694452587866e-05, "loss": 0.0106, "step": 29620 }, { "grad_norm": 0.13494651019573212, "learning_rate": 7.44689223177091e-05, "loss": 0.009, "step": 29630 }, { "grad_norm": 0.1474994570016861, "learning_rate": 7.445089592846994e-05, "loss": 0.0086, "step": 29640 }, { "grad_norm": 0.11156781017780304, "learning_rate": 7.443286536124141e-05, "loss": 0.0093, "step": 29650 }, { "grad_norm": 0.17099298536777496, "learning_rate": 7.441483061910443e-05, "loss": 0.0101, "step": 29660 }, { "grad_norm": 0.15582074224948883, "learning_rate": 7.439679170514064e-05, "loss": 0.0086, "step": 29670 }, { "grad_norm": 0.17759603261947632, "learning_rate": 7.43787486224324e-05, "loss": 0.0087, "step": 29680 }, { "grad_norm": 0.11852706223726273, "learning_rate": 7.436070137406276e-05, "loss": 0.0101, "step": 29690 }, { "grad_norm": 0.15488971769809723, "learning_rate": 7.434264996311556e-05, "loss": 0.0107, "step": 29700 }, { "grad_norm": 0.13976728916168213, "learning_rate": 7.432459439267525e-05, "loss": 0.0107, "step": 29710 }, { "grad_norm": 0.189415842294693, "learning_rate": 7.430653466582701e-05, "loss": 0.0099, "step": 29720 }, { "grad_norm": 0.16768702864646912, "learning_rate": 7.42884707856568e-05, "loss": 0.0145, "step": 29730 }, { "grad_norm": 0.14040061831474304, "learning_rate": 7.427040275525122e-05, "loss": 0.0088, "step": 29740 }, { "grad_norm": 0.14521640539169312, "learning_rate": 7.42523305776976e-05, "loss": 0.0112, "step": 29750 }, { "grad_norm": 0.11094605177640915, "learning_rate": 7.4234254256084e-05, "loss": 0.0109, "step": 29760 }, { "grad_norm": 0.1621251106262207, "learning_rate": 7.421617379349915e-05, "loss": 0.0118, "step": 29770 }, { "grad_norm": 0.18237492442131042, "learning_rate": 7.41980891930325e-05, "loss": 0.0098, "step": 29780 }, { "grad_norm": 0.1362476497888565, "learning_rate": 7.418000045777425e-05, "loss": 0.0101, "step": 29790 }, { "grad_norm": 0.16306710243225098, "learning_rate": 7.416190759081523e-05, "loss": 0.0136, "step": 29800 }, { "grad_norm": 0.1538466364145279, "learning_rate": 7.414381059524704e-05, "loss": 0.0075, "step": 29810 }, { "grad_norm": 0.12303240597248077, "learning_rate": 7.412570947416195e-05, "loss": 0.0098, "step": 29820 }, { "grad_norm": 0.15452131628990173, "learning_rate": 7.410760423065295e-05, "loss": 0.011, "step": 29830 }, { "grad_norm": 0.14393258094787598, "learning_rate": 7.408949486781372e-05, "loss": 0.0092, "step": 29840 }, { "grad_norm": 0.17425672709941864, "learning_rate": 7.407138138873868e-05, "loss": 0.0116, "step": 29850 }, { "grad_norm": 0.1395934373140335, "learning_rate": 7.405326379652292e-05, "loss": 0.0124, "step": 29860 }, { "grad_norm": 0.1532888561487198, "learning_rate": 7.403514209426222e-05, "loss": 0.01, "step": 29870 }, { "grad_norm": 0.15779347717761993, "learning_rate": 7.40170162850531e-05, "loss": 0.0121, "step": 29880 }, { "grad_norm": 0.21654589474201202, "learning_rate": 7.399888637199278e-05, "loss": 0.0146, "step": 29890 }, { "grad_norm": 0.1863301396369934, "learning_rate": 7.398075235817914e-05, "loss": 0.011, "step": 29900 }, { "grad_norm": 0.15596964955329895, "learning_rate": 7.39626142467108e-05, "loss": 0.0126, "step": 29910 }, { "grad_norm": 0.11879423260688782, "learning_rate": 7.394447204068706e-05, "loss": 0.0098, "step": 29920 }, { "grad_norm": 0.13370148837566376, "learning_rate": 7.392632574320793e-05, "loss": 0.0131, "step": 29930 }, { "grad_norm": 0.10196653753519058, "learning_rate": 7.390817535737411e-05, "loss": 0.0108, "step": 29940 }, { "grad_norm": 0.1466955840587616, "learning_rate": 7.389002088628703e-05, "loss": 0.0105, "step": 29950 }, { "grad_norm": 0.12497834116220474, "learning_rate": 7.387186233304877e-05, "loss": 0.008, "step": 29960 }, { "grad_norm": 0.19707807898521423, "learning_rate": 7.385369970076212e-05, "loss": 0.0092, "step": 29970 }, { "grad_norm": 0.1281168907880783, "learning_rate": 7.38355329925306e-05, "loss": 0.0086, "step": 29980 }, { "grad_norm": 0.14240752160549164, "learning_rate": 7.381736221145838e-05, "loss": 0.0094, "step": 29990 }, { "grad_norm": 0.12559981644153595, "learning_rate": 7.37991873606504e-05, "loss": 0.01, "step": 30000 }, { "grad_norm": 0.12247290462255478, "learning_rate": 7.378100844321218e-05, "loss": 0.009, "step": 30010 }, { "grad_norm": 0.12190864235162735, "learning_rate": 7.376282546225004e-05, "loss": 0.0132, "step": 30020 }, { "grad_norm": 0.19021061062812805, "learning_rate": 7.374463842087094e-05, "loss": 0.0103, "step": 30030 }, { "grad_norm": 0.1413145363330841, "learning_rate": 7.372644732218254e-05, "loss": 0.0093, "step": 30040 }, { "grad_norm": 0.10064435005187988, "learning_rate": 7.370825216929322e-05, "loss": 0.0108, "step": 30050 }, { "grad_norm": 0.13091261684894562, "learning_rate": 7.369005296531205e-05, "loss": 0.012, "step": 30060 }, { "grad_norm": 0.17330880463123322, "learning_rate": 7.367184971334873e-05, "loss": 0.0097, "step": 30070 }, { "grad_norm": 0.11907614767551422, "learning_rate": 7.365364241651371e-05, "loss": 0.009, "step": 30080 }, { "grad_norm": 0.13789552450180054, "learning_rate": 7.363543107791815e-05, "loss": 0.0119, "step": 30090 }, { "grad_norm": 0.15399034321308136, "learning_rate": 7.361721570067384e-05, "loss": 0.0111, "step": 30100 }, { "grad_norm": 0.19152666628360748, "learning_rate": 7.359899628789331e-05, "loss": 0.0114, "step": 30110 }, { "grad_norm": 0.18562863767147064, "learning_rate": 7.358077284268974e-05, "loss": 0.0104, "step": 30120 }, { "grad_norm": 0.2022068053483963, "learning_rate": 7.356254536817702e-05, "loss": 0.0101, "step": 30130 }, { "grad_norm": 0.18514026701450348, "learning_rate": 7.354431386746973e-05, "loss": 0.0143, "step": 30140 }, { "grad_norm": 0.16905206441879272, "learning_rate": 7.352607834368316e-05, "loss": 0.0108, "step": 30150 }, { "grad_norm": 0.14083974063396454, "learning_rate": 7.350783879993324e-05, "loss": 0.0106, "step": 30160 }, { "grad_norm": 0.13744837045669556, "learning_rate": 7.348959523933658e-05, "loss": 0.0119, "step": 30170 }, { "grad_norm": 0.15321429073810577, "learning_rate": 7.347134766501057e-05, "loss": 0.0099, "step": 30180 }, { "grad_norm": 0.1499767303466797, "learning_rate": 7.345309608007315e-05, "loss": 0.01, "step": 30190 }, { "grad_norm": 0.12421757727861404, "learning_rate": 7.343484048764308e-05, "loss": 0.0105, "step": 30200 }, { "grad_norm": 0.12865960597991943, "learning_rate": 7.341658089083972e-05, "loss": 0.0112, "step": 30210 }, { "grad_norm": 0.15955275297164917, "learning_rate": 7.339831729278313e-05, "loss": 0.0125, "step": 30220 }, { "grad_norm": 0.14201632142066956, "learning_rate": 7.338004969659404e-05, "loss": 0.0081, "step": 30230 }, { "grad_norm": 0.14257210493087769, "learning_rate": 7.336177810539391e-05, "loss": 0.0101, "step": 30240 }, { "grad_norm": 0.17603705823421478, "learning_rate": 7.334350252230485e-05, "loss": 0.0135, "step": 30250 }, { "grad_norm": 0.12620800733566284, "learning_rate": 7.332522295044965e-05, "loss": 0.0102, "step": 30260 }, { "grad_norm": 0.14642645418643951, "learning_rate": 7.33069393929518e-05, "loss": 0.0107, "step": 30270 }, { "grad_norm": 0.18030916154384613, "learning_rate": 7.328865185293545e-05, "loss": 0.0137, "step": 30280 }, { "grad_norm": 0.1480218917131424, "learning_rate": 7.327036033352546e-05, "loss": 0.0102, "step": 30290 }, { "grad_norm": 0.18247032165527344, "learning_rate": 7.325206483784733e-05, "loss": 0.01, "step": 30300 }, { "grad_norm": 0.10856296122074127, "learning_rate": 7.323376536902724e-05, "loss": 0.0091, "step": 30310 }, { "grad_norm": 0.12150953710079193, "learning_rate": 7.321546193019213e-05, "loss": 0.0111, "step": 30320 }, { "grad_norm": 0.12719030678272247, "learning_rate": 7.31971545244695e-05, "loss": 0.0096, "step": 30330 }, { "grad_norm": 0.17062142491340637, "learning_rate": 7.31788431549876e-05, "loss": 0.0114, "step": 30340 }, { "grad_norm": 0.22049644589424133, "learning_rate": 7.316052782487534e-05, "loss": 0.0105, "step": 30350 }, { "grad_norm": 0.1580382138490677, "learning_rate": 7.314220853726234e-05, "loss": 0.0117, "step": 30360 }, { "grad_norm": 0.1428556591272354, "learning_rate": 7.312388529527884e-05, "loss": 0.0111, "step": 30370 }, { "grad_norm": 0.1488487869501114, "learning_rate": 7.310555810205577e-05, "loss": 0.0095, "step": 30380 }, { "grad_norm": 0.19483499228954315, "learning_rate": 7.308722696072476e-05, "loss": 0.0103, "step": 30390 }, { "grad_norm": 0.16903670132160187, "learning_rate": 7.306889187441811e-05, "loss": 0.0083, "step": 30400 }, { "grad_norm": 0.12800252437591553, "learning_rate": 7.305055284626876e-05, "loss": 0.01, "step": 30410 }, { "grad_norm": 0.13174408674240112, "learning_rate": 7.303220987941037e-05, "loss": 0.0094, "step": 30420 }, { "grad_norm": 0.1391110122203827, "learning_rate": 7.301386297697726e-05, "loss": 0.0071, "step": 30430 }, { "grad_norm": 0.1287643313407898, "learning_rate": 7.299551214210438e-05, "loss": 0.0085, "step": 30440 }, { "grad_norm": 0.2642654776573181, "learning_rate": 7.297715737792738e-05, "loss": 0.0099, "step": 30450 }, { "grad_norm": 0.14245469868183136, "learning_rate": 7.295879868758265e-05, "loss": 0.0081, "step": 30460 }, { "grad_norm": 0.22934500873088837, "learning_rate": 7.294043607420713e-05, "loss": 0.0094, "step": 30470 }, { "grad_norm": 0.16102565824985504, "learning_rate": 7.292206954093852e-05, "loss": 0.0085, "step": 30480 }, { "grad_norm": 0.15882433950901031, "learning_rate": 7.290369909091515e-05, "loss": 0.0094, "step": 30490 }, { "grad_norm": 0.15911105275154114, "learning_rate": 7.2885324727276e-05, "loss": 0.0105, "step": 30500 }, { "grad_norm": 0.12862643599510193, "learning_rate": 7.286694645316076e-05, "loss": 0.0097, "step": 30510 }, { "grad_norm": 0.12056498974561691, "learning_rate": 7.284856427170982e-05, "loss": 0.0097, "step": 30520 }, { "grad_norm": 0.14526857435703278, "learning_rate": 7.283017818606414e-05, "loss": 0.0112, "step": 30530 }, { "grad_norm": 0.10831858962774277, "learning_rate": 7.28117881993654e-05, "loss": 0.0091, "step": 30540 }, { "grad_norm": 0.20354315638542175, "learning_rate": 7.279339431475598e-05, "loss": 0.0096, "step": 30550 }, { "grad_norm": 0.16021223366260529, "learning_rate": 7.277499653537887e-05, "loss": 0.0117, "step": 30560 }, { "grad_norm": 0.17448562383651733, "learning_rate": 7.275659486437776e-05, "loss": 0.0138, "step": 30570 }, { "grad_norm": 0.13088364899158478, "learning_rate": 7.273818930489695e-05, "loss": 0.0111, "step": 30580 }, { "grad_norm": 0.13561756908893585, "learning_rate": 7.271977986008151e-05, "loss": 0.0094, "step": 30590 }, { "grad_norm": 0.1799735426902771, "learning_rate": 7.270136653307705e-05, "loss": 0.0093, "step": 30600 }, { "grad_norm": 0.1337338089942932, "learning_rate": 7.268294932702994e-05, "loss": 0.011, "step": 30610 }, { "grad_norm": 0.15293146669864655, "learning_rate": 7.266452824508719e-05, "loss": 0.0077, "step": 30620 }, { "grad_norm": 0.12001408636569977, "learning_rate": 7.264610329039643e-05, "loss": 0.0098, "step": 30630 }, { "grad_norm": 0.16131356358528137, "learning_rate": 7.262767446610599e-05, "loss": 0.0108, "step": 30640 }, { "grad_norm": 0.09855906665325165, "learning_rate": 7.260924177536485e-05, "loss": 0.0091, "step": 30650 }, { "grad_norm": 0.16762153804302216, "learning_rate": 7.259080522132265e-05, "loss": 0.0115, "step": 30660 }, { "grad_norm": 0.11058001965284348, "learning_rate": 7.257236480712972e-05, "loss": 0.0094, "step": 30670 }, { "grad_norm": 0.13760539889335632, "learning_rate": 7.255392053593697e-05, "loss": 0.0103, "step": 30680 }, { "grad_norm": 0.12205514311790466, "learning_rate": 7.253547241089607e-05, "loss": 0.0104, "step": 30690 }, { "grad_norm": 0.16361494362354279, "learning_rate": 7.251702043515927e-05, "loss": 0.0096, "step": 30700 }, { "grad_norm": 0.13857725262641907, "learning_rate": 7.249856461187952e-05, "loss": 0.0104, "step": 30710 }, { "grad_norm": 0.13616977632045746, "learning_rate": 7.248010494421042e-05, "loss": 0.0104, "step": 30720 }, { "grad_norm": 0.1683320701122284, "learning_rate": 7.246164143530622e-05, "loss": 0.0123, "step": 30730 }, { "grad_norm": 0.1910884529352188, "learning_rate": 7.244317408832181e-05, "loss": 0.0115, "step": 30740 }, { "grad_norm": 0.11875306814908981, "learning_rate": 7.242470290641279e-05, "loss": 0.0091, "step": 30750 }, { "grad_norm": 0.17406558990478516, "learning_rate": 7.240622789273536e-05, "loss": 0.0101, "step": 30760 }, { "grad_norm": 0.173521488904953, "learning_rate": 7.238774905044638e-05, "loss": 0.0104, "step": 30770 }, { "grad_norm": 0.16208630800247192, "learning_rate": 7.236926638270341e-05, "loss": 0.0114, "step": 30780 }, { "grad_norm": 0.2094365358352661, "learning_rate": 7.23507798926646e-05, "loss": 0.0091, "step": 30790 }, { "grad_norm": 0.14164678752422333, "learning_rate": 7.23322895834888e-05, "loss": 0.0113, "step": 30800 }, { "grad_norm": 0.15087303519248962, "learning_rate": 7.231379545833552e-05, "loss": 0.0107, "step": 30810 }, { "grad_norm": 0.15373937785625458, "learning_rate": 7.229529752036487e-05, "loss": 0.0111, "step": 30820 }, { "grad_norm": 0.11169006675481796, "learning_rate": 7.227679577273765e-05, "loss": 0.0112, "step": 30830 }, { "grad_norm": 0.14221233129501343, "learning_rate": 7.225829021861529e-05, "loss": 0.0087, "step": 30840 }, { "grad_norm": 0.13900712132453918, "learning_rate": 7.223978086115992e-05, "loss": 0.0087, "step": 30850 }, { "grad_norm": 0.11942726373672485, "learning_rate": 7.222126770353425e-05, "loss": 0.0106, "step": 30860 }, { "grad_norm": 0.14966051280498505, "learning_rate": 7.22027507489017e-05, "loss": 0.0132, "step": 30870 }, { "grad_norm": 0.1862052083015442, "learning_rate": 7.218423000042627e-05, "loss": 0.0115, "step": 30880 }, { "grad_norm": 0.1558045744895935, "learning_rate": 7.216570546127268e-05, "loss": 0.0102, "step": 30890 }, { "grad_norm": 0.19751767814159393, "learning_rate": 7.214717713460626e-05, "loss": 0.0105, "step": 30900 }, { "grad_norm": 0.16769397258758545, "learning_rate": 7.2128645023593e-05, "loss": 0.0104, "step": 30910 }, { "grad_norm": 0.1680925190448761, "learning_rate": 7.211010913139951e-05, "loss": 