{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 80000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.4151914417743683, "learning_rate": 2.25e-07, "loss": 0.0389, "step": 10 }, { "grad_norm": 0.45794206857681274, "learning_rate": 4.75e-07, "loss": 0.023, "step": 20 }, { "grad_norm": 0.3701542913913727, "learning_rate": 7.25e-07, "loss": 0.0307, "step": 30 }, { "grad_norm": 0.3362779915332794, "learning_rate": 9.75e-07, "loss": 0.0302, "step": 40 }, { "grad_norm": 0.3698456585407257, "learning_rate": 1.2250000000000001e-06, "loss": 0.0218, "step": 50 }, { "grad_norm": 0.3265153169631958, "learning_rate": 1.475e-06, "loss": 0.0241, "step": 60 }, { "grad_norm": 0.7132813930511475, "learning_rate": 1.7250000000000002e-06, "loss": 0.0248, "step": 70 }, { "grad_norm": 0.22214633226394653, "learning_rate": 1.975e-06, "loss": 0.0153, "step": 80 }, { "grad_norm": 0.7616195678710938, "learning_rate": 2.225e-06, "loss": 0.0187, "step": 90 }, { "grad_norm": 0.3667467534542084, "learning_rate": 2.4750000000000004e-06, "loss": 0.0227, "step": 100 }, { "grad_norm": 0.30280157923698425, "learning_rate": 2.725e-06, "loss": 0.0222, "step": 110 }, { "grad_norm": 0.42388901114463806, "learning_rate": 2.975e-06, "loss": 0.0211, "step": 120 }, { "grad_norm": 0.7851755023002625, "learning_rate": 3.225e-06, "loss": 0.016, "step": 130 }, { "grad_norm": 0.27732235193252563, "learning_rate": 3.4750000000000006e-06, "loss": 0.0148, "step": 140 }, { "grad_norm": 0.3041248917579651, "learning_rate": 3.725e-06, "loss": 0.0184, "step": 150 }, { "grad_norm": 0.4418157935142517, "learning_rate": 3.975e-06, "loss": 0.0143, "step": 160 }, { "grad_norm": 0.1583632379770279, "learning_rate": 4.225e-06, "loss": 0.0115, "step": 170 }, { "grad_norm": 0.19200001657009125, "learning_rate": 4.475e-06, "loss": 0.0156, "step": 180 }, { "grad_norm": 0.49035272002220154, "learning_rate": 4.7250000000000005e-06, "loss": 0.022, "step": 190 }, { "grad_norm": 0.3685218393802643, "learning_rate": 4.975000000000001e-06, "loss": 0.0152, "step": 200 }, { "grad_norm": 0.31504198908805847, "learning_rate": 5.225e-06, "loss": 0.0148, "step": 210 }, { "grad_norm": 3.8550913333892822, "learning_rate": 5.475e-06, "loss": 0.0154, "step": 220 }, { "grad_norm": 0.17421358823776245, "learning_rate": 5.725e-06, "loss": 0.0136, "step": 230 }, { "grad_norm": 0.2083987444639206, "learning_rate": 5.975e-06, "loss": 0.0192, "step": 240 }, { "grad_norm": 0.17531993985176086, "learning_rate": 6.2250000000000005e-06, "loss": 0.0111, "step": 250 }, { "grad_norm": 0.31158995628356934, "learning_rate": 6.475000000000001e-06, "loss": 0.0136, "step": 260 }, { "grad_norm": 0.3297537863254547, "learning_rate": 6.725000000000001e-06, "loss": 0.0138, "step": 270 }, { "grad_norm": 0.17748145759105682, "learning_rate": 6.975000000000001e-06, "loss": 0.0129, "step": 280 }, { "grad_norm": 0.3574780821800232, "learning_rate": 7.2249999999999994e-06, "loss": 0.0148, "step": 290 }, { "grad_norm": 0.22364838421344757, "learning_rate": 7.4750000000000004e-06, "loss": 0.0147, "step": 300 }, { "grad_norm": 0.25133275985717773, "learning_rate": 7.725e-06, "loss": 0.0131, "step": 310 }, { "grad_norm": 0.21385912597179413, "learning_rate": 7.975e-06, "loss": 0.0127, "step": 320 }, { "grad_norm": 0.3158475458621979, "learning_rate": 8.225e-06, "loss": 0.0125, "step": 330 }, { "grad_norm": 0.2439994513988495, "learning_rate": 8.475000000000001e-06, "loss": 0.0122, "step": 340 }, { "grad_norm": 0.26948079466819763, "learning_rate": 8.725e-06, "loss": 0.0142, "step": 350 }, { "grad_norm": 0.21210238337516785, "learning_rate": 8.975e-06, "loss": 0.009, "step": 360 }, { "grad_norm": 0.29093098640441895, "learning_rate": 9.225e-06, "loss": 0.012, "step": 370 }, { "grad_norm": 0.11808878928422928, "learning_rate": 9.475e-06, "loss": 0.015, "step": 380 }, { "grad_norm": 0.4430798888206482, "learning_rate": 9.725000000000001e-06, "loss": 0.0134, "step": 390 }, { "grad_norm": 0.22632800042629242, "learning_rate": 9.975e-06, "loss": 0.0136, "step": 400 }, { "grad_norm": 0.14313599467277527, "learning_rate": 1.0225e-05, "loss": 0.0134, "step": 410 }, { "grad_norm": 0.12392386794090271, "learning_rate": 1.0475e-05, "loss": 0.0131, "step": 420 }, { "grad_norm": 0.18089509010314941, "learning_rate": 1.0725e-05, "loss": 0.0138, "step": 430 }, { "grad_norm": 0.13577978312969208, "learning_rate": 1.0975e-05, "loss": 0.0125, "step": 440 }, { "grad_norm": 0.2347945272922516, "learning_rate": 1.1225e-05, "loss": 0.0115, "step": 450 }, { "grad_norm": 0.1955430507659912, "learning_rate": 1.1475000000000001e-05, "loss": 0.0125, "step": 460 }, { "grad_norm": 0.35329654812812805, "learning_rate": 1.1725e-05, "loss": 0.0173, "step": 470 }, { "grad_norm": 0.25761696696281433, "learning_rate": 1.1975e-05, "loss": 0.0166, "step": 480 }, { "grad_norm": 0.7765153646469116, "learning_rate": 1.2225e-05, "loss": 0.016, "step": 490 }, { "grad_norm": 0.27257344126701355, "learning_rate": 1.2475e-05, "loss": 0.015, "step": 500 }, { "grad_norm": 0.2846745550632477, "learning_rate": 1.2725000000000001e-05, "loss": 0.015, "step": 510 }, { "grad_norm": 0.487277090549469, "learning_rate": 1.2975e-05, "loss": 0.0164, "step": 520 }, { "grad_norm": 0.28140988945961, "learning_rate": 1.3225000000000001e-05, "loss": 0.0161, "step": 530 }, { "grad_norm": 0.2776981592178345, "learning_rate": 1.3475000000000002e-05, "loss": 0.009, "step": 540 }, { "grad_norm": 0.2664620280265808, "learning_rate": 1.3725000000000002e-05, "loss": 0.0154, "step": 550 }, { "grad_norm": 0.2877046465873718, "learning_rate": 1.3975000000000003e-05, "loss": 0.0141, "step": 560 }, { "grad_norm": 0.12096092849969864, "learning_rate": 1.4225e-05, "loss": 0.0125, "step": 570 }, { "grad_norm": 0.21074801683425903, "learning_rate": 1.4475e-05, "loss": 0.0146, "step": 580 }, { "grad_norm": 0.1559780091047287, "learning_rate": 1.4725e-05, "loss": 0.012, "step": 590 }, { "grad_norm": 0.1732572466135025, "learning_rate": 1.4975e-05, "loss": 0.0188, "step": 600 }, { "grad_norm": 0.1851796954870224, "learning_rate": 1.5225e-05, "loss": 0.0135, "step": 610 }, { "grad_norm": 0.36023154854774475, "learning_rate": 1.5475e-05, "loss": 0.0156, "step": 620 }, { "grad_norm": 0.3450355529785156, "learning_rate": 1.5725e-05, "loss": 0.0129, "step": 630 }, { "grad_norm": 0.2465735673904419, "learning_rate": 1.5975000000000002e-05, "loss": 0.0122, "step": 640 }, { "grad_norm": 0.5348679423332214, "learning_rate": 1.6225e-05, "loss": 0.0143, "step": 650 }, { "grad_norm": 0.3632704019546509, "learning_rate": 1.6475e-05, "loss": 0.0218, "step": 660 }, { "grad_norm": 0.18474546074867249, "learning_rate": 1.6725000000000003e-05, "loss": 0.0127, "step": 670 }, { "grad_norm": 0.2067318558692932, "learning_rate": 1.6975000000000003e-05, "loss": 0.0174, "step": 680 }, { "grad_norm": 0.23953920602798462, "learning_rate": 1.7225e-05, "loss": 0.0118, "step": 690 }, { "grad_norm": 0.2460770308971405, "learning_rate": 1.7475e-05, "loss": 0.0131, "step": 700 }, { "grad_norm": 0.23125529289245605, "learning_rate": 1.7725e-05, "loss": 0.0117, "step": 710 }, { "grad_norm": 0.3071388602256775, "learning_rate": 1.7975e-05, "loss": 0.0134, "step": 720 }, { "grad_norm": 0.20726922154426575, "learning_rate": 1.8225e-05, "loss": 0.0124, "step": 730 }, { "grad_norm": 0.11487627029418945, "learning_rate": 1.8475000000000002e-05, "loss": 0.0107, "step": 740 }, { "grad_norm": 0.19039256870746613, "learning_rate": 1.8725e-05, "loss": 0.0106, "step": 750 }, { "grad_norm": 0.18727752566337585, "learning_rate": 1.8975e-05, "loss": 0.0113, "step": 760 }, { "grad_norm": 0.1487288624048233, "learning_rate": 1.9225e-05, "loss": 0.0157, "step": 770 }, { "grad_norm": 0.14843779802322388, "learning_rate": 1.9475000000000002e-05, "loss": 0.0161, "step": 780 }, { "grad_norm": 0.31058263778686523, "learning_rate": 1.9725000000000002e-05, "loss": 0.0149, "step": 790 }, { "grad_norm": 0.29171404242515564, "learning_rate": 1.9975e-05, "loss": 0.0106, "step": 800 }, { "grad_norm": 0.2115887850522995, "learning_rate": 2.0225000000000004e-05, "loss": 0.0155, "step": 810 }, { "grad_norm": 0.3826092481613159, "learning_rate": 2.0475e-05, "loss": 0.0212, "step": 820 }, { "grad_norm": 0.2510721683502197, "learning_rate": 2.0725e-05, "loss": 0.0277, "step": 830 }, { "grad_norm": 0.1936909556388855, "learning_rate": 2.0975e-05, "loss": 0.0137, "step": 840 }, { "grad_norm": 0.4749101996421814, "learning_rate": 2.1225e-05, "loss": 0.0202, "step": 850 }, { "grad_norm": 0.2741926610469818, "learning_rate": 2.1475e-05, "loss": 0.0182, "step": 860 }, { "grad_norm": 0.264329731464386, "learning_rate": 2.1725e-05, "loss": 0.0183, "step": 870 }, { "grad_norm": 0.24461659789085388, "learning_rate": 2.1975000000000002e-05, "loss": 0.0129, "step": 880 }, { "grad_norm": 0.21706221997737885, "learning_rate": 2.2225e-05, "loss": 0.0146, "step": 890 }, { "grad_norm": 0.30357810854911804, "learning_rate": 2.2475e-05, "loss": 0.0145, "step": 900 }, { "grad_norm": 0.16196976602077484, "learning_rate": 2.2725000000000003e-05, "loss": 0.0186, "step": 910 }, { "grad_norm": 0.24720242619514465, "learning_rate": 2.2975000000000003e-05, "loss": 0.011, "step": 920 }, { "grad_norm": 0.2639351785182953, "learning_rate": 2.3225000000000002e-05, "loss": 0.0161, "step": 930 }, { "grad_norm": 0.260273814201355, "learning_rate": 2.3475e-05, "loss": 0.0191, "step": 940 }, { "grad_norm": 0.17270052433013916, "learning_rate": 2.3725e-05, "loss": 0.0131, "step": 950 }, { "grad_norm": 0.2040589451789856, "learning_rate": 2.3975e-05, "loss": 0.012, "step": 960 }, { "grad_norm": 0.2640938460826874, "learning_rate": 2.4225e-05, "loss": 0.0175, "step": 970 }, { "grad_norm": 0.3148156404495239, "learning_rate": 2.4475000000000002e-05, "loss": 0.0107, "step": 980 }, { "grad_norm": 0.20724202692508698, "learning_rate": 2.4725e-05, "loss": 0.0107, "step": 990 }, { "grad_norm": 0.2002474069595337, "learning_rate": 2.4975e-05, "loss": 0.0139, "step": 1000 }, { "grad_norm": 0.2987534701824188, "learning_rate": 2.5225e-05, "loss": 0.0174, "step": 1010 }, { "grad_norm": 0.2734834551811218, "learning_rate": 2.5475e-05, "loss": 0.0145, "step": 1020 }, { "grad_norm": 0.16589055955410004, "learning_rate": 2.5725e-05, "loss": 0.0111, "step": 1030 }, { "grad_norm": 0.3975799083709717, "learning_rate": 2.5974999999999998e-05, "loss": 0.0164, "step": 1040 }, { "grad_norm": 0.5580929517745972, "learning_rate": 2.6225e-05, "loss": 0.0214, "step": 1050 }, { "grad_norm": 0.19564585387706757, "learning_rate": 2.6475e-05, "loss": 0.0155, "step": 1060 }, { "grad_norm": 0.3258846700191498, "learning_rate": 2.6725e-05, "loss": 0.0218, "step": 1070 }, { "grad_norm": 0.3583710193634033, "learning_rate": 2.6975000000000002e-05, "loss": 0.0199, "step": 1080 }, { "grad_norm": 0.34365367889404297, "learning_rate": 2.7225e-05, "loss": 0.0173, "step": 1090 }, { "grad_norm": 0.29315048456192017, "learning_rate": 2.7475e-05, "loss": 0.0132, "step": 1100 }, { "grad_norm": 0.23530106246471405, "learning_rate": 2.7725e-05, "loss": 0.0198, "step": 1110 }, { "grad_norm": 0.13386934995651245, "learning_rate": 2.7975000000000002e-05, "loss": 0.0105, "step": 1120 }, { "grad_norm": 0.2516616880893707, "learning_rate": 2.8225e-05, "loss": 0.0146, "step": 1130 }, { "grad_norm": 0.39129048585891724, "learning_rate": 2.8475e-05, "loss": 0.0295, "step": 1140 }, { "grad_norm": 0.2741755545139313, "learning_rate": 2.8725e-05, "loss": 0.0205, "step": 1150 }, { "grad_norm": 0.24253880977630615, "learning_rate": 2.8975000000000003e-05, "loss": 0.0148, "step": 1160 }, { "grad_norm": 0.21871517598628998, "learning_rate": 2.9225000000000002e-05, "loss": 0.0156, "step": 1170 }, { "grad_norm": 0.24068371951580048, "learning_rate": 2.9475e-05, "loss": 0.0138, "step": 1180 }, { "grad_norm": 0.2534858286380768, "learning_rate": 2.9725000000000004e-05, "loss": 0.0174, "step": 1190 }, { "grad_norm": 0.4467220902442932, "learning_rate": 2.9975000000000004e-05, "loss": 0.0205, "step": 1200 }, { "grad_norm": 0.4496861696243286, "learning_rate": 3.0225000000000003e-05, "loss": 0.0165, "step": 1210 }, { "grad_norm": 0.3796267509460449, "learning_rate": 3.0475000000000002e-05, "loss": 0.0193, "step": 1220 }, { "grad_norm": 0.26194918155670166, "learning_rate": 3.0725e-05, "loss": 0.0138, "step": 1230 }, { "grad_norm": 0.1994919627904892, "learning_rate": 3.0975e-05, "loss": 0.0167, "step": 1240 }, { "grad_norm": 0.33258944749832153, "learning_rate": 3.122500000000001e-05, "loss": 0.0146, "step": 1250 }, { "grad_norm": 0.2840392589569092, "learning_rate": 3.1475e-05, "loss": 0.0146, "step": 1260 }, { "grad_norm": 0.2743649184703827, "learning_rate": 3.1725e-05, "loss": 0.0152, "step": 1270 }, { "grad_norm": 0.23264087736606598, "learning_rate": 3.1975e-05, "loss": 0.0199, "step": 1280 }, { "grad_norm": 0.24968069791793823, "learning_rate": 3.2225e-05, "loss": 0.0204, "step": 1290 }, { "grad_norm": 0.19101153314113617, "learning_rate": 3.2474999999999997e-05, "loss": 0.0152, "step": 1300 }, { "grad_norm": 0.35133692622184753, "learning_rate": 3.2725e-05, "loss": 0.0131, "step": 1310 }, { "grad_norm": 0.17899993062019348, "learning_rate": 3.2975e-05, "loss": 0.0173, "step": 1320 }, { "grad_norm": 0.16797500848770142, "learning_rate": 3.3225e-05, "loss": 0.0135, "step": 1330 }, { "grad_norm": 0.20468012988567352, "learning_rate": 3.3475e-05, "loss": 0.0108, "step": 1340 }, { "grad_norm": 0.23725639283657074, "learning_rate": 3.3725e-05, "loss": 0.0225, "step": 1350 }, { "grad_norm": 0.3764455020427704, "learning_rate": 3.3975e-05, "loss": 0.0167, "step": 1360 }, { "grad_norm": 0.30603086948394775, "learning_rate": 3.4225e-05, "loss": 0.0121, "step": 1370 }, { "grad_norm": 0.293377548456192, "learning_rate": 3.4475000000000005e-05, "loss": 0.0155, "step": 1380 }, { "grad_norm": 0.2823452651500702, "learning_rate": 3.4725000000000004e-05, "loss": 0.0144, "step": 1390 }, { "grad_norm": 0.2759988009929657, "learning_rate": 3.4975e-05, "loss": 0.0131, "step": 1400 }, { "grad_norm": 0.223181813955307, "learning_rate": 3.5225e-05, "loss": 0.0172, "step": 1410 }, { "grad_norm": 0.26566067337989807, "learning_rate": 3.5475e-05, "loss": 0.0202, "step": 1420 }, { "grad_norm": 0.21965520083904266, "learning_rate": 3.5725e-05, "loss": 0.0136, "step": 1430 }, { "grad_norm": 0.2696687877178192, "learning_rate": 3.5975e-05, "loss": 0.0182, "step": 1440 }, { "grad_norm": 0.24795526266098022, "learning_rate": 3.6225000000000006e-05, "loss": 0.0164, "step": 1450 }, { "grad_norm": 0.1370963305234909, "learning_rate": 3.6475000000000006e-05, "loss": 0.017, "step": 1460 }, { "grad_norm": 0.19740216434001923, "learning_rate": 3.6725000000000005e-05, "loss": 0.0186, "step": 1470 }, { "grad_norm": 0.15543779730796814, "learning_rate": 3.6975000000000004e-05, "loss": 0.014, "step": 1480 }, { "grad_norm": 0.2023056298494339, "learning_rate": 3.7225000000000004e-05, "loss": 0.0132, "step": 1490 }, { "grad_norm": 0.1715230941772461, "learning_rate": 3.7475e-05, "loss": 0.0158, "step": 1500 }, { "grad_norm": 0.2440866082906723, "learning_rate": 3.7725e-05, "loss": 0.0125, "step": 1510 }, { "grad_norm": 0.34850990772247314, "learning_rate": 3.7975e-05, "loss": 0.0156, "step": 1520 }, { "grad_norm": 0.2406880408525467, "learning_rate": 3.8225e-05, "loss": 0.0167, "step": 1530 }, { "grad_norm": 0.2601943910121918, "learning_rate": 3.8475e-05, "loss": 0.0149, "step": 1540 }, { "grad_norm": 0.23841121792793274, "learning_rate": 3.8725e-05, "loss": 0.0129, "step": 1550 }, { "grad_norm": 0.22083096206188202, "learning_rate": 3.8975e-05, "loss": 0.0263, "step": 1560 }, { "grad_norm": 0.20647065341472626, "learning_rate": 3.9225e-05, "loss": 0.0131, "step": 1570 }, { "grad_norm": 0.2309848815202713, "learning_rate": 3.9475000000000004e-05, "loss": 0.0173, "step": 1580 }, { "grad_norm": 0.3098561763763428, "learning_rate": 3.9725e-05, "loss": 0.0207, "step": 1590 }, { "grad_norm": 0.27607250213623047, "learning_rate": 3.9975e-05, "loss": 0.0126, "step": 1600 }, { "grad_norm": 0.25508585572242737, "learning_rate": 4.0225e-05, "loss": 0.0123, "step": 1610 }, { "grad_norm": 0.669927716255188, "learning_rate": 4.0475e-05, "loss": 0.0154, "step": 1620 }, { "grad_norm": 0.27618804574012756, "learning_rate": 4.0725e-05, "loss": 0.0276, "step": 1630 }, { "grad_norm": 0.40290433168411255, "learning_rate": 4.0975e-05, "loss": 0.019, "step": 1640 }, { "grad_norm": 0.23392800986766815, "learning_rate": 4.1225e-05, "loss": 0.0181, "step": 1650 }, { "grad_norm": 0.31702709197998047, "learning_rate": 4.1475000000000005e-05, "loss": 0.019, "step": 1660 }, { "grad_norm": 0.15787199139595032, "learning_rate": 4.1725000000000005e-05, "loss": 0.0168, "step": 1670 }, { "grad_norm": 0.34117332100868225, "learning_rate": 4.1975000000000004e-05, "loss": 0.0146, "step": 1680 }, { "grad_norm": 0.17394587397575378, "learning_rate": 4.2225e-05, "loss": 0.0172, "step": 1690 }, { "grad_norm": 0.30725163221359253, "learning_rate": 4.2475e-05, "loss": 0.0187, "step": 1700 }, { "grad_norm": 0.35600772500038147, "learning_rate": 4.2725e-05, "loss": 0.0199, "step": 1710 }, { "grad_norm": 0.3167257606983185, "learning_rate": 4.2975e-05, "loss": 0.0209, "step": 1720 }, { "grad_norm": 0.2302026003599167, "learning_rate": 4.322500000000001e-05, "loss": 0.0176, "step": 1730 }, { "grad_norm": 0.22493158280849457, "learning_rate": 4.3475000000000006e-05, "loss": 0.0151, "step": 1740 }, { "grad_norm": 0.43045344948768616, "learning_rate": 4.3725000000000006e-05, "loss": 0.0189, "step": 1750 }, { "grad_norm": 0.3488284945487976, "learning_rate": 4.3975e-05, "loss": 0.0172, "step": 1760 }, { "grad_norm": 0.22455598413944244, "learning_rate": 4.4225e-05, "loss": 0.0151, "step": 1770 }, { "grad_norm": 0.21068716049194336, "learning_rate": 4.4475e-05, "loss": 0.0202, "step": 1780 }, { "grad_norm": 0.28559187054634094, "learning_rate": 4.4725e-05, "loss": 0.0155, "step": 1790 }, { "grad_norm": 0.2639520466327667, "learning_rate": 4.4975e-05, "loss": 0.0174, "step": 1800 }, { "grad_norm": 0.3791493773460388, "learning_rate": 4.5225e-05, "loss": 0.019, "step": 1810 }, { "grad_norm": 0.40655726194381714, "learning_rate": 4.5475e-05, "loss": 0.0197, "step": 1820 }, { "grad_norm": 0.6505765914916992, "learning_rate": 4.5725e-05, "loss": 0.0216, "step": 1830 }, { "grad_norm": 0.3765034079551697, "learning_rate": 4.5975e-05, "loss": 0.0184, "step": 1840 }, { "grad_norm": 0.18848441541194916, "learning_rate": 4.6225e-05, "loss": 0.023, "step": 1850 }, { "grad_norm": 0.3533344268798828, "learning_rate": 4.6475000000000005e-05, "loss": 0.0245, "step": 1860 }, { "grad_norm": 0.2374913990497589, "learning_rate": 4.6725000000000004e-05, "loss": 0.0197, "step": 1870 }, { "grad_norm": 0.2381051480770111, "learning_rate": 4.6975000000000003e-05, "loss": 0.0162, "step": 1880 }, { "grad_norm": 0.19937638938426971, "learning_rate": 4.7225e-05, "loss": 0.0192, "step": 1890 }, { "grad_norm": 0.2857486605644226, "learning_rate": 4.7475e-05, "loss": 0.0159, "step": 1900 }, { "grad_norm": 0.3074304759502411, "learning_rate": 4.7725e-05, "loss": 0.0178, "step": 1910 }, { "grad_norm": 0.22565898299217224, "learning_rate": 4.7975e-05, "loss": 0.0175, "step": 1920 }, { "grad_norm": 0.2487826943397522, "learning_rate": 4.822500000000001e-05, "loss": 0.014, "step": 1930 }, { "grad_norm": 0.2955079972743988, "learning_rate": 4.8475000000000006e-05, "loss": 0.02, "step": 1940 }, { "grad_norm": 0.15529514849185944, "learning_rate": 4.8725000000000005e-05, "loss": 0.0141, "step": 1950 }, { "grad_norm": 0.3888922929763794, "learning_rate": 4.8975000000000005e-05, "loss": 0.0219, "step": 1960 }, { "grad_norm": 0.2670777440071106, "learning_rate": 4.9225000000000004e-05, "loss": 0.0162, "step": 1970 }, { "grad_norm": 0.38619422912597656, "learning_rate": 4.9475e-05, "loss": 0.0231, "step": 1980 }, { "grad_norm": 0.22968466579914093, "learning_rate": 4.9725e-05, "loss": 0.0175, "step": 1990 }, { "grad_norm": 0.30427032709121704, "learning_rate": 4.9975e-05, "loss": 0.0216, "step": 2000 }, { "grad_norm": 0.3700944483280182, "learning_rate": 5.0225e-05, "loss": 0.0321, "step": 2010 }, { "grad_norm": 0.3043968677520752, "learning_rate": 5.047500000000001e-05, "loss": 0.0209, "step": 2020 }, { "grad_norm": 0.25605836510658264, "learning_rate": 5.0725e-05, "loss": 0.0159, "step": 2030 }, { "grad_norm": 0.25170618295669556, "learning_rate": 5.0975000000000006e-05, "loss": 0.0231, "step": 2040 }, { "grad_norm": 0.3698650896549225, "learning_rate": 5.1225e-05, "loss": 0.0166, "step": 2050 }, { "grad_norm": 0.3471932113170624, "learning_rate": 5.1475000000000004e-05, "loss": 0.0166, "step": 2060 }, { "grad_norm": 0.19152724742889404, "learning_rate": 5.1725000000000004e-05, "loss": 0.0135, "step": 2070 }, { "grad_norm": 0.29330867528915405, "learning_rate": 5.197500000000001e-05, "loss": 0.0196, "step": 2080 }, { "grad_norm": 0.17844903469085693, "learning_rate": 5.2225e-05, "loss": 0.0182, "step": 2090 }, { "grad_norm": 0.29118838906288147, "learning_rate": 5.247500000000001e-05, "loss": 0.0193, "step": 2100 }, { "grad_norm": 0.31702014803886414, "learning_rate": 5.2725e-05, "loss": 0.0293, "step": 2110 }, { "grad_norm": 0.3972291052341461, "learning_rate": 5.297500000000001e-05, "loss": 0.0187, "step": 2120 }, { "grad_norm": 0.35534411668777466, "learning_rate": 5.3225e-05, "loss": 0.0224, "step": 2130 }, { "grad_norm": 0.2001568228006363, "learning_rate": 5.3475e-05, "loss": 0.0163, "step": 2140 }, { "grad_norm": 0.39658883213996887, "learning_rate": 5.3725000000000005e-05, "loss": 0.0219, "step": 2150 }, { "grad_norm": 0.224977046251297, "learning_rate": 5.3975e-05, "loss": 0.0171, "step": 2160 }, { "grad_norm": 0.5860757827758789, "learning_rate": 5.4225000000000003e-05, "loss": 0.0173, "step": 2170 }, { "grad_norm": 0.2469227910041809, "learning_rate": 5.4474999999999996e-05, "loss": 0.0218, "step": 2180 }, { "grad_norm": 0.37040242552757263, "learning_rate": 5.4725e-05, "loss": 0.0211, "step": 2190 }, { "grad_norm": 0.3192334771156311, "learning_rate": 5.4975e-05, "loss": 0.0189, "step": 2200 }, { "grad_norm": 0.19988077878952026, "learning_rate": 5.522500000000001e-05, "loss": 0.0199, "step": 2210 }, { "grad_norm": 0.44981491565704346, "learning_rate": 5.5475e-05, "loss": 0.0208, "step": 2220 }, { "grad_norm": 0.36382657289505005, "learning_rate": 5.5725000000000006e-05, "loss": 0.0163, "step": 2230 }, { "grad_norm": 0.20044194161891937, "learning_rate": 5.5975e-05, "loss": 0.0178, "step": 2240 }, { "grad_norm": 0.30450183153152466, "learning_rate": 5.6225000000000005e-05, "loss": 0.0201, "step": 2250 }, { "grad_norm": 0.27106884121894836, "learning_rate": 5.6475e-05, "loss": 0.0183, "step": 2260 }, { "grad_norm": 0.26910969614982605, "learning_rate": 5.6725e-05, "loss": 0.0173, "step": 2270 }, { "grad_norm": 0.2282923460006714, "learning_rate": 5.6975e-05, "loss": 0.0212, "step": 2280 }, { "grad_norm": 0.3314507305622101, "learning_rate": 5.722500000000001e-05, "loss": 0.0247, "step": 2290 }, { "grad_norm": 0.35453081130981445, "learning_rate": 5.7475e-05, "loss": 0.0249, "step": 2300 }, { "grad_norm": 0.25059011578559875, "learning_rate": 5.772500000000001e-05, "loss": 0.0204, "step": 2310 }, { "grad_norm": 0.29233190417289734, "learning_rate": 5.7975e-05, "loss": 0.0149, "step": 2320 }, { "grad_norm": 0.549227774143219, "learning_rate": 5.8225000000000006e-05, "loss": 0.028, "step": 2330 }, { "grad_norm": 0.31378069519996643, "learning_rate": 5.8475000000000005e-05, "loss": 0.0195, "step": 2340 }, { "grad_norm": 0.2032504379749298, "learning_rate": 5.8725000000000004e-05, "loss": 0.0189, "step": 2350 }, { "grad_norm": 0.22169329226016998, "learning_rate": 5.8975000000000004e-05, "loss": 0.0194, "step": 2360 }, { "grad_norm": 0.18885675072669983, "learning_rate": 5.922500000000001e-05, "loss": 0.0192, "step": 2370 }, { "grad_norm": 0.3611180782318115, "learning_rate": 5.9475e-05, "loss": 0.0263, "step": 2380 }, { "grad_norm": 0.3449031710624695, "learning_rate": 5.9724999999999995e-05, "loss": 0.0228, "step": 2390 }, { "grad_norm": 0.287273645401001, "learning_rate": 5.9975e-05, "loss": 0.0345, "step": 2400 }, { "grad_norm": 0.34233564138412476, "learning_rate": 6.0225e-05, "loss": 0.0224, "step": 2410 }, { "grad_norm": 0.48270800709724426, "learning_rate": 6.0475000000000006e-05, "loss": 0.035, "step": 2420 }, { "grad_norm": 0.23786301910877228, "learning_rate": 6.0725e-05, "loss": 0.0222, "step": 2430 }, { "grad_norm": 0.32582470774650574, "learning_rate": 6.0975000000000005e-05, "loss": 0.0222, "step": 2440 }, { "grad_norm": 0.29652464389801025, "learning_rate": 6.1225e-05, "loss": 0.016, "step": 2450 }, { "grad_norm": 0.3680236339569092, "learning_rate": 6.1475e-05, "loss": 0.0227, "step": 2460 }, { "grad_norm": 0.2105773687362671, "learning_rate": 6.1725e-05, "loss": 0.0173, "step": 2470 }, { "grad_norm": 0.27500438690185547, "learning_rate": 6.1975e-05, "loss": 0.0238, "step": 2480 }, { "grad_norm": 0.2836405634880066, "learning_rate": 6.2225e-05, "loss": 0.0177, "step": 2490 }, { "grad_norm": 0.15291047096252441, "learning_rate": 6.2475e-05, "loss": 0.0157, "step": 2500 }, { "grad_norm": 0.30264371633529663, "learning_rate": 6.2725e-05, "loss": 0.0165, "step": 2510 }, { "grad_norm": 0.17121624946594238, "learning_rate": 6.297500000000001e-05, "loss": 0.0177, "step": 2520 }, { "grad_norm": 0.29929065704345703, "learning_rate": 6.3225e-05, "loss": 0.0134, "step": 2530 }, { "grad_norm": 0.16253378987312317, "learning_rate": 6.347500000000001e-05, "loss": 0.0148, "step": 2540 }, { "grad_norm": 0.21832650899887085, "learning_rate": 6.3725e-05, "loss": 0.0172, "step": 2550 }, { "grad_norm": 0.3210463523864746, "learning_rate": 6.397500000000001e-05, "loss": 0.0227, "step": 2560 }, { "grad_norm": 0.31201961636543274, "learning_rate": 6.4225e-05, "loss": 0.0169, "step": 2570 }, { "grad_norm": 0.24349340796470642, "learning_rate": 6.447500000000001e-05, "loss": 0.0236, "step": 2580 }, { "grad_norm": 0.3496743440628052, "learning_rate": 6.4725e-05, "loss": 0.0159, "step": 2590 }, { "grad_norm": 0.23485195636749268, "learning_rate": 6.497500000000001e-05, "loss": 0.0212, "step": 2600 }, { "grad_norm": 0.3386506140232086, "learning_rate": 6.5225e-05, "loss": 0.02, "step": 2610 }, { "grad_norm": 0.1988300234079361, "learning_rate": 6.5475e-05, "loss": 0.0181, "step": 2620 }, { "grad_norm": 0.2828742563724518, "learning_rate": 6.5725e-05, "loss": 0.0181, "step": 2630 }, { "grad_norm": 0.35073861479759216, "learning_rate": 6.5975e-05, "loss": 0.0244, "step": 2640 }, { "grad_norm": 0.30278998613357544, "learning_rate": 6.6225e-05, "loss": 0.0228, "step": 2650 }, { "grad_norm": 0.33118388056755066, "learning_rate": 6.6475e-05, "loss": 0.0232, "step": 2660 }, { "grad_norm": 0.26037999987602234, "learning_rate": 6.672500000000001e-05, "loss": 0.0245, "step": 2670 }, { "grad_norm": 1.2005596160888672, "learning_rate": 6.6975e-05, "loss": 0.0341, "step": 2680 }, { "grad_norm": 0.21156920492649078, "learning_rate": 6.722500000000001e-05, "loss": 0.0239, "step": 2690 }, { "grad_norm": 0.3726041316986084, "learning_rate": 6.7475e-05, "loss": 0.0208, "step": 2700 }, { "grad_norm": 0.3427751362323761, "learning_rate": 6.7725e-05, "loss": 0.0233, "step": 2710 }, { "grad_norm": 0.45008042454719543, "learning_rate": 6.7975e-05, "loss": 0.0373, "step": 2720 }, { "grad_norm": 0.3463032841682434, "learning_rate": 6.8225e-05, "loss": 0.0225, "step": 2730 }, { "grad_norm": 0.30263206362724304, "learning_rate": 6.8475e-05, "loss": 0.0199, "step": 2740 }, { "grad_norm": 0.2695619761943817, "learning_rate": 6.8725e-05, "loss": 0.0195, "step": 2750 }, { "grad_norm": 0.3872945010662079, "learning_rate": 6.8975e-05, "loss": 0.019, "step": 2760 }, { "grad_norm": 0.22936345636844635, "learning_rate": 6.9225e-05, "loss": 0.022, "step": 2770 }, { "grad_norm": 0.3054845333099365, "learning_rate": 6.9475e-05, "loss": 0.0245, "step": 2780 }, { "grad_norm": 0.22854548692703247, "learning_rate": 6.9725e-05, "loss": 0.0221, "step": 2790 }, { "grad_norm": 0.4766150712966919, "learning_rate": 6.997500000000001e-05, "loss": 0.0202, "step": 2800 }, { "grad_norm": 0.3722187578678131, "learning_rate": 7.022500000000001e-05, "loss": 0.0342, "step": 2810 }, { "grad_norm": 0.25570985674858093, "learning_rate": 7.0475e-05, "loss": 0.0217, "step": 2820 }, { "grad_norm": 0.2234734743833542, "learning_rate": 7.072500000000001e-05, "loss": 0.0209, "step": 2830 }, { "grad_norm": 0.334346741437912, "learning_rate": 7.0975e-05, "loss": 0.0315, "step": 2840 }, { "grad_norm": 0.2891973853111267, "learning_rate": 7.122500000000001e-05, "loss": 0.0254, "step": 2850 }, { "grad_norm": 0.33743810653686523, "learning_rate": 7.1475e-05, "loss": 0.0251, "step": 2860 }, { "grad_norm": 0.18732210993766785, "learning_rate": 7.172500000000001e-05, "loss": 0.016, "step": 2870 }, { "grad_norm": 0.708278477191925, "learning_rate": 7.1975e-05, "loss": 0.0204, "step": 2880 }, { "grad_norm": 0.1786675602197647, "learning_rate": 7.2225e-05, "loss": 0.0165, "step": 2890 }, { "grad_norm": 0.36573660373687744, "learning_rate": 7.2475e-05, "loss": 0.0307, "step": 2900 }, { "grad_norm": 0.46843478083610535, "learning_rate": 7.272499999999999e-05, "loss": 0.0193, "step": 2910 }, { "grad_norm": 0.173651322722435, "learning_rate": 7.2975e-05, "loss": 0.0238, "step": 2920 }, { "grad_norm": 0.3022743761539459, "learning_rate": 7.3225e-05, "loss": 0.016, "step": 2930 }, { "grad_norm": 0.24419209361076355, "learning_rate": 7.347500000000001e-05, "loss": 0.0213, "step": 2940 }, { "grad_norm": 0.229168102145195, "learning_rate": 7.3725e-05, "loss": 0.0145, "step": 2950 }, { "grad_norm": 0.21088118851184845, "learning_rate": 7.397500000000001e-05, "loss": 0.0267, "step": 2960 }, { "grad_norm": 0.2709707021713257, "learning_rate": 7.4225e-05, "loss": 0.0194, "step": 2970 }, { "grad_norm": 0.24641430377960205, "learning_rate": 7.447500000000001e-05, "loss": 0.0238, "step": 2980 }, { "grad_norm": 0.2908385097980499, "learning_rate": 7.4725e-05, "loss": 0.0233, "step": 2990 }, { "grad_norm": 0.23369471728801727, "learning_rate": 7.4975e-05, "loss": 0.022, "step": 3000 }, { "grad_norm": 0.22918955981731415, "learning_rate": 7.5225e-05, "loss": 0.0161, "step": 3010 }, { "grad_norm": 0.20502886176109314, "learning_rate": 7.5475e-05, "loss": 0.018, "step": 3020 }, { "grad_norm": 0.34633657336235046, "learning_rate": 7.5725e-05, "loss": 0.02, "step": 3030 }, { "grad_norm": 0.34502118825912476, "learning_rate": 7.5975e-05, "loss": 0.0205, "step": 3040 }, { "grad_norm": 0.3271520137786865, "learning_rate": 7.6225e-05, "loss": 0.0271, "step": 3050 }, { "grad_norm": 0.2793262302875519, "learning_rate": 7.6475e-05, "loss": 0.0243, "step": 3060 }, { "grad_norm": 0.4010169208049774, "learning_rate": 7.672500000000001e-05, "loss": 0.0235, "step": 3070 }, { "grad_norm": 0.2355024516582489, "learning_rate": 7.697500000000001e-05, "loss": 0.0275, "step": 3080 }, { "grad_norm": 0.4558001756668091, "learning_rate": 7.722500000000001e-05, "loss": 0.0308, "step": 3090 }, { "grad_norm": 0.36604300141334534, "learning_rate": 7.747500000000001e-05, "loss": 0.0188, "step": 3100 }, { "grad_norm": 0.3211832046508789, "learning_rate": 7.7725e-05, "loss": 0.0183, "step": 3110 }, { "grad_norm": 0.23686030507087708, "learning_rate": 7.797500000000001e-05, "loss": 0.017, "step": 3120 }, { "grad_norm": 0.4162594676017761, "learning_rate": 7.8225e-05, "loss": 0.0175, "step": 3130 }, { "grad_norm": 0.2035870999097824, "learning_rate": 7.8475e-05, "loss": 0.0237, "step": 3140 }, { "grad_norm": 0.23911459743976593, "learning_rate": 7.8725e-05, "loss": 0.0193, "step": 3150 }, { "grad_norm": 0.3446142077445984, "learning_rate": 7.8975e-05, "loss": 0.0244, "step": 3160 }, { "grad_norm": 0.201637402176857, "learning_rate": 7.9225e-05, "loss": 0.0154, "step": 3170 }, { "grad_norm": 0.2688902020454407, "learning_rate": 7.9475e-05, "loss": 0.0291, "step": 3180 }, { "grad_norm": 0.27691948413848877, "learning_rate": 7.9725e-05, "loss": 0.0284, "step": 3190 }, { "grad_norm": 0.2846969962120056, "learning_rate": 7.9975e-05, "loss": 0.0248, "step": 3200 }, { "grad_norm": 0.3370182514190674, "learning_rate": 8.022500000000001e-05, "loss": 0.0197, "step": 3210 }, { "grad_norm": 0.48870766162872314, "learning_rate": 8.0475e-05, "loss": 0.0314, "step": 3220 }, { "grad_norm": 0.4210657775402069, "learning_rate": 8.072500000000001e-05, "loss": 0.0222, "step": 3230 }, { "grad_norm": 0.1794518232345581, "learning_rate": 8.0975e-05, "loss": 0.0194, "step": 3240 }, { "grad_norm": 0.21340401470661163, "learning_rate": 8.122500000000001e-05, "loss": 0.0334, "step": 3250 }, { "grad_norm": 0.3817152678966522, "learning_rate": 8.1475e-05, "loss": 0.0215, "step": 3260 }, { "grad_norm": 0.26284059882164, "learning_rate": 8.172500000000001e-05, "loss": 0.0193, "step": 3270 }, { "grad_norm": 0.2569011151790619, "learning_rate": 8.1975e-05, "loss": 0.018, "step": 3280 }, { "grad_norm": 0.31211021542549133, "learning_rate": 8.2225e-05, "loss": 0.0211, "step": 3290 }, { "grad_norm": 0.3401640057563782, "learning_rate": 8.2475e-05, "loss": 0.0273, "step": 3300 }, { "grad_norm": 0.49753811955451965, "learning_rate": 8.2725e-05, "loss": 0.0334, "step": 3310 }, { "grad_norm": 0.253807008266449, "learning_rate": 8.2975e-05, "loss": 0.0317, "step": 3320 }, { "grad_norm": 0.28469276428222656, "learning_rate": 8.3225e-05, "loss": 0.027, "step": 3330 }, { "grad_norm": 0.3768412172794342, "learning_rate": 8.347500000000001e-05, "loss": 0.0238, "step": 3340 }, { "grad_norm": 0.242493137717247, "learning_rate": 8.3725e-05, "loss": 0.0188, "step": 3350 }, { "grad_norm": 0.3826657831668854, "learning_rate": 8.397500000000001e-05, "loss": 0.0183, "step": 3360 }, { "grad_norm": 0.3275626599788666, "learning_rate": 8.422500000000001e-05, "loss": 0.0288, "step": 3370 }, { "grad_norm": 0.4674336612224579, "learning_rate": 8.447500000000001e-05, "loss": 0.044, "step": 3380 }, { "grad_norm": 0.28299981355667114, "learning_rate": 8.4725e-05, "loss": 0.0255, "step": 3390 }, { "grad_norm": 0.34027865529060364, "learning_rate": 8.4975e-05, "loss": 0.0274, "step": 3400 }, { "grad_norm": 0.3117365837097168, "learning_rate": 8.5225e-05, "loss": 0.0265, "step": 3410 }, { "grad_norm": 0.3808150589466095, "learning_rate": 8.5475e-05, "loss": 0.0186, "step": 3420 }, { "grad_norm": 0.27859798073768616, "learning_rate": 8.5725e-05, "loss": 0.0215, "step": 3430 }, { "grad_norm": 0.30101850628852844, "learning_rate": 8.5975e-05, "loss": 0.0326, "step": 3440 }, { "grad_norm": 0.21741880476474762, "learning_rate": 8.6225e-05, "loss": 0.0228, "step": 3450 }, { "grad_norm": 0.2278010994195938, "learning_rate": 8.6475e-05, "loss": 0.0179, "step": 3460 }, { "grad_norm": 0.24561958014965057, "learning_rate": 8.672500000000001e-05, "loss": 0.0315, "step": 3470 }, { "grad_norm": 0.46005627512931824, "learning_rate": 8.6975e-05, "loss": 0.0285, "step": 3480 }, { "grad_norm": 0.24922208487987518, "learning_rate": 8.7225e-05, "loss": 0.0274, "step": 3490 }, { "grad_norm": 0.40026605129241943, "learning_rate": 8.747500000000001e-05, "loss": 0.0184, "step": 3500 }, { "grad_norm": 0.3433447480201721, "learning_rate": 8.7725e-05, "loss": 0.0251, "step": 3510 }, { "grad_norm": 0.3053664565086365, "learning_rate": 8.797500000000001e-05, "loss": 0.0215, "step": 3520 }, { "grad_norm": 0.31211933493614197, "learning_rate": 8.8225e-05, "loss": 0.0218, "step": 3530 }, { "grad_norm": 0.2842128276824951, "learning_rate": 8.847500000000001e-05, "loss": 0.0202, "step": 3540 }, { "grad_norm": 0.19863305985927582, "learning_rate": 8.8725e-05, "loss": 0.0185, "step": 3550 }, { "grad_norm": 0.26524844765663147, "learning_rate": 8.897500000000001e-05, "loss": 0.0177, "step": 3560 }, { "grad_norm": 0.29895225167274475, "learning_rate": 8.9225e-05, "loss": 0.0223, "step": 3570 }, { "grad_norm": 0.39050787687301636, "learning_rate": 8.9475e-05, "loss": 0.0279, "step": 3580 }, { "grad_norm": 0.37314730882644653, "learning_rate": 8.9725e-05, "loss": 0.0282, "step": 3590 }, { "grad_norm": 0.30075669288635254, "learning_rate": 8.9975e-05, "loss": 0.025, "step": 3600 }, { "grad_norm": 0.2968229651451111, "learning_rate": 9.0225e-05, "loss": 0.0225, "step": 3610 }, { "grad_norm": 0.2996150553226471, "learning_rate": 9.0475e-05, "loss": 0.0221, "step": 3620 }, { "grad_norm": 0.36501702666282654, "learning_rate": 9.072500000000001e-05, "loss": 0.0184, "step": 3630 }, { "grad_norm": 0.16352909803390503, "learning_rate": 9.0975e-05, "loss": 0.0203, "step": 3640 }, { "grad_norm": 0.41416221857070923, "learning_rate": 9.122500000000001e-05, "loss": 0.0189, "step": 3650 }, { "grad_norm": 0.3035706579685211, "learning_rate": 9.1475e-05, "loss": 0.0207, "step": 3660 }, { "grad_norm": 0.326854407787323, "learning_rate": 9.172500000000001e-05, "loss": 0.0269, "step": 3670 }, { "grad_norm": 0.3140025734901428, "learning_rate": 9.1975e-05, "loss": 0.0271, "step": 3680 }, { "grad_norm": 0.43138718605041504, "learning_rate": 9.2225e-05, "loss": 0.0276, "step": 3690 }, { "grad_norm": 0.3243575394153595, "learning_rate": 9.2475e-05, "loss": 0.0243, "step": 3700 }, { "grad_norm": 0.35402241349220276, "learning_rate": 9.2725e-05, "loss": 0.0224, "step": 3710 }, { "grad_norm": 0.29233554005622864, "learning_rate": 9.2975e-05, "loss": 0.0243, "step": 3720 }, { "grad_norm": 0.21791400015354156, "learning_rate": 9.3225e-05, "loss": 0.0219, "step": 3730 }, { "grad_norm": 0.29340073466300964, "learning_rate": 9.3475e-05, "loss": 0.0291, "step": 3740 }, { "grad_norm": 0.3430042564868927, "learning_rate": 9.3725e-05, "loss": 0.0326, "step": 3750 }, { "grad_norm": 0.2454020380973816, "learning_rate": 9.397500000000001e-05, "loss": 0.0314, "step": 3760 }, { "grad_norm": 0.35556092858314514, "learning_rate": 9.422500000000001e-05, "loss": 0.0252, "step": 3770 }, { "grad_norm": 0.26724496483802795, "learning_rate": 9.4475e-05, "loss": 0.0269, "step": 3780 }, { "grad_norm": 0.2787385880947113, "learning_rate": 9.472500000000001e-05, "loss": 0.0274, "step": 3790 }, { "grad_norm": 0.39022231101989746, "learning_rate": 9.4975e-05, "loss": 0.0418, "step": 3800 }, { "grad_norm": 0.205674409866333, "learning_rate": 9.522500000000001e-05, "loss": 0.021, "step": 3810 }, { "grad_norm": 0.39544081687927246, "learning_rate": 9.5475e-05, "loss": 0.0272, "step": 3820 }, { "grad_norm": 0.2493753731250763, "learning_rate": 9.572500000000001e-05, "loss": 0.0215, "step": 3830 }, { "grad_norm": 0.4904313087463379, "learning_rate": 9.5975e-05, "loss": 0.0221, "step": 3840 }, { "grad_norm": 0.33078533411026, "learning_rate": 9.622500000000001e-05, "loss": 0.0316, "step": 3850 }, { "grad_norm": 0.39157068729400635, "learning_rate": 9.6475e-05, "loss": 0.0197, "step": 3860 }, { "grad_norm": 0.5064399838447571, "learning_rate": 9.6725e-05, "loss": 0.0271, "step": 3870 }, { "grad_norm": 0.4266607463359833, "learning_rate": 9.6975e-05, "loss": 0.0253, "step": 3880 }, { "grad_norm": 0.34235653281211853, "learning_rate": 9.7225e-05, "loss": 0.0265, "step": 3890 }, { "grad_norm": 0.42793887853622437, "learning_rate": 9.747500000000001e-05, "loss": 0.0281, "step": 3900 }, { "grad_norm": 0.3670414388179779, "learning_rate": 9.7725e-05, "loss": 0.0234, "step": 3910 }, { "grad_norm": 0.719860851764679, "learning_rate": 9.797500000000001e-05, "loss": 0.0302, "step": 3920 }, { "grad_norm": 0.2781868875026703, "learning_rate": 9.8225e-05, "loss": 0.0201, "step": 3930 }, { "grad_norm": 0.3784807622432709, "learning_rate": 9.847500000000001e-05, "loss": 0.0329, "step": 3940 }, { "grad_norm": 0.2727004885673523, "learning_rate": 9.8725e-05, "loss": 0.0287, "step": 3950 }, { "grad_norm": 0.3367099165916443, "learning_rate": 9.897500000000001e-05, "loss": 0.0374, "step": 3960 }, { "grad_norm": 0.31174275279045105, "learning_rate": 9.9225e-05, "loss": 0.0229, "step": 3970 }, { "grad_norm": 0.40122276544570923, "learning_rate": 9.9475e-05, "loss": 0.0278, "step": 3980 }, { "grad_norm": 0.41432562470436096, "learning_rate": 9.9725e-05, "loss": 0.0266, "step": 3990 }, { "grad_norm": 0.3610788583755493, "learning_rate": 9.9975e-05, "loss": 0.0277, "step": 4000 }, { "grad_norm": 0.38288208842277527, "learning_rate": 9.999999653982884e-05, "loss": 0.0247, "step": 4010 }, { "grad_norm": 0.7986325025558472, "learning_rate": 9.999998457874392e-05, "loss": 0.0242, "step": 4020 }, { "grad_norm": 0.30207476019859314, "learning_rate": 9.999996407402913e-05, "loss": 0.029, "step": 4030 }, { "grad_norm": 0.25106462836265564, "learning_rate": 9.999993502568801e-05, "loss": 0.0175, "step": 4040 }, { "grad_norm": 0.40821757912635803, "learning_rate": 9.999989743372548e-05, "loss": 0.0258, "step": 4050 }, { "grad_norm": 0.28312110900878906, "learning_rate": 9.999985129814798e-05, "loss": 0.0239, "step": 4060 }, { "grad_norm": 0.2822045087814331, "learning_rate": 9.99997966189634e-05, "loss": 0.0209, "step": 4070 }, { "grad_norm": 0.22758616507053375, "learning_rate": 9.999973339618107e-05, "loss": 0.0251, "step": 4080 }, { "grad_norm": 0.36409491300582886, "learning_rate": 9.999966162981179e-05, "loss": 0.0289, "step": 4090 }, { "grad_norm": 0.3023951053619385, "learning_rate": 9.999958131986784e-05, "loss": 0.0229, "step": 4100 }, { "grad_norm": 0.32163479924201965, "learning_rate": 9.999949246636293e-05, "loss": 0.0253, "step": 4110 }, { "grad_norm": 0.5454026460647583, "learning_rate": 9.999939506931224e-05, "loss": 0.0331, "step": 4120 }, { "grad_norm": 0.2420809119939804, "learning_rate": 9.999928912873243e-05, "loss": 0.0242, "step": 4130 }, { "grad_norm": 0.32276761531829834, "learning_rate": 9.999917464464159e-05, "loss": 0.0304, "step": 4140 }, { "grad_norm": 0.9476318955421448, "learning_rate": 9.999905161705929e-05, "loss": 0.037, "step": 4150 }, { "grad_norm": 0.3511921763420105, "learning_rate": 9.999892004600653e-05, "loss": 0.0246, "step": 4160 }, { "grad_norm": 0.37165728211402893, "learning_rate": 9.999877993150581e-05, "loss": 0.0223, "step": 4170 }, { "grad_norm": 0.2757931053638458, "learning_rate": 9.999863127358108e-05, "loss": 0.021, "step": 4180 }, { "grad_norm": 0.398709237575531, "learning_rate": 9.999847407225773e-05, "loss": 0.0256, "step": 4190 }, { "grad_norm": 0.24709248542785645, "learning_rate": 9.999830832756262e-05, "loss": 0.0228, "step": 4200 }, { "grad_norm": 0.2311801314353943, "learning_rate": 9.999813403952407e-05, "loss": 0.0229, "step": 4210 }, { "grad_norm": 0.2827320098876953, "learning_rate": 9.999795120817187e-05, "loss": 0.0315, "step": 4220 }, { "grad_norm": 0.5579318404197693, "learning_rate": 9.999775983353725e-05, "loss": 0.0415, "step": 4230 }, { "grad_norm": 0.3959321975708008, "learning_rate": 9.999755991565292e-05, "loss": 0.0324, "step": 4240 }, { "grad_norm": 0.33035707473754883, "learning_rate": 9.999735145455303e-05, "loss": 0.035, "step": 4250 }, { "grad_norm": 0.38905590772628784, "learning_rate": 9.99971344502732e-05, "loss": 0.0299, "step": 4260 }, { "grad_norm": 0.3741612732410431, "learning_rate": 9.999690890285053e-05, "loss": 0.0299, "step": 4270 }, { "grad_norm": 0.3880236744880676, "learning_rate": 9.999667481232356e-05, "loss": 0.023, "step": 4280 }, { "grad_norm": 0.25325942039489746, "learning_rate": 9.999643217873225e-05, "loss": 0.0236, "step": 4290 }, { "grad_norm": 0.2768535614013672, "learning_rate": 9.999618100211809e-05, "loss": 0.02, "step": 4300 }, { "grad_norm": 0.26310157775878906, "learning_rate": 9.999592128252402e-05, "loss": 0.0251, "step": 4310 }, { "grad_norm": 0.5444860458374023, "learning_rate": 9.999565301999437e-05, "loss": 0.0407, "step": 4320 }, { "grad_norm": 0.3433971703052521, "learning_rate": 9.999537621457502e-05, "loss": 0.0257, "step": 4330 }, { "grad_norm": 0.32996198534965515, "learning_rate": 9.999509086631323e-05, "loss": 0.021, "step": 4340 }, { "grad_norm": 0.24119959771633148, "learning_rate": 9.99947969752578e-05, "loss": 0.0227, "step": 4350 }, { "grad_norm": 0.2698550522327423, "learning_rate": 9.999449454145891e-05, "loss": 0.0245, "step": 4360 }, { "grad_norm": 0.36075693368911743, "learning_rate": 9.999418356496827e-05, "loss": 0.0232, "step": 4370 }, { "grad_norm": 0.27091237902641296, "learning_rate": 9.999386404583899e-05, "loss": 0.0238, "step": 4380 }, { "grad_norm": 0.2742140293121338, "learning_rate": 9.999353598412568e-05, "loss": 0.0217, "step": 4390 }, { "grad_norm": 0.393194317817688, "learning_rate": 9.999319937988442e-05, "loss": 0.0266, "step": 4400 }, { "grad_norm": 0.3747082054615021, "learning_rate": 9.999285423317268e-05, "loss": 0.0352, "step": 4410 }, { "grad_norm": 0.26385170221328735, "learning_rate": 9.999250054404947e-05, "loss": 0.0418, "step": 4420 }, { "grad_norm": 0.30820563435554504, "learning_rate": 9.99921383125752e-05, "loss": 0.0293, "step": 4430 }, { "grad_norm": 0.3904531002044678, "learning_rate": 9.99917675388118e-05, "loss": 0.0303, "step": 4440 }, { "grad_norm": 0.49986350536346436, "learning_rate": 9.99913882228226e-05, "loss": 0.0325, "step": 4450 }, { "grad_norm": 0.23618778586387634, "learning_rate": 9.999100036467242e-05, "loss": 0.0301, "step": 4460 }, { "grad_norm": 0.24064171314239502, "learning_rate": 9.999060396442753e-05, "loss": 0.0229, "step": 4470 }, { "grad_norm": 0.29355746507644653, "learning_rate": 9.999019902215566e-05, "loss": 0.0292, "step": 4480 }, { "grad_norm": 0.42953965067863464, "learning_rate": 9.998978553792602e-05, "loss": 0.0274, "step": 4490 }, { "grad_norm": 0.2826063632965088, "learning_rate": 9.998936351180926e-05, "loss": 0.0244, "step": 4500 }, { "grad_norm": 0.24044360220432281, "learning_rate": 9.998893294387747e-05, "loss": 0.0422, "step": 4510 }, { "grad_norm": 0.4006167948246002, "learning_rate": 9.998849383420426e-05, "loss": 0.0322, "step": 4520 }, { "grad_norm": 0.21610654890537262, "learning_rate": 9.998804618286465e-05, "loss": 0.0242, "step": 4530 }, { "grad_norm": 0.4394013285636902, "learning_rate": 9.99875899899351e-05, "loss": 0.0344, "step": 4540 }, { "grad_norm": 0.3721255362033844, "learning_rate": 9.99871252554936e-05, "loss": 0.0351, "step": 4550 }, { "grad_norm": 0.25116756558418274, "learning_rate": 9.998665197961955e-05, "loss": 0.022, "step": 4560 }, { "grad_norm": 0.322481632232666, "learning_rate": 9.998617016239379e-05, "loss": 0.0312, "step": 4570 }, { "grad_norm": 0.40373337268829346, "learning_rate": 9.998567980389869e-05, "loss": 0.0302, "step": 4580 }, { "grad_norm": 0.2638698220252991, "learning_rate": 9.998518090421802e-05, "loss": 0.0269, "step": 4590 }, { "grad_norm": 0.27653175592422485, "learning_rate": 9.998467346343703e-05, "loss": 0.0261, "step": 4600 }, { "grad_norm": 0.29895225167274475, "learning_rate": 9.998415748164243e-05, "loss": 0.0285, "step": 4610 }, { "grad_norm": 0.3292335867881775, "learning_rate": 9.998363295892238e-05, "loss": 0.0235, "step": 4620 }, { "grad_norm": 0.33138012886047363, "learning_rate": 9.998309989536652e-05, "loss": 0.0344, "step": 4630 }, { "grad_norm": 0.30174076557159424, "learning_rate": 9.998255829106593e-05, "loss": 0.0339, "step": 4640 }, { "grad_norm": 0.34379804134368896, "learning_rate": 9.998200814611316e-05, "loss": 0.0215, "step": 4650 }, { "grad_norm": 0.30285435914993286, "learning_rate": 9.998144946060219e-05, "loss": 0.0273, "step": 4660 }, { "grad_norm": 0.42801573872566223, "learning_rate": 9.998088223462852e-05, "loss": 0.0233, "step": 4670 }, { "grad_norm": 0.31989458203315735, "learning_rate": 9.998030646828905e-05, "loss": 0.0241, "step": 4680 }, { "grad_norm": 0.3235442340373993, "learning_rate": 9.997972216168217e-05, "loss": 0.0274, "step": 4690 }, { "grad_norm": 0.27618658542633057, "learning_rate": 9.997912931490771e-05, "loss": 0.0247, "step": 4700 }, { "grad_norm": 0.2647986114025116, "learning_rate": 9.9978527928067e-05, "loss": 0.0287, "step": 4710 }, { "grad_norm": 0.2828710675239563, "learning_rate": 9.997791800126277e-05, "loss": 0.022, "step": 4720 }, { "grad_norm": 0.2537602186203003, "learning_rate": 9.997729953459927e-05, "loss": 0.0301, "step": 4730 }, { "grad_norm": 0.2809496223926544, "learning_rate": 9.997667252818214e-05, "loss": 0.0267, "step": 4740 }, { "grad_norm": 0.3296492099761963, "learning_rate": 9.997603698211855e-05, "loss": 0.0217, "step": 4750 }, { "grad_norm": 0.45058000087738037, "learning_rate": 9.99753928965171e-05, "loss": 0.0275, "step": 4760 }, { "grad_norm": 0.3515799045562744, "learning_rate": 9.997474027148781e-05, "loss": 0.0257, "step": 4770 }, { "grad_norm": 0.3848544955253601, "learning_rate": 9.997407910714223e-05, "loss": 0.0313, "step": 4780 }, { "grad_norm": 0.2649591565132141, "learning_rate": 9.997340940359332e-05, "loss": 0.0217, "step": 4790 }, { "grad_norm": 0.2594106197357178, "learning_rate": 9.997273116095552e-05, "loss": 0.023, "step": 4800 }, { "grad_norm": 0.3387317359447479, "learning_rate": 9.997204437934473e-05, "loss": 0.0268, "step": 4810 }, { "grad_norm": 0.3899420499801636, "learning_rate": 9.997134905887829e-05, "loss": 0.0235, "step": 4820 }, { "grad_norm": 0.32392871379852295, "learning_rate": 9.997064519967501e-05, "loss": 0.0275, "step": 4830 }, { "grad_norm": 0.2737424373626709, "learning_rate": 9.996993280185517e-05, "loss": 0.0277, "step": 4840 }, { "grad_norm": 0.2577958106994629, "learning_rate": 9.99692118655405e-05, "loss": 