0.0108, "step": 30920 }, { "grad_norm": 0.12405595928430557, "learning_rate": 7.209156946119308e-05, "loss": 0.0094, "step": 30930 }, { "grad_norm": 0.15723852813243866, "learning_rate": 7.207302601614166e-05, "loss": 0.0088, "step": 30940 }, { "grad_norm": 0.10489574819803238, "learning_rate": 7.205447879941378e-05, "loss": 0.0086, "step": 30950 }, { "grad_norm": 0.14458104968070984, "learning_rate": 7.203592781417866e-05, "loss": 0.0093, "step": 30960 }, { "grad_norm": 0.14104925096035004, "learning_rate": 7.201737306360617e-05, "loss": 0.0086, "step": 30970 }, { "grad_norm": 0.14155161380767822, "learning_rate": 7.19988145508668e-05, "loss": 0.0098, "step": 30980 }, { "grad_norm": 0.1269342005252838, "learning_rate": 7.198025227913168e-05, "loss": 0.012, "step": 30990 }, { "grad_norm": 0.11726046353578568, "learning_rate": 7.196168625157261e-05, "loss": 0.0093, "step": 31000 }, { "grad_norm": 0.21063566207885742, "learning_rate": 7.194311647136201e-05, "loss": 0.0109, "step": 31010 }, { "grad_norm": 0.1217351108789444, "learning_rate": 7.192454294167297e-05, "loss": 0.0103, "step": 31020 }, { "grad_norm": 0.10544305294752121, "learning_rate": 7.190596566567917e-05, "loss": 0.0084, "step": 31030 }, { "grad_norm": 0.17081066966056824, "learning_rate": 7.188738464655496e-05, "loss": 0.0125, "step": 31040 }, { "grad_norm": 0.12715192139148712, "learning_rate": 7.186879988747533e-05, "loss": 0.0098, "step": 31050 }, { "grad_norm": 0.15666955709457397, "learning_rate": 7.185021139161592e-05, "loss": 0.0093, "step": 31060 }, { "grad_norm": 0.11022387444972992, "learning_rate": 7.1831619162153e-05, "loss": 0.0098, "step": 31070 }, { "grad_norm": 0.10881957411766052, "learning_rate": 7.181302320226345e-05, "loss": 0.0107, "step": 31080 }, { "grad_norm": 0.14409098029136658, "learning_rate": 7.179442351512482e-05, "loss": 0.0082, "step": 31090 }, { "grad_norm": 0.15492083132266998, "learning_rate": 7.177582010391528e-05, "loss": 0.0106, "step": 31100 }, { "grad_norm": 0.17072349786758423, "learning_rate": 7.175721297181366e-05, "loss": 0.009, "step": 31110 }, { "grad_norm": 0.1693984568119049, "learning_rate": 7.173860212199942e-05, "loss": 0.0114, "step": 31120 }, { "grad_norm": 0.15449437499046326, "learning_rate": 7.171998755765263e-05, "loss": 0.0092, "step": 31130 }, { "grad_norm": 0.2344992756843567, "learning_rate": 7.170136928195398e-05, "loss": 0.0109, "step": 31140 }, { "grad_norm": 0.17691485583782196, "learning_rate": 7.168274729808489e-05, "loss": 0.0105, "step": 31150 }, { "grad_norm": 0.21510912477970123, "learning_rate": 7.166412160922728e-05, "loss": 0.0114, "step": 31160 }, { "grad_norm": 0.17449386417865753, "learning_rate": 7.164549221856382e-05, "loss": 0.0102, "step": 31170 }, { "grad_norm": 0.15240997076034546, "learning_rate": 7.162685912927775e-05, "loss": 0.0101, "step": 31180 }, { "grad_norm": 0.15712589025497437, "learning_rate": 7.160822234455294e-05, "loss": 0.0086, "step": 31190 }, { "grad_norm": 0.19175483286380768, "learning_rate": 7.158958186757391e-05, "loss": 0.0086, "step": 31200 }, { "grad_norm": 0.14832976460456848, "learning_rate": 7.157093770152582e-05, "loss": 0.0097, "step": 31210 }, { "grad_norm": 0.1305330991744995, "learning_rate": 7.155228984959446e-05, "loss": 0.0114, "step": 31220 }, { "grad_norm": 0.14465154707431793, "learning_rate": 7.153363831496621e-05, "loss": 0.0102, "step": 31230 }, { "grad_norm": 0.12081125378608704, "learning_rate": 7.151498310082811e-05, "loss": 0.0128, "step": 31240 }, { "grad_norm": 0.16002075374126434, "learning_rate": 7.149632421036784e-05, "loss": 0.0102, "step": 31250 }, { "grad_norm": 0.13746467232704163, "learning_rate": 7.147766164677369e-05, "loss": 0.0095, "step": 31260 }, { "grad_norm": 0.12816974520683289, "learning_rate": 7.145899541323459e-05, "loss": 0.01, "step": 31270 }, { "grad_norm": 0.1409575343132019, "learning_rate": 7.144032551294007e-05, "loss": 0.0127, "step": 31280 }, { "grad_norm": 0.15441519021987915, "learning_rate": 7.14216519490803e-05, "loss": 0.0119, "step": 31290 }, { "grad_norm": 0.19077113270759583, "learning_rate": 7.140297472484609e-05, "loss": 0.0132, "step": 31300 }, { "grad_norm": 0.11353395134210587, "learning_rate": 7.138429384342891e-05, "loss": 0.0085, "step": 31310 }, { "grad_norm": 0.1666625589132309, "learning_rate": 7.136560930802074e-05, "loss": 0.0095, "step": 31320 }, { "grad_norm": 0.145858496427536, "learning_rate": 7.134692112181431e-05, "loss": 0.0102, "step": 31330 }, { "grad_norm": 0.1586606651544571, "learning_rate": 7.13282292880029e-05, "loss": 0.009, "step": 31340 }, { "grad_norm": 0.13411352038383484, "learning_rate": 7.130953380978043e-05, "loss": 0.0109, "step": 31350 }, { "grad_norm": 0.13764946162700653, "learning_rate": 7.129083469034144e-05, "loss": 0.0111, "step": 31360 }, { "grad_norm": 0.10341604053974152, "learning_rate": 7.127213193288112e-05, "loss": 0.0115, "step": 31370 }, { "grad_norm": 0.12222939729690552, "learning_rate": 7.125342554059522e-05, "loss": 0.0097, "step": 31380 }, { "grad_norm": 0.15291906893253326, "learning_rate": 7.12347155166802e-05, "loss": 0.0125, "step": 31390 }, { "grad_norm": 0.1592274010181427, "learning_rate": 7.121600186433306e-05, "loss": 0.0103, "step": 31400 }, { "grad_norm": 0.12684722244739532, "learning_rate": 7.119728458675148e-05, "loss": 0.011, "step": 31410 }, { "grad_norm": 0.18070395290851593, "learning_rate": 7.117856368713369e-05, "loss": 0.0115, "step": 31420 }, { "grad_norm": 0.15655310451984406, "learning_rate": 7.115983916867861e-05, "loss": 0.0088, "step": 31430 }, { "grad_norm": 0.1639319807291031, "learning_rate": 7.114111103458574e-05, "loss": 0.0096, "step": 31440 }, { "grad_norm": 0.20598921179771423, "learning_rate": 7.11223792880552e-05, "loss": 0.0096, "step": 31450 }, { "grad_norm": 0.17609480023384094, "learning_rate": 7.110364393228773e-05, "loss": 0.0112, "step": 31460 }, { "grad_norm": 0.1467633992433548, "learning_rate": 7.108490497048471e-05, "loss": 0.0095, "step": 31470 }, { "grad_norm": 0.14859512448310852, "learning_rate": 7.10661624058481e-05, "loss": 0.0096, "step": 31480 }, { "grad_norm": 0.14991316199302673, "learning_rate": 7.10474162415805e-05, "loss": 0.0104, "step": 31490 }, { "grad_norm": 0.123890720307827, "learning_rate": 7.102866648088511e-05, "loss": 0.0106, "step": 31500 }, { "grad_norm": 0.14436811208724976, "learning_rate": 7.100991312696576e-05, "loss": 0.0084, "step": 31510 }, { "grad_norm": 0.14426110684871674, "learning_rate": 7.099115618302686e-05, "loss": 0.0083, "step": 31520 }, { "grad_norm": 0.12517429888248444, "learning_rate": 7.097239565227349e-05, "loss": 0.0088, "step": 31530 }, { "grad_norm": 0.1371229887008667, "learning_rate": 7.09536315379113e-05, "loss": 0.0125, "step": 31540 }, { "grad_norm": 0.14845675230026245, "learning_rate": 7.093486384314656e-05, "loss": 0.0135, "step": 31550 }, { "grad_norm": 0.16464760899543762, "learning_rate": 7.091609257118616e-05, "loss": 0.0092, "step": 31560 }, { "grad_norm": 0.11783231049776077, "learning_rate": 7.08973177252376e-05, "loss": 0.0074, "step": 31570 }, { "grad_norm": 0.1924896091222763, "learning_rate": 7.087853930850898e-05, "loss": 0.0082, "step": 31580 }, { "grad_norm": 0.15091653168201447, "learning_rate": 7.085975732420903e-05, "loss": 0.0095, "step": 31590 }, { "grad_norm": 0.18002264201641083, "learning_rate": 7.084097177554706e-05, "loss": 0.0101, "step": 31600 }, { "grad_norm": 0.18597422540187836, "learning_rate": 7.082218266573301e-05, "loss": 0.0085, "step": 31610 }, { "grad_norm": 0.20558638870716095, "learning_rate": 7.080338999797743e-05, "loss": 0.0138, "step": 31620 }, { "grad_norm": 0.154714435338974, "learning_rate": 7.07845937754915e-05, "loss": 0.0076, "step": 31630 }, { "grad_norm": 0.14681676030158997, "learning_rate": 7.076579400148693e-05, "loss": 0.0111, "step": 31640 }, { "grad_norm": 0.163153737783432, "learning_rate": 7.074699067917611e-05, "loss": 0.0106, "step": 31650 }, { "grad_norm": 0.15745000541210175, "learning_rate": 7.072818381177201e-05, "loss": 0.0105, "step": 31660 }, { "grad_norm": 0.16997988522052765, "learning_rate": 7.070937340248823e-05, "loss": 0.0102, "step": 31670 }, { "grad_norm": 0.11601332575082779, "learning_rate": 7.069055945453893e-05, "loss": 0.0097, "step": 31680 }, { "grad_norm": 0.1341714709997177, "learning_rate": 7.067174197113892e-05, "loss": 0.0085, "step": 31690 }, { "grad_norm": 0.1367424577474594, "learning_rate": 7.065292095550355e-05, "loss": 0.0105, "step": 31700 }, { "grad_norm": 0.1670609563589096, "learning_rate": 7.063409641084887e-05, "loss": 0.0086, "step": 31710 }, { "grad_norm": 0.14361847937107086, "learning_rate": 7.061526834039145e-05, "loss": 0.0078, "step": 31720 }, { "grad_norm": 0.15444575250148773, "learning_rate": 7.05964367473485e-05, "loss": 0.0091, "step": 31730 }, { "grad_norm": 0.1316268891096115, "learning_rate": 7.057760163493783e-05, "loss": 0.0098, "step": 31740 }, { "grad_norm": 0.13820184767246246, "learning_rate": 7.055876300637783e-05, "loss": 0.0103, "step": 31750 }, { "grad_norm": 0.16436459124088287, "learning_rate": 7.053992086488753e-05, "loss": 0.01, "step": 31760 }, { "grad_norm": 0.17623649537563324, "learning_rate": 7.052107521368651e-05, "loss": 0.0106, "step": 31770 }, { "grad_norm": 0.12496256828308105, "learning_rate": 7.0502226055995e-05, "loss": 0.0092, "step": 31780 }, { "grad_norm": 0.20174214243888855, "learning_rate": 7.048337339503379e-05, "loss": 0.01, "step": 31790 }, { "grad_norm": 0.2541510760784149, "learning_rate": 7.046451723402427e-05, "loss": 0.0097, "step": 31800 }, { "grad_norm": 0.13960206508636475, "learning_rate": 7.044565757618848e-05, "loss": 0.0105, "step": 31810 }, { "grad_norm": 0.11563035100698471, "learning_rate": 7.042679442474899e-05, "loss": 0.0094, "step": 31820 }, { "grad_norm": 0.11504814028739929, "learning_rate": 7.040792778292902e-05, "loss": 0.0101, "step": 31830 }, { "grad_norm": 0.15209539234638214, "learning_rate": 7.038905765395234e-05, "loss": 0.009, "step": 31840 }, { "grad_norm": 0.15141913294792175, "learning_rate": 7.037018404104334e-05, "loss": 0.0097, "step": 31850 }, { "grad_norm": 0.20797328650951385, "learning_rate": 7.035130694742702e-05, "loss": 0.0103, "step": 31860 }, { "grad_norm": 0.19751589000225067, "learning_rate": 7.033242637632897e-05, "loss": 0.0097, "step": 31870 }, { "grad_norm": 0.13620199263095856, "learning_rate": 7.031354233097534e-05, "loss": 0.0093, "step": 31880 }, { "grad_norm": 0.12577155232429504, "learning_rate": 7.029465481459289e-05, "loss": 0.0089, "step": 31890 }, { "grad_norm": 0.16605618596076965, "learning_rate": 7.027576383040898e-05, "loss": 0.0089, "step": 31900 }, { "grad_norm": 0.1550745964050293, "learning_rate": 7.025686938165159e-05, "loss": 0.0108, "step": 31910 }, { "grad_norm": 0.15096521377563477, "learning_rate": 7.023797147154924e-05, "loss": 0.0092, "step": 31920 }, { "grad_norm": 0.1647184193134308, "learning_rate": 7.021907010333111e-05, "loss": 0.0104, "step": 31930 }, { "grad_norm": 0.17099153995513916, "learning_rate": 7.020016528022685e-05, "loss": 0.0118, "step": 31940 }, { "grad_norm": 0.15949979424476624, "learning_rate": 7.018125700546683e-05, "loss": 0.0096, "step": 31950 }, { "grad_norm": 0.1600237786769867, "learning_rate": 7.016234528228196e-05, "loss": 0.0124, "step": 31960 }, { "grad_norm": 0.1547948271036148, "learning_rate": 7.014343011390372e-05, "loss": 0.0095, "step": 31970 }, { "grad_norm": 0.14490550756454468, "learning_rate": 7.01245115035642e-05, "loss": 0.0111, "step": 31980 }, { "grad_norm": 0.1215967983007431, "learning_rate": 7.010558945449606e-05, "loss": 0.008, "step": 31990 }, { "grad_norm": 0.12568768858909607, "learning_rate": 7.008666396993258e-05, "loss": 0.009, "step": 32000 }, { "grad_norm": 0.12314353883266449, "learning_rate": 7.006773505310759e-05, "loss": 0.0121, "step": 32010 }, { "grad_norm": 0.12521690130233765, "learning_rate": 7.004880270725553e-05, "loss": 0.0094, "step": 32020 }, { "grad_norm": 0.14443525671958923, "learning_rate": 7.002986693561144e-05, "loss": 0.0094, "step": 32030 }, { "grad_norm": 0.15078824758529663, "learning_rate": 7.001092774141089e-05, "loss": 0.0084, "step": 32040 }, { "grad_norm": 0.13598960638046265, "learning_rate": 6.999198512789009e-05, "loss": 0.008, "step": 32050 }, { "grad_norm": 0.12739694118499756, "learning_rate": 6.997303909828584e-05, "loss": 0.0099, "step": 32060 }, { "grad_norm": 0.17474254965782166, "learning_rate": 6.995408965583544e-05, "loss": 0.0076, "step": 32070 }, { "grad_norm": 0.17818069458007812, "learning_rate": 6.993513680377688e-05, "loss": 0.0116, "step": 32080 }, { "grad_norm": 0.12897537648677826, "learning_rate": 6.991618054534868e-05, "loss": 0.0102, "step": 32090 }, { "grad_norm": 0.15041647851467133, "learning_rate": 6.989722088378991e-05, "loss": 0.0085, "step": 32100 }, { "grad_norm": 0.09120882302522659, "learning_rate": 6.987825782234027e-05, "loss": 0.0088, "step": 32110 }, { "grad_norm": 0.1491437703371048, "learning_rate": 6.985929136424006e-05, "loss": 0.0101, "step": 32120 }, { "grad_norm": 0.20392024517059326, "learning_rate": 6.984032151273012e-05, "loss": 0.0096, "step": 32130 }, { "grad_norm": 0.11613885313272476, "learning_rate": 6.982134827105186e-05, "loss": 0.0107, "step": 32140 }, { "grad_norm": 0.12744367122650146, "learning_rate": 6.980237164244729e-05, "loss": 0.0109, "step": 32150 }, { "grad_norm": 0.14806519448757172, "learning_rate": 6.9783391630159e-05, "loss": 0.0087, "step": 32160 }, { "grad_norm": 0.19934958219528198, "learning_rate": 6.976440823743015e-05, "loss": 0.0092, "step": 32170 }, { "grad_norm": 0.17526279389858246, "learning_rate": 6.974542146750451e-05, "loss": 0.0097, "step": 32180 }, { "grad_norm": 0.1253579556941986, "learning_rate": 6.972643132362637e-05, "loss": 0.0084, "step": 32190 }, { "grad_norm": 0.09768456220626831, "learning_rate": 6.970743780904064e-05, "loss": 0.0083, "step": 32200 }, { "grad_norm": 0.11860575526952744, "learning_rate": 6.968844092699277e-05, "loss": 0.0133, "step": 32210 }, { "grad_norm": 0.15975503623485565, "learning_rate": 6.966944068072883e-05, "loss": 0.0095, "step": 32220 }, { "grad_norm": 0.14570730924606323, "learning_rate": 6.965043707349545e-05, "loss": 0.0097, "step": 32230 }, { "grad_norm": 0.15770244598388672, "learning_rate": 6.963143010853982e-05, "loss": 0.0097, "step": 32240 }, { "grad_norm": 0.14917880296707153, "learning_rate": 6.961241978910971e-05, "loss": 0.0076, "step": 32250 }, { "grad_norm": 0.1446174681186676, "learning_rate": 6.959340611845344e-05, "loss": 0.0109, "step": 32260 }, { "grad_norm": 0.13506169617176056, "learning_rate": 6.957438909981995e-05, "loss": 0.0088, "step": 32270 }, { "grad_norm": 0.17182590067386627, "learning_rate": 6.955536873645872e-05, "loss": 0.0109, "step": 32280 }, { "grad_norm": 0.2001619189977646, "learning_rate": 6.953634503161982e-05, "loss": 0.0107, "step": 32290 }, { "grad_norm": 0.10881545394659042, "learning_rate": 6.951731798855387e-05, "loss": 0.0095, "step": 32300 }, { "grad_norm": 0.1521860808134079, "learning_rate": 6.949828761051208e-05, "loss": 0.0086, "step": 32310 }, { "grad_norm": 0.12343744933605194, "learning_rate": 6.947925390074622e-05, "loss": 0.0075, "step": 32320 }, { "grad_norm": 0.1221390962600708, "learning_rate": 6.946021686250863e-05, "loss": 0.0104, "step": 32330 }, { "grad_norm": 0.12894852459430695, "learning_rate": 6.94411764990522e-05, "loss": 0.0086, "step": 32340 }, { "grad_norm": 0.0941469743847847, "learning_rate": 6.942213281363044e-05, "loss": 0.0092, "step": 32350 }, { "grad_norm": 0.15601637959480286, "learning_rate": 6.940308580949737e-05, "loss": 0.0105, "step": 32360 }, { "grad_norm": 0.16998398303985596, "learning_rate": 6.93840354899076e-05, "loss": 0.01, "step": 32370 }, { "grad_norm": 0.15261583030223846, "learning_rate": 6.936498185811633e-05, "loss": 0.0083, "step": 32380 }, { "grad_norm": 0.16543962061405182, "learning_rate": 6.93459249173793e-05, "loss": 0.0121, "step": 32390 }, { "grad_norm": 0.1368863582611084, "learning_rate": 6.932686467095279e-05, "loss": 0.0111, "step": 32400 }, { "grad_norm": 0.129129558801651, "learning_rate": 6.930780112209373e-05, "loss": 0.0131, "step": 32410 }, { "grad_norm": 0.16824014484882355, "learning_rate": 6.92887342740595e-05, "loss": 0.0102, "step": 32420 }, { "grad_norm": 0.17963621020317078, "learning_rate": 6.926966413010816e-05, "loss": 0.0091, "step": 32430 }, { "grad_norm": 0.1382555514574051, "learning_rate": 6.925059069349824e-05, "loss": 0.0087, "step": 32440 }, { "grad_norm": 0.20469003915786743, "learning_rate": 6.923151396748886e-05, "loss": 0.0101, "step": 32450 }, { "grad_norm": 0.12353713065385818, "learning_rate": 6.921243395533974e-05, "loss": 0.0109, "step": 32460 }, { "grad_norm": 0.21350930631160736, "learning_rate": 6.919335066031109e-05, "loss": 0.011, "step": 32470 }, { "grad_norm": 0.15360423922538757, "learning_rate": 6.917426408566379e-05, "loss": 0.0122, "step": 32480 }, { "grad_norm": 0.1510186791419983, "learning_rate": 6.915517423465916e-05, "loss": 0.0104, "step": 32490 }, { "grad_norm": 0.15875492990016937, "learning_rate": 6.913608111055914e-05, "loss": 0.01, "step": 32500 }, { "grad_norm": 0.12596704065799713, "learning_rate": 6.911698471662623e-05, "loss": 0.0105, "step": 32510 }, { "grad_norm": 0.190636545419693, "learning_rate": 6.90978850561235e-05, "loss": 0.0111, "step": 32520 }, { "grad_norm": 0.20660656690597534, "learning_rate": 6.907878213231454e-05, "loss": 0.0092, "step": 32530 }, { "grad_norm": 0.193669393658638, "learning_rate": 6.90596759484635e-05, "loss": 0.015, "step": 32540 }, { "grad_norm": 0.19529886543750763, "learning_rate": 6.904056650783514e-05, "loss": 0.01, "step": 32550 }, { "grad_norm": 0.11822694540023804, "learning_rate": 6.902145381369471e-05, "loss": 0.0087, "step": 32560 }, { "grad_norm": 0.1765553206205368, "learning_rate": 6.900233786930808e-05, "loss": 0.0099, "step": 32570 }, { "grad_norm": 0.17085327208042145, "learning_rate": 6.898321867794161e-05, "loss": 0.0126, "step": 32580 }, { "grad_norm": 0.10486014932394028, "learning_rate": 6.896409624286226e-05, "loss": 0.0117, "step": 32590 }, { "grad_norm": 0.11800210922956467, "learning_rate": 6.894497056733754e-05, "loss": 0.0132, "step": 32600 }, { "grad_norm": 0.1171235665678978, "learning_rate": 6.89258416546355e-05, "loss": 0.0079, "step": 32610 }, { "grad_norm": 0.17119453847408295, "learning_rate": 6.890670950802474e-05, "loss": 0.0073, "step": 32620 }, { "grad_norm": 0.10589562356472015, "learning_rate": 6.88875741307744e-05, "loss": 0.0116, "step": 32630 }, { "grad_norm": 0.12326796352863312, "learning_rate": 6.886843552615425e-05, "loss": 0.0092, "step": 32640 }, { "grad_norm": 0.15018807351589203, "learning_rate": 6.884929369743451e-05, "loss": 0.0074, "step": 32650 }, { "grad_norm": 0.1737043261528015, "learning_rate": 6.8830148647886e-05, "loss": 0.0093, "step": 32660 }, { "grad_norm": 0.11259948462247849, "learning_rate": 6.88110003807801e-05, "loss": 0.0092, "step": 32670 }, { "grad_norm": 0.18054337799549103, "learning_rate": 6.87918488993887e-05, "loss": 0.0115, "step": 32680 }, { "grad_norm": 0.15292784571647644, "learning_rate": 6.877269420698431e-05, "loss": 0.0097, "step": 32690 }, { "grad_norm": 0.1602875143289566, "learning_rate": 6.875353630683989e-05, "loss": 0.0093, "step": 32700 }, { "grad_norm": 0.1336761862039566, "learning_rate": 6.873437520222905e-05, "loss": 0.0072, "step": 32710 }, { "grad_norm": 0.15069471299648285, "learning_rate": 6.871521089642585e-05, "loss": 0.0085, "step": 32720 }, { "grad_norm": 0.1500498354434967, "learning_rate": 6.869604339270498e-05, "loss": 0.0116, "step": 32730 }, { "grad_norm": 0.11108548194169998, "learning_rate": 6.867687269434164e-05, "loss": 0.0074, "step": 32740 }, { "grad_norm": 0.16207346320152283, "learning_rate": 6.865769880461156e-05, "loss": 0.0106, "step": 32750 }, { "grad_norm": 0.15207701921463013, "learning_rate": 6.863852172679104e-05, "loss": 0.009, "step": 32760 }, { "grad_norm": 0.15926259756088257, "learning_rate": 6.861934146415693e-05, "loss": 0.0107, "step": 32770 }, { "grad_norm": 0.16672514379024506, "learning_rate": 6.86001580199866e-05, "loss": 0.0097, "step": 32780 }, { "grad_norm": 0.1964118927717209, "learning_rate": 6.858097139755798e-05, "loss": 0.0081, "step": 32790 }, { "grad_norm": 0.13912762701511383, "learning_rate": 6.856178160014955e-05, "loss": 0.0097, "step": 32800 }, { "grad_norm": 0.11338198184967041, "learning_rate": 6.85425886310403e-05, "loss": 0.0087, "step": 32810 }, { "grad_norm": 0.1454532891511917, "learning_rate": 6.852339249350979e-05, "loss": 0.0081, "step": 32820 }, { "grad_norm": 0.19506563246250153, "learning_rate": 6.850419319083812e-05, "loss": 0.0115, "step": 32830 }, { "grad_norm": 0.1583508998155594, "learning_rate": 6.848499072630592e-05, "loss": 0.0085, "step": 32840 }, { "grad_norm": 0.15419580042362213, "learning_rate": 6.846578510319439e-05, "loss": 0.0086, "step": 32850 }, { "grad_norm": 0.127447709441185, "learning_rate": 6.844657632478519e-05, "loss": 0.0084, "step": 32860 }, { "grad_norm": 0.10701601207256317, "learning_rate": 6.842736439436063e-05, "loss": 0.008, "step": 32870 }, { "grad_norm": 0.1355786919593811, "learning_rate": 6.84081493152035e-05, "loss": 0.011, "step": 32880 }, { "grad_norm": 0.09929969906806946, "learning_rate": 6.83889310905971e-05, "loss": 0.0074, "step": 32890 }, { "grad_norm": 0.17259781062602997, "learning_rate": 6.836970972382533e-05, "loss": 0.0095, "step": 32900 }, { "grad_norm": 0.1331409513950348, "learning_rate": 6.835048521817257e-05, "loss": 0.0078, "step": 32910 }, { "grad_norm": 0.18727624416351318, "learning_rate": 6.833125757692379e-05, "loss": 0.01, "step": 32920 }, { "grad_norm": 0.17636609077453613, "learning_rate": 6.831202680336441e-05, "loss": 0.0089, "step": 32930 }, { "grad_norm": 0.12117599695920944, "learning_rate": 6.829279290078052e-05, "loss": 0.0122, "step": 32940 }, { "grad_norm": 0.1145276203751564, "learning_rate": 6.827355587245863e-05, "loss": 0.0095, "step": 32950 }, { "grad_norm": 0.12285180389881134, "learning_rate": 6.82543157216858e-05, "loss": 0.0079, "step": 32960 }, { "grad_norm": 0.141244038939476, "learning_rate": 6.823507245174969e-05, "loss": 0.0077, "step": 32970 }, { "grad_norm": 0.14715351164340973, "learning_rate": 6.821582606593841e-05, "loss": 0.0088, "step": 32980 }, { "grad_norm": 0.1307455450296402, "learning_rate": 6.81965765675407e-05, "loss": 0.0106, "step": 32990 }, { "grad_norm": 0.11634676158428192, "learning_rate": 6.81773239598457e-05, "loss": 0.0099, "step": 33000 }, { "grad_norm": 0.1372176557779312, "learning_rate": 6.815806824614319e-05, "loss": 0.0085, "step": 33010 }, { "grad_norm": 0.13415268063545227, "learning_rate": 6.813880942972343e-05, "loss": 0.0098, "step": 33020 }, { "grad_norm": 0.10428724437952042, "learning_rate": 6.811954751387726e-05, "loss": 0.0107, "step": 33030 }, { "grad_norm": 0.1008610874414444, "learning_rate": 6.810028250189598e-05, "loss": 0.0101, "step": 33040 }, { "grad_norm": 0.3307223320007324, "learning_rate": 6.808101439707147e-05, "loss": 0.0112, "step": 33050 }, { "grad_norm": 0.15680155158042908, "learning_rate": 6.806174320269609e-05, "loss": 0.0105, "step": 33060 }, { "grad_norm": 0.20708544552326202, "learning_rate": 6.804246892206281e-05, "loss": 0.0088, "step": 33070 }, { "grad_norm": 0.14032046496868134, "learning_rate": 6.802319155846506e-05, "loss": 0.0113, "step": 33080 }, { "grad_norm": 0.14682486653327942, "learning_rate": 6.800391111519679e-05, "loss": 0.0112, "step": 33090 }, { "grad_norm": 0.18665967881679535, "learning_rate": 6.798462759555253e-05, "loss": 0.0107, "step": 33100 }, { "grad_norm": 0.13366474211215973, "learning_rate": 6.79653410028273e-05, "loss": 0.009, "step": 33110 }, { "grad_norm": 0.12042402476072311, "learning_rate": 6.794605134031663e-05, "loss": 0.0091, "step": 33120 }, { "grad_norm": 0.1339382827281952, "learning_rate": 6.792675861131661e-05, "loss": 0.0087, "step": 33130 }, { "grad_norm": 0.21724584698677063, "learning_rate": 6.790746281912386e-05, "loss": 0.0133, "step": 33140 }, { "grad_norm": 0.2057529091835022, "learning_rate": 6.788816396703546e-05, "loss": 0.0111, "step": 33150 }, { "grad_norm": 0.14063411951065063, "learning_rate": 6.78688620583491e-05, "loss": 0.0103, "step": 33160 }, { "grad_norm": 0.17455236613750458, "learning_rate": 6.784955709636292e-05, "loss": 0.0116, "step": 33170 }, { "grad_norm": 0.16365961730480194, "learning_rate": 6.783024908437564e-05, "loss": 0.0107, "step": 33180 }, { "grad_norm": 0.16553997993469238, "learning_rate": 6.781093802568641e-05, "loss": 0.0093, "step": 33190 }, { "grad_norm": 0.10894989967346191, "learning_rate": 6.779162392359504e-05, "loss": 0.008, "step": 33200 }, { "grad_norm": 0.17319215834140778, "learning_rate": 6.777230678140172e-05, "loss": 0.0123, "step": 33210 }, { "grad_norm": 0.14725926518440247, "learning_rate": 6.775298660240726e-05, "loss": 0.0092, "step": 33220 }, { "grad_norm": 0.13342955708503723, "learning_rate": 6.773366338991292e-05, "loss": 0.0104, "step": 33230 }, { "grad_norm": 0.15063922107219696, "learning_rate": 6.771433714722052e-05, "loss": 0.0101, "step": 33240 }, { "grad_norm": 0.1127677634358406, "learning_rate": 6.769500787763239e-05, "loss": 0.0088, "step": 33250 }, { "grad_norm": 0.13880257308483124, "learning_rate": 6.76756755844514e-05, "loss": 0.0129, "step": 33260 }, { "grad_norm": 0.1282007098197937, "learning_rate": 6.765634027098087e-05, "loss": 0.0097, "step": 33270 }, { "grad_norm": 0.13041812181472778, "learning_rate": 6.763700194052468e-05, "loss": 0.0094, "step": 33280 }, { "grad_norm": 0.15796546638011932, "learning_rate": 6.761766059638723e-05, "loss": 0.0095, "step": 33290 }, { "grad_norm": 0.1487676203250885, "learning_rate": 6.759831624187345e-05, "loss": 0.0086, "step": 33300 }, { "grad_norm": 0.12898582220077515, "learning_rate": 6.757896888028871e-05, "loss": 0.01, "step": 33310 }, { "grad_norm": 0.1506514847278595, "learning_rate": 6.7559618514939e-05, "loss": 0.0077, "step": 33320 }, { "grad_norm": 0.15945830941200256, "learning_rate": 6.754026514913073e-05, "loss": 0.0128, "step": 33330 }, { "grad_norm": 0.14082463085651398, "learning_rate": 6.752090878617087e-05, "loss": 0.0089, "step": 33340 }, { "grad_norm": 0.13778366148471832, "learning_rate": 6.75015494293669e-05, "loss": 0.0085, "step": 33350 }, { "grad_norm": 0.11462924629449844, "learning_rate": 6.74821870820268e-05, "loss": 0.0089, "step": 33360 }, { "grad_norm": 0.12538793683052063, "learning_rate": 6.746282174745907e-05, "loss": 0.0104, "step": 33370 }, { "grad_norm": 0.14479024708271027, "learning_rate": 6.744345342897271e-05, "loss": 0.0102, "step": 33380 }, { "grad_norm": 0.15631906688213348, "learning_rate": 6.742408212987724e-05, "loss": 0.0101, "step": 33390 }, { "grad_norm": 0.1307409107685089, "learning_rate": 6.740470785348269e-05, "loss": 0.0091, "step": 33400 }, { "grad_norm": 0.15634463727474213, "learning_rate": 6.738533060309958e-05, "loss": 0.0103, "step": 33410 }, { "grad_norm": 0.17863161861896515, "learning_rate": 6.736595038203894e-05, "loss": 0.008, "step": 33420 }, { "grad_norm": 0.13843084871768951, "learning_rate": 6.734656719361236e-05, "loss": 0.0091, "step": 33430 }, { "grad_norm": 0.25827521085739136, "learning_rate": 6.732718104113189e-05, "loss": 0.0136, "step": 33440 }, { "grad_norm": 0.17368008196353912, "learning_rate": 6.730779192791006e-05, "loss": 0.0092, "step": 33450 }, { "grad_norm": 0.1419590562582016, "learning_rate": 6.728839985725997e-05, "loss": 0.0081, "step": 33460 }, { "grad_norm": 0.13999846577644348, "learning_rate": 6.726900483249517e-05, "loss": 0.0115, "step": 33470 }, { "grad_norm": 0.15279605984687805, "learning_rate": 6.724960685692976e-05, "loss": 0.0116, "step": 33480 }, { "grad_norm": 0.1046672835946083, "learning_rate": 6.723020593387833e-05, "loss": 0.0083, "step": 33490 }, { "grad_norm": 0.11627921462059021, "learning_rate": 6.721080206665593e-05, "loss": 0.0082, "step": 33500 }, { "grad_norm": 0.16656610369682312, "learning_rate": 6.719139525857819e-05, "loss": 0.0125, "step": 33510 }, { "grad_norm": 0.12809346616268158, "learning_rate": 6.717198551296117e-05, "loss": 0.0079, "step": 33520 }, { "grad_norm": 