0.0257, "step": 4850 }, { "grad_norm": 0.6332340240478516, "learning_rate": 9.996848239085417e-05, "loss": 0.0246, "step": 4860 }, { "grad_norm": 0.28354203701019287, "learning_rate": 9.996774437792085e-05, "loss": 0.0206, "step": 4870 }, { "grad_norm": 0.30445876717567444, "learning_rate": 9.996699782686664e-05, "loss": 0.0194, "step": 4880 }, { "grad_norm": 0.34719935059547424, "learning_rate": 9.996624273781909e-05, "loss": 0.0378, "step": 4890 }, { "grad_norm": 0.3616904020309448, "learning_rate": 9.996547911090725e-05, "loss": 0.0208, "step": 4900 }, { "grad_norm": 0.2288486659526825, "learning_rate": 9.996470694626157e-05, "loss": 0.0286, "step": 4910 }, { "grad_norm": 0.27179649472236633, "learning_rate": 9.996392624401403e-05, "loss": 0.0235, "step": 4920 }, { "grad_norm": 0.3171641528606415, "learning_rate": 9.996313700429801e-05, "loss": 0.0279, "step": 4930 }, { "grad_norm": 0.2817358076572418, "learning_rate": 9.996233922724836e-05, "loss": 0.0187, "step": 4940 }, { "grad_norm": 0.30165746808052063, "learning_rate": 9.996153291300141e-05, "loss": 0.0304, "step": 4950 }, { "grad_norm": 0.20021414756774902, "learning_rate": 9.996071806169494e-05, "loss": 0.0279, "step": 4960 }, { "grad_norm": 0.31168800592422485, "learning_rate": 9.995989467346817e-05, "loss": 0.0307, "step": 4970 }, { "grad_norm": 0.3067048490047455, "learning_rate": 9.995906274846183e-05, "loss": 0.0266, "step": 4980 }, { "grad_norm": 0.3829570412635803, "learning_rate": 9.995822228681803e-05, "loss": 0.0248, "step": 4990 }, { "grad_norm": 0.29557082056999207, "learning_rate": 9.99573732886804e-05, "loss": 0.0346, "step": 5000 }, { "grad_norm": 0.3663511872291565, "learning_rate": 9.995651575419402e-05, "loss": 0.0321, "step": 5010 }, { "grad_norm": 0.5087868571281433, "learning_rate": 9.995564968350541e-05, "loss": 0.0307, "step": 5020 }, { "grad_norm": 0.28503045439720154, "learning_rate": 9.995477507676256e-05, "loss": 0.019, "step": 5030 }, { "grad_norm": 0.3032188415527344, "learning_rate": 9.995389193411493e-05, "loss": 0.0264, "step": 5040 }, { "grad_norm": 0.31787097454071045, "learning_rate": 9.995300025571339e-05, "loss": 0.0233, "step": 5050 }, { "grad_norm": 0.16842243075370789, "learning_rate": 9.995210004171034e-05, "loss": 0.0164, "step": 5060 }, { "grad_norm": 0.3796558082103729, "learning_rate": 9.995119129225956e-05, "loss": 0.0299, "step": 5070 }, { "grad_norm": 0.24024978280067444, "learning_rate": 9.995027400751637e-05, "loss": 0.0277, "step": 5080 }, { "grad_norm": 0.39480409026145935, "learning_rate": 9.994934818763751e-05, "loss": 0.023, "step": 5090 }, { "grad_norm": 0.3678234815597534, "learning_rate": 9.994841383278115e-05, "loss": 0.0224, "step": 5100 }, { "grad_norm": 0.4681873321533203, "learning_rate": 9.994747094310695e-05, "loss": 0.0266, "step": 5110 }, { "grad_norm": 0.34140995144844055, "learning_rate": 9.994651951877604e-05, "loss": 0.0352, "step": 5120 }, { "grad_norm": 0.3284502327442169, "learning_rate": 9.994555955995099e-05, "loss": 0.0315, "step": 5130 }, { "grad_norm": 0.2608861029148102, "learning_rate": 9.994459106679581e-05, "loss": 0.0272, "step": 5140 }, { "grad_norm": 0.2242289036512375, "learning_rate": 9.994361403947603e-05, "loss": 0.0252, "step": 5150 }, { "grad_norm": 0.3108457326889038, "learning_rate": 9.994262847815854e-05, "loss": 0.0345, "step": 5160 }, { "grad_norm": 0.36117956042289734, "learning_rate": 9.99416343830118e-05, "loss": 0.0264, "step": 5170 }, { "grad_norm": 0.32842716574668884, "learning_rate": 9.994063175420565e-05, "loss": 0.0205, "step": 5180 }, { "grad_norm": 0.35476699471473694, "learning_rate": 9.99396205919114e-05, "loss": 0.0373, "step": 5190 }, { "grad_norm": 0.2380100041627884, "learning_rate": 9.993860089630185e-05, "loss": 0.0226, "step": 5200 }, { "grad_norm": 0.2779267430305481, "learning_rate": 9.993757266755123e-05, "loss": 0.0293, "step": 5210 }, { "grad_norm": 0.3813204765319824, "learning_rate": 9.993653590583522e-05, "loss": 0.0241, "step": 5220 }, { "grad_norm": 0.3203929364681244, "learning_rate": 9.993549061133102e-05, "loss": 0.03, "step": 5230 }, { "grad_norm": 0.31544411182403564, "learning_rate": 9.993443678421719e-05, "loss": 0.0344, "step": 5240 }, { "grad_norm": 0.47993004322052, "learning_rate": 9.993337442467384e-05, "loss": 0.03, "step": 5250 }, { "grad_norm": 0.41438522934913635, "learning_rate": 9.993230353288248e-05, "loss": 0.0259, "step": 5260 }, { "grad_norm": 0.3009912669658661, "learning_rate": 9.993122410902608e-05, "loss": 0.0259, "step": 5270 }, { "grad_norm": 0.3660680651664734, "learning_rate": 9.993013615328912e-05, "loss": 0.0306, "step": 5280 }, { "grad_norm": 0.337118536233902, "learning_rate": 9.992903966585747e-05, "loss": 0.0274, "step": 5290 }, { "grad_norm": 0.3129914700984955, "learning_rate": 9.992793464691852e-05, "loss": 0.0206, "step": 5300 }, { "grad_norm": 0.30026182532310486, "learning_rate": 9.992682109666105e-05, "loss": 0.0249, "step": 5310 }, { "grad_norm": 0.3250908851623535, "learning_rate": 9.992569901527538e-05, "loss": 0.0222, "step": 5320 }, { "grad_norm": 0.34453174471855164, "learning_rate": 9.99245684029532e-05, "loss": 0.0316, "step": 5330 }, { "grad_norm": 0.3776600658893585, "learning_rate": 9.992342925988774e-05, "loss": 0.0272, "step": 5340 }, { "grad_norm": 0.33015596866607666, "learning_rate": 9.992228158627361e-05, "loss": 0.0249, "step": 5350 }, { "grad_norm": 0.2825014293193817, "learning_rate": 9.992112538230693e-05, "loss": 0.0291, "step": 5360 }, { "grad_norm": 0.22545777261257172, "learning_rate": 9.991996064818527e-05, "loss": 0.0215, "step": 5370 }, { "grad_norm": 0.3263667821884155, "learning_rate": 9.991878738410768e-05, "loss": 0.0211, "step": 5380 }, { "grad_norm": 0.27026450634002686, "learning_rate": 9.991760559027457e-05, "loss": 0.0189, "step": 5390 }, { "grad_norm": 0.22199241816997528, "learning_rate": 9.991641526688793e-05, "loss": 0.0296, "step": 5400 }, { "grad_norm": 0.3871157467365265, "learning_rate": 9.991521641415113e-05, "loss": 0.0252, "step": 5410 }, { "grad_norm": 0.2307693362236023, "learning_rate": 9.991400903226904e-05, "loss": 0.021, "step": 5420 }, { "grad_norm": 0.3975049555301666, "learning_rate": 9.991279312144794e-05, "loss": 0.0239, "step": 5430 }, { "grad_norm": 0.2844218909740448, "learning_rate": 9.991156868189564e-05, "loss": 0.028, "step": 5440 }, { "grad_norm": 0.32111209630966187, "learning_rate": 9.991033571382131e-05, "loss": 0.0167, "step": 5450 }, { "grad_norm": 0.24878114461898804, "learning_rate": 9.990909421743569e-05, "loss": 0.0213, "step": 5460 }, { "grad_norm": 0.23194493353366852, "learning_rate": 9.990784419295085e-05, "loss": 0.0204, "step": 5470 }, { "grad_norm": 0.2439504712820053, "learning_rate": 9.990658564058044e-05, "loss": 0.0208, "step": 5480 }, { "grad_norm": 0.32902398705482483, "learning_rate": 9.990531856053948e-05, "loss": 0.0259, "step": 5490 }, { "grad_norm": 0.2874191701412201, "learning_rate": 9.99040429530445e-05, "loss": 0.0301, "step": 5500 }, { "grad_norm": 0.35623982548713684, "learning_rate": 9.990275881831346e-05, "loss": 0.0328, "step": 5510 }, { "grad_norm": 0.3565194606781006, "learning_rate": 9.990146615656577e-05, "loss": 0.0223, "step": 5520 }, { "grad_norm": 0.2818768620491028, "learning_rate": 9.990016496802233e-05, "loss": 0.0202, "step": 5530 }, { "grad_norm": 0.36808231472969055, "learning_rate": 9.989885525290548e-05, "loss": 0.0198, "step": 5540 }, { "grad_norm": 0.26355093717575073, "learning_rate": 9.989753701143897e-05, "loss": 0.0214, "step": 5550 }, { "grad_norm": 0.24700123071670532, "learning_rate": 9.989621024384812e-05, "loss": 0.0196, "step": 5560 }, { "grad_norm": 0.3401612341403961, "learning_rate": 9.989487495035959e-05, "loss": 0.0225, "step": 5570 }, { "grad_norm": 0.3749087154865265, "learning_rate": 9.989353113120156e-05, "loss": 0.03, "step": 5580 }, { "grad_norm": 0.31492000818252563, "learning_rate": 9.989217878660366e-05, "loss": 0.023, "step": 5590 }, { "grad_norm": 0.3835098445415497, "learning_rate": 9.989081791679695e-05, "loss": 0.0228, "step": 5600 }, { "grad_norm": 0.30857720971107483, "learning_rate": 9.988944852201397e-05, "loss": 0.0238, "step": 5610 }, { "grad_norm": 0.35604390501976013, "learning_rate": 9.988807060248873e-05, "loss": 0.0214, "step": 5620 }, { "grad_norm": 0.6073826551437378, "learning_rate": 9.988668415845665e-05, "loss": 0.0211, "step": 5630 }, { "grad_norm": 0.3861088156700134, "learning_rate": 9.988528919015466e-05, "loss": 0.026, "step": 5640 }, { "grad_norm": 0.3742545545101166, "learning_rate": 9.988388569782112e-05, "loss": 0.0407, "step": 5650 }, { "grad_norm": 0.3974781334400177, "learning_rate": 9.988247368169583e-05, "loss": 0.0226, "step": 5660 }, { "grad_norm": 0.3039540648460388, "learning_rate": 9.988105314202007e-05, "loss": 0.0236, "step": 5670 }, { "grad_norm": 0.23351937532424927, "learning_rate": 9.987962407903659e-05, "loss": 0.0309, "step": 5680 }, { "grad_norm": 0.5037024617195129, "learning_rate": 9.987818649298957e-05, "loss": 0.0269, "step": 5690 }, { "grad_norm": 0.2776654362678528, "learning_rate": 9.987674038412465e-05, "loss": 0.0246, "step": 5700 }, { "grad_norm": 0.30847012996673584, "learning_rate": 9.987528575268891e-05, "loss": 0.0236, "step": 5710 }, { "grad_norm": 0.2258274406194687, "learning_rate": 9.987382259893095e-05, "loss": 0.0231, "step": 5720 }, { "grad_norm": 0.27822044491767883, "learning_rate": 9.987235092310074e-05, "loss": 0.0231, "step": 5730 }, { "grad_norm": 0.29677367210388184, "learning_rate": 9.987087072544978e-05, "loss": 0.0283, "step": 5740 }, { "grad_norm": 0.20815400779247284, "learning_rate": 9.9869382006231e-05, "loss": 0.0257, "step": 5750 }, { "grad_norm": 0.2247232347726822, "learning_rate": 9.986788476569875e-05, "loss": 0.0305, "step": 5760 }, { "grad_norm": 0.33511486649513245, "learning_rate": 9.986637900410887e-05, "loss": 0.0272, "step": 5770 }, { "grad_norm": 0.35042980313301086, "learning_rate": 9.986486472171869e-05, "loss": 0.0286, "step": 5780 }, { "grad_norm": 0.3804785907268524, "learning_rate": 9.986334191878692e-05, "loss": 0.0209, "step": 5790 }, { "grad_norm": 0.36064738035202026, "learning_rate": 9.986181059557378e-05, "loss": 0.0333, "step": 5800 }, { "grad_norm": 0.21554580330848694, "learning_rate": 9.986027075234094e-05, "loss": 0.0244, "step": 5810 }, { "grad_norm": 0.25786352157592773, "learning_rate": 9.985872238935152e-05, "loss": 0.0266, "step": 5820 }, { "grad_norm": 0.28966793417930603, "learning_rate": 9.985716550687008e-05, "loss": 0.0259, "step": 5830 }, { "grad_norm": 0.303935706615448, "learning_rate": 9.985560010516264e-05, "loss": 0.0218, "step": 5840 }, { "grad_norm": 0.2772881090641022, "learning_rate": 9.985402618449668e-05, "loss": 0.0212, "step": 5850 }, { "grad_norm": 0.2974008619785309, "learning_rate": 9.985244374514118e-05, "loss": 0.0322, "step": 5860 }, { "grad_norm": 0.2841437757015228, "learning_rate": 9.985085278736651e-05, "loss": 0.0274, "step": 5870 }, { "grad_norm": 0.4945041239261627, "learning_rate": 9.984925331144452e-05, "loss": 0.0263, "step": 5880 }, { "grad_norm": 0.2835965156555176, "learning_rate": 9.984764531764851e-05, "loss": 0.0301, "step": 5890 }, { "grad_norm": 0.30312374234199524, "learning_rate": 9.984602880625326e-05, "loss": 0.0362, "step": 5900 }, { "grad_norm": 0.34292125701904297, "learning_rate": 9.9844403777535e-05, "loss": 0.0328, "step": 5910 }, { "grad_norm": 0.37531447410583496, "learning_rate": 9.984277023177135e-05, "loss": 0.0247, "step": 5920 }, { "grad_norm": 0.3024936616420746, "learning_rate": 9.984112816924148e-05, "loss": 0.0202, "step": 5930 }, { "grad_norm": 0.26810422539711, "learning_rate": 9.983947759022596e-05, "loss": 0.0229, "step": 5940 }, { "grad_norm": 0.296197772026062, "learning_rate": 9.983781849500682e-05, "loss": 0.0227, "step": 5950 }, { "grad_norm": 0.4674935042858124, "learning_rate": 9.98361508838676e-05, "loss": 0.0211, "step": 5960 }, { "grad_norm": 0.31716373562812805, "learning_rate": 9.98344747570932e-05, "loss": 0.0259, "step": 5970 }, { "grad_norm": 0.334885835647583, "learning_rate": 9.983279011497004e-05, "loss": 0.0295, "step": 5980 }, { "grad_norm": 0.49123847484588623, "learning_rate": 9.983109695778596e-05, "loss": 0.0339, "step": 5990 }, { "grad_norm": 0.33976373076438904, "learning_rate": 9.982939528583032e-05, "loss": 0.0411, "step": 6000 }, { "grad_norm": 0.33573395013809204, "learning_rate": 9.982768509939385e-05, "loss": 0.0266, "step": 6010 }, { "grad_norm": 0.27268892526626587, "learning_rate": 9.982596639876879e-05, "loss": 0.023, "step": 6020 }, { "grad_norm": 0.2639053165912628, "learning_rate": 9.982423918424881e-05, "loss": 0.0263, "step": 6030 }, { "grad_norm": 0.37173476815223694, "learning_rate": 9.982250345612908e-05, "loss": 0.0222, "step": 6040 }, { "grad_norm": 0.30356159806251526, "learning_rate": 9.982075921470611e-05, "loss": 0.0279, "step": 6050 }, { "grad_norm": 0.3019671142101288, "learning_rate": 9.981900646027802e-05, "loss": 0.0242, "step": 6060 }, { "grad_norm": 0.2821747660636902, "learning_rate": 9.981724519314425e-05, "loss": 0.0239, "step": 6070 }, { "grad_norm": 0.3390778601169586, "learning_rate": 9.981547541360581e-05, "loss": 0.028, "step": 6080 }, { "grad_norm": 0.29512086510658264, "learning_rate": 9.981369712196508e-05, "loss": 0.0285, "step": 6090 }, { "grad_norm": 0.4215666651725769, "learning_rate": 9.981191031852592e-05, "loss": 0.0324, "step": 6100 }, { "grad_norm": 0.2872384190559387, "learning_rate": 9.981011500359362e-05, "loss": 0.0217, "step": 6110 }, { "grad_norm": 0.31518086791038513, "learning_rate": 9.9808311177475e-05, "loss": 0.0258, "step": 6120 }, { "grad_norm": 0.3098181188106537, "learning_rate": 9.980649884047826e-05, "loss": 0.0151, "step": 6130 }, { "grad_norm": 0.29330357909202576, "learning_rate": 9.980467799291307e-05, "loss": 0.0229, "step": 6140 }, { "grad_norm": 0.4269962012767792, "learning_rate": 9.980284863509058e-05, "loss": 0.0444, "step": 6150 }, { "grad_norm": 0.470743328332901, "learning_rate": 9.980101076732334e-05, "loss": 0.0409, "step": 6160 }, { "grad_norm": 0.3394226133823395, "learning_rate": 9.979916438992544e-05, "loss": 0.0305, "step": 6170 }, { "grad_norm": 0.23585660755634308, "learning_rate": 9.979730950321237e-05, "loss": 0.0255, "step": 6180 }, { "grad_norm": 0.26976123452186584, "learning_rate": 9.979544610750104e-05, "loss": 0.0324, "step": 6190 }, { "grad_norm": 0.40041959285736084, "learning_rate": 9.97935742031099e-05, "loss": 0.023, "step": 6200 }, { "grad_norm": 0.37199142575263977, "learning_rate": 9.979169379035878e-05, "loss": 0.0334, "step": 6210 }, { "grad_norm": 0.39168471097946167, "learning_rate": 9.978980486956899e-05, "loss": 0.0242, "step": 6220 }, { "grad_norm": 0.23276805877685547, "learning_rate": 9.978790744106332e-05, "loss": 0.025, "step": 6230 }, { "grad_norm": 0.3504984378814697, "learning_rate": 9.978600150516594e-05, "loss": 0.0283, "step": 6240 }, { "grad_norm": 0.30579274892807007, "learning_rate": 9.978408706220259e-05, "loss": 0.0268, "step": 6250 }, { "grad_norm": 0.2973839342594147, "learning_rate": 9.978216411250032e-05, "loss": 0.0223, "step": 6260 }, { "grad_norm": 0.23549269139766693, "learning_rate": 9.978023265638778e-05, "loss": 0.0271, "step": 6270 }, { "grad_norm": 0.36465194821357727, "learning_rate": 9.977829269419495e-05, "loss": 0.0285, "step": 6280 }, { "grad_norm": 0.34856608510017395, "learning_rate": 9.977634422625335e-05, "loss": 0.028, "step": 6290 }, { "grad_norm": 0.3235543668270111, "learning_rate": 9.97743872528959e-05, "loss": 0.0336, "step": 6300 }, { "grad_norm": 0.3579272925853729, "learning_rate": 9.9772421774457e-05, "loss": 0.0284, "step": 6310 }, { "grad_norm": 0.27436864376068115, "learning_rate": 9.977044779127252e-05, "loss": 0.0277, "step": 6320 }, { "grad_norm": 0.4755459427833557, "learning_rate": 9.976846530367971e-05, "loss": 0.0237, "step": 6330 }, { "grad_norm": 0.30209097266197205, "learning_rate": 9.976647431201735e-05, "loss": 0.0392, "step": 6340 }, { "grad_norm": 0.2156686782836914, "learning_rate": 9.976447481662568e-05, "loss": 0.0247, "step": 6350 }, { "grad_norm": 0.3082420527935028, "learning_rate": 9.976246681784629e-05, "loss": 0.0281, "step": 6360 }, { "grad_norm": 0.32310208678245544, "learning_rate": 9.976045031602234e-05, "loss": 0.0259, "step": 6370 }, { "grad_norm": 0.316683292388916, "learning_rate": 9.975842531149837e-05, "loss": 0.0234, "step": 6380 }, { "grad_norm": 0.36064866185188293, "learning_rate": 9.975639180462043e-05, "loss": 0.0255, "step": 6390 }, { "grad_norm": 0.23980356752872467, "learning_rate": 9.975434979573596e-05, "loss": 0.0268, "step": 6400 }, { "grad_norm": 0.3378514051437378, "learning_rate": 9.97522992851939e-05, "loss": 0.0189, "step": 6410 }, { "grad_norm": 0.27263355255126953, "learning_rate": 9.975024027334461e-05, "loss": 0.0199, "step": 6420 }, { "grad_norm": 0.24975524842739105, "learning_rate": 9.974817276053993e-05, "loss": 0.0246, "step": 6430 }, { "grad_norm": 0.328136682510376, "learning_rate": 9.974609674713315e-05, "loss": 0.026, "step": 6440 }, { "grad_norm": 0.33349186182022095, "learning_rate": 9.9744012233479e-05, "loss": 0.0259, "step": 6450 }, { "grad_norm": 0.23469573259353638, "learning_rate": 9.974191921993366e-05, "loss": 0.026, "step": 6460 }, { "grad_norm": 0.28139033913612366, "learning_rate": 9.973981770685474e-05, "loss": 0.0232, "step": 6470 }, { "grad_norm": 0.34074878692626953, "learning_rate": 9.97377076946014e-05, "loss": 0.0222, "step": 6480 }, { "grad_norm": 0.3225656747817993, "learning_rate": 9.973558918353412e-05, "loss": 0.0333, "step": 6490 }, { "grad_norm": 0.1937873363494873, "learning_rate": 9.973346217401494e-05, "loss": 0.0255, "step": 6500 }, { "grad_norm": 0.3959302008152008, "learning_rate": 9.973132666640726e-05, "loss": 0.0195, "step": 6510 }, { "grad_norm": 0.3326432704925537, "learning_rate": 9.972918266107602e-05, "loss": 0.0242, "step": 6520 }, { "grad_norm": 0.31818845868110657, "learning_rate": 9.972703015838756e-05, "loss": 0.0239, "step": 6530 }, { "grad_norm": 0.3447970449924469, "learning_rate": 9.97248691587097e-05, "loss": 0.0309, "step": 6540 }, { "grad_norm": 0.3925192356109619, "learning_rate": 9.972269966241166e-05, "loss": 0.0232, "step": 6550 }, { "grad_norm": 0.302316278219223, "learning_rate": 9.972052166986417e-05, "loss": 0.027, "step": 6560 }, { "grad_norm": 0.41797587275505066, "learning_rate": 9.971833518143938e-05, "loss": 0.0276, "step": 6570 }, { "grad_norm": 0.33703017234802246, "learning_rate": 9.971614019751093e-05, "loss": 0.0232, "step": 6580 }, { "grad_norm": 0.2984570264816284, "learning_rate": 9.971393671845383e-05, "loss": 0.0278, "step": 6590 }, { "grad_norm": 0.281360924243927, "learning_rate": 9.971172474464464e-05, "loss": 0.0307, "step": 6600 }, { "grad_norm": 0.25909358263015747, "learning_rate": 9.97095042764613e-05, "loss": 0.0205, "step": 6610 }, { "grad_norm": 0.21794918179512024, "learning_rate": 9.970727531428324e-05, "loss": 0.0234, "step": 6620 }, { "grad_norm": 0.3068667948246002, "learning_rate": 9.970503785849132e-05, "loss": 0.0231, "step": 6630 }, { "grad_norm": 0.5969708561897278, "learning_rate": 9.970279190946788e-05, "loss": 0.022, "step": 6640 }, { "grad_norm": 0.36967214941978455, "learning_rate": 9.970053746759667e-05, "loss": 0.0257, "step": 6650 }, { "grad_norm": 0.29632386565208435, "learning_rate": 9.969827453326292e-05, "loss": 0.0285, "step": 6660 }, { "grad_norm": 0.30132317543029785, "learning_rate": 9.969600310685332e-05, "loss": 0.0206, "step": 6670 }, { "grad_norm": 0.21707379817962646, "learning_rate": 9.969372318875596e-05, "loss": 0.0231, "step": 6680 }, { "grad_norm": 0.2648886442184448, "learning_rate": 9.969143477936043e-05, "loss": 0.03, "step": 6690 }, { "grad_norm": 0.4484325051307678, "learning_rate": 9.968913787905775e-05, "loss": 0.0279, "step": 6700 }, { "grad_norm": 0.425790935754776, "learning_rate": 9.968683248824045e-05, "loss": 0.0233, "step": 6710 }, { "grad_norm": 0.3802105188369751, "learning_rate": 9.968451860730238e-05, "loss": 0.0291, "step": 6720 }, { "grad_norm": 0.2826952040195465, "learning_rate": 9.968219623663896e-05, "loss": 0.0235, "step": 6730 }, { "grad_norm": 0.30529749393463135, "learning_rate": 9.967986537664702e-05, "loss": 0.034, "step": 6740 }, { "grad_norm": 0.21491549909114838, "learning_rate": 9.967752602772483e-05, "loss": 0.0222, "step": 6750 }, { "grad_norm": 0.21811552345752716, "learning_rate": 9.967517819027212e-05, "loss": 0.0262, "step": 6760 }, { "grad_norm": 0.48537692427635193, "learning_rate": 9.967282186469009e-05, "loss": 0.0201, "step": 6770 }, { "grad_norm": 0.3632664382457733, "learning_rate": 9.967045705138135e-05, "loss": 0.0276, "step": 6780 }, { "grad_norm": 0.5571557879447937, "learning_rate": 9.966808375074998e-05, "loss": 0.0259, "step": 6790 }, { "grad_norm": 0.2299354374408722, "learning_rate": 9.966570196320154e-05, "loss": 0.0288, "step": 6800 }, { "grad_norm": 0.29970675706863403, "learning_rate": 9.966331168914299e-05, "loss": 0.0237, "step": 6810 }, { "grad_norm": 0.39356955885887146, "learning_rate": 9.966091292898277e-05, "loss": 0.0284, "step": 6820 }, { "grad_norm": 0.3709717094898224, "learning_rate": 9.965850568313076e-05, "loss": 0.0239, "step": 6830 }, { "grad_norm": 0.3585985004901886, "learning_rate": 9.965608995199827e-05, "loss": 0.0257, "step": 6840 }, { "grad_norm": 0.29174572229385376, "learning_rate": 9.965366573599812e-05, "loss": 0.0263, "step": 6850 }, { "grad_norm": 0.2394421547651291, "learning_rate": 9.965123303554453e-05, "loss": 0.0268, "step": 6860 }, { "grad_norm": 0.41423091292381287, "learning_rate": 9.964879185105317e-05, "loss": 0.0296, "step": 6870 }, { "grad_norm": 0.26926925778388977, "learning_rate": 9.964634218294119e-05, "loss": 0.0283, "step": 6880 }, { "grad_norm": 0.3657136857509613, "learning_rate": 9.964388403162714e-05, "loss": 0.0289, "step": 6890 }, { "grad_norm": 0.30807650089263916, "learning_rate": 9.96414173975311e-05, "loss": 0.0308, "step": 6900 }, { "grad_norm": 0.2792010009288788, "learning_rate": 9.963894228107451e-05, "loss": 0.0242, "step": 6910 }, { "grad_norm": 0.33703121542930603, "learning_rate": 9.963645868268032e-05, "loss": 0.025, "step": 6920 }, { "grad_norm": 0.29493027925491333, "learning_rate": 9.963396660277289e-05, "loss": 0.0299, "step": 6930 }, { "grad_norm": 0.40612319111824036, "learning_rate": 9.963146604177807e-05, "loss": 0.024, "step": 6940 }, { "grad_norm": 0.2966022491455078, "learning_rate": 9.962895700012311e-05, "loss": 0.0249, "step": 6950 }, { "grad_norm": 0.3560218811035156, "learning_rate": 9.962643947823677e-05, "loss": 0.0279, "step": 6960 }, { "grad_norm": 0.3578282594680786, "learning_rate": 9.962391347654921e-05, "loss": 0.0304, "step": 6970 }, { "grad_norm": 0.2897525131702423, "learning_rate": 9.962137899549204e-05, "loss": 0.0376, "step": 6980 }, { "grad_norm": 0.350568950176239, "learning_rate": 9.961883603549835e-05, "loss": 0.0234, "step": 6990 }, { "grad_norm": 0.522214949131012, "learning_rate": 9.961628459700267e-05, "loss": 0.0234, "step": 7000 }, { "grad_norm": 0.30630114674568176, "learning_rate": 9.961372468044095e-05, "loss": 0.0319, "step": 7010 }, { "grad_norm": 0.30590441823005676, "learning_rate": 9.961115628625062e-05, "loss": 0.026, "step": 7020 }, { "grad_norm": 0.377021849155426, "learning_rate": 9.960857941487056e-05, "loss": 0.0328, "step": 7030 }, { "grad_norm": 0.28158360719680786, "learning_rate": 9.960599406674106e-05, "loss": 0.0273, "step": 7040 }, { "grad_norm": 0.2819320559501648, "learning_rate": 9.960340024230393e-05, "loss": 0.0235, "step": 7050 }, { "grad_norm": 0.27649402618408203, "learning_rate": 9.960079794200232e-05, "loss": 0.0244, "step": 7060 }, { "grad_norm": 0.37171879410743713, "learning_rate": 9.959818716628096e-05, "loss": 0.0289, "step": 7070 }, { "grad_norm": 0.22903884947299957, "learning_rate": 9.95955679155859e-05, "loss": 0.0375, "step": 7080 }, { "grad_norm": 0.4475005865097046, "learning_rate": 9.959294019036472e-05, "loss": 0.0347, "step": 7090 }, { "grad_norm": 0.4079831838607788, "learning_rate": 9.959030399106646e-05, "loss": 0.0311, "step": 7100 }, { "grad_norm": 0.2385255992412567, "learning_rate": 9.958765931814153e-05, "loss": 0.0307, "step": 7110 }, { "grad_norm": 0.32138165831565857, "learning_rate": 9.958500617204184e-05, "loss": 0.0191, "step": 7120 }, { "grad_norm": 0.27935197949409485, "learning_rate": 9.958234455322075e-05, "loss": 0.0248, "step": 7130 }, { "grad_norm": 0.31096529960632324, "learning_rate": 9.957967446213308e-05, "loss": 0.0353, "step": 7140 }, { "grad_norm": 0.25014302134513855, "learning_rate": 9.957699589923501e-05, "loss": 0.0301, "step": 7150 }, { "grad_norm": 0.41219428181648254, "learning_rate": 9.957430886498431e-05, "loss": 0.0251, "step": 7160 }, { "grad_norm": 0.2608598470687866, "learning_rate": 9.957161335984008e-05, "loss": 0.0273, "step": 7170 }, { "grad_norm": 0.6962363719940186, "learning_rate": 9.956890938426291e-05, "loss": 0.0219, "step": 7180 }, { "grad_norm": 0.2050987333059311, "learning_rate": 9.956619693871482e-05, "loss": 0.0285, "step": 7190 }, { "grad_norm": 0.38084107637405396, "learning_rate": 9.956347602365934e-05, "loss": 0.0252, "step": 7200 }, { "grad_norm": 0.28264617919921875, "learning_rate": 9.956074663956135e-05, "loss": 0.0241, "step": 7210 }, { "grad_norm": 0.2922685146331787, "learning_rate": 9.955800878688726e-05, "loss": 0.0227, "step": 7220 }, { "grad_norm": 0.32018452882766724, "learning_rate": 9.955526246610489e-05, "loss": 0.0268, "step": 7230 }, { "grad_norm": 0.2911505103111267, "learning_rate": 9.955250767768349e-05, "loss": 0.0276, "step": 7240 }, { "grad_norm": 0.24504217505455017, "learning_rate": 9.95497444220938e-05, "loss": 0.0239, "step": 7250 }, { "grad_norm": 0.23866236209869385, "learning_rate": 9.954697269980797e-05, "loss": 0.0264, "step": 7260 }, { "grad_norm": 0.5427371263504028, "learning_rate": 9.954419251129962e-05, "loss": 0.0253, "step": 7270 }, { "grad_norm": 0.2264024019241333, "learning_rate": 9.95414038570438e-05, "loss": 0.0241, "step": 7280 }, { "grad_norm": 0.3028821647167206, "learning_rate": 9.953860673751703e-05, "loss": 0.0243, "step": 7290 }, { "grad_norm": 0.416890412569046, "learning_rate": 9.953580115319725e-05, "loss": 0.0239, "step": 7300 }, { "grad_norm": 0.21750475466251373, "learning_rate": 9.953298710456387e-05, "loss": 0.023, "step": 7310 }, { "grad_norm": 0.32348260283470154, "learning_rate": 9.953016459209771e-05, "loss": 0.0277, "step": 7320 }, { "grad_norm": 0.27046510577201843, "learning_rate": 9.952733361628108e-05, "loss": 0.0189, "step": 7330 }, { "grad_norm": 0.3077075779438019, "learning_rate": 9.952449417759772e-05, "loss": 0.0285, "step": 7340 }, { "grad_norm": 0.333192378282547, "learning_rate": 9.952164627653279e-05, "loss": 0.0267, "step": 7350 }, { "grad_norm": 0.28885963559150696, "learning_rate": 9.951878991357292e-05, "loss": 0.0258, "step": 7360 }, { "grad_norm": 0.3338298201560974, "learning_rate": 9.951592508920622e-05, "loss": 0.0284, "step": 7370 }, { "grad_norm": 0.24942368268966675, "learning_rate": 9.951305180392219e-05, "loss": 0.0238, "step": 7380 }, { "grad_norm": 0.2500746548175812, "learning_rate": 9.951017005821178e-05, "loss": 0.0252, "step": 7390 }, { "grad_norm": 0.2577601671218872, "learning_rate": 9.95072798525674e-05, "loss": 0.022, "step": 7400 }, { "grad_norm": 0.3308383822441101, "learning_rate": 9.950438118748293e-05, "loss": 0.0284, "step": 7410 }, { "grad_norm": 0.3934571146965027, "learning_rate": 9.950147406345366e-05, "loss": 0.033, "step": 7420 }, { "grad_norm": 0.4905903935432434, "learning_rate": 9.949855848097635e-05, "loss": 0.0319, "step": 7430 }, { "grad_norm": 0.37276390194892883, "learning_rate": 9.949563444054916e-05, "loss": 0.027, "step": 7440 }, { "grad_norm": 0.26602238416671753, "learning_rate": 9.949270194267178e-05, "loss": 0.0249, "step": 7450 }, { "grad_norm": 0.2987116575241089, "learning_rate": 9.948976098784526e-05, "loss": 0.0249, "step": 7460 }, { "grad_norm": 0.25136202573776245, "learning_rate": 9.948681157657213e-05, "loss": 0.0274, "step": 7470 }, { "grad_norm": 0.3980383574962616, "learning_rate": 9.948385370935638e-05, "loss": 0.0349, "step": 7480 }, { "grad_norm": 0.3260123133659363, "learning_rate": 9.94808873867034e-05, "loss": 0.0276, "step": 7490 }, { "grad_norm": 0.362826406955719, "learning_rate": 9.947791260912009e-05, "loss": 0.0211, "step": 7500 }, { "grad_norm": 0.32152876257896423, "learning_rate": 9.947492937711474e-05, "loss": 0.0221, "step": 7510 }, { "grad_norm": 0.2261449247598648, "learning_rate": 9.947193769119707e-05, "loss": 0.0272, "step": 7520 }, { "grad_norm": 0.2306436151266098, "learning_rate": 9.946893755187834e-05, "loss": 0.0234, "step": 7530 }, { "grad_norm": 0.19335104525089264, "learning_rate": 9.946592895967115e-05, "loss": 0.0236, "step": 7540 }, { "grad_norm": 0.4490503668785095, "learning_rate": 9.94629119150896e-05, "loss": 0.0218, "step": 7550 }, { "grad_norm": 0.2082783728837967, "learning_rate": 9.94598864186492e-05, "loss": 0.0237, "step": 7560 }, { "grad_norm": 0.2634128928184509, "learning_rate": 9.945685247086696e-05, "loss": 0.0291, "step": 7570 }, { "grad_norm": 0.37327149510383606, "learning_rate": 9.945381007226129e-05, "loss": 0.0293, "step": 7580 }, { "grad_norm": 0.2535640001296997, "learning_rate": 9.945075922335203e-05, "loss": 0.0243, "step": 7590 }, { "grad_norm": 0.3020217716693878, "learning_rate": 9.944769992466049e-05, "loss": 0.0327, "step": 7600 }, { "grad_norm": 0.41731390357017517, "learning_rate": 9.944463217670945e-05, "loss": 0.029, "step": 7610 }, { "grad_norm": 0.2384582906961441, "learning_rate": 9.944155598002307e-05, "loss": 0.0282, "step": 7620 }, { "grad_norm": 0.21733491122722626, "learning_rate": 9.943847133512701e-05, "loss": 0.023, "step": 7630 }, { "grad_norm": 0.3847140669822693, "learning_rate": 9.943537824254834e-05, "loss": 0.0254, "step": 7640 }, { "grad_norm": 0.3055148124694824, "learning_rate": 9.943227670281559e-05, "loss": 0.0228, "step": 7650 }, { "grad_norm": 0.3465765118598938, "learning_rate": 9.942916671645873e-05, "loss": 0.0209, "step": 7660 }, { "grad_norm": 0.34906119108200073, "learning_rate": 9.942604828400916e-05, "loss": 0.0318, "step": 7670 }, { "grad_norm": 0.3057197332382202, "learning_rate": 9.942292140599975e-05, "loss": 0.0287, "step": 7680 }, { "grad_norm": 0.3486880660057068, "learning_rate": 9.94197860829648e-05, "loss": 0.0276, "step": 7690 }, { "grad_norm": 0.24566268920898438, "learning_rate": 9.941664231544004e-05, "loss": 0.0182, "step": 7700 }, { "grad_norm": 0.3375215232372284, "learning_rate": 9.941349010396264e-05, "loss": 0.0336, "step": 7710 }, { "grad_norm": 0.528278112411499, "learning_rate": 9.941032944907125e-05, "loss": 0.0271, "step": 7720 }, { "grad_norm": 0.24615323543548584, "learning_rate": 9.940716035130596e-05, "loss": 0.0244, "step": 7730 }, { "grad_norm": 0.31370046734809875, "learning_rate": 9.940398281120821e-05, "loss": 0.0283, "step": 7740 }, { "grad_norm": 0.21236950159072876, "learning_rate": 9.940079682932102e-05, "loss": 0.0229, "step": 7750 }, { "grad_norm": 0.27626028656959534, "learning_rate": 9.939760240618877e-05, "loss": 0.0251, "step": 7760 }, { "grad_norm": 0.44044482707977295, "learning_rate": 9.939439954235729e-05, "loss": 0.0314, "step": 7770 }, { "grad_norm": 0.396626353263855, "learning_rate": 9.939118823837387e-05, "loss": 0.0309, "step": 7780 }, { "grad_norm": 0.40265169739723206, "learning_rate": 9.938796849478725e-05, "loss": 0.0251, "step": 7790 }, { "grad_norm": 0.36130303144454956, "learning_rate": 9.938474031214755e-05, "loss": 0.0252, "step": 7800 }, { "grad_norm": 0.32522767782211304, "learning_rate": 9.938150369100643e-05, "loss": 0.0278, "step": 7810 }, { "grad_norm": 0.49650391936302185, "learning_rate": 9.93782586319169e-05, "loss": 0.0172, "step": 7820 }, { "grad_norm": 0.4088502526283264, "learning_rate": 9.937500513543348e-05, "loss": 0.0287, "step": 7830 }, { "grad_norm": 0.34122195839881897, "learning_rate": 9.937174320211207e-05, "loss": 0.0245, "step": 7840 }, { "grad_norm": 0.28373780846595764, "learning_rate": 9.936847283251009e-05, "loss": 0.025, "step": 7850 }, { "grad_norm": 0.3715327978134155, "learning_rate": 9.936519402718632e-05, "loss": 0.0266, "step": 7860 }, { "grad_norm": 0.5219953060150146, "learning_rate": 9.936190678670102e-05, "loss": 0.0275, "step": 7870 }, { "grad_norm": 0.3253597319126129, "learning_rate": 9.935861111161593e-05, "loss": 0.0276, "step": 7880 }, { "grad_norm": 0.22741805016994476, "learning_rate": 9.935530700249416e-05, "loss": 0.0265, "step": 7890 }, { "grad_norm": 0.28555038571357727, "learning_rate": 9.935199445990028e-05, "loss": 0.0357, "step": 7900 }, { "grad_norm": 0.23432308435440063, "learning_rate": 9.934867348440033e-05, "loss": 0.0302, "step": 7910 }, { "grad_norm": 0.3919669985771179, "learning_rate": 9.934534407656176e-05, "loss": 0.037, "step": 7920 }, { "grad_norm": 0.37751027941703796, "learning_rate": 9.93420062369535e-05, "loss": 0.0274, "step": 7930 }, { "grad_norm": 0.30618107318878174, "learning_rate": 9.933865996614589e-05, "loss": 0.027, "step": 7940 }, { "grad_norm": 0.3086288869380951, "learning_rate": 9.933530526471068e-05, "loss": 0.0218, "step": 7950 }, { "grad_norm": 0.20336692035198212, "learning_rate": 9.933194213322114e-05, "loss": 0.0197, "step": 7960 }, { "grad_norm": 0.2657483220100403, "learning_rate": 9.932857057225192e-05, "loss": 0.0225, "step": 7970 }, { "grad_norm": 0.25749731063842773, "learning_rate": 9.932519058237912e-05, "loss": 0.0195, "step": 7980 }, { "grad_norm": 0.2366318702697754, "learning_rate": 9.932180216418032e-05, "loss": 0.0215, "step": 7990 }, { "grad_norm": 0.3456028699874878, "learning_rate": 9.931840531823446e-05, "loss": 0.0224, "step": 8000 }, { "grad_norm": 0.33359551429748535, "learning_rate": 9.9315000045122e-05, "loss": 0.0173, "step": 8010 }, { "grad_norm": 0.3928941488265991, "learning_rate": 9.931158634542481e-05, "loss": 0.026, "step": 8020 }, { "grad_norm": 0.2803351879119873, "learning_rate": 9.930816421972617e-05, "loss": 0.0249, "step": 8030 }, { "grad_norm": 0.3534027338027954, "learning_rate": 9.930473366861086e-05, "loss": 0.021, "step": 8040 }, { "grad_norm": 0.4378472864627838, "learning_rate": 9.930129469266505e-05, "loss": 0.0261, "step": 8050 }, { "grad_norm": 0.29399871826171875, "learning_rate": 9.929784729247638e-05, "loss": 0.0228, "step": 8060 }, { "grad_norm": 0.24281194806098938, "learning_rate": 9.929439146863389e-05, "loss": 0.0208, "step": 8070 }, { "grad_norm": 0.44008201360702515, "learning_rate": 9.92909272217281e-05, "loss": 0.0358, "step": 8080 }, { "grad_norm": 0.35107260942459106, "learning_rate": 9.928745455235097e-05, "loss": 0.0277, "step": 8090 }, { "grad_norm": 0.3176348805427551, "learning_rate": 9.928397346109588e-05, "loss": 0.0269, "step": 8100 }, { "grad_norm": 0.33402740955352783, "learning_rate": 9.928048394855762e-05, "loss": 0.0314, "step": 8110 }, { "grad_norm": 0.25868427753448486, "learning_rate": 9.92769860153325e-05, "loss": 0.0286, "step": 8120 }, { "grad_norm": 0.4206448495388031, "learning_rate": 9.927347966201819e-05, "loss": 0.0361, "step": 8130 }, { "grad_norm": 0.2597289979457855, "learning_rate": 9.926996488921383e-05, "loss": 0.0288, "step": 8140 }, { "grad_norm": 0.29048389196395874, "learning_rate": 9.926644169752001e-05, "loss": 0.0227, "step": 8150 }, { "grad_norm": 0.3533458709716797, "learning_rate": 9.926291008753875e-05, "loss": 0.0277, "step": 8160 }, { "grad_norm": 0.2918403744697571, "learning_rate": 9.92593700598735e-05, "loss": 0.0261, "step": 8170 }, { "grad_norm": 0.2761167585849762, "learning_rate": 9.925582161512915e-05, "loss": 0.0298, "step": 8180 }, { "grad_norm": 0.3152838349342346, "learning_rate": 9.925226475391205e-05, "loss": 0.0253, "step": 8190 }, { "grad_norm": 0.4226949214935303, "learning_rate": 9.924869947682993e-05, "loss": 0.0263, "step": 8200 }, { "grad_norm": 0.3569718599319458, "learning_rate": 9.924512578449204e-05, "loss": 0.0199, "step": 8210 }, { "grad_norm": 0.408809095621109, "learning_rate": 9.924154367750901e-05, "loss": 0.0252, "step": 8220 }, { "grad_norm": 0.2778615951538086, "learning_rate": 9.923795315649293e-05, "loss": 0.0271, "step": 8230 }, { "grad_norm": 0.25509271025657654, "learning_rate": 9.92343542220573e-05, "loss": 0.0255, "step": 8240 }, { "grad_norm": 0.6777804493904114, "learning_rate": 9.92307468748171e-05, "loss": 0.0284, "step": 8250 }, { "grad_norm": 0.3354063332080841, "learning_rate": 9.922713111538873e-05, "loss": 0.022, "step": 8260 }, { "grad_norm": 0.29923808574676514, "learning_rate": 9.922350694439003e-05, "loss": 0.0218, "step": 8270 }, { "grad_norm": 0.5179187059402466, "learning_rate": 9.921987436244024e-05, "loss": 0.0328, "step": 8280 }, { "grad_norm": 0.3448713421821594, "learning_rate": 9.921623337016008e-05, "loss": 0.0199, "step": 8290 }, { "grad_norm": 0.2502465844154358, "learning_rate": 9.921258396817172e-05, "loss": 0.0234, "step": 8300 }, { "grad_norm": 0.3079729974269867, "learning_rate": 9.920892615709874e-05, "loss": 0.0256, "step": 8310 }, { "grad_norm": 0.2946290671825409, "learning_rate": 9.920525993756612e-05, "loss": 0.0188, "step": 8320 }, { "grad_norm": 0.7406721711158752, "learning_rate": 9.920158531020036e-05, "loss": 0.0202, "step": 8330 }, { "grad_norm": 0.6367990374565125, "learning_rate": 9.919790227562933e-05, "loss": 0.0199, "step": 8340 }, { "grad_norm": 0.363202303647995, "learning_rate": 9.919421083448237e-05, "loss": 0.0322, "step": 8350 }, { "grad_norm": 0.5022500157356262, "learning_rate": 9.919051098739022e-05, "loss": 0.0228, "step": 8360 }, { "grad_norm": 0.40896108746528625, "learning_rate": 9.918680273498514e-05, "loss": 0.0216, "step": 8370 }, { "grad_norm": 0.29762619733810425, "learning_rate": 9.918308607790072e-05, "loss": 0.0248, "step": 8380 }, { "grad_norm": 0.2876676917076111, "learning_rate": 9.917936101677205e-05, "loss": 0.0233, "step": 8390 }, { "grad_norm": 0.2690202295780182, "learning_rate": 9.917562755223564e-05, "loss": 0.024, "step": 8400 }, { "grad_norm": 0.33210039138793945, "learning_rate": 9.917188568492944e-05, "loss": 0.0264, "step": 8410 }, { "grad_norm": 0.3196198046207428, "learning_rate": 9.916813541549283e-05, "loss": 0.0225, "step": 8420 }, { "grad_norm": 0.3815516531467438, "learning_rate": 9.916437674456663e-05, "loss": 0.0267, "step": 8430 }, { "grad_norm": 0.33480745553970337, "learning_rate": 9.916060967279308e-05, "loss": 0.0245, "step": 8440 }, { "grad_norm": 0.37890952825546265, "learning_rate": 9.91568342008159e-05, "loss": 0.0225, "step": 8450 }, { "grad_norm": 0.3291693925857544, "learning_rate": 9.915305032928019e-05, "loss": 0.0238, "step": 8460 }, { "grad_norm": 0.2486797422170639, "learning_rate": 9.914925805883253e-05, "loss": 0.0204, "step": 8470 }, { "grad_norm": 0.4885823130607605, "learning_rate": 9.914545739012088e-05, "loss": 0.0279, "step": 8480 }, { "grad_norm": 0.9454846978187561, "learning_rate": 9.91416483237947e-05, "loss": 0.0256, "step": 8490 }, { "grad_norm": 0.4337364733219147, "learning_rate": 9.913783086050485e-05, "loss": 0.0253, "step": 8500 }, { "grad_norm": 0.2898540496826172, "learning_rate": 9.913400500090364e-05, "loss": 0.0267, "step": 8510 }, { "grad_norm": 0.30469954013824463, "learning_rate": 9.913017074564479e-05, "loss": 0.0195, "step": 8520 }, { "grad_norm": 0.21991068124771118, "learning_rate": 9.912632809538348e-05, "loss": 0.0282, "step": 8530 }, { "grad_norm": 0.3752175271511078, "learning_rate": 9.912247705077629e-05, "loss": 0.0206, "step": 8540 }, { "grad_norm": 0.2514849305152893, "learning_rate": 9.911861761248127e-05, "loss": 0.024, "step": 8550 }, { "grad_norm": 0.32769885659217834, "learning_rate": 9.91147497811579e-05, "loss": 0.029, "step": 8560 }, { "grad_norm": 0.3373855650424957, "learning_rate": 9.911087355746709e-05, "loss": 0.0429, "step": 8570 }, { "grad_norm": 0.3004923462867737, "learning_rate": 9.910698894207117e-05, "loss": 0.0292, "step": 8580 }, { "grad_norm": 0.3056347668170929, "learning_rate": 9.910309593563392e-05, "loss": 0.0272, "step": 8590 }, { "grad_norm": 0.32792991399765015, "learning_rate": 9.909919453882057e-05, "loss": 0.0194, "step": 8600 }, { "grad_norm": 0.3567076027393341, "learning_rate": 9.90952847522977e-05, "loss": 0.0207, "step": 8610 }, { "grad_norm": 0.4308660924434662, "learning_rate": 9.909136657673346e-05, "loss": 0.029, "step": 8620 }, { "grad_norm": 0.28881916403770447, "learning_rate": 9.908744001279731e-05, "loss": 0.0221, "step": 8630 }, { "grad_norm": 0.3821750581264496, "learning_rate": 9.90835050611602e-05, "loss": 0.0273, "step": 8640 }, { "grad_norm": 0.3821294605731964, "learning_rate": 9.90795617224945e-05, "loss": 0.025, "step": 8650 }, { "grad_norm": 0.3270058035850525, "learning_rate": 9.907560999747405e-05, "loss": 0.0236, "step": 8660 }, { "grad_norm": 0.4309808611869812, "learning_rate": 9.907164988677408e-05, "loss": 0.0211, "step": 8670 }, { "grad_norm": 0.2068018764257431, "learning_rate": 9.906768139107124e-05, "loss": 0.0202, "step": 8680 }, { "grad_norm": 0.3315526843070984, "learning_rate": 9.906370451104367e-05, "loss": 0.0263, "step": 8690 }, { "grad_norm": 0.3090708255767822, "learning_rate": 9.905971924737088e-05, "loss": 0.0313, "step": 8700 }, { "grad_norm": 0.3265189826488495, "learning_rate": 9.905572560073387e-05, "loss": 0.0362, "step": 8710 }, { "grad_norm": 0.5861369371414185, "learning_rate": 9.905172357181501e-05, "loss": 0.038, "step": 8720 }, { "grad_norm": 0.2444690316915512, "learning_rate": 9.904771316129817e-05, "loss": 0.0327, "step": 8730 }, { "grad_norm": 0.2562212646007538, "learning_rate": 9.904369436986862e-05, "loss": 0.0268, "step": 8740 }, { "grad_norm": 0.3362107276916504, "learning_rate": 9.903966719821303e-05, "loss": 0.0322, "step": 8750 }, { "grad_norm": 0.40186935663223267, "learning_rate": 9.903563164701956e-05, "loss": 0.0249, "step": 8760 }, { "grad_norm": 0.28412914276123047, "learning_rate": 9.903158771697778e-05, "loss": 0.0259, "step": 8770 }, { "grad_norm": 0.3006318211555481, "learning_rate": 9.902753540877867e-05, "loss": 0.0326, "step": 8780 }, { "grad_norm": 0.23904986679553986, "learning_rate": 9.902347472311466e-05, "loss": 0.0213, "step": 8790 }, { "grad_norm": 0.2395918071269989, "learning_rate": 9.901940566067962e-05, "loss": 0.0165, "step": 8800 }, { "grad_norm": 0.21185749769210815, "learning_rate": 9.901532822216883e-05, "loss": 0.0183, "step": 8810 }, { "grad_norm": 0.2915710508823395, "learning_rate": 9.901124240827904e-05, "loss": 0.025, "step": 8820 }, { "grad_norm": 0.2826468348503113, "learning_rate": 9.900714821970835e-05, "loss": 0.0208, "step": 8830 }, { "grad_norm": 0.551578938961029, "learning_rate": 9.900304565715641e-05, "loss": 0.0249, "step": 8840 }, { "grad_norm": 0.24776390194892883, "learning_rate": 9.899893472132419e-05, "loss": 0.0223, "step": 8850 }, { "grad_norm": 0.22090394794940948, "learning_rate": 9.899481541291415e-05, "loss": 0.0254, "step": 8860 }, { "grad_norm": 0.390190064907074, "learning_rate": 9.899068773263016e-05, "loss": 0.0285, "step": 8870 }, { "grad_norm": 0.298201322555542, "learning_rate": 9.898655168117754e-05, "loss": 0.0245, "step": 8880 }, { "grad_norm": 0.23551464080810547, "learning_rate": 9.898240725926302e-05, "loss": 0.025, "step": 8890 }, { "grad_norm": 0.22583426535129547, "learning_rate": 9.897825446759478e-05, "loss": 0.0229, "step": 8900 }, { "grad_norm": 0.2120397388935089, "learning_rate": 9.897409330688241e-05, "loss": 0.0226, "step": 8910 }, { "grad_norm": 0.32496869564056396, "learning_rate": 9.896992377783692e-05, "loss": 0.029, "step": 8920 }, { "grad_norm": 0.3818524181842804, "learning_rate": 9.89657458811708e-05, "loss": 0.0195, "step": 8930 }, { "grad_norm": 0.29234910011291504, "learning_rate": 9.896155961759792e-05, "loss": 0.0235, "step": 8940 }, { "grad_norm": 0.24733224511146545, "learning_rate": 9.895736498783361e-05, "loss": 0.0269, "step": 8950 }, { "grad_norm": 0.2297264039516449, "learning_rate": 9.895316199259462e-05, "loss": 0.0204, "step": 8960 }, { "grad_norm": 0.3085598647594452, "learning_rate": 9.894895063259909e-05, "loss": 0.0178, "step": 8970 }, { "grad_norm": 0.2831011712551117, "learning_rate": 9.894473090856667e-05, "loss": 0.0226, "step": 8980 }, { "grad_norm": 0.3279501497745514, "learning_rate": 9.894050282121839e-05, "loss": 0.0371, "step": 8990 }, { "grad_norm": 0.36309295892715454, "learning_rate": 9.893626637127668e-05, "loss": 0.0268, "step": 9000 }, { "grad_norm": 0.4440087378025055, "learning_rate": 9.893202155946546e-05, "loss": 0.0213, "step": 9010 }, { "grad_norm": 0.34444093704223633, "learning_rate": 9.892776838651006e-05, "loss": 0.0246, "step": 9020 }, { "grad_norm": 0.33478111028671265, "learning_rate": 9.892350685313722e-05, "loss": 0.0301, "step": 9030 }, { "grad_norm": 0.35282471776008606, "learning_rate": 9.891923696007513e-05, "loss": 0.0327, "step": 9040 }, { "grad_norm": 0.22571660578250885, "learning_rate": 9.891495870805336e-05, "loss": 0.0208, "step": 9050 }, { "grad_norm": 0.2798762321472168, "learning_rate": 9.891067209780298e-05, "loss": 0.0351, "step": 9060 }, { "grad_norm": 0.3619536757469177, "learning_rate": 9.890637713005646e-05, "loss": 0.0247, "step": 9070 }, { "grad_norm": 0.42670607566833496, "learning_rate": 9.890207380554767e-05, "loss": 0.0305, "step": 9080 }, { "grad_norm": 0.2911798655986786, "learning_rate": 9.889776212501196e-05, "loss": 0.0324, "step": 9090 }, { "grad_norm": 0.31699302792549133, "learning_rate": 9.889344208918605e-05, "loss": 0.0337, "step": 9100 }, { "grad_norm": 0.43698394298553467, "learning_rate": 9.888911369880812e-05, "loss": 0.0312, "step": 9110 }, { "grad_norm": 0.3104386627674103, "learning_rate": 9.888477695461777e-05, "loss": 0.0229, "step": 9120 }, { "grad_norm": 0.31708335876464844, "learning_rate": 9.888043185735607e-05, "loss": 0.0325, "step": 9130 }, { "grad_norm": 0.3249432444572449, "learning_rate": 9.887607840776542e-05, "loss": 0.0232, "step": 9140 }, { "grad_norm": 0.34056198596954346, "learning_rate": 9.887171660658975e-05, "loss": 0.0289, "step": 9150 }, { "grad_norm": 0.20617790520191193, "learning_rate": 9.886734645457435e-05, "loss": 0.0174, "step": 9160 }, { "grad_norm": 0.3023681640625, "learning_rate": 9.886296795246597e-05, "loss": 0.0244, "step": 9170 }, { "grad_norm": 0.3019917905330658, "learning_rate": 9.885858110101276e-05, "loss": 0.0211, "step": 9180 }, { "grad_norm": 0.24671971797943115, "learning_rate": 9.885418590096434e-05, "loss": 0.0219, "step": 9190 }, { "grad_norm": 0.2386152297258377, "learning_rate": 9.88497823530717e-05, "loss": 0.0202, "step": 9200 }, { "grad_norm": 0.3665713667869568, "learning_rate": 9.884537045808732e-05, "loss": 0.0358, "step": 9210 }, { "grad_norm": 0.39650169014930725, "learning_rate": 9.884095021676502e-05, "loss": 0.0345, "step": 9220 }, { "grad_norm": 0.42952582240104675, "learning_rate": 9.883652162986017e-05, "loss": 0.0306, "step": 9230 }, { "grad_norm": 0.3498530983924866, "learning_rate": 9.883208469812943e-05, "loss": 0.0395, "step": 9240 }, { "grad_norm": 0.34247907996177673, "learning_rate": 9.882763942233098e-05, "loss": 0.0244, "step": 9250 }, { "grad_norm": 0.25043928623199463, "learning_rate": 9.882318580322441e-05, "loss": 0.0219, "step": 9260 }, { "grad_norm": 0.37990686297416687, "learning_rate": 9.881872384157067e-05, "loss": 0.0242, "step": 9270 }, { "grad_norm": 0.2776334583759308, "learning_rate": 9.881425353813225e-05, "loss": 0.0228, "step": 9280 }, { "grad_norm": 0.43198713660240173, "learning_rate": 9.880977489367296e-05, "loss": 0.0259, "step": 9290 }, { "grad_norm": 0.20615756511688232, "learning_rate": 9.88052879089581e-05, "loss": 0.0314, "step": 9300 }, { "grad_norm": 0.312440425157547, "learning_rate": 9.880079258475434e-05, "loss": 0.0183, "step": 9310 }, { "grad_norm": 0.33391454815864563, "learning_rate": 9.879628892182985e-05, "loss": 0.0396, "step": 9320 }, { "grad_norm": 0.4110356867313385, "learning_rate": 9.879177692095416e-05, "loss": 0.0272, "step": 9330 }, { "grad_norm": 0.32569408416748047, "learning_rate": 9.878725658289825e-05, "loss": 0.0302, "step": 9340 }, { "grad_norm": 0.3590027987957001, "learning_rate": 9.878272790843454e-05, "loss": 0.0231, "step": 9350 }, { "grad_norm": 0.257819265127182, "learning_rate": 9.877819089833682e-05, "loss": 0.0225, "step": 9360 }, { "grad_norm": 0.23439010977745056, "learning_rate": 9.877364555338038e-05, "loss": 0.0393, "step": 9370 }, { "grad_norm": 0.24776573479175568, "learning_rate": 9.876909187434186e-05, "loss": 0.0261, "step": 9380 }, { "grad_norm": 0.3794482946395874, "learning_rate": 9.876452986199939e-05, "loss": 0.0255, "step": 9390 }, { "grad_norm": 0.25806599855422974, "learning_rate": 9.875995951713248e-05, "loss": 0.0319, "step": 9400 }, { "grad_norm": 0.3576003611087799, "learning_rate": 9.875538084052207e-05, "loss": 0.027, "step": 9410 }, { "grad_norm": 0.3603156507015228, "learning_rate": 9.875079383295053e-05, "loss": 0.03, "step": 9420 }, { "grad_norm": 0.2053857445716858, "learning_rate": 9.874619849520167e-05, "loss": 0.0203, "step": 9430 }, { "grad_norm": 0.4625720679759979, "learning_rate": 9.874159482806069e-05, "loss": 0.0263, "step": 9440 }, { "grad_norm": 0.318570613861084, "learning_rate": 9.873698283231426e-05, "loss": 0.0245, "step": 9450 }, { "grad_norm": 0.2749320864677429, "learning_rate": 9.87323625087504e-05, "loss": 0.0218, "step": 9460 }, { "grad_norm": 0.32356899976730347, "learning_rate": 9.872773385815863e-05, "loss": 0.0303, "step": 9470 }, { "grad_norm": 0.36036744713783264, "learning_rate": 9.872309688132986e-05, "loss": 0.0227, "step": 9480 }, { "grad_norm": 0.28682583570480347, "learning_rate": 9.871845157905639e-05, "loss": 0.0233, "step": 9490 }, { "grad_norm": 0.297262579202652, "learning_rate": 9.871379795213201e-05, "loss": 0.0182, "step": 9500 }, { "grad_norm": 0.29007887840270996, "learning_rate": 9.87091360013519e-05, "loss": 0.0286, "step": 9510 }, { "grad_norm": 0.22966386377811432, "learning_rate": 9.870446572751262e-05, "loss": 0.0279, "step": 9520 }, { "grad_norm": 0.20912982523441315, "learning_rate": 9.869978713141224e-05, "loss": 0.0262, "step": 9530 }, { "grad_norm": 0.39959776401519775, "learning_rate": 9.869510021385016e-05, "loss": 0.0269, "step": 9540 }, { "grad_norm": 0.1905774474143982, "learning_rate": 9.869040497562727e-05, "loss": 0.0236, "step": 9550 }, { "grad_norm": 0.27303868532180786, "learning_rate": 9.868570141754587e-05, "loss": 0.0291, "step": 9560 }, { "grad_norm": 0.3384583294391632, "learning_rate": 9.868098954040965e-05, "loss": 0.0332, "step": 9570 }, { "grad_norm": 0.25526735186576843, "learning_rate": 9.867626934502374e-05, "loss": 0.0244, "step": 9580 }, { "grad_norm": 0.3814257085323334, "learning_rate": 9.86715408321947e-05, "loss": 0.0282, "step": 9590 }, { "grad_norm": 0.5022850632667542, "learning_rate": 9.86668040027305e-05, "loss": 0.0291, "step": 9600 }, { "grad_norm": 0.3048495352268219, "learning_rate": 9.866205885744053e-05, "loss": 0.0216, "step": 9610 }, { "grad_norm": 0.30851733684539795, "learning_rate": 9.865730539713563e-05, "loss": 0.0296, "step": 9620 }, { "grad_norm": 0.3310665786266327, "learning_rate": 9.8652543622628e-05, "loss": 0.0238, "step": 9630 }, { "grad_norm": 0.2556701898574829, "learning_rate": 9.864777353473132e-05, "loss": 0.0259, "step": 9640 }, { "grad_norm": 0.8833345770835876, "learning_rate": 9.864299513426068e-05, "loss": 0.0258, "step": 9650 }, { "grad_norm": 0.47137656807899475, "learning_rate": 9.863820842203254e-05, "loss": 0.0259, "step": 9660 }, { "grad_norm": 0.4027319550514221, "learning_rate": 9.863341339886483e-05, "loss": 0.0326, "step": 9670 }, { "grad_norm": 0.27978312969207764, "learning_rate": 9.86286100655769e-05, "loss": 0.0239, "step": 9680 }, { "grad_norm": 0.3040481507778168, "learning_rate": 9.862379842298953e-05, "loss": 0.0265, "step": 9690 }, { "grad_norm": 0.2726801931858063, "learning_rate": 9.861897847192485e-05, "loss": 0.0184, "step": 9700 }, { "grad_norm": 0.2690298855304718, "learning_rate": 9.86141502132065e-05, "loss": 0.0262, "step": 9710 }, { "grad_norm": 0.35205188393592834, "learning_rate": 9.860931364765946e-05, "loss": 0.0228, "step": 9720 }, { "grad_norm": 0.3364746570587158, "learning_rate": 9.860446877611021e-05, "loss": 0.029, "step": 9730 }, { "grad_norm": 0.32678163051605225, "learning_rate": 9.859961559938655e-05, "loss": 0.0262, "step": 9740 }, { "grad_norm": 0.2728565037250519, "learning_rate": 9.85947541183178e-05, "loss": 0.0222, "step": 9750 }, { "grad_norm": 0.40571480989456177, "learning_rate": 9.858988433373463e-05, "loss": 0.0258, "step": 9760 }, { "grad_norm": 0.3148847818374634, "learning_rate": 9.858500624646918e-05, "loss": 0.0219, "step": 9770 }, { "grad_norm": 0.3375985026359558, "learning_rate": 9.858011985735497e-05, "loss": 0.0268, "step": 9780 }, { "grad_norm": 0.34183502197265625, "learning_rate": 9.857522516722693e-05, "loss": 0.0253, "step": 9790 }, { "grad_norm": 0.3640196919441223, "learning_rate": 9.857032217692145e-05, "loss": 0.0234, "step": 9800 }, { "grad_norm": 0.29364579916000366, "learning_rate": 9.856541088727631e-05, "loss": 0.0284, "step": 9810 }, { "grad_norm": 0.6245757341384888, "learning_rate": 9.856049129913072e-05, "loss": 0.0309, "step": 9820 }, { "grad_norm": 0.21725523471832275, "learning_rate": 9.85555634133253e-05, "loss": 0.0303, "step": 9830 }, { "grad_norm": 0.23234492540359497, "learning_rate": 9.855062723070208e-05, "loss": 0.028, "step": 9840 }, { "grad_norm": 0.2365596890449524, "learning_rate": 9.854568275210454e-05, "loss": 0.0277, "step": 9850 }, { "grad_norm": 0.30411621928215027, "learning_rate": 9.854072997837754e-05, "loss": 0.0206, "step": 9860 }, { "grad_norm": 0.25620338320732117, "learning_rate": 9.853576891036737e-05, "loss": 0.0226, "step": 9870 }, { "grad_norm": 0.40247172117233276, "learning_rate": 9.853079954892177e-05, "loss": 0.0268, "step": 9880 }, { "grad_norm": 0.22807855904102325, "learning_rate": 9.852582189488983e-05, "loss": 0.0225, "step": 9890 }, { "grad_norm": 0.45759302377700806, "learning_rate": 9.852083594912212e-05, "loss": 0.0273, "step": 9900 }, { "grad_norm": 0.2975000739097595, "learning_rate": 9.851584171247058e-05, "loss": 0.0265, "step": 9910 }, { "grad_norm": 0.26116400957107544, "learning_rate": 9.851083918578863e-05, "loss": 0.0335, "step": 9920 }, { "grad_norm": 0.47403162717819214, "learning_rate": 9.850582836993103e-05, "loss": 0.0255, "step": 9930 }, { "grad_norm": 0.3095027208328247, "learning_rate": 9.850080926575397e-05, "loss": 0.0277, "step": 9940 }, { "grad_norm": 0.4436231851577759, "learning_rate": 9.849578187411515e-05, "loss": 0.0333, "step": 9950 }, { "grad_norm": 0.22959382832050323, "learning_rate": 9.849074619587354e-05, "loss": 0.0208, "step": 9960 }, { "grad_norm": 0.25826510787010193, "learning_rate": 9.848570223188964e-05, "loss": 0.0289, "step": 9970 }, { "grad_norm": 0.32063770294189453, "learning_rate": 9.848064998302531e-05, "loss": 0.0245, "step": 9980 }, { "grad_norm": 0.2805885672569275, "learning_rate": 9.847558945014386e-05, "loss": 0.0262, "step": 9990 }, { "grad_norm": 0.3955390155315399, "learning_rate": 9.847052063410996e-05, "loss": 0.0307, "step": 10000 }, { "grad_norm": 0.33458781242370605, "learning_rate": 9.846544353578977e-05, "loss": 0.0258, "step": 10010 }, { "grad_norm": 0.29087093472480774, "learning_rate": 9.846035815605081e-05, "loss": 0.0256, "step": 10020 }, { "grad_norm": 0.33825501799583435, "learning_rate": 9.845526449576204e-05, "loss": 0.0223, "step": 10030 }, { "grad_norm": 0.3577125072479248, "learning_rate": 9.845016255579383e-05, "loss": 0.0262, "step": 10040 }, { "grad_norm": 0.3374944031238556, "learning_rate": 9.844505233701794e-05, "loss": 0.0255, "step": 10050 }, { "grad_norm": 0.33784183859825134, "learning_rate": 9.843993384030757e-05, "loss": 0.0277, "step": 10060 }, { "grad_norm": 0.26742011308670044, "learning_rate": 9.843480706653737e-05, "loss": 0.0173, "step": 10070 }, { "grad_norm": 0.290140300989151, "learning_rate": 9.84296720165833e-05, "loss": 0.0253, "step": 10080 }, { "grad_norm": 0.3599807322025299, "learning_rate": 9.842452869132286e-05, "loss": 0.024, "step": 10090 }, { "grad_norm": 0.2977087199687958, "learning_rate": 9.841937709163489e-05, "loss": 0.0276, "step": 10100 }, { "grad_norm": 0.25296634435653687, "learning_rate": 9.841421721839962e-05, "loss": 0.019, "step": 10110 }, { "grad_norm": 0.2170797735452652, "learning_rate": 9.840904907249879e-05, "loss": 0.0188, "step": 10120 }, { "grad_norm": 0.4580096900463104, "learning_rate": 9.840387265481545e-05, "loss": 0.0304, "step": 10130 }, { "grad_norm": 0.31924667954444885, "learning_rate": 9.839868796623411e-05, "loss": 0.0216, "step": 10140 }, { "grad_norm": 0.30561354756355286, "learning_rate": 9.839349500764072e-05, "loss": 0.0341, "step": 10150 }, { "grad_norm": 0.32710930705070496, "learning_rate": 9.83882937799226e-05, "loss": 0.0275, "step": 10160 }, { "grad_norm": 0.2960193157196045, "learning_rate": 9.838308428396849e-05, "loss": 0.0248, "step": 10170 }, { "grad_norm": 0.3361365497112274, "learning_rate": 9.837786652066854e-05, "loss": 0.0306, "step": 10180 }, { "grad_norm": 0.3650586009025574, "learning_rate": 9.837264049091437e-05, "loss": 0.032, "step": 10190 }, { "grad_norm": 0.4456435441970825, "learning_rate": 9.836740619559893e-05, "loss": 0.0275, "step": 10200 }, { "grad_norm": 0.22082625329494476, "learning_rate": 9.836216363561659e-05, "loss": 0.0233, "step": 10210 }, { "grad_norm": 0.24274390935897827, "learning_rate": 9.835691281186322e-05, "loss": 0.0245, "step": 10220 }, { "grad_norm": 0.2784874737262726, "learning_rate": 9.8351653725236e-05, "loss": 0.0291, "step": 10230 }, { "grad_norm": 0.26512014865875244, "learning_rate": 9.83463863766336e-05, "loss": 0.0226, "step": 10240 }, { "grad_norm": 0.19383904337882996, "learning_rate": 9.834111076695602e-05, "loss": 0.0197, "step": 10250 }, { "grad_norm": 0.2634434998035431, "learning_rate": 9.833582689710477e-05, "loss": 0.019, "step": 10260 }, { "grad_norm": 0.42971929907798767, "learning_rate": 9.833053476798268e-05, "loss": 0.0256, "step": 10270 }, { "grad_norm": 0.37699252367019653, "learning_rate": 9.832523438049404e-05, "loss": 0.0283, "step": 10280 }, { "grad_norm": 0.21138913929462433, "learning_rate": 9.831992573554454e-05, "loss": 0.0328, "step": 10290 }, { "grad_norm": 0.37842848896980286, "learning_rate": 9.831460883404128e-05, "loss": 0.0239, "step": 10300 }, { "grad_norm": 0.30250707268714905, "learning_rate": 9.830928367689278e-05, "loss": 0.0225, "step": 10310 }, { "grad_norm": 0.22055499255657196, "learning_rate": 9.830395026500896e-05, "loss": 0.0207, "step": 10320 }, { "grad_norm": 0.25073254108428955, "learning_rate": 9.829860859930115e-05, "loss": 0.0226, "step": 10330 }, { "grad_norm": 0.3375846743583679, "learning_rate": 9.829325868068212e-05, "loss": 0.0213, "step": 10340 }, { "grad_norm": 0.33178791403770447, "learning_rate": 9.8287900510066e-05, "loss": 0.0258, "step": 10350 }, { "grad_norm": 0.3023558557033539, "learning_rate": 9.828253408836834e-05, "loss": 0.0222, "step": 10360 }, { "grad_norm": 0.2828664481639862, "learning_rate": 9.827715941650615e-05, "loss": 0.0189, "step": 10370 }, { "grad_norm": 0.2561330199241638, "learning_rate": 9.82717764953978e-05, "loss": 0.0211, "step": 10380 }, { "grad_norm": 0.24938279390335083, "learning_rate": 9.826638532596308e-05, "loss": 0.021, "step": 10390 }, { "grad_norm": 0.25451165437698364, "learning_rate": 9.82609859091232e-05, "loss": 0.0217, "step": 10400 }, { "grad_norm": 0.3265402019023895, "learning_rate": 9.825557824580076e-05, "loss": 0.0268, "step": 10410 }, { "grad_norm": 0.4605780243873596, "learning_rate": 9.82501623369198e-05, "loss": 0.025, "step": 10420 }, { "grad_norm": 0.28024032711982727, "learning_rate": 9.824473818340574e-05, "loss": 0.023, "step": 10430 }, { "grad_norm": 0.32186150550842285, "learning_rate": 9.823930578618541e-05, "loss": 0.0214, "step": 10440 }, { "grad_norm": 0.29954060912132263, "learning_rate": 9.823386514618709e-05, "loss": 0.0218, "step": 10450 }, { "grad_norm": 0.3176597058773041, "learning_rate": 9.82284162643404e-05, "loss": 0.0245, "step": 10460 }, { "grad_norm": 0.30501797795295715, "learning_rate": 9.822295914157642e-05, "loss": 0.0209, "step": 10470 }, { "grad_norm": 0.22111496329307556, "learning_rate": 9.821749377882763e-05, "loss": 0.0222, "step": 10480 }, { "grad_norm": 0.276327908039093, "learning_rate": 9.821202017702791e-05, "loss": 0.0242, "step": 10490 }, { "grad_norm": 0.2515116035938263, "learning_rate": 9.820653833711253e-05, "loss": 0.0211, "step": 10500 }, { "grad_norm": 0.23675008118152618, "learning_rate": 9.820104826001822e-05, "loss": 0.0289, "step": 10510 }, { "grad_norm": 0.3811260163784027, "learning_rate": 9.819554994668305e-05, "loss": 0.0239, "step": 10520 }, { "grad_norm": 0.2106151431798935, "learning_rate": 9.819004339804654e-05, "loss": 0.0317, "step": 10530 }, { "grad_norm": 0.26484403014183044, "learning_rate": 9.818452861504961e-05, "loss": 0.0243, "step": 10540 }, { "grad_norm": 0.22874492406845093, "learning_rate": 9.81790055986346e-05, "loss": 0.0238, "step": 10550 }, { "grad_norm": 0.5234336853027344, "learning_rate": 9.817347434974523e-05, "loss": 0.0221, "step": 10560 }, { "grad_norm": 0.32359492778778076, "learning_rate": 9.816793486932664e-05, "loss": 0.017, "step": 10570 }, { "grad_norm": 0.21409520506858826, "learning_rate": 9.816238715832538e-05, "loss": 0.0215, "step": 10580 }, { "grad_norm": 0.3598271608352661, "learning_rate": 9.815683121768939e-05, "loss": 0.0212, "step": 10590 }, { "grad_norm": 0.18945914506912231, "learning_rate": 9.815126704836804e-05, "loss": 0.0286, "step": 10600 }, { "grad_norm": 0.4681123197078705, "learning_rate": 9.81456946513121e-05, "loss": 0.0372, "step": 10610 }, { "grad_norm": 0.41015878319740295, "learning_rate": 9.814011402747373e-05, "loss": 0.0408, "step": 10620 }, { "grad_norm": 0.3202025592327118, "learning_rate": 9.813452517780651e-05, "loss": 0.0244, "step": 10630 }, { "grad_norm": 0.19268231093883514, "learning_rate": 9.81289281032654e-05, "loss": 0.0303, "step": 10640 }, { "grad_norm": 0.34543922543525696, "learning_rate": 9.812332280480683e-05, "loss": 0.0289, "step": 10650 }, { "grad_norm": 0.2777855694293976, "learning_rate": 9.811770928338854e-05, "loss": 0.0268, "step": 10660 }, { "grad_norm": 0.3349146246910095, "learning_rate": 9.811208753996979e-05, "loss": 0.022, "step": 10670 }, { "grad_norm": 0.3126879036426544, "learning_rate": 9.810645757551113e-05, "loss": 0.0165, "step": 10680 }, { "grad_norm": 0.2900192439556122, "learning_rate": 9.810081939097459e-05, "loss": 0.0237, "step": 10690 }, { "grad_norm": 0.3199375569820404, "learning_rate": 9.809517298732356e-05, "loss": 0.0415, "step": 10700 }, { "grad_norm": 0.2362624704837799, "learning_rate": 9.80895183655229e-05, "loss": 0.0249, "step": 10710 }, { "grad_norm": 0.40468499064445496, "learning_rate": 9.808385552653877e-05, "loss": 0.0243, "step": 10720 }, { "grad_norm": 0.3125559389591217, "learning_rate": 9.807818447133886e-05, "loss": 0.0207, "step": 10730 }, { "grad_norm": 0.6441357135772705, "learning_rate": 9.807250520089215e-05, "loss": 0.0187, "step": 10740 }, { "grad_norm": 0.203190416097641, "learning_rate": 9.806681771616908e-05, "loss": 0.0232, "step": 10750 }, { "grad_norm": 0.313574880361557, "learning_rate": 9.80611220181415e-05, "loss": 0.0257, "step": 10760 }, { "grad_norm": 0.4280308187007904, "learning_rate": 9.805541810778264e-05, "loss": 0.0286, "step": 10770 }, { "grad_norm": 0.40113821625709534, "learning_rate": 9.804970598606716e-05, "loss": 0.0306, "step": 10780 }, { "grad_norm": 0.3814411461353302, "learning_rate": 9.804398565397106e-05, "loss": 0.024, "step": 10790 }, { "grad_norm": 0.41336140036582947, "learning_rate": 9.803825711247183e-05, "loss": 0.0294, "step": 10800 }, { "grad_norm": 0.2614530324935913, "learning_rate": 9.803252036254831e-05, "loss": 0.0298, "step": 10810 }, { "grad_norm": 0.30276423692703247, "learning_rate": 9.802677540518076e-05, "loss": 0.0269, "step": 10820 }, { "grad_norm": 0.29920750856399536, "learning_rate": 9.802102224135081e-05, "loss": 0.0252, "step": 10830 }, { "grad_norm": 0.24407854676246643, "learning_rate": 9.801526087204155e-05, "loss": 0.0329, "step": 10840 }, { "grad_norm": 0.4775392711162567, "learning_rate": 9.800949129823743e-05, "loss": 0.0288, "step": 10850 }, { "grad_norm": 0.28990691900253296, "learning_rate": 9.80037135209243e-05, "loss": 0.0245, "step": 10860 }, { "grad_norm": 0.41714003682136536, "learning_rate": 9.799792754108946e-05, "loss": 0.026, "step": 10870 }, { "grad_norm": 0.2637101411819458, "learning_rate": 9.799213335972152e-05, "loss": 0.0245, "step": 10880 }, { "grad_norm": 0.3026460111141205, "learning_rate": 9.798633097781058e-05, "loss": 0.028, "step": 10890 }, { "grad_norm": 0.28150099515914917, "learning_rate": 9.79805203963481e-05, "loss": 0.0323, "step": 10900 }, { "grad_norm": 0.24346113204956055, "learning_rate": 9.797470161632697e-05, "loss": 0.0242, "step": 10910 }, { "grad_norm": 0.4075450003147125, "learning_rate": 9.796887463874145e-05, "loss": 0.0301, "step": 10920 }, { "grad_norm": 0.3592146039009094, "learning_rate": 9.796303946458718e-05, "loss": 0.0251, "step": 10930 }, { "grad_norm": 0.2919843792915344, "learning_rate": 9.795719609486127e-05, "loss": 0.029, "step": 10940 }, { "grad_norm": 0.3288089334964752, "learning_rate": 9.795134453056219e-05, "loss": 0.0306, "step": 10950 }, { "grad_norm": 0.27931922674179077, "learning_rate": 9.794548477268979e-05, "loss": 0.0204, "step": 10960 }, { "grad_norm": 0.17872262001037598, "learning_rate": 9.793961682224537e-05, "loss": 0.0244, "step": 10970 }, { "grad_norm": 0.4187450408935547, "learning_rate": 9.793374068023156e-05, "loss": 0.0304, "step": 10980 }, { "grad_norm": 0.24545417726039886, "learning_rate": 9.792785634765247e-05, "loss": 0.0259, "step": 10990 }, { "grad_norm": 0.45517557859420776, "learning_rate": 9.792196382551357e-05, "loss": 0.0238, "step": 11000 }, { "grad_norm": 0.2442779541015625, "learning_rate": 9.791606311482171e-05, "loss": 0.0245, "step": 11010 }, { "grad_norm": 0.37696877121925354, "learning_rate": 9.791015421658518e-05, "loss": 0.0269, "step": 11020 }, { "grad_norm": 0.34225666522979736, "learning_rate": 9.790423713181362e-05, "loss": 0.0326, "step": 11030 }, { "grad_norm": 0.2861303389072418, "learning_rate": 9.789831186151814e-05, "loss": 0.0257, "step": 11040 }, { "grad_norm": 0.2685452401638031, "learning_rate": 9.789237840671118e-05, "loss": 0.0219, "step": 11050 }, { "grad_norm": 0.26536306738853455, "learning_rate": 9.78864367684066e-05, "loss": 0.0216, "step": 11060 }, { "grad_norm": 0.3137136399745941, "learning_rate": 9.788048694761968e-05, "loss": 0.0193, "step": 11070 }, { "grad_norm": 0.2784353196620941, "learning_rate": 9.787452894536709e-05, "loss": 0.0245, "step": 11080 }, { "grad_norm": 0.21915574371814728, "learning_rate": 9.786856276266685e-05, "loss": 0.0222, "step": 11090 }, { "grad_norm": 0.2737283706665039, "learning_rate": 9.786258840053845e-05, "loss": 0.0323, "step": 11100 }, { "grad_norm": 0.4709624946117401, "learning_rate": 9.785660586000273e-05, "loss": 0.0276, "step": 11110 }, { "grad_norm": 0.36991244554519653, "learning_rate": 9.785061514208196e-05, "loss": 0.0221, "step": 11120 }, { "grad_norm": 0.31286314129829407, "learning_rate": 9.784461624779977e-05, "loss": 0.0245, "step": 11130 }, { "grad_norm": 0.2583399713039398, "learning_rate": 9.783860917818123e-05, "loss": 0.0229, "step": 11140 }, { "grad_norm": 0.26896408200263977, "learning_rate": 9.783259393425277e-05, "loss": 0.0173, "step": 11150 }, { "grad_norm": 0.34716787934303284, "learning_rate": 9.782657051704221e-05, "loss": 0.0208, "step": 11160 }, { "grad_norm": 0.2603589594364166, "learning_rate": 9.782053892757883e-05, "loss": 0.022, "step": 11170 }, { "grad_norm": 0.2635310888290405, "learning_rate": 9.781449916689324e-05, "loss": 0.0235, "step": 11180 }, { "grad_norm": 0.29204002022743225, "learning_rate": 9.780845123601746e-05, "loss": 0.026, "step": 11190 }, { "grad_norm": 0.33701613545417786, "learning_rate": 9.780239513598492e-05, "loss": 0.0264, "step": 11200 }, { "grad_norm": 0.3557259440422058, "learning_rate": 9.779633086783047e-05, "loss": 0.0272, "step": 11210 }, { "grad_norm": 0.3019830584526062, "learning_rate": 9.779025843259031e-05, "loss": 0.0245, "step": 11220 }, { "grad_norm": 0.3422353267669678, "learning_rate": 9.778417783130204e-05, "loss": 0.0297, "step": 11230 }, { "grad_norm": 0.3259344696998596, "learning_rate": 9.777808906500468e-05, "loss": 0.0282, "step": 11240 }, { "grad_norm": 0.35665977001190186, "learning_rate": 9.777199213473862e-05, "loss": 0.0166, "step": 11250 }, { "grad_norm": 0.3068661093711853, "learning_rate": 9.77658870415457e-05, "loss": 0.0191, "step": 11260 }, { "grad_norm": 0.16826577484607697, "learning_rate": 9.775977378646906e-05, "loss": 0.0191, "step": 11270 }, { "grad_norm": 0.36785322427749634, "learning_rate": 9.775365237055331e-05, "loss": 0.0288, "step": 11280 }, { "grad_norm": 0.284995973110199, "learning_rate": 9.774752279484445e-05, "loss": 0.0225, "step": 11290 }, { "grad_norm": 0.4566374719142914, "learning_rate": 9.774138506038984e-05, "loss": 0.0243, "step": 11300 }, { "grad_norm": 0.411964476108551, "learning_rate": 9.773523916823826e-05, "loss": 0.0223, "step": 11310 }, { "grad_norm": 0.2791539132595062, "learning_rate": 9.772908511943986e-05, "loss": 0.0223, "step": 11320 }, { "grad_norm": 0.274950236082077, "learning_rate": 9.77229229150462e-05, "loss": 0.0176, "step": 11330 }, { "grad_norm": 0.3278381824493408, "learning_rate": 9.771675255611024e-05, "loss": 0.0188, "step": 11340 }, { "grad_norm": 0.3260321617126465, "learning_rate": 9.771057404368632e-05, "loss": 0.0386, "step": 11350 }, { "grad_norm": 0.2981843948364258, "learning_rate": 9.770438737883018e-05, "loss": 0.021, "step": 11360 }, { "grad_norm": 0.34058132767677307, "learning_rate": 9.769819256259898e-05, "loss": 0.0182, "step": 11370 }, { "grad_norm": 0.29395848512649536, "learning_rate": 9.769198959605119e-05, "loss": 0.0252, "step": 11380 }, { "grad_norm": 0.33158883452415466, "learning_rate": 9.768577848024678e-05, "loss": 0.0239, "step": 11390 }, { "grad_norm": 0.32967686653137207, "learning_rate": 9.767955921624702e-05, "loss": 0.0325, "step": 11400 }, { "grad_norm": 0.3555016815662384, "learning_rate": 9.767333180511465e-05, "loss": 0.0185, "step": 11410 }, { "grad_norm": 0.34591910243034363, "learning_rate": 9.766709624791373e-05, "loss": 0.0293, "step": 11420 }, { "grad_norm": 0.2599866986274719, "learning_rate": 9.766085254570975e-05, "loss": 0.0214, "step": 11430 }, { "grad_norm": 0.31621047854423523, "learning_rate": 9.76546006995696e-05, "loss": 0.0176, "step": 11440 }, { "grad_norm": 0.3594781756401062, "learning_rate": 9.764834071056155e-05, "loss": 0.0256, "step": 11450 }, { "grad_norm": 0.3044229745864868, "learning_rate": 9.764207257975526e-05, "loss": 0.0292, "step": 11460 }, { "grad_norm": 0.3572138845920563, "learning_rate": 9.763579630822179e-05, "loss": 0.0187, "step": 11470 }, { "grad_norm": 0.450417160987854, "learning_rate": 9.762951189703356e-05, "loss": 0.0273, "step": 11480 }, { "grad_norm": 0.3392789661884308, "learning_rate": 9.762321934726442e-05, "loss": 0.0239, "step": 11490 }, { "grad_norm": 0.3502764403820038, "learning_rate": 9.761691865998959e-05, "loss": 0.0249, "step": 11500 }, { "grad_norm": 0.3046119511127472, "learning_rate": 9.76106098362857e-05, "loss": 0.0203, "step": 11510 }, { "grad_norm": 0.3923623263835907, "learning_rate": 9.760429287723072e-05, "loss": 0.0279, "step": 11520 }, { "grad_norm": 0.3310619592666626, "learning_rate": 9.759796778390406e-05, "loss": 0.0216, "step": 11530 }, { "grad_norm": 0.30520397424697876, "learning_rate": 9.759163455738653e-05, "loss": 0.0223, "step": 11540 }, { "grad_norm": 0.4248286485671997, "learning_rate": 9.75852931987603e-05, "loss": 0.0217, "step": 11550 }, { "grad_norm": 0.5656601190567017, "learning_rate": 9.757894370910891e-05, "loss": 0.0212, "step": 11560 }, { "grad_norm": 0.33392998576164246, "learning_rate": 9.757258608951733e-05, "loss": 0.0243, "step": 11570 }, { "grad_norm": 0.4599064588546753, "learning_rate": 9.75662203410719e-05, "loss": 0.0318, "step": 11580 }, { "grad_norm": 0.25468334555625916, "learning_rate": 9.755984646486034e-05, "loss": 0.0292, "step": 11590 }, { "grad_norm": 0.3394075036048889, "learning_rate": 9.75534644619718e-05, "loss": 0.0266, "step": 11600 }, { "grad_norm": 0.1987941414117813, "learning_rate": 9.754707433349676e-05, "loss": 0.0296, "step": 11610 }, { "grad_norm": 0.38624194264411926, "learning_rate": 9.754067608052715e-05, "loss": 0.0291, "step": 11620 }, { "grad_norm": 0.32510456442832947, "learning_rate": 9.753426970415622e-05, "loss": 0.0372, "step": 11630 }, { "grad_norm": 0.3301006555557251, "learning_rate": 9.752785520547868e-05, "loss": 0.0231, "step": 11640 }, { "grad_norm": 0.40944352746009827, "learning_rate": 9.752143258559056e-05, "loss": 0.0308, "step": 11650 }, { "grad_norm": 0.26617226004600525, "learning_rate": 9.751500184558933e-05, "loss": 0.025, "step": 11660 }, { "grad_norm": 0.23338474333286285, "learning_rate": 9.750856298657383e-05, "loss": 0.0214, "step": 11670 }, { "grad_norm": 0.3086144030094147, "learning_rate": 9.750211600964428e-05, "loss": 0.0247, "step": 11680 }, { "grad_norm": 0.27870285511016846, "learning_rate": 9.749566091590226e-05, "loss": 0.022, "step": 11690 }, { "grad_norm": 0.2694042921066284, "learning_rate": 9.748919770645083e-05, "loss": 0.0227, "step": 11700 }, { "grad_norm": 0.29091548919677734, "learning_rate": 9.748272638239432e-05, "loss": 0.0215, "step": 11710 }, { "grad_norm": 0.39788997173309326, "learning_rate": 9.747624694483855e-05, "loss": 0.0231, "step": 11720 }, { "grad_norm": 0.2276439368724823, "learning_rate": 9.746975939489065e-05, "loss": 0.0192, "step": 11730 }, { "grad_norm": 0.27576595544815063, "learning_rate": 9.746326373365918e-05, "loss": 0.0273, "step": 11740 }, { "grad_norm": 0.2866297662258148, "learning_rate": 9.745675996225403e-05, "loss": 0.0275, "step": 11750 }, { "grad_norm": 0.19523926079273224, "learning_rate": 9.745024808178657e-05, "loss": 0.0233, "step": 11760 }, { "grad_norm": 0.4059751331806183, "learning_rate": 9.744372809336947e-05, "loss": 0.0259, "step": 11770 }, { "grad_norm": 0.43276646733283997, "learning_rate": 9.743719999811682e-05, "loss": 0.0225, "step": 11780 }, { "grad_norm": 0.20850950479507446, "learning_rate": 9.743066379714412e-05, "loss": 0.0212, "step": 11790 }, { "grad_norm": 0.1906794011592865, "learning_rate": 9.74241194915682e-05, "loss": 0.0215, "step": 11800 }, { "grad_norm": 0.3023849129676819, "learning_rate": 9.741756708250731e-05, "loss": 0.0288, "step": 11810 }, { "grad_norm": 0.24043144285678864, "learning_rate": 9.741100657108109e-05, "loss": 0.0211, "step": 11820 }, { "grad_norm": 0.3994988799095154, "learning_rate": 9.740443795841054e-05, "loss": 0.0279, "step": 11830 }, { "grad_norm": 0.30430707335472107, "learning_rate": 9.739786124561805e-05, "loss": 0.0185, "step": 11840 }, { "grad_norm": 0.17964760959148407, "learning_rate": 9.73912764338274e-05, "loss": 0.0191, "step": 11850 }, { "grad_norm": 0.2813236117362976, "learning_rate": 9.738468352416377e-05, "loss": 0.0228, "step": 11860 }, { "grad_norm": 0.23637111485004425, "learning_rate": 9.737808251775369e-05, "loss": 0.0205, "step": 11870 }, { "grad_norm": 0.22051045298576355, "learning_rate": 9.737147341572512e-05, "loss": 0.0319, "step": 11880 }, { "grad_norm": 0.3339405357837677, "learning_rate": 9.736485621920735e-05, "loss": 0.0212, "step": 11890 }, { "grad_norm": 0.25371554493904114, "learning_rate": 9.735823092933108e-05, "loss": 0.0248, "step": 11900 }, { "grad_norm": 0.22587251663208008, "learning_rate": 9.735159754722838e-05, "loss": 0.0263, "step": 11910 }, { "grad_norm": 0.25667548179626465, "learning_rate": 9.734495607403275e-05, "loss": 0.0199, "step": 11920 }, { "grad_norm": 0.3088631331920624, "learning_rate": 9.733830651087901e-05, "loss": 0.0242, "step": 11930 }, { "grad_norm": 0.33494865894317627, "learning_rate": 9.733164885890338e-05, "loss": 0.0227, "step": 11940 }, { "grad_norm": 0.26097357273101807, "learning_rate": 9.732498311924349e-05, "loss": 0.015, "step": 11950 }, { "grad_norm": 0.29431307315826416, "learning_rate": 9.731830929303833e-05, "loss": 0.0178, "step": 11960 }, { "grad_norm": 0.2787224352359772, "learning_rate": 9.731162738142827e-05, "loss": 0.0266, "step": 11970 }, { "grad_norm": 0.3049173355102539, "learning_rate": 9.730493738555506e-05, "loss": 0.0242, "step": 11980 }, { "grad_norm": 0.18292729556560516, "learning_rate": 9.729823930656186e-05, "loss": 0.0262, "step": 11990 }, { "grad_norm": 0.32706257700920105, "learning_rate": 9.729153314559316e-05, "loss": 0.0263, "step": 12000 }, { "grad_norm": 0.3631448447704315, "learning_rate": 9.728481890379486e-05, "loss": 0.0278, "step": 12010 }, { "grad_norm": 0.2254846692085266, "learning_rate": 9.727809658231428e-05, "loss": 0.0241, "step": 12020 }, { "grad_norm": 0.27699175477027893, "learning_rate": 9.727136618230003e-05, "loss": 0.0233, "step": 12030 }, { "grad_norm": 0.579538881778717, "learning_rate": 9.726462770490219e-05, "loss": 0.0286, "step": 12040 }, { "grad_norm": 0.33228254318237305, "learning_rate": 9.725788115127214e-05, "loss": 0.0294, "step": 12050 }, { "grad_norm": 0.27988991141319275, "learning_rate": 9.725112652256274e-05, "loss": 0.0253, "step": 12060 }, { "grad_norm": 0.21687909960746765, "learning_rate": 9.724436381992812e-05, "loss": 0.0361, "step": 12070 }, { "grad_norm": 0.3997022211551666, "learning_rate": 9.723759304452387e-05, "loss": 0.023, "step": 12080 }, { "grad_norm": 0.28098368644714355, "learning_rate": 9.72308141975069e-05, "loss": 0.024, "step": 12090 }, { "grad_norm": 0.2551739811897278, "learning_rate": 9.722402728003557e-05, "loss": 0.0263, "step": 12100 }, { "grad_norm": 0.33159777522087097, "learning_rate": 9.721723229326953e-05, "loss": 0.0291, "step": 12110 }, { "grad_norm": 0.2785235047340393, "learning_rate": 9.721042923836992e-05, "loss": 0.0215, "step": 12120 }, { "grad_norm": 0.2602385878562927, "learning_rate": 9.720361811649914e-05, "loss": 0.0224, "step": 12130 }, { "grad_norm": 0.3155183792114258, "learning_rate": 9.719679892882106e-05, "loss": 0.0202, "step": 12140 }, { "grad_norm": 0.32861950993537903, "learning_rate": 9.718997167650085e-05, "loss": 0.02, "step": 12150 }, { "grad_norm": 0.228573739528656, "learning_rate": 9.718313636070515e-05, "loss": 0.0176, "step": 12160 }, { "grad_norm": 0.21219545602798462, "learning_rate": 9.717629298260192e-05, "loss": 0.0217, "step": 12170 }, { "grad_norm": 0.44793522357940674, "learning_rate": 9.716944154336047e-05, "loss": 0.0326, "step": 12180 }, { "grad_norm": 0.2800026535987854, "learning_rate": 9.716258204415157e-05, "loss": 0.0212, "step": 12190 }, { "grad_norm": 0.22730140388011932, "learning_rate": 9.715571448614728e-05, "loss": 0.0312, "step": 12200 }, { "grad_norm": 0.1865977793931961, "learning_rate": 9.71488388705211e-05, "loss": 0.0342, "step": 12210 }, { "grad_norm": 0.36689311265945435, "learning_rate": 9.714195519844788e-05, "loss": 0.0216, "step": 12220 }, { "grad_norm": 0.24878574907779694, "learning_rate": 9.713506347110386e-05, "loss": 0.0256, "step": 12230 }, { "grad_norm": 0.17421536147594452, "learning_rate": 9.712816368966662e-05, "loss": 0.0203, "step": 12240 }, { "grad_norm": 0.23690682649612427, "learning_rate": 9.712125585531517e-05, "loss": 0.0303, "step": 12250 }, { "grad_norm": 0.30845099687576294, "learning_rate": 9.711433996922988e-05, "loss": 0.0361, "step": 12260 }, { "grad_norm": 0.35339298844337463, "learning_rate": 9.710741603259245e-05, "loss": 0.0252, "step": 12270 }, { "grad_norm": 0.38159453868865967, "learning_rate": 9.710048404658603e-05, "loss": 0.0233, "step": 12280 }, { "grad_norm": 0.3450494110584259, "learning_rate": 9.709354401239508e-05, "loss": 0.0287, "step": 12290 }, { "grad_norm": 0.25067979097366333, "learning_rate": 9.708659593120546e-05, "loss": 0.016, "step": 12300 }, { "grad_norm": 0.32074233889579773, "learning_rate": 9.707963980420443e-05, "loss": 0.0215, "step": 12310 }, { "grad_norm": 0.48796001076698303, "learning_rate": 9.707267563258058e-05, "loss": 0.0279, "step": 12320 }, { "grad_norm": 0.25022226572036743, "learning_rate": 9.70657034175239e-05, "loss": 0.0244, "step": 12330 }, { "grad_norm": 0.42827272415161133, "learning_rate": 9.705872316022577e-05, "loss": 0.0267, "step": 12340 }, { "grad_norm": 0.28047019243240356, "learning_rate": 9.705173486187891e-05, "loss": 0.0243, "step": 12350 }, { "grad_norm": 0.2986035943031311, "learning_rate": 9.704473852367741e-05, "loss": 0.0305, "step": 12360 }, { "grad_norm": 0.203781858086586, "learning_rate": 9.70377341468168e-05, "loss": 0.0191, "step": 12370 }, { "grad_norm": 0.2310526818037033, "learning_rate": 9.703072173249389e-05, "loss": 0.0222, "step": 12380 }, { "grad_norm": 0.36542123556137085, "learning_rate": 9.702370128190693e-05, "loss": 0.0239, "step": 12390 }, { "grad_norm": 0.3205857574939728, "learning_rate": 9.701667279625552e-05, "loss": 0.026, "step": 12400 }, { "grad_norm": 0.36810797452926636, "learning_rate": 9.700963627674065e-05, "loss": 0.0287, "step": 12410 }, { "grad_norm": 0.24383792281150818, "learning_rate": 9.700259172456466e-05, "loss": 0.0182, "step": 12420 }, { "grad_norm": 0.521928071975708, "learning_rate": 9.699553914093124e-05, "loss": 0.0316, "step": 12430 }, { "grad_norm": 0.2990049719810486, "learning_rate": 9.698847852704553e-05, "loss": 0.0273, "step": 12440 }, { "grad_norm": 0.4244701862335205, "learning_rate": 9.6981409884114e-05, "loss": 0.0264, "step": 12450 }, { "grad_norm": 0.2849758565425873, "learning_rate": 9.697433321334443e-05, "loss": 0.0265, "step": 12460 }, { "grad_norm": 0.396987646818161, "learning_rate": 9.696724851594607e-05, "loss": 0.0238, "step": 12470 }, { "grad_norm": 0.1717742532491684, "learning_rate": 9.696015579312952e-05, "loss": 0.0197, "step": 12480 }, { "grad_norm": 0.26801058650016785, "learning_rate": 9.695305504610668e-05, "loss": 0.0191, "step": 12490 }, { "grad_norm": 0.28952234983444214, "learning_rate": 9.694594627609092e-05, "loss": 0.0233, "step": 12500 }, { "grad_norm": 0.24543651938438416, "learning_rate": 9.693882948429691e-05, "loss": 0.0304, "step": 12510 }, { "grad_norm": 0.23367737233638763, "learning_rate": 9.693170467194071e-05, "loss": 0.0192, "step": 12520 }, { "grad_norm": 0.201235830783844, "learning_rate": 9.692457184023977e-05, "loss": 0.0219, "step": 12530 }, { "grad_norm": 0.32650771737098694, "learning_rate": 9.691743099041291e-05, "loss": 0.0262, "step": 12540 }, { "grad_norm": 0.29742807149887085, "learning_rate": 9.691028212368027e-05, "loss": 0.0255, "step": 12550 }, { "grad_norm": 0.3648836612701416, "learning_rate": 9.690312524126342e-05, "loss": 0.0218, "step": 12560 }, { "grad_norm": 0.22191715240478516, "learning_rate": 9.689596034438527e-05, "loss": 0.0228, "step": 12570 }, { "grad_norm": 0.14653372764587402, "learning_rate": 9.688878743427012e-05, "loss": 0.0199, "step": 12580 }, { "grad_norm": 0.2255648374557495, "learning_rate": 9.688160651214359e-05, "loss": 0.0283, "step": 12590 }, { "grad_norm": 0.28975799679756165, "learning_rate": 9.687441757923273e-05, "loss": 0.0247, "step": 12600 }, { "grad_norm": 0.20374447107315063, "learning_rate": 9.68672206367659e-05, "loss": 0.0194, "step": 12610 }, { "grad_norm": 0.2938326895236969, "learning_rate": 9.686001568597291e-05, "loss": 0.0189, "step": 12620 }, { "grad_norm": 0.3075612187385559, "learning_rate": 9.685280272808486e-05, "loss": 0.02, "step": 12630 }, { "grad_norm": 0.4020783007144928, "learning_rate": 9.684558176433424e-05, "loss": 0.0278, "step": 12640 }, { "grad_norm": 0.31895536184310913, "learning_rate": 9.683835279595495e-05, "loss": 0.025, "step": 12650 }, { "grad_norm": 0.32626503705978394, "learning_rate": 9.683111582418216e-05, "loss": 0.0327, "step": 12660 }, { "grad_norm": 0.29727479815483093, "learning_rate": 9.682387085025254e-05, "loss": 0.0214, "step": 12670 }, { "grad_norm": 0.2723451554775238, "learning_rate": 9.681661787540401e-05, "loss": 0.0258, "step": 12680 }, { "grad_norm": 0.261320561170578, "learning_rate": 9.680935690087593e-05, "loss": 0.0337, "step": 12690 }, { "grad_norm": 0.267751008272171, "learning_rate": 9.680208792790901e-05, "loss": 0.0272, "step": 12700 }, { "grad_norm": 0.22698968648910522, "learning_rate": 9.679481095774529e-05, "loss": 0.0193, "step": 12710 }, { "grad_norm": 0.34060215950012207, "learning_rate": 9.678752599162822e-05, "loss": 0.0232, "step": 12720 }, { "grad_norm": 0.37164485454559326, "learning_rate": 9.678023303080259e-05, "loss": 0.0219, "step": 12730 }, { "grad_norm": 0.35995736718177795, "learning_rate": 9.677293207651459e-05, "loss": 0.0256, "step": 12740 }, { "grad_norm": 0.248500794172287, "learning_rate": 9.676562313001173e-05, "loss": 0.018, "step": 12750 }, { "grad_norm": 0.5148802399635315, "learning_rate": 9.675830619254293e-05, "loss": 0.0294, "step": 12760 }, { "grad_norm": 0.40824371576309204, "learning_rate": 9.675098126535843e-05, "loss": 0.0344, "step": 12770 }, { "grad_norm": 0.33065903186798096, "learning_rate": 9.674364834970988e-05, "loss": 0.0279, "step": 12780 }, { "grad_norm": 0.3212364912033081, "learning_rate": 9.673630744685028e-05, "loss": 0.0274, "step": 12790 }, { "grad_norm": 0.21698260307312012, "learning_rate": 9.672895855803397e-05, "loss": 0.0226, "step": 12800 }, { "grad_norm": 0.24838607013225555, "learning_rate": 9.672160168451667e-05, "loss": 0.0225, "step": 12810 }, { "grad_norm": 0.5837302803993225, "learning_rate": 9.671423682755549e-05, "loss": 0.03, "step": 12820 }, { "grad_norm": 0.6438053846359253, "learning_rate": 9.670686398840888e-05, "loss": 0.0357, "step": 12830 }, { "grad_norm": 0.2621347904205322, "learning_rate": 9.669948316833664e-05, "loss": 0.0209, "step": 12840 }, { "grad_norm": 0.21400175988674164, "learning_rate": 9.669209436859997e-05, "loss": 0.0226, "step": 12850 }, { "grad_norm": 0.38587644696235657, "learning_rate": 9.66846975904614e-05, "loss": 0.0222, "step": 12860 }, { "grad_norm": 0.3819001317024231, "learning_rate": 9.667729283518483e-05, "loss": 0.0238, "step": 12870 }, { "grad_norm": 0.3110886514186859, "learning_rate": 9.666988010403557e-05, "loss": 0.0238, "step": 12880 }, { "grad_norm": 0.3146665692329407, "learning_rate": 9.66624593982802e-05, "loss": 0.0212, "step": 12890 }, { "grad_norm": 0.25264158844947815, "learning_rate": 9.665503071918675e-05, "loss": 0.0297, "step": 12900 }, { "grad_norm": 0.29969117045402527, "learning_rate": 9.664759406802456e-05, "loss": 0.0197, "step": 12910 }, { "grad_norm": 0.27436524629592896, "learning_rate": 9.664014944606437e-05, "loss": 0.0146, "step": 12920 }, { "grad_norm": 0.1972460150718689, "learning_rate": 9.663269685457822e-05, "loss": 0.0199, "step": 12930 }, { "grad_norm": 0.34136074781417847, "learning_rate": 9.662523629483962e-05, "loss": 0.0225, "step": 12940 }, { "grad_norm": 0.19914859533309937, "learning_rate": 9.661776776812333e-05, "loss": 0.0184, "step": 12950 }, { "grad_norm": 0.32973822951316833, "learning_rate": 9.661029127570553e-05, "loss": 0.0217, "step": 12960 }, { "grad_norm": 0.3050682544708252, "learning_rate": 9.660280681886373e-05, "loss": 0.0195, "step": 12970 }, { "grad_norm": 0.2538144290447235, "learning_rate": 9.659531439887685e-05, "loss": 0.0156, "step": 12980 }, { "grad_norm": 0.28937527537345886, "learning_rate": 9.658781401702511e-05, "loss": 0.0233, "step": 12990 }, { "grad_norm": 0.30720746517181396, "learning_rate": 9.658030567459015e-05, "loss": 0.0216, "step": 13000 }, { "grad_norm": 0.3193855583667755, "learning_rate": 9.65727893728549e-05, "loss": 0.0387, "step": 13010 }, { "grad_norm": 0.3132772743701935, "learning_rate": 9.656526511310375e-05, "loss": 0.0299, "step": 13020 }, { "grad_norm": 0.31523922085762024, "learning_rate": 9.655773289662233e-05, "loss": 0.0264, "step": 13030 }, { "grad_norm": 0.3834327757358551, "learning_rate": 9.655019272469772e-05, "loss": 0.0263, "step": 13040 }, { "grad_norm": 0.23515942692756653, "learning_rate": 9.654264459861832e-05, "loss": 0.017, "step": 13050 }, { "grad_norm": 0.3151296377182007, "learning_rate": 9.653508851967391e-05, "loss": 0.021, "step": 13060 }, { "grad_norm": 0.3465896248817444, "learning_rate": 9.65275244891556e-05, "loss": 0.0276, "step": 13070 }, { "grad_norm": 0.26790913939476013, "learning_rate": 9.651995250835591e-05, "loss": 0.0207, "step": 13080 }, { "grad_norm": 0.27479469776153564, "learning_rate": 9.651237257856862e-05, "loss": 0.0246, "step": 13090 }, { "grad_norm": 0.2645755708217621, "learning_rate": 9.6504784701089e-05, "loss": 0.0224, "step": 13100 }, { "grad_norm": 0.28728803992271423, "learning_rate": 9.649718887721357e-05, "loss": 0.0183, "step": 13110 }, { "grad_norm": 0.26865604519844055, "learning_rate": 9.648958510824028e-05, "loss": 0.021, "step": 13120 }, { "grad_norm": 0.3110477924346924, "learning_rate": 9.648197339546837e-05, "loss": 0.0277, "step": 13130 }, { "grad_norm": 0.31133517622947693, "learning_rate": 9.647435374019851e-05, "loss": 0.0244, "step": 13140 }, { "grad_norm": 0.8020980954170227, "learning_rate": 9.646672614373266e-05, "loss": 0.0351, "step": 13150 }, { "grad_norm": 0.5853726863861084, "learning_rate": 9.645909060737418e-05, "loss": 0.0258, "step": 13160 }, { "grad_norm": 0.22455738484859467, "learning_rate": 9.645144713242778e-05, "loss": 0.0233, "step": 13170 }, { "grad_norm": 0.2643216550350189, "learning_rate": 9.64437957201995e-05, "loss": 0.0258, "step": 13180 }, { "grad_norm": 0.29261642694473267, "learning_rate": 9.643613637199678e-05, "loss": 0.0183, "step": 13190 }, { "grad_norm": 0.2701864242553711, "learning_rate": 9.642846908912839e-05, "loss": 0.029, "step": 13200 }, { "grad_norm": 0.23113039135932922, "learning_rate": 9.642079387290444e-05, "loss": 0.0179, "step": 13210 }, { "grad_norm": 0.20170219242572784, "learning_rate": 9.641311072463644e-05, "loss": 0.0185, "step": 13220 }, { "grad_norm": 0.26214832067489624, "learning_rate": 9.640541964563722e-05, "loss": 0.0288, "step": 13230 }, { "grad_norm": 0.3137769401073456, "learning_rate": 9.639772063722096e-05, "loss": 0.0288, "step": 13240 }, { "grad_norm": 0.26078665256500244, "learning_rate": 9.639001370070324e-05, "loss": 0.0217, "step": 13250 }, { "grad_norm": 0.41304123401641846, "learning_rate": 9.638229883740095e-05, "loss": 0.0204, "step": 13260 }, { "grad_norm": 0.5294812321662903, "learning_rate": 9.637457604863233e-05, "loss": 0.0187, "step": 13270 }, { "grad_norm": 0.31818702816963196, "learning_rate": 9.636684533571703e-05, "loss": 0.0259, "step": 13280 }, { "grad_norm": 0.2647780776023865, "learning_rate": 9.635910669997599e-05, "loss": 0.02, "step": 13290 }, { "grad_norm": 0.247688889503479, "learning_rate": 9.635136014273154e-05, "loss": 0.0228, "step": 13300 }, { "grad_norm": 0.25587671995162964, "learning_rate": 9.634360566530735e-05, "loss": 0.0277, "step": 13310 }, { "grad_norm": 0.40175455808639526, "learning_rate": 9.633584326902845e-05, "loss": 0.0318, "step": 13320 }, { "grad_norm": 0.29588326811790466, "learning_rate": 9.632807295522124e-05, "loss": 0.0235, "step": 13330 }, { "grad_norm": 0.27941519021987915, "learning_rate": 9.632029472521342e-05, "loss": 0.0211, "step": 13340 }, { "grad_norm": 0.31558454036712646, "learning_rate": 9.631250858033409e-05, "loss": 0.0207, "step": 13350 }, { "grad_norm": 0.2909889221191406, "learning_rate": 9.630471452191371e-05, "loss": 0.0262, "step": 13360 }, { "grad_norm": 0.552111029624939, "learning_rate": 9.629691255128405e-05, "loss": 0.0294, "step": 13370 }, { "grad_norm": 0.5317257046699524, "learning_rate": 9.628910266977825e-05, "loss": 0.0233, "step": 13380 }, { "grad_norm": 0.33093592524528503, "learning_rate": 9.628128487873083e-05, "loss": 0.0261, "step": 13390 }, { "grad_norm": 0.28064170479774475, "learning_rate": 9.627345917947761e-05, "loss": 0.0312, "step": 13400 }, { "grad_norm": 0.2168944776058197, "learning_rate": 9.626562557335579e-05, "loss": 0.0234, "step": 13410 }, { "grad_norm": 0.2015625536441803, "learning_rate": 9.625778406170393e-05, "loss": 0.0206, "step": 13420 }, { "grad_norm": 0.2616831362247467, "learning_rate": 9.624993464586193e-05, "loss": 0.0214, "step": 13430 }, { "grad_norm": 0.29530370235443115, "learning_rate": 9.624207732717105e-05, "loss": 0.024, "step": 13440 }, { "grad_norm": 0.5304392576217651, "learning_rate": 9.623421210697386e-05, "loss": 0.0292, "step": 13450 }, { "grad_norm": 0.2859516739845276, "learning_rate": 9.622633898661434e-05, "loss": 0.0305, "step": 13460 }, { "grad_norm": 0.2933354079723358, "learning_rate": 9.621845796743778e-05, "loss": 0.0329, "step": 13470 }, { "grad_norm": 0.26539677381515503, "learning_rate": 9.621056905079082e-05, "loss": 0.0325, "step": 13480 }, { "grad_norm": 0.24846602976322174, "learning_rate": 9.620267223802149e-05, "loss": 0.0257, "step": 13490 }, { "grad_norm": 0.3724942207336426, "learning_rate": 9.619476753047911e-05, "loss": 0.0226, "step": 13500 }, { "grad_norm": 0.34150230884552, "learning_rate": 9.618685492951438e-05, "loss": 0.0268, "step": 13510 }, { "grad_norm": 0.2843014597892761, "learning_rate": 9.617893443647938e-05, "loss": 0.0308, "step": 13520 }, { "grad_norm": 0.35825422406196594, "learning_rate": 9.617100605272746e-05, "loss": 0.0312, "step": 13530 }, { "grad_norm": 0.25264281034469604, "learning_rate": 9.616306977961338e-05, "loss": 0.025, "step": 13540 }, { "grad_norm": 0.2732946574687958, "learning_rate": 9.615512561849326e-05, "loss": 0.0215, "step": 13550 }, { "grad_norm": 0.27659037709236145, "learning_rate": 9.61471735707245e-05, "loss": 0.027, "step": 13560 }, { "grad_norm": 0.2820923328399658, "learning_rate": 9.613921363766592e-05, "loss": 0.0215, "step": 13570 }, { "grad_norm": 0.4080924391746521, "learning_rate": 9.613124582067763e-05, "loss": 0.0308, "step": 13580 }, { "grad_norm": 0.3774625062942505, "learning_rate": 9.612327012112112e-05, "loss": 0.0276, "step": 13590 }, { "grad_norm": 0.2715129554271698, "learning_rate": 9.611528654035921e-05, "loss": 0.0354, "step": 13600 }, { "grad_norm": 0.2032877802848816, "learning_rate": 9.610729507975611e-05, "loss": 0.0212, "step": 13610 }, { "grad_norm": 0.2648811936378479, "learning_rate": 9.609929574067731e-05, "loss": 0.0311, "step": 13620 }, { "grad_norm": 0.2508719861507416, "learning_rate": 9.609128852448967e-05, "loss": 0.0192, "step": 13630 }, { "grad_norm": 0.2081982046365738, "learning_rate": 9.608327343256143e-05, "loss": 0.0233, "step": 13640 }, { "grad_norm": 0.37065109610557556, "learning_rate": 9.607525046626216e-05, "loss": 0.0241, "step": 13650 }, { "grad_norm": 0.38215431571006775, "learning_rate": 9.606721962696272e-05, "loss": 0.0302, "step": 13660 }, { "grad_norm": 0.23335258662700653, "learning_rate": 9.60591809160354e-05, "loss": 0.0218, "step": 13670 }, { "grad_norm": 0.2483726143836975, "learning_rate": 9.605113433485378e-05, "loss": 0.0271, "step": 13680 }, { "grad_norm": 0.39725732803344727, "learning_rate": 9.604307988479279e-05, "loss": 0.0229, "step": 13690 }, { "grad_norm": 0.30350592732429504, "learning_rate": 9.603501756722876e-05, "loss": 0.0244, "step": 13700 }, { "grad_norm": 0.19130076467990875, "learning_rate": 9.602694738353927e-05, "loss": 0.0205, "step": 13710 }, { "grad_norm": 0.2632615566253662, "learning_rate": 9.601886933510331e-05, "loss": 0.019, "step": 13720 }, { "grad_norm": 0.31542444229125977, "learning_rate": 9.60107834233012e-05, "loss": 0.0174, "step": 13730 }, { "grad_norm": 0.27313968539237976, "learning_rate": 9.60026896495146e-05, "loss": 0.0187, "step": 13740 }, { "grad_norm": 0.2130453735589981, "learning_rate": 9.599458801512652e-05, "loss": 0.0191, "step": 13750 }, { "grad_norm": 0.31196314096450806, "learning_rate": 9.598647852152129e-05, "loss": 0.0339, "step": 13760 }, { "grad_norm": 0.28570619225502014, "learning_rate": 9.597836117008462e-05, "loss": 0.0288, "step": 13770 }, { "grad_norm": 0.29575642943382263, "learning_rate": 9.597023596220356e-05, "loss": 0.022, "step": 13780 }, { "grad_norm": 0.3790680468082428, "learning_rate": 9.596210289926643e-05, "loss": 0.0229, "step": 13790 }, { "grad_norm": 0.276998907327652, "learning_rate": 9.5953961982663e-05, "loss": 0.0222, "step": 13800 }, { "grad_norm": 0.26410895586013794, "learning_rate": 9.594581321378431e-05, "loss": 0.0313, "step": 13810 }, { "grad_norm": 0.23868428170681, "learning_rate": 9.593765659402276e-05, "loss": 0.0256, "step": 13820 }, { "grad_norm": 0.3912113308906555, "learning_rate": 9.59294921247721e-05, "loss": 0.0265, "step": 13830 }, { "grad_norm": 0.23749832808971405, "learning_rate": 9.59213198074274e-05, "loss": 0.0308, "step": 13840 }, { "grad_norm": 0.32836371660232544, "learning_rate": 9.59131396433851e-05, "loss": 0.0229, "step": 13850 }, { "grad_norm": 0.3529205620288849, "learning_rate": 9.590495163404297e-05, "loss": 0.0237, "step": 13860 }, { "grad_norm": 0.2755950689315796, "learning_rate": 9.589675578080009e-05, "loss": 0.0259, "step": 13870 }, { "grad_norm": 0.31931933760643005, "learning_rate": 9.588855208505694e-05, "loss": 0.017, "step": 13880 }, { "grad_norm": 0.41511988639831543, "learning_rate": 9.588034054821529e-05, "loss": 0.0312, "step": 13890 }, { "grad_norm": 0.2880419194698334, "learning_rate": 9.587212117167826e-05, "loss": 0.0187, "step": 13900 }, { "grad_norm": 0.29219573736190796, "learning_rate": 9.586389395685033e-05, "loss": 0.0232, "step": 13910 }, { "grad_norm": 0.5746498703956604, "learning_rate": 9.585565890513733e-05, "loss": 0.0282, "step": 13920 }, { "grad_norm": 0.25299298763275146, "learning_rate": 9.584741601794636e-05, "loss": 0.0221, "step": 13930 }, { "grad_norm": 0.3440326750278473, "learning_rate": 9.58391652966859e-05, "loss": 0.024, "step": 13940 }, { "grad_norm": 0.2839961349964142, "learning_rate": 9.583090674276583e-05, "loss": 0.0227, "step": 13950 }, { "grad_norm": 0.19907796382904053, "learning_rate": 9.582264035759726e-05, "loss": 0.0222, "step": 13960 }, { "grad_norm": 0.2957085967063904, "learning_rate": 9.58143661425927e-05, "loss": 0.0419, "step": 13970 }, { "grad_norm": 0.28418904542922974, "learning_rate": 