0.14021438360214233, "learning_rate": 6.715257283312148e-05, "loss": 0.0112, "step": 33530 }, { "grad_norm": 0.10460172593593597, "learning_rate": 6.713315722237623e-05, "loss": 0.0072, "step": 33540 }, { "grad_norm": 0.13981465995311737, "learning_rate": 6.7113738684043e-05, "loss": 0.0096, "step": 33550 }, { "grad_norm": 0.12737508118152618, "learning_rate": 6.709431722143989e-05, "loss": 0.0085, "step": 33560 }, { "grad_norm": 0.1554042398929596, "learning_rate": 6.70748928378855e-05, "loss": 0.0081, "step": 33570 }, { "grad_norm": 0.1561456024646759, "learning_rate": 6.705546553669891e-05, "loss": 0.0093, "step": 33580 }, { "grad_norm": 0.12537553906440735, "learning_rate": 6.703603532119974e-05, "loss": 0.0096, "step": 33590 }, { "grad_norm": 0.15369457006454468, "learning_rate": 6.701660219470808e-05, "loss": 0.0085, "step": 33600 }, { "grad_norm": 0.13286104798316956, "learning_rate": 6.69971661605445e-05, "loss": 0.01, "step": 33610 }, { "grad_norm": 0.10178805142641068, "learning_rate": 6.697772722203008e-05, "loss": 0.008, "step": 33620 }, { "grad_norm": 0.15407225489616394, "learning_rate": 6.695828538248643e-05, "loss": 0.0104, "step": 33630 }, { "grad_norm": 0.12220622599124908, "learning_rate": 6.693884064523563e-05, "loss": 0.0088, "step": 33640 }, { "grad_norm": 0.14228974282741547, "learning_rate": 6.691939301360023e-05, "loss": 0.01, "step": 33650 }, { "grad_norm": 0.14268584549427032, "learning_rate": 6.689994249090333e-05, "loss": 0.0094, "step": 33660 }, { "grad_norm": 0.16623318195343018, "learning_rate": 6.688048908046845e-05, "loss": 0.0093, "step": 33670 }, { "grad_norm": 0.11787770688533783, "learning_rate": 6.686103278561969e-05, "loss": 0.0081, "step": 33680 }, { "grad_norm": 0.12796586751937866, "learning_rate": 6.684157360968156e-05, "loss": 0.0116, "step": 33690 }, { "grad_norm": 0.12233679741621017, "learning_rate": 6.682211155597911e-05, "loss": 0.0084, "step": 33700 }, { "grad_norm": 0.14471998810768127, "learning_rate": 6.680264662783789e-05, "loss": 0.0095, "step": 33710 }, { "grad_norm": 0.1308470070362091, "learning_rate": 6.678317882858391e-05, "loss": 0.0087, "step": 33720 }, { "grad_norm": 0.1231297105550766, "learning_rate": 6.67637081615437e-05, "loss": 0.0084, "step": 33730 }, { "grad_norm": 0.12459918111562729, "learning_rate": 6.674423463004427e-05, "loss": 0.0075, "step": 33740 }, { "grad_norm": 0.1356281042098999, "learning_rate": 6.672475823741308e-05, "loss": 0.0093, "step": 33750 }, { "grad_norm": 0.09535478800535202, "learning_rate": 6.670527898697811e-05, "loss": 0.0093, "step": 33760 }, { "grad_norm": 0.14758138358592987, "learning_rate": 6.668579688206788e-05, "loss": 0.0096, "step": 33770 }, { "grad_norm": 0.22482632100582123, "learning_rate": 6.666631192601131e-05, "loss": 0.0102, "step": 33780 }, { "grad_norm": 0.15489141643047333, "learning_rate": 6.664682412213785e-05, "loss": 0.0095, "step": 33790 }, { "grad_norm": 0.1673680543899536, "learning_rate": 6.662733347377745e-05, "loss": 0.0109, "step": 33800 }, { "grad_norm": 0.12424441426992416, "learning_rate": 6.660783998426051e-05, "loss": 0.0096, "step": 33810 }, { "grad_norm": 0.10256950557231903, "learning_rate": 6.658834365691794e-05, "loss": 0.0074, "step": 33820 }, { "grad_norm": 0.16971971094608307, "learning_rate": 6.656884449508115e-05, "loss": 0.0106, "step": 33830 }, { "grad_norm": 0.10503862053155899, "learning_rate": 6.654934250208198e-05, "loss": 0.0085, "step": 33840 }, { "grad_norm": 0.1243133619427681, "learning_rate": 6.65298376812528e-05, "loss": 0.0103, "step": 33850 }, { "grad_norm": 0.07928191125392914, "learning_rate": 6.651033003592646e-05, "loss": 0.0068, "step": 33860 }, { "grad_norm": 0.17052589356899261, "learning_rate": 6.649081956943626e-05, "loss": 0.0078, "step": 33870 }, { "grad_norm": 0.17748844623565674, "learning_rate": 6.647130628511604e-05, "loss": 0.0098, "step": 33880 }, { "grad_norm": 0.17784391343593597, "learning_rate": 6.645179018630005e-05, "loss": 0.0104, "step": 33890 }, { "grad_norm": 0.16446058452129364, "learning_rate": 6.643227127632309e-05, "loss": 0.0106, "step": 33900 }, { "grad_norm": 0.14555759727954865, "learning_rate": 6.641274955852038e-05, "loss": 0.0075, "step": 33910 }, { "grad_norm": 0.10893594473600388, "learning_rate": 6.639322503622768e-05, "loss": 0.0083, "step": 33920 }, { "grad_norm": 0.14448335766792297, "learning_rate": 6.637369771278116e-05, "loss": 0.0116, "step": 33930 }, { "grad_norm": 0.09004832804203033, "learning_rate": 6.635416759151751e-05, "loss": 0.0083, "step": 33940 }, { "grad_norm": 0.12230554968118668, "learning_rate": 6.633463467577394e-05, "loss": 0.0078, "step": 33950 }, { "grad_norm": 0.18234393000602722, "learning_rate": 6.631509896888803e-05, "loss": 0.0109, "step": 33960 }, { "grad_norm": 0.13875548541545868, "learning_rate": 6.629556047419794e-05, "loss": 0.0098, "step": 33970 }, { "grad_norm": 0.1650933027267456, "learning_rate": 6.627601919504223e-05, "loss": 0.012, "step": 33980 }, { "grad_norm": 0.17847949266433716, "learning_rate": 6.625647513476001e-05, "loss": 0.0089, "step": 33990 }, { "grad_norm": 0.14346329867839813, "learning_rate": 6.62369282966908e-05, "loss": 0.0107, "step": 34000 }, { "grad_norm": 0.12578707933425903, "learning_rate": 6.621737868417464e-05, "loss": 0.0081, "step": 34010 }, { "grad_norm": 0.13689404726028442, "learning_rate": 6.619782630055198e-05, "loss": 0.0093, "step": 34020 }, { "grad_norm": 0.16743294894695282, "learning_rate": 6.617827114916382e-05, "loss": 0.0119, "step": 34030 }, { "grad_norm": 0.14442627131938934, "learning_rate": 6.615871323335161e-05, "loss": 0.0097, "step": 34040 }, { "grad_norm": 0.18368712067604065, "learning_rate": 6.613915255645725e-05, "loss": 0.0092, "step": 34050 }, { "grad_norm": 0.13889211416244507, "learning_rate": 6.611958912182312e-05, "loss": 0.0122, "step": 34060 }, { "grad_norm": 0.14706556499004364, "learning_rate": 6.610002293279207e-05, "loss": 0.0091, "step": 34070 }, { "grad_norm": 0.1328681856393814, "learning_rate": 6.608045399270746e-05, "loss": 0.009, "step": 34080 }, { "grad_norm": 0.1496371030807495, "learning_rate": 6.606088230491304e-05, "loss": 0.0107, "step": 34090 }, { "grad_norm": 0.13955534994602203, "learning_rate": 6.604130787275312e-05, "loss": 0.008, "step": 34100 }, { "grad_norm": 0.15207427740097046, "learning_rate": 6.602173069957242e-05, "loss": 0.0113, "step": 34110 }, { "grad_norm": 0.15022210776805878, "learning_rate": 6.600215078871612e-05, "loss": 0.0103, "step": 34120 }, { "grad_norm": 0.15665730834007263, "learning_rate": 6.598256814352992e-05, "loss": 0.0107, "step": 34130 }, { "grad_norm": 0.13729609549045563, "learning_rate": 6.596298276735995e-05, "loss": 0.008, "step": 34140 }, { "grad_norm": 0.1320171058177948, "learning_rate": 6.594339466355282e-05, "loss": 0.0094, "step": 34150 }, { "grad_norm": 0.13610759377479553, "learning_rate": 6.592380383545558e-05, "loss": 0.0085, "step": 34160 }, { "grad_norm": 0.11617247015237808, "learning_rate": 6.590421028641577e-05, "loss": 0.0082, "step": 34170 }, { "grad_norm": 0.11724571883678436, "learning_rate": 6.588461401978143e-05, "loss": 0.0097, "step": 34180 }, { "grad_norm": 0.13037537038326263, "learning_rate": 6.586501503890099e-05, "loss": 0.0078, "step": 34190 }, { "grad_norm": 0.1661536544561386, "learning_rate": 6.584541334712338e-05, "loss": 0.008, "step": 34200 }, { "grad_norm": 0.13047325611114502, "learning_rate": 6.582580894779802e-05, "loss": 0.0076, "step": 34210 }, { "grad_norm": 0.14221909642219543, "learning_rate": 6.580620184427473e-05, "loss": 0.0089, "step": 34220 }, { "grad_norm": 0.19660213589668274, "learning_rate": 6.578659203990385e-05, "loss": 0.0091, "step": 34230 }, { "grad_norm": 0.1656796634197235, "learning_rate": 6.576697953803615e-05, "loss": 0.0107, "step": 34240 }, { "grad_norm": 0.12912112474441528, "learning_rate": 6.57473643420229e-05, "loss": 0.0098, "step": 34250 }, { "grad_norm": 0.1648557037115097, "learning_rate": 6.572774645521574e-05, "loss": 0.0107, "step": 34260 }, { "grad_norm": 0.11930884420871735, "learning_rate": 6.570812588096688e-05, "loss": 0.0093, "step": 34270 }, { "grad_norm": 0.14400029182434082, "learning_rate": 6.568850262262893e-05, "loss": 0.0102, "step": 34280 }, { "grad_norm": 0.15290462970733643, "learning_rate": 6.566887668355497e-05, "loss": 0.0094, "step": 34290 }, { "grad_norm": 0.11513786762952805, "learning_rate": 6.564924806709851e-05, "loss": 0.0077, "step": 34300 }, { "grad_norm": 0.13857384026050568, "learning_rate": 6.562961677661359e-05, "loss": 0.0076, "step": 34310 }, { "grad_norm": 0.11669662594795227, "learning_rate": 6.56099828154546e-05, "loss": 0.0105, "step": 34320 }, { "grad_norm": 0.10823102295398712, "learning_rate": 6.55903461869765e-05, "loss": 0.0082, "step": 34330 }, { "grad_norm": 0.13486427068710327, "learning_rate": 6.557070689453465e-05, "loss": 0.0073, "step": 34340 }, { "grad_norm": 0.12435145676136017, "learning_rate": 6.555106494148482e-05, "loss": 0.0097, "step": 34350 }, { "grad_norm": 0.12251950055360794, "learning_rate": 6.553142033118333e-05, "loss": 0.0101, "step": 34360 }, { "grad_norm": 0.1482730358839035, "learning_rate": 6.551177306698688e-05, "loss": 0.0082, "step": 34370 }, { "grad_norm": 0.11652394384145737, "learning_rate": 6.549212315225267e-05, "loss": 0.0092, "step": 34380 }, { "grad_norm": 0.22502011060714722, "learning_rate": 6.547247059033833e-05, "loss": 0.0132, "step": 34390 }, { "grad_norm": 0.13397333025932312, "learning_rate": 6.545281538460193e-05, "loss": 0.0083, "step": 34400 }, { "grad_norm": 0.10034628212451935, "learning_rate": 6.543315753840202e-05, "loss": 0.0091, "step": 34410 }, { "grad_norm": 0.10986277461051941, "learning_rate": 6.541349705509758e-05, "loss": 0.0081, "step": 34420 }, { "grad_norm": 0.13045291602611542, "learning_rate": 6.539383393804805e-05, "loss": 0.0096, "step": 34430 }, { "grad_norm": 0.09377928078174591, "learning_rate": 6.537416819061333e-05, "loss": 0.007, "step": 34440 }, { "grad_norm": 0.12198180705308914, "learning_rate": 6.535449981615375e-05, "loss": 0.0099, "step": 34450 }, { "grad_norm": 0.12302320450544357, "learning_rate": 6.53348288180301e-05, "loss": 0.0083, "step": 34460 }, { "grad_norm": 0.12201714515686035, "learning_rate": 6.531515519960361e-05, "loss": 0.0082, "step": 34470 }, { "grad_norm": 0.12814918160438538, "learning_rate": 6.529547896423597e-05, "loss": 0.0093, "step": 34480 }, { "grad_norm": 0.15677960216999054, "learning_rate": 6.52758001152893e-05, "loss": 0.0083, "step": 34490 }, { "grad_norm": 0.11884045600891113, "learning_rate": 6.525611865612618e-05, "loss": 0.0088, "step": 34500 }, { "grad_norm": 0.1341465562582016, "learning_rate": 6.523643459010966e-05, "loss": 0.0088, "step": 34510 }, { "grad_norm": 0.14921103417873383, "learning_rate": 6.521674792060317e-05, "loss": 0.0094, "step": 34520 }, { "grad_norm": 0.13157403469085693, "learning_rate": 6.519705865097063e-05, "loss": 0.0099, "step": 34530 }, { "grad_norm": 0.15233686566352844, "learning_rate": 6.517736678457641e-05, "loss": 0.0082, "step": 34540 }, { "grad_norm": 0.10295775532722473, "learning_rate": 6.515767232478534e-05, "loss": 0.0082, "step": 34550 }, { "grad_norm": 0.10244769603013992, "learning_rate": 6.51379752749626e-05, "loss": 0.0084, "step": 34560 }, { "grad_norm": 0.11519049108028412, "learning_rate": 6.511827563847393e-05, "loss": 0.0094, "step": 34570 }, { "grad_norm": 0.12929442524909973, "learning_rate": 6.509857341868542e-05, "loss": 0.0105, "step": 34580 }, { "grad_norm": 0.13542453944683075, "learning_rate": 6.507886861896367e-05, "loss": 0.0083, "step": 34590 }, { "grad_norm": 0.16855137050151825, "learning_rate": 6.505916124267567e-05, "loss": 0.0077, "step": 34600 }, { "grad_norm": 0.1382591724395752, "learning_rate": 6.503945129318891e-05, "loss": 0.0087, "step": 34610 }, { "grad_norm": 0.1658170372247696, "learning_rate": 6.501973877387122e-05, "loss": 0.0097, "step": 34620 }, { "grad_norm": 0.11988728493452072, "learning_rate": 6.500002368809098e-05, "loss": 0.0098, "step": 34630 }, { "grad_norm": 0.12693259119987488, "learning_rate": 6.498030603921694e-05, "loss": 0.0106, "step": 34640 }, { "grad_norm": 0.1624719202518463, "learning_rate": 6.496058583061832e-05, "loss": 0.0118, "step": 34650 }, { "grad_norm": 0.10605248808860779, "learning_rate": 6.494086306566475e-05, "loss": 0.0073, "step": 34660 }, { "grad_norm": 0.14744645357131958, "learning_rate": 6.492113774772632e-05, "loss": 0.0076, "step": 34670 }, { "grad_norm": 0.10494917631149292, "learning_rate": 6.490140988017354e-05, "loss": 0.01, "step": 34680 }, { "grad_norm": 0.15417024493217468, "learning_rate": 6.488167946637736e-05, "loss": 0.0088, "step": 34690 }, { "grad_norm": 0.14695946872234344, "learning_rate": 6.486194650970915e-05, "loss": 0.0106, "step": 34700 }, { "grad_norm": 0.11935192346572876, "learning_rate": 6.48422110135408e-05, "loss": 0.0094, "step": 34710 }, { "grad_norm": 0.1584669053554535, "learning_rate": 6.482247298124451e-05, "loss": 0.0094, "step": 34720 }, { "grad_norm": 0.1539417952299118, "learning_rate": 6.480273241619297e-05, "loss": 0.0085, "step": 34730 }, { "grad_norm": 0.14121435582637787, "learning_rate": 6.478298932175933e-05, "loss": 0.0096, "step": 34740 }, { "grad_norm": 0.14900429546833038, "learning_rate": 6.476324370131712e-05, "loss": 0.0098, "step": 34750 }, { "grad_norm": 0.12535446882247925, "learning_rate": 6.474349555824036e-05, "loss": 0.0079, "step": 34760 }, { "grad_norm": 0.15935951471328735, "learning_rate": 6.472374489590342e-05, "loss": 0.0094, "step": 34770 }, { "grad_norm": 0.1363370567560196, "learning_rate": 6.470399171768118e-05, "loss": 0.0075, "step": 34780 }, { "grad_norm": 0.14685259759426117, "learning_rate": 6.468423602694891e-05, "loss": 0.0084, "step": 34790 }, { "grad_norm": 0.1344071328639984, "learning_rate": 6.466447782708232e-05, "loss": 0.0079, "step": 34800 }, { "grad_norm": 0.1212378740310669, "learning_rate": 6.464471712145754e-05, "loss": 0.0103, "step": 34810 }, { "grad_norm": 0.15359720587730408, "learning_rate": 6.462495391345114e-05, "loss": 0.0079, "step": 34820 }, { "grad_norm": 