9.580608409916601e-05, "loss": 0.0302, "step": 13980 }, { "grad_norm": 0.3105297386646271, "learning_rate": 9.579779422873233e-05, "loss": 0.0303, "step": 13990 }, { "grad_norm": 0.38968971371650696, "learning_rate": 9.578949653270819e-05, "loss": 0.0256, "step": 14000 }, { "grad_norm": 0.40872225165367126, "learning_rate": 9.578119101251144e-05, "loss": 0.0251, "step": 14010 }, { "grad_norm": 0.29238978028297424, "learning_rate": 9.577287766956127e-05, "loss": 0.0295, "step": 14020 }, { "grad_norm": 0.40705811977386475, "learning_rate": 9.57645565052782e-05, "loss": 0.0356, "step": 14030 }, { "grad_norm": 0.4780283272266388, "learning_rate": 9.575622752108407e-05, "loss": 0.0237, "step": 14040 }, { "grad_norm": 0.32147306203842163, "learning_rate": 9.57478907184021e-05, "loss": 0.0227, "step": 14050 }, { "grad_norm": 0.25067654252052307, "learning_rate": 9.573954609865681e-05, "loss": 0.0229, "step": 14060 }, { "grad_norm": 0.2983788847923279, "learning_rate": 9.573119366327408e-05, "loss": 0.0205, "step": 14070 }, { "grad_norm": 0.23809589445590973, "learning_rate": 9.57228334136811e-05, "loss": 0.0206, "step": 14080 }, { "grad_norm": 0.2838675081729889, "learning_rate": 9.571446535130641e-05, "loss": 0.0232, "step": 14090 }, { "grad_norm": 0.18362201750278473, "learning_rate": 9.570608947757988e-05, "loss": 0.024, "step": 14100 }, { "grad_norm": 0.2252350002527237, "learning_rate": 9.569770579393274e-05, "loss": 0.0233, "step": 14110 }, { "grad_norm": 0.2508869469165802, "learning_rate": 9.56893143017975e-05, "loss": 0.0226, "step": 14120 }, { "grad_norm": 0.30034539103507996, "learning_rate": 9.568091500260806e-05, "loss": 0.0273, "step": 14130 }, { "grad_norm": 0.30352553725242615, "learning_rate": 9.567250789779961e-05, "loss": 0.026, "step": 14140 }, { "grad_norm": 0.8080687522888184, "learning_rate": 9.566409298880872e-05, "loss": 0.0232, "step": 14150 }, { "grad_norm": 0.3606376647949219, "learning_rate": 9.565567027707326e-05, "loss": 0.0405, "step": 14160 }, { "grad_norm": 0.28583699464797974, "learning_rate": 9.56472397640324e-05, "loss": 0.0257, "step": 14170 }, { "grad_norm": 0.3199436068534851, "learning_rate": 9.563880145112675e-05, "loss": 0.026, "step": 14180 }, { "grad_norm": 0.3445911407470703, "learning_rate": 9.563035533979814e-05, "loss": 0.026, "step": 14190 }, { "grad_norm": 0.3994375765323639, "learning_rate": 9.562190143148981e-05, "loss": 0.0231, "step": 14200 }, { "grad_norm": 0.22041818499565125, "learning_rate": 9.561343972764627e-05, "loss": 0.0202, "step": 14210 }, { "grad_norm": 0.44144636392593384, "learning_rate": 9.560497022971343e-05, "loss": 0.0337, "step": 14220 }, { "grad_norm": 0.24791496992111206, "learning_rate": 9.559649293913847e-05, "loss": 0.0228, "step": 14230 }, { "grad_norm": 0.3043995201587677, "learning_rate": 9.558800785736993e-05, "loss": 0.0221, "step": 14240 }, { "grad_norm": 0.22914564609527588, "learning_rate": 9.557951498585767e-05, "loss": 0.0257, "step": 14250 }, { "grad_norm": 0.30935221910476685, "learning_rate": 9.557101432605293e-05, "loss": 0.0206, "step": 14260 }, { "grad_norm": 0.3761961758136749, "learning_rate": 9.556250587940818e-05, "loss": 0.0261, "step": 14270 }, { "grad_norm": 0.25872206687927246, "learning_rate": 9.555398964737734e-05, "loss": 0.022, "step": 14280 }, { "grad_norm": 0.3298124670982361, "learning_rate": 9.554546563141555e-05, "loss": 0.0209, "step": 14290 }, { "grad_norm": 0.30787158012390137, "learning_rate": 9.553693383297937e-05, "loss": 0.0287, "step": 14300 }, { "grad_norm": 0.3508327305316925, "learning_rate": 9.552839425352663e-05, "loss": 0.0211, "step": 14310 }, { "grad_norm": 0.22550620138645172, "learning_rate": 9.551984689451652e-05, "loss": 0.0325, "step": 14320 }, { "grad_norm": 0.27113115787506104, "learning_rate": 9.551129175740953e-05, "loss": 0.024, "step": 14330 }, { "grad_norm": 0.19265729188919067, "learning_rate": 9.550272884366754e-05, "loss": 0.0257, "step": 14340 }, { "grad_norm": 0.25516000390052795, "learning_rate": 9.549415815475369e-05, "loss": 0.0207, "step": 14350 }, { "grad_norm": 0.15886569023132324, "learning_rate": 9.548557969213247e-05, "loss": 0.0194, "step": 14360 }, { "grad_norm": 0.3015773296356201, "learning_rate": 9.547699345726972e-05, "loss": 0.0244, "step": 14370 }, { "grad_norm": 0.21709127724170685, "learning_rate": 9.546839945163257e-05, "loss": 0.0235, "step": 14380 }, { "grad_norm": 0.31098270416259766, "learning_rate": 9.545979767668953e-05, "loss": 0.0318, "step": 14390 }, { "grad_norm": 0.2551332712173462, "learning_rate": 9.54511881339104e-05, "loss": 0.0235, "step": 14400 }, { "grad_norm": 0.319835901260376, "learning_rate": 9.54425708247663e-05, "loss": 0.0199, "step": 14410 }, { "grad_norm": 0.5061445832252502, "learning_rate": 9.543394575072972e-05, "loss": 0.0229, "step": 14420 }, { "grad_norm": 0.3227766454219818, "learning_rate": 9.542531291327441e-05, "loss": 0.0231, "step": 14430 }, { "grad_norm": 0.28639042377471924, "learning_rate": 9.541667231387552e-05, "loss": 0.0181, "step": 14440 }, { "grad_norm": 0.2749623954296112, "learning_rate": 9.540802395400949e-05, "loss": 0.0263, "step": 14450 }, { "grad_norm": 0.21590326726436615, "learning_rate": 9.539936783515406e-05, "loss": 0.0245, "step": 14460 }, { "grad_norm": 0.29182153940200806, "learning_rate": 9.539070395878835e-05, "loss": 0.0253, "step": 14470 }, { "grad_norm": 0.3400528132915497, "learning_rate": 9.538203232639277e-05, "loss": 0.0203, "step": 14480 }, { "grad_norm": 0.3762671947479248, "learning_rate": 9.537335293944907e-05, "loss": 0.0264, "step": 14490 }, { "grad_norm": 0.2210482656955719, "learning_rate": 9.536466579944032e-05, "loss": 0.0268, "step": 14500 }, { "grad_norm": 0.2557812035083771, "learning_rate": 9.535597090785091e-05, "loss": 0.0225, "step": 14510 }, { "grad_norm": 0.3022112250328064, "learning_rate": 9.534726826616656e-05, "loss": 0.0192, "step": 14520 }, { "grad_norm": 0.29399046301841736, "learning_rate": 9.53385578758743e-05, "loss": 0.0357, "step": 14530 }, { "grad_norm": 0.2367454618215561, "learning_rate": 9.532983973846252e-05, "loss": 0.0203, "step": 14540 }, { "grad_norm": 0.2722536027431488, "learning_rate": 9.53211138554209e-05, "loss": 0.0285, "step": 14550 }, { "grad_norm": 0.28378522396087646, "learning_rate": 9.531238022824047e-05, "loss": 0.0266, "step": 14560 }, { "grad_norm": 0.29288867115974426, "learning_rate": 9.530363885841355e-05, "loss": 0.0217, "step": 14570 }, { "grad_norm": 0.4586062431335449, "learning_rate": 9.52948897474338e-05, "loss": 0.0308, "step": 14580 }, { "grad_norm": 0.26705387234687805, "learning_rate": 9.528613289679622e-05, "loss": 0.0302, "step": 14590 }, { "grad_norm": 0.4488847851753235, "learning_rate": 9.52773683079971e-05, "loss": 0.0239, "step": 14600 }, { "grad_norm": 0.2700198292732239, "learning_rate": 9.526859598253407e-05, "loss": 0.0251, "step": 14610 }, { "grad_norm": 0.3268066346645355, "learning_rate": 9.525981592190609e-05, "loss": 0.0214, "step": 14620 }, { "grad_norm": 0.3534747064113617, "learning_rate": 9.525102812761342e-05, "loss": 0.0209, "step": 14630 }, { "grad_norm": 0.28202807903289795, "learning_rate": 9.524223260115768e-05, "loss": 0.0177, "step": 14640 }, { "grad_norm": 0.2706359028816223, "learning_rate": 9.523342934404175e-05, "loss": 0.0206, "step": 14650 }, { "grad_norm": 0.32480382919311523, "learning_rate": 9.522461835776989e-05, "loss": 0.0206, "step": 14660 }, { "grad_norm": 0.5830492973327637, "learning_rate": 9.521579964384764e-05, "loss": 0.031, "step": 14670 }, { "grad_norm": 0.3401317596435547, "learning_rate": 9.52069732037819e-05, "loss": 0.0216, "step": 14680 }, { "grad_norm": 0.3991366922855377, "learning_rate": 9.519813903908083e-05, "loss": 0.0233, "step": 14690 }, { "grad_norm": 0.2812100648880005, "learning_rate": 9.5189297151254e-05, "loss": 0.0276, "step": 14700 }, { "grad_norm": 0.4208744168281555, "learning_rate": 9.518044754181218e-05, "loss": 0.0304, "step": 14710 }, { "grad_norm": 0.37054112553596497, "learning_rate": 9.51715902122676e-05, "loss": 0.0272, "step": 14720 }, { "grad_norm": 0.22007520496845245, "learning_rate": 9.516272516413368e-05, "loss": 0.0284, "step": 14730 }, { "grad_norm": 0.2845207750797272, "learning_rate": 9.515385239892525e-05, "loss": 0.03, "step": 14740 }, { "grad_norm": 0.2854002118110657, "learning_rate": 9.514497191815839e-05, "loss": 0.0229, "step": 14750 }, { "grad_norm": 0.2732314169406891, "learning_rate": 9.513608372335055e-05, "loss": 0.0211, "step": 14760 }, { "grad_norm": 0.2786048650741577, "learning_rate": 9.512718781602045e-05, "loss": 0.0244, "step": 14770 }, { "grad_norm": 0.35592979192733765, "learning_rate": 9.511828419768823e-05, "loss": 0.0303, "step": 14780 }, { "grad_norm": 0.26318761706352234, "learning_rate": 9.510937286987521e-05, "loss": 0.0195, "step": 14790 }, { "grad_norm": 0.310833215713501, "learning_rate": 9.510045383410408e-05, "loss": 0.0224, "step": 14800 }, { "grad_norm": 0.24398335814476013, "learning_rate": 9.509152709189892e-05, "loss": 0.0234, "step": 14810 }, { "grad_norm": 0.2066742479801178, "learning_rate": 9.508259264478504e-05, "loss": 0.0303, "step": 14820 }, { "grad_norm": 0.21204587817192078, "learning_rate": 9.507365049428909e-05, "loss": 0.032, "step": 14830 }, { "grad_norm": 0.21971803903579712, "learning_rate": 9.506470064193902e-05, "loss": 0.0364, "step": 14840 }, { "grad_norm": 0.17025791108608246, "learning_rate": 9.505574308926414e-05, "loss": 0.0356, "step": 14850 }, { "grad_norm": 0.23878014087677002, "learning_rate": 9.504677783779505e-05, "loss": 0.0215, "step": 14860 }, { "grad_norm": 0.3531917631626129, "learning_rate": 9.503780488906365e-05, "loss": 0.0251, "step": 14870 }, { "grad_norm": 0.3328186571598053, "learning_rate": 9.502882424460319e-05, "loss": 0.0255, "step": 14880 }, { "grad_norm": 0.2955200672149658, "learning_rate": 9.501983590594821e-05, "loss": 0.0243, "step": 14890 }, { "grad_norm": 0.35281580686569214, "learning_rate": 9.501083987463455e-05, "loss": 0.021, "step": 14900 }, { "grad_norm": 0.38492998480796814, "learning_rate": 9.500183615219942e-05, "loss": 0.0248, "step": 14910 }, { "grad_norm": 0.5416027307510376, "learning_rate": 9.49928247401813e-05, "loss": 0.0196, "step": 14920 }, { "grad_norm": 0.324672669172287, "learning_rate": 9.498380564011997e-05, "loss": 0.0198, "step": 14930 }, { "grad_norm": 0.2303190976381302, "learning_rate": 9.497477885355656e-05, "loss": 0.025, "step": 14940 }, { "grad_norm": 0.3585049510002136, "learning_rate": 9.496574438203353e-05, "loss": 0.0308, "step": 14950 }, { "grad_norm": 0.2249257117509842, "learning_rate": 9.495670222709459e-05, "loss": 0.0227, "step": 14960 }, { "grad_norm": 0.2820444703102112, "learning_rate": 9.494765239028483e-05, "loss": 0.0202, "step": 14970 }, { "grad_norm": 0.20778638124465942, "learning_rate": 9.493859487315057e-05, "loss": 0.0221, "step": 14980 }, { "grad_norm": 0.28960564732551575, "learning_rate": 9.492952967723953e-05, "loss": 0.0233, "step": 14990 }, { "grad_norm": 0.43466511368751526, "learning_rate": 9.492045680410068e-05, "loss": 0.0254, "step": 15000 }, { "grad_norm": 0.32852432131767273, "learning_rate": 9.491137625528436e-05, "loss": 0.0218, "step": 15010 }, { "grad_norm": 0.3870449960231781, "learning_rate": 9.490228803234215e-05, "loss": 0.0212, "step": 15020 }, { "grad_norm": 0.43505367636680603, "learning_rate": 9.489319213682701e-05, "loss": 0.024, "step": 15030 }, { "grad_norm": 0.30302122235298157, "learning_rate": 9.488408857029316e-05, "loss": 0.0303, "step": 15040 }, { "grad_norm": 0.269224613904953, "learning_rate": 9.487497733429616e-05, "loss": 0.0319, "step": 15050 }, { "grad_norm": 0.5278037190437317, "learning_rate": 9.486585843039286e-05, "loss": 0.0265, "step": 15060 }, { "grad_norm": 0.25871747732162476, "learning_rate": 9.485673186014143e-05, "loss": 0.0229, "step": 15070 }, { "grad_norm": 0.3643636703491211, "learning_rate": 9.484759762510137e-05, "loss": 0.0357, "step": 15080 }, { "grad_norm": 0.30000433325767517, "learning_rate": 9.483845572683346e-05, "loss": 0.0257, "step": 15090 }, { "grad_norm": 0.20812669396400452, "learning_rate": 9.48293061668998e-05, "loss": 0.0244, "step": 15100 }, { "grad_norm": 0.3539418876171112, "learning_rate": 9.48201489468638e-05, "loss": 0.0233, "step": 15110 }, { "grad_norm": 0.23382973670959473, "learning_rate": 9.481098406829016e-05, "loss": 0.0239, "step": 15120 }, { "grad_norm": 0.28167885541915894, "learning_rate": 9.480181153274495e-05, "loss": 0.0251, "step": 15130 }, { "grad_norm": 0.2919912040233612, "learning_rate": 9.479263134179548e-05, "loss": 0.0198, "step": 15140 }, { "grad_norm": 0.38288405537605286, "learning_rate": 9.478344349701039e-05, "loss": 0.0242, "step": 15150 }, { "grad_norm": 0.35793590545654297, "learning_rate": 9.477424799995964e-05, "loss": 0.0278, "step": 15160 }, { "grad_norm": 0.3429708182811737, "learning_rate": 9.476504485221448e-05, "loss": 0.0266, "step": 15170 }, { "grad_norm": 0.25597718358039856, "learning_rate": 9.475583405534748e-05, "loss": 0.0215, "step": 15180 }, { "grad_norm": 0.373812735080719, "learning_rate": 9.474661561093251e-05, "loss": 0.0234, "step": 15190 }, { "grad_norm": 0.306182324886322, "learning_rate": 9.473738952054478e-05, "loss": 0.0266, "step": 15200 }, { "grad_norm": 0.22109819948673248, "learning_rate": 9.472815578576073e-05, "loss": 0.0215, "step": 15210 }, { "grad_norm": 0.3684481680393219, "learning_rate": 9.471891440815817e-05, "loss": 0.0215, "step": 15220 }, { "grad_norm": 0.4334743022918701, "learning_rate": 9.470966538931621e-05, "loss": 0.0154, "step": 15230 }, { "grad_norm": 0.24577650427818298, "learning_rate": 9.470040873081525e-05, "loss": 0.0206, "step": 15240 }, { "grad_norm": 0.3626476526260376, "learning_rate": 9.469114443423698e-05, "loss": 0.0358, "step": 15250 }, { "grad_norm": 0.21584616601467133, "learning_rate": 9.468187250116445e-05, "loss": 0.024, "step": 15260 }, { "grad_norm": 0.20101168751716614, "learning_rate": 9.467259293318197e-05, "loss": 0.0181, "step": 15270 }, { "grad_norm": 0.2244011014699936, "learning_rate": 9.466330573187514e-05, "loss": 0.0242, "step": 15280 }, { "grad_norm": 0.2865752875804901, "learning_rate": 9.46540108988309e-05, "loss": 0.0339, "step": 15290 }, { "grad_norm": 0.5556091070175171, "learning_rate": 9.46447084356375e-05, "loss": 0.0284, "step": 15300 }, { "grad_norm": 0.24781322479248047, "learning_rate": 9.463539834388447e-05, "loss": 0.0297, "step": 15310 }, { "grad_norm": 0.22154077887535095, "learning_rate": 9.462608062516263e-05, "loss": 0.0175, "step": 15320 }, { "grad_norm": 0.32565367221832275, "learning_rate": 9.461675528106413e-05, "loss": 0.018, "step": 15330 }, { "grad_norm": 0.32823050022125244, "learning_rate": 9.460742231318244e-05, "loss": 0.0251, "step": 15340 }, { "grad_norm": 0.3198299705982208, "learning_rate": 9.459808172311229e-05, "loss": 0.0325, "step": 15350 }, { "grad_norm": 0.2260040044784546, "learning_rate": 9.458873351244972e-05, "loss": 0.0211, "step": 15360 }, { "grad_norm": 0.2504154443740845, "learning_rate": 9.457937768279211e-05, "loss": 0.024, "step": 15370 }, { "grad_norm": 0.3722923696041107, "learning_rate": 9.45700142357381e-05, "loss": 0.0254, "step": 15380 }, { "grad_norm": 0.23622547090053558, "learning_rate": 9.456064317288765e-05, "loss": 0.021, "step": 15390 }, { "grad_norm": 0.3701287806034088, "learning_rate": 9.455126449584201e-05, "loss": 0.0245, "step": 15400 }, { "grad_norm": 0.2735409438610077, "learning_rate": 9.454187820620375e-05, "loss": 0.0188, "step": 15410 }, { "grad_norm": 0.3040124177932739, "learning_rate": 9.453248430557673e-05, "loss": 0.0237, "step": 15420 }, { "grad_norm": 0.36125120520591736, "learning_rate": 9.452308279556611e-05, "loss": 0.0363, "step": 15430 }, { "grad_norm": 0.24317869544029236, "learning_rate": 9.451367367777835e-05, "loss": 0.0183, "step": 15440 }, { "grad_norm": 0.24431179463863373, "learning_rate": 9.450425695382122e-05, "loss": 0.0243, "step": 15450 }, { "grad_norm": 0.2905408442020416, "learning_rate": 9.449483262530375e-05, "loss": 0.024, "step": 15460 }, { "grad_norm": 0.5940244197845459, "learning_rate": 9.448540069383633e-05, "loss": 0.0262, "step": 15470 }, { "grad_norm": 0.2535865306854248, "learning_rate": 9.447596116103061e-05, "loss": 0.024, "step": 15480 }, { "grad_norm": 0.3494795560836792, "learning_rate": 9.446651402849955e-05, "loss": 0.0293, "step": 15490 }, { "grad_norm": 0.24991020560264587, "learning_rate": 9.44570592978574e-05, "loss": 0.0251, "step": 15500 }, { "grad_norm": 0.375363826751709, "learning_rate": 9.444759697071972e-05, "loss": 0.0253, "step": 15510 }, { "grad_norm": 0.27724742889404297, "learning_rate": 9.443812704870336e-05, "loss": 0.0169, "step": 15520 }, { "grad_norm": 0.2622544765472412, "learning_rate": 9.442864953342649e-05, "loss": 0.0238, "step": 15530 }, { "grad_norm": 0.296096533536911, "learning_rate": 9.441916442650852e-05, "loss": 0.0235, "step": 15540 }, { "grad_norm": 0.30336448550224304, "learning_rate": 9.440967172957023e-05, "loss": 0.0191, "step": 15550 }, { "grad_norm": 0.2639497220516205, "learning_rate": 9.440017144423364e-05, "loss": 0.0256, "step": 15560 }, { "grad_norm": 0.3580288589000702, "learning_rate": 9.439066357212209e-05, "loss": 0.0223, "step": 15570 }, { "grad_norm": 0.4067288339138031, "learning_rate": 9.438114811486022e-05, "loss": 0.0249, "step": 15580 }, { "grad_norm": 0.26019781827926636, "learning_rate": 9.4371625074074e-05, "loss": 0.0198, "step": 15590 }, { "grad_norm": 0.24837851524353027, "learning_rate": 9.436209445139059e-05, "loss": 0.0193, "step": 15600 }, { "grad_norm": 0.26641130447387695, "learning_rate": 9.435255624843855e-05, "loss": 0.0233, "step": 15610 }, { "grad_norm": 0.19642241299152374, "learning_rate": 9.43430104668477e-05, "loss": 0.0199, "step": 15620 }, { "grad_norm": 0.45064789056777954, "learning_rate": 9.433345710824914e-05, "loss": 0.0227, "step": 15630 }, { "grad_norm": 0.3247368335723877, "learning_rate": 9.432389617427529e-05, "loss": 0.0203, "step": 15640 }, { "grad_norm": 0.1899435669183731, "learning_rate": 9.431432766655984e-05, "loss": 0.0185, "step": 15650 }, { "grad_norm": 0.4916006922721863, "learning_rate": 9.430475158673778e-05, "loss": 0.0252, "step": 15660 }, { "grad_norm": 0.37497544288635254, "learning_rate": 9.429516793644542e-05, "loss": 0.0218, "step": 15670 }, { "grad_norm": 0.31889909505844116, "learning_rate": 9.428557671732034e-05, "loss": 0.0162, "step": 15680 }, { "grad_norm": 0.22378021478652954, "learning_rate": 9.42759779310014e-05, "loss": 0.0186, "step": 15690 }, { "grad_norm": 0.3129652440547943, "learning_rate": 9.426637157912879e-05, "loss": 0.0216, "step": 15700 }, { "grad_norm": 0.2895345985889435, "learning_rate": 9.425675766334397e-05, "loss": 0.0221, "step": 15710 }, { "grad_norm": 0.3386744558811188, "learning_rate": 9.424713618528968e-05, "loss": 0.0164, "step": 15720 }, { "grad_norm": 0.3345943093299866, "learning_rate": 9.423750714661e-05, "loss": 0.024, "step": 15730 }, { "grad_norm": 0.39087989926338196, "learning_rate": 9.422787054895022e-05, "loss": 0.0242, "step": 15740 }, { "grad_norm": 0.3087981939315796, "learning_rate": 9.4218226393957e-05, "loss": 0.0305, "step": 15750 }, { "grad_norm": 0.376688152551651, "learning_rate": 9.420857468327828e-05, "loss": 0.0241, "step": 15760 }, { "grad_norm": 0.44301140308380127, "learning_rate": 9.419891541856323e-05, "loss": 0.0277, "step": 15770 }, { "grad_norm": 0.34904050827026367, "learning_rate": 9.41892486014624e-05, "loss": 0.0245, "step": 15780 }, { "grad_norm": 0.34978529810905457, "learning_rate": 9.417957423362756e-05, "loss": 0.0303, "step": 15790 }, { "grad_norm": 0.3344935476779938, "learning_rate": 9.416989231671178e-05, "loss": 0.0264, "step": 15800 }, { "grad_norm": 0.40132850408554077, "learning_rate": 9.416020285236946e-05, "loss": 0.0221, "step": 15810 }, { "grad_norm": 0.31098997592926025, "learning_rate": 9.415050584225626e-05, "loss": 0.022, "step": 15820 }, { "grad_norm": 0.2457628846168518, "learning_rate": 9.414080128802914e-05, "loss": 0.0213, "step": 15830 }, { "grad_norm": 0.3999299705028534, "learning_rate": 9.413108919134632e-05, "loss": 0.0226, "step": 15840 }, { "grad_norm": 0.3848758041858673, "learning_rate": 9.412136955386734e-05, "loss": 0.0218, "step": 15850 }, { "grad_norm": 0.2913174331188202, "learning_rate": 9.411164237725303e-05, "loss": 0.0275, "step": 15860 }, { "grad_norm": 0.3016570210456848, "learning_rate": 9.41019076631655e-05, "loss": 0.0279, "step": 15870 }, { "grad_norm": 0.28602761030197144, "learning_rate": 9.409216541326815e-05, "loss": 0.0267, "step": 15880 }, { "grad_norm": 0.33079081773757935, "learning_rate": 9.408241562922564e-05, "loss": 0.0238, "step": 15890 }, { "grad_norm": 0.18850690126419067, "learning_rate": 9.407265831270395e-05, "loss": 0.019, "step": 15900 }, { "grad_norm": 0.3141925036907196, "learning_rate": 9.406289346537035e-05, "loss": 0.0227, "step": 15910 }, { "grad_norm": 0.5599552989006042, "learning_rate": 9.405312108889339e-05, "loss": 0.026, "step": 15920 }, { "grad_norm": 0.4378877282142639, "learning_rate": 9.404334118494288e-05, "loss": 0.0223, "step": 15930 }, { "grad_norm": 0.2886340618133545, "learning_rate": 9.403355375518995e-05, "loss": 0.0215, "step": 15940 }, { "grad_norm": 0.22202357649803162, "learning_rate": 9.4023758801307e-05, "loss": 0.0197, "step": 15950 }, { "grad_norm": 0.20377717912197113, "learning_rate": 9.401395632496774e-05, "loss": 0.0196, "step": 15960 }, { "grad_norm": 0.2958905100822449, "learning_rate": 9.400414632784711e-05, "loss": 0.0244, "step": 15970 }, { "grad_norm": 0.3680046498775482, "learning_rate": 9.39943288116214e-05, "loss": 0.0242, "step": 15980 }, { "grad_norm": 0.24100269377231598, "learning_rate": 9.398450377796815e-05, "loss": 0.0228, "step": 15990 }, { "grad_norm": 0.26605933904647827, "learning_rate": 9.397467122856616e-05, "loss": 0.0282, "step": 16000 }, { "grad_norm": 0.2743891775608063, "learning_rate": 9.396483116509558e-05, "loss": 0.0235, "step": 16010 }, { "grad_norm": 0.3129275143146515, "learning_rate": 9.39549835892378e-05, "loss": 0.0259, "step": 16020 }, { "grad_norm": 0.25562208890914917, "learning_rate": 9.39451285026755e-05, "loss": 0.0235, "step": 16030 }, { "grad_norm": 0.22785280644893646, "learning_rate": 9.393526590709262e-05, "loss": 0.0211, "step": 16040 }, { "grad_norm": 0.372177392244339, "learning_rate": 9.392539580417444e-05, "loss": 0.0206, "step": 16050 }, { "grad_norm": 0.2901802361011505, "learning_rate": 9.391551819560747e-05, "loss": 0.0237, "step": 16060 }, { "grad_norm": 0.29500216245651245, "learning_rate": 9.390563308307955e-05, "loss": 0.0203, "step": 16070 }, { "grad_norm": 0.256770521402359, "learning_rate": 9.389574046827974e-05, "loss": 0.02, "step": 16080 }, { "grad_norm": 0.3110209107398987, "learning_rate": 9.388584035289845e-05, "loss": 0.033, "step": 16090 }, { "grad_norm": 0.38705208897590637, "learning_rate": 9.387593273862732e-05, "loss": 0.0298, "step": 16100 }, { "grad_norm": 0.2887970209121704, "learning_rate": 9.386601762715929e-05, "loss": 0.024, "step": 16110 }, { "grad_norm": 0.2730705440044403, "learning_rate": 9.38560950201886e-05, "loss": 0.0266, "step": 16120 }, { "grad_norm": 0.44827088713645935, "learning_rate": 9.384616491941071e-05, "loss": 0.0254, "step": 16130 }, { "grad_norm": 0.2695498764514923, "learning_rate": 9.383622732652245e-05, "loss": 0.0203, "step": 16140 }, { "grad_norm": 0.21471652388572693, "learning_rate": 9.382628224322187e-05, "loss": 0.0282, "step": 16150 }, { "grad_norm": 0.20789922773838043, "learning_rate": 9.381632967120829e-05, "loss": 0.0199, "step": 16160 }, { "grad_norm": 0.4491157531738281, "learning_rate": 9.380636961218235e-05, "loss": 0.0212, "step": 16170 }, { "grad_norm": 0.21141260862350464, "learning_rate": 9.379640206784597e-05, "loss": 0.0156, "step": 16180 }, { "grad_norm": 0.31250834465026855, "learning_rate": 9.378642703990229e-05, "loss": 0.0177, "step": 16190 }, { "grad_norm": 0.21966156363487244, "learning_rate": 9.37764445300558e-05, "loss": 0.0221, "step": 16200 }, { "grad_norm": 0.2646046578884125, "learning_rate": 9.376645454001222e-05, "loss": 0.0287, "step": 16210 }, { "grad_norm": 0.30358636379241943, "learning_rate": 9.375645707147858e-05, "loss": 0.0199, "step": 16220 }, { "grad_norm": 0.22266437113285065, "learning_rate": 9.374645212616316e-05, "loss": 0.0285, "step": 16230 }, { "grad_norm": 0.3764084577560425, "learning_rate": 9.373643970577555e-05, "loss": 0.024, "step": 16240 }, { "grad_norm": 0.315202921628952, "learning_rate": 9.372641981202659e-05, "loss": 0.0223, "step": 16250 }, { "grad_norm": 0.5233691930770874, "learning_rate": 9.37163924466284e-05, "loss": 0.0344, "step": 16260 }, { "grad_norm": 0.31272467970848083, "learning_rate": 9.370635761129438e-05, "loss": 0.0221, "step": 16270 }, { "grad_norm": 0.23522011935710907, "learning_rate": 9.36963153077392e-05, "loss": 0.0219, "step": 16280 }, { "grad_norm": 0.22016197443008423, "learning_rate": 9.368626553767888e-05, "loss": 0.0221, "step": 16290 }, { "grad_norm": 0.1758708655834198, "learning_rate": 9.367620830283057e-05, "loss": 0.017, "step": 16300 }, { "grad_norm": 0.41084039211273193, "learning_rate": 9.366614360491281e-05, "loss": 0.0227, "step": 16310 }, { "grad_norm": 0.29721060395240784, "learning_rate": 9.365607144564539e-05, "loss": 0.0207, "step": 16320 }, { "grad_norm": 0.2751075327396393, "learning_rate": 9.364599182674934e-05, "loss": 0.0254, "step": 16330 }, { "grad_norm": 0.2469491809606552, "learning_rate": 9.3635904749947e-05, "loss": 0.0192, "step": 16340 }, { "grad_norm": 0.31959640979766846, "learning_rate": 9.362581021696202e-05, "loss": 0.0215, "step": 16350 }, { "grad_norm": 0.2719712257385254, "learning_rate": 9.361570822951921e-05, "loss": 0.025, "step": 16360 }, { "grad_norm": 0.25059035420417786, "learning_rate": 9.360559878934476e-05, "loss": 0.0225, "step": 16370 }, { "grad_norm": 0.30227935314178467, "learning_rate": 9.359548189816611e-05, "loss": 0.0208, "step": 16380 }, { "grad_norm": 0.27189478278160095, "learning_rate": 9.358535755771193e-05, "loss": 0.0284, "step": 16390 }, { "grad_norm": 0.24217656254768372, "learning_rate": 9.357522576971221e-05, "loss": 0.0218, "step": 16400 }, { "grad_norm": 0.2866979241371155, "learning_rate": 9.356508653589819e-05, "loss": 0.0245, "step": 16410 }, { "grad_norm": 0.34250280261039734, "learning_rate": 9.355493985800237e-05, "loss": 0.0232, "step": 16420 }, { "grad_norm": 0.28716525435447693, "learning_rate": 9.354478573775857e-05, "loss": 0.0217, "step": 16430 }, { "grad_norm": 0.39318153262138367, "learning_rate": 9.353462417690186e-05, "loss": 0.0223, "step": 16440 }, { "grad_norm": 0.34977248311042786, "learning_rate": 9.352445517716853e-05, "loss": 0.0203, "step": 16450 }, { "grad_norm": 0.2825795114040375, "learning_rate": 9.351427874029621e-05, "loss": 0.0206, "step": 16460 }, { "grad_norm": 0.2817252576351166, "learning_rate": 9.350409486802379e-05, "loss": 0.0191, "step": 16470 }, { "grad_norm": 0.32749584317207336, "learning_rate": 9.349390356209138e-05, "loss": 0.028, "step": 16480 }, { "grad_norm": 0.27639204263687134, "learning_rate": 9.348370482424042e-05, "loss": 0.0223, "step": 16490 }, { "grad_norm": 0.33925962448120117, "learning_rate": 9.347349865621357e-05, "loss": 0.0219, "step": 16500 }, { "grad_norm": 0.32002076506614685, "learning_rate": 9.346328505975481e-05, "loss": 0.023, "step": 16510 }, { "grad_norm": 0.23593029379844666, "learning_rate": 9.345306403660936e-05, "loss": 0.0226, "step": 16520 }, { "grad_norm": 0.3815995156764984, "learning_rate": 9.344283558852371e-05, "loss": 0.0209, "step": 16530 }, { "grad_norm": 0.3503400981426239, "learning_rate": 9.343259971724563e-05, "loss": 0.0188, "step": 16540 }, { "grad_norm": 0.3807649612426758, "learning_rate": 9.342235642452413e-05, "loss": 0.0266, "step": 16550 }, { "grad_norm": 0.28936800360679626, "learning_rate": 9.341210571210954e-05, "loss": 0.023, "step": 16560 }, { "grad_norm": 0.29847782850265503, "learning_rate": 9.340184758175338e-05, "loss": 0.0191, "step": 16570 }, { "grad_norm": 0.2949077785015106, "learning_rate": 9.339158203520854e-05, "loss": 0.0272, "step": 16580 }, { "grad_norm": 0.4637305736541748, "learning_rate": 9.338130907422908e-05, "loss": 0.0174, "step": 16590 }, { "grad_norm": 0.3117484748363495, "learning_rate": 9.337102870057037e-05, "loss": 0.0168, "step": 16600 }, { "grad_norm": 0.23664914071559906, "learning_rate": 9.336074091598907e-05, "loss": 0.0162, "step": 16610 }, { "grad_norm": 0.24119846522808075, "learning_rate": 9.335044572224306e-05, "loss": 0.025, "step": 16620 }, { "grad_norm": 0.2678243815898895, "learning_rate": 9.334014312109151e-05, "loss": 0.0173, "step": 16630 }, { "grad_norm": 0.41588473320007324, "learning_rate": 9.332983311429486e-05, "loss": 0.0259, "step": 16640 }, { "grad_norm": 0.2910511791706085, "learning_rate": 9.33195157036148e-05, "loss": 0.0237, "step": 16650 }, { "grad_norm": 0.24877576529979706, "learning_rate": 9.330919089081432e-05, "loss": 0.0216, "step": 16660 }, { "grad_norm": 0.23855385184288025, "learning_rate": 9.32988586776576e-05, "loss": 0.0357, "step": 16670 }, { "grad_norm": 0.20319271087646484, "learning_rate": 9.328851906591016e-05, "loss": 0.0227, "step": 16680 }, { "grad_norm": 0.28009384870529175, "learning_rate": 9.327817205733875e-05, "loss": 0.022, "step": 16690 }, { "grad_norm": 0.3369509279727936, "learning_rate": 9.326781765371142e-05, "loss": 0.0224, "step": 16700 }, { "grad_norm": 0.322878897190094, "learning_rate": 9.325745585679741e-05, "loss": 0.024, "step": 16710 }, { "grad_norm": 0.38612550497055054, "learning_rate": 9.32470866683673e-05, "loss": 0.0324, "step": 16720 }, { "grad_norm": 0.29078346490859985, "learning_rate": 9.323671009019288e-05, "loss": 0.0308, "step": 16730 }, { "grad_norm": 0.25616300106048584, "learning_rate": 9.322632612404725e-05, "loss": 0.0231, "step": 16740 }, { "grad_norm": 0.31412458419799805, "learning_rate": 9.321593477170471e-05, "loss": 0.0299, "step": 16750 }, { "grad_norm": 0.33858463168144226, "learning_rate": 9.320553603494088e-05, "loss": 0.0205, "step": 16760 }, { "grad_norm": 0.21430368721485138, "learning_rate": 9.319512991553261e-05, "loss": 0.0177, "step": 16770 }, { "grad_norm": 0.3916029632091522, "learning_rate": 9.318471641525803e-05, "loss": 0.0254, "step": 16780 }, { "grad_norm": 0.3370897173881531, "learning_rate": 9.317429553589652e-05, "loss": 0.0211, "step": 16790 }, { "grad_norm": 0.22586753964424133, "learning_rate": 9.316386727922873e-05, "loss": 0.0225, "step": 16800 }, { "grad_norm": 0.2950020432472229, "learning_rate": 9.315343164703656e-05, "loss": 0.0213, "step": 16810 }, { "grad_norm": 0.23533634841442108, "learning_rate": 9.314298864110316e-05, "loss": 0.0183, "step": 16820 }, { "grad_norm": 0.23514220118522644, "learning_rate": 9.313253826321295e-05, "loss": 0.0191, "step": 16830 }, { "grad_norm": 0.23761019110679626, "learning_rate": 9.312208051515165e-05, "loss": 0.0194, "step": 16840 }, { "grad_norm": 0.2705020606517792, "learning_rate": 9.311161539870618e-05, "loss": 0.0216, "step": 16850 }, { "grad_norm": 0.19979971647262573, "learning_rate": 9.310114291566474e-05, "loss": 0.0228, "step": 16860 }, { "grad_norm": 0.2750105857849121, "learning_rate": 9.309066306781679e-05, "loss": 0.0166, "step": 16870 }, { "grad_norm": 0.5594354271888733, "learning_rate": 9.308017585695306e-05, "loss": 0.0214, "step": 16880 }, { "grad_norm": 0.3117850422859192, "learning_rate": 9.306968128486552e-05, "loss": 0.0266, "step": 16890 }, { "grad_norm": 0.29863280057907104, "learning_rate": 9.30591793533474e-05, "loss": 0.0199, "step": 16900 }, { "grad_norm": 0.2920501232147217, "learning_rate": 9.304867006419321e-05, "loss": 0.0215, "step": 16910 }, { "grad_norm": 0.280240535736084, "learning_rate": 9.303815341919868e-05, "loss": 0.0198, "step": 16920 }, { "grad_norm": 0.2796175479888916, "learning_rate": 9.302762942016084e-05, "loss": 0.0282, "step": 16930 }, { "grad_norm": 0.5249584317207336, "learning_rate": 9.301709806887792e-05, "loss": 0.0219, "step": 16940 }, { "grad_norm": 0.40264594554901123, "learning_rate": 9.300655936714948e-05, "loss": 0.0386, "step": 16950 }, { "grad_norm": 0.2764070928096771, "learning_rate": 9.299601331677627e-05, "loss": 0.0291, "step": 16960 }, { "grad_norm": 0.2928406894207001, "learning_rate": 9.298545991956033e-05, "loss": 0.0287, "step": 16970 }, { "grad_norm": 0.22154463827610016, "learning_rate": 9.297489917730493e-05, "loss": 0.0253, "step": 16980 }, { "grad_norm": 0.49733966588974, "learning_rate": 9.296433109181464e-05, "loss": 0.0273, "step": 16990 }, { "grad_norm": 0.4432713985443115, "learning_rate": 9.295375566489523e-05, "loss": 0.0291, "step": 17000 }, { "grad_norm": 0.35778260231018066, "learning_rate": 9.294317289835379e-05, "loss": 0.0309, "step": 17010 }, { "grad_norm": 0.32930925488471985, "learning_rate": 9.293258279399859e-05, "loss": 0.0281, "step": 17020 }, { "grad_norm": 0.2586813271045685, "learning_rate": 9.292198535363919e-05, "loss": 0.0215, "step": 17030 }, { "grad_norm": 0.25767266750335693, "learning_rate": 9.291138057908641e-05, "loss": 0.022, "step": 17040 }, { "grad_norm": 0.2546219527721405, "learning_rate": 9.290076847215234e-05, "loss": 0.0199, "step": 17050 }, { "grad_norm": 0.5195832848548889, "learning_rate": 9.289014903465025e-05, "loss": 0.0323, "step": 17060 }, { "grad_norm": 0.2802042067050934, "learning_rate": 9.287952226839475e-05, "loss": 0.0195, "step": 17070 }, { "grad_norm": 0.3560789227485657, "learning_rate": 9.286888817520164e-05, "loss": 0.0213, "step": 17080 }, { "grad_norm": 0.2463877648115158, "learning_rate": 9.285824675688803e-05, "loss": 0.0206, "step": 17090 }, { "grad_norm": 0.38499489426612854, "learning_rate": 9.28475980152722e-05, "loss": 0.0238, "step": 17100 }, { "grad_norm": 0.2949081063270569, "learning_rate": 9.283694195217379e-05, "loss": 0.0232, "step": 17110 }, { "grad_norm": 0.30886197090148926, "learning_rate": 9.282627856941356e-05, "loss": 0.0197, "step": 17120 }, { "grad_norm": 0.31626588106155396, "learning_rate": 9.281560786881363e-05, "loss": 0.0206, "step": 17130 }, { "grad_norm": 0.2827546298503876, "learning_rate": 9.280492985219733e-05, "loss": 0.0184, "step": 17140 }, { "grad_norm": 0.18333573639392853, "learning_rate": 9.279424452138924e-05, "loss": 0.022, "step": 17150 }, { "grad_norm": 0.28785958886146545, "learning_rate": 9.278355187821517e-05, "loss": 0.022, "step": 17160 }, { "grad_norm": 0.265992134809494, "learning_rate": 9.277285192450224e-05, "loss": 0.0246, "step": 17170 }, { "grad_norm": 0.3700374364852905, "learning_rate": 9.276214466207875e-05, "loss": 0.0212, "step": 17180 }, { "grad_norm": 0.31767886877059937, "learning_rate": 9.275143009277427e-05, "loss": 0.0218, "step": 17190 }, { "grad_norm": 0.30861449241638184, "learning_rate": 9.274070821841964e-05, "loss": 0.0226, "step": 17200 }, { "grad_norm": 0.29810407757759094, "learning_rate": 9.272997904084696e-05, "loss": 0.0201, "step": 17210 }, { "grad_norm": 0.3003561198711395, "learning_rate": 9.271924256188951e-05, "loss": 0.0217, "step": 17220 }, { "grad_norm": 0.3029332458972931, "learning_rate": 9.270849878338189e-05, "loss": 0.0223, "step": 17230 }, { "grad_norm": 0.2659651041030884, "learning_rate": 9.269774770715991e-05, "loss": 0.0192, "step": 17240 }, { "grad_norm": 0.31458723545074463, "learning_rate": 9.268698933506061e-05, "loss": 0.0337, "step": 17250 }, { "grad_norm": 0.21939994394779205, "learning_rate": 9.267622366892235e-05, "loss": 0.0246, "step": 17260 }, { "grad_norm": 0.28294387459754944, "learning_rate": 9.266545071058465e-05, "loss": 0.021, "step": 17270 }, { "grad_norm": 0.31208381056785583, "learning_rate": 9.265467046188833e-05, "loss": 0.0196, "step": 17280 }, { "grad_norm": 0.24926722049713135, "learning_rate": 9.264388292467543e-05, "loss": 0.0216, "step": 17290 }, { "grad_norm": 0.333952933549881, "learning_rate": 9.263308810078926e-05, "loss": 0.0197, "step": 17300 }, { "grad_norm": 0.669254720211029, "learning_rate": 9.262228599207434e-05, "loss": 0.0205, "step": 17310 }, { "grad_norm": 0.2309129536151886, "learning_rate": 9.261147660037647e-05, "loss": 0.0178, "step": 17320 }, { "grad_norm": 0.23464888334274292, "learning_rate": 9.26006599275427e-05, "loss": 0.0207, "step": 17330 }, { "grad_norm": 0.20017355680465698, "learning_rate": 9.258983597542124e-05, "loss": 0.0226, "step": 17340 }, { "grad_norm": 0.299110472202301, "learning_rate": 9.257900474586167e-05, "loss": 0.0172, "step": 17350 }, { "grad_norm": 0.317192018032074, "learning_rate": 9.256816624071471e-05, "loss": 0.0283, "step": 17360 }, { "grad_norm": 0.3202327787876129, "learning_rate": 9.25573204618324e-05, "loss": 0.0203, "step": 17370 }, { "grad_norm": 0.18569566309452057, "learning_rate": 9.254646741106796e-05, "loss": 0.021, "step": 17380 }, { "grad_norm": 0.32494428753852844, "learning_rate": 9.253560709027589e-05, "loss": 0.0221, "step": 17390 }, { "grad_norm": 0.3854230046272278, "learning_rate": 9.252473950131192e-05, "loss": 0.0238, "step": 17400 }, { "grad_norm": 0.2533405125141144, "learning_rate": 9.251386464603302e-05, "loss": 0.0186, "step": 17410 }, { "grad_norm": 0.2848255932331085, "learning_rate": 9.250298252629741e-05, "loss": 0.025, "step": 17420 }, { "grad_norm": 0.2347521185874939, "learning_rate": 9.249209314396454e-05, "loss": 0.0198, "step": 17430 }, { "grad_norm": 0.36695390939712524, "learning_rate": 9.248119650089513e-05, "loss": 0.0243, "step": 17440 }, { "grad_norm": 0.25706732273101807, "learning_rate": 9.247029259895108e-05, "loss": 0.0225, "step": 17450 }, { "grad_norm": 0.2936131954193115, "learning_rate": 9.24593814399956e-05, "loss": 0.0199, "step": 17460 }, { "grad_norm": 0.3507643938064575, "learning_rate": 9.244846302589309e-05, "loss": 0.0184, "step": 17470 }, { "grad_norm": 0.2660659849643707, "learning_rate": 9.243753735850923e-05, "loss": 0.0225, "step": 17480 }, { "grad_norm": 0.25438690185546875, "learning_rate": 9.24266044397109e-05, "loss": 0.0209, "step": 17490 }, { "grad_norm": 0.21420782804489136, "learning_rate": 9.241566427136624e-05, "loss": 0.0175, "step": 17500 }, { "grad_norm": 0.3154640197753906, "learning_rate": 9.240471685534463e-05, "loss": 0.021, "step": 17510 }, { "grad_norm": 0.2241480052471161, "learning_rate": 9.239376219351667e-05, "loss": 0.0262, "step": 17520 }, { "grad_norm": 0.42467525601387024, "learning_rate": 9.238280028775425e-05, "loss": 0.0242, "step": 17530 }, { "grad_norm": 0.34355536103248596, "learning_rate": 9.237183113993041e-05, "loss": 0.0275, "step": 17540 }, { "grad_norm": 0.2886054217815399, "learning_rate": 9.236085475191952e-05, "loss": 0.0235, "step": 17550 }, { "grad_norm": 0.2616133689880371, "learning_rate": 9.234987112559709e-05, "loss": 0.0372, "step": 17560 }, { "grad_norm": 0.31680646538734436, "learning_rate": 9.233888026283999e-05, "loss": 0.0262, "step": 17570 }, { "grad_norm": 0.28798162937164307, "learning_rate": 9.232788216552619e-05, "loss": 0.0266, "step": 17580 }, { "grad_norm": 0.45188620686531067, "learning_rate": 9.231687683553502e-05, "loss": 0.026, "step": 17590 }, { "grad_norm": 0.22754788398742676, "learning_rate": 9.230586427474698e-05, "loss": 0.0294, "step": 17600 }, { "grad_norm": 0.3494235575199127, "learning_rate": 9.229484448504379e-05, "loss": 0.0294, "step": 17610 }, { "grad_norm": 0.32745495438575745, "learning_rate": 9.228381746830843e-05, "loss": 0.0223, "step": 17620 }, { "grad_norm": 0.26313599944114685, "learning_rate": 9.227278322642514e-05, "loss": 0.0199, "step": 17630 }, { "grad_norm": 0.3675142824649811, "learning_rate": 9.226174176127937e-05, "loss": 0.0251, "step": 17640 }, { "grad_norm": 0.22664876282215118, "learning_rate": 9.22506930747578e-05, "loss": 0.0254, "step": 17650 }, { "grad_norm": 0.4151473939418793, "learning_rate": 9.223963716874831e-05, "loss": 0.032, "step": 17660 }, { "grad_norm": 0.23020151257514954, "learning_rate": 9.222857404514012e-05, "loss": 0.0279, "step": 17670 }, { "grad_norm": 0.4085593521595001, "learning_rate": 9.221750370582355e-05, "loss": 0.0293, "step": 17680 }, { "grad_norm": 0.30133599042892456, "learning_rate": 9.220642615269028e-05, "loss": 0.0273, "step": 17690 }, { "grad_norm": 0.2304711490869522, "learning_rate": 9.219534138763311e-05, "loss": 0.0324, "step": 17700 }, { "grad_norm": 0.35102108120918274, "learning_rate": 9.218424941254613e-05, "loss": 0.0252, "step": 17710 }, { "grad_norm": 0.2529018521308899, "learning_rate": 9.217315022932468e-05, "loss": 0.0262, "step": 17720 }, { "grad_norm": 0.44861510396003723, "learning_rate": 9.216204383986528e-05, "loss": 0.0251, "step": 17730 }, { "grad_norm": 0.2664146423339844, "learning_rate": 9.215093024606574e-05, "loss": 0.0303, "step": 17740 }, { "grad_norm": 0.34633389115333557, "learning_rate": 9.213980944982506e-05, "loss": 0.0397, "step": 17750 }, { "grad_norm": 0.26788076758384705, "learning_rate": 9.212868145304346e-05, "loss": 0.0281, "step": 17760 }, { "grad_norm": 0.28623369336128235, "learning_rate": 9.211754625762241e-05, "loss": 0.0202, "step": 17770 }, { "grad_norm": 0.2690451145172119, "learning_rate": 9.210640386546463e-05, "loss": 0.0289, "step": 17780 }, { "grad_norm": 0.23524393141269684, "learning_rate": 9.209525427847405e-05, "loss": 0.0154, "step": 17790 }, { "grad_norm": 0.3385899066925049, "learning_rate": 9.208409749855583e-05, "loss": 0.0182, "step": 17800 }, { "grad_norm": 0.30945783853530884, "learning_rate": 9.207293352761633e-05, "loss": 0.0279, "step": 17810 }, { "grad_norm": 0.41189950704574585, "learning_rate": 9.206176236756319e-05, "loss": 0.02, "step": 17820 }, { "grad_norm": 0.20669864118099213, "learning_rate": 9.205058402030525e-05, "loss": 0.0222, "step": 17830 }, { "grad_norm": 0.2106238156557083, "learning_rate": 9.203939848775259e-05, "loss": 0.0183, "step": 17840 }, { "grad_norm": 0.22628748416900635, "learning_rate": 9.202820577181652e-05, "loss": 0.0181, "step": 17850 }, { "grad_norm": 0.385009229183197, "learning_rate": 9.201700587440953e-05, "loss": 0.0215, "step": 17860 }, { "grad_norm": 0.24141602218151093, "learning_rate": 9.200579879744544e-05, "loss": 0.023, "step": 17870 }, { "grad_norm": 0.3289232552051544, "learning_rate": 9.199458454283918e-05, "loss": 0.0215, "step": 17880 }, { "grad_norm": 1.1511528491973877, "learning_rate": 9.198336311250697e-05, "loss": 0.0202, "step": 17890 }, { "grad_norm": 0.379599392414093, "learning_rate": 9.197213450836626e-05, "loss": 0.0197, "step": 17900 }, { "grad_norm": 0.707097589969635, "learning_rate": 9.19608987323357e-05, "loss": 0.0288, "step": 17910 }, { "grad_norm": 0.27003705501556396, "learning_rate": 9.194965578633517e-05, "loss": 0.0198, "step": 17920 }, { "grad_norm": 0.23742875456809998, "learning_rate": 9.193840567228582e-05, "loss": 0.0243, "step": 17930 }, { "grad_norm": 0.32869842648506165, "learning_rate": 9.192714839210994e-05, "loss": 0.0221, "step": 17940 }, { "grad_norm": 0.278580904006958, "learning_rate": 9.19158839477311e-05, "loss": 0.0303, "step": 17950 }, { "grad_norm": 0.22290673851966858, "learning_rate": 9.190461234107411e-05, "loss": 0.0222, "step": 17960 }, { "grad_norm": 0.3861420750617981, "learning_rate": 9.189333357406496e-05, "loss": 0.0229, "step": 17970 }, { "grad_norm": 0.229803204536438, "learning_rate": 9.188204764863089e-05, "loss": 0.029, "step": 17980 }, { "grad_norm": 0.46872055530548096, "learning_rate": 9.187075456670033e-05, "loss": 0.0209, "step": 17990 }, { "grad_norm": 0.29535767436027527, "learning_rate": 9.1859454330203e-05, "loss": 0.0238, "step": 18000 }, { "grad_norm": 0.33310118317604065, "learning_rate": 9.18481469410698e-05, "loss": 0.03, "step": 18010 }, { "grad_norm": 0.2629270553588867, "learning_rate": 9.183683240123281e-05, "loss": 0.0223, "step": 18020 }, { "grad_norm": 0.30109521746635437, "learning_rate": 9.182551071262541e-05, "loss": 0.0241, "step": 18030 }, { "grad_norm": 0.19557221233844757, "learning_rate": 9.181418187718218e-05, "loss": 0.0223, "step": 18040 }, { "grad_norm": 0.45916756987571716, "learning_rate": 9.180284589683888e-05, "loss": 0.0249, "step": 18050 }, { "grad_norm": 0.34962520003318787, "learning_rate": 9.17915027735325e-05, "loss": 0.0198, "step": 18060 }, { "grad_norm": 0.2801700234413147, "learning_rate": 9.178015250920133e-05, "loss": 0.0225, "step": 18070 }, { "grad_norm": 0.31492629647254944, "learning_rate": 9.176879510578477e-05, "loss": 0.0216, "step": 18080 }, { "grad_norm": 0.3464377224445343, "learning_rate": 9.17574305652235e-05, "loss": 0.0208, "step": 18090 }, { "grad_norm": 0.385761022567749, "learning_rate": 9.174605888945942e-05, "loss": 0.0259, "step": 18100 }, { "grad_norm": 0.29493287205696106, "learning_rate": 9.173468008043564e-05, "loss": 0.0212, "step": 18110 }, { "grad_norm": 0.3724227249622345, "learning_rate": 9.172329414009648e-05, "loss": 0.027, "step": 18120 }, { "grad_norm": 0.32023361325263977, "learning_rate": 9.171190107038747e-05, "loss": 0.0221, "step": 18130 }, { "grad_norm": 0.24108126759529114, "learning_rate": 9.170050087325541e-05, "loss": 0.0196, "step": 18140 }, { "grad_norm": 0.25790199637413025, "learning_rate": 9.168909355064824e-05, "loss": 0.0225, "step": 18150 }, { "grad_norm": 0.2637295126914978, "learning_rate": 9.167767910451519e-05, "loss": 0.0216, "step": 18160 }, { "grad_norm": 0.2707362771034241, "learning_rate": 9.166625753680669e-05, "loss": 0.0234, "step": 18170 }, { "grad_norm": 0.2886747121810913, "learning_rate": 9.165482884947431e-05, "loss": 0.0184, "step": 18180 }, { "grad_norm": 0.2770044803619385, "learning_rate": 9.164339304447098e-05, "loss": 0.0246, "step": 18190 }, { "grad_norm": 0.29288816452026367, "learning_rate": 9.163195012375072e-05, "loss": 0.0157, "step": 18200 }, { "grad_norm": 0.3566961884498596, "learning_rate": 9.16205000892688e-05, "loss": 0.0243, "step": 18210 }, { "grad_norm": 0.369265615940094, "learning_rate": 9.160904294298175e-05, "loss": 0.0171, "step": 18220 }, { "grad_norm": 0.27055951952934265, "learning_rate": 9.159757868684727e-05, "loss": 0.0222, "step": 18230 }, { "grad_norm": 0.32912832498550415, "learning_rate": 9.15861073228243e-05, "loss": 0.0192, "step": 18240 }, { "grad_norm": 0.46373307704925537, "learning_rate": 9.157462885287296e-05, "loss": 0.0253, "step": 18250 }, { "grad_norm": 0.33104637265205383, "learning_rate": 9.156314327895461e-05, "loss": 0.0178, "step": 18260 }, { "grad_norm": 0.2112276703119278, "learning_rate": 9.155165060303185e-05, "loss": 0.0253, "step": 18270 }, { "grad_norm": 0.3035043478012085, "learning_rate": 9.154015082706841e-05, "loss": 0.0186, "step": 18280 }, { "grad_norm": 0.3966767489910126, "learning_rate": 9.152864395302936e-05, "loss": 0.027, "step": 18290 }, { "grad_norm": 0.3224141001701355, "learning_rate": 9.151712998288085e-05, "loss": 0.0224, "step": 18300 }, { "grad_norm": 0.2421838641166687, "learning_rate": 9.150560891859031e-05, "loss": 0.0237, "step": 18310 }, { "grad_norm": 0.29264035820961, "learning_rate": 9.14940807621264e-05, "loss": 0.0216, "step": 18320 }, { "grad_norm": 0.22511424124240875, "learning_rate": 9.148254551545894e-05, "loss": 0.0195, "step": 18330 }, { "grad_norm": 0.3064204454421997, "learning_rate": 9.147100318055901e-05, "loss": 0.0231, "step": 18340 }, { "grad_norm": 0.2737096846103668, "learning_rate": 9.145945375939888e-05, "loss": 0.019, "step": 18350 }, { "grad_norm": 0.2701784372329712, "learning_rate": 9.144789725395203e-05, "loss": 0.0208, "step": 18360 }, { "grad_norm": 0.35836485028266907, "learning_rate": 9.14363336661931e-05, "loss": 0.0215, "step": 18370 }, { "grad_norm": 0.37014660239219666, "learning_rate": 9.142476299809806e-05, "loss": 0.0246, "step": 18380 }, { "grad_norm": 0.2251604050397873, "learning_rate": 9.1413185251644e-05, "loss": 0.0182, "step": 18390 }, { "grad_norm": 0.27927812933921814, "learning_rate": 9.140160042880923e-05, "loss": 0.0237, "step": 18400 }, { "grad_norm": 0.316463828086853, "learning_rate": 9.139000853157327e-05, "loss": 0.0265, "step": 18410 }, { "grad_norm": 0.3287767171859741, "learning_rate": 9.137840956191688e-05, "loss": 0.0255, "step": 18420 }, { "grad_norm": 0.22087910771369934, "learning_rate": 9.136680352182199e-05, "loss": 0.018, "step": 18430 }, { "grad_norm": 0.2749236822128296, "learning_rate": 9.135519041327177e-05, "loss": 0.021, "step": 18440 }, { "grad_norm": 0.2720319926738739, "learning_rate": 9.134357023825058e-05, "loss": 0.0248, "step": 18450 }, { "grad_norm": 0.35608193278312683, "learning_rate": 9.133194299874398e-05, "loss": 0.0232, "step": 18460 }, { "grad_norm": 0.39379218220710754, "learning_rate": 9.132030869673876e-05, "loss": 0.0253, "step": 18470 }, { "grad_norm": 0.3310406506061554, "learning_rate": 9.130866733422288e-05, "loss": 0.0194, "step": 18480 }, { "grad_norm": 0.2478247731924057, "learning_rate": 9.129701891318556e-05, "loss": 0.0174, "step": 18490 }, { "grad_norm": 0.3100341558456421, "learning_rate": 9.128536343561718e-05, "loss": 0.028, "step": 18500 }, { "grad_norm": 0.2828509211540222, "learning_rate": 9.127370090350934e-05, "loss": 0.0271, "step": 18510 }, { "grad_norm": 0.35936877131462097, "learning_rate": 9.126203131885487e-05, "loss": 