0.12712037563323975, "learning_rate": 6.46051882064401e-05, "loss": 0.0102, "step": 34830 }, { "grad_norm": 0.16927583515644073, "learning_rate": 6.458542000380186e-05, "loss": 0.0085, "step": 34840 }, { "grad_norm": 0.16026203334331512, "learning_rate": 6.456564930891424e-05, "loss": 0.0088, "step": 34850 }, { "grad_norm": 0.12430262565612793, "learning_rate": 6.454587612515555e-05, "loss": 0.0076, "step": 34860 }, { "grad_norm": 0.1329944133758545, "learning_rate": 6.452610045590444e-05, "loss": 0.0111, "step": 34870 }, { "grad_norm": 0.12858663499355316, "learning_rate": 6.450632230454005e-05, "loss": 0.0084, "step": 34880 }, { "grad_norm": 0.12561506032943726, "learning_rate": 6.448654167444195e-05, "loss": 0.0101, "step": 34890 }, { "grad_norm": 0.1382557451725006, "learning_rate": 6.446675856899005e-05, "loss": 0.0084, "step": 34900 }, { "grad_norm": 0.15168380737304688, "learning_rate": 6.444697299156481e-05, "loss": 0.0081, "step": 34910 }, { "grad_norm": 0.12174972146749496, "learning_rate": 6.442718494554701e-05, "loss": 0.0075, "step": 34920 }, { "grad_norm": 0.17862722277641296, "learning_rate": 6.440739443431787e-05, "loss": 0.0089, "step": 34930 }, { "grad_norm": 0.14247073233127594, "learning_rate": 6.438760146125906e-05, "loss": 0.0105, "step": 34940 }, { "grad_norm": 0.12511543929576874, "learning_rate": 6.436780602975267e-05, "loss": 0.0084, "step": 34950 }, { "grad_norm": 0.1368945837020874, "learning_rate": 6.43480081431812e-05, "loss": 0.0088, "step": 34960 }, { "grad_norm": 0.12283777445554733, "learning_rate": 6.432820780492756e-05, "loss": 0.0073, "step": 34970 }, { "grad_norm": 0.12204957753419876, "learning_rate": 6.430840501837506e-05, "loss": 0.0114, "step": 34980 }, { "grad_norm": 0.12697738409042358, "learning_rate": 6.428859978690748e-05, "loss": 0.0084, "step": 34990 }, { "grad_norm": 0.1125757172703743, "learning_rate": 6.426879211390901e-05, "loss": 0.0083, "step": 35000 }, { "grad_norm": 0.14210273325443268, "learning_rate": 6.424898200276422e-05, "loss": 0.0132, "step": 35010 }, { "grad_norm": 0.2094593197107315, "learning_rate": 6.42291694568581e-05, "loss": 0.0085, "step": 35020 }, { "grad_norm": 0.13938985764980316, "learning_rate": 6.42093544795761e-05, "loss": 0.0104, "step": 35030 }, { "grad_norm": 0.16380178928375244, "learning_rate": 6.418953707430403e-05, "loss": 0.0087, "step": 35040 }, { "grad_norm": 0.14230948686599731, "learning_rate": 6.416971724442819e-05, "loss": 0.0079, "step": 35050 }, { "grad_norm": 0.12472235411405563, "learning_rate": 6.414989499333519e-05, "loss": 0.0076, "step": 35060 }, { "grad_norm": 0.15447811782360077, "learning_rate": 6.413007032441214e-05, "loss": 0.0096, "step": 35070 }, { "grad_norm": 0.13741165399551392, "learning_rate": 6.411024324104653e-05, "loss": 0.0131, "step": 35080 }, { "grad_norm": 0.13897819817066193, "learning_rate": 6.409041374662628e-05, "loss": 0.0096, "step": 35090 }, { "grad_norm": 0.1737438440322876, "learning_rate": 6.407058184453967e-05, "loss": 0.0096, "step": 35100 }, { "grad_norm": 0.10456886142492294, "learning_rate": 6.405074753817548e-05, "loss": 0.0126, "step": 35110 }, { "grad_norm": 0.1704474240541458, "learning_rate": 6.40309108309228e-05, "loss": 0.0097, "step": 35120 }, { "grad_norm": 0.2414419800043106, "learning_rate": 6.401107172617122e-05, "loss": 0.0096, "step": 35130 }, { "grad_norm": 0.1676580309867859, "learning_rate": 6.399123022731068e-05, "loss": 0.0094, "step": 35140 }, { "grad_norm": 0.10263657569885254, "learning_rate": 6.397138633773157e-05, "loss": 0.0086, "step": 35150 }, { "grad_norm": 0.1502906233072281, "learning_rate": 6.395154006082463e-05, "loss": 0.0089, "step": 35160 }, { "grad_norm": 0.13381001353263855, "learning_rate": 6.393169139998109e-05, "loss": 0.008, "step": 35170 }, { "grad_norm": 0.11768896877765656, "learning_rate": 6.39118403585925e-05, "loss": 0.0115, "step": 35180 }, { "grad_norm": 0.15243925154209137, "learning_rate": 6.38919869400509e-05, "loss": 0.0107, "step": 35190 }, { "grad_norm": 0.17871706187725067, "learning_rate": 6.387213114774865e-05, "loss": 0.0119, "step": 35200 }, { "grad_norm": 0.0913848727941513, "learning_rate": 6.385227298507863e-05, "loss": 0.0086, "step": 35210 }, { "grad_norm": 0.18187960982322693, "learning_rate": 6.3832412455434e-05, "loss": 0.0091, "step": 35220 }, { "grad_norm": 0.15215906500816345, "learning_rate": 6.381254956220841e-05, "loss": 0.0083, "step": 35230 }, { "grad_norm": 0.10002002120018005, "learning_rate": 6.379268430879586e-05, "loss": 0.0084, "step": 35240 }, { "grad_norm": 0.12205713987350464, "learning_rate": 6.37728166985908e-05, "loss": 0.0087, "step": 35250 }, { "grad_norm": 0.15849970281124115, "learning_rate": 6.375294673498804e-05, "loss": 0.0096, "step": 35260 }, { "grad_norm": 0.17920680344104767, "learning_rate": 6.373307442138284e-05, "loss": 0.0085, "step": 35270 }, { "grad_norm": 0.11312532424926758, "learning_rate": 6.371319976117081e-05, "loss": 0.0068, "step": 35280 }, { "grad_norm": 0.13220909237861633, "learning_rate": 6.3693322757748e-05, "loss": 0.0096, "step": 35290 }, { "grad_norm": 0.10413862764835358, "learning_rate": 6.367344341451086e-05, "loss": 0.0093, "step": 35300 }, { "grad_norm": 0.21206112205982208, "learning_rate": 6.36535617348562e-05, "loss": 0.0098, "step": 35310 }, { "grad_norm": 0.168814554810524, "learning_rate": 6.363367772218128e-05, "loss": 0.0101, "step": 35320 }, { "grad_norm": 0.14503447711467743, "learning_rate": 6.36137913798837e-05, "loss": 0.0106, "step": 35330 }, { "grad_norm": 0.0955452173948288, "learning_rate": 6.359390271136151e-05, "loss": 0.009, "step": 35340 }, { "grad_norm": 0.13157613575458527, "learning_rate": 6.357401172001314e-05, "loss": 0.0094, "step": 35350 }, { "grad_norm": 0.14532716572284698, "learning_rate": 6.355411840923742e-05, "loss": 0.0078, "step": 35360 }, { "grad_norm": 0.10055302083492279, "learning_rate": 6.353422278243358e-05, "loss": 0.0123, "step": 35370 }, { "grad_norm": 0.11085980385541916, "learning_rate": 6.351432484300121e-05, "loss": 0.0072, "step": 35380 }, { "grad_norm": 0.15192778408527374, "learning_rate": 6.349442459434036e-05, "loss": 0.0099, "step": 35390 }, { "grad_norm": 0.19227972626686096, "learning_rate": 6.34745220398514e-05, "loss": 0.0097, "step": 35400 }, { "grad_norm": 0.1146320253610611, "learning_rate": 6.345461718293518e-05, "loss": 0.0085, "step": 35410 }, { "grad_norm": 0.15516884624958038, "learning_rate": 6.343471002699286e-05, "loss": 0.0119, "step": 35420 }, { "grad_norm": 0.11823661625385284, "learning_rate": 6.341480057542602e-05, "loss": 0.0097, "step": 35430 }, { "grad_norm": 0.15691117942333221, "learning_rate": 6.339488883163667e-05, "loss": 0.0124, "step": 35440 }, { "grad_norm": 0.12878940999507904, "learning_rate": 6.337497479902716e-05, "loss": 0.0092, "step": 35450 }, { "grad_norm": 0.1141677275300026, "learning_rate": 6.335505848100027e-05, "loss": 0.0094, "step": 35460 }, { "grad_norm": 0.13808117806911469, "learning_rate": 6.333513988095915e-05, "loss": 0.0096, "step": 35470 }, { "grad_norm": 0.12560348212718964, "learning_rate": 6.331521900230735e-05, "loss": 0.0089, "step": 35480 }, { "grad_norm": 0.10222937166690826, "learning_rate": 6.329529584844878e-05, "loss": 0.0059, "step": 35490 }, { "grad_norm": 0.09705215692520142, "learning_rate": 6.327537042278777e-05, "loss": 0.0101, "step": 35500 }, { "grad_norm": 0.12820617854595184, "learning_rate": 6.325544272872905e-05, "loss": 0.012, "step": 35510 }, { "grad_norm": 0.10545045882463455, "learning_rate": 6.323551276967771e-05, "loss": 0.0099, "step": 35520 }, { "grad_norm": 0.15922445058822632, "learning_rate": 6.321558054903922e-05, "loss": 0.0096, "step": 35530 }, { "grad_norm": 0.156324103474617, "learning_rate": 6.319564607021947e-05, "loss": 0.0096, "step": 35540 }, { "grad_norm": 0.11224402487277985, "learning_rate": 6.31757093366247e-05, "loss": 0.008, "step": 35550 }, { "grad_norm": 0.10560940951108932, "learning_rate": 6.315577035166154e-05, "loss": 0.0092, "step": 35560 }, { "grad_norm": 0.13868668675422668, "learning_rate": 6.313582911873708e-05, "loss": 0.0097, "step": 35570 }, { "grad_norm": 0.15615582466125488, "learning_rate": 6.311588564125865e-05, "loss": 0.009, "step": 35580 }, { "grad_norm": 0.1531221717596054, "learning_rate": 6.30959399226341e-05, "loss": 0.0092, "step": 35590 }, { "grad_norm": 0.1290268748998642, "learning_rate": 6.30759919662716e-05, "loss": 0.0089, "step": 35600 }, { "grad_norm": 0.12209105491638184, "learning_rate": 6.30560417755797e-05, "loss": 0.0088, "step": 35610 }, { "grad_norm": 0.15522348880767822, "learning_rate": 6.303608935396735e-05, "loss": 0.0105, "step": 35620 }, { "grad_norm": 0.14668560028076172, "learning_rate": 6.301613470484386e-05, "loss": 0.011, "step": 35630 }, { "grad_norm": 0.2115555852651596, "learning_rate": 6.299617783161893e-05, "loss": 0.0117, "step": 35640 }, { "grad_norm": 0.17237722873687744, "learning_rate": 6.297621873770266e-05, "loss": 0.0085, "step": 35650 }, { "grad_norm": 0.1100766733288765, "learning_rate": 6.29562574265055e-05, "loss": 0.0093, "step": 35660 }, { "grad_norm": 0.10020703822374344, "learning_rate": 6.293629390143834e-05, "loss": 0.0086, "step": 35670 }, { "grad_norm": 0.15818756818771362, "learning_rate": 6.291632816591232e-05, "loss": 0.0088, "step": 35680 }, { "grad_norm": 0.1276216208934784, "learning_rate": 6.28963602233391e-05, "loss": 0.0102, "step": 35690 }, { "grad_norm": 0.1415749192237854, "learning_rate": 6.287639007713062e-05, "loss": 0.0091, "step": 35700 }, { "grad_norm": 0.12485335022211075, "learning_rate": 6.285641773069926e-05, "loss": 0.0078, "step": 35710 }, { "grad_norm": 0.18548114597797394, "learning_rate": 6.283644318745773e-05, "loss": 0.0116, "step": 35720 }, { "grad_norm": 0.1725097894668579, "learning_rate": 6.281646645081912e-05, "loss": 0.0101, "step": 35730 }, { "grad_norm": 0.1408088505268097, "learning_rate": 6.279648752419693e-05, "loss": 0.0082, "step": 35740 }, { "grad_norm": 0.12632834911346436, "learning_rate": 6.2776506411005e-05, "loss": 0.0084, "step": 35750 }, { "grad_norm": 0.12369148433208466, "learning_rate": 6.275652311465758e-05, "loss": 0.0081, "step": 35760 }, { "grad_norm": 0.13290973007678986, "learning_rate": 6.273653763856926e-05, "loss": 0.0105, "step": 35770 }, { "grad_norm": 0.10419554263353348, "learning_rate": 6.271654998615501e-05, "loss": 0.0064, "step": 35780 }, { "grad_norm": 0.17064379155635834, "learning_rate": 6.269656016083013e-05, "loss": 0.0075, "step": 35790 }, { "grad_norm": 0.15773093700408936, "learning_rate": 6.267656816601038e-05, "loss": 0.0104, "step": 35800 }, { "grad_norm": 0.13736794888973236, "learning_rate": 6.265657400511185e-05, "loss": 0.0116, "step": 35810 }, { "grad_norm": 0.1961161345243454, "learning_rate": 6.263657768155098e-05, "loss": 0.0086, "step": 35820 }, { "grad_norm": 0.12354956567287445, "learning_rate": 6.261657919874457e-05, "loss": 0.008, "step": 35830 }, { "grad_norm": 0.12546786665916443, "learning_rate": 6.259657856010986e-05, "loss": 0.0087, "step": 35840 }, { "grad_norm": 0.1086384579539299, "learning_rate": 6.257657576906439e-05, "loss": 0.0082, "step": 35850 }, { "grad_norm": 0.10780695825815201, "learning_rate": 6.255657082902609e-05, "loss": 0.0067, "step": 35860 }, { "grad_norm": 0.15165577828884125, "learning_rate": 6.253656374341325e-05, "loss": 0.0104, "step": 35870 }, { "grad_norm": 0.15838229656219482, "learning_rate": 6.251655451564457e-05, "loss": 0.0073, "step": 35880 }, { "grad_norm": 0.15244066715240479, "learning_rate": 6.249654314913902e-05, "loss": 0.0079, "step": 35890 }, { "grad_norm": 0.12485591322183609, "learning_rate": 6.247652964731604e-05, "loss": 0.0079, "step": 35900 }, { "grad_norm": 0.1341056376695633, "learning_rate": 6.245651401359537e-05, "loss": 0.0114, "step": 35910 }, { "grad_norm": 0.12241022288799286, "learning_rate": 6.243649625139715e-05, "loss": 0.0073, "step": 35920 }, { "grad_norm": 0.130552738904953, "learning_rate": 6.241647636414185e-05, "loss": 0.0088, "step": 35930 }, { "grad_norm": 0.10372506827116013, "learning_rate": 6.239645435525034e-05, "loss": 0.008, "step": 35940 }, { "grad_norm": 0.10312580317258835, "learning_rate": 6.237643022814381e-05, "loss": 0.0081, "step": 35950 }, { "grad_norm": 0.0871843546628952, "learning_rate": 6.235640398624386e-05, "loss": 0.0073, "step": 35960 }, { "grad_norm": 0.1477137804031372, "learning_rate": 6.233637563297243e-05, "loss": 0.009, "step": 35970 }, { "grad_norm": 0.12268123030662537, "learning_rate": 6.23163451717518e-05, "loss": 0.0088, "step": 35980 }, { "grad_norm": 0.17496608197689056, "learning_rate": 6.229631260600463e-05, "loss": 0.0091, "step": 35990 }, { "grad_norm": 0.12648387253284454, "learning_rate": 6.227627793915392e-05, "loss": 0.0097, "step": 36000 }, { "grad_norm": 0.16076917946338654, "learning_rate": 6.225624117462309e-05, "loss": 0.0082, "step": 36010 }, { "grad_norm": 0.17100003361701965, "learning_rate": 6.223620231583586e-05, "loss": 0.0081, "step": 36020 }, { "grad_norm": 0.11457639187574387, "learning_rate": 6.221616136621629e-05, "loss": 0.0105, "step": 36030 }, { "grad_norm": 0.09258747845888138, "learning_rate": 6.219611832918887e-05, "loss": 0.009, "step": 36040 }, { "grad_norm": 0.12518896162509918, "learning_rate": 6.217607320817838e-05, "loss": 0.0079, "step": 36050 }, { "grad_norm": 0.10506247729063034, "learning_rate": 6.215602600661001e-05, "loss": 0.0099, "step": 36060 }, { "grad_norm": 0.1907317042350769, "learning_rate": 6.213597672790925e-05, "loss": 0.0081, "step": 36070 }, { "grad_norm": 0.12940534949302673, "learning_rate": 6.2115925375502e-05, "loss": 0.0092, "step": 36080 }, { "grad_norm": 0.14913329482078552, "learning_rate": 6.209587195281447e-05, "loss": 0.0105, "step": 36090 }, { "grad_norm": 0.09381312131881714, "learning_rate": 6.207581646327324e-05, "loss": 0.0116, "step": 36100 }, { "grad_norm": 0.1018788143992424, "learning_rate": 6.205575891030526e-05, "loss": 0.0096, "step": 36110 }, { "grad_norm": 0.13465067744255066, "learning_rate": 6.203569929733781e-05, "loss": 0.0082, "step": 36120 }, { "grad_norm": 0.16407831013202667, "learning_rate": 6.201563762779852e-05, "loss": 0.0086, "step": 36130 }, { "grad_norm": 0.15172937512397766, "learning_rate": 6.199557390511538e-05, "loss": 0.0082, "step": 36140 }, { "grad_norm": 0.22558535635471344, "learning_rate": 6.197550813271675e-05, "loss": 0.0097, "step": 36150 }, { "grad_norm": 0.11398304253816605, "learning_rate": 6.195544031403131e-05, "loss": 0.0082, "step": 36160 }, { "grad_norm": 0.1176052987575531, "learning_rate": 6.19353704524881e-05, "loss": 0.0097, "step": 36170 }, { "grad_norm": 0.13120967149734497, "learning_rate": 6.191529855151652e-05, "loss": 0.0073, "step": 36180 }, { "grad_norm": 0.1001669317483902, "learning_rate": 6.189522461454629e-05, "loss": 0.0084, "step": 36190 }, { "grad_norm": 0.1761484295129776, "learning_rate": 6.187514864500752e-05, "loss": 0.007, "step": 36200 }, { "grad_norm": 0.1486390084028244, "learning_rate": 6.185507064633062e-05, "loss": 0.0071, "step": 36210 }, { "grad_norm": 0.15756072103977203, "learning_rate": 6.18349906219464e-05, "loss": 0.007, "step": 36220 }, { "grad_norm": 0.13416817784309387, "learning_rate": 6.181490857528596e-05, "loss": 0.0111, "step": 36230 }, { "grad_norm": 0.1950540393590927, "learning_rate": 6.179482450978077e-05, "loss": 0.0093, "step": 36240 }, { "grad_norm": 0.10494721680879593, "learning_rate": 6.177473842886269e-05, "loss": 0.0078, "step": 36250 }, { "grad_norm": 0.13058842718601227, "learning_rate": 6.175465033596382e-05, "loss": 0.0092, "step": 36260 }, { "grad_norm": 0.12373717129230499, "learning_rate": 6.173456023451671e-05, "loss": 0.0088, "step": 36270 }, { "grad_norm": 0.1645311564207077, "learning_rate": 6.171446812795422e-05, "loss": 0.0095, "step": 36280 }, { "grad_norm": 0.1661631464958191, "learning_rate": 6.169437401970949e-05, "loss": 0.0086, "step": 36290 }, { "grad_norm": 0.13120441138744354, "learning_rate": 6.16742779132161e-05, "loss": 0.008, "step": 36300 }, { "grad_norm": 0.12249085307121277, "learning_rate": 6.165417981190789e-05, "loss": 0.0076, "step": 36310 }, { "grad_norm": 0.13291433453559875, "learning_rate": 6.16340797192191e-05, "loss": 0.0077, "step": 36320 }, { "grad_norm": 0.16833274066448212, "learning_rate": 6.161397763858427e-05, "loss": 0.0088, "step": 36330 }, { "grad_norm": 0.18036136031150818, "learning_rate": 6.159387357343834e-05, "loss": 0.0101, "step": 36340 }, { "grad_norm": 0.12259277701377869, "learning_rate": 6.157376752721648e-05, "loss": 0.0077, "step": 36350 }, { "grad_norm": 0.07193491607904434, "learning_rate": 6.155365950335428e-05, "loss": 0.0068, "step": 36360 }, { "grad_norm": 0.13941489160060883, "learning_rate": 6.153354950528768e-05, "loss": 0.0088, "step": 36370 }, { "grad_norm": 0.1785261034965515, "learning_rate": 6.151343753645293e-05, "loss": 0.0101, "step": 36380 }, { "grad_norm": 0.13040737807750702, "learning_rate": 6.149332360028657e-05, "loss": 0.0111, "step": 36390 }, { "grad_norm": 0.15043127536773682, "learning_rate": 6.147320770022555e-05, "loss": 0.0079, "step": 36400 }, { "grad_norm": 0.18633012473583221, "learning_rate": 6.145308983970715e-05, "loss": 0.0084, "step": 36410 }, { "grad_norm": 0.13613896071910858, "learning_rate": 6.143297002216892e-05, "loss": 0.0079, "step": 36420 }, { "grad_norm": 0.1305338740348816, "learning_rate": 6.141284825104882e-05, "loss": 0.0089, "step": 36430 }, { "grad_norm": 0.18531322479248047, "learning_rate": 6.13927245297851e-05, "loss": 0.0089, "step": 36440 }, { "grad_norm": 0.18532317876815796, "learning_rate": 6.137259886181633e-05, "loss": 0.008, "step": 36450 }, { "grad_norm": 0.13682027161121368, "learning_rate": 6.135247125058145e-05, "loss": 0.0073, "step": 36460 }, { "grad_norm": 0.12773285806179047, "learning_rate": 6.133234169951974e-05, "loss": 0.0084, "step": 36470 }, { "grad_norm": 0.09094521403312683, "learning_rate": 6.131221021207078e-05, "loss": 0.0083, "step": 36480 }, { "grad_norm": 0.1165165975689888, "learning_rate": 6.129207679167448e-05, "loss": 0.0077, "step": 36490 }, { "grad_norm": 0.16297565400600433, "learning_rate": 6.127194144177109e-05, "loss": 0.0093, "step": 36500 }, { "grad_norm": 0.170453280210495, "learning_rate": 6.125180416580118e-05, "loss": 0.0069, "step": 36510 }, { "grad_norm": 0.11692190170288086, "learning_rate": 6.123166496720571e-05, "loss": 0.007, "step": 36520 }, { "grad_norm": 0.08600393682718277, "learning_rate": 6.121152384942588e-05, "loss": 0.0077, "step": 36530 }, { "grad_norm": 0.19637005031108856, "learning_rate": 6.119138081590324e-05, "loss": 0.0107, "step": 36540 }, { "grad_norm": 0.12217891216278076, "learning_rate": 6.117123587007971e-05, "loss": 0.0078, "step": 36550 }, { "grad_norm": 0.18455511331558228, "learning_rate": 6.11510890153975e-05, "loss": 0.008, "step": 36560 }, { "grad_norm": 0.12534068524837494, "learning_rate": 6.113094025529916e-05, "loss": 0.0086, "step": 36570 }, { "grad_norm": 0.1564287543296814, "learning_rate": 6.111078959322757e-05, "loss": 0.012, "step": 36580 }, { "grad_norm": 0.12697550654411316, "learning_rate": 6.109063703262592e-05, "loss": 0.0107, "step": 36590 }, { "grad_norm": 0.13358870148658752, "learning_rate": 6.107048257693772e-05, "loss": 0.0096, "step": 36600 }, { "grad_norm": 0.1329762488603592, "learning_rate": 6.105032622960683e-05, "loss": 0.0105, "step": 36610 }, { "grad_norm": 0.11209385097026825, "learning_rate": 6.103016799407743e-05, "loss": 0.0091, "step": 36620 }, { "grad_norm": 0.11796753108501434, "learning_rate": 6.1010007873793984e-05, "loss": 0.0079, "step": 36630 }, { "grad_norm": 0.19447940587997437, "learning_rate": 6.098984587220131e-05, "loss": 0.0079, "step": 36640 }, { "grad_norm": 0.11912007629871368, "learning_rate": 6.096968199274456e-05, "loss": 0.01, "step": 36650 }, { "grad_norm": 0.1283235251903534, "learning_rate": 6.0949516238869166e-05, "loss": 0.007, "step": 36660 }, { "grad_norm": 0.12733525037765503, "learning_rate": 6.092934861402092e-05, "loss": 0.0075, "step": 36670 }, { "grad_norm": 0.10976648330688477, "learning_rate": 6.0909179121645924e-05, "loss": 0.0083, "step": 36680 }, { "grad_norm": 0.10324689000844955, "learning_rate": 6.0889007765190576e-05, "loss": 0.0109, "step": 36690 }, { "grad_norm": 0.10734514892101288, "learning_rate": 6.086883454810162e-05, "loss": 0.0087, "step": 36700 }, { "grad_norm": 0.11816003173589706, "learning_rate": 6.0848659473826084e-05, "loss": 0.0084, "step": 36710 }, { "grad_norm": 0.11556678265333176, "learning_rate": 6.082848254581138e-05, "loss": 0.0092, "step": 36720 }, { "grad_norm": 0.10856729745864868, "learning_rate": 6.080830376750517e-05, "loss": 0.0077, "step": 36730 }, { "grad_norm": 0.11678047478199005, "learning_rate": 6.0788123142355445e-05, "loss": 0.0074, "step": 36740 }, { "grad_norm": 0.1029219925403595, "learning_rate": 6.076794067381052e-05, "loss": 0.0103, "step": 36750 }, { "grad_norm": 0.1252070963382721, "learning_rate": 6.074775636531905e-05, "loss": 0.0107, "step": 36760 }, { "grad_norm": 0.15547941625118256, "learning_rate": 6.072757022032997e-05, "loss": 0.0067, "step": 36770 }, { "grad_norm": 0.12304626405239105, "learning_rate": 6.070738224229253e-05, "loss": 0.008, "step": 36780 }, { "grad_norm": 0.10932935774326324, "learning_rate": 6.0687192434656314e-05, "loss": 0.009, "step": 36790 }, { "grad_norm": 0.1646518111228943, "learning_rate": 6.066700080087121e-05, "loss": 0.0103, "step": 36800 }, { "grad_norm": 0.10974925011396408, "learning_rate": 6.0646807344387424e-05, "loss": 0.0078, "step": 36810 }, { "grad_norm": 0.18214310705661774, "learning_rate": 6.062661206865543e-05, "loss": 0.0101, "step": 36820 }, { "grad_norm": 0.1692999005317688, "learning_rate": 6.06064149771261e-05, "loss": 0.0109, "step": 36830 }, { "grad_norm": 0.17869237065315247, "learning_rate": 6.058621607325051e-05, "loss": 0.0085, "step": 36840 }, { "grad_norm": 0.1419384926557541, "learning_rate": 6.056601536048014e-05, "loss": 0.0102, "step": 36850 }, { "grad_norm": 0.13223132491111755, "learning_rate": 6.0545812842266725e-05, "loss": 0.009, "step": 36860 }, { "grad_norm": 0.13946594297885895, "learning_rate": 6.052560852206232e-05, "loss": 0.0094, "step": 36870 }, { "grad_norm": 0.1396101415157318, "learning_rate": 6.05054024033193e-05, "loss": 0.0092, "step": 36880 }, { "grad_norm": 0.13158199191093445, "learning_rate": 6.048519448949032e-05, "loss": 0.0075, "step": 36890 }, { "grad_norm": 0.16905081272125244, "learning_rate": 6.046498478402839e-05, "loss": 0.0086, "step": 36900 }, { "grad_norm": 0.1605108231306076, "learning_rate": 6.044477329038677e-05, "loss": 0.0119, "step": 36910 }, { "grad_norm": 0.12447147071361542, "learning_rate": 6.042456001201906e-05, "loss": 0.0076, "step": 36920 }, { "grad_norm": 0.09576017409563065, "learning_rate": 6.040434495237917e-05, "loss": 0.0071, "step": 36930 }, { "grad_norm": 0.16420400142669678, "learning_rate": 6.0384128114921256e-05, "loss": 0.0088, "step": 36940 }, { "grad_norm": 0.17074888944625854, "learning_rate": 6.036390950309987e-05, "loss": 0.0112, "step": 36950 }, { "grad_norm": 0.13635173439979553, "learning_rate": 6.0343689120369805e-05, "loss": 0.0127, "step": 36960 }, { "grad_norm": 0.12598192691802979, "learning_rate": 6.032346697018616e-05, "loss": 0.0093, "step": 36970 }, { "grad_norm": 0.1276840716600418, "learning_rate": 6.0303243056004375e-05, "loss": 0.0082, "step": 36980 }, { "grad_norm": 0.14002707600593567, "learning_rate": 6.0283017381280136e-05, "loss": 0.008, "step": 36990 }, { "grad_norm": 0.12799382209777832, "learning_rate": 6.026278994946948e-05, "loss": 0.0106, "step": 37000 }, { "grad_norm": 0.1533951461315155, "learning_rate": 6.02425607640287e-05, "loss": 0.0089, "step": 37010 }, { "grad_norm": 0.09702210873365402, "learning_rate": 6.022232982841441e-05, "loss": 0.0072, "step": 37020 }, { "grad_norm": 0.11647437512874603, "learning_rate": 6.020209714608355e-05, "loss": 0.0077, "step": 37030 }, { "grad_norm": 0.10031693428754807, "learning_rate": 6.018186272049331e-05, "loss": 0.0109, "step": 37040 }, { "grad_norm": 0.1491677165031433, "learning_rate": 6.01616265551012e-05, "loss": 0.0098, "step": 37050 }, { "grad_norm": 0.15318743884563446, "learning_rate": 6.014138865336503e-05, "loss": 0.0107, "step": 37060 }, { "grad_norm": 0.11817675828933716, "learning_rate": 6.0121149018742905e-05, "loss": 0.0084, "step": 37070 }, { "grad_norm": 0.19897401332855225, "learning_rate": 6.010090765469325e-05, "loss": 0.0083, "step": 37080 }, { "grad_norm": 0.20964005589485168, "learning_rate": 6.008066456467473e-05, "loss": 0.0105, "step": 37090 }, { "grad_norm": 0.15269435942173004, "learning_rate": 6.0060419752146335e-05, "loss": 0.0097, "step": 37100 }, { "grad_norm": 0.12824104726314545, "learning_rate": 6.0040173220567353e-05, "loss": 0.0077, "step": 37110 }, { "grad_norm": 0.12498539686203003, "learning_rate": 6.001992497339737e-05, "loss": 0.0071, "step": 37120 }, { "grad_norm": 0.09243949502706528, "learning_rate": 5.999967501409626e-05, "loss": 0.0086, "step": 37130 }, { "grad_norm": 0.10654932260513306, "learning_rate": 5.997942334612418e-05, "loss": 0.0076, "step": 37140 }, { "grad_norm": 0.15246382355690002, "learning_rate": 5.995916997294158e-05, "loss": 0.0087, "step": 37150 }, { "grad_norm": 0.12895075976848602, "learning_rate": 5.9938914898009214e-05, "loss": 0.0085, "step": 37160 }, { "grad_norm": 0.09668994694948196, "learning_rate": 5.991865812478813e-05, "loss": 0.0097, "step": 37170 }, { "grad_norm": 0.10391804575920105, "learning_rate": 5.989839965673964e-05, "loss": 0.0091, "step": 37180 }, { "grad_norm": 0.17619097232818604, "learning_rate": 5.987813949732539e-05, "loss": 0.0117, "step": 37190 }, { "grad_norm": 0.15786287188529968, "learning_rate": 5.9857877650007255e-05, "loss": 0.0092, "step": 37200 }, { "grad_norm": 0.11621573567390442, "learning_rate": 5.983761411824744e-05, "loss": 0.0113, "step": 37210 }, { "grad_norm": 0.12975792586803436, "learning_rate": 5.981734890550844e-05, "loss": 0.0095, "step": 37220 }, { "grad_norm": 0.10192407667636871, "learning_rate": 5.979708201525301e-05, "loss": 0.0102, "step": 37230 }, { "grad_norm": 0.14386028051376343, "learning_rate": 5.977681345094422e-05, "loss": 0.008, "step": 37240 }, { "grad_norm": 0.08631476759910583, "learning_rate": 5.97565432160454e-05, "loss": 0.0083, "step": 37250 }, { "grad_norm": 0.11584726721048355, "learning_rate": 5.9736271314020186e-05, "loss": 0.0074, "step": 37260 }, { "grad_norm": 0.19242286682128906, "learning_rate": 5.971599774833251e-05, "loss": 0.0071, "step": 37270 }, { "grad_norm": 0.18834468722343445, "learning_rate": 5.9695722522446525e-05, "loss": 0.0099, "step": 37280 }, { "grad_norm": 0.11898092180490494, "learning_rate": 5.9675445639826765e-05, "loss": 0.0067, "step": 37290 }, { "grad_norm": 0.1176624745130539, "learning_rate": 5.965516710393796e-05, "loss": 0.0088, "step": 37300 }, { "grad_norm": 0.131114199757576, "learning_rate": 5.963488691824516e-05, "loss": 0.0075, "step": 37310 }, { "grad_norm": 0.14352324604988098, "learning_rate": 5.96146050862137e-05, "loss": 0.0106, "step": 37320 }, { "grad_norm": 0.1900099366903305, "learning_rate": 5.959432161130919e-05, "loss": 0.01, "step": 37330 }, { "grad_norm": 0.15076713263988495, "learning_rate": 5.9574036496997545e-05, "loss": 0.0097, "step": 37340 }, { "grad_norm": 0.1043129488825798, "learning_rate": 5.955374974674488e-05, "loss": 0.0103, "step": 37350 }, { "grad_norm": 0.12711885571479797, "learning_rate": 5.9533461364017696e-05, "loss": 0.0089, "step": 37360 }, { "grad_norm": 0.11508873850107193, "learning_rate": 5.9513171352282716e-05, "loss": 0.0066, "step": 37370 }, { "grad_norm": 0.13071036338806152, "learning_rate": 5.949287971500692e-05, "loss": 0.0088, "step": 37380 }, { "grad_norm": 0.10257460922002792, "learning_rate": 5.947258645565762e-05, "loss": 0.0098, "step": 37390 }, { "grad_norm": 0.13569973409175873, "learning_rate": 5.945229157770237e-05, "loss": 0.0082, "step": 37400 }, { "grad_norm": 0.09403935819864273, "learning_rate": 5.9431995084609006e-05, "loss": 0.0114, "step": 37410 }, { "grad_norm": 0.1188686192035675, "learning_rate": 5.941169697984564e-05, "loss": 0.0076, "step": 37420 }, { "grad_norm": 0.13915368914604187, "learning_rate": 