0.0317, "step": 18520 }, { "grad_norm": 0.3250263035297394, "learning_rate": 9.125035468364775e-05, "loss": 0.0281, "step": 18530 }, { "grad_norm": 0.2535099685192108, "learning_rate": 9.123867099988322e-05, "loss": 0.0192, "step": 18540 }, { "grad_norm": 0.24572333693504333, "learning_rate": 9.122698026955769e-05, "loss": 0.0215, "step": 18550 }, { "grad_norm": 0.16233623027801514, "learning_rate": 9.12152824946688e-05, "loss": 0.03, "step": 18560 }, { "grad_norm": 0.4071321189403534, "learning_rate": 9.120357767721538e-05, "loss": 0.0241, "step": 18570 }, { "grad_norm": 0.35766080021858215, "learning_rate": 9.119186581919745e-05, "loss": 0.0225, "step": 18580 }, { "grad_norm": 0.3625970780849457, "learning_rate": 9.118014692261624e-05, "loss": 0.0244, "step": 18590 }, { "grad_norm": 0.43096184730529785, "learning_rate": 9.116842098947422e-05, "loss": 0.0376, "step": 18600 }, { "grad_norm": 0.3904597759246826, "learning_rate": 9.115668802177499e-05, "loss": 0.0214, "step": 18610 }, { "grad_norm": 0.20529530942440033, "learning_rate": 9.114494802152342e-05, "loss": 0.0275, "step": 18620 }, { "grad_norm": 0.4667549133300781, "learning_rate": 9.113320099072555e-05, "loss": 0.0445, "step": 18630 }, { "grad_norm": 0.2885618507862091, "learning_rate": 9.112144693138864e-05, "loss": 0.0235, "step": 18640 }, { "grad_norm": 0.3121265172958374, "learning_rate": 9.110968584552111e-05, "loss": 0.0204, "step": 18650 }, { "grad_norm": 0.22478875517845154, "learning_rate": 9.109791773513264e-05, "loss": 0.0226, "step": 18660 }, { "grad_norm": 0.3616558909416199, "learning_rate": 9.108614260223403e-05, "loss": 0.0261, "step": 18670 }, { "grad_norm": 0.23383969068527222, "learning_rate": 9.107436044883738e-05, "loss": 0.0214, "step": 18680 }, { "grad_norm": 0.22809040546417236, "learning_rate": 9.10625712769559e-05, "loss": 0.0193, "step": 18690 }, { "grad_norm": 0.20072880387306213, "learning_rate": 9.105077508860406e-05, "loss": 0.017, "step": 18700 }, { "grad_norm": 0.2588958442211151, "learning_rate": 9.103897188579751e-05, "loss": 0.0204, "step": 18710 }, { "grad_norm": 0.23247916996479034, "learning_rate": 9.102716167055308e-05, "loss": 0.0158, "step": 18720 }, { "grad_norm": 0.32007983326911926, "learning_rate": 9.10153444448888e-05, "loss": 0.0415, "step": 18730 }, { "grad_norm": 0.3141101598739624, "learning_rate": 9.100352021082393e-05, "loss": 0.016, "step": 18740 }, { "grad_norm": 0.3524780571460724, "learning_rate": 9.099168897037891e-05, "loss": 0.0194, "step": 18750 }, { "grad_norm": 0.2749752998352051, "learning_rate": 9.097985072557538e-05, "loss": 0.0181, "step": 18760 }, { "grad_norm": 0.2637425661087036, "learning_rate": 9.096800547843615e-05, "loss": 0.0371, "step": 18770 }, { "grad_norm": 0.3606577515602112, "learning_rate": 9.095615323098526e-05, "loss": 0.0288, "step": 18780 }, { "grad_norm": 0.2993919849395752, "learning_rate": 9.094429398524795e-05, "loss": 0.0186, "step": 18790 }, { "grad_norm": 0.2748119533061981, "learning_rate": 9.093242774325061e-05, "loss": 0.0227, "step": 18800 }, { "grad_norm": 0.2491217702627182, "learning_rate": 9.092055450702088e-05, "loss": 0.0224, "step": 18810 }, { "grad_norm": 0.34897318482398987, "learning_rate": 9.090867427858756e-05, "loss": 0.0252, "step": 18820 }, { "grad_norm": 0.30906012654304504, "learning_rate": 9.089678705998066e-05, "loss": 0.023, "step": 18830 }, { "grad_norm": 0.25951361656188965, "learning_rate": 9.088489285323139e-05, "loss": 0.0205, "step": 18840 }, { "grad_norm": 0.26203951239585876, "learning_rate": 9.087299166037212e-05, "loss": 0.022, "step": 18850 }, { "grad_norm": 0.3208867311477661, "learning_rate": 9.086108348343647e-05, "loss": 0.0234, "step": 18860 }, { "grad_norm": 0.5511046051979065, "learning_rate": 9.08491683244592e-05, "loss": 0.0287, "step": 18870 }, { "grad_norm": 0.2785114645957947, "learning_rate": 9.08372461854763e-05, "loss": 0.0226, "step": 18880 }, { "grad_norm": 0.32241612672805786, "learning_rate": 9.082531706852492e-05, "loss": 0.0225, "step": 18890 }, { "grad_norm": 0.24880684912204742, "learning_rate": 9.081338097564342e-05, "loss": 0.0263, "step": 18900 }, { "grad_norm": 0.20859064161777496, "learning_rate": 9.080143790887137e-05, "loss": 0.0181, "step": 18910 }, { "grad_norm": 0.3531864881515503, "learning_rate": 9.07894878702495e-05, "loss": 0.0228, "step": 18920 }, { "grad_norm": 0.37518996000289917, "learning_rate": 9.077753086181974e-05, "loss": 0.0275, "step": 18930 }, { "grad_norm": 0.31789857149124146, "learning_rate": 9.076556688562524e-05, "loss": 0.0187, "step": 18940 }, { "grad_norm": 0.19195973873138428, "learning_rate": 9.075359594371029e-05, "loss": 0.0204, "step": 18950 }, { "grad_norm": 0.3378305435180664, "learning_rate": 9.07416180381204e-05, "loss": 0.021, "step": 18960 }, { "grad_norm": 0.2489228993654251, "learning_rate": 9.072963317090228e-05, "loss": 0.0243, "step": 18970 }, { "grad_norm": 0.22002947330474854, "learning_rate": 9.071764134410382e-05, "loss": 0.0206, "step": 18980 }, { "grad_norm": 0.28526201844215393, "learning_rate": 9.070564255977407e-05, "loss": 0.0189, "step": 18990 }, { "grad_norm": 0.28785017132759094, "learning_rate": 9.06936368199633e-05, "loss": 0.0223, "step": 19000 }, { "grad_norm": 0.276728093624115, "learning_rate": 9.0681624126723e-05, "loss": 0.0249, "step": 19010 }, { "grad_norm": 0.34936973452568054, "learning_rate": 9.066960448210576e-05, "loss": 0.0228, "step": 19020 }, { "grad_norm": 0.24068906903266907, "learning_rate": 9.065757788816543e-05, "loss": 0.016, "step": 19030 }, { "grad_norm": 0.349953830242157, "learning_rate": 9.064554434695705e-05, "loss": 0.024, "step": 19040 }, { "grad_norm": 0.20583190023899078, "learning_rate": 9.063350386053677e-05, "loss": 0.0243, "step": 19050 }, { "grad_norm": 0.3219214379787445, "learning_rate": 9.062145643096202e-05, "loss": 0.0229, "step": 19060 }, { "grad_norm": 0.2724219262599945, "learning_rate": 9.060940206029136e-05, "loss": 0.0258, "step": 19070 }, { "grad_norm": 0.20835107564926147, "learning_rate": 9.059734075058457e-05, "loss": 0.019, "step": 19080 }, { "grad_norm": 0.21628671884536743, "learning_rate": 9.058527250390257e-05, "loss": 0.0247, "step": 19090 }, { "grad_norm": 0.2961932420730591, "learning_rate": 9.057319732230752e-05, "loss": 0.0281, "step": 19100 }, { "grad_norm": 0.21793560683727264, "learning_rate": 9.056111520786273e-05, "loss": 0.022, "step": 19110 }, { "grad_norm": 0.2504262328147888, "learning_rate": 9.054902616263268e-05, "loss": 0.0247, "step": 19120 }, { "grad_norm": 0.2257266789674759, "learning_rate": 9.05369301886831e-05, "loss": 0.0284, "step": 19130 }, { "grad_norm": 0.3216484487056732, "learning_rate": 9.052482728808083e-05, "loss": 0.0212, "step": 19140 }, { "grad_norm": 0.19675633311271667, "learning_rate": 9.051271746289391e-05, "loss": 0.0197, "step": 19150 }, { "grad_norm": 0.26878485083580017, "learning_rate": 9.050060071519162e-05, "loss": 0.0278, "step": 19160 }, { "grad_norm": 0.275544673204422, "learning_rate": 9.048847704704437e-05, "loss": 0.0219, "step": 19170 }, { "grad_norm": 0.30634739995002747, "learning_rate": 9.047634646052376e-05, "loss": 0.0256, "step": 19180 }, { "grad_norm": 0.2744450867176056, "learning_rate": 9.046420895770256e-05, "loss": 0.0369, "step": 19190 }, { "grad_norm": 0.4155887961387634, "learning_rate": 9.045206454065473e-05, "loss": 0.0217, "step": 19200 }, { "grad_norm": 0.42042964696884155, "learning_rate": 9.043991321145546e-05, "loss": 0.0214, "step": 19210 }, { "grad_norm": 0.35197365283966064, "learning_rate": 9.042775497218105e-05, "loss": 0.0168, "step": 19220 }, { "grad_norm": 0.19169758260250092, "learning_rate": 9.041558982490901e-05, "loss": 0.0203, "step": 19230 }, { "grad_norm": 0.32044675946235657, "learning_rate": 9.040341777171805e-05, "loss": 0.0205, "step": 19240 }, { "grad_norm": 0.2803666293621063, "learning_rate": 9.039123881468802e-05, "loss": 0.0225, "step": 19250 }, { "grad_norm": 0.21390147507190704, "learning_rate": 9.037905295589998e-05, "loss": 0.0175, "step": 19260 }, { "grad_norm": 0.24545705318450928, "learning_rate": 9.036686019743617e-05, "loss": 0.0158, "step": 19270 }, { "grad_norm": 0.302874356508255, "learning_rate": 9.035466054137997e-05, "loss": 0.0235, "step": 19280 }, { "grad_norm": 0.3344654440879822, "learning_rate": 9.0342453989816e-05, "loss": 0.0292, "step": 19290 }, { "grad_norm": 0.24884511530399323, "learning_rate": 9.033024054483e-05, "loss": 0.0232, "step": 19300 }, { "grad_norm": 0.4136523902416229, "learning_rate": 9.031802020850894e-05, "loss": 0.0193, "step": 19310 }, { "grad_norm": 0.3695865273475647, "learning_rate": 9.030579298294092e-05, "loss": 0.0282, "step": 19320 }, { "grad_norm": 0.29444006085395813, "learning_rate": 9.029355887021524e-05, "loss": 0.0324, "step": 19330 }, { "grad_norm": 0.4116823077201843, "learning_rate": 9.028131787242238e-05, "loss": 0.0401, "step": 19340 }, { "grad_norm": 0.23359936475753784, "learning_rate": 9.026906999165399e-05, "loss": 0.0324, "step": 19350 }, { "grad_norm": 0.2704618275165558, "learning_rate": 9.025681523000291e-05, "loss": 0.0196, "step": 19360 }, { "grad_norm": 0.3112216293811798, "learning_rate": 9.024455358956315e-05, "loss": 0.0231, "step": 19370 }, { "grad_norm": 0.30699074268341064, "learning_rate": 9.023228507242984e-05, "loss": 0.0312, "step": 19380 }, { "grad_norm": 0.22946177423000336, "learning_rate": 9.022000968069937e-05, "loss": 0.0234, "step": 19390 }, { "grad_norm": 0.22974485158920288, "learning_rate": 9.020772741646928e-05, "loss": 0.0202, "step": 19400 }, { "grad_norm": 0.3091638386249542, "learning_rate": 9.019543828183826e-05, "loss": 0.0188, "step": 19410 }, { "grad_norm": 0.2210260033607483, "learning_rate": 9.018314227890616e-05, "loss": 0.0189, "step": 19420 }, { "grad_norm": 0.3828868269920349, "learning_rate": 9.017083940977408e-05, "loss": 0.0369, "step": 19430 }, { "grad_norm": 0.26458093523979187, "learning_rate": 9.015852967654422e-05, "loss": 0.0198, "step": 19440 }, { "grad_norm": 0.5617436170578003, "learning_rate": 9.014621308131996e-05, "loss": 0.0248, "step": 19450 }, { "grad_norm": 0.21639616787433624, "learning_rate": 9.01338896262059e-05, "loss": 0.021, "step": 19460 }, { "grad_norm": 0.2751805782318115, "learning_rate": 9.012155931330777e-05, "loss": 0.0233, "step": 19470 }, { "grad_norm": 0.22752758860588074, "learning_rate": 9.010922214473246e-05, "loss": 0.0188, "step": 19480 }, { "grad_norm": 0.25460657477378845, "learning_rate": 9.009687812258808e-05, "loss": 0.0272, "step": 19490 }, { "grad_norm": 0.27391666173934937, "learning_rate": 9.00845272489839e-05, "loss": 0.0193, "step": 19500 }, { "grad_norm": 0.2616334557533264, "learning_rate": 9.007216952603031e-05, "loss": 0.027, "step": 19510 }, { "grad_norm": 0.3255758285522461, "learning_rate": 9.005980495583894e-05, "loss": 0.0154, "step": 19520 }, { "grad_norm": 0.25346142053604126, "learning_rate": 9.004743354052252e-05, "loss": 0.0207, "step": 19530 }, { "grad_norm": 0.26397615671157837, "learning_rate": 9.003505528219503e-05, "loss": 0.0194, "step": 19540 }, { "grad_norm": 0.22185653448104858, "learning_rate": 9.002267018297154e-05, "loss": 0.0244, "step": 19550 }, { "grad_norm": 0.22450658679008484, "learning_rate": 9.001027824496834e-05, "loss": 0.0288, "step": 19560 }, { "grad_norm": 0.2200223058462143, "learning_rate": 8.999787947030287e-05, "loss": 0.0211, "step": 19570 }, { "grad_norm": 0.25147050619125366, "learning_rate": 8.998547386109376e-05, "loss": 0.0275, "step": 19580 }, { "grad_norm": 0.283186674118042, "learning_rate": 8.997306141946073e-05, "loss": 0.025, "step": 19590 }, { "grad_norm": 0.3043783903121948, "learning_rate": 8.996064214752481e-05, "loss": 0.0294, "step": 19600 }, { "grad_norm": 0.33012786507606506, "learning_rate": 8.994821604740806e-05, "loss": 0.0251, "step": 19610 }, { "grad_norm": 0.3029673397541046, "learning_rate": 8.993578312123377e-05, "loss": 0.0236, "step": 19620 }, { "grad_norm": 0.2751491367816925, "learning_rate": 8.992334337112639e-05, "loss": 0.0201, "step": 19630 }, { "grad_norm": 0.22664937376976013, "learning_rate": 8.991089679921154e-05, "loss": 0.0173, "step": 19640 }, { "grad_norm": 0.2766452133655548, "learning_rate": 8.989844340761599e-05, "loss": 0.0222, "step": 19650 }, { "grad_norm": 0.24953271448612213, "learning_rate": 8.988598319846768e-05, "loss": 0.0213, "step": 19660 }, { "grad_norm": 0.33756786584854126, "learning_rate": 8.987351617389574e-05, "loss": 0.0244, "step": 19670 }, { "grad_norm": 0.2126266062259674, "learning_rate": 8.98610423360304e-05, "loss": 0.0218, "step": 19680 }, { "grad_norm": 0.22382375597953796, "learning_rate": 8.984856168700317e-05, "loss": 0.0221, "step": 19690 }, { "grad_norm": 0.21334001421928406, "learning_rate": 8.983607422894658e-05, "loss": 0.021, "step": 19700 }, { "grad_norm": 0.16462919116020203, "learning_rate": 8.982357996399442e-05, "loss": 0.0227, "step": 19710 }, { "grad_norm": 0.29305794835090637, "learning_rate": 8.981107889428164e-05, "loss": 0.0188, "step": 19720 }, { "grad_norm": 0.23747536540031433, "learning_rate": 8.979857102194428e-05, "loss": 0.0173, "step": 19730 }, { "grad_norm": 0.25284087657928467, "learning_rate": 8.978605634911968e-05, "loss": 0.019, "step": 19740 }, { "grad_norm": 0.30594170093536377, "learning_rate": 8.977353487794616e-05, "loss": 0.0175, "step": 19750 }, { "grad_norm": 0.33026477694511414, "learning_rate": 8.976100661056334e-05, "loss": 0.0264, "step": 19760 }, { "grad_norm": 0.2552105486392975, "learning_rate": 8.974847154911197e-05, "loss": 0.0188, "step": 19770 }, { "grad_norm": 0.27273720502853394, "learning_rate": 8.973592969573393e-05, "loss": 0.0262, "step": 19780 }, { "grad_norm": 0.2752652168273926, "learning_rate": 8.972338105257228e-05, "loss": 0.0264, "step": 19790 }, { "grad_norm": 0.29948729276657104, "learning_rate": 8.971082562177125e-05, "loss": 0.0191, "step": 19800 }, { "grad_norm": 0.27980369329452515, "learning_rate": 8.96982634054762e-05, "loss": 0.0202, "step": 19810 }, { "grad_norm": 0.2207041233778, "learning_rate": 8.96856944058337e-05, "loss": 0.0202, "step": 19820 }, { "grad_norm": 0.41226136684417725, "learning_rate": 8.967311862499144e-05, "loss": 0.0227, "step": 19830 }, { "grad_norm": 0.3200736343860626, "learning_rate": 8.966053606509825e-05, "loss": 0.0191, "step": 19840 }, { "grad_norm": 0.30821874737739563, "learning_rate": 8.964794672830417e-05, "loss": 0.0302, "step": 19850 }, { "grad_norm": 0.29787707328796387, "learning_rate": 8.963535061676038e-05, "loss": 0.0263, "step": 19860 }, { "grad_norm": 0.33439600467681885, "learning_rate": 8.962274773261918e-05, "loss": 0.0246, "step": 19870 }, { "grad_norm": 0.24712873995304108, "learning_rate": 8.961013807803409e-05, "loss": 0.0252, "step": 19880 }, { "grad_norm": 0.2162124514579773, "learning_rate": 8.959752165515973e-05, "loss": 0.0238, "step": 19890 }, { "grad_norm": 0.41523733735084534, "learning_rate": 8.958489846615193e-05, "loss": 0.0378, "step": 19900 }, { "grad_norm": 0.3843728005886078, "learning_rate": 8.957226851316762e-05, "loss": 0.0251, "step": 19910 }, { "grad_norm": 0.2694084346294403, "learning_rate": 8.955963179836493e-05, "loss": 0.0209, "step": 19920 }, { "grad_norm": 0.21042931079864502, "learning_rate": 8.954698832390312e-05, "loss": 0.0238, "step": 19930 }, { "grad_norm": 0.29124072194099426, "learning_rate": 8.953433809194263e-05, "loss": 0.025, "step": 19940 }, { "grad_norm": 0.297915518283844, "learning_rate": 8.9521681104645e-05, "loss": 0.0225, "step": 19950 }, { "grad_norm": 0.3799668252468109, "learning_rate": 8.9509017364173e-05, "loss": 0.0251, "step": 19960 }, { "grad_norm": 0.2670363783836365, "learning_rate": 8.949634687269052e-05, "loss": 0.0211, "step": 19970 }, { "grad_norm": 0.3391839563846588, "learning_rate": 8.948366963236259e-05, "loss": 0.0247, "step": 19980 }, { "grad_norm": 0.592803418636322, "learning_rate": 8.947098564535538e-05, "loss": 0.0266, "step": 19990 }, { "grad_norm": 0.3857142925262451, "learning_rate": 8.945829491383627e-05, "loss": 0.0203, "step": 20000 }, { "grad_norm": 0.4013942778110504, "learning_rate": 8.944559743997374e-05, "loss": 0.0234, "step": 20010 }, { "grad_norm": 0.33015695214271545, "learning_rate": 8.943289322593746e-05, "loss": 0.0228, "step": 20020 }, { "grad_norm": 0.2647242844104767, "learning_rate": 8.942018227389821e-05, "loss": 0.0167, "step": 20030 }, { "grad_norm": 0.38892197608947754, "learning_rate": 8.940746458602795e-05, "loss": 0.0357, "step": 20040 }, { "grad_norm": 0.3510943055152893, "learning_rate": 8.939474016449979e-05, "loss": 0.0256, "step": 20050 }, { "grad_norm": 0.2203434258699417, "learning_rate": 8.938200901148799e-05, "loss": 0.0185, "step": 20060 }, { "grad_norm": 0.3679850101470947, "learning_rate": 8.936927112916795e-05, "loss": 0.0263, "step": 20070 }, { "grad_norm": 0.27088651061058044, "learning_rate": 8.935652651971622e-05, "loss": 0.0216, "step": 20080 }, { "grad_norm": 0.2596835196018219, "learning_rate": 8.934377518531052e-05, "loss": 0.0258, "step": 20090 }, { "grad_norm": 0.35824379324913025, "learning_rate": 8.933101712812967e-05, "loss": 0.0276, "step": 20100 }, { "grad_norm": 0.26733121275901794, "learning_rate": 8.931825235035374e-05, "loss": 0.0208, "step": 20110 }, { "grad_norm": 0.34949445724487305, "learning_rate": 8.930548085416382e-05, "loss": 0.0232, "step": 20120 }, { "grad_norm": 0.3497052490711212, "learning_rate": 8.92927026417422e-05, "loss": 0.0231, "step": 20130 }, { "grad_norm": 0.4183312654495239, "learning_rate": 8.92799177152724e-05, "loss": 0.0204, "step": 20140 }, { "grad_norm": 0.17804402112960815, "learning_rate": 8.926712607693895e-05, "loss": 0.0205, "step": 20150 }, { "grad_norm": 0.20498326420783997, "learning_rate": 8.925432772892762e-05, "loss": 0.0269, "step": 20160 }, { "grad_norm": 0.27863141894340515, "learning_rate": 8.924152267342529e-05, "loss": 0.0285, "step": 20170 }, { "grad_norm": 0.2969316840171814, "learning_rate": 8.922871091261998e-05, "loss": 0.0297, "step": 20180 }, { "grad_norm": 0.3434338867664337, "learning_rate": 8.92158924487009e-05, "loss": 0.0198, "step": 20190 }, { "grad_norm": 0.3117678165435791, "learning_rate": 8.920306728385834e-05, "loss": 0.0194, "step": 20200 }, { "grad_norm": 0.26721999049186707, "learning_rate": 8.919023542028379e-05, "loss": 0.0178, "step": 20210 }, { "grad_norm": 0.3028925657272339, "learning_rate": 8.917739686016988e-05, "loss": 0.0184, "step": 20220 }, { "grad_norm": 0.27210283279418945, "learning_rate": 8.916455160571033e-05, "loss": 0.0265, "step": 20230 }, { "grad_norm": 0.3024541437625885, "learning_rate": 8.915169965910008e-05, "loss": 0.0196, "step": 20240 }, { "grad_norm": 0.25653377175331116, "learning_rate": 8.913884102253514e-05, "loss": 0.0259, "step": 20250 }, { "grad_norm": 0.1904931664466858, "learning_rate": 8.912597569821273e-05, "loss": 0.0171, "step": 20260 }, { "grad_norm": 0.27236080169677734, "learning_rate": 8.911310368833118e-05, "loss": 0.0197, "step": 20270 }, { "grad_norm": 0.3149193525314331, "learning_rate": 8.910022499508994e-05, "loss": 0.0158, "step": 20280 }, { "grad_norm": 0.2385721355676651, "learning_rate": 8.908733962068965e-05, "loss": 0.0245, "step": 20290 }, { "grad_norm": 0.16787073016166687, "learning_rate": 8.907444756733207e-05, "loss": 0.0191, "step": 20300 }, { "grad_norm": 0.2526399493217468, "learning_rate": 8.906154883722006e-05, "loss": 0.0244, "step": 20310 }, { "grad_norm": 0.17215198278427124, "learning_rate": 8.904864343255773e-05, "loss": 0.0194, "step": 20320 }, { "grad_norm": 0.27811330556869507, "learning_rate": 8.90357313555502e-05, "loss": 0.0216, "step": 20330 }, { "grad_norm": 0.38823726773262024, "learning_rate": 8.90228126084038e-05, "loss": 0.0262, "step": 20340 }, { "grad_norm": 0.38852590322494507, "learning_rate": 8.900988719332601e-05, "loss": 0.0243, "step": 20350 }, { "grad_norm": 0.29927685856819153, "learning_rate": 8.899695511252542e-05, "loss": 0.0212, "step": 20360 }, { "grad_norm": 0.34608250856399536, "learning_rate": 8.898401636821176e-05, "loss": 0.0286, "step": 20370 }, { "grad_norm": 0.27416035532951355, "learning_rate": 8.897107096259593e-05, "loss": 0.0319, "step": 20380 }, { "grad_norm": 0.2835012972354889, "learning_rate": 8.895811889788994e-05, "loss": 0.0199, "step": 20390 }, { "grad_norm": 0.17059899866580963, "learning_rate": 8.894516017630692e-05, "loss": 0.0168, "step": 20400 }, { "grad_norm": 0.3718080520629883, "learning_rate": 8.893219480006118e-05, "loss": 0.0178, "step": 20410 }, { "grad_norm": 0.23298372328281403, "learning_rate": 8.891922277136817e-05, "loss": 0.0184, "step": 20420 }, { "grad_norm": 0.31947794556617737, "learning_rate": 8.890624409244441e-05, "loss": 0.019, "step": 20430 }, { "grad_norm": 0.3371874690055847, "learning_rate": 8.889325876550763e-05, "loss": 0.0232, "step": 20440 }, { "grad_norm": 0.21985116600990295, "learning_rate": 8.888026679277666e-05, "loss": 0.0248, "step": 20450 }, { "grad_norm": 0.24457986652851105, "learning_rate": 8.886726817647147e-05, "loss": 0.0344, "step": 20460 }, { "grad_norm": 0.2642762362957001, "learning_rate": 8.885426291881319e-05, "loss": 0.0303, "step": 20470 }, { "grad_norm": 0.25586235523223877, "learning_rate": 8.884125102202401e-05, "loss": 0.0228, "step": 20480 }, { "grad_norm": 0.25165319442749023, "learning_rate": 8.882823248832736e-05, "loss": 0.0156, "step": 20490 }, { "grad_norm": 0.24011798202991486, "learning_rate": 8.881520731994772e-05, "loss": 0.021, "step": 20500 }, { "grad_norm": 0.4415116608142853, "learning_rate": 8.880217551911077e-05, "loss": 0.0214, "step": 20510 }, { "grad_norm": 0.2705262005329132, "learning_rate": 8.878913708804323e-05, "loss": 0.0201, "step": 20520 }, { "grad_norm": 0.2678286135196686, "learning_rate": 8.877609202897308e-05, "loss": 0.0239, "step": 20530 }, { "grad_norm": 0.327898234128952, "learning_rate": 8.876304034412933e-05, "loss": 0.0212, "step": 20540 }, { "grad_norm": 0.3313930332660675, "learning_rate": 8.874998203574214e-05, "loss": 0.0271, "step": 20550 }, { "grad_norm": 0.2914118766784668, "learning_rate": 8.873691710604284e-05, "loss": 0.0206, "step": 20560 }, { "grad_norm": 0.36240988969802856, "learning_rate": 8.872384555726387e-05, "loss": 0.0194, "step": 20570 }, { "grad_norm": 0.2743500769138336, "learning_rate": 8.871076739163878e-05, "loss": 0.0211, "step": 20580 }, { "grad_norm": 0.30773410201072693, "learning_rate": 8.86976826114023e-05, "loss": 0.0256, "step": 20590 }, { "grad_norm": 0.31567251682281494, "learning_rate": 8.868459121879023e-05, "loss": 0.0261, "step": 20600 }, { "grad_norm": 0.2343410849571228, "learning_rate": 8.867149321603956e-05, "loss": 0.0279, "step": 20610 }, { "grad_norm": 0.2490563690662384, "learning_rate": 8.865838860538835e-05, "loss": 0.0194, "step": 20620 }, { "grad_norm": 0.25674235820770264, "learning_rate": 8.864527738907585e-05, "loss": 0.0242, "step": 20630 }, { "grad_norm": 0.3485383093357086, "learning_rate": 8.863215956934239e-05, "loss": 0.0243, "step": 20640 }, { "grad_norm": 0.2564593553543091, "learning_rate": 8.861903514842947e-05, "loss": 0.0182, "step": 20650 }, { "grad_norm": 0.31622588634490967, "learning_rate": 8.860590412857966e-05, "loss": 0.0219, "step": 20660 }, { "grad_norm": 0.2057475447654724, "learning_rate": 8.85927665120367e-05, "loss": 0.0154, "step": 20670 }, { "grad_norm": 0.31032097339630127, "learning_rate": 8.857962230104546e-05, "loss": 0.0194, "step": 20680 }, { "grad_norm": 0.31302526593208313, "learning_rate": 8.856647149785193e-05, "loss": 0.0196, "step": 20690 }, { "grad_norm": 0.2610245645046234, "learning_rate": 8.855331410470322e-05, "loss": 0.0203, "step": 20700 }, { "grad_norm": 0.32821816205978394, "learning_rate": 8.854015012384756e-05, "loss": 0.0182, "step": 20710 }, { "grad_norm": 0.2899799048900604, "learning_rate": 8.852697955753433e-05, "loss": 0.0244, "step": 20720 }, { "grad_norm": 0.2132079303264618, "learning_rate": 8.851380240801399e-05, "loss": 0.0291, "step": 20730 }, { "grad_norm": 0.19830608367919922, "learning_rate": 8.850061867753818e-05, "loss": 0.0159, "step": 20740 }, { "grad_norm": 0.3496856391429901, "learning_rate": 8.848742836835963e-05, "loss": 0.0257, "step": 20750 }, { "grad_norm": 0.2555263042449951, "learning_rate": 8.847423148273221e-05, "loss": 0.0281, "step": 20760 }, { "grad_norm": 0.24850600957870483, "learning_rate": 8.846102802291092e-05, "loss": 0.0288, "step": 20770 }, { "grad_norm": 0.303943395614624, "learning_rate": 8.844781799115183e-05, "loss": 0.0269, "step": 20780 }, { "grad_norm": 0.3045172095298767, "learning_rate": 8.84346013897122e-05, "loss": 0.0238, "step": 20790 }, { "grad_norm": 0.35029852390289307, "learning_rate": 8.842137822085038e-05, "loss": 0.0257, "step": 20800 }, { "grad_norm": 0.24869997799396515, "learning_rate": 8.840814848682585e-05, "loss": 0.0181, "step": 20810 }, { "grad_norm": 0.20954564213752747, "learning_rate": 8.83949121898992e-05, "loss": 0.0298, "step": 20820 }, { "grad_norm": 0.2063262015581131, "learning_rate": 8.838166933233217e-05, "loss": 0.0184, "step": 20830 }, { "grad_norm": 0.2939109206199646, "learning_rate": 8.83684199163876e-05, "loss": 0.025, "step": 20840 }, { "grad_norm": 0.23119299113750458, "learning_rate": 8.835516394432943e-05, "loss": 0.0166, "step": 20850 }, { "grad_norm": 0.21086730062961578, "learning_rate": 8.834190141842276e-05, "loss": 0.0266, "step": 20860 }, { "grad_norm": 0.41871026158332825, "learning_rate": 8.83286323409338e-05, "loss": 0.0228, "step": 20870 }, { "grad_norm": 0.23900355398654938, "learning_rate": 8.831535671412986e-05, "loss": 0.0217, "step": 20880 }, { "grad_norm": 0.292807012796402, "learning_rate": 8.830207454027938e-05, "loss": 0.0205, "step": 20890 }, { "grad_norm": 0.240286722779274, "learning_rate": 8.828878582165192e-05, "loss": 0.0202, "step": 20900 }, { "grad_norm": 0.27691730856895447, "learning_rate": 8.827549056051818e-05, "loss": 0.0178, "step": 20910 }, { "grad_norm": 0.2116340547800064, "learning_rate": 8.826218875914993e-05, "loss": 0.0179, "step": 20920 }, { "grad_norm": 0.17614302039146423, "learning_rate": 8.82488804198201e-05, "loss": 0.0178, "step": 20930 }, { "grad_norm": 0.185320183634758, "learning_rate": 8.82355655448027e-05, "loss": 0.0188, "step": 20940 }, { "grad_norm": 0.3700696527957916, "learning_rate": 8.822224413637293e-05, "loss": 0.0234, "step": 20950 }, { "grad_norm": 0.2786369323730469, "learning_rate": 8.820891619680697e-05, "loss": 0.0223, "step": 20960 }, { "grad_norm": 0.3473937511444092, "learning_rate": 8.819558172838227e-05, "loss": 0.0248, "step": 20970 }, { "grad_norm": 0.20534370839595795, "learning_rate": 8.818224073337731e-05, "loss": 0.0179, "step": 20980 }, { "grad_norm": 0.2551165819168091, "learning_rate": 8.816889321407169e-05, "loss": 0.018, "step": 20990 }, { "grad_norm": 0.20855213701725006, "learning_rate": 8.815553917274615e-05, "loss": 0.0195, "step": 21000 }, { "grad_norm": 0.2240505963563919, "learning_rate": 8.81421786116825e-05, "loss": 0.0276, "step": 21010 }, { "grad_norm": 0.24886363744735718, "learning_rate": 8.812881153316373e-05, "loss": 0.0182, "step": 21020 }, { "grad_norm": 0.4305388033390045, "learning_rate": 8.81154379394739e-05, "loss": 0.0269, "step": 21030 }, { "grad_norm": 0.3951822519302368, "learning_rate": 8.810205783289818e-05, "loss": 0.0232, "step": 21040 }, { "grad_norm": 0.23572510480880737, "learning_rate": 8.808867121572286e-05, "loss": 0.0221, "step": 21050 }, { "grad_norm": 0.21599508821964264, "learning_rate": 8.807527809023537e-05, "loss": 0.0247, "step": 21060 }, { "grad_norm": 0.21824337542057037, "learning_rate": 8.80618784587242e-05, "loss": 0.025, "step": 21070 }, { "grad_norm": 0.3330705165863037, "learning_rate": 8.804847232347902e-05, "loss": 0.024, "step": 21080 }, { "grad_norm": 0.24199561774730682, "learning_rate": 8.803505968679054e-05, "loss": 0.0245, "step": 21090 }, { "grad_norm": 0.3158712387084961, "learning_rate": 8.802164055095061e-05, "loss": 0.0254, "step": 21100 }, { "grad_norm": 0.3103267550468445, "learning_rate": 8.80082149182522e-05, "loss": 0.021, "step": 21110 }, { "grad_norm": 0.29973194003105164, "learning_rate": 8.79947827909894e-05, "loss": 0.0192, "step": 21120 }, { "grad_norm": 0.27920156717300415, "learning_rate": 8.798134417145738e-05, "loss": 0.0242, "step": 21130 }, { "grad_norm": 0.2486690878868103, "learning_rate": 8.796789906195243e-05, "loss": 0.0221, "step": 21140 }, { "grad_norm": 0.222757488489151, "learning_rate": 8.795444746477195e-05, "loss": 0.0173, "step": 21150 }, { "grad_norm": 0.32885387539863586, "learning_rate": 8.794098938221446e-05, "loss": 0.0238, "step": 21160 }, { "grad_norm": 0.16412191092967987, "learning_rate": 8.792752481657957e-05, "loss": 0.0214, "step": 21170 }, { "grad_norm": 0.28660985827445984, "learning_rate": 8.791405377016802e-05, "loss": 0.0192, "step": 21180 }, { "grad_norm": 0.308133065700531, "learning_rate": 8.790057624528163e-05, "loss": 0.0241, "step": 21190 }, { "grad_norm": 0.375044584274292, "learning_rate": 8.788709224422333e-05, "loss": 0.022, "step": 21200 }, { "grad_norm": 0.2568941116333008, "learning_rate": 8.787360176929717e-05, "loss": 0.0216, "step": 21210 }, { "grad_norm": 0.16273753345012665, "learning_rate": 8.786010482280834e-05, "loss": 0.023, "step": 21220 }, { "grad_norm": 0.21814559400081635, "learning_rate": 8.784660140706306e-05, "loss": 0.0221, "step": 21230 }, { "grad_norm": 0.29850688576698303, "learning_rate": 8.783309152436872e-05, "loss": 0.0199, "step": 21240 }, { "grad_norm": 0.2696208357810974, "learning_rate": 8.781957517703375e-05, "loss": 0.0224, "step": 21250 }, { "grad_norm": 0.30490821599960327, "learning_rate": 8.780605236736776e-05, "loss": 0.0234, "step": 21260 }, { "grad_norm": 0.2556997835636139, "learning_rate": 8.779252309768142e-05, "loss": 0.0192, "step": 21270 }, { "grad_norm": 0.22920268774032593, "learning_rate": 8.777898737028652e-05, "loss": 0.0338, "step": 21280 }, { "grad_norm": 0.22102083265781403, "learning_rate": 8.776544518749591e-05, "loss": 0.0314, "step": 21290 }, { "grad_norm": 0.29594242572784424, "learning_rate": 8.775189655162364e-05, "loss": 0.0247, "step": 21300 }, { "grad_norm": 0.6023618578910828, "learning_rate": 8.773834146498474e-05, "loss": 0.027, "step": 21310 }, { "grad_norm": 0.40669670701026917, "learning_rate": 8.772477992989545e-05, "loss": 0.0256, "step": 21320 }, { "grad_norm": 0.3248502016067505, "learning_rate": 8.771121194867304e-05, "loss": 0.0323, "step": 21330 }, { "grad_norm": 0.251546710729599, "learning_rate": 8.769763752363589e-05, "loss": 0.024, "step": 21340 }, { "grad_norm": 0.27525678277015686, "learning_rate": 8.768405665710352e-05, "loss": 0.0293, "step": 21350 }, { "grad_norm": 0.365037202835083, "learning_rate": 8.767046935139655e-05, "loss": 0.0233, "step": 21360 }, { "grad_norm": 0.25999340415000916, "learning_rate": 8.765687560883666e-05, "loss": 0.0224, "step": 21370 }, { "grad_norm": 0.3242656886577606, "learning_rate": 8.764327543174664e-05, "loss": 0.0194, "step": 21380 }, { "grad_norm": 0.21485400199890137, "learning_rate": 8.762966882245038e-05, "loss": 0.0184, "step": 21390 }, { "grad_norm": 0.36546677350997925, "learning_rate": 8.761605578327291e-05, "loss": 0.0253, "step": 21400 }, { "grad_norm": 0.17575201392173767, "learning_rate": 8.76024363165403e-05, "loss": 0.0268, "step": 21410 }, { "grad_norm": 0.2936478555202484, "learning_rate": 8.758881042457976e-05, "loss": 0.0256, "step": 21420 }, { "grad_norm": 0.42189323902130127, "learning_rate": 8.757517810971957e-05, "loss": 0.0222, "step": 21430 }, { "grad_norm": 0.2553100883960724, "learning_rate": 8.756153937428913e-05, "loss": 0.0228, "step": 21440 }, { "grad_norm": 0.32751700282096863, "learning_rate": 8.754789422061889e-05, "loss": 0.0273, "step": 21450 }, { "grad_norm": 0.20284898579120636, "learning_rate": 8.753424265104052e-05, "loss": 0.0191, "step": 21460 }, { "grad_norm": 0.34058520197868347, "learning_rate": 8.752058466788659e-05, "loss": 0.029, "step": 21470 }, { "grad_norm": 0.2529947757720947, "learning_rate": 8.750692027349097e-05, "loss": 0.0202, "step": 21480 }, { "grad_norm": 0.2978547513484955, "learning_rate": 8.749324947018847e-05, "loss": 0.0214, "step": 21490 }, { "grad_norm": 0.29794248938560486, "learning_rate": 8.747957226031507e-05, "loss": 0.0212, "step": 21500 }, { "grad_norm": 0.26380157470703125, "learning_rate": 8.746588864620787e-05, "loss": 0.0235, "step": 21510 }, { "grad_norm": 0.25338953733444214, "learning_rate": 8.745219863020498e-05, "loss": 0.0184, "step": 21520 }, { "grad_norm": 0.2745751142501831, "learning_rate": 8.743850221464564e-05, "loss": 0.0201, "step": 21530 }, { "grad_norm": 0.269561767578125, "learning_rate": 8.742479940187026e-05, "loss": 0.0183, "step": 21540 }, { "grad_norm": 0.31195011734962463, "learning_rate": 8.74110901942202e-05, "loss": 0.02, "step": 21550 }, { "grad_norm": 0.2709837555885315, "learning_rate": 8.739737459403803e-05, "loss": 0.0211, "step": 21560 }, { "grad_norm": 0.1688566654920578, "learning_rate": 8.738365260366737e-05, "loss": 0.0195, "step": 21570 }, { "grad_norm": 0.19035963714122772, "learning_rate": 8.736992422545292e-05, "loss": 0.0189, "step": 21580 }, { "grad_norm": 0.35111212730407715, "learning_rate": 8.73561894617405e-05, "loss": 0.0305, "step": 21590 }, { "grad_norm": 0.20522426068782806, "learning_rate": 8.734244831487697e-05, "loss": 0.0183, "step": 21600 }, { "grad_norm": 0.31080666184425354, "learning_rate": 8.732870078721035e-05, "loss": 0.0324, "step": 21610 }, { "grad_norm": 0.23753510415554047, "learning_rate": 8.731494688108972e-05, "loss": 0.0187, "step": 21620 }, { "grad_norm": 0.2555677890777588, "learning_rate": 8.730118659886523e-05, "loss": 0.0279, "step": 21630 }, { "grad_norm": 0.4072701930999756, "learning_rate": 8.728741994288814e-05, "loss": 0.0245, "step": 21640 }, { "grad_norm": 0.33244603872299194, "learning_rate": 8.727364691551079e-05, "loss": 0.0202, "step": 21650 }, { "grad_norm": 0.35422956943511963, "learning_rate": 8.725986751908661e-05, "loss": 0.0203, "step": 21660 }, { "grad_norm": 0.2367301881313324, "learning_rate": 8.724608175597016e-05, "loss": 0.0196, "step": 21670 }, { "grad_norm": 0.3205711543560028, "learning_rate": 8.723228962851699e-05, "loss": 0.0168, "step": 21680 }, { "grad_norm": 0.5015128254890442, "learning_rate": 8.721849113908385e-05, "loss": 0.0193, "step": 21690 }, { "grad_norm": 0.26769009232521057, "learning_rate": 8.720468629002848e-05, "loss": 0.0235, "step": 21700 }, { "grad_norm": 0.33028125762939453, "learning_rate": 8.719087508370978e-05, "loss": 0.017, "step": 21710 }, { "grad_norm": 0.273175984621048, "learning_rate": 8.717705752248772e-05, "loss": 0.0248, "step": 21720 }, { "grad_norm": 0.290962815284729, "learning_rate": 8.71632336087233e-05, "loss": 0.021, "step": 21730 }, { "grad_norm": 0.33819854259490967, "learning_rate": 8.71494033447787e-05, "loss": 0.0242, "step": 21740 }, { "grad_norm": 0.26080429553985596, "learning_rate": 8.713556673301708e-05, "loss": 0.0161, "step": 21750 }, { "grad_norm": 0.2169567495584488, "learning_rate": 8.712172377580278e-05, "loss": 0.0306, "step": 21760 }, { "grad_norm": 0.32935357093811035, "learning_rate": 8.710787447550114e-05, "loss": 0.0264, "step": 21770 }, { "grad_norm": 0.27488917112350464, "learning_rate": 8.70940188344787e-05, "loss": 0.0277, "step": 21780 }, { "grad_norm": 0.26643800735473633, "learning_rate": 8.708015685510293e-05, "loss": 0.0224, "step": 21790 }, { "grad_norm": 0.29477164149284363, "learning_rate": 8.706628853974252e-05, "loss": 0.0234, "step": 21800 }, { "grad_norm": 0.2306334227323532, "learning_rate": 8.705241389076715e-05, "loss": 0.022, "step": 21810 }, { "grad_norm": 0.29633548855781555, "learning_rate": 8.703853291054764e-05, "loss": 0.0197, "step": 21820 }, { "grad_norm": 0.36893653869628906, "learning_rate": 8.702464560145587e-05, "loss": 0.0211, "step": 21830 }, { "grad_norm": 0.17578653991222382, "learning_rate": 8.701075196586476e-05, "loss": 0.0198, "step": 21840 }, { "grad_norm": 0.20989133417606354, "learning_rate": 8.699685200614842e-05, "loss": 0.0207, "step": 21850 }, { "grad_norm": 0.3326020836830139, "learning_rate": 8.698294572468193e-05, "loss": 0.0239, "step": 21860 }, { "grad_norm": 0.1357812136411667, "learning_rate": 8.696903312384148e-05, "loss": 0.021, "step": 21870 }, { "grad_norm": 0.2222125232219696, "learning_rate": 8.695511420600439e-05, "loss": 0.016, "step": 21880 }, { "grad_norm": 0.21386447548866272, "learning_rate": 8.694118897354901e-05, "loss": 0.0139, "step": 21890 }, { "grad_norm": 0.21344082057476044, "learning_rate": 8.692725742885478e-05, "loss": 0.0173, "step": 21900 }, { "grad_norm": 0.347724586725235, "learning_rate": 8.691331957430221e-05, "loss": 0.0225, "step": 21910 }, { "grad_norm": 0.2831481099128723, "learning_rate": 8.68993754122729e-05, "loss": 0.0212, "step": 21920 }, { "grad_norm": 0.2610733211040497, "learning_rate": 8.688542494514955e-05, "loss": 0.0159, "step": 21930 }, { "grad_norm": 0.25461581349372864, "learning_rate": 8.68714681753159e-05, "loss": 0.0194, "step": 21940 }, { "grad_norm": 0.2821119427680969, "learning_rate": 8.685750510515676e-05, "loss": 0.0213, "step": 21950 }, { "grad_norm": 0.17238378524780273, "learning_rate": 8.684353573705805e-05, "loss": 0.0223, "step": 21960 }, { "grad_norm": 0.2979683578014374, "learning_rate": 8.682956007340677e-05, "loss": 0.0149, "step": 21970 }, { "grad_norm": 0.2772475481033325, "learning_rate": 8.681557811659095e-05, "loss": 0.0172, "step": 21980 }, { "grad_norm": 0.23061369359493256, "learning_rate": 8.680158986899974e-05, "loss": 0.0178, "step": 21990 }, { "grad_norm": 0.20854471623897552, "learning_rate": 8.678759533302335e-05, "loss": 0.0194, "step": 22000 }, { "grad_norm": 0.30077287554740906, "learning_rate": 8.677359451105308e-05, "loss": 0.0187, "step": 22010 }, { "grad_norm": 0.16850218176841736, "learning_rate": 8.675958740548123e-05, "loss": 0.0227, "step": 22020 }, { "grad_norm": 0.234857439994812, "learning_rate": 8.674557401870129e-05, "loss": 0.024, "step": 22030 }, { "grad_norm": 0.3339470624923706, "learning_rate": 8.673155435310775e-05, "loss": 0.0183, "step": 22040 }, { "grad_norm": 0.16968227922916412, "learning_rate": 8.671752841109617e-05, "loss": 0.0143, "step": 22050 }, { "grad_norm": 0.24588841199874878, "learning_rate": 8.670349619506321e-05, "loss": 0.0197, "step": 22060 }, { "grad_norm": 0.4280160069465637, "learning_rate": 8.66894577074066e-05, "loss": 0.0182, "step": 22070 }, { "grad_norm": 0.23398204147815704, "learning_rate": 8.667541295052513e-05, "loss": 0.017, "step": 22080 }, { "grad_norm": 0.4953756034374237, "learning_rate": 8.666136192681865e-05, "loss": 0.0254, "step": 22090 }, { "grad_norm": 0.2480010986328125, "learning_rate": 8.664730463868811e-05, "loss": 0.0181, "step": 22100 }, { "grad_norm": 0.32929205894470215, "learning_rate": 8.663324108853552e-05, "loss": 0.0301, "step": 22110 }, { "grad_norm": 0.3542686998844147, "learning_rate": 8.661917127876395e-05, "loss": 0.0307, "step": 22120 }, { "grad_norm": 0.3338368833065033, "learning_rate": 8.660509521177754e-05, "loss": 0.0237, "step": 22130 }, { "grad_norm": 0.17052635550498962, "learning_rate": 8.65910128899815e-05, "loss": 0.02, "step": 22140 }, { "grad_norm": 0.2325194776058197, "learning_rate": 8.657692431578214e-05, "loss": 0.0228, "step": 22150 }, { "grad_norm": 0.2672604024410248, "learning_rate": 8.656282949158679e-05, "loss": 0.0182, "step": 22160 }, { "grad_norm": 0.13783900439739227, "learning_rate": 8.654872841980388e-05, "loss": 0.0123, "step": 22170 }, { "grad_norm": 0.2765243947505951, "learning_rate": 8.653462110284289e-05, "loss": 0.016, "step": 22180 }, { "grad_norm": 0.235557422041893, "learning_rate": 8.652050754311437e-05, "loss": 0.0172, "step": 22190 }, { "grad_norm": 0.25641822814941406, "learning_rate": 8.650638774302995e-05, "loss": 0.018, "step": 22200 }, { "grad_norm": 0.3288574814796448, "learning_rate": 8.649226170500233e-05, "loss": 0.0187, "step": 22210 }, { "grad_norm": 0.2552492022514343, "learning_rate": 8.647812943144524e-05, "loss": 0.0184, "step": 22220 }, { "grad_norm": 0.19641347229480743, "learning_rate": 8.646399092477351e-05, "loss": 0.0173, "step": 22230 }, { "grad_norm": 0.35531294345855713, "learning_rate": 8.644984618740301e-05, "loss": 0.0234, "step": 22240 }, { "grad_norm": 0.20782583951950073, "learning_rate": 8.643569522175073e-05, "loss": 0.0178, "step": 22250 }, { "grad_norm": 0.3068031072616577, "learning_rate": 8.642153803023463e-05, "loss": 0.0215, "step": 22260 }, { "grad_norm": 0.18884772062301636, "learning_rate": 8.640737461527383e-05, "loss": 0.0147, "step": 22270 }, { "grad_norm": 0.25819480419158936, "learning_rate": 8.639320497928845e-05, "loss": 0.0209, "step": 22280 }, { "grad_norm": 0.240977942943573, "learning_rate": 8.637902912469969e-05, "loss": 0.0151, "step": 22290 }, { "grad_norm": 0.2363945096731186, "learning_rate": 8.636484705392982e-05, "loss": 0.0192, "step": 22300 }, { "grad_norm": 0.29450303316116333, "learning_rate": 8.635065876940216e-05, "loss": 0.0211, "step": 22310 }, { "grad_norm": 0.24422839283943176, "learning_rate": 8.633646427354112e-05, "loss": 0.023, "step": 22320 }, { "grad_norm": 0.27524179220199585, "learning_rate": 8.632226356877213e-05, "loss": 0.021, "step": 22330 }, { "grad_norm": 0.2532022297382355, "learning_rate": 8.630805665752173e-05, "loss": 0.0258, "step": 22340 }, { "grad_norm": 0.23081374168395996, "learning_rate": 8.629384354221748e-05, "loss": 0.0204, "step": 22350 }, { "grad_norm": 0.17682915925979614, "learning_rate": 8.627962422528797e-05, "loss": 0.0196, "step": 22360 }, { "grad_norm": 0.2657669186592102, "learning_rate": 8.626539870916296e-05, "loss": 0.0294, "step": 22370 }, { "grad_norm": 0.24773019552230835, "learning_rate": 8.625116699627317e-05, "loss": 0.0199, "step": 22380 }, { "grad_norm": 0.1787988543510437, "learning_rate": 8.623692908905041e-05, "loss": 0.0149, "step": 22390 }, { "grad_norm": 0.2621360421180725, "learning_rate": 8.622268498992755e-05, "loss": 0.023, "step": 22400 }, { "grad_norm": 0.23394529521465302, "learning_rate": 8.620843470133851e-05, "loss": 0.0307, "step": 22410 }, { "grad_norm": 0.34414470195770264, "learning_rate": 8.619417822571829e-05, "loss": 0.0209, "step": 22420 }, { "grad_norm": 0.32732346653938293, "learning_rate": 8.617991556550292e-05, "loss": 0.0158, "step": 22430 }, { "grad_norm": 0.31804656982421875, "learning_rate": 8.616564672312952e-05, "loss": 0.02, "step": 22440 }, { "grad_norm": 0.2847036123275757, "learning_rate": 8.61513717010362e-05, "loss": 0.0222, "step": 22450 }, { "grad_norm": 0.30172064900398254, "learning_rate": 8.613709050166221e-05, "loss": 0.0214, "step": 22460 }, { "grad_norm": 0.18362043797969818, "learning_rate": 8.61228031274478e-05, "loss": 0.0237, "step": 22470 }, { "grad_norm": 0.31232213973999023, "learning_rate": 8.610850958083431e-05, "loss": 0.0186, "step": 22480 }, { "grad_norm": 0.3111746609210968, "learning_rate": 8.609420986426409e-05, "loss": 0.0198, "step": 22490 }, { "grad_norm": 0.24927851557731628, "learning_rate": 8.60799039801806e-05, "loss": 0.0174, "step": 22500 }, { "grad_norm": 0.22824938595294952, "learning_rate": 8.606559193102828e-05, "loss": 0.0142, "step": 22510 }, { "grad_norm": 0.20457705855369568, "learning_rate": 8.605127371925273e-05, "loss": 0.025, "step": 22520 }, { "grad_norm": 0.2756165862083435, "learning_rate": 8.603694934730047e-05, "loss": 0.0217, "step": 22530 }, { "grad_norm": 0.22447803616523743, "learning_rate": 8.602261881761919e-05, "loss": 0.0148, "step": 22540 }, { "grad_norm": 0.16177941858768463, "learning_rate": 8.600828213265759e-05, "loss": 0.0228, "step": 22550 }, { "grad_norm": 0.37597334384918213, "learning_rate": 8.599393929486539e-05, "loss": 0.0282, "step": 22560 }, { "grad_norm": 0.2722380757331848, "learning_rate": 8.59795903066934e-05, "loss": 0.0178, "step": 22570 }, { "grad_norm": 0.2738266587257385, "learning_rate": 8.596523517059347e-05, "loss": 0.0176, "step": 22580 }, { "grad_norm": 0.19838255643844604, "learning_rate": 8.59508738890185e-05, "loss": 0.015, "step": 22590 }, { "grad_norm": 0.23529289662837982, "learning_rate": 8.593650646442246e-05, "loss": 0.0226, "step": 22600 }, { "grad_norm": 0.2976261079311371, "learning_rate": 8.59221328992603e-05, "loss": 0.0156, "step": 22610 }, { "grad_norm": 0.25055745244026184, "learning_rate": 8.590775319598813e-05, "loss": 0.0187, "step": 22620 }, { "grad_norm": 0.33982110023498535, "learning_rate": 8.589336735706301e-05, "loss": 0.0157, "step": 22630 }, { "grad_norm": 0.3126983940601349, "learning_rate": 8.587897538494307e-05, "loss": 0.0262, "step": 22640 }, { "grad_norm": 0.4371914565563202, "learning_rate": 8.586457728208756e-05, "loss": 0.0231, "step": 22650 }, { "grad_norm": 0.4238954484462738, "learning_rate": 8.585017305095667e-05, "loss": 0.0229, "step": 22660 }, { "grad_norm": 0.2610109746456146, "learning_rate": 8.583576269401173e-05, "loss": 0.0245, "step": 22670 }, { "grad_norm": 0.3130834996700287, "learning_rate": 8.582134621371504e-05, "loss": 0.0202, "step": 22680 }, { "grad_norm": 0.3890957534313202, "learning_rate": 8.580692361253e-05, "loss": 0.0176, "step": 22690 }, { "grad_norm": 0.3315679728984833, "learning_rate": 8.579249489292104e-05, "loss": 0.0221, "step": 22700 }, { "grad_norm": 0.3438046872615814, "learning_rate": 8.577806005735363e-05, "loss": 0.0286, "step": 22710 }, { "grad_norm": 0.30052998661994934, "learning_rate": 8.576361910829429e-05, "loss": 0.0184, "step": 22720 }, { "grad_norm": 0.3912333548069, "learning_rate": 8.574917204821057e-05, "loss": 0.0211, "step": 22730 }, { "grad_norm": 0.175326406955719, "learning_rate": 8.57347188795711e-05, "loss": 0.0219, "step": 22740 }, { "grad_norm": 0.30474546551704407, "learning_rate": 8.572025960484551e-05, "loss": 0.0255, "step": 22750 }, { "grad_norm": 0.26719024777412415, "learning_rate": 8.57057942265045e-05, "loss": 0.0259, "step": 22760 }, { "grad_norm": 0.34910809993743896, "learning_rate": 8.569132274701984e-05, "loss": 0.021, "step": 22770 }, { "grad_norm": 0.21335235238075256, "learning_rate": 8.567684516886427e-05, "loss": 0.0173, "step": 22780 }, { "grad_norm": 0.2937045097351074, "learning_rate": 8.56623614945116e-05, "loss": 0.015, "step": 22790 }, { "grad_norm": 0.3680862784385681, "learning_rate": 8.564787172643675e-05, "loss": 0.0192, "step": 22800 }, { "grad_norm": 0.22422055900096893, "learning_rate": 8.563337586711559e-05, "loss": 0.0188, "step": 22810 }, { "grad_norm": 0.26851850748062134, "learning_rate": 8.561887391902506e-05, "loss": 0.0273, "step": 22820 }, { "grad_norm": 0.2719905376434326, "learning_rate": 8.560436588464316e-05, "loss": 0.0182, "step": 22830 }, { "grad_norm": 0.26575303077697754, "learning_rate": 8.55898517664489e-05, "loss": 0.0206, "step": 22840 }, { "grad_norm": 0.4742496907711029, "learning_rate": 8.557533156692236e-05, "loss": 0.0334, "step": 22850 }, { "grad_norm": 0.34199193120002747, "learning_rate": 8.556080528854467e-05, "loss": 0.0159, "step": 22860 }, { "grad_norm": 0.3067469298839569, "learning_rate": 8.554627293379791e-05, "loss": 0.0174, "step": 22870 }, { "grad_norm": 0.28949472308158875, "learning_rate": 8.553173450516531e-05, "loss": 0.0254, "step": 22880 }, { "grad_norm": 0.2008538693189621, "learning_rate": 8.551719000513108e-05, "loss": 0.0177, "step": 22890 }, { "grad_norm": 0.3595437705516815, "learning_rate": 8.550263943618049e-05, "loss": 0.0333, "step": 22900 }, { "grad_norm": 0.26002010703086853, "learning_rate": 8.54880828007998e-05, "loss": 0.0309, "step": 22910 }, { "grad_norm": 0.24201878905296326, "learning_rate": 8.547352010147637e-05, "loss": 0.0264, "step": 22920 }, { "grad_norm": 0.31640708446502686, "learning_rate": 8.545895134069855e-05, "loss": 0.0201, "step": 22930 }, { "grad_norm": 0.26701056957244873, "learning_rate": 8.544437652095576e-05, "loss": 0.0157, "step": 22940 }, { "grad_norm": 0.469032347202301, "learning_rate": 8.542979564473843e-05, "loss": 0.03, "step": 22950 }, { "grad_norm": 0.23094671964645386, "learning_rate": 8.541520871453802e-05, "loss": 0.0163, "step": 22960 }, { "grad_norm": 0.29773491621017456, "learning_rate": 8.540061573284705e-05, "loss": 0.0153, "step": 22970 }, { "grad_norm": 0.2474329173564911, "learning_rate": 8.538601670215906e-05, "loss": 0.0193, "step": 22980 }, { "grad_norm": 0.295822411775589, "learning_rate": 8.537141162496864e-05, "loss": 0.0201, "step": 22990 }, { "grad_norm": 0.35455188155174255, "learning_rate": 8.535680050377137e-05, "loss": 0.0219, "step": 23000 }, { "grad_norm": 0.2940465807914734, "learning_rate": 8.534218334106391e-05, "loss": 0.0191, "step": 23010 }, { "grad_norm": 0.22678548097610474, "learning_rate": 8.532756013934393e-05, "loss": 0.0225, "step": 23020 }, { "grad_norm": 0.20559193193912506, "learning_rate": 8.531293090111012e-05, "loss": 0.0232, "step": 23030 }, { "grad_norm": 0.2238209992647171, "learning_rate": 8.529829562886225e-05, "loss": 0.0184, "step": 23040 }, { "grad_norm": 0.347922146320343, "learning_rate": 8.528365432510105e-05, "loss": 0.0243, "step": 23050 }, { "grad_norm": 0.3005368113517761, "learning_rate": 8.526900699232833e-05, "loss": 