5.9391397266880675e-05, "loss": 0.007, "step": 37430 }, { "grad_norm": 0.118211530148983, "learning_rate": 5.937109594918279e-05, "loss": 0.0109, "step": 37440 }, { "grad_norm": 0.14324310421943665, "learning_rate": 5.9350793030220884e-05, "loss": 0.0093, "step": 37450 }, { "grad_norm": 0.12435400485992432, "learning_rate": 5.933048851346421e-05, "loss": 0.0102, "step": 37460 }, { "grad_norm": 0.14176398515701294, "learning_rate": 5.931018240238224e-05, "loss": 0.0099, "step": 37470 }, { "grad_norm": 0.1493224948644638, "learning_rate": 5.928987470044471e-05, "loss": 0.0087, "step": 37480 }, { "grad_norm": 0.14318817853927612, "learning_rate": 5.9269565411121695e-05, "loss": 0.0097, "step": 37490 }, { "grad_norm": 0.157594233751297, "learning_rate": 5.924925453788347e-05, "loss": 0.0098, "step": 37500 }, { "grad_norm": 0.12711955606937408, "learning_rate": 5.92289420842006e-05, "loss": 0.0085, "step": 37510 }, { "grad_norm": 0.12676739692687988, "learning_rate": 5.9208628053543945e-05, "loss": 0.0102, "step": 37520 }, { "grad_norm": 0.1625799834728241, "learning_rate": 5.918831244938462e-05, "loss": 0.0085, "step": 37530 }, { "grad_norm": 0.11297602951526642, "learning_rate": 5.9167995275194e-05, "loss": 0.0079, "step": 37540 }, { "grad_norm": 0.13009093701839447, "learning_rate": 5.914767653444373e-05, "loss": 0.0086, "step": 37550 }, { "grad_norm": 0.08674248307943344, "learning_rate": 5.912735623060572e-05, "loss": 0.0085, "step": 37560 }, { "grad_norm": 0.12214253097772598, "learning_rate": 5.910703436715217e-05, "loss": 0.0103, "step": 37570 }, { "grad_norm": 0.12113066017627716, "learning_rate": 5.908671094755552e-05, "loss": 0.007, "step": 37580 }, { "grad_norm": 0.1475069373846054, "learning_rate": 5.906638597528851e-05, "loss": 0.0084, "step": 37590 }, { "grad_norm": 0.09455163031816483, "learning_rate": 5.9046059453824076e-05, "loss": 0.0054, "step": 37600 }, { "grad_norm": 0.12304693460464478, "learning_rate": 5.9025731386635505e-05, "loss": 0.0081, "step": 37610 }, { "grad_norm": 0.08666323870420456, "learning_rate": 5.900540177719629e-05, "loss": 0.0065, "step": 37620 }, { "grad_norm": 0.13202466070652008, "learning_rate": 5.898507062898021e-05, "loss": 0.0063, "step": 37630 }, { "grad_norm": 0.1194833293557167, "learning_rate": 5.8964737945461316e-05, "loss": 0.006, "step": 37640 }, { "grad_norm": 0.14852683246135712, "learning_rate": 5.8944403730113885e-05, "loss": 0.0088, "step": 37650 }, { "grad_norm": 0.10496039688587189, "learning_rate": 5.892406798641248e-05, "loss": 0.0102, "step": 37660 }, { "grad_norm": 0.1525622010231018, "learning_rate": 5.890373071783193e-05, "loss": 0.0094, "step": 37670 }, { "grad_norm": 0.15104328095912933, "learning_rate": 5.888339192784732e-05, "loss": 0.0107, "step": 37680 }, { "grad_norm": 0.21534445881843567, "learning_rate": 5.8863051619934003e-05, "loss": 0.0097, "step": 37690 }, { "grad_norm": 0.11742999404668808, "learning_rate": 5.8842709797567554e-05, "loss": 0.0076, "step": 37700 }, { "grad_norm": 0.13416649401187897, "learning_rate": 5.8822366464223855e-05, "loss": 0.0084, "step": 37710 }, { "grad_norm": 0.1112830638885498, "learning_rate": 5.880202162337901e-05, "loss": 0.0078, "step": 37720 }, { "grad_norm": 0.08889603614807129, "learning_rate": 5.8781675278509405e-05, "loss": 0.0073, "step": 37730 }, { "grad_norm": 0.1518532782793045, "learning_rate": 5.8761327433091696e-05, "loss": 0.009, "step": 37740 }, { "grad_norm": 0.08784615993499756, "learning_rate": 5.874097809060275e-05, "loss": 0.0086, "step": 37750 }, { "grad_norm": 0.12612704932689667, "learning_rate": 5.87206272545197e-05, "loss": 0.0085, "step": 37760 }, { "grad_norm": 0.1905517727136612, "learning_rate": 5.8700274928319955e-05, "loss": 0.0071, "step": 37770 }, { "grad_norm": 0.12246732413768768, "learning_rate": 5.867992111548118e-05, "loss": 0.0095, "step": 37780 }, { "grad_norm": 0.12845957279205322, "learning_rate": 5.865956581948131e-05, "loss": 0.007, "step": 37790 }, { "grad_norm": 0.13942645490169525, "learning_rate": 5.863920904379845e-05, "loss": 0.0075, "step": 37800 }, { "grad_norm": 0.11607850342988968, "learning_rate": 5.8618850791911064e-05, "loss": 0.0085, "step": 37810 }, { "grad_norm": 0.14103439450263977, "learning_rate": 5.859849106729779e-05, "loss": 0.008, "step": 37820 }, { "grad_norm": 0.15912018716335297, "learning_rate": 5.857812987343758e-05, "loss": 0.006, "step": 37830 }, { "grad_norm": 0.21275241672992706, "learning_rate": 5.855776721380957e-05, "loss": 0.0082, "step": 37840 }, { "grad_norm": 0.1612008661031723, "learning_rate": 5.8537403091893217e-05, "loss": 0.0102, "step": 37850 }, { "grad_norm": 0.10854454338550568, "learning_rate": 5.851703751116816e-05, "loss": 0.0069, "step": 37860 }, { "grad_norm": 0.12717851996421814, "learning_rate": 5.8496670475114336e-05, "loss": 0.0067, "step": 37870 }, { "grad_norm": 0.11966116726398468, "learning_rate": 5.84763019872119e-05, "loss": 0.0075, "step": 37880 }, { "grad_norm": 0.12926612794399261, "learning_rate": 5.845593205094131e-05, "loss": 0.0082, "step": 37890 }, { "grad_norm": 0.13883426785469055, "learning_rate": 5.843556066978318e-05, "loss": 0.0068, "step": 37900 }, { "grad_norm": 0.13230754435062408, "learning_rate": 5.8415187847218455e-05, "loss": 0.0076, "step": 37910 }, { "grad_norm": 0.14899234473705292, "learning_rate": 5.839481358672827e-05, "loss": 0.0077, "step": 37920 }, { "grad_norm": 0.14530940353870392, "learning_rate": 5.837443789179407e-05, "loss": 0.0082, "step": 37930 }, { "grad_norm": 0.1013341024518013, "learning_rate": 5.8354060765897445e-05, "loss": 0.0073, "step": 37940 }, { "grad_norm": 0.1309262365102768, "learning_rate": 5.8333682212520334e-05, "loss": 0.0073, "step": 37950 }, { "grad_norm": 0.117154560983181, "learning_rate": 5.831330223514486e-05, "loss": 0.0082, "step": 37960 }, { "grad_norm": 0.10104481875896454, "learning_rate": 5.8292920837253396e-05, "loss": 0.0092, "step": 37970 }, { "grad_norm": 0.17911477386951447, "learning_rate": 5.827253802232857e-05, "loss": 0.0077, "step": 37980 }, { "grad_norm": 0.08474351465702057, "learning_rate": 5.825215379385327e-05, "loss": 0.0064, "step": 37990 }, { "grad_norm": 0.1435088813304901, "learning_rate": 5.823176815531057e-05, "loss": 0.0108, "step": 38000 }, { "grad_norm": 0.13353709876537323, "learning_rate": 5.8211381110183826e-05, "loss": 0.0088, "step": 38010 }, { "grad_norm": 0.13132940232753754, "learning_rate": 5.8190992661956645e-05, "loss": 0.0075, "step": 38020 }, { "grad_norm": 0.13172857463359833, "learning_rate": 5.817060281411284e-05, "loss": 0.0064, "step": 38030 }, { "grad_norm": 0.09954889863729477, "learning_rate": 5.815021157013647e-05, "loss": 0.0083, "step": 38040 }, { "grad_norm": 0.14659057557582855, "learning_rate": 5.8129818933511856e-05, "loss": 0.0095, "step": 38050 }, { "grad_norm": 0.12570641934871674, "learning_rate": 5.8109424907723544e-05, "loss": 0.0076, "step": 38060 }, { "grad_norm": 0.12568943202495575, "learning_rate": 5.80890294962563e-05, "loss": 0.0066, "step": 38070 }, { "grad_norm": 0.15419664978981018, "learning_rate": 5.806863270259515e-05, "loss": 0.0088, "step": 38080 }, { "grad_norm": 0.11768282204866409, "learning_rate": 5.804823453022536e-05, "loss": 0.0104, "step": 38090 }, { "grad_norm": 0.1297290176153183, "learning_rate": 5.80278349826324e-05, "loss": 0.0112, "step": 38100 }, { "grad_norm": 0.08925539255142212, "learning_rate": 5.8007434063302014e-05, "loss": 0.0084, "step": 38110 }, { "grad_norm": 0.1770194172859192, "learning_rate": 5.7987031775720136e-05, "loss": 0.0101, "step": 38120 }, { "grad_norm": 0.19221223890781403, "learning_rate": 5.7966628123372976e-05, "loss": 0.0104, "step": 38130 }, { "grad_norm": 0.14225485920906067, "learning_rate": 5.7946223109746956e-05, "loss": 0.0073, "step": 38140 }, { "grad_norm": 0.1494607925415039, "learning_rate": 5.7925816738328754e-05, "loss": 0.0074, "step": 38150 }, { "grad_norm": 0.13243231177330017, "learning_rate": 5.790540901260521e-05, "loss": 0.0085, "step": 38160 }, { "grad_norm": 0.13112181425094604, "learning_rate": 5.788499993606351e-05, "loss": 0.0101, "step": 38170 }, { "grad_norm": 0.10225106030702591, "learning_rate": 5.786458951219096e-05, "loss": 0.0067, "step": 38180 }, { "grad_norm": 0.16236044466495514, "learning_rate": 5.784417774447517e-05, "loss": 0.0088, "step": 38190 }, { "grad_norm": 0.1082768514752388, "learning_rate": 5.782376463640393e-05, "loss": 0.0074, "step": 38200 }, { "grad_norm": 0.13239072263240814, "learning_rate": 5.780335019146531e-05, "loss": 0.0079, "step": 38210 }, { "grad_norm": 0.15268808603286743, "learning_rate": 5.778293441314755e-05, "loss": 0.0084, "step": 38220 }, { "grad_norm": 0.11625352501869202, "learning_rate": 5.776251730493917e-05, "loss": 0.0097, "step": 38230 }, { "grad_norm": 0.14770077168941498, "learning_rate": 5.774209887032887e-05, "loss": 0.0078, "step": 38240 }, { "grad_norm": 0.1503550410270691, "learning_rate": 5.772167911280565e-05, "loss": 0.01, "step": 38250 }, { "grad_norm": 0.1302075833082199, "learning_rate": 5.770125803585864e-05, "loss": 0.0092, "step": 38260 }, { "grad_norm": 0.1452380120754242, "learning_rate": 5.768083564297726e-05, "loss": 0.0083, "step": 38270 }, { "grad_norm": 0.16805988550186157, "learning_rate": 5.766041193765114e-05, "loss": 0.0082, "step": 38280 }, { "grad_norm": 0.15298278629779816, "learning_rate": 5.763998692337015e-05, "loss": 0.0097, "step": 38290 }, { "grad_norm": 0.1031808853149414, "learning_rate": 5.761956060362433e-05, "loss": 0.0087, "step": 38300 }, { "grad_norm": 0.15460820496082306, "learning_rate": 5.7599132981904e-05, "loss": 0.0089, "step": 38310 }, { "grad_norm": 0.09165116399526596, "learning_rate": 5.75787040616997e-05, "loss": 0.0072, "step": 38320 }, { "grad_norm": 0.1321437954902649, "learning_rate": 5.755827384650212e-05, "loss": 0.0099, "step": 38330 }, { "grad_norm": 0.10542434453964233, "learning_rate": 5.753784233980228e-05, "loss": 0.0123, "step": 38340 }, { "grad_norm": 0.12611450254917145, "learning_rate": 5.751740954509135e-05, "loss": 0.0076, "step": 38350 }, { "grad_norm": 0.13851965963840485, "learning_rate": 5.7496975465860715e-05, "loss": 0.0094, "step": 38360 }, { "grad_norm": 0.2052443027496338, "learning_rate": 5.747654010560202e-05, "loss": 0.0101, "step": 38370 }, { "grad_norm": 0.23639099299907684, "learning_rate": 5.7456103467807097e-05, "loss": 0.009, "step": 38380 }, { "grad_norm": 0.14510086178779602, "learning_rate": 5.7435665555968046e-05, "loss": 0.0073, "step": 38390 }, { "grad_norm": 0.13127638399600983, "learning_rate": 5.74152263735771e-05, "loss": 0.0078, "step": 38400 }, { "grad_norm": 0.14711689949035645, "learning_rate": 5.739478592412677e-05, "loss": 0.0097, "step": 38410 }, { "grad_norm": 0.1606992483139038, "learning_rate": 5.7374344211109766e-05, "loss": 0.0092, "step": 38420 }, { "grad_norm": 0.14986790716648102, "learning_rate": 5.735390123801904e-05, "loss": 0.0083, "step": 38430 }, { "grad_norm": 0.13546369969844818, "learning_rate": 5.7333457008347704e-05, "loss": 0.007, "step": 38440 }, { "grad_norm": 0.11763894557952881, "learning_rate": 5.7313011525589156e-05, "loss": 0.0083, "step": 38450 }, { "grad_norm": 0.09327908605337143, "learning_rate": 5.729256479323694e-05, "loss": 0.0098, "step": 38460 }, { "grad_norm": 0.1461433321237564, "learning_rate": 5.727211681478485e-05, "loss": 0.0093, "step": 38470 }, { "grad_norm": 0.1315276026725769, "learning_rate": 5.7251667593726886e-05, "loss": 0.0086, "step": 38480 }, { "grad_norm": 0.14127832651138306, "learning_rate": 5.723121713355728e-05, "loss": 0.0097, "step": 38490 }, { "grad_norm": 0.10474518686532974, "learning_rate": 5.721076543777044e-05, "loss": 0.0101, "step": 38500 }, { "grad_norm": 0.1348392218351364, "learning_rate": 5.7190312509860986e-05, "loss": 0.0099, "step": 38510 }, { "grad_norm": 0.12232063710689545, "learning_rate": 5.716985835332379e-05, "loss": 0.0068, "step": 38520 }, { "grad_norm": 0.1320660412311554, "learning_rate": 5.714940297165389e-05, "loss": 0.0085, "step": 38530 }, { "grad_norm": 0.1408431977033615, "learning_rate": 5.712894636834656e-05, "loss": 0.0101, "step": 38540 }, { "grad_norm": 0.1250707060098648, "learning_rate": 5.7108488546897276e-05, "loss": 0.0101, "step": 38550 }, { "grad_norm": 0.12276914715766907, "learning_rate": 5.708802951080172e-05, "loss": 0.0067, "step": 38560 }, { "grad_norm": 0.10975661873817444, "learning_rate": 5.706756926355576e-05, "loss": 0.0076, "step": 38570 }, { "grad_norm": 0.11326982825994492, "learning_rate": 5.704710780865554e-05, "loss": 0.0078, "step": 38580 }, { "grad_norm": 0.1332738995552063, "learning_rate": 5.7026645149597325e-05, "loss": 0.0088, "step": 38590 }, { "grad_norm": 0.10771164298057556, "learning_rate": 5.700618128987764e-05, "loss": 0.009, "step": 38600 }, { "grad_norm": 0.15156763792037964, "learning_rate": 5.698571623299317e-05, "loss": 0.0086, "step": 38610 }, { "grad_norm": 0.13514606654644012, "learning_rate": 5.696524998244086e-05, "loss": 0.0121, "step": 38620 }, { "grad_norm": 0.10317163169384003, "learning_rate": 5.6944782541717836e-05, "loss": 0.0079, "step": 38630 }, { "grad_norm": 0.15377886593341827, "learning_rate": 5.69243139143214e-05, "loss": 0.0086, "step": 38640 }, { "grad_norm": 0.10601552575826645, "learning_rate": 5.6903844103749125e-05, "loss": 0.0077, "step": 38650 }, { "grad_norm": 0.14581571519374847, "learning_rate": 5.688337311349869e-05, "loss": 0.0081, "step": 38660 }, { "grad_norm": 0.1492283046245575, "learning_rate": 5.6862900947068074e-05, "loss": 0.0068, "step": 38670 }, { "grad_norm": 0.16401119530200958, "learning_rate": 5.6842427607955374e-05, "loss": 0.0087, "step": 38680 }, { "grad_norm": 0.1441788524389267, "learning_rate": 5.682195309965893e-05, "loss": 0.0078, "step": 38690 }, { "grad_norm": 0.10449833422899246, "learning_rate": 5.6801477425677294e-05, "loss": 0.0083, "step": 38700 }, { "grad_norm": 0.11630330234766006, "learning_rate": 5.678100058950917e-05, "loss": 0.008, "step": 38710 }, { "grad_norm": 0.09923841804265976, "learning_rate": 5.676052259465352e-05, "loss": 0.0083, "step": 38720 }, { "grad_norm": 