0.0215, "step": 23060 }, { "grad_norm": 0.22844147682189941, "learning_rate": 8.525435363304695e-05, "loss": 0.0195, "step": 23070 }, { "grad_norm": 0.21565838158130646, "learning_rate": 8.523969424976072e-05, "loss": 0.0197, "step": 23080 }, { "grad_norm": 0.1953432410955429, "learning_rate": 8.522502884497457e-05, "loss": 0.0186, "step": 23090 }, { "grad_norm": 0.2850029170513153, "learning_rate": 8.521035742119437e-05, "loss": 0.0211, "step": 23100 }, { "grad_norm": 0.16477616131305695, "learning_rate": 8.519567998092712e-05, "loss": 0.0299, "step": 23110 }, { "grad_norm": 0.24266399443149567, "learning_rate": 8.518099652668075e-05, "loss": 0.0221, "step": 23120 }, { "grad_norm": 0.22462747991085052, "learning_rate": 8.516630706096429e-05, "loss": 0.0202, "step": 23130 }, { "grad_norm": 0.17002832889556885, "learning_rate": 8.515161158628773e-05, "loss": 0.014, "step": 23140 }, { "grad_norm": 0.29216864705085754, "learning_rate": 8.513691010516216e-05, "loss": 0.038, "step": 23150 }, { "grad_norm": 0.23973964154720306, "learning_rate": 8.512220262009966e-05, "loss": 0.0185, "step": 23160 }, { "grad_norm": 0.33614346385002136, "learning_rate": 8.510748913361332e-05, "loss": 0.0223, "step": 23170 }, { "grad_norm": 0.2143629938364029, "learning_rate": 8.509276964821726e-05, "loss": 0.0178, "step": 23180 }, { "grad_norm": 0.23475515842437744, "learning_rate": 8.507804416642669e-05, "loss": 0.0179, "step": 23190 }, { "grad_norm": 0.2249014675617218, "learning_rate": 8.506331269075774e-05, "loss": 0.0165, "step": 23200 }, { "grad_norm": 0.37065160274505615, "learning_rate": 8.504857522372765e-05, "loss": 0.0151, "step": 23210 }, { "grad_norm": 0.18049314618110657, "learning_rate": 8.503383176785461e-05, "loss": 0.0162, "step": 23220 }, { "grad_norm": 0.27915680408477783, "learning_rate": 8.501908232565792e-05, "loss": 0.02, "step": 23230 }, { "grad_norm": 0.24973466992378235, "learning_rate": 8.50043268996578e-05, "loss": 0.0158, "step": 23240 }, { "grad_norm": 0.266806960105896, "learning_rate": 8.498956549237562e-05, "loss": 0.0203, "step": 23250 }, { "grad_norm": 0.19460061192512512, "learning_rate": 8.497479810633366e-05, "loss": 0.0192, "step": 23260 }, { "grad_norm": 0.3716616630554199, "learning_rate": 8.496002474405525e-05, "loss": 0.0249, "step": 23270 }, { "grad_norm": 0.5043413639068604, "learning_rate": 8.494524540806478e-05, "loss": 0.0231, "step": 23280 }, { "grad_norm": 0.2268788367509842, "learning_rate": 8.493046010088761e-05, "loss": 0.0306, "step": 23290 }, { "grad_norm": 0.3259357810020447, "learning_rate": 8.491566882505018e-05, "loss": 0.0271, "step": 23300 }, { "grad_norm": 0.3473004400730133, "learning_rate": 8.490087158307988e-05, "loss": 0.0241, "step": 23310 }, { "grad_norm": 0.4013957381248474, "learning_rate": 8.488606837750518e-05, "loss": 0.0184, "step": 23320 }, { "grad_norm": 0.2958561182022095, "learning_rate": 8.487125921085552e-05, "loss": 0.0212, "step": 23330 }, { "grad_norm": 0.22024399042129517, "learning_rate": 8.485644408566141e-05, "loss": 0.0227, "step": 23340 }, { "grad_norm": 0.2842771112918854, "learning_rate": 8.484162300445431e-05, "loss": 0.0226, "step": 23350 }, { "grad_norm": 0.32579272985458374, "learning_rate": 8.482679596976676e-05, "loss": 0.0165, "step": 23360 }, { "grad_norm": 0.20883150398731232, "learning_rate": 8.48119629841323e-05, "loss": 0.0258, "step": 23370 }, { "grad_norm": 0.3422904908657074, "learning_rate": 8.479712405008547e-05, "loss": 0.0262, "step": 23380 }, { "grad_norm": 0.24051839113235474, "learning_rate": 8.478227917016184e-05, "loss": 0.0308, "step": 23390 }, { "grad_norm": 0.2353856861591339, "learning_rate": 8.476742834689801e-05, "loss": 0.0287, "step": 23400 }, { "grad_norm": 0.3203820586204529, "learning_rate": 8.475257158283157e-05, "loss": 0.0255, "step": 23410 }, { "grad_norm": 0.24932941794395447, "learning_rate": 8.473770888050112e-05, "loss": 0.017, "step": 23420 }, { "grad_norm": 0.6497485637664795, "learning_rate": 8.47228402424463e-05, "loss": 0.0257, "step": 23430 }, { "grad_norm": 0.16316381096839905, "learning_rate": 8.470796567120775e-05, "loss": 0.022, "step": 23440 }, { "grad_norm": 0.34890496730804443, "learning_rate": 8.469308516932714e-05, "loss": 0.026, "step": 23450 }, { "grad_norm": 0.29934802651405334, "learning_rate": 8.467819873934714e-05, "loss": 0.0229, "step": 23460 }, { "grad_norm": 0.33653661608695984, "learning_rate": 8.466330638381143e-05, "loss": 0.0249, "step": 23470 }, { "grad_norm": 0.2138463854789734, "learning_rate": 8.464840810526469e-05, "loss": 0.0214, "step": 23480 }, { "grad_norm": 0.3251016438007355, "learning_rate": 8.463350390625264e-05, "loss": 0.03, "step": 23490 }, { "grad_norm": 0.5502961874008179, "learning_rate": 8.4618593789322e-05, "loss": 0.035, "step": 23500 }, { "grad_norm": 0.2600526213645935, "learning_rate": 8.46036777570205e-05, "loss": 0.0248, "step": 23510 }, { "grad_norm": 0.3108114004135132, "learning_rate": 8.458875581189688e-05, "loss": 0.0338, "step": 23520 }, { "grad_norm": 0.19760750234127045, "learning_rate": 8.457382795650092e-05, "loss": 0.018, "step": 23530 }, { "grad_norm": 0.2266707420349121, "learning_rate": 8.455889419338335e-05, "loss": 0.0236, "step": 23540 }, { "grad_norm": 0.29956281185150146, "learning_rate": 8.454395452509593e-05, "loss": 0.0167, "step": 23550 }, { "grad_norm": 0.2539774179458618, "learning_rate": 8.452900895419146e-05, "loss": 0.0217, "step": 23560 }, { "grad_norm": 0.18916146457195282, "learning_rate": 8.451405748322376e-05, "loss": 0.0202, "step": 23570 }, { "grad_norm": 0.2972065508365631, "learning_rate": 8.449910011474759e-05, "loss": 0.0162, "step": 23580 }, { "grad_norm": 0.31213194131851196, "learning_rate": 8.448413685131876e-05, "loss": 0.0307, "step": 23590 }, { "grad_norm": 0.30759334564208984, "learning_rate": 8.446916769549407e-05, "loss": 0.0265, "step": 23600 }, { "grad_norm": 0.6185351610183716, "learning_rate": 8.445419264983136e-05, "loss": 0.0258, "step": 23610 }, { "grad_norm": 0.1789563000202179, "learning_rate": 8.443921171688947e-05, "loss": 0.0216, "step": 23620 }, { "grad_norm": 0.2751973569393158, "learning_rate": 8.442422489922819e-05, "loss": 0.0181, "step": 23630 }, { "grad_norm": 0.2900049388408661, "learning_rate": 8.440923219940838e-05, "loss": 0.0274, "step": 23640 }, { "grad_norm": 0.3142226040363312, "learning_rate": 8.439423361999189e-05, "loss": 0.0257, "step": 23650 }, { "grad_norm": 0.2680749297142029, "learning_rate": 8.437922916354155e-05, "loss": 0.0275, "step": 23660 }, { "grad_norm": 0.16548344492912292, "learning_rate": 8.436421883262123e-05, "loss": 0.0204, "step": 23670 }, { "grad_norm": 0.21444925665855408, "learning_rate": 8.434920262979577e-05, "loss": 0.0264, "step": 23680 }, { "grad_norm": 0.26693978905677795, "learning_rate": 8.433418055763104e-05, "loss": 0.0208, "step": 23690 }, { "grad_norm": 0.3728101849555969, "learning_rate": 8.431915261869389e-05, "loss": 0.0232, "step": 23700 }, { "grad_norm": 0.18841128051280975, "learning_rate": 8.43041188155522e-05, "loss": 0.0173, "step": 23710 }, { "grad_norm": 0.23141443729400635, "learning_rate": 8.428907915077481e-05, "loss": 0.0167, "step": 23720 }, { "grad_norm": 0.4112662374973297, "learning_rate": 8.42740336269316e-05, "loss": 0.0235, "step": 23730 }, { "grad_norm": 0.31347617506980896, "learning_rate": 8.425898224659345e-05, "loss": 0.0193, "step": 23740 }, { "grad_norm": 0.2520157992839813, "learning_rate": 8.42439250123322e-05, "loss": 0.0214, "step": 23750 }, { "grad_norm": 0.24475860595703125, "learning_rate": 8.422886192672076e-05, "loss": 0.0237, "step": 23760 }, { "grad_norm": 0.2946791350841522, "learning_rate": 8.421379299233297e-05, "loss": 0.0222, "step": 23770 }, { "grad_norm": 0.4082163870334625, "learning_rate": 8.419871821174371e-05, "loss": 0.0244, "step": 23780 }, { "grad_norm": 0.24678771197795868, "learning_rate": 8.418363758752884e-05, "loss": 0.0198, "step": 23790 }, { "grad_norm": 0.21139785647392273, "learning_rate": 8.416855112226523e-05, "loss": 0.022, "step": 23800 }, { "grad_norm": 0.1983000487089157, "learning_rate": 8.415345881853075e-05, "loss": 0.0192, "step": 23810 }, { "grad_norm": 0.3476252257823944, "learning_rate": 8.413836067890426e-05, "loss": 0.0203, "step": 23820 }, { "grad_norm": 0.32397449016571045, "learning_rate": 8.41232567059656e-05, "loss": 0.022, "step": 23830 }, { "grad_norm": 0.3197854161262512, "learning_rate": 8.410814690229565e-05, "loss": 0.0307, "step": 23840 }, { "grad_norm": 0.2802533805370331, "learning_rate": 8.409303127047626e-05, "loss": 0.0187, "step": 23850 }, { "grad_norm": 0.25843989849090576, "learning_rate": 8.407790981309028e-05, "loss": 0.0176, "step": 23860 }, { "grad_norm": 0.3758639991283417, "learning_rate": 8.406278253272153e-05, "loss": 0.0246, "step": 23870 }, { "grad_norm": 0.2298264503479004, "learning_rate": 8.404764943195487e-05, "loss": 0.0194, "step": 23880 }, { "grad_norm": 0.29022252559661865, "learning_rate": 8.403251051337613e-05, "loss": 0.0236, "step": 23890 }, { "grad_norm": 0.3402405083179474, "learning_rate": 8.401736577957214e-05, "loss": 0.0233, "step": 23900 }, { "grad_norm": 0.33267536759376526, "learning_rate": 8.40022152331307e-05, "loss": 0.0215, "step": 23910 }, { "grad_norm": 0.29528993368148804, "learning_rate": 8.398705887664064e-05, "loss": 0.0246, "step": 23920 }, { "grad_norm": 0.20585881173610687, "learning_rate": 8.397189671269177e-05, "loss": 0.0157, "step": 23930 }, { "grad_norm": 0.34609538316726685, "learning_rate": 8.395672874387488e-05, "loss": 0.0165, "step": 23940 }, { "grad_norm": 0.20311695337295532, "learning_rate": 8.394155497278177e-05, "loss": 0.0145, "step": 23950 }, { "grad_norm": 0.3709016740322113, "learning_rate": 8.392637540200523e-05, "loss": 0.0232, "step": 23960 }, { "grad_norm": 0.16357974708080292, "learning_rate": 8.391119003413902e-05, "loss": 0.0173, "step": 23970 }, { "grad_norm": 0.31488004326820374, "learning_rate": 8.38959988717779e-05, "loss": 0.0192, "step": 23980 }, { "grad_norm": 0.22559063136577606, "learning_rate": 8.388080191751764e-05, "loss": 0.0223, "step": 23990 }, { "grad_norm": 0.2511318027973175, "learning_rate": 8.386559917395496e-05, "loss": 0.0165, "step": 24000 }, { "grad_norm": 0.2489529252052307, "learning_rate": 8.385039064368761e-05, "loss": 0.0145, "step": 24010 }, { "grad_norm": 0.16250209510326385, "learning_rate": 8.383517632931431e-05, "loss": 0.0133, "step": 24020 }, { "grad_norm": 0.27020907402038574, "learning_rate": 8.381995623343477e-05, "loss": 0.0168, "step": 24030 }, { "grad_norm": 0.2455199956893921, "learning_rate": 8.380473035864968e-05, "loss": 0.0213, "step": 24040 }, { "grad_norm": 0.35953080654144287, "learning_rate": 8.378949870756076e-05, "loss": 0.0212, "step": 24050 }, { "grad_norm": 0.19313526153564453, "learning_rate": 8.377426128277063e-05, "loss": 0.022, "step": 24060 }, { "grad_norm": 0.1716238558292389, "learning_rate": 8.375901808688298e-05, "loss": 0.0133, "step": 24070 }, { "grad_norm": 0.2637847065925598, "learning_rate": 8.374376912250246e-05, "loss": 0.021, "step": 24080 }, { "grad_norm": 0.21577699482440948, "learning_rate": 8.372851439223468e-05, "loss": 0.0177, "step": 24090 }, { "grad_norm": 0.27237722277641296, "learning_rate": 8.371325389868627e-05, "loss": 0.0215, "step": 24100 }, { "grad_norm": 0.2829805910587311, "learning_rate": 8.369798764446482e-05, "loss": 0.0173, "step": 24110 }, { "grad_norm": 0.35649335384368896, "learning_rate": 8.368271563217893e-05, "loss": 0.0223, "step": 24120 }, { "grad_norm": 0.2055388242006302, "learning_rate": 8.366743786443817e-05, "loss": 0.0301, "step": 24130 }, { "grad_norm": 0.43956172466278076, "learning_rate": 8.365215434385309e-05, "loss": 0.0266, "step": 24140 }, { "grad_norm": 0.31485435366630554, "learning_rate": 8.36368650730352e-05, "loss": 0.0205, "step": 24150 }, { "grad_norm": 0.3093578815460205, "learning_rate": 8.362157005459705e-05, "loss": 0.0225, "step": 24160 }, { "grad_norm": 0.30670276284217834, "learning_rate": 8.360626929115213e-05, "loss": 0.0174, "step": 24170 }, { "grad_norm": 0.4289565682411194, "learning_rate": 8.359096278531492e-05, "loss": 0.0198, "step": 24180 }, { "grad_norm": 0.21379756927490234, "learning_rate": 8.357565053970088e-05, "loss": 0.019, "step": 24190 }, { "grad_norm": 0.21102073788642883, "learning_rate": 8.356033255692647e-05, "loss": 0.025, "step": 24200 }, { "grad_norm": 0.351997435092926, "learning_rate": 8.354500883960911e-05, "loss": 0.0187, "step": 24210 }, { "grad_norm": 0.2904318571090698, "learning_rate": 8.352967939036717e-05, "loss": 0.0249, "step": 24220 }, { "grad_norm": 0.2970268130302429, "learning_rate": 8.35143442118201e-05, "loss": 0.0264, "step": 24230 }, { "grad_norm": 0.18592414259910583, "learning_rate": 8.349900330658819e-05, "loss": 0.0158, "step": 24240 }, { "grad_norm": 0.17343491315841675, "learning_rate": 8.348365667729284e-05, "loss": 0.0182, "step": 24250 }, { "grad_norm": 0.299835741519928, "learning_rate": 8.346830432655633e-05, "loss": 0.0143, "step": 24260 }, { "grad_norm": 0.30639541149139404, "learning_rate": 8.345294625700195e-05, "loss": 0.019, "step": 24270 }, { "grad_norm": 0.30251607298851013, "learning_rate": 8.343758247125402e-05, "loss": 0.0188, "step": 24280 }, { "grad_norm": 0.3991411626338959, "learning_rate": 8.342221297193776e-05, "loss": 0.0264, "step": 24290 }, { "grad_norm": 0.22148996591567993, "learning_rate": 8.34068377616794e-05, "loss": 0.0275, "step": 24300 }, { "grad_norm": 0.4251110851764679, "learning_rate": 8.339145684310615e-05, "loss": 0.0228, "step": 24310 }, { "grad_norm": 0.3188556134700775, "learning_rate": 8.337607021884618e-05, "loss": 0.0205, "step": 24320 }, { "grad_norm": 0.18970739841461182, "learning_rate": 8.336067789152867e-05, "loss": 0.0146, "step": 24330 }, { "grad_norm": 0.1768442541360855, "learning_rate": 8.334527986378369e-05, "loss": 0.0213, "step": 24340 }, { "grad_norm": 0.2324650138616562, "learning_rate": 8.332987613824239e-05, "loss": 0.0281, "step": 24350 }, { "grad_norm": 0.2758481800556183, "learning_rate": 8.331446671753685e-05, "loss": 0.0278, "step": 24360 }, { "grad_norm": 0.2899630069732666, "learning_rate": 8.329905160430007e-05, "loss": 0.0152, "step": 24370 }, { "grad_norm": 0.21864722669124603, "learning_rate": 8.328363080116611e-05, "loss": 0.0186, "step": 24380 }, { "grad_norm": 0.35790783166885376, "learning_rate": 8.326820431076997e-05, "loss": 0.0168, "step": 24390 }, { "grad_norm": 0.21935933828353882, "learning_rate": 8.325277213574759e-05, "loss": 0.017, "step": 24400 }, { "grad_norm": 0.26554983854293823, "learning_rate": 8.32373342787359e-05, "loss": 0.0161, "step": 24410 }, { "grad_norm": 0.22757457196712494, "learning_rate": 8.322189074237285e-05, "loss": 0.0179, "step": 24420 }, { "grad_norm": 0.3252151906490326, "learning_rate": 8.32064415292973e-05, "loss": 0.0214, "step": 24430 }, { "grad_norm": 0.15627668797969818, "learning_rate": 8.319098664214907e-05, "loss": 0.0199, "step": 24440 }, { "grad_norm": 0.29760730266571045, "learning_rate": 8.3175526083569e-05, "loss": 0.0232, "step": 24450 }, { "grad_norm": 0.41905704140663147, "learning_rate": 8.316005985619889e-05, "loss": 0.0233, "step": 24460 }, { "grad_norm": 0.22960175573825836, "learning_rate": 8.314458796268147e-05, "loss": 0.0214, "step": 24470 }, { "grad_norm": 0.2409752905368805, "learning_rate": 8.312911040566047e-05, "loss": 0.0189, "step": 24480 }, { "grad_norm": 0.28986042737960815, "learning_rate": 8.31136271877806e-05, "loss": 0.0204, "step": 24490 }, { "grad_norm": 0.38849765062332153, "learning_rate": 8.309813831168748e-05, "loss": 0.0374, "step": 24500 }, { "grad_norm": 0.26755478978157043, "learning_rate": 8.308264378002777e-05, "loss": 0.019, "step": 24510 }, { "grad_norm": 0.2637883126735687, "learning_rate": 8.306714359544906e-05, "loss": 0.0248, "step": 24520 }, { "grad_norm": 0.16797871887683868, "learning_rate": 8.30516377605999e-05, "loss": 0.0171, "step": 24530 }, { "grad_norm": 0.32446637749671936, "learning_rate": 8.30361262781298e-05, "loss": 0.0132, "step": 24540 }, { "grad_norm": 0.24436712265014648, "learning_rate": 8.302060915068924e-05, "loss": 0.0225, "step": 24550 }, { "grad_norm": 0.27362966537475586, "learning_rate": 8.300508638092972e-05, "loss": 0.0132, "step": 24560 }, { "grad_norm": 0.33652055263519287, "learning_rate": 8.298955797150361e-05, "loss": 0.0193, "step": 24570 }, { "grad_norm": 0.3007338345050812, "learning_rate": 8.297402392506433e-05, "loss": 0.017, "step": 24580 }, { "grad_norm": 0.3219527006149292, "learning_rate": 8.295848424426617e-05, "loss": 0.0189, "step": 24590 }, { "grad_norm": 0.2170255482196808, "learning_rate": 8.29429389317645e-05, "loss": 0.0158, "step": 24600 }, { "grad_norm": 0.28203821182250977, "learning_rate": 8.292738799021556e-05, "loss": 0.0244, "step": 24610 }, { "grad_norm": 0.4656982719898224, "learning_rate": 8.291183142227656e-05, "loss": 0.0176, "step": 24620 }, { "grad_norm": 0.23034143447875977, "learning_rate": 8.289626923060572e-05, "loss": 0.0304, "step": 24630 }, { "grad_norm": 0.37665972113609314, "learning_rate": 8.288070141786218e-05, "loss": 0.0237, "step": 24640 }, { "grad_norm": 0.29806095361709595, "learning_rate": 8.286512798670605e-05, "loss": 0.0221, "step": 24650 }, { "grad_norm": 0.24399499595165253, "learning_rate": 8.284954893979842e-05, "loss": 0.0194, "step": 24660 }, { "grad_norm": 0.17363958060741425, "learning_rate": 8.283396427980131e-05, "loss": 0.0245, "step": 24670 }, { "grad_norm": 0.26061832904815674, "learning_rate": 8.281837400937771e-05, "loss": 0.0205, "step": 24680 }, { "grad_norm": 0.36477282643318176, "learning_rate": 8.28027781311916e-05, "loss": 0.0206, "step": 24690 }, { "grad_norm": 0.27645012736320496, "learning_rate": 8.278717664790785e-05, "loss": 0.0239, "step": 24700 }, { "grad_norm": 0.37614670395851135, "learning_rate": 8.277156956219234e-05, "loss": 0.0212, "step": 24710 }, { "grad_norm": 0.29524171352386475, "learning_rate": 8.275595687671189e-05, "loss": 0.0202, "step": 24720 }, { "grad_norm": 0.3329417109489441, "learning_rate": 8.27403385941343e-05, "loss": 0.0197, "step": 24730 }, { "grad_norm": 0.30340418219566345, "learning_rate": 8.272471471712828e-05, "loss": 0.0209, "step": 24740 }, { "grad_norm": 0.24184922873973846, "learning_rate": 8.270908524836355e-05, "loss": 0.0231, "step": 24750 }, { "grad_norm": 0.3315296173095703, "learning_rate": 8.269345019051074e-05, "loss": 0.0257, "step": 24760 }, { "grad_norm": 0.51337069272995, "learning_rate": 8.267780954624147e-05, "loss": 0.0268, "step": 24770 }, { "grad_norm": 0.20838041603565216, "learning_rate": 8.266216331822827e-05, "loss": 0.0223, "step": 24780 }, { "grad_norm": 0.36887916922569275, "learning_rate": 8.264651150914469e-05, "loss": 0.0208, "step": 24790 }, { "grad_norm": 0.21538299322128296, "learning_rate": 8.263085412166517e-05, "loss": 0.0148, "step": 24800 }, { "grad_norm": 0.1845175176858902, "learning_rate": 8.261519115846514e-05, "loss": 0.0176, "step": 24810 }, { "grad_norm": 0.5811129212379456, "learning_rate": 8.259952262222096e-05, "loss": 0.0201, "step": 24820 }, { "grad_norm": 0.22741280496120453, "learning_rate": 8.258384851560997e-05, "loss": 0.0172, "step": 24830 }, { "grad_norm": 0.3004263639450073, "learning_rate": 8.256816884131044e-05, "loss": 0.0229, "step": 24840 }, { "grad_norm": 0.5004527568817139, "learning_rate": 8.255248360200159e-05, "loss": 0.0222, "step": 24850 }, { "grad_norm": 0.26638996601104736, "learning_rate": 8.253679280036359e-05, "loss": 0.0207, "step": 24860 }, { "grad_norm": 0.3552076518535614, "learning_rate": 8.252109643907762e-05, "loss": 0.0303, "step": 24870 }, { "grad_norm": 0.31367576122283936, "learning_rate": 8.250539452082569e-05, "loss": 0.0191, "step": 24880 }, { "grad_norm": 0.32391858100891113, "learning_rate": 8.248968704829087e-05, "loss": 0.0196, "step": 24890 }, { "grad_norm": 0.5432723164558411, "learning_rate": 8.247397402415714e-05, "loss": 0.0232, "step": 24900 }, { "grad_norm": 0.2887305021286011, "learning_rate": 8.24582554511094e-05, "loss": 0.0194, "step": 24910 }, { "grad_norm": 0.1807195395231247, "learning_rate": 8.244253133183355e-05, "loss": 0.0188, "step": 24920 }, { "grad_norm": 0.2619110345840454, "learning_rate": 8.24268016690164e-05, "loss": 0.0262, "step": 24930 }, { "grad_norm": 0.4214327931404114, "learning_rate": 8.241106646534571e-05, "loss": 0.0223, "step": 24940 }, { "grad_norm": 0.3041341006755829, "learning_rate": 8.23953257235102e-05, "loss": 0.0197, "step": 24950 }, { "grad_norm": 0.31668123602867126, "learning_rate": 8.237957944619956e-05, "loss": 0.0224, "step": 24960 }, { "grad_norm": 0.29945293068885803, "learning_rate": 8.236382763610437e-05, "loss": 0.0185, "step": 24970 }, { "grad_norm": 0.33029577136039734, "learning_rate": 8.234807029591619e-05, "loss": 0.021, "step": 24980 }, { "grad_norm": 0.2254749983549118, "learning_rate": 8.233230742832752e-05, "loss": 0.0313, "step": 24990 }, { "grad_norm": 0.28923189640045166, "learning_rate": 8.231653903603178e-05, "loss": 0.0188, "step": 25000 }, { "grad_norm": 0.23617051541805267, "learning_rate": 8.23007651217234e-05, "loss": 0.0171, "step": 25010 }, { "grad_norm": 0.6354823112487793, "learning_rate": 8.228498568809769e-05, "loss": 0.0223, "step": 25020 }, { "grad_norm": 0.36551642417907715, "learning_rate": 8.22692007378509e-05, "loss": 0.0322, "step": 25030 }, { "grad_norm": 0.2956200838088989, "learning_rate": 8.225341027368028e-05, "loss": 0.0213, "step": 25040 }, { "grad_norm": 0.2620048224925995, "learning_rate": 8.223761429828399e-05, "loss": 0.0228, "step": 25050 }, { "grad_norm": 0.200799822807312, "learning_rate": 8.22218128143611e-05, "loss": 0.0239, "step": 25060 }, { "grad_norm": 0.3516783118247986, "learning_rate": 8.220600582461166e-05, "loss": 0.0198, "step": 25070 }, { "grad_norm": 0.2089529037475586, "learning_rate": 8.219019333173668e-05, "loss": 0.0188, "step": 25080 }, { "grad_norm": 0.30519670248031616, "learning_rate": 8.217437533843805e-05, "loss": 0.0239, "step": 25090 }, { "grad_norm": 0.2521950304508209, "learning_rate": 8.215855184741867e-05, "loss": 0.0221, "step": 25100 }, { "grad_norm": 0.9679659605026245, "learning_rate": 8.21427228613823e-05, "loss": 0.0281, "step": 25110 }, { "grad_norm": 0.2867315411567688, "learning_rate": 8.21268883830337e-05, "loss": 0.0291, "step": 25120 }, { "grad_norm": 0.3115343749523163, "learning_rate": 8.211104841507855e-05, "loss": 0.0323, "step": 25130 }, { "grad_norm": 0.23080140352249146, "learning_rate": 8.209520296022346e-05, "loss": 0.0286, "step": 25140 }, { "grad_norm": 0.26643097400665283, "learning_rate": 8.207935202117599e-05, "loss": 0.0241, "step": 25150 }, { "grad_norm": 0.1962260603904724, "learning_rate": 8.206349560064463e-05, "loss": 0.0212, "step": 25160 }, { "grad_norm": 0.34213146567344666, "learning_rate": 8.204763370133881e-05, "loss": 0.0151, "step": 25170 }, { "grad_norm": 0.22995232045650482, "learning_rate": 8.203176632596892e-05, "loss": 0.0193, "step": 25180 }, { "grad_norm": 0.24950946867465973, "learning_rate": 8.20158934772462e-05, "loss": 0.0217, "step": 25190 }, { "grad_norm": 0.3513086140155792, "learning_rate": 8.200001515788294e-05, "loss": 0.0203, "step": 25200 }, { "grad_norm": 0.26027560234069824, "learning_rate": 8.198413137059228e-05, "loss": 0.0226, "step": 25210 }, { "grad_norm": 0.18777264654636383, "learning_rate": 8.196824211808835e-05, "loss": 0.0168, "step": 25220 }, { "grad_norm": 0.28866079449653625, "learning_rate": 8.195234740308617e-05, "loss": 0.023, "step": 25230 }, { "grad_norm": 0.40640053153038025, "learning_rate": 8.193644722830171e-05, "loss": 0.0176, "step": 25240 }, { "grad_norm": 0.27026742696762085, "learning_rate": 8.19205415964519e-05, "loss": 0.0207, "step": 25250 }, { "grad_norm": 0.32923877239227295, "learning_rate": 8.190463051025456e-05, "loss": 0.0223, "step": 25260 }, { "grad_norm": 0.1916404366493225, "learning_rate": 8.188871397242843e-05, "loss": 0.0221, "step": 25270 }, { "grad_norm": 0.21517008543014526, "learning_rate": 8.187279198569326e-05, "loss": 0.0212, "step": 25280 }, { "grad_norm": 0.23651222884655, "learning_rate": 8.185686455276966e-05, "loss": 0.0209, "step": 25290 }, { "grad_norm": 0.17570394277572632, "learning_rate": 8.184093167637921e-05, "loss": 0.0172, "step": 25300 }, { "grad_norm": 0.34486448764801025, "learning_rate": 8.182499335924437e-05, "loss": 0.0206, "step": 25310 }, { "grad_norm": 0.28212565183639526, "learning_rate": 8.18090496040886e-05, "loss": 0.017, "step": 25320 }, { "grad_norm": 0.30003321170806885, "learning_rate": 8.179310041363621e-05, "loss": 0.0237, "step": 25330 }, { "grad_norm": 0.2614038586616516, "learning_rate": 8.17771457906125e-05, "loss": 0.0167, "step": 25340 }, { "grad_norm": 0.3118605613708496, "learning_rate": 8.176118573774371e-05, "loss": 0.017, "step": 25350 }, { "grad_norm": 0.34054115414619446, "learning_rate": 8.174522025775692e-05, "loss": 0.0193, "step": 25360 }, { "grad_norm": 0.4708649814128876, "learning_rate": 8.172924935338022e-05, "loss": 0.0211, "step": 25370 }, { "grad_norm": 0.31702372431755066, "learning_rate": 8.171327302734262e-05, "loss": 0.0182, "step": 25380 }, { "grad_norm": 0.2174578309059143, "learning_rate": 8.169729128237401e-05, "loss": 0.0156, "step": 25390 }, { "grad_norm": 0.21970833837985992, "learning_rate": 8.168130412120525e-05, "loss": 0.0164, "step": 25400 }, { "grad_norm": 0.3391003906726837, "learning_rate": 8.16653115465681e-05, "loss": 0.0194, "step": 25410 }, { "grad_norm": 0.29108691215515137, "learning_rate": 8.164931356119526e-05, "loss": 0.0147, "step": 25420 }, { "grad_norm": 0.33069372177124023, "learning_rate": 8.163331016782032e-05, "loss": 0.0253, "step": 25430 }, { "grad_norm": 0.26194775104522705, "learning_rate": 8.161730136917785e-05, "loss": 0.0135, "step": 25440 }, { "grad_norm": 0.377118319272995, "learning_rate": 8.160128716800333e-05, "loss": 0.0219, "step": 25450 }, { "grad_norm": 0.32569438219070435, "learning_rate": 8.158526756703313e-05, "loss": 0.0198, "step": 25460 }, { "grad_norm": 0.21143971383571625, "learning_rate": 8.156924256900455e-05, "loss": 0.0226, "step": 25470 }, { "grad_norm": 0.2361375242471695, "learning_rate": 8.155321217665584e-05, "loss": 0.0163, "step": 25480 }, { "grad_norm": 0.3624820113182068, "learning_rate": 8.153717639272614e-05, "loss": 0.0156, "step": 25490 }, { "grad_norm": 0.2305428683757782, "learning_rate": 8.152113521995555e-05, "loss": 0.0214, "step": 25500 }, { "grad_norm": 0.14925804734230042, "learning_rate": 8.150508866108505e-05, "loss": 0.0177, "step": 25510 }, { "grad_norm": 0.43241947889328003, "learning_rate": 8.148903671885657e-05, "loss": 0.0178, "step": 25520 }, { "grad_norm": 0.2753165662288666, "learning_rate": 8.147297939601292e-05, "loss": 0.0263, "step": 25530 }, { "grad_norm": 0.3250594735145569, "learning_rate": 8.145691669529792e-05, "loss": 0.0218, "step": 25540 }, { "grad_norm": 0.2793024778366089, "learning_rate": 8.144084861945618e-05, "loss": 0.0181, "step": 25550 }, { "grad_norm": 0.25945308804512024, "learning_rate": 8.142477517123333e-05, "loss": 0.019, "step": 25560 }, { "grad_norm": 0.41804951429367065, "learning_rate": 8.140869635337586e-05, "loss": 0.0205, "step": 25570 }, { "grad_norm": 0.2192186415195465, "learning_rate": 8.139261216863123e-05, "loss": 0.0215, "step": 25580 }, { "grad_norm": 0.22642092406749725, "learning_rate": 8.137652261974776e-05, "loss": 0.0184, "step": 25590 }, { "grad_norm": 0.3571400046348572, "learning_rate": 8.136042770947472e-05, "loss": 0.0215, "step": 25600 }, { "grad_norm": 0.16496755182743073, "learning_rate": 8.134432744056228e-05, "loss": 0.0206, "step": 25610 }, { "grad_norm": 0.3079025447368622, "learning_rate": 8.132822181576158e-05, "loss": 0.0245, "step": 25620 }, { "grad_norm": 0.5990504026412964, "learning_rate": 8.131211083782459e-05, "loss": 0.0231, "step": 25630 }, { "grad_norm": 0.2785904109477997, "learning_rate": 8.129599450950424e-05, "loss": 0.0211, "step": 25640 }, { "grad_norm": 0.34893420338630676, "learning_rate": 8.127987283355438e-05, "loss": 0.0249, "step": 25650 }, { "grad_norm": 0.20694296061992645, "learning_rate": 8.126374581272976e-05, "loss": 0.0138, "step": 25660 }, { "grad_norm": 0.24560612440109253, "learning_rate": 8.124761344978605e-05, "loss": 0.0259, "step": 25670 }, { "grad_norm": 0.24143144488334656, "learning_rate": 8.12314757474798e-05, "loss": 0.0171, "step": 25680 }, { "grad_norm": 0.1599140763282776, "learning_rate": 8.121533270856856e-05, "loss": 0.021, "step": 25690 }, { "grad_norm": 0.3737129867076874, "learning_rate": 8.119918433581069e-05, "loss": 0.0216, "step": 25700 }, { "grad_norm": 0.31050440669059753, "learning_rate": 8.118303063196551e-05, "loss": 0.018, "step": 25710 }, { "grad_norm": 0.25102221965789795, "learning_rate": 8.116687159979326e-05, "loss": 0.0171, "step": 25720 }, { "grad_norm": 0.29690369963645935, "learning_rate": 8.115070724205508e-05, "loss": 0.0271, "step": 25730 }, { "grad_norm": 0.2894830107688904, "learning_rate": 8.113453756151296e-05, "loss": 0.0188, "step": 25740 }, { "grad_norm": 0.23980213701725006, "learning_rate": 8.111836256092995e-05, "loss": 0.0204, "step": 25750 }, { "grad_norm": 0.19439952075481415, "learning_rate": 8.110218224306985e-05, "loss": 0.0217, "step": 25760 }, { "grad_norm": 0.2596328854560852, "learning_rate": 8.108599661069745e-05, "loss": 0.0194, "step": 25770 }, { "grad_norm": 0.2067980021238327, "learning_rate": 8.106980566657845e-05, "loss": 0.0188, "step": 25780 }, { "grad_norm": 0.19738776981830597, "learning_rate": 8.10536094134794e-05, "loss": 0.0168, "step": 25790 }, { "grad_norm": 0.19735084474086761, "learning_rate": 8.103740785416783e-05, "loss": 0.0177, "step": 25800 }, { "grad_norm": 0.393203467130661, "learning_rate": 8.102120099141212e-05, "loss": 0.0192, "step": 25810 }, { "grad_norm": 0.424282044172287, "learning_rate": 8.100498882798163e-05, "loss": 0.022, "step": 25820 }, { "grad_norm": 0.3216343820095062, "learning_rate": 8.09887713666465e-05, "loss": 0.0222, "step": 25830 }, { "grad_norm": 0.32584065198898315, "learning_rate": 8.09725486101779e-05, "loss": 0.0224, "step": 25840 }, { "grad_norm": 0.27167248725891113, "learning_rate": 8.095632056134784e-05, "loss": 0.0258, "step": 25850 }, { "grad_norm": 0.7303150296211243, "learning_rate": 8.094008722292925e-05, "loss": 0.0185, "step": 25860 }, { "grad_norm": 0.18767932057380676, "learning_rate": 8.092384859769598e-05, "loss": 0.0206, "step": 25870 }, { "grad_norm": 0.295819491147995, "learning_rate": 8.090760468842275e-05, "loss": 0.0263, "step": 25880 }, { "grad_norm": 0.18262165784835815, "learning_rate": 8.089135549788521e-05, "loss": 0.0162, "step": 25890 }, { "grad_norm": 0.2906375825405121, "learning_rate": 8.087510102885987e-05, "loss": 0.0179, "step": 25900 }, { "grad_norm": 0.2235007882118225, "learning_rate": 8.085884128412422e-05, "loss": 0.0288, "step": 25910 }, { "grad_norm": 0.3792654871940613, "learning_rate": 8.084257626645659e-05, "loss": 0.0226, "step": 25920 }, { "grad_norm": 0.5611668825149536, "learning_rate": 8.08263059786362e-05, "loss": 0.0219, "step": 25930 }, { "grad_norm": 0.280480295419693, "learning_rate": 8.081003042344325e-05, "loss": 0.0181, "step": 25940 }, { "grad_norm": 0.2298857420682907, "learning_rate": 8.079374960365872e-05, "loss": 0.0222, "step": 25950 }, { "grad_norm": 0.2985594868659973, "learning_rate": 8.077746352206463e-05, "loss": 0.0251, "step": 25960 }, { "grad_norm": 0.2674814462661743, "learning_rate": 8.076117218144377e-05, "loss": 0.0196, "step": 25970 }, { "grad_norm": 0.3998815417289734, "learning_rate": 8.074487558457991e-05, "loss": 0.0224, "step": 25980 }, { "grad_norm": 0.2605524957180023, "learning_rate": 8.072857373425768e-05, "loss": 0.0261, "step": 25990 }, { "grad_norm": 0.4000841975212097, "learning_rate": 8.071226663326264e-05, "loss": 0.0191, "step": 26000 }, { "grad_norm": 0.36379900574684143, "learning_rate": 8.069595428438121e-05, "loss": 0.0255, "step": 26010 }, { "grad_norm": 0.24646404385566711, "learning_rate": 8.067963669040072e-05, "loss": 0.025, "step": 26020 }, { "grad_norm": 0.2819545269012451, "learning_rate": 8.066331385410942e-05, "loss": 0.0197, "step": 26030 }, { "grad_norm": 0.2891572415828705, "learning_rate": 8.064698577829641e-05, "loss": 0.0224, "step": 26040 }, { "grad_norm": 0.28667518496513367, "learning_rate": 8.063065246575175e-05, "loss": 0.0222, "step": 26050 }, { "grad_norm": 0.2319047749042511, "learning_rate": 8.061431391926631e-05, "loss": 0.0198, "step": 26060 }, { "grad_norm": 0.24075089395046234, "learning_rate": 8.059797014163195e-05, "loss": 0.0212, "step": 26070 }, { "grad_norm": 0.487061083316803, "learning_rate": 8.058162113564133e-05, "loss": 0.0305, "step": 26080 }, { "grad_norm": 0.28968340158462524, "learning_rate": 8.056526690408806e-05, "loss": 0.0247, "step": 26090 }, { "grad_norm": 0.3252113163471222, "learning_rate": 8.054890744976666e-05, "loss": 0.0227, "step": 26100 }, { "grad_norm": 0.2819698750972748, "learning_rate": 8.053254277547248e-05, "loss": 0.0183, "step": 26110 }, { "grad_norm": 0.26827579736709595, "learning_rate": 8.051617288400182e-05, "loss": 0.0191, "step": 26120 }, { "grad_norm": 0.19216202199459076, "learning_rate": 8.049979777815182e-05, "loss": 0.0197, "step": 26130 }, { "grad_norm": 0.2156507968902588, "learning_rate": 8.048341746072054e-05, "loss": 0.0195, "step": 26140 }, { "grad_norm": 0.3184807300567627, "learning_rate": 8.046703193450696e-05, "loss": 0.024, "step": 26150 }, { "grad_norm": 0.3085029125213623, "learning_rate": 8.04506412023109e-05, "loss": 0.0215, "step": 26160 }, { "grad_norm": 0.23038165271282196, "learning_rate": 8.043424526693306e-05, "loss": 0.0218, "step": 26170 }, { "grad_norm": 0.3011395037174225, "learning_rate": 8.04178441311751e-05, "loss": 0.0217, "step": 26180 }, { "grad_norm": 0.37185126543045044, "learning_rate": 8.04014377978395e-05, "loss": 0.0188, "step": 26190 }, { "grad_norm": 0.2700944244861603, "learning_rate": 8.038502626972967e-05, "loss": 0.0269, "step": 26200 }, { "grad_norm": 0.23819169402122498, "learning_rate": 8.036860954964989e-05, "loss": 0.0213, "step": 26210 }, { "grad_norm": 0.41704198718070984, "learning_rate": 8.035218764040531e-05, "loss": 0.022, "step": 26220 }, { "grad_norm": 0.3433437943458557, "learning_rate": 8.033576054480199e-05, "loss": 0.0229, "step": 26230 }, { "grad_norm": 0.3627871870994568, "learning_rate": 8.031932826564688e-05, "loss": 0.0216, "step": 26240 }, { "grad_norm": 0.2723202109336853, "learning_rate": 8.030289080574782e-05, "loss": 0.0204, "step": 26250 }, { "grad_norm": 0.2815989553928375, "learning_rate": 8.028644816791349e-05, "loss": 0.0257, "step": 26260 }, { "grad_norm": 0.35010701417922974, "learning_rate": 8.027000035495351e-05, "loss": 0.0179, "step": 26270 }, { "grad_norm": 0.24994643032550812, "learning_rate": 8.025354736967836e-05, "loss": 0.0325, "step": 26280 }, { "grad_norm": 0.20446687936782837, "learning_rate": 8.023708921489941e-05, "loss": 0.0278, "step": 26290 }, { "grad_norm": 0.30494022369384766, "learning_rate": 8.022062589342887e-05, "loss": 0.0326, "step": 26300 }, { "grad_norm": 0.3066459596157074, "learning_rate": 8.020415740807993e-05, "loss": 0.0288, "step": 26310 }, { "grad_norm": 0.4017309248447418, "learning_rate": 8.018768376166656e-05, "loss": 0.0199, "step": 26320 }, { "grad_norm": 0.46581128239631653, "learning_rate": 8.017120495700368e-05, "loss": 0.0203, "step": 26330 }, { "grad_norm": 0.42507511377334595, "learning_rate": 8.015472099690704e-05, "loss": 0.0226, "step": 26340 }, { "grad_norm": 0.20134082436561584, "learning_rate": 8.013823188419332e-05, "loss": 0.0157, "step": 26350 }, { "grad_norm": 0.35020574927330017, "learning_rate": 8.012173762168006e-05, "loss": 0.0249, "step": 26360 }, { "grad_norm": 0.24426797032356262, "learning_rate": 8.010523821218567e-05, "loss": 0.0176, "step": 26370 }, { "grad_norm": 0.20803868770599365, "learning_rate": 8.008873365852945e-05, "loss": 0.0211, "step": 26380 }, { "grad_norm": 0.27664995193481445, "learning_rate": 8.007222396353157e-05, "loss": 0.0256, "step": 26390 }, { "grad_norm": 0.2981477379798889, "learning_rate": 8.00557091300131e-05, "loss": 0.0222, "step": 26400 }, { "grad_norm": 0.22433608770370483, "learning_rate": 8.003918916079597e-05, "loss": 0.0202, "step": 26410 }, { "grad_norm": 0.2614423632621765, "learning_rate": 8.002266405870298e-05, "loss": 0.0132, "step": 26420 }, { "grad_norm": 0.2629789412021637, "learning_rate": 8.000613382655782e-05, "loss": 0.0215, "step": 26430 }, { "grad_norm": 0.2730265259742737, "learning_rate": 7.998959846718505e-05, "loss": 0.0219, "step": 26440 }, { "grad_norm": 0.2579699158668518, "learning_rate": 7.997305798341012e-05, "loss": 0.0225, "step": 26450 }, { "grad_norm": 0.22797498106956482, "learning_rate": 7.995651237805937e-05, "loss": 0.0226, "step": 26460 }, { "grad_norm": 0.3593367636203766, "learning_rate": 7.993996165395996e-05, "loss": 0.0153, "step": 26470 }, { "grad_norm": 0.4373835027217865, "learning_rate": 7.992340581393996e-05, "loss": 0.0203, "step": 26480 }, { "grad_norm": 0.24217084050178528, "learning_rate": 7.990684486082831e-05, "loss": 0.0182, "step": 26490 }, { "grad_norm": 0.38580521941185, "learning_rate": 7.989027879745482e-05, "loss": 0.0235, "step": 26500 }, { "grad_norm": 0.25521233677864075, "learning_rate": 7.98737076266502e-05, "loss": 0.0276, "step": 26510 }, { "grad_norm": 0.18104882538318634, "learning_rate": 7.985713135124598e-05, "loss": 0.0195, "step": 26520 }, { "grad_norm": 0.31318753957748413, "learning_rate": 7.98405499740746e-05, "loss": 0.0187, "step": 26530 }, { "grad_norm": 0.2613826394081116, "learning_rate": 7.98239634979694e-05, "loss": 0.0188, "step": 26540 }, { "grad_norm": 0.291973352432251, "learning_rate": 7.98073719257645e-05, "loss": 0.0154, "step": 26550 }, { "grad_norm": 0.40452349185943604, "learning_rate": 7.979077526029499e-05, "loss": 0.0227, "step": 26560 }, { "grad_norm": 0.17577427625656128, "learning_rate": 7.977417350439675e-05, "loss": 0.0143, "step": 26570 }, { "grad_norm": 0.24201564490795135, "learning_rate": 7.97575666609066e-05, "loss": 0.0197, "step": 26580 }, { "grad_norm": 0.31104710698127747, "learning_rate": 7.974095473266216e-05, "loss": 0.0179, "step": 26590 }, { "grad_norm": 0.30180659890174866, "learning_rate": 7.972433772250198e-05, "loss": 0.0249, "step": 26600 }, { "grad_norm": 0.49309778213500977, "learning_rate": 7.970771563326544e-05, "loss": 0.0166, "step": 26610 }, { "grad_norm": 0.23417672514915466, "learning_rate": 7.96910884677928e-05, "loss": 0.0203, "step": 26620 }, { "grad_norm": 0.3380349278450012, "learning_rate": 7.967445622892523e-05, "loss": 0.0253, "step": 26630 }, { "grad_norm": 0.2118843048810959, "learning_rate": 7.965781891950465e-05, "loss": 0.0172, "step": 26640 }, { "grad_norm": 0.22063207626342773, "learning_rate": 7.964117654237397e-05, "loss": 0.0149, "step": 26650 }, { "grad_norm": 0.3531607985496521, "learning_rate": 7.962452910037692e-05, "loss": 0.0225, "step": 26660 }, { "grad_norm": 0.34568172693252563, "learning_rate": 7.96078765963581e-05, "loss": 0.0218, "step": 26670 }, { "grad_norm": 0.2962368130683899, "learning_rate": 7.95912190331629e-05, "loss": 0.0205, "step": 26680 }, { "grad_norm": 0.21283529698848724, "learning_rate": 7.957455641363772e-05, "loss": 0.0203, "step": 26690 }, { "grad_norm": 0.29331716895103455, "learning_rate": 7.955788874062968e-05, "loss": 0.0253, "step": 26700 }, { "grad_norm": 0.381337970495224, "learning_rate": 7.95412160169869e-05, "loss": 0.0239, "step": 26710 }, { "grad_norm": 0.34690043330192566, "learning_rate": 7.952453824555824e-05, "loss": 0.0172, "step": 26720 }, { "grad_norm": 0.21735769510269165, "learning_rate": 7.95078554291935e-05, "loss": 0.0134, "step": 26730 }, { "grad_norm": 0.2649267315864563, "learning_rate": 7.94911675707433e-05, "loss": 0.0176, "step": 26740 }, { "grad_norm": 0.26249635219573975, "learning_rate": 7.947447467305915e-05, "loss": 0.0222, "step": 26750 }, { "grad_norm": 0.1621110588312149, "learning_rate": 7.94577767389934e-05, "loss": 0.0185, "step": 26760 }, { "grad_norm": 0.1933879852294922, "learning_rate": 7.944107377139928e-05, "loss": 0.0175, "step": 26770 }, { "grad_norm": 0.36330658197402954, "learning_rate": 7.942436577313088e-05, "loss": 0.0219, "step": 26780 }, { "grad_norm": 0.36706992983818054, "learning_rate": 7.940765274704312e-05, "loss": 0.0252, "step": 26790 }, { "grad_norm": 0.40621107816696167, "learning_rate": 7.939093469599181e-05, "loss": 0.0265, "step": 26800 }, { "grad_norm": 0.22648018598556519, "learning_rate": 7.93742116228336e-05, "loss": 0.0203, "step": 26810 }, { "grad_norm": 0.39709118008613586, "learning_rate": 7.935748353042602e-05, "loss": 0.0243, "step": 26820 }, { "grad_norm": 0.2451341301202774, "learning_rate": 7.934075042162744e-05, "loss": 0.0205, "step": 26830 }, { "grad_norm": 0.3720207214355469, "learning_rate": 7.932401229929705e-05, "loss": 0.0227, "step": 26840 }, { "grad_norm": 0.2630385160446167, "learning_rate": 7.9307269166295e-05, "loss": 0.0232, "step": 26850 }, { "grad_norm": 0.20876957476139069, "learning_rate": 7.92905210254822e-05, "loss": 0.0175, "step": 26860 }, { "grad_norm": 0.25898072123527527, "learning_rate": 7.927376787972045e-05, "loss": 0.0151, "step": 26870 }, { "grad_norm": 0.22996215522289276, "learning_rate": 7.92570097318724e-05, "loss": 0.0224, "step": 26880 }, { "grad_norm": 0.3102685809135437, "learning_rate": 7.924024658480158e-05, "loss": 0.0184, "step": 26890 }, { "grad_norm": 0.333110511302948, "learning_rate": 7.922347844137233e-05, "loss": 0.0167, "step": 26900 }, { "grad_norm": 0.23957529664039612, "learning_rate": 7.92067053044499e-05, "loss": 0.0223, "step": 26910 }, { "grad_norm": 0.42017054557800293, "learning_rate": 7.918992717690031e-05, "loss": 0.0231, "step": 26920 }, { "grad_norm": 0.26834261417388916, "learning_rate": 7.917314406159053e-05, "loss": 0.0215, "step": 26930 }, { "grad_norm": 0.49077433347702026, "learning_rate": 7.915635596138832e-05, "loss": 0.0312, "step": 26940 }, { "grad_norm": 0.20037707686424255, "learning_rate": 7.913956287916228e-05, "loss": 0.0198, "step": 26950 }, { "grad_norm": 0.3135901391506195, "learning_rate": 7.912276481778193e-05, "loss": 0.0178, "step": 26960 }, { "grad_norm": 0.2563845217227936, "learning_rate": 7.910596178011759e-05, "loss": 0.0227, "step": 26970 }, { "grad_norm": 0.17428338527679443, "learning_rate": 7.908915376904043e-05, "loss": 0.0167, "step": 26980 }, { "grad_norm": 0.2633456885814667, "learning_rate": 7.907234078742247e-05, "loss": 0.0201, "step": 26990 }, { "grad_norm": 0.5089501142501831, "learning_rate": 7.90555228381366e-05, "loss": 0.0188, "step": 27000 }, { "grad_norm": 0.3636196553707123, "learning_rate": 7.903869992405656e-05, "loss": 0.0162, "step": 27010 }, { "grad_norm": 0.4846072196960449, "learning_rate": 7.902187204805691e-05, "loss": 0.0218, "step": 27020 }, { "grad_norm": 0.30610567331314087, "learning_rate": 7.900503921301308e-05, "loss": 0.0168, "step": 27030 }, { "grad_norm": 0.26181039214134216, "learning_rate": 7.898820142180133e-05, "loss": 0.0256, "step": 27040 }, { "grad_norm": 0.23646153509616852, "learning_rate": 7.897135867729879e-05, "loss": 0.0192, "step": 27050 }, { "grad_norm": 0.2095041126012802, "learning_rate": 7.89545109823834e-05, "loss": 0.0165, "step": 27060 }, { "grad_norm": 0.21874770522117615, "learning_rate": 7.8937658339934e-05, "loss": 0.0137, "step": 27070 }, { "grad_norm": 0.24229517579078674, "learning_rate": 7.892080075283026e-05, "loss": 0.0157, "step": 27080 }, { "grad_norm": 0.1790338158607483, "learning_rate": 7.890393822395263e-05, "loss": 0.0176, "step": 27090 }, { "grad_norm": 0.1928258240222931, "learning_rate": 7.88870707561825e-05, "loss": 0.0144, "step": 27100 }, { "grad_norm": 0.3536090850830078, "learning_rate": 7.887019835240203e-05, "loss": 0.0211, "step": 27110 }, { "grad_norm": 0.30683326721191406, "learning_rate": 7.885332101549427e-05, "loss": 0.0194, "step": 27120 }, { "grad_norm": 0.24485377967357635, "learning_rate": 7.883643874834308e-05, "loss": 0.0184, "step": 27130 }, { "grad_norm": 0.32176733016967773, "learning_rate": 7.881955155383321e-05, "loss": 0.0249, "step": 27140 }, { "grad_norm": 0.2905500531196594, "learning_rate": 7.880265943485017e-05, "loss": 0.0186, "step": 27150 }, { "grad_norm": 0.13109396398067474, "learning_rate": 7.878576239428038e-05, "loss": 0.0183, "step": 27160 }, { "grad_norm": 0.2667747437953949, "learning_rate": 7.87688604350111e-05, "loss": 0.0263, "step": 27170 }, { "grad_norm": 0.38821855187416077, "learning_rate": 7.875195355993042e-05, "loss": 0.0188, "step": 27180 }, { "grad_norm": 0.17038226127624512, "learning_rate": 7.873504177192724e-05, "loss": 0.0236, "step": 27190 }, { "grad_norm": 0.30217695236206055, "learning_rate": 7.87181250738913e-05, "loss": 0.0228, "step": 27200 }, { "grad_norm": 0.2605116367340088, "learning_rate": 7.870120346871324e-05, "loss": 0.0193, "step": 27210 }, { "grad_norm": 0.2608157694339752, "learning_rate": 7.86842769592845e-05, "loss": 0.0187, "step": 27220 }, { "grad_norm": 0.2717829644680023, "learning_rate": 7.866734554849732e-05, "loss": 0.0168, "step": 27230 }, { "grad_norm": 0.23833128809928894, "learning_rate": 7.865040923924486e-05, "loss": 0.0174, "step": 27240 }, { "grad_norm": 0.3884611129760742, "learning_rate": 7.863346803442104e-05, "loss": 0.0225, "step": 27250 }, { "grad_norm": 0.2896218001842499, "learning_rate": 7.861652193692067e-05, "loss": 0.0178, "step": 27260 }, { "grad_norm": 0.24975943565368652, "learning_rate": 7.859957094963937e-05, "loss": 0.0127, "step": 27270 }, { "grad_norm": 0.19674886763095856, "learning_rate": 7.858261507547357e-05, "loss": 0.0155, "step": 27280 }, { "grad_norm": 0.23782069981098175, "learning_rate": 7.856565431732061e-05, "loss": 0.018, "step": 27290 }, { "grad_norm": 0.21505291759967804, "learning_rate": 7.854868867807859e-05, "loss": 0.0189, "step": 27300 }, { "grad_norm": 0.42873185873031616, "learning_rate": 7.85317181606465e-05, "loss": 0.0217, "step": 27310 }, { "grad_norm": 0.20647193491458893, "learning_rate": 7.85147427679241e-05, "loss": 0.017, "step": 27320 }, { "grad_norm": 0.18658412992954254, "learning_rate": 7.849776250281205e-05, "loss": 0.0116, "step": 27330 }, { "grad_norm": 0.1991843432188034, "learning_rate": 7.84807773682118e-05, "loss": 0.0172, "step": 27340 }, { "grad_norm": 0.17840301990509033, "learning_rate": 7.846378736702565e-05, "loss": 0.0191, "step": 27350 }, { "grad_norm": 0.963623583316803, "learning_rate": 7.844679250215671e-05, "loss": 0.0248, "step": 27360 }, { "grad_norm": 0.19066570699214935, "learning_rate": 7.842979277650898e-05, "loss": 0.0197, "step": 27370 }, { "grad_norm": 0.3172701895236969, "learning_rate": 7.84127881929872e-05, "loss": 0.0189, "step": 27380 }, { "grad_norm": 0.24550004303455353, "learning_rate": 7.839577875449704e-05, "loss": 0.0222, "step": 27390 }, { "grad_norm": 0.21051159501075745, "learning_rate": 7.837876446394489e-05, "loss": 0.0199, "step": 27400 }, { "grad_norm": 0.23020252585411072, "learning_rate": 7.836174532423805e-05, "loss": 0.0258, "step": 27410 }, { "grad_norm": 0.2972840964794159, "learning_rate": 7.834472133828466e-05, "loss": 0.0168, "step": 27420 }, { "grad_norm": 0.4126165211200714, "learning_rate": 7.832769250899359e-05, "loss": 0.0255, "step": 27430 }, { "grad_norm": 0.3085152208805084, "learning_rate": 7.831065883927464e-05, "loss": 0.0205, "step": 27440 }, { "grad_norm": 0.31143906712532043, "learning_rate": 7.829362033203841e-05, "loss": 0.02, "step": 27450 }, { "grad_norm": 0.35126224160194397, "learning_rate": 7.827657699019628e-05, "loss": 0.023, "step": 27460 }, { "grad_norm": 0.22938603162765503, "learning_rate": 7.825952881666052e-05, "loss": 0.0246, "step": 27470 }, { "grad_norm": 0.33396121859550476, "learning_rate": 7.824247581434418e-05, "loss": 0.0247, "step": 27480 }, { "grad_norm": 0.30851104855537415, "learning_rate": 7.822541798616116e-05, "loss": 0.02, "step": 27490 }, { "grad_norm": 0.43156322836875916, "learning_rate": 7.820835533502617e-05, "loss": 0.0272, "step": 27500 }, { "grad_norm": 0.3573181629180908, "learning_rate": 7.819128786385475e-05, "loss": 0.0233, "step": 27510 }, { "grad_norm": 0.3348078429698944, "learning_rate": 7.817421557556329e-05, "loss": 0.0246, "step": 27520 }, { "grad_norm": 0.24050259590148926, "learning_rate": 7.815713847306893e-05, "loss": 0.0282, "step": 27530 }, { "grad_norm": 0.31676703691482544, "learning_rate": 7.81400565592897e-05, "loss": 0.0195, "step": 27540 }, { "grad_norm": 0.3050762712955475, "learning_rate": 7.812296983714444e-05, "loss": 0.0227, "step": 27550 }, { "grad_norm": 0.19715599715709686, "learning_rate": 7.810587830955281e-05, "loss": 0.0174, "step": 27560 }, { "grad_norm": 0.2353196144104004, "learning_rate": 7.808878197943528e-05, "loss": 0.0207, "step": 27570 }, { "grad_norm": 0.21633973717689514, "learning_rate": 7.807168084971312e-05, "loss": 0.0184, "step": 27580 }, { "grad_norm": 0.3301539719104767, "learning_rate": 7.805457492330849e-05, "loss": 0.0203, "step": 27590 }, { "grad_norm": 0.2548258304595947, "learning_rate": 7.803746420314428e-05, "loss": 0.0219, "step": 27600 }, { "grad_norm": 0.27844053506851196, "learning_rate": 7.802034869214428e-05, "loss": 0.0265, "step": 27610 }, { "grad_norm": 0.30753016471862793, "learning_rate": 7.800322839323303e-05, "loss": 0.0186, "step": 27620 }, { "grad_norm": 0.14943960309028625, "learning_rate": 7.798610330933593e-05, "loss": 0.0194, "step": 27630 }, { "grad_norm": 0.17586427927017212, "learning_rate": 7.796897344337922e-05, "loss": 0.0178, "step": 27640 }, { "grad_norm": 0.29924532771110535, "learning_rate": 7.795183879828989e-05, "loss": 0.0217, "step": 27650 }, { "grad_norm": 0.2545079290866852, "learning_rate": 7.793469937699579e-05, "loss": 0.0277, "step": 27660 }, { "grad_norm": 0.313750684261322, "learning_rate": 7.791755518242558e-05, "loss": 0.0247, "step": 27670 }, { "grad_norm": 0.2848942279815674, "learning_rate": 7.790040621750876e-05, "loss": 0.0209, "step": 27680 }, { "grad_norm": 0.25456467270851135, "learning_rate": 7.788325248517558e-05, "loss": 0.0199, "step": 27690 }, { "grad_norm": 0.2823868691921234, "learning_rate": 7.786609398835715e-05, "loss": 0.0219, "step": 27700 }, { "grad_norm": 0.34775570034980774, "learning_rate": 7.784893072998541e-05, "loss": 0.0198, "step": 27710 }, { "grad_norm": 0.20552372932434082, "learning_rate": 7.783176271299306e-05, "loss": 0.0204, "step": 27720 }, { "grad_norm": 0.24393051862716675, "learning_rate": 7.781458994031368e-05, "loss": 0.0177, "step": 27730 }, { "grad_norm": 0.3368784189224243, "learning_rate": 7.779741241488161e-05, "loss": 0.0225, "step": 27740 }, { "grad_norm": 0.2714061141014099, "learning_rate": 7.7780230139632e-05, "loss": 0.0185, "step": 27750 }, { "grad_norm": 0.22795945405960083, "learning_rate": 7.776304311750087e-05, "loss": 0.0154, "step": 27760 }, { "grad_norm": 0.2502574920654297, "learning_rate": 7.7745851351425e-05, "loss": 0.0239, "step": 27770 }, { "grad_norm": 0.22933940589427948, "learning_rate": 7.772865484434197e-05, "loss": 0.0208, "step": 27780 }, { "grad_norm": 0.3426024913787842, "learning_rate": 7.77114535991902e-05, "loss": 0.0212, "step": 27790 }, { "grad_norm": 0.23686358332633972, "learning_rate": 7.769424761890893e-05, "loss": 0.0173, "step": 27800 }, { "grad_norm": 0.3485773205757141, "learning_rate": 7.767703690643817e-05, "loss": 0.0199, "step": 27810 }, { "grad_norm": 0.12148944288492203, "learning_rate": 7.76598214647188e-05, "loss": 0.0151, "step": 27820 }, { "grad_norm": 0.23372554779052734, "learning_rate": 7.764260129669241e-05, "loss": 0.0174, "step": 27830 }, { "grad_norm": 0.20091512799263, "learning_rate": 7.76253764053015e-05, "loss": 0.0259, "step": 27840 }, { "grad_norm": 0.2616726756095886, "learning_rate": 7.760814679348932e-05, "loss": 0.0165, "step": 27850 }, { "grad_norm": 0.22875502705574036, "learning_rate": 7.759091246419992e-05, "loss": 0.0168, "step": 27860 }, { "grad_norm": 0.3570299446582794, "learning_rate": 7.75736734203782e-05, "loss": 0.0204, "step": 27870 }, { "grad_norm": 0.33566418290138245, "learning_rate": 7.755642966496985e-05, "loss": 0.0277, "step": 27880 }, { "grad_norm": 0.23333804309368134, "learning_rate": 7.753918120092132e-05, "loss": 0.0204, "step": 27890 }, { "grad_norm": 0.24059391021728516, "learning_rate": 7.752192803117993e-05, "loss": 0.0205, "step": 27900 }, { "grad_norm": 0.3355652987957001, "learning_rate": 7.750467015869377e-05, "loss": 0.0188, "step": 27910 }, { "grad_norm": 0.41131892800331116, "learning_rate": 7.748740758641174e-05, "loss": 0.0191, "step": 27920 }, { "grad_norm": 0.2360055297613144, "learning_rate": 7.74701403172835e-05, "loss": 0.0225, "step": 27930 }, { "grad_norm": 0.1913212388753891, "learning_rate": 7.745286835425962e-05, "loss": 0.0214, "step": 27940 }, { "grad_norm": 0.28977423906326294, "learning_rate": 7.743559170029138e-05, "loss": 0.0208, "step": 27950 }, { "grad_norm": 0.28325021266937256, "learning_rate": 7.741831035833087e-05, "loss": 0.03, "step": 27960 }, { "grad_norm": 0.19174376130104065, "learning_rate": 7.740102433133102e-05, "loss": 0.0163, "step": 27970 }, { "grad_norm": 0.19392544031143188, "learning_rate": 7.738373362224553e-05, "loss": 0.0202, "step": 27980 }, { "grad_norm": 0.23529425263404846, "learning_rate": 7.73664382340289e-05, "loss": 0.018, "step": 27990 }, { "grad_norm": 0.2960596978664398, "learning_rate": 7.734913816963647e-05, "loss": 0.0263, "step": 28000 }, { "grad_norm": 0.1882481575012207, "learning_rate": 7.73318334320243e-05, "loss": 0.0208, "step": 28010 }, { "grad_norm": 0.36898094415664673, "learning_rate": 7.731452402414934e-05, "loss": 0.0218, "step": 28020 }, { "grad_norm": 0.31141769886016846, "learning_rate": 7.729720994896928e-05, "loss": 0.0211, "step": 28030 }, { "grad_norm": 0.23579595983028412, "learning_rate": 7.727989120944262e-05, "loss": 0.0212, "step": 28040 }, { "grad_norm": 0.36124321818351746, "learning_rate": 7.726256780852865e-05, "loss": 0.0155, "step": 28050 }, { "grad_norm": 0.23316653072834015, "learning_rate": 7.724523974918749e-05, "loss": 0.0146, "step": 28060 }, { "grad_norm": 0.21874810755252838, "learning_rate": 7.722790703438002e-05, "loss": 0.0142, "step": 28070 }, { "grad_norm": 0.14622528851032257, "learning_rate": 7.72105696670679e-05, "loss": 0.0153, "step": 28080 }, { "grad_norm": 0.3231786787509918, "learning_rate": 7.719322765021364e-05, "loss": 0.0222, "step": 28090 }, { "grad_norm": 0.29934585094451904, "learning_rate": 7.717588098678051e-05, "loss": 0.0212, "step": 28100 }, { "grad_norm": 0.1631719470024109, "learning_rate": 7.715852967973258e-05, "loss": 0.0183, "step": 28110 }, { "grad_norm": 0.27846768498420715, "learning_rate": 7.714117373203474e-05, "loss": 0.022, "step": 28120 }, { "grad_norm": 0.2344742715358734, "learning_rate": 7.712381314665259e-05, "loss": 0.0177, "step": 28130 }, { "grad_norm": 0.3195285499095917, "learning_rate": 7.710644792655261e-05, "loss": 0.0168, "step": 28140 }, { "grad_norm": 0.4722442328929901, "learning_rate": 7.708907807470207e-05, "loss": 0.0233, "step": 28150 }, { "grad_norm": 0.17679420113563538, "learning_rate": 7.707170359406896e-05, "loss": 0.0182, "step": 28160 }, { "grad_norm": 0.23808377981185913, "learning_rate": 7.705432448762213e-05, "loss": 0.0184, "step": 28170 }, { "grad_norm": 0.2879447042942047, "learning_rate": 7.703694075833117e-05, "loss": 0.025, "step": 28180 }, { "grad_norm": 0.4072037935256958, "learning_rate": 7.70195524091665e-05, "loss": 0.034, "step": 28190 }, { "grad_norm": 0.3279232084751129, "learning_rate": 7.70021594430993e-05, "loss": 0.0264, "step": 28200 }, { "grad_norm": 0.23085063695907593, "learning_rate": 7.698476186310157e-05, "loss": 0.0172, "step": 28210 }, { "grad_norm": 0.22585822641849518, "learning_rate": 7.696735967214608e-05, "loss": 0.0175, "step": 28220 }, { "grad_norm": 0.30028194189071655, "learning_rate": 7.694995287320636e-05, "loss": 0.0175, "step": 28230 }, { "grad_norm": 0.20972062647342682, "learning_rate": 7.693254146925679e-05, "loss": 0.0172, "step": 28240 }, { "grad_norm": 0.38341301679611206, "learning_rate": 7.691512546327251e-05, "loss": 0.0265, "step": 28250 }, { "grad_norm": 0.21658962965011597, "learning_rate": 7.689770485822939e-05, "loss": 0.0195, "step": 28260 }, { "grad_norm": 0.2020283043384552, "learning_rate": 7.688027965710416e-05, "loss": 0.0144, "step": 28270 }, { "grad_norm": 0.22705435752868652, "learning_rate": 7.686284986287433e-05, "loss": 0.0198, "step": 28280 }, { "grad_norm": 0.3350352644920349, "learning_rate": 7.684541547851817e-05, "loss": 0.0155, "step": 28290 }, { "grad_norm": 0.2609689235687256, "learning_rate": 7.68279765070147e-05, "loss": 0.0175, "step": 28300 }, { "grad_norm": 0.2467597872018814, "learning_rate": 7.68105329513438e-05, "loss": 0.0188, "step": 28310 }, { "grad_norm": 0.1955839842557907, "learning_rate": 7.67930848144861e-05, "loss": 0.0172, "step": 28320 }, { "grad_norm": 0.19047921895980835, "learning_rate": 7.6775632099423e-05, "loss": 0.0221, "step": 28330 }, { "grad_norm": 0.16704332828521729, "learning_rate": 7.675817480913667e-05, "loss": 0.0157, "step": 28340 }, { "grad_norm": 0.2525384724140167, "learning_rate": 7.674071294661011e-05, "loss": 0.0194, "step": 28350 }, { "grad_norm": 0.1918746829032898, "learning_rate": 7.672324651482707e-05, "loss": 0.0147, "step": 28360 }, { "grad_norm": 0.33101168274879456, "learning_rate": 7.670577551677209e-05, "loss": 0.0217, "step": 28370 }, { "grad_norm": 0.29048240184783936, "learning_rate": 7.668829995543047e-05, "loss": 0.0191, "step": 28380 }, { "grad_norm": 0.24451005458831787, "learning_rate": 7.667081983378832e-05, "loss": 0.0233, "step": 28390 }, { "grad_norm": 0.21630100905895233, "learning_rate": 7.66533351548325e-05, "loss": 0.018, "step": 28400 }, { "grad_norm": 0.32656291127204895, "learning_rate": 7.663584592155069e-05, "loss": 0.018, "step": 28410 }, { "grad_norm": 0.38678446412086487, "learning_rate": 7.661835213693129e-05, "loss": 0.02, "step": 28420 }, { "grad_norm": 0.29738572239875793, "learning_rate": 7.660085380396353e-05, "loss": 0.0166, "step": 28430 }, { "grad_norm": 0.27562278509140015, "learning_rate": 7.658335092563738e-05, "loss": 0.0183, "step": 28440 }, { "grad_norm": 0.2918901741504669, "learning_rate": 7.656584350494362e-05, "loss": 0.0186, "step": 28450 }, { "grad_norm": 0.2530454397201538, "learning_rate": 7.654833154487378e-05, "loss": 0.0191, "step": 28460 }, { "grad_norm": 0.3753783702850342, "learning_rate": 7.653081504842017e-05, "loss": 0.0155, "step": 28470 }, { "grad_norm": 0.41173017024993896, "learning_rate": 7.65132940185759e-05, "loss": 0.0233, "step": 28480 }, { "grad_norm": 0.19701634347438812, "learning_rate": 7.649576845833481e-05, "loss": 0.0159, "step": 28490 }, { "grad_norm": 0.14547684788703918, "learning_rate": 7.647823837069156e-05, "loss": 0.0163, "step": 28500 }, { "grad_norm": 0.27235832810401917, "learning_rate": 7.646070375864156e-05, "loss": 0.0156, "step": 28510 }, { "grad_norm": 0.2710810601711273, "learning_rate": 7.644316462518099e-05, "loss": 0.0219, "step": 28520 }, { "grad_norm": 0.2652335464954376, "learning_rate": 7.642562097330679e-05, "loss": 0.0245, "step": 28530 }, { "grad_norm": 0.18311099708080292, "learning_rate": 7.640807280601671e-05, "loss": 0.0197, "step": 28540 }, { "grad_norm": 0.2660517394542694, "learning_rate": 7.639052012630927e-05, "loss": 0.0212, "step": 28550 }, { "grad_norm": 0.35375627875328064, "learning_rate": 7.63729629371837e-05, "loss": 0.026, "step": 28560 }, { "grad_norm": 0.2651803493499756, "learning_rate": 7.635540124164009e-05, "loss": 0.0279, "step": 28570 }, { "grad_norm": 0.3270946145057678, "learning_rate": 7.633783504267922e-05, "loss": 0.0218, "step": 28580 }, { "grad_norm": 0.22031807899475098, "learning_rate": 7.632026434330269e-05, "loss": 0.0193, "step": 28590 }, { "grad_norm": 0.20398852229118347, "learning_rate": 7.630268914651282e-05, "loss": 0.0161, "step": 28600 }, { "grad_norm": 0.30589646100997925, "learning_rate": 7.628510945531278e-05, "loss": 0.0265, "step": 28610 }, { "grad_norm": 0.23692834377288818, "learning_rate": 7.626752527270641e-05, "loss": 0.0204, "step": 28620 }, { "grad_norm": 0.20728044211864471, "learning_rate": 7.62499366016984e-05, "loss": 0.0156, "step": 28630 }, { "grad_norm": 0.13669922947883606, "learning_rate": 7.623234344529416e-05, "loss": 0.0175, "step": 28640 }, { "grad_norm": 0.3615644872188568, "learning_rate": 7.62147458064999e-05, "loss": 0.0194, "step": 28650 }, { "grad_norm": 0.34837326407432556, "learning_rate": 7.619714368832254e-05, "loss": 0.0248, "step": 28660 }, { "grad_norm": 0.21380621194839478, "learning_rate": 7.61795370937698e-05, "loss": 0.0169, "step": 28670 }, { "grad_norm": 0.24063439667224884, "learning_rate": 7.61619260258502e-05, "loss": 0.0228, "step": 28680 }, { "grad_norm": 0.3557019829750061, "learning_rate": 7.614431048757298e-05, "loss": 0.0236, "step": 28690 }, { "grad_norm": 0.24676744639873505, "learning_rate": 7.612669048194814e-05, "loss": 0.0225, "step": 28700 }, { "grad_norm": 0.2614726126194, "learning_rate": 7.610906601198646e-05, "loss": 0.018, "step": 28710 }, { "grad_norm": 0.24595856666564941, "learning_rate": 7.60914370806995e-05, "loss": 0.0221, "step": 28720 }, { "grad_norm": 0.2090669423341751, "learning_rate": 7.607380369109953e-05, "loss": 0.0312, "step": 28730 }, { "grad_norm": 0.19612327218055725, "learning_rate": 7.605616584619961e-05, "loss": 0.0169, "step": 28740 }, { "grad_norm": 0.3346056640148163, "learning_rate": 7.603852354901362e-05, "loss": 0.0159, "step": 28750 }, { "grad_norm": 0.18903854489326477, "learning_rate": 7.602087680255609e-05, "loss": 0.0299, "step": 28760 }, { "grad_norm": 0.35900241136550903, "learning_rate": 7.600322560984238e-05, "loss": 0.0237, "step": 28770 }, { "grad_norm": 0.28182119131088257, "learning_rate": 7.598556997388863e-05, "loss": 0.0207, "step": 28780 }, { "grad_norm": 0.18749414384365082, "learning_rate": 7.596790989771166e-05, "loss": 0.0201, "step": 28790 }, { "grad_norm": 0.3248055875301361, "learning_rate": 7.595024538432914e-05, "loss": 0.0184, "step": 28800 }, { "grad_norm": 0.14948183298110962, "learning_rate": 7.59325764367594e-05, "loss": 0.0149, "step": 28810 }, { "grad_norm": 0.2014910876750946, "learning_rate": 7.59149030580216e-05, "loss": 0.0203, "step": 28820 }, { "grad_norm": 0.2561303973197937, "learning_rate": 7.589722525113562e-05, "loss": 0.0192, "step": 28830 }, { "grad_norm": 0.35969629883766174, "learning_rate": 7.587954301912216e-05, "loss": 0.0161, "step": 28840 }, { "grad_norm": 0.3926233947277069, "learning_rate": 7.586185636500263e-05, "loss": 0.0182, "step": 28850 }, { "grad_norm": 0.23158839344978333, "learning_rate": 7.584416529179914e-05, "loss": 0.024, "step": 28860 }, { "grad_norm": 0.33735230565071106, "learning_rate": 7.582646980253465e-05, "loss": 0.0184, "step": 28870 }, { "grad_norm": 0.23383130133152008, "learning_rate": 7.580876990023282e-05, "loss": 0.026, "step": 28880 }, { "grad_norm": 0.20622916519641876, "learning_rate": 7.579106558791809e-05, "loss": 0.0144, "step": 28890 }, { "grad_norm": 0.30504581332206726, "learning_rate": 7.577335686861565e-05, "loss": 0.0207, "step": 28900 }, { "grad_norm": 0.3176715075969696, "learning_rate": 7.575564374535141e-05, "loss": 0.0214, "step": 28910 }, { "grad_norm": 0.17891567945480347, "learning_rate": 7.573792622115207e-05, "loss": 0.0147, "step": 28920 }, { "grad_norm": 0.23873436450958252, "learning_rate": 7.572020429904507e-05, "loss": 0.0162, "step": 28930 }, { "grad_norm": 0.2764529585838318, "learning_rate": 7.570247798205861e-05, "loss": 0.022, "step": 28940 }, { "grad_norm": 0.11673741042613983, "learning_rate": 7.568474727322164e-05, "loss": 0.024, "step": 28950 }, { "grad_norm": 0.2170984447002411, "learning_rate": 7.566701217556384e-05, "loss": 0.0174, "step": 28960 }, { "grad_norm": 0.31141427159309387, "learning_rate": 7.564927269211564e-05, "loss": 0.0192, "step": 28970 }, { "grad_norm": 0.24301964044570923, "learning_rate": 7.563152882590824e-05, "loss": 0.019, "step": 28980 }, { "grad_norm": 0.19139990210533142, "learning_rate": 7.56137805799736e-05, "loss": 0.0157, "step": 28990 }, { "grad_norm": 0.265377938747406, "learning_rate": 7.559602795734439e-05, "loss": 0.0169, "step": 29000 }, { "grad_norm": 0.1807323396205902, "learning_rate": 7.557827096105408e-05, "loss": 0.017, "step": 29010 }, { "grad_norm": 0.22698423266410828, "learning_rate": 7.55605095941368e-05, "loss": 0.0231, "step": 29020 }, { "grad_norm": 0.18533578515052795, "learning_rate": 7.554274385962752e-05, "loss": 0.0203, "step": 29030 }, { "grad_norm": 0.26207399368286133, "learning_rate": 7.552497376056191e-05, "loss": 0.0199, "step": 29040 }, { "grad_norm": 0.21165470778942108, "learning_rate": 7.550719929997639e-05, "loss": 0.0184, "step": 29050 }, { "grad_norm": 0.3736170828342438, "learning_rate": 7.548942048090813e-05, "loss": 0.0188, "step": 29060 }, { "grad_norm": 0.32336580753326416, "learning_rate": 7.547163730639506e-05, "loss": 0.0206, "step": 29070 }, { "grad_norm": 0.16255797445774078, "learning_rate": 7.545384977947583e-05, "loss": 0.016, "step": 29080 }, { "grad_norm": 0.15886250138282776, "learning_rate": 7.543605790318981e-05, "loss": 0.0164, "step": 29090 }, { "grad_norm": 0.34672999382019043, "learning_rate": 7.54182616805772e-05, "loss": 0.0221, "step": 29100 }, { "grad_norm": 0.18575887382030487, "learning_rate": 7.540046111467885e-05, "loss": 0.018, "step": 29110 }, { "grad_norm": 0.20490334928035736, "learning_rate": 7.53826562085364e-05, "loss": 0.0156, "step": 29120 }, { "grad_norm": 0.29863467812538147, "learning_rate": 7.536484696519221e-05, "loss": 0.0182, "step": 29130 }, { "grad_norm": 0.22679303586483002, "learning_rate": 7.534703338768942e-05, "loss": 0.0168, "step": 29140 }, { "grad_norm": 0.3986552655696869, "learning_rate": 7.532921547907185e-05, "loss": 0.0211, "step": 29150 }, { "grad_norm": 0.2569337487220764, "learning_rate": 7.531139324238412e-05, "loss": 0.0216, "step": 29160 }, { "grad_norm": 0.349020779132843, "learning_rate": 7.529356668067157e-05, "loss": 0.0263, "step": 29170 }, { "grad_norm": 0.3210863471031189, "learning_rate": 7.527573579698023e-05, "loss": 0.0201, "step": 29180 }, { "grad_norm": 0.2616508901119232, "learning_rate": 7.525790059435693e-05, "loss": 0.0199, "step": 29190 }, { "grad_norm": 0.2504475712776184, "learning_rate": 7.524006107584926e-05, "loss": 0.0162, "step": 29200 }, { "grad_norm": 0.3219624161720276, "learning_rate": 7.522221724450544e-05, "loss": 0.0198, "step": 29210 }, { "grad_norm": 0.235950767993927, "learning_rate": 7.520436910337451e-05, "loss": 0.0242, "step": 29220 }, { "grad_norm": 0.23938928544521332, "learning_rate": 7.518651665550627e-05, "loss": 0.0174, "step": 29230 }, { "grad_norm": 0.17766723036766052, "learning_rate": 7.516865990395117e-05, "loss": 0.0266, "step": 29240 }, { "grad_norm": 0.24876680970191956, "learning_rate": 7.515079885176047e-05, "loss": 0.0147, "step": 29250 }, { "grad_norm": 0.14763405919075012, "learning_rate": 7.513293350198612e-05, "loss": 0.0201, "step": 29260 }, { "grad_norm": 0.34143444895744324, "learning_rate": 7.511506385768081e-05, "loss": 0.0205, "step": 29270 }, { "grad_norm": 0.3000483214855194, "learning_rate": 7.509718992189801e-05, "loss": 0.0183, "step": 29280 }, { "grad_norm": 0.2574169337749481, "learning_rate": 7.507931169769182e-05, "loss": 0.0217, "step": 29290 }, { "grad_norm": 0.27438095211982727, "learning_rate": 7.506142918811722e-05, "loss": 0.0193, "step": 29300 }, { "grad_norm": 0.17253287136554718, "learning_rate": 7.504354239622978e-05, "loss": 0.0211, "step": 29310 }, { "grad_norm": 0.2217591106891632, "learning_rate": 7.50256513250859e-05, "loss": 0.0191, "step": 29320 }, { "grad_norm": 0.2613363564014435, "learning_rate": 7.500775597774265e-05, "loss": 0.0139, "step": 29330 }, { "grad_norm": 0.22051416337490082, "learning_rate": 7.498985635725788e-05, "loss": 0.0149, "step": 29340 }, { "grad_norm": 0.3381454348564148, "learning_rate": 7.497195246669012e-05, "loss": 0.0234, "step": 29350 }, { "grad_norm": 0.33686667680740356, "learning_rate": 7.495404430909868e-05, "loss": 0.0198, "step": 29360 }, { "grad_norm": 0.3014897406101227, "learning_rate": 7.493613188754356e-05, "loss": 0.021, "step": 29370 }, { "grad_norm": 0.223870187997818, "learning_rate": 7.49182152050855e-05, "loss": 0.0179, "step": 29380 }, { "grad_norm": 0.24985529482364655, "learning_rate": 7.490029426478598e-05, "loss": 0.0153, "step": 29390 }, { "grad_norm": 0.18644031882286072, "learning_rate": 7.488236906970719e-05, "loss": 0.0218, "step": 29400 }, { "grad_norm": 0.28821006417274475, "learning_rate": 7.486443962291207e-05, "loss": 0.0202, "step": 29410 }, { "grad_norm": 0.24930429458618164, "learning_rate": 7.484650592746424e-05, "loss": 0.0169, "step": 29420 }, { "grad_norm": 0.2333497852087021, "learning_rate": 7.482856798642811e-05, "loss": 0.0158, "step": 29430 }, { "grad_norm": 0.15394464135169983, "learning_rate": 7.481062580286878e-05, "loss": 0.0221, "step": 29440 }, { "grad_norm": 0.19698357582092285, "learning_rate": 7.479267937985208e-05, "loss": 0.0157, "step": 29450 }, { "grad_norm": 0.30384722352027893, "learning_rate": 7.477472872044456e-05, "loss": 0.013, "step": 29460 }, { "grad_norm": 0.2291005253791809, "learning_rate": 7.475677382771347e-05, "loss": 0.0181, "step": 29470 }, { "grad_norm": 0.18301540613174438, "learning_rate": 7.473881470472683e-05, "loss": 0.0172, "step": 29480 }, { "grad_norm": 0.3031880855560303, "learning_rate": 7.472085135455337e-05, "loss": 0.0199, "step": 29490 }, { "grad_norm": 0.2799389660358429, "learning_rate": 7.470288378026256e-05, "loss": 0.0195, "step": 29500 }, { "grad_norm": 0.32545018196105957, "learning_rate": 7.468491198492451e-05, "loss": 0.0185, "step": 29510 }, { "grad_norm": 0.49778205156326294, "learning_rate": 7.466693597161013e-05, "loss": 0.017, "step": 29520 }, { "grad_norm": 0.14379428327083588, "learning_rate": 7.464895574339104e-05, "loss": 0.0234, "step": 29530 }, { "grad_norm": 0.20985932648181915, "learning_rate": 7.463097130333958e-05, "loss": 0.0251, "step": 29540 }, { "grad_norm": 0.17914758622646332, "learning_rate": 7.461298265452876e-05, "loss": 0.0204, "step": 29550 }, { "grad_norm": 0.21826505661010742, "learning_rate": 7.459498980003239e-05, "loss": 0.0188, "step": 29560 }, { "grad_norm": 0.27087539434432983, "learning_rate": 7.457699274292493e-05, "loss": 0.0194, "step": 29570 }, { "grad_norm": 0.1737523376941681, "learning_rate": 7.455899148628159e-05, "loss": 0.0209, "step": 29580 }, { "grad_norm": 0.24623358249664307, "learning_rate": 7.45409860331783e-05, "loss": 0.0175, "step": 29590 }, { "grad_norm": 0.19910885393619537, "learning_rate": 7.452297638669169e-05, "loss": 0.0149, "step": 29600 }, { "grad_norm": 0.20775634050369263, "learning_rate": 7.450496254989911e-05, "loss": 0.0129, "step": 29610 }, { "grad_norm": 0.2705439627170563, "learning_rate": 7.448694452587866e-05, "loss": 0.0175, "step": 29620 }, { "grad_norm": 0.27694007754325867, "learning_rate": 7.44689223177091e-05, "loss": 0.0201, "step": 29630 }, { "grad_norm": 0.26718804240226746, "learning_rate": 7.445089592846994e-05, "loss": 0.0183, "step": 29640 }, { "grad_norm": 0.2436395138502121, "learning_rate": 7.443286536124141e-05, "loss": 0.0225, "step": 29650 }, { "grad_norm": 0.2274554967880249, "learning_rate": 7.441483061910443e-05, "loss": 0.0266, "step": 29660 }, { "grad_norm": 0.2318359613418579, "learning_rate": 7.439679170514064e-05, "loss": 0.0186, "step": 29670 }, { "grad_norm": 0.3239384889602661, "learning_rate": 7.43787486224324e-05, "loss": 0.02, "step": 29680 }, { "grad_norm": 0.26227232813835144, "learning_rate": 7.436070137406276e-05, "loss": 0.0171, "step": 29690 }, { "grad_norm": 0.23601464927196503, "learning_rate": 7.434264996311556e-05, "loss": 0.0162, "step": 29700 }, { "grad_norm": 0.47106361389160156, "learning_rate": 7.432459439267525e-05, "loss": 0.0224, "step": 29710 }, { "grad_norm": 0.20122060179710388, "learning_rate": 7.430653466582701e-05, "loss": 0.0162, "step": 29720 }, { "grad_norm": 0.21246330440044403, "learning_rate": 7.42884707856568e-05, "loss": 0.0289, "step": 29730 }, { "grad_norm": 0.24950140714645386, "learning_rate": 7.427040275525122e-05, "loss": 0.0178, "step": 29740 }, { "grad_norm": 0.3132196366786957, "learning_rate": 7.42523305776976e-05, "loss": 0.0182, "step": 29750 }, { "grad_norm": 0.35403573513031006, "learning_rate": 7.4234254256084e-05, "loss": 0.0165, "step": 29760 }, { "grad_norm": 0.17286527156829834, "learning_rate": 7.421617379349915e-05, "loss": 0.0198, "step": 29770 }, { "grad_norm": 0.3020942211151123, "learning_rate": 7.41980891930325e-05, "loss": 0.0181, "step": 29780 }, { "grad_norm": 0.16778910160064697, "learning_rate": 7.418000045777425e-05, "loss": 0.0161, "step": 29790 }, { "grad_norm": 0.25194329023361206, "learning_rate": 7.416190759081523e-05, "loss": 0.0181, "step": 29800 }, { "grad_norm": 0.3616977035999298, "learning_rate": 7.414381059524704e-05, "loss": 0.0173, "step": 29810 }, { "grad_norm": 0.2536938488483429, "learning_rate": 7.412570947416195e-05, "loss": 0.0135, "step": 29820 }, { "grad_norm": 0.36727175116539, "learning_rate": 7.410760423065295e-05, "loss": 0.0289, "step": 29830 }, { "grad_norm": 0.3250160217285156, "learning_rate": 7.408949486781372e-05, "loss": 0.0221, "step": 29840 }, { "grad_norm": 0.26972776651382446, "learning_rate": 7.407138138873868e-05, "loss": 0.0215, "step": 29850 }, { "grad_norm": 0.3060078024864197, "learning_rate": 7.405326379652292e-05, "loss": 0.0191, "step": 29860 }, { "grad_norm": 0.2122138887643814, "learning_rate": 7.403514209426222e-05, "loss": 0.0171, "step": 29870 }, { "grad_norm": 0.3220767378807068, "learning_rate": 7.40170162850531e-05, "loss": 0.022, "step": 29880 }, { "grad_norm": 0.25131940841674805, "learning_rate": 7.399888637199278e-05, "loss": 0.0155, "step": 29890 }, { "grad_norm": 0.23816384375095367, "learning_rate": 7.398075235817914e-05, "loss": 0.0291, "step": 29900 }, { "grad_norm": 0.2522256672382355, "learning_rate": 7.39626142467108e-05, "loss": 0.019, "step": 29910 }, { "grad_norm": 0.54916912317276, "learning_rate": 7.394447204068706e-05, "loss": 0.0151, "step": 29920 }, { "grad_norm": 0.36369410157203674, "learning_rate": 7.392632574320793e-05, "loss": 0.0287, "step": 29930 }, { "grad_norm": 0.1970641016960144, "learning_rate": 7.390817535737411e-05, "loss": 0.0218, "step": 29940 }, { "grad_norm": 0.3084077537059784, "learning_rate": 7.389002088628703e-05, "loss": 0.0195, "step": 29950 }, { "grad_norm": 0.31938356161117554, "learning_rate": 7.387186233304877e-05, "loss": 0.0228, "step": 29960 }, { "grad_norm": 0.3712559640407562, "learning_rate": 7.385369970076212e-05, "loss": 0.0238, "step": 29970 }, { "grad_norm": 0.3098253309726715, "learning_rate": 7.38355329925306e-05, "loss": 0.0229, "step": 29980 }, { "grad_norm": 0.28117334842681885, "learning_rate": 7.381736221145838e-05, "loss": 0.0213, "step": 29990 }, { "grad_norm": 0.24433766305446625, "learning_rate": 7.37991873606504e-05, "loss": 0.0167, "step": 30000 }, { "grad_norm": 0.31572461128234863, "learning_rate": 7.378100844321218e-05, "loss": 0.0183, "step": 30010 }, { "grad_norm": 0.24103377759456635, "learning_rate": 7.376282546225004e-05, "loss": 0.014, "step": 30020 }, { "grad_norm": 0.2778218984603882, "learning_rate": 7.374463842087094e-05, "loss": 0.0141, "step": 30030 }, { "grad_norm": 0.321073055267334, "learning_rate": 7.372644732218254e-05, "loss": 0.0183, "step": 30040 }, { "grad_norm": 0.2812645435333252, "learning_rate": 7.370825216929322e-05, "loss": 0.0254, "step": 30050 }, { "grad_norm": 0.1769513338804245, "learning_rate": 7.369005296531205e-05, "loss": 0.0183, "step": 30060 }, { "grad_norm": 0.22306890785694122, "learning_rate": 7.367184971334873e-05, "loss": 0.0211, "step": 30070 }, { "grad_norm": 0.20482686161994934, "learning_rate": 7.365364241651371e-05, "loss": 0.018, "step": 30080 }, { "grad_norm": 0.2649124264717102, "learning_rate": 7.363543107791815e-05, "loss": 0.0163, "step": 30090 }, { "grad_norm": 0.2625294029712677, "learning_rate": 7.361721570067384e-05, "loss": 0.0168, "step": 30100 }, { "grad_norm": 0.33800098299980164, "learning_rate": 7.359899628789331e-05, "loss": 0.0208, "step": 30110 }, { "grad_norm": 0.34074270725250244, "learning_rate": 7.358077284268974e-05, "loss": 0.0233, "step": 30120 }, { "grad_norm": 0.4620625674724579, "learning_rate": 7.356254536817702e-05, "loss": 0.0182, "step": 30130 }, { "grad_norm": 0.2699247896671295, "learning_rate": 7.354431386746973e-05, "loss": 0.0198, "step": 30140 }, { "grad_norm": 0.20643192529678345, "learning_rate": 7.352607834368316e-05, "loss": 0.019, "step": 30150 }, { "grad_norm": 0.2581424415111542, "learning_rate": 7.350783879993324e-05, "loss": 0.0166, "step": 30160 }, { "grad_norm": 0.44409266114234924, "learning_rate": 7.348959523933658e-05, "loss": 0.0223, "step": 30170 }, { "grad_norm": 0.20869077742099762, "learning_rate": 7.347134766501057e-05, "loss": 0.0143, "step": 30180 }, { "grad_norm": 0.28594210743904114, "learning_rate": 7.345309608007315e-05, "loss": 0.0176, "step": 30190 }, { "grad_norm": 0.2081761658191681, "learning_rate": 7.343484048764308e-05, "loss": 0.0178, "step": 30200 }, { "grad_norm": 0.1998537927865982, "learning_rate": 7.341658089083972e-05, "loss": 0.0153, "step": 30210 }, { "grad_norm": 0.17315256595611572, "learning_rate": 7.339831729278313e-05, "loss": 0.0143, "step": 30220 }, { "grad_norm": 0.21133142709732056, "learning_rate": 7.338004969659404e-05, "loss": 0.0148, "step": 30230 }, { "grad_norm": 0.19828496873378754, "learning_rate": 7.336177810539391e-05, "loss": 0.0193, "step": 30240 }, { "grad_norm": 0.2374514490365982, "learning_rate": 7.334350252230485e-05, "loss": 0.0133, "step": 30250 }, { "grad_norm": 0.18512989580631256, "learning_rate": 7.332522295044965e-05, "loss": 0.0141, "step": 30260 }, { "grad_norm": 0.26530975103378296, "learning_rate": 7.33069393929518e-05, "loss": 0.0218, "step": 30270 }, { "grad_norm": 0.29311874508857727, "learning_rate": 7.328865185293545e-05, "loss": 0.0317, "step": 30280 }, { "grad_norm": 0.21172171831130981, "learning_rate": 7.327036033352546e-05, "loss": 0.0181, "step": 30290 }, { "grad_norm": 0.2526291608810425, "learning_rate": 7.325206483784733e-05, "loss": 0.0155, "step": 30300 }, { "grad_norm": 0.4665715992450714, "learning_rate": 7.323376536902724e-05, "loss": 0.0242, "step": 30310 }, { "grad_norm": 0.30619338154792786, "learning_rate": 7.321546193019213e-05, "loss": 0.0243, "step": 30320 }, { "grad_norm": 0.2842245399951935, "learning_rate": 7.31971545244695e-05, "loss": 0.0225, "step": 30330 }, { "grad_norm": 0.31815189123153687, "learning_rate": 7.31788431549876e-05, "loss": 0.019, "step": 30340 }, { "grad_norm": 0.1669236272573471, "learning_rate": 7.316052782487534e-05, "loss": 0.016, "step": 30350 }, { "grad_norm": 0.2406177669763565, "learning_rate": 7.314220853726234e-05, "loss": 0.0227, "step": 30360 }, { "grad_norm": 0.2638116180896759, "learning_rate": 7.312388529527884e-05, "loss": 0.0187, "step": 30370 }, { "grad_norm": 0.286092609167099, "learning_rate": 7.310555810205577e-05, "loss": 0.018, "step": 30380 }, { "grad_norm": 0.20434631407260895, "learning_rate": 7.308722696072476e-05, "loss": 0.0206, "step": 30390 }, { "grad_norm": 0.5139673352241516, "learning_rate": 7.306889187441811e-05, "loss": 0.0189, "step": 30400 }, { "grad_norm": 0.3531641960144043, "learning_rate": 7.305055284626876e-05, "loss": 0.0176, "step": 30410 }, { "grad_norm": 0.23672333359718323, "learning_rate": 7.303220987941037e-05, "loss": 0.0198, "step": 30420 }, { "grad_norm": 0.2352401465177536, "learning_rate": 7.301386297697726e-05, "loss": 0.0244, "step": 30430 }, { "grad_norm": 0.3071494996547699, "learning_rate": 7.299551214210438e-05, "loss": 0.021, "step": 30440 }, { "grad_norm": 0.17227515578269958, "learning_rate": 7.297715737792738e-05, "loss": 0.0167, "step": 30450 }, { "grad_norm": 0.30639833211898804, "learning_rate": 7.295879868758265e-05, "loss": 0.0214, "step": 30460 }, { "grad_norm": 0.22379691898822784, "learning_rate": 7.294043607420713e-05, "loss": 0.0135, "step": 30470 }, { "grad_norm": 0.20531412959098816, "learning_rate": 7.292206954093852e-05, "loss": 0.0229, "step": 30480 }, { "grad_norm": 0.3232993185520172, "learning_rate": 7.290369909091515e-05, "loss": 0.016, "step": 30490 }, { "grad_norm": 0.27384185791015625, "learning_rate": 7.2885324727276e-05, "loss": 0.0185, "step": 30500 }, { "grad_norm": 0.19505755603313446, "learning_rate": 7.286694645316076e-05, "loss": 0.0187, "step": 30510 }, { "grad_norm": 0.2591778635978699, "learning_rate": 7.284856427170982e-05, "loss": 0.0194, "step": 30520 }, { "grad_norm": 0.3350532650947571, "learning_rate": 7.283017818606414e-05, "loss": 0.018, "step": 30530 }, { "grad_norm": 0.3563095033168793, "learning_rate": 7.28117881993654e-05, "loss": 0.0188, "step": 30540 }, { "grad_norm": 0.16725032031536102, "learning_rate": 7.279339431475598e-05, "loss": 0.0144, "step": 30550 }, { "grad_norm": 0.2845793068408966, "learning_rate": 7.277499653537887e-05, "loss": 0.0133, "step": 30560 }, { "grad_norm": 0.28459835052490234, "learning_rate": 7.275659486437776e-05, "loss": 0.0234, "step": 30570 }, { "grad_norm": 0.2493591159582138, "learning_rate": 7.273818930489695e-05, "loss": 0.0232, "step": 30580 }, { "grad_norm": 0.3039996922016144, "learning_rate": 7.271977986008151e-05, "loss": 0.0169, "step": 30590 }, { "grad_norm": 0.21216711401939392, "learning_rate": 7.270136653307705e-05, "loss": 0.0187, "step": 30600 }, { "grad_norm": 0.38385283946990967, "learning_rate": 7.268294932702994e-05, "loss": 0.0217, "step": 30610 }, { "grad_norm": 0.37141770124435425, "learning_rate": 7.266452824508719e-05, "loss": 0.0285, "step": 30620 }, { "grad_norm": 0.2195863127708435, "learning_rate": 7.264610329039643e-05, "loss": 0.0194, "step": 30630 }, { "grad_norm": 0.27819135785102844, "learning_rate": 7.262767446610599e-05, "loss": 0.0217, "step": 30640 }, { "grad_norm": 0.3558798134326935, "learning_rate": 7.260924177536485e-05, "loss": 0.0286, "step": 30650 }, { "grad_norm": 0.216474711894989, "learning_rate": 7.259080522132265e-05, "loss": 0.0257, "step": 30660 }, { "grad_norm": 0.21759247779846191, "learning_rate": 7.257236480712972e-05, "loss": 0.0184, "step": 30670 }, { "grad_norm": 0.24781188368797302, "learning_rate": 7.255392053593697e-05, "loss": 0.0193, "step": 30680 }, { "grad_norm": 0.3624063730239868, "learning_rate": 7.253547241089607e-05, "loss": 0.0149, "step": 30690 }, { "grad_norm": 0.22332347929477692, "learning_rate": 7.251702043515927e-05, "loss": 0.0208, "step": 30700 }, { "grad_norm": 0.31294959783554077, "learning_rate": 7.249856461187952e-05, "loss": 0.0182, "step": 30710 }, { "grad_norm": 0.32456836104393005, "learning_rate": 7.248010494421042e-05, "loss": 0.016, "step": 30720 }, { "grad_norm": 0.21854400634765625, "learning_rate": 7.246164143530622e-05, "loss": 0.0273, "step": 30730 }, { "grad_norm": 0.2730295956134796, "learning_rate": 7.244317408832181e-05, "loss": 0.0174, "step": 30740 }, { "grad_norm": 0.18114148080348969, "learning_rate": 7.242470290641279e-05, "loss": 0.0185, "step": 30750 }, { "grad_norm": 0.34308719635009766, "learning_rate": 7.240622789273536e-05, "loss": 0.0211, "step": 30760 }, { "grad_norm": 0.19657395780086517, "learning_rate": 7.238774905044638e-05, "loss": 0.0143, "step": 30770 }, { "grad_norm": 0.3282098174095154, "learning_rate": 7.236926638270341e-05, "loss": 0.02, "step": 30780 }, { "grad_norm": 0.8460372090339661, "learning_rate": 7.23507798926646e-05, "loss": 0.0272, "step": 30790 }, { "grad_norm": 0.23858210444450378, "learning_rate": 7.23322895834888e-05, "loss": 0.019, "step": 30800 }, { "grad_norm": 0.26672911643981934, "learning_rate": 7.231379545833552e-05, "loss": 0.0199, "step": 30810 }, { "grad_norm": 0.2910730540752411, "learning_rate": 7.229529752036487e-05, "loss": 0.0202, "step": 30820 }, { "grad_norm": 0.2514723241329193, "learning_rate": 7.227679577273765e-05, "loss": 0.0196, "step": 30830 }, { "grad_norm": 0.2705495059490204, "learning_rate": 7.225829021861529e-05, "loss": 0.0169, "step": 30840 }, { "grad_norm": 0.20653505623340607, "learning_rate": 7.223978086115992e-05, "loss": 0.015, "step": 30850 }, { "grad_norm": 0.2105703353881836, "learning_rate": 7.222126770353425e-05, "loss": 0.0202, "step": 30860 }, { "grad_norm": 0.24972344934940338, "learning_rate": 7.22027507489017e-05, "loss": 0.0177, "step": 30870 }, { "grad_norm": 0.2352248579263687, "learning_rate": 7.218423000042627e-05, "loss": 0.0177, "step": 30880 }, { "grad_norm": 0.23992744088172913, "learning_rate": 7.216570546127268e-05, "loss": 0.0178, "step": 30890 }, { "grad_norm": 0.29359176754951477, "learning_rate": 7.214717713460626e-05, "loss": 0.0273, "step": 30900 }, { "grad_norm": 0.21462035179138184, "learning_rate": 7.2128645023593e-05, "loss": 0.0226, "step": 30910 }, { "grad_norm": 0.17671848833560944, "learning_rate": 7.211010913139951e-05, "loss": 0.0173, "step": 30920 }, { "grad_norm": 0.32496708631515503, "learning_rate": 7.209156946119308e-05, "loss": 0.0183, "step": 30930 }, { "grad_norm": 0.4345152974128723, "learning_rate": 7.207302601614166e-05, "loss": 0.0184, "step": 30940 }, { "grad_norm": 0.3546402156352997, "learning_rate": 7.205447879941378e-05, "loss": 0.02, "step": 30950 }, { "grad_norm": 0.39642149209976196, "learning_rate": 7.203592781417866e-05, "loss": 0.0203, "step": 30960 }, { "grad_norm": 0.30386781692504883, "learning_rate": 7.201737306360617e-05, "loss": 0.0173, "step": 30970 }, { "grad_norm": 0.22486765682697296, "learning_rate": 7.19988145508668e-05, "loss": 0.0209, "step": 30980 }, { "grad_norm": 0.2765046954154968, "learning_rate": 7.198025227913168e-05, "loss": 0.0208, "step": 30990 }, { "grad_norm": 0.377691388130188, "learning_rate": 7.196168625157261e-05, "loss": 0.0215, "step": 31000 }, { "grad_norm": 0.31433579325675964, "learning_rate": 7.194311647136201e-05, "loss": 0.0189, "step": 31010 }, { "grad_norm": 0.384944885969162, "learning_rate": 7.192454294167297e-05, "loss": 0.0149, "step": 31020 }, { "grad_norm": 0.21003460884094238, "learning_rate": 7.190596566567917e-05, "loss": 0.0176, "step": 31030 }, { "grad_norm": 0.3185444176197052, "learning_rate": 7.188738464655496e-05, "loss": 0.0186, "step": 31040 }, { "grad_norm": 0.23683902621269226, "learning_rate": 7.186879988747533e-05, "loss": 0.0139, "step": 31050 }, { "grad_norm": 0.15943822264671326, "learning_rate": 7.185021139161592e-05, "loss": 0.0154, "step": 31060 }, { "grad_norm": 0.22478145360946655, "learning_rate": 7.1831619162153e-05, "loss": 0.0146, "step": 31070 }, { "grad_norm": 0.33216696977615356, "learning_rate": 7.181302320226345e-05, "loss": 0.0162, "step": 31080 }, { "grad_norm": 0.2630360722541809, "learning_rate": 7.179442351512482e-05, "loss": 0.0223, "step": 31090 }, { "grad_norm": 0.21080853044986725, "learning_rate": 7.177582010391528e-05, "loss": 0.0176, "step": 31100 }, { "grad_norm": 0.44293510913848877, "learning_rate": 7.175721297181366e-05, "loss": 0.0177, "step": 31110 }, { "grad_norm": 0.2891785502433777, "learning_rate": 7.173860212199942e-05, "loss": 0.0178, "step": 31120 }, { "grad_norm": 0.22599560022354126, "learning_rate": 7.171998755765263e-05, "loss": 0.0151, "step": 31130 }, { "grad_norm": 0.1972692310810089, "learning_rate": 7.170136928195398e-05, "loss": 0.0155, "step": 31140 }, { "grad_norm": 0.2037699818611145, "learning_rate": 7.168274729808489e-05, "loss": 0.0166, "step": 31150 }, { "grad_norm": 0.21426032483577728, "learning_rate": 7.166412160922728e-05, "loss": 0.0201, "step": 31160 }, { "grad_norm": 0.1692609190940857, "learning_rate": 7.164549221856382e-05, "loss": 0.0141, "step": 31170 }, { "grad_norm": 0.364926815032959, "learning_rate": 7.162685912927775e-05, "loss": 0.0255, "step": 31180 }, { "grad_norm": 0.1896856278181076, "learning_rate": 7.160822234455294e-05, "loss": 0.0163, "step": 31190 }, { "grad_norm": 0.22316046059131622, "learning_rate": 7.158958186757391e-05, "loss": 0.0177, "step": 31200 }, { "grad_norm": 0.2273014485836029, "learning_rate": 7.157093770152582e-05, "loss": 0.0195, "step": 31210 }, { "grad_norm": 0.20479652285575867, "learning_rate": 7.155228984959446e-05, "loss": 0.0177, "step": 31220 }, { "grad_norm": 0.28119808435440063, "learning_rate": 7.153363831496621e-05, "loss": 0.0226, "step": 31230 }, { "grad_norm": 0.288003534078598, "learning_rate": 7.151498310082811e-05, "loss": 0.021, "step": 31240 }, { "grad_norm": 0.21062123775482178, "learning_rate": 7.149632421036784e-05, "loss": 0.0144, "step": 31250 }, { "grad_norm": 0.3363015353679657, "learning_rate": 7.147766164677369e-05, "loss": 0.0326, "step": 31260 }, { "grad_norm": 0.24900926649570465, "learning_rate": 7.145899541323459e-05, "loss": 0.0148, "step": 31270 }, { "grad_norm": 0.2584385573863983, "learning_rate": 7.144032551294007e-05, "loss": 0.0254, "step": 31280 }, { "grad_norm": 0.29824844002723694, "learning_rate": 7.14216519490803e-05, "loss": 0.0224, "step": 31290 }, { "grad_norm": 0.3333509564399719, "learning_rate": 7.140297472484609e-05, "loss": 0.0251, "step": 31300 }, { "grad_norm": 0.17311808466911316, "learning_rate": 7.138429384342891e-05, "loss": 0.0305, "step": 31310 }, { "grad_norm": 0.26383304595947266, "learning_rate": 7.136560930802074e-05, "loss": 0.0176, "step": 31320 }, { "grad_norm": 0.41910362243652344, "learning_rate": 7.134692112181431e-05, "loss": 0.017, "step": 31330 }, { "grad_norm": 0.2387366145849228, "learning_rate": 7.13282292880029e-05, "loss": 0.0182, "step": 31340 }, { "grad_norm": 0.21133583784103394, "learning_rate": 7.130953380978043e-05, "loss": 0.0159, "step": 31350 }, { "grad_norm": 0.16985878348350525, "learning_rate": 7.129083469034144e-05, "loss": 0.0235, "step": 31360 }, { "grad_norm": 0.3695753812789917, "learning_rate": 7.127213193288112e-05, "loss": 0.0239, "step": 31370 }, { "grad_norm": 0.24409912526607513, "learning_rate": 7.125342554059522e-05, "loss": 0.0186, "step": 31380 }, { "grad_norm": 0.2849862277507782, "learning_rate": 7.12347155166802e-05, "loss": 0.02, "step": 31390 }, { "grad_norm": 0.33507245779037476, "learning_rate": 7.121600186433306e-05, "loss": 0.0182, "step": 31400 }, { "grad_norm": 0.31751006841659546, "learning_rate": 7.119728458675148e-05, "loss": 0.0166, "step": 31410 }, { "grad_norm": 0.18113826215267181, "learning_rate": 7.117856368713369e-05, "loss": 0.0146, "step": 31420 }, { "grad_norm": 0.31336304545402527, "learning_rate": 7.115983916867861e-05, "loss": 0.0196, "step": 31430 }, { "grad_norm": 0.17323319613933563, "learning_rate": 7.114111103458574e-05, "loss": 0.0131, "step": 31440 }, { "grad_norm": 0.14255847036838531, "learning_rate": 7.11223792880552e-05, "loss": 0.0141, "step": 31450 }, { "grad_norm": 0.21102970838546753, "learning_rate": 7.110364393228773e-05, "loss": 0.0199, "step": 31460 }, { "grad_norm": 0.3708130121231079, "learning_rate": 7.108490497048471e-05, "loss": 0.0154, "step": 31470 }, { "grad_norm": 0.23537948727607727, "learning_rate": 7.10661624058481e-05, "loss": 0.0178, "step": 31480 }, { "grad_norm": 0.225441113114357, "learning_rate": 7.10474162415805e-05, "loss": 0.0183, "step": 31490 }, { "grad_norm": 0.15539252758026123, "learning_rate": 7.102866648088511e-05, "loss": 0.0199, "step": 31500 }, { "grad_norm": 0.25001469254493713, "learning_rate": 7.100991312696576e-05, "loss": 0.0195, "step": 31510 }, { "grad_norm": 0.25796815752983093, "learning_rate": 7.099115618302686e-05, "loss": 0.0169, "step": 31520 }, { "grad_norm": 0.26998767256736755, "learning_rate": 7.097239565227349e-05, "loss": 0.0233, "step": 31530 }, { "grad_norm": 0.2695040702819824, "learning_rate": 7.09536315379113e-05, "loss": 0.015, "step": 31540 }, { "grad_norm": 0.22098268568515778, "learning_rate": 7.093486384314656e-05, "loss": 0.0151, "step": 31550 }, { "grad_norm": 0.18717791140079498, "learning_rate": 7.091609257118616e-05, "loss": 0.0189, "step": 31560 }, { "grad_norm": 0.24947437644004822, "learning_rate": 7.08973177252376e-05, "loss": 0.0195, "step": 31570 }, { "grad_norm": 0.48958733677864075, "learning_rate": 7.087853930850898e-05, "loss": 0.0156, "step": 31580 }, { "grad_norm": 0.24500097334384918, "learning_rate": 7.085975732420903e-05, "loss": 0.0135, "step": 31590 }, { "grad_norm": 1.0753036737442017, "learning_rate": 7.084097177554706e-05, "loss": 0.0173, "step": 31600 }, { "grad_norm": 0.33279359340667725, "learning_rate": 7.082218266573301e-05, "loss": 0.0223, "step": 31610 }, { "grad_norm": 0.37640270590782166, "learning_rate": 7.080338999797743e-05, "loss": 0.0193, "step": 31620 }, { "grad_norm": 0.26788151264190674, "learning_rate": 7.07845937754915e-05, "loss": 0.0162, "step": 31630 }, { "grad_norm": 0.21762318909168243, "learning_rate": 7.076579400148693e-05, "loss": 0.0219, "step": 31640 }, { "grad_norm": 0.15415215492248535, "learning_rate": 7.074699067917611e-05, "loss": 0.0215, "step": 31650 }, { "grad_norm": 0.32427364587783813, "learning_rate": 7.072818381177201e-05, "loss": 0.0171, "step": 31660 }, { "grad_norm": 0.2114148586988449, "learning_rate": 7.070937340248823e-05, "loss": 0.0184, "step": 31670 }, { "grad_norm": 0.1194198727607727, "learning_rate": 7.069055945453893e-05, "loss": 0.0171, "step": 31680 }, { "grad_norm": 0.21330900490283966, "learning_rate": 7.067174197113892e-05, "loss": 0.0136, "step": 31690 }, { "grad_norm": 0.22379836440086365, "learning_rate": 7.065292095550355e-05, "loss": 0.0199, "step": 31700 }, { "grad_norm": 0.2621617615222931, "learning_rate": 7.063409641084887e-05, "loss": 0.0164, "step": 31710 }, { "grad_norm": 0.3301861584186554, "learning_rate": 7.061526834039145e-05, "loss": 0.0209, "step": 31720 }, { "grad_norm": 0.24256175756454468, "learning_rate": 7.05964367473485e-05, "loss": 0.0164, "step": 31730 }, { "grad_norm": 0.24743296205997467, "learning_rate": 7.057760163493783e-05, "loss": 0.0175, "step": 31740 }, { "grad_norm": 0.30079707503318787, "learning_rate": 7.055876300637783e-05, "loss": 0.0165, "step": 31750 }, { "grad_norm": 0.21288777887821198, "learning_rate": 7.053992086488753e-05, "loss": 0.0237, "step": 31760 }, { "grad_norm": 0.22249987721443176, "learning_rate": 7.052107521368651e-05, "loss": 0.0184, "step": 31770 }, { "grad_norm": 0.24811409413814545, "learning_rate": 7.0502226055995e-05, "loss": 0.0246, "step": 31780 }, { "grad_norm": 0.19565653800964355, "learning_rate": 7.048337339503379e-05, "loss": 0.0281, "step": 31790 }, { "grad_norm": 0.279206246137619, "learning_rate": 7.046451723402427e-05, "loss": 0.0187, "step": 31800 }, { "grad_norm": 0.4869987368583679, "learning_rate": 7.044565757618848e-05, "loss": 0.0316, "step": 31810 }, { "grad_norm": 0.47276097536087036, "learning_rate": 7.042679442474899e-05, "loss": 0.0199, "step": 31820 }, { "grad_norm": 0.199188232421875, "learning_rate": 7.040792778292902e-05, "loss": 0.0282, "step": 31830 }, { "grad_norm": 0.30780503153800964, "learning_rate": 7.038905765395234e-05, "loss": 0.0255, "step": 31840 }, { "grad_norm": 0.26371634006500244, "learning_rate": 7.037018404104334e-05, "loss": 0.0201, "step": 31850 }, { "grad_norm": 0.231680765748024, "learning_rate": 7.035130694742702e-05, "loss": 0.0163, "step": 31860 }, { "grad_norm": 0.24180202186107635, "learning_rate": 7.033242637632897e-05, "loss": 0.0216, "step": 31870 }, { "grad_norm": 0.3053404688835144, "learning_rate": 7.031354233097534e-05, "loss": 0.0191, "step": 31880 }, { "grad_norm": 0.2885149419307709, "learning_rate": 7.029465481459289e-05, "loss": 0.0316, "step": 31890 }, { "grad_norm": 0.20773009955883026, "learning_rate": 7.027576383040898e-05, "loss": 0.0162, "step": 31900 }, { "grad_norm": 0.419087678194046, "learning_rate": 7.025686938165159e-05, "loss": 0.0278, "step": 31910 }, { "grad_norm": 0.2819332182407379, "learning_rate": 7.023797147154924e-05, "loss": 0.0302, "step": 31920 }, { "grad_norm": 0.2684289515018463, "learning_rate": 7.021907010333111e-05, "loss": 0.016, "step": 31930 }, { "grad_norm": 0.20355653762817383, "learning_rate": 7.020016528022685e-05, "loss": 0.018, "step": 31940 }, { "grad_norm": 0.2853761911392212, "learning_rate": 7.018125700546683e-05, "loss": 0.0224, "step": 31950 }, { "grad_norm": 0.22042229771614075, "learning_rate": 7.016234528228196e-05, "loss": 0.0217, "step": 31960 }, { "grad_norm": 0.36489400267601013, "learning_rate": 7.014343011390372e-05, "loss": 0.0233, "step": 31970 }, { "grad_norm": 0.21877817809581757, "learning_rate": 7.01245115035642e-05, "loss": 0.0131, "step": 31980 }, { "grad_norm": 0.21904365718364716, "learning_rate": 7.010558945449606e-05, "loss": 0.0235, "step": 31990 }, { "grad_norm": 0.22692641615867615, "learning_rate": 7.008666396993258e-05, "loss": 0.0163, "step": 32000 }, { "grad_norm": 0.3795824944972992, "learning_rate": 7.006773505310759e-05, "loss": 0.0246, "step": 32010 }, { "grad_norm": 0.35023608803749084, "learning_rate": 7.004880270725553e-05, "loss": 0.0213, "step": 32020 }, { "grad_norm": 0.2475120574235916, "learning_rate": 7.002986693561144e-05, "loss": 0.0173, "step": 32030 }, { "grad_norm": 0.23130694031715393, "learning_rate": 7.001092774141089e-05, "loss": 0.0238, "step": 32040 }, { "grad_norm": 0.3940589129924774, "learning_rate": 6.999198512789009e-05, "loss": 0.0212, "step": 32050 }, { "grad_norm": 0.38828009366989136, "learning_rate": 6.997303909828584e-05, "loss": 0.0211, "step": 32060 }, { "grad_norm": 0.28564515709877014, "learning_rate": 6.995408965583544e-05, "loss": 0.0197, "step": 32070 }, { "grad_norm": 0.2965945899486542, "learning_rate": 6.993513680377688e-05, "loss": 0.0137, "step": 32080 }, { "grad_norm": 0.28085264563560486, "learning_rate": 6.991618054534868e-05, "loss": 0.0241, "step": 32090 }, { "grad_norm": 0.3413423001766205, "learning_rate": 6.989722088378991e-05, "loss": 0.0231, "step": 32100 }, { "grad_norm": 0.3005920946598053, "learning_rate": 6.987825782234027e-05, "loss": 0.0192, "step": 32110 }, { "grad_norm": 0.19351057708263397, "learning_rate": 6.985929136424006e-05, "loss": 0.0165, "step": 32120 }, { "grad_norm": 0.5061119794845581, "learning_rate": 6.984032151273012e-05, "loss": 0.0242, "step": 32130 }, { "grad_norm": 0.22077836096286774, "learning_rate": 6.982134827105186e-05, "loss": 0.0169, "step": 32140 }, { "grad_norm": 0.2827928960323334, "learning_rate": 6.980237164244729e-05, "loss": 0.0166, "step": 32150 }, { "grad_norm": 0.1900634616613388, "learning_rate": 6.9783391630159e-05, "loss": 0.0135, "step": 32160 }, { "grad_norm": 0.22494660317897797, "learning_rate": 6.976440823743015e-05, "loss": 0.014, "step": 32170 }, { "grad_norm": 0.2722202241420746, "learning_rate": 6.974542146750451e-05, "loss": 0.0117, "step": 32180 }, { "grad_norm": 0.19327767193317413, "learning_rate": 6.972643132362637e-05, "loss": 0.0122, "step": 32190 }, { "grad_norm": 0.21945199370384216, "learning_rate": 6.970743780904064e-05, "loss": 0.0182, "step": 32200 }, { "grad_norm": 0.30854594707489014, "learning_rate": 6.968844092699277e-05, "loss": 0.0164, "step": 32210 }, { "grad_norm": 0.355297327041626, "learning_rate": 6.966944068072883e-05, "loss": 0.0174, "step": 32220 }, { "grad_norm": 0.25741028785705566, "learning_rate": 6.965043707349545e-05, "loss": 0.0132, "step": 32230 }, { "grad_norm": 0.2886500954627991, "learning_rate": 6.963143010853982e-05, "loss": 0.0211, "step": 32240 }, { "grad_norm": 0.3882254660129547, "learning_rate": 6.961241978910971e-05, "loss": 0.0233, "step": 32250 }, { "grad_norm": 0.31061384081840515, "learning_rate": 6.959340611845344e-05, "loss": 0.0146, "step": 32260 }, { "grad_norm": 0.26649075746536255, "learning_rate": 6.957438909981995e-05, "loss": 0.0178, "step": 32270 }, { "grad_norm": 0.5151455998420715, "learning_rate": 6.955536873645872e-05, "loss": 0.0136, "step": 32280 }, { "grad_norm": 0.2809252142906189, "learning_rate": 6.953634503161982e-05, "loss": 0.0161, "step": 32290 }, { "grad_norm": 0.25542163848876953, "learning_rate": 6.951731798855387e-05, "loss": 0.0199, "step": 32300 }, { "grad_norm": 0.2154480218887329, "learning_rate": 6.949828761051208e-05, "loss": 0.0205, "step": 32310 }, { "grad_norm": 0.22759488224983215, "learning_rate": 6.947925390074622e-05, "loss": 0.0164, "step": 32320 }, { "grad_norm": 0.20619724690914154, "learning_rate": 6.946021686250863e-05, "loss": 0.014, "step": 32330 }, { "grad_norm": 0.21012485027313232, "learning_rate": 6.94411764990522e-05, "loss": 0.0169, "step": 32340 }, { "grad_norm": 0.19191640615463257, "learning_rate": 6.942213281363044e-05, "loss": 0.0151, "step": 32350 }, { "grad_norm": 0.17735959589481354, "learning_rate": 6.940308580949737e-05, "loss": 0.0171, "step": 32360 }, { "grad_norm": 0.4679580330848694, "learning_rate": 6.93840354899076e-05, "loss": 0.0163, "step": 32370 }, { "grad_norm": 0.20326845347881317, "learning_rate": 6.936498185811633e-05, "loss": 0.0166, "step": 32380 }, { "grad_norm": 0.2447061985731125, "learning_rate": 6.93459249173793e-05, "loss": 0.0185, "step": 32390 }, { "grad_norm": 0.3013680577278137, "learning_rate": 6.932686467095279e-05, "loss": 0.0173, "step": 32400 }, { "grad_norm": 0.3221130967140198, "learning_rate": 6.930780112209373e-05, "loss": 0.0206, "step": 32410 }, { "grad_norm": 0.3192709982395172, "learning_rate": 6.92887342740595e-05, "loss": 0.0214, "step": 32420 }, { "grad_norm": 0.2865724563598633, "learning_rate": 6.926966413010816e-05, "loss": 0.0202, "step": 32430 }, { "grad_norm": 0.21334028244018555, "learning_rate": 6.925059069349824e-05, "loss": 0.0152, "step": 32440 }, { "grad_norm": 0.22972576320171356, "learning_rate": 6.923151396748886e-05, "loss": 0.0182, "step": 32450 }, { "grad_norm": 0.21712228655815125, "learning_rate": 6.921243395533974e-05, "loss": 0.0203, "step": 32460 }, { "grad_norm": 0.2699779272079468, "learning_rate": 6.919335066031109e-05, "loss": 0.0199, "step": 32470 }, { "grad_norm": 0.24428389966487885, "learning_rate": 6.917426408566379e-05, "loss": 0.0181, "step": 32480 }, { "grad_norm": 0.3171919584274292, "learning_rate": 6.915517423465916e-05, "loss": 0.0199, "step": 32490 }, { "grad_norm": 0.27771613001823425, "learning_rate": 6.913608111055914e-05, "loss": 0.0193, "step": 32500 }, { "grad_norm": 0.18371766805648804, "learning_rate": 6.911698471662623e-05, "loss": 0.0138, "step": 32510 }, { "grad_norm": 0.13860183954238892, "learning_rate": 6.90978850561235e-05, "loss": 0.0138, "step": 32520 }, { "grad_norm": 0.25178247690200806, "learning_rate": 6.907878213231454e-05, "loss": 0.0139, "step": 32530 }, { "grad_norm": 0.2552264332771301, "learning_rate": 6.90596759484635e-05, "loss": 0.0133, "step": 32540 }, { "grad_norm": 0.22694450616836548, "learning_rate": 6.904056650783514e-05, "loss": 0.0185, "step": 32550 }, { "grad_norm": 0.2562071979045868, "learning_rate": 6.902145381369471e-05, "loss": 0.0197, "step": 32560 }, { "grad_norm": 0.2814638018608093, "learning_rate": 6.900233786930808e-05, "loss": 0.0191, "step": 32570 }, { "grad_norm": 0.4831608831882477, "learning_rate": 6.898321867794161e-05, "loss": 0.0205, "step": 32580 }, { "grad_norm": 0.18324311077594757, "learning_rate": 6.896409624286226e-05, "loss": 0.022, "step": 32590 }, { "grad_norm": 0.2684771716594696, "learning_rate": 6.894497056733754e-05, "loss": 0.0154, "step": 32600 }, { "grad_norm": 0.24694277346134186, "learning_rate": 6.89258416546355e-05, "loss": 0.0217, "step": 32610 }, { "grad_norm": 0.20710551738739014, "learning_rate": 6.890670950802474e-05, "loss": 0.0144, "step": 32620 }, { "grad_norm": 0.23172058165073395, "learning_rate": 6.88875741307744e-05, "loss": 0.0151, "step": 32630 }, { "grad_norm": 0.16623470187187195, "learning_rate": 6.886843552615425e-05, "loss": 0.0205, "step": 32640 }, { "grad_norm": 0.15670645236968994, "learning_rate": 6.884929369743451e-05, "loss": 0.0114, "step": 32650 }, { "grad_norm": 0.2873477637767792, "learning_rate": 6.8830148647886e-05, "loss": 0.0155, "step": 32660 }, { "grad_norm": 0.23779399693012238, "learning_rate": 6.88110003807801e-05, "loss": 0.0244, "step": 32670 }, { "grad_norm": 0.2746293842792511, "learning_rate": 6.87918488993887e-05, "loss": 0.0142, "step": 32680 }, { "grad_norm": 0.40167880058288574, "learning_rate": 6.877269420698431e-05, "loss": 0.017, "step": 32690 }, { "grad_norm": 0.2770039141178131, "learning_rate": 6.875353630683989e-05, "loss": 0.0201, "step": 32700 }, { "grad_norm": 0.40259042382240295, "learning_rate": 6.873437520222905e-05, "loss": 0.0248, "step": 32710 }, { "grad_norm": 0.24130462110042572, "learning_rate": 6.871521089642585e-05, "loss": 0.0182, "step": 32720 }, { "grad_norm": 0.2759177088737488, "learning_rate": 6.869604339270498e-05, "loss": 0.0165, "step": 32730 }, { "grad_norm": 0.32365500926971436, "learning_rate": 6.867687269434164e-05, "loss": 0.0255, "step": 32740 }, { "grad_norm": 0.20607751607894897, "learning_rate": 6.865769880461156e-05, "loss": 0.0172, "step": 32750 }, { "grad_norm": 0.23864968121051788, "learning_rate": 6.863852172679104e-05, "loss": 0.0173, "step": 32760 }, { "grad_norm": 0.2747195065021515, "learning_rate": 6.861934146415693e-05, "loss": 0.0204, "step": 32770 }, { "grad_norm": 0.30510783195495605, "learning_rate": 6.86001580199866e-05, "loss": 0.023, "step": 32780 }, { "grad_norm": 0.16970863938331604, "learning_rate": 6.858097139755798e-05, "loss": 0.0141, "step": 32790 }, { "grad_norm": 0.3172622323036194, "learning_rate": 6.856178160014955e-05, "loss": 0.0189, "step": 32800 }, { "grad_norm": 0.27722224593162537, "learning_rate": 6.85425886310403e-05, "loss": 0.0156, "step": 32810 }, { "grad_norm": 0.2644611597061157, "learning_rate": 6.852339249350979e-05, "loss": 0.0155, "step": 32820 }, { "grad_norm": 0.1963493973016739, "learning_rate": 6.850419319083812e-05, "loss": 0.0254, "step": 32830 }, { "grad_norm": 0.30975770950317383, "learning_rate": 6.848499072630592e-05, "loss": 0.0193, "step": 32840 }, { "grad_norm": 0.3366585671901703, "learning_rate": 6.846578510319439e-05, "loss": 0.0218, "step": 32850 }, { "grad_norm": 0.2701798379421234, "learning_rate": 6.844657632478519e-05, "loss": 0.0218, "step": 32860 }, { "grad_norm": 0.24472245573997498, "learning_rate": 6.842736439436063e-05, "loss": 0.0142, "step": 32870 }, { "grad_norm": 0.24459607899188995, "learning_rate": 6.84081493152035e-05, "loss": 0.0166, "step": 32880 }, { "grad_norm": 0.19520457088947296, "learning_rate": 6.83889310905971e-05, "loss": 0.0117, "step": 32890 }, { "grad_norm": 0.1992533653974533, "learning_rate": 6.836970972382533e-05, "loss": 0.0237, "step": 32900 }, { "grad_norm": 0.22349479794502258, "learning_rate": 6.835048521817257e-05, "loss": 0.0189, "step": 32910 }, { "grad_norm": 0.25218287110328674, "learning_rate": 6.833125757692379e-05, "loss": 0.0195, "step": 32920 }, { "grad_norm": 0.22000931203365326, "learning_rate": 6.831202680336441e-05, "loss": 0.0147, "step": 32930 }, { "grad_norm": 0.2007831633090973, "learning_rate": 6.829279290078052e-05, "loss": 0.0172, "step": 32940 }, { "grad_norm": 0.24928973615169525, "learning_rate": 6.827355587245863e-05, "loss": 0.0186, "step": 32950 }, { "grad_norm": 0.19858305156230927, "learning_rate": 6.82543157216858e-05, "loss": 0.0159, "step": 32960 }, { "grad_norm": 0.18736451864242554, "learning_rate": 6.823507245174969e-05, "loss": 0.02, "step": 32970 }, { "grad_norm": 0.25431036949157715, "learning_rate": 6.821582606593841e-05, "loss": 0.0168, "step": 32980 }, { "grad_norm": 0.20685584843158722, "learning_rate": 6.81965765675407e-05, "loss": 0.014, "step": 32990 }, { "grad_norm": 0.23674467206001282, "learning_rate": 6.81773239598457e-05, "loss": 0.0182, "step": 33000 }, { "grad_norm": 0.25097930431365967, "learning_rate": 6.815806824614319e-05, "loss": 0.018, "step": 33010 }, { "grad_norm": 0.17501477897167206, "learning_rate": 6.813880942972343e-05, "loss": 0.0129, "step": 33020 }, { "grad_norm": 0.1754416972398758, "learning_rate": 6.811954751387726e-05, "loss": 0.0152, "step": 33030 }, { "grad_norm": 0.17655189335346222, "learning_rate": 6.810028250189598e-05, "loss": 0.0162, "step": 33040 }, { "grad_norm": 0.3389626145362854, "learning_rate": 6.808101439707147e-05, "loss": 0.0301, "step": 33050 }, { "grad_norm": 0.25106388330459595, "learning_rate": 6.806174320269609e-05, "loss": 0.0219, "step": 33060 }, { "grad_norm": 0.33338671922683716, "learning_rate": 6.804246892206281e-05, "loss": 0.022, "step": 33070 }, { "grad_norm": 0.19931989908218384, "learning_rate": 6.802319155846506e-05, "loss": 0.0322, "step": 33080 }, { "grad_norm": 0.3155302107334137, "learning_rate": 6.800391111519679e-05, "loss": 0.0185, "step": 33090 }, { "grad_norm": 0.4383586049079895, "learning_rate": 6.798462759555253e-05, "loss": 0.0221, "step": 33100 }, { "grad_norm": 0.1704431176185608, "learning_rate": 6.79653410028273e-05, "loss": 0.0172, "step": 33110 }, { "grad_norm": 0.2265135645866394, "learning_rate": 6.794605134031663e-05, "loss": 0.0193, "step": 33120 }, { "grad_norm": 0.28629234433174133, "learning_rate": 6.792675861131661e-05, "loss": 0.021, "step": 33130 }, { "grad_norm": 0.32695937156677246, "learning_rate": 6.790746281912386e-05, "loss": 0.0262, "step": 33140 }, { "grad_norm": 0.2064063847064972, "learning_rate": 6.788816396703546e-05, "loss": 0.0172, "step": 33150 }, { "grad_norm": 0.20655223727226257, "learning_rate": 6.78688620583491e-05, "loss": 0.016, "step": 33160 }, { "grad_norm": 0.2511707842350006, "learning_rate": 6.784955709636292e-05, "loss": 0.0167, "step": 33170 }, { "grad_norm": 0.3179965615272522, "learning_rate": 6.783024908437564e-05, "loss": 0.0167, "step": 33180 }, { "grad_norm": 0.2823767066001892, "learning_rate": 6.781093802568641e-05, "loss": 0.0131, "step": 33190 }, { "grad_norm": 0.20979632437229156, "learning_rate": 6.779162392359504e-05, "loss": 0.0162, "step": 33200 }, { "grad_norm": 0.34387141466140747, "learning_rate": 6.777230678140172e-05, "loss": 0.0216, "step": 33210 }, { "grad_norm": 0.27383309602737427, "learning_rate": 6.775298660240726e-05, "loss": 0.0211, "step": 33220 }, { "grad_norm": 0.203102707862854, "learning_rate": 6.773366338991292e-05, "loss": 0.0255, "step": 33230 }, { "grad_norm": 0.24271419644355774, "learning_rate": 6.771433714722052e-05, "loss": 0.017, "step": 33240 }, { "grad_norm": 0.19398175179958344, "learning_rate": 6.769500787763239e-05, "loss": 0.0214, "step": 33250 }, { "grad_norm": 0.18437603116035461, "learning_rate": 6.76756755844514e-05, "loss": 0.0218, "step": 33260 }, { "grad_norm": 0.25093191862106323, "learning_rate": 6.765634027098087e-05, "loss": 0.0193, "step": 33270 }, { "grad_norm": 0.22507672011852264, "learning_rate": 6.763700194052468e-05, "loss": 0.0171, "step": 33280 }, { "grad_norm": 0.20060886442661285, "learning_rate": 6.761766059638723e-05, "loss": 0.0193, "step": 33290 }, { "grad_norm": 0.30337846279144287, "learning_rate": 6.759831624187345e-05, "loss": 0.0187, "step": 33300 }, { "grad_norm": 0.39477425813674927, "learning_rate": 6.757896888028871e-05, "loss": 0.0211, "step": 33310 }, { "grad_norm": 0.28216084837913513, "learning_rate": 6.7559618514939e-05, "loss": 0.0209, "step": 33320 }, { "grad_norm": 0.23585198819637299, "learning_rate": 6.754026514913073e-05, "loss": 0.0168, "step": 33330 }, { "grad_norm": 0.21455217897891998, "learning_rate": 6.752090878617087e-05, "loss": 0.0195, "step": 33340 }, { "grad_norm": 0.3454999029636383, "learning_rate": 6.75015494293669e-05, "loss": 0.0189, "step": 33350 }, { "grad_norm": 0.793411135673523, "learning_rate": 6.74821870820268e-05, "loss": 0.0148, "step": 33360 }, { "grad_norm": 0.2565869987010956, "learning_rate": 6.746282174745907e-05, "loss": 0.0255, "step": 33370 }, { "grad_norm": 0.2631659507751465, "learning_rate": 6.744345342897271e-05, "loss": 0.0196, "step": 33380 }, { "grad_norm": 0.2619173228740692, "learning_rate": 6.742408212987724e-05, "loss": 0.0203, "step": 33390 }, { "grad_norm": 0.26631203293800354, "learning_rate": 6.740470785348269e-05, "loss": 0.0163, "step": 33400 }, { "grad_norm": 0.23913508653640747, "learning_rate": 6.738533060309958e-05, "loss": 0.0217, "step": 33410 }, { "grad_norm": 0.16200226545333862, "learning_rate": 6.736595038203894e-05, "loss": 0.0165, "step": 33420 }, { "grad_norm": 0.2015252411365509, "learning_rate": 6.734656719361236e-05, "loss": 0.0192, "step": 33430 }, { "grad_norm": 0.34403228759765625, "learning_rate": 6.732718104113189e-05, "loss": 0.0212, "step": 33440 }, { "grad_norm": 0.2793932557106018, "learning_rate": 6.730779192791006e-05, "loss": 0.0179, "step": 33450 }, { "grad_norm": 0.28329747915267944, "learning_rate": 6.728839985725997e-05, "loss": 0.0179, "step": 33460 }, { "grad_norm": 0.17822910845279694, "learning_rate": 6.726900483249517e-05, "loss": 0.0163, "step": 33470 }, { "grad_norm": 0.27550452947616577, "learning_rate": 6.724960685692976e-05, "loss": 0.0178, "step": 33480 }, { "grad_norm": 0.20001333951950073, "learning_rate": 6.723020593387833e-05, "loss": 0.0151, "step": 33490 }, { "grad_norm": 0.3364485800266266, "learning_rate": 6.721080206665593e-05, "loss": 0.0177, "step": 33500 }, { "grad_norm": 0.3410690128803253, "learning_rate": 6.719139525857819e-05, "loss": 0.0251, "step": 33510 }, { "grad_norm": 0.17650653421878815, "learning_rate": 6.717198551296117e-05, "loss": 0.013, "step": 33520 }, { "grad_norm": 0.24550195038318634, "learning_rate": 6.715257283312148e-05, "loss": 0.0156, "step": 33530 }, { "grad_norm": 0.18029868602752686, "learning_rate": 6.713315722237623e-05, "loss": 0.0132, "step": 33540 }, { "grad_norm": 0.2898443043231964, "learning_rate": 6.7113738684043e-05, "loss": 0.0251, "step": 33550 }, { "grad_norm": 0.2894138693809509, "learning_rate": 6.709431722143989e-05, "loss": 0.0159, "step": 33560 }, { "grad_norm": 0.22104091942310333, "learning_rate": 6.70748928378855e-05, "loss": 0.0215, "step": 33570 }, { "grad_norm": 0.16587811708450317, "learning_rate": 6.705546553669891e-05, "loss": 0.0144, "step": 33580 }, { "grad_norm": 0.18068526685237885, "learning_rate": 6.703603532119974e-05, "loss": 0.0116, "step": 33590 }, { "grad_norm": 0.18304041028022766, "learning_rate": 6.701660219470808e-05, "loss": 0.0129, "step": 33600 }, { "grad_norm": 0.3193587064743042, "learning_rate": 6.69971661605445e-05, "loss": 0.0166, "step": 33610 }, { "grad_norm": 0.1717967540025711, "learning_rate": 6.697772722203008e-05, "loss": 0.0159, "step": 33620 }, { "grad_norm": 0.2623186707496643, "learning_rate": 6.695828538248643e-05, "loss": 0.0181, "step": 33630 }, { "grad_norm": 0.2766178250312805, "learning_rate": 6.693884064523563e-05, "loss": 0.0225, "step": 33640 }, { "grad_norm": 0.21528150141239166, "learning_rate": 6.691939301360023e-05, "loss": 0.0177, "step": 33650 }, { "grad_norm": 0.1873156577348709, "learning_rate": 6.689994249090333e-05, "loss": 0.0175, "step": 33660 }, { "grad_norm": 0.5115038156509399, "learning_rate": 6.688048908046845e-05, "loss": 0.0207, "step": 33670 }, { "grad_norm": 0.16842764616012573, "learning_rate": 6.686103278561969e-05, "loss": 0.0135, "step": 33680 }, { "grad_norm": 0.284441202878952, "learning_rate": 6.684157360968156e-05, "loss": 0.0188, "step": 33690 }, { "grad_norm": 0.2859804332256317, "learning_rate": 6.682211155597911e-05, "loss": 0.015, "step": 33700 }, { "grad_norm": 0.25514036417007446, "learning_rate": 6.680264662783789e-05, "loss": 0.0169, "step": 33710 }, { "grad_norm": 0.3100246787071228, "learning_rate": 6.678317882858391e-05, "loss": 0.0238, "step": 33720 }, { "grad_norm": 0.2768777012825012, "learning_rate": 6.67637081615437e-05, "loss": 0.0148, "step": 33730 }, { "grad_norm": 0.2349393218755722, "learning_rate": 6.674423463004427e-05, "loss": 0.0117, "step": 33740 }, { "grad_norm": 0.33103570342063904, "learning_rate": 6.672475823741308e-05, "loss": 0.0159, "step": 33750 }, { "grad_norm": 0.20768041908740997, "learning_rate": 6.670527898697811e-05, "loss": 0.0181, "step": 33760 }, { "grad_norm": 0.19614072144031525, "learning_rate": 6.668579688206788e-05, "loss": 0.0176, "step": 33770 }, { "grad_norm": 0.2667481601238251, "learning_rate": 6.666631192601131e-05, "loss": 0.0145, "step": 33780 }, { "grad_norm": 0.23610569536685944, "learning_rate": 6.664682412213785e-05, "loss": 0.0186, "step": 33790 }, { "grad_norm": 0.273759663105011, "learning_rate": 6.662733347377745e-05, "loss": 0.0154, "step": 33800 }, { "grad_norm": 0.1811472475528717, "learning_rate": 6.660783998426051e-05, "loss": 0.0168, "step": 33810 }, { "grad_norm": 0.341657817363739, "learning_rate": 6.658834365691794e-05, "loss": 0.0195, "step": 33820 }, { "grad_norm": 0.17219941318035126, "learning_rate": 6.656884449508115e-05, "loss": 0.0143, "step": 33830 }, { "grad_norm": 0.23471564054489136, "learning_rate": 6.654934250208198e-05, "loss": 0.0152, "step": 33840 }, { "grad_norm": 0.17482300102710724, "learning_rate": 6.65298376812528e-05, "loss": 0.0156, "step": 33850 }, { "grad_norm": 0.3270610272884369, "learning_rate": 6.651033003592646e-05, "loss": 0.0164, "step": 33860 }, { "grad_norm": 0.3051730990409851, "learning_rate": 6.649081956943626e-05, "loss": 0.0153, "step": 33870 }, { "grad_norm": 0.33846545219421387, "learning_rate": 6.647130628511604e-05, "loss": 0.0155, "step": 33880 }, { "grad_norm": 0.2396858185529709, "learning_rate": 6.645179018630005e-05, "loss": 0.0152, "step": 33890 }, { "grad_norm": 0.1914719194173813, "learning_rate": 6.643227127632309e-05, "loss": 0.0157, "step": 33900 }, { "grad_norm": 0.2771962285041809, "learning_rate": 6.641274955852038e-05, "loss": 0.0183, "step": 33910 }, { "grad_norm": 0.3012855052947998, "learning_rate": 6.639322503622768e-05, "loss": 0.0362, "step": 33920 }, { "grad_norm": 0.29081660509109497, "learning_rate": 6.637369771278116e-05, "loss": 0.0164, "step": 33930 }, { "grad_norm": 0.2957113981246948, "learning_rate": 6.635416759151751e-05, "loss": 0.0177, "step": 33940 }, { "grad_norm": 0.2088092863559723, "learning_rate": 6.633463467577394e-05, "loss": 0.0291, "step": 33950 }, { "grad_norm": 0.4336487650871277, "learning_rate": 6.631509896888803e-05, "loss": 0.0166, "step": 33960 }, { "grad_norm": 0.19012956321239471, "learning_rate": 6.629556047419794e-05, "loss": 0.0157, "step": 33970 }, { "grad_norm": 0.4547116458415985, "learning_rate": 6.627601919504223e-05, "loss": 0.0191, "step": 33980 }, { "grad_norm": 0.22523699700832367, "learning_rate": 6.625647513476001e-05, "loss": 0.0143, "step": 33990 }, { "grad_norm": 0.4779549539089203, "learning_rate": 6.62369282966908e-05, "loss": 0.0257, "step": 34000 }, { "grad_norm": 0.1976788491010666, "learning_rate": 6.621737868417464e-05, "loss": 0.0173, "step": 34010 }, { "grad_norm": 0.1955166757106781, "learning_rate": 6.619782630055198e-05, "loss": 0.0169, "step": 34020 }, { "grad_norm": 0.2803741693496704, "learning_rate": 6.617827114916382e-05, "loss": 0.0166, "step": 34030 }, { "grad_norm": 0.15508933365345, "learning_rate": 6.615871323335161e-05, "loss": 0.0166, "step": 34040 }, { "grad_norm": 0.3793627619743347, "learning_rate": 6.613915255645725e-05, "loss": 0.0179, "step": 34050 }, { "grad_norm": 0.2449440211057663, "learning_rate": 6.611958912182312e-05, "loss": 0.0152, "step": 34060 }, { "grad_norm": 0.23813636600971222, "learning_rate": 6.610002293279207e-05, "loss": 0.0178, "step": 34070 }, { "grad_norm": 0.20391777157783508, "learning_rate": 6.608045399270746e-05, "loss": 0.0129, "step": 34080 }, { "grad_norm": 0.3871323764324188, "learning_rate": 6.606088230491304e-05, "loss": 0.0167, "step": 34090 }, { "grad_norm": 0.21997587382793427, "learning_rate": 6.604130787275312e-05, "loss": 0.0168, "step": 34100 }, { "grad_norm": 0.22108539938926697, "learning_rate": 6.602173069957242e-05, "loss": 0.0148, "step": 34110 }, { "grad_norm": 0.20930485427379608, "learning_rate": 6.600215078871612e-05, "loss": 0.0157, "step": 34120 }, { "grad_norm": 0.2896011769771576, "learning_rate": 6.598256814352992e-05, "loss": 0.0145, "step": 34130 }, { "grad_norm": 0.286432683467865, "learning_rate": 6.596298276735995e-05, "loss": 0.0136, "step": 34140 }, { "grad_norm": 0.413879930973053, "learning_rate": 6.594339466355282e-05, "loss": 0.0165, "step": 34150 }, { "grad_norm": 0.23911884427070618, "learning_rate": 6.592380383545558e-05, "loss": 0.019, "step": 34160 }, { "grad_norm": 0.25520649552345276, "learning_rate": 6.590421028641577e-05, "loss": 0.0229, "step": 34170 }, { "grad_norm": 0.25487765669822693, "learning_rate": 6.588461401978143e-05, "loss": 0.0179, "step": 34180 }, { "grad_norm": 0.3382783532142639, "learning_rate": 6.586501503890099e-05, "loss": 0.0198, "step": 34190 }, { "grad_norm": 0.23334427177906036, "learning_rate": 6.584541334712338e-05, "loss": 0.0171, "step": 34200 }, { "grad_norm": 0.42301100492477417, "learning_rate": 6.582580894779802e-05, "loss": 0.0196, "step": 34210 }, { "grad_norm": 0.2253515124320984, "learning_rate": 6.580620184427473e-05, "loss": 0.0164, "step": 34220 }, { "grad_norm": 0.2204551100730896, "learning_rate": 6.578659203990385e-05, "loss": 0.0193, "step": 34230 }, { "grad_norm": 0.1717977523803711, "learning_rate": 6.576697953803615e-05, "loss": 0.0145, "step": 34240 }, { "grad_norm": 0.23315982520580292, "learning_rate": 6.57473643420229e-05, "loss": 0.017, "step": 34250 }, { "grad_norm": 0.24390673637390137, "learning_rate": 6.572774645521574e-05, "loss": 0.019, "step": 34260 }, { "grad_norm": 0.4213845133781433, "learning_rate": 6.570812588096688e-05, "loss": 0.0221, "step": 34270 }, { "grad_norm": 0.2269153594970703, "learning_rate": 6.568850262262893e-05, "loss": 0.0179, "step": 34280 }, { "grad_norm": 0.28063637018203735, "learning_rate": 6.566887668355497e-05, "loss": 0.0163, "step": 34290 }, { "grad_norm": 0.2552875578403473, "learning_rate": 6.564924806709851e-05, "loss": 0.0165, "step": 34300 }, { "grad_norm": 0.22756516933441162, "learning_rate": 6.562961677661359e-05, "loss": 0.0193, "step": 34310 }, { "grad_norm": 0.3726707994937897, "learning_rate": 6.56099828154546e-05, "loss": 0.0216, "step": 34320 }, { "grad_norm": 0.2525690793991089, "learning_rate": 6.55903461869765e-05, "loss": 0.0138, "step": 34330 }, { "grad_norm": 0.35032668709754944, "learning_rate": 6.557070689453465e-05, "loss": 0.0162, "step": 34340 }, { "grad_norm": 0.12044884264469147, "learning_rate": 6.555106494148482e-05, "loss": 0.021, "step": 34350 }, { "grad_norm": 0.4264647960662842, "learning_rate": 6.553142033118333e-05, "loss": 0.0177, "step": 34360 }, { "grad_norm": 0.4235162138938904, "learning_rate": 6.551177306698688e-05, "loss": 0.0215, "step": 34370 }, { "grad_norm": 0.16828298568725586, "learning_rate": 6.549212315225267e-05, "loss": 0.0189, "step": 34380 }, { "grad_norm": 0.20291124284267426, "learning_rate": 6.547247059033833e-05, "loss": 0.0176, "step": 34390 }, { "grad_norm": 0.2910388112068176, "learning_rate": 6.545281538460193e-05, "loss": 0.0158, "step": 34400 }, { "grad_norm": 0.22826723754405975, "learning_rate": 6.543315753840202e-05, "loss": 0.0155, "step": 34410 }, { "grad_norm": 0.19436335563659668, "learning_rate": 6.541349705509758e-05, "loss": 0.0134, "step": 34420 }, { "grad_norm": 0.16802982985973358, "learning_rate": 6.539383393804805e-05, "loss": 0.0164, "step": 34430 }, { "grad_norm": 0.23762549459934235, "learning_rate": 6.537416819061333e-05, "loss": 0.0219, "step": 34440 }, { "grad_norm": 0.20382508635520935, "learning_rate": 6.535449981615375e-05, "loss": 0.0184, "step": 34450 }, { "grad_norm": 0.20743896067142487, "learning_rate": 6.53348288180301e-05, "loss": 0.0143, "step": 34460 }, { "grad_norm": 0.25979188084602356, "learning_rate": 6.531515519960361e-05, "loss": 0.0231, "step": 34470 }, { "grad_norm": 0.16141048073768616, "learning_rate": 6.529547896423597e-05, "loss": 0.0141, "step": 34480 }, { "grad_norm": 0.22479389607906342, "learning_rate": 6.52758001152893e-05, "loss": 0.0192, "step": 34490 }, { "grad_norm": 0.15140089392662048, "learning_rate": 6.525611865612618e-05, "loss": 0.0139, "step": 34500 }, { "grad_norm": 0.24932844936847687, "learning_rate": 6.523643459010966e-05, "loss": 0.0196, "step": 34510 }, { "grad_norm": 0.2872313857078552, "learning_rate": 6.521674792060317e-05, "loss": 0.0207, "step": 34520 }, { "grad_norm": 0.26273998618125916, "learning_rate": 6.519705865097063e-05, "loss": 0.0179, "step": 34530 }, { "grad_norm": 0.2622193396091461, "learning_rate": 6.517736678457641e-05, "loss": 0.0157, "step": 34540 }, { "grad_norm": 0.19481387734413147, "learning_rate": 6.515767232478534e-05, "loss": 0.0154, "step": 34550 }, { "grad_norm": 0.2552516460418701, "learning_rate": 6.51379752749626e-05, "loss": 0.0151, "step": 34560 }, { "grad_norm": 0.27605563402175903, "learning_rate": 6.511827563847393e-05, "loss": 0.021, "step": 34570 }, { "grad_norm": 0.29293951392173767, "learning_rate": 6.509857341868542e-05, "loss": 0.0165, "step": 34580 }, { "grad_norm": 0.264728844165802, "learning_rate": 6.507886861896367e-05, "loss": 0.0167, "step": 34590 }, { "grad_norm": 0.25813910365104675, "learning_rate": 6.505916124267567e-05, "loss": 0.0195, "step": 34600 }, { "grad_norm": 0.2896479368209839, "learning_rate": 6.503945129318891e-05, "loss": 0.0175, "step": 34610 }, { "grad_norm": 0.26197510957717896, "learning_rate": 6.501973877387122e-05, "loss": 0.0109, "step": 34620 }, { "grad_norm": 0.30585145950317383, "learning_rate": 6.500002368809098e-05, "loss": 0.0187, "step": 34630 }, { "grad_norm": 0.3290345072746277, "learning_rate": 6.498030603921694e-05, "loss": 0.0173, "step": 34640 }, { "grad_norm": 0.27100709080696106, "learning_rate": 6.496058583061832e-05, "loss": 0.0196, "step": 34650 }, { "grad_norm": 0.27859795093536377, "learning_rate": 6.494086306566475e-05, "loss": 0.0165, "step": 34660 }, { "grad_norm": 0.27446097135543823, "learning_rate": 6.492113774772632e-05, "loss": 0.0169, "step": 34670 }, { "grad_norm": 0.671587347984314, "learning_rate": 6.490140988017354e-05, "loss": 0.0185, "step": 34680 }, { "grad_norm": 0.37452155351638794, "learning_rate": 6.488167946637736e-05, "loss": 0.0186, "step": 34690 }, { "grad_norm": 0.28022196888923645, "learning_rate": 6.486194650970915e-05, "loss": 0.0157, "step": 34700 }, { "grad_norm": 0.4327683746814728, "learning_rate": 6.48422110135408e-05, "loss": 0.0184, "step": 34710 }, { "grad_norm": 0.18967023491859436, "learning_rate": 6.482247298124451e-05, "loss": 0.0199, "step": 34720 }, { "grad_norm": 0.13842090964317322, "learning_rate": 6.480273241619297e-05, "loss": 0.0157, "step": 34730 }, { "grad_norm": 0.15690307319164276, "learning_rate": 6.478298932175933e-05, "loss": 0.0162, "step": 34740 }, { "grad_norm": 0.2024831473827362, "learning_rate": 6.476324370131712e-05, "loss": 0.0172, "step": 34750 }, { "grad_norm": 0.32928523421287537, "learning_rate": 6.474349555824036e-05, "loss": 0.0195, "step": 34760 }, { "grad_norm": 0.2273278832435608, "learning_rate": 6.472374489590342e-05, "loss": 0.0129, "step": 34770 }, { "grad_norm": 0.14896845817565918, "learning_rate": 6.470399171768118e-05, "loss": 0.0264, "step": 34780 }, { "grad_norm": 0.2648784816265106, "learning_rate": 6.468423602694891e-05, "loss": 0.0132, "step": 34790 }, { "grad_norm": 0.6121752262115479, "learning_rate": 6.466447782708232e-05, "loss": 0.0122, "step": 34800 }, { "grad_norm": 0.16827361285686493, "learning_rate": 6.464471712145754e-05, "loss": 0.0124, "step": 34810 }, { "grad_norm": 0.17176416516304016, "learning_rate": 6.462495391345114e-05, "loss": 0.0166, "step": 34820 }, { "grad_norm": 0.23843656480312347, "learning_rate": 6.46051882064401e-05, "loss": 0.0137, "step": 34830 }, { "grad_norm": 0.2119520902633667, "learning_rate": 6.458542000380186e-05, "loss": 0.0172, "step": 34840 }, { "grad_norm": 0.24776510894298553, "learning_rate": 6.456564930891424e-05, "loss": 0.0158, "step": 34850 }, { "grad_norm": 0.26025906205177307, "learning_rate": 6.454587612515555e-05, "loss": 0.0189, "step": 34860 }, { "grad_norm": 0.17791317403316498, "learning_rate": 6.452610045590444e-05, "loss": 0.015, "step": 34870 }, { "grad_norm": 0.29739490151405334, "learning_rate": 6.450632230454005e-05, "loss": 0.0167, "step": 34880 }, { "grad_norm": 0.37785467505455017, "learning_rate": 6.448654167444195e-05, "loss": 0.0232, "step": 34890 }, { "grad_norm": 0.2601141035556793, "learning_rate": 6.446675856899005e-05, "loss": 0.0137, "step": 34900 }, { "grad_norm": 0.1328122317790985, "learning_rate": 6.444697299156481e-05, "loss": 0.0147, "step": 34910 }, { "grad_norm": 0.25890403985977173, "learning_rate": 6.442718494554701e-05, "loss": 0.0164, "step": 34920 }, { "grad_norm": 0.24864864349365234, "learning_rate": 6.440739443431787e-05, "loss": 0.0137, "step": 34930 }, { "grad_norm": 0.21753914654254913, "learning_rate": 6.438760146125906e-05, "loss": 0.0121, "step": 34940 }, { "grad_norm": 0.26551681756973267, "learning_rate": 6.436780602975267e-05, "loss": 0.0208, "step": 34950 }, { "grad_norm": 0.225264310836792, "learning_rate": 6.43480081431812e-05, "loss": 0.019, "step": 34960 }, { "grad_norm": 0.23547770082950592, "learning_rate": 6.432820780492756e-05, "loss": 0.0207, "step": 34970 }, { "grad_norm": 0.23894792795181274, "learning_rate": 6.430840501837506e-05, "loss": 0.0208, "step": 34980 }, { "grad_norm": 0.24389924108982086, "learning_rate": 6.428859978690748e-05, "loss": 0.022, "step": 34990 }, { "grad_norm": 0.25403228402137756, "learning_rate": 6.426879211390901e-05, "loss": 0.0166, "step": 35000 }, { "grad_norm": 0.31132400035858154, "learning_rate": 6.424898200276422e-05, "loss": 0.0202, "step": 35010 }, { "grad_norm": 0.34082096815109253, "learning_rate": 6.42291694568581e-05, "loss": 0.0154, "step": 35020 }, { "grad_norm": 0.34931671619415283, "learning_rate": 6.42093544795761e-05, "loss": 0.0257, "step": 35030 }, { "grad_norm": 0.19136978685855865, "learning_rate": 6.418953707430403e-05, "loss": 0.0191, "step": 35040 }, { "grad_norm": 0.3282100558280945, "learning_rate": 6.416971724442819e-05, "loss": 0.015, "step": 35050 }, { "grad_norm": 0.2199256718158722, "learning_rate": 6.414989499333519e-05, "loss": 0.0153, "step": 35060 }, { "grad_norm": 0.2808989882469177, "learning_rate": 6.413007032441214e-05, "loss": 0.0176, "step": 35070 }, { "grad_norm": 0.25855326652526855, "learning_rate": 6.411024324104653e-05, "loss": 0.0191, "step": 35080 }, { "grad_norm": 0.35465753078460693, "learning_rate": 6.409041374662628e-05, "loss": 0.0148, "step": 35090 }, { "grad_norm": 0.2121053785085678, "learning_rate": 6.407058184453967e-05, "loss": 0.0268, "step": 35100 }, { "grad_norm": 0.3713197112083435, "learning_rate": 6.405074753817548e-05, "loss": 0.0172, "step": 35110 }, { "grad_norm": 0.22725190222263336, "learning_rate": 6.40309108309228e-05, "loss": 0.016, "step": 35120 }, { "grad_norm": 0.2648414671421051, "learning_rate": 6.401107172617122e-05, "loss": 0.0169, "step": 35130 }, { "grad_norm": 0.264274537563324, "learning_rate": 6.399123022731068e-05, "loss": 0.0289, "step": 35140 }, { "grad_norm": 0.25725066661834717, "learning_rate": 6.397138633773157e-05, "loss": 0.0169, "step": 35150 }, { "grad_norm": 0.37970829010009766, "learning_rate": 6.395154006082463e-05, "loss": 0.0154, "step": 35160 }, { "grad_norm": 0.27186617255210876, "learning_rate": 6.393169139998109e-05, "loss": 0.0221, "step": 35170 }, { "grad_norm": 0.2282378226518631, "learning_rate": 6.39118403585925e-05, "loss": 0.0156, "step": 35180 }, { "grad_norm": 0.25934353470802307, "learning_rate": 6.38919869400509e-05, "loss": 0.0182, "step": 35190 }, { "grad_norm": 0.1981322169303894, "learning_rate": 6.387213114774865e-05, "loss": 0.0189, "step": 35200 }, { "grad_norm": 0.17717325687408447, "learning_rate": 6.385227298507863e-05, "loss": 0.0299, "step": 35210 }, { "grad_norm": 0.18354295194149017, "learning_rate": 6.3832412455434e-05, "loss": 0.0132, "step": 35220 }, { "grad_norm": 0.2077685296535492, "learning_rate": 6.381254956220841e-05, "loss": 0.0164, "step": 35230 }, { "grad_norm": 0.1765270233154297, "learning_rate": 6.379268430879586e-05, "loss": 0.019, "step": 35240 }, { "grad_norm": 0.25150272250175476, "learning_rate": 6.37728166985908e-05, "loss": 0.015, "step": 35250 }, { "grad_norm": 0.38550323247909546, "learning_rate": 6.375294673498804e-05, "loss": 0.0154, "step": 35260 }, { "grad_norm": 0.2260049283504486, "learning_rate": 6.373307442138284e-05, "loss": 0.0171, "step": 35270 }, { "grad_norm": 0.3465149402618408, "learning_rate": 6.371319976117081e-05, "loss": 0.0275, "step": 35280 }, { "grad_norm": 0.2366846650838852, "learning_rate": 6.3693322757748e-05, "loss": 0.0161, "step": 35290 }, { "grad_norm": 0.2814082205295563, "learning_rate": 6.367344341451086e-05, "loss": 0.016, "step": 35300 }, { "grad_norm": 0.3336634337902069, "learning_rate": 6.36535617348562e-05, "loss": 0.0181, "step": 35310 }, { "grad_norm": 0.2977210581302643, "learning_rate": 6.363367772218128e-05, "loss": 0.0186, "step": 35320 }, { "grad_norm": 0.24494001269340515, "learning_rate": 6.36137913798837e-05, "loss": 0.019, "step": 35330 }, { "grad_norm": 0.2202952802181244, "learning_rate": 6.359390271136151e-05, "loss": 0.0146, "step": 35340 }, { "grad_norm": 0.26180997490882874, "learning_rate": 6.357401172001314e-05, "loss": 0.0155, "step": 35350 }, { "grad_norm": 0.30617755651474, "learning_rate": 6.355411840923742e-05, "loss": 0.0142, "step": 35360 }, { "grad_norm": 0.2534100413322449, "learning_rate": 6.353422278243358e-05, "loss": 0.0136, "step": 35370 }, { "grad_norm": 0.27914005517959595, "learning_rate": 6.351432484300121e-05, "loss": 0.0179, "step": 35380 }, { "grad_norm": 0.23470503091812134, "learning_rate": 6.349442459434036e-05, "loss": 0.0193, "step": 35390 }, { "grad_norm": 0.23889128863811493, "learning_rate": 6.34745220398514e-05, "loss": 0.0199, "step": 35400 }, { "grad_norm": 0.2832832336425781, "learning_rate": 6.345461718293518e-05, "loss": 0.0157, "step": 35410 }, { "grad_norm": 0.49419018626213074, "learning_rate": 6.343471002699286e-05, "loss": 0.0261, "step": 35420 }, { "grad_norm": 0.3910737633705139, "learning_rate": 6.341480057542602e-05, "loss": 0.0217, "step": 35430 }, { "grad_norm": 0.20738263428211212, "learning_rate": 6.339488883163667e-05, "loss": 0.015, "step": 35440 }, { "grad_norm": 0.4051576852798462, "learning_rate": 6.337497479902716e-05, "loss": 0.0157, "step": 35450 }, { "grad_norm": 0.25487369298934937, "learning_rate": 6.335505848100027e-05, "loss": 0.0151, "step": 35460 }, { "grad_norm": 0.2225569188594818, "learning_rate": 6.333513988095915e-05, "loss": 0.0159, "step": 35470 }, { "grad_norm": 0.1883213371038437, "learning_rate": 6.331521900230735e-05, "loss": 0.0167, "step": 35480 }, { "grad_norm": 0.2433248907327652, "learning_rate": 6.329529584844878e-05, "loss": 0.0163, "step": 35490 }, { "grad_norm": 0.34500375390052795, "learning_rate": 6.327537042278777e-05, "loss": 0.021, "step": 35500 }, { "grad_norm": 0.15854038298130035, "learning_rate": 6.325544272872905e-05, "loss": 0.0167, "step": 35510 }, { "grad_norm": 0.22824490070343018, "learning_rate": 6.323551276967771e-05, "loss": 0.0192, "step": 35520 }, { "grad_norm": 0.2598012089729309, "learning_rate": 6.321558054903922e-05, "loss": 0.0148, "step": 35530 }, { "grad_norm": 0.24305681884288788, "learning_rate": 6.319564607021947e-05, "loss": 0.0142, "step": 35540 }, { "grad_norm": 0.263654887676239, "learning_rate": 6.31757093366247e-05, "loss": 0.0129, "step": 35550 }, { "grad_norm": 0.2352122813463211, "learning_rate": 6.315577035166154e-05, "loss": 0.0147, "step": 35560 }, { "grad_norm": 0.2252400815486908, "learning_rate": 6.313582911873708e-05, "loss": 0.0151, "step": 35570 }, { "grad_norm": 0.24148184061050415, "learning_rate": 6.311588564125865e-05, "loss": 0.0136, "step": 35580 }, { "grad_norm": 0.31903186440467834, "learning_rate": 6.30959399226341e-05, "loss": 0.0172, "step": 35590 }, { "grad_norm": 0.22047732770442963, "learning_rate": 6.30759919662716e-05, "loss": 0.0241, "step": 35600 }, { "grad_norm": 0.24946579337120056, "learning_rate": 6.30560417755797e-05, "loss": 0.0163, "step": 35610 }, { "grad_norm": 0.15844807028770447, "learning_rate": 6.303608935396735e-05, "loss": 0.0125, "step": 35620 }, { "grad_norm": 0.16324588656425476, "learning_rate": 6.301613470484386e-05, "loss": 0.0161, "step": 35630 }, { "grad_norm": 0.3061490058898926, "learning_rate": 6.299617783161893e-05, "loss": 0.0155, "step": 35640 }, { "grad_norm": 0.23714475333690643, "learning_rate": 6.297621873770266e-05, "loss": 0.0133, "step": 35650 }, { "grad_norm": 0.4816874861717224, "learning_rate": 6.29562574265055e-05, "loss": 0.0146, "step": 35660 }, { "grad_norm": 0.2776983976364136, "learning_rate": 6.293629390143834e-05, "loss": 0.016, "step": 35670 }, { "grad_norm": 0.15628252923488617, "learning_rate": 6.291632816591232e-05, "loss": 0.014, "step": 35680 }, { "grad_norm": 0.20505349338054657, "learning_rate": 6.28963602233391e-05, "loss": 0.0185, "step": 35690 }, { "grad_norm": 0.2601027488708496, "learning_rate": 6.287639007713062e-05, "loss": 0.0192, "step": 35700 }, { "grad_norm": 0.18443447351455688, "learning_rate": 6.285641773069926e-05, "loss": 0.0161, "step": 35710 }, { "grad_norm": 0.24740369617938995, "learning_rate": 6.283644318745773e-05, "loss": 0.0129, "step": 35720 }, { "grad_norm": 0.1890406310558319, "learning_rate": 6.281646645081912e-05, "loss": 0.0184, "step": 35730 }, { "grad_norm": 0.3424789309501648, "learning_rate": 6.279648752419693e-05, "loss": 0.0136, "step": 35740 }, { "grad_norm": 0.24297797679901123, "learning_rate": 6.2776506411005e-05, "loss": 0.0132, "step": 35750 }, { "grad_norm": 0.2070161998271942, "learning_rate": 6.275652311465758e-05, "loss": 0.0201, "step": 35760 }, { "grad_norm": 0.19793620705604553, "learning_rate": 6.273653763856926e-05, "loss": 0.0222, "step": 35770 }, { "grad_norm": 0.3265361189842224, "learning_rate": 6.271654998615501e-05, "loss": 0.017, "step": 35780 }, { "grad_norm": 0.3076416850090027, "learning_rate": 6.269656016083013e-05, "loss": 0.0167, "step": 35790 }, { "grad_norm": 0.1802975982427597, "learning_rate": 6.267656816601038e-05, "loss": 0.0176, "step": 35800 }, { "grad_norm": 0.18931163847446442, "learning_rate": 6.265657400511185e-05, "loss": 0.0234, "step": 35810 }, { "grad_norm": 0.24589066207408905, "learning_rate": 6.263657768155098e-05, "loss": 0.0246, "step": 35820 }, { "grad_norm": 0.22281232476234436, "learning_rate": 6.261657919874457e-05, "loss": 0.023, "step": 35830 }, { "grad_norm": 0.13900120556354523, "learning_rate": 6.259657856010986e-05, "loss": 0.0217, "step": 35840 }, { "grad_norm": 0.26881009340286255, "learning_rate": 6.257657576906439e-05, "loss": 0.0155, "step": 35850 }, { "grad_norm": 0.2672398090362549, "learning_rate": 6.255657082902609e-05, "loss": 0.013, "step": 35860 }, { "grad_norm": 0.2540406286716461, "learning_rate": 6.253656374341325e-05, "loss": 0.011, "step": 35870 }, { "grad_norm": 0.3631613552570343, "learning_rate": 6.251655451564457e-05, "loss": 0.0186, "step": 35880 }, { "grad_norm": 0.21014469861984253, "learning_rate": 6.249654314913902e-05, "loss": 0.0162, "step": 35890 }, { "grad_norm": 0.43604013323783875, "learning_rate": 6.247652964731604e-05, "loss": 0.017, "step": 35900 }, { "grad_norm": 0.15626850724220276, "learning_rate": 6.245651401359537e-05, "loss": 0.0313, "step": 35910 }, { "grad_norm": 0.21945036947727203, "learning_rate": 6.243649625139715e-05, "loss": 0.0196, "step": 35920 }, { "grad_norm": 0.278209924697876, "learning_rate": 6.241647636414185e-05, "loss": 0.0172, "step": 35930 }, { "grad_norm": 0.2817797362804413, "learning_rate": 6.239645435525034e-05, "loss": 0.022, "step": 35940 }, { "grad_norm": 0.19903361797332764, "learning_rate": 6.237643022814381e-05, "loss": 0.0177, "step": 35950 }, { "grad_norm": 0.2943781018257141, "learning_rate": 6.235640398624386e-05, "loss": 0.015, "step": 35960 }, { "grad_norm": 0.1894761323928833, "learning_rate": 6.233637563297243e-05, "loss": 0.0174, "step": 35970 }, { "grad_norm": 0.3147399127483368, "learning_rate": 6.23163451717518e-05, "loss": 0.0192, "step": 35980 }, { "grad_norm": 0.3083718419075012, "learning_rate": 6.229631260600463e-05, "loss": 0.0172, "step": 35990 }, { "grad_norm": 0.21293333172798157, "learning_rate": 6.227627793915392e-05, "loss": 0.0258, "step": 36000 }, { "grad_norm": 0.26668375730514526, "learning_rate": 6.225624117462309e-05, "loss": 0.0233, "step": 36010 }, { "grad_norm": 0.2605346441268921, "learning_rate": 6.223620231583586e-05, "loss": 0.0169, "step": 36020 }, { "grad_norm": 0.3008548319339752, "learning_rate": 6.221616136621629e-05, "loss": 0.0154, "step": 36030 }, { "grad_norm": 0.28199976682662964, "learning_rate": 6.219611832918887e-05, "loss": 0.0197, "step": 36040 }, { "grad_norm": 0.22200605273246765, "learning_rate": 6.217607320817838e-05, "loss": 0.0137, "step": 36050 }, { "grad_norm": 0.25856244564056396, "learning_rate": 6.215602600661001e-05, "loss": 0.0152, "step": 36060 }, { "grad_norm": 0.20869866013526917, "learning_rate": 6.213597672790925e-05, "loss": 0.014, "step": 36070 }, { "grad_norm": 0.17118442058563232, "learning_rate": 6.2115925375502e-05, "loss": 0.0172, "step": 36080 }, { "grad_norm": 0.21503028273582458, "learning_rate": 6.209587195281447e-05, "loss": 0.0138, "step": 36090 }, { "grad_norm": 0.2809486985206604, "learning_rate": 6.207581646327324e-05, "loss": 0.0198, "step": 36100 }, { "grad_norm": 0.131461501121521, "learning_rate": 6.205575891030526e-05, "loss": 0.0134, "step": 36110 }, { "grad_norm": 0.22416947782039642, "learning_rate": 6.203569929733781e-05, "loss": 0.0146, "step": 36120 }, { "grad_norm": 0.25739532709121704, "learning_rate": 6.201563762779852e-05, "loss": 0.0165, "step": 36130 }, { "grad_norm": 0.18617792427539825, "learning_rate": 6.199557390511538e-05, "loss": 0.0134, "step": 36140 }, { "grad_norm": 0.1589193195104599, "learning_rate": 6.197550813271675e-05, "loss": 0.0122, "step": 36150 }, { "grad_norm": 0.21989412605762482, "learning_rate": 6.195544031403131e-05, "loss": 0.0157, "step": 36160 }, { "grad_norm": 0.3034663200378418, "learning_rate": 6.19353704524881e-05, "loss": 0.017, "step": 36170 }, { "grad_norm": 0.22568091750144958, "learning_rate": 6.191529855151652e-05, "loss": 0.0152, "step": 36180 }, { "grad_norm": 0.2076205462217331, "learning_rate": 6.189522461454629e-05, "loss": 0.018, "step": 36190 }, { "grad_norm": 0.2075408399105072, "learning_rate": 6.187514864500752e-05, "loss": 0.0201, "step": 36200 }, { "grad_norm": 0.27550458908081055, "learning_rate": 6.185507064633062e-05, "loss": 0.0144, "step": 36210 }, { "grad_norm": 0.22756528854370117, "learning_rate": 6.18349906219464e-05, "loss": 0.0178, "step": 36220 }, { "grad_norm": 0.2349841445684433, "learning_rate": 6.181490857528596e-05, "loss": 0.0163, "step": 36230 }, { "grad_norm": 0.1522255688905716, "learning_rate": 6.179482450978077e-05, "loss": 0.0106, "step": 36240 }, { "grad_norm": 0.17989957332611084, "learning_rate": 6.177473842886269e-05, "loss": 0.0126, "step": 36250 }, { "grad_norm": 0.2979081869125366, "learning_rate": 6.175465033596382e-05, "loss": 0.0168, "step": 36260 }, { "grad_norm": 0.1957027167081833, "learning_rate": 6.173456023451671e-05, "loss": 0.0169, "step": 36270 }, { "grad_norm": 0.21242044866085052, "learning_rate": 6.171446812795422e-05, "loss": 0.02, "step": 36280 }, { "grad_norm": 0.1984703093767166, "learning_rate": 6.169437401970949e-05, "loss": 0.0297, "step": 36290 }, { "grad_norm": 0.241494283080101, "learning_rate": 6.16742779132161e-05, "loss": 0.0172, "step": 36300 }, { "grad_norm": 0.3901979327201843, "learning_rate": 6.165417981190789e-05, "loss": 0.0243, "step": 36310 }, { "grad_norm": 0.1863996386528015, "learning_rate": 6.16340797192191e-05, "loss": 0.0158, "step": 36320 }, { "grad_norm": 0.5114160776138306, "learning_rate": 6.161397763858427e-05, "loss": 0.0222, "step": 36330 }, { "grad_norm": 0.19298596680164337, "learning_rate": 6.159387357343834e-05, "loss": 0.0138, "step": 36340 }, { "grad_norm": 0.24858775734901428, "learning_rate": 6.157376752721648e-05, "loss": 0.0175, "step": 36350 }, { "grad_norm": 0.20275087654590607, "learning_rate": 6.155365950335428e-05, "loss": 0.0199, "step": 36360 }, { "grad_norm": 0.3281090557575226, "learning_rate": 6.153354950528768e-05, "loss": 0.0201, "step": 36370 }, { "grad_norm": 0.28510618209838867, "learning_rate": 6.151343753645293e-05, "loss": 0.0201, "step": 36380 }, { "grad_norm": 0.23217126727104187, "learning_rate": 6.149332360028657e-05, "loss": 0.0167, "step": 36390 }, { "grad_norm": 0.21050044894218445, "learning_rate": 6.147320770022555e-05, "loss": 0.0185, "step": 36400 }, { "grad_norm": 0.2324826568365097, "learning_rate": 6.145308983970715e-05, "loss": 0.0202, "step": 36410 }, { "grad_norm": 0.7374110817909241, "learning_rate": 6.143297002216892e-05, "loss": 0.0175, "step": 36420 }, { "grad_norm": 0.30170372128486633, "learning_rate": 6.141284825104882e-05, "loss": 0.0189, "step": 36430 }, { "grad_norm": 0.20753750205039978, "learning_rate": 6.13927245297851e-05, "loss": 0.0189, "step": 36440 }, { "grad_norm": 0.2698177099227905, "learning_rate": 6.137259886181633e-05, "loss": 0.0178, "step": 36450 }, { "grad_norm": 0.2272961437702179, "learning_rate": 6.135247125058145e-05, "loss": 0.0232, "step": 36460 }, { "grad_norm": 0.16449978947639465, "learning_rate": 6.133234169951974e-05, "loss": 0.0112, "step": 36470 }, { "grad_norm": 0.17148935794830322, "learning_rate": 6.131221021207078e-05, "loss": 0.0197, "step": 36480 }, { "grad_norm": 0.23721005022525787, "learning_rate": 6.129207679167448e-05, "loss": 0.0161, "step": 36490 }, { "grad_norm": 0.16281671822071075, "learning_rate": 6.127194144177109e-05, "loss": 0.0144, "step": 36500 }, { "grad_norm": 0.14413337409496307, "learning_rate": 6.125180416580118e-05, "loss": 0.0123, "step": 36510 }, { "grad_norm": 0.17805540561676025, "learning_rate": 6.123166496720571e-05, "loss": 0.0169, "step": 36520 }, { "grad_norm": 0.5534072518348694, "learning_rate": 6.121152384942588e-05, "loss": 0.016, "step": 36530 }, { "grad_norm": 0.2742452621459961, "learning_rate": 6.119138081590324e-05, "loss": 0.0183, "step": 36540 }, { "grad_norm": 0.20119212567806244, "learning_rate": 6.117123587007971e-05, "loss": 0.0128, "step": 36550 }, { "grad_norm": 0.22628647089004517, "learning_rate": 6.11510890153975e-05, "loss": 0.0176, "step": 36560 }, { "grad_norm": 0.3157965838909149, "learning_rate": 6.113094025529916e-05, "loss": 0.0163, "step": 36570 }, { "grad_norm": 0.17764049768447876, "learning_rate": 6.111078959322757e-05, "loss": 0.0249, "step": 36580 }, { "grad_norm": 0.30650249123573303, "learning_rate": 6.109063703262592e-05, "loss": 0.0242, "step": 36590 }, { "grad_norm": 0.26642221212387085, "learning_rate": 6.107048257693772e-05, "loss": 0.0167, "step": 36600 }, { "grad_norm": 0.20050948858261108, "learning_rate": 6.105032622960683e-05, "loss": 0.0161, "step": 36610 }, { "grad_norm": 0.2335335910320282, "learning_rate": 6.103016799407743e-05, "loss": 0.015, "step": 36620 }, { "grad_norm": 0.2896946668624878, "learning_rate": 6.1010007873793984e-05, "loss": 0.0213, "step": 36630 }, { "grad_norm": 0.22437888383865356, "learning_rate": 6.098984587220131e-05, "loss": 0.0198, "step": 36640 }, { "grad_norm": 0.20183493196964264, "learning_rate": 6.096968199274456e-05, "loss": 0.0158, "step": 36650 }, { "grad_norm": 0.23727433383464813, "learning_rate": 6.0949516238869166e-05, "loss": 0.0212, "step": 36660 }, { "grad_norm": 0.3442381024360657, "learning_rate": 6.092934861402092e-05, "loss": 0.0216, "step": 36670 }, { "grad_norm": 0