0.15655097365379333, "learning_rate": 5.674004344460945e-05, "loss": 0.0097, "step": 38730 }, { "grad_norm": 0.12764811515808105, "learning_rate": 5.6719563142876295e-05, "loss": 0.0086, "step": 38740 }, { "grad_norm": 0.15430478751659393, "learning_rate": 5.669908169295359e-05, "loss": 0.0108, "step": 38750 }, { "grad_norm": 0.1418597251176834, "learning_rate": 5.667859909834105e-05, "loss": 0.0072, "step": 38760 }, { "grad_norm": 0.1579897552728653, "learning_rate": 5.6658115362538565e-05, "loss": 0.0102, "step": 38770 }, { "grad_norm": 0.15964896976947784, "learning_rate": 5.663763048904628e-05, "loss": 0.008, "step": 38780 }, { "grad_norm": 0.12360034137964249, "learning_rate": 5.661714448136447e-05, "loss": 0.0078, "step": 38790 }, { "grad_norm": 0.10880400985479355, "learning_rate": 5.659665734299366e-05, "loss": 0.0079, "step": 38800 }, { "grad_norm": 0.15064382553100586, "learning_rate": 5.6576169077434516e-05, "loss": 0.0087, "step": 38810 }, { "grad_norm": 0.09867972880601883, "learning_rate": 5.6555679688187944e-05, "loss": 0.0086, "step": 38820 }, { "grad_norm": 0.14019128680229187, "learning_rate": 5.6535189178755e-05, "loss": 0.0096, "step": 38830 }, { "grad_norm": 0.10362327843904495, "learning_rate": 5.651469755263698e-05, "loss": 0.0083, "step": 38840 }, { "grad_norm": 0.11175315082073212, "learning_rate": 5.6494204813335316e-05, "loss": 0.0088, "step": 38850 }, { "grad_norm": 0.11431559175252914, "learning_rate": 5.647371096435168e-05, "loss": 0.0116, "step": 38860 }, { "grad_norm": 0.18980002403259277, "learning_rate": 5.645321600918788e-05, "loss": 0.0082, "step": 38870 }, { "grad_norm": 0.17012323439121246, "learning_rate": 5.643271995134597e-05, "loss": 0.0091, "step": 38880 }, { "grad_norm": 0.19574229419231415, "learning_rate": 5.641222279432814e-05, "loss": 0.0087, "step": 38890 }, { "grad_norm": 0.13731220364570618, "learning_rate": 5.6391724541636834e-05, "loss": 0.0095, "step": 38900 }, { "grad_norm": 0.1846318542957306, "learning_rate": 5.6371225196774605e-05, "loss": 0.0087, "step": 38910 }, { "grad_norm": 0.1713385283946991, "learning_rate": 5.635072476324423e-05, "loss": 0.0073, "step": 38920 }, { "grad_norm": 0.11622654646635056, "learning_rate": 5.63302232445487e-05, "loss": 0.0063, "step": 38930 }, { "grad_norm": 0.11573334783315659, "learning_rate": 5.6309720644191144e-05, "loss": 0.0073, "step": 38940 }, { "grad_norm": 0.13294155895709991, "learning_rate": 5.628921696567491e-05, "loss": 0.0104, "step": 38950 }, { "grad_norm": 0.11544280499219894, "learning_rate": 5.62687122125035e-05, "loss": 0.0076, "step": 38960 }, { "grad_norm": 0.14521905779838562, "learning_rate": 5.624820638818062e-05, "loss": 0.0076, "step": 38970 }, { "grad_norm": 0.15377934277057648, "learning_rate": 5.6227699496210164e-05, "loss": 0.0081, "step": 38980 }, { "grad_norm": 0.09791140258312225, "learning_rate": 5.6207191540096195e-05, "loss": 0.0091, "step": 38990 }, { "grad_norm": 0.145713672041893, "learning_rate": 5.618668252334296e-05, "loss": 0.0074, "step": 39000 }, { "grad_norm": 0.15040557086467743, "learning_rate": 5.616617244945488e-05, "loss": 0.0103, "step": 39010 }, { "grad_norm": 0.12508955597877502, "learning_rate": 5.614566132193656e-05, "loss": 0.0075, "step": 39020 }, { "grad_norm": 0.09079552441835403, "learning_rate": 5.612514914429282e-05, "loss": 0.0086, "step": 39030 }, { "grad_norm": 0.164314404129982, "learning_rate": 5.610463592002863e-05, "loss": 0.0098, "step": 39040 }, { "grad_norm": 0.12666764855384827, "learning_rate": 5.608412165264909e-05, "loss": 0.0062, "step": 39050 }, { "grad_norm": 0.10689163953065872, "learning_rate": 5.606360634565959e-05, "loss": 0.01, "step": 39060 }, { "grad_norm": 0.12926341593265533, "learning_rate": 5.604309000256559e-05, "loss": 0.0092, "step": 39070 }, { "grad_norm": 0.1517905741930008, "learning_rate": 5.6022572626872785e-05, "loss": 0.0059, "step": 39080 }, { "grad_norm": 0.09797976166009903, "learning_rate": 5.600205422208704e-05, "loss": 0.009, "step": 39090 }, { "grad_norm": 0.12282688170671463, "learning_rate": 5.59815347917144e-05, "loss": 0.0086, "step": 39100 }, { "grad_norm": 0.16133897006511688, "learning_rate": 5.596101433926103e-05, "loss": 0.0061, "step": 39110 }, { "grad_norm": 0.12606652081012726, "learning_rate": 5.5940492868233364e-05, "loss": 0.0106, "step": 39120 }, { "grad_norm": 0.1131080761551857, "learning_rate": 5.591997038213793e-05, "loss": 0.0105, "step": 39130 }, { "grad_norm": 0.14385151863098145, "learning_rate": 5.5899446884481475e-05, "loss": 0.0096, "step": 39140 }, { "grad_norm": 0.1242058277130127, "learning_rate": 5.5878922378770906e-05, "loss": 0.0067, "step": 39150 }, { "grad_norm": 0.1328054666519165, "learning_rate": 5.5858396868513285e-05, "loss": 0.0085, "step": 39160 }, { "grad_norm": 0.10831483453512192, "learning_rate": 5.583787035721586e-05, "loss": 0.0069, "step": 39170 }, { "grad_norm": 0.11651147156953812, "learning_rate": 5.581734284838606e-05, "loss": 0.0069, "step": 39180 }, { "grad_norm": 0.11339947581291199, "learning_rate": 5.579681434553147e-05, "loss": 0.0093, "step": 39190 }, { "grad_norm": 0.1404985785484314, "learning_rate": 5.5776284852159854e-05, "loss": 0.0075, "step": 39200 }, { "grad_norm": 0.13193054497241974, "learning_rate": 5.575575437177913e-05, "loss": 0.0089, "step": 39210 }, { "grad_norm": 0.11529065668582916, "learning_rate": 5.573522290789742e-05, "loss": 0.008, "step": 39220 }, { "grad_norm": 0.13257116079330444, "learning_rate": 5.571469046402298e-05, "loss": 0.0078, "step": 39230 }, { "grad_norm": 0.24272620677947998, "learning_rate": 5.5694157043664205e-05, "loss": 0.0122, "step": 39240 }, { "grad_norm": 0.10446315258741379, "learning_rate": 5.567362265032975e-05, "loss": 0.0066, "step": 39250 }, { "grad_norm": 0.10167956352233887, "learning_rate": 5.565308728752836e-05, "loss": 0.0063, "step": 39260 }, { "grad_norm": 0.1389751136302948, "learning_rate": 5.5632550958768956e-05, "loss": 0.0069, "step": 39270 }, { "grad_norm": 0.12259645015001297, "learning_rate": 5.5612013667560636e-05, "loss": 0.0099, "step": 39280 }, { "grad_norm": 0.11530276387929916, "learning_rate": 5.5591475417412676e-05, "loss": 0.0094, "step": 39290 }, { "grad_norm": 0.109172523021698, "learning_rate": 5.557093621183451e-05, "loss": 0.0068, "step": 39300 }, { "grad_norm": 0.13965748250484467, "learning_rate": 5.55503960543357e-05, "loss": 0.0093, "step": 39310 }, { "grad_norm": 0.16931284964084625, "learning_rate": 5.552985494842601e-05, "loss": 0.0086, "step": 39320 }, { "grad_norm": 0.09393413364887238, "learning_rate": 5.550931289761534e-05, "loss": 0.0091, "step": 39330 }, { "grad_norm": 0.09801337867975235, "learning_rate": 5.548876990541378e-05, "loss": 0.0068, "step": 39340 }, { "grad_norm": 0.13822461664676666, "learning_rate": 5.5468225975331564e-05, "loss": 0.01, "step": 39350 }, { "grad_norm": 0.10959275811910629, "learning_rate": 5.544768111087909e-05, "loss": 0.0092, "step": 39360 }, { "grad_norm": 0.1245531216263771, "learning_rate": 5.54271353155669e-05, "loss": 0.0071, "step": 39370 }, { "grad_norm": 0.10822148621082306, "learning_rate": 5.5406588592905715e-05, "loss": 0.0072, "step": 39380 }, { "grad_norm": 0.16509169340133667, "learning_rate": 5.5386040946406416e-05, "loss": 0.0069, "step": 39390 }, { "grad_norm": 0.13380639255046844, "learning_rate": 5.536549237958004e-05, "loss": 0.006, "step": 39400 }, { "grad_norm": 0.14752648770809174, "learning_rate": 5.5344942895937744e-05, "loss": 0.008, "step": 39410 }, { "grad_norm": 0.08073003590106964, "learning_rate": 5.5324392498990904e-05, "loss": 0.008, "step": 39420 }, { "grad_norm": 0.15315969288349152, "learning_rate": 5.5303841192251e-05, "loss": 0.0093, "step": 39430 }, { "grad_norm": 0.14828188717365265, "learning_rate": 5.52832889792297e-05, "loss": 0.0105, "step": 39440 }, { "grad_norm": 0.10254231840372086, "learning_rate": 5.526273586343881e-05, "loss": 0.0099, "step": 39450 }, { "grad_norm": 0.16446474194526672, "learning_rate": 5.5242181848390306e-05, "loss": 0.008, "step": 39460 }, { "grad_norm": 0.14096814393997192, "learning_rate": 5.5221626937596285e-05, "loss": 0.0084, "step": 39470 }, { "grad_norm": 0.09840019047260284, "learning_rate": 5.520107113456903e-05, "loss": 0.0072, "step": 39480 }, { "grad_norm": 0.1456964910030365, "learning_rate": 5.5180514442820974e-05, "loss": 0.0083, "step": 39490 }, { "grad_norm": 0.1260497272014618, "learning_rate": 5.515995686586469e-05, "loss": 0.0071, "step": 39500 }, { "grad_norm": 0.17191778123378754, "learning_rate": 5.5139398407212916e-05, "loss": 0.0115, "step": 39510 }, { "grad_norm": 0.13337606191635132, "learning_rate": 5.511883907037849e-05, "loss": 0.0101, "step": 39520 }, { "grad_norm": 0.11056991666555405, "learning_rate": 5.509827885887449e-05, "loss": 0.0064, "step": 39530 }, { "grad_norm": 0.09735719114542007, "learning_rate": 5.507771777621406e-05, "loss": 0.0072, "step": 39540 }, { "grad_norm": 0.11122003942728043, "learning_rate": 5.505715582591052e-05, "loss": 0.0071, "step": 39550 }, { "grad_norm": 0.09627022594213486, "learning_rate": 5.50365930114774e-05, "loss": 0.007, "step": 39560 }, { "grad_norm": 0.09947533905506134, "learning_rate": 5.5016029336428255e-05, "loss": 0.0067, "step": 39570 }, { "grad_norm": 0.12490514665842056, "learning_rate": 5.49954648042769e-05, "loss": 0.0096, "step": 39580 }, { "grad_norm": 0.1669202297925949, "learning_rate": 5.4974899418537226e-05, "loss": 0.009, "step": 39590 }, { "grad_norm": 0.1748838573694229, "learning_rate": 5.4954333182723316e-05, "loss": 0.0069, "step": 39600 }, { "grad_norm": 0.36534973978996277, "learning_rate": 5.493376610034937e-05, "loss": 0.009, "step": 39610 }, { "grad_norm": 0.2986612915992737, "learning_rate": 5.4913198174929735e-05, "loss": 0.0072, "step": 39620 }, { "grad_norm": 0.1308041214942932, "learning_rate": 5.489262940997889e-05, "loss": 0.0077, "step": 39630 }, { "grad_norm": 0.12928612530231476, "learning_rate": 5.487205980901151e-05, "loss": 0.0079, "step": 39640 }, { "grad_norm": 0.14859315752983093, "learning_rate": 5.485148937554234e-05, "loss": 0.0061, "step": 39650 }, { "grad_norm": 0.15063698589801788, "learning_rate": 5.483091811308635e-05, "loss": 0.0083, "step": 39660 }, { "grad_norm": 0.1492464542388916, "learning_rate": 5.4810346025158564e-05, "loss": 0.0084, "step": 39670 }, { "grad_norm": 0.13825371861457825, "learning_rate": 5.478977311527421e-05, "loss": 0.0093, "step": 39680 }, { "grad_norm": 0.12305626273155212, "learning_rate": 5.476919938694863e-05, "loss": 0.009, "step": 39690 }, { "grad_norm": 0.13901282846927643, "learning_rate": 5.474862484369733e-05, "loss": 0.0085, "step": 39700 }, { "grad_norm": 0.1538352370262146, "learning_rate": 5.472804948903589e-05, "loss": 0.0096, "step": 39710 }, { "grad_norm": 0.11860021948814392, "learning_rate": 5.470747332648013e-05, "loss": 0.0075, "step": 39720 }, { "grad_norm": 0.16955609619617462, "learning_rate": 5.468689635954591e-05, "loss": 0.009, "step": 39730 }, { "grad_norm": 0.12817369401454926, "learning_rate": 5.46663185917493e-05, "loss": 0.0105, "step": 39740 }, { "grad_norm": 0.14705945551395416, "learning_rate": 5.464574002660645e-05, "loss": 0.0061, "step": 39750 }, { "grad_norm": 0.15771161019802094, "learning_rate": 5.4625160667633724e-05, "loss": 0.0079, "step": 39760 }, { "grad_norm": 0.14001500606536865, "learning_rate": 5.4604580518347505e-05, "loss": 0.0081, "step": 39770 }, { "grad_norm": 0.0991816595196724, "learning_rate": 5.458399958226442e-05, "loss": 0.0082, "step": 39780 }, { "grad_norm": 0.1257786750793457, "learning_rate": 5.456341786290119e-05, "loss": 0.0107, "step": 39790 }, { "grad_norm": 0.10764738917350769, "learning_rate": 5.454283536377465e-05, "loss": 0.008, "step": 39800 }, { "grad_norm": 0.1545497328042984, "learning_rate": 5.452225208840179e-05, "loss": 0.008, "step": 39810 }, { "grad_norm": 0.12399772554636002, "learning_rate": 5.450166804029975e-05, "loss": 0.0101, "step": 39820 }, { "grad_norm": 0.13350822031497955, "learning_rate": 5.448108322298574e-05, "loss": 0.0074, "step": 39830 }, { "grad_norm": 0.16470323503017426, "learning_rate": 5.446049763997717e-05, "loss": 0.0083, "step": 39840 }, { "grad_norm": 0.13024139404296875, "learning_rate": 5.4439911294791546e-05, "loss": 0.0095, "step": 39850 }, { "grad_norm": 0.14172613620758057, "learning_rate": 5.441932419094652e-05, "loss": 0.0078, "step": 39860 }, { "grad_norm": 0.12813010811805725, "learning_rate": 5.439873633195985e-05, "loss": 0.0068, "step": 39870 }, { "grad_norm": 0.11418073624372482, "learning_rate": 5.437814772134947e-05, "loss": 0.0076, "step": 39880 }, { "grad_norm": 0.12535367906093597, "learning_rate": 5.4357558362633366e-05, "loss": 0.0104, "step": 39890 }, { "grad_norm": 0.10447458922863007, "learning_rate": 5.4336968259329726e-05, "loss": 0.0098, "step": 39900 }, { "grad_norm": 0.15735048055648804, "learning_rate": 5.431637741495681e-05, "loss": 0.0081, "step": 39910 }, { "grad_norm": 0.07298723608255386, "learning_rate": 5.429578583303307e-05, "loss": 0.0089, "step": 39920 }, { "grad_norm": 0.12718166410923004, "learning_rate": 5.427519351707701e-05, "loss": 0.0102, "step": 39930 }, { "grad_norm": 0.1168239414691925, "learning_rate": 5.42546004706073e-05, "loss": 0.0073, "step": 39940 }, { "grad_norm": 0.10301476716995239, "learning_rate": 5.4234006697142735e-05, "loss": 0.0066, "step": 39950 }, { "grad_norm": 0.16008873283863068, "learning_rate": 5.421341220020224e-05, "loss": 0.0094, "step": 39960 }, { "grad_norm": 0.11227688193321228, "learning_rate": 5.419281698330482e-05, "loss": 0.0065, "step": 39970 }, { "grad_norm": 0.10422108322381973, "learning_rate": 5.4172221049969665e-05, "loss": 0.0078, "step": 39980 }, { "grad_norm": 0.13038909435272217, "learning_rate": 5.415162440371604e-05, "loss": 0.0067, "step": 39990 }, { "grad_norm": 0.14961174130439758, "learning_rate": 5.413102704806334e-05, "loss": 0.0068, "step": 40000 } ], "logging_steps": 10, "max_steps": 80000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 20000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 48, "trial_name": null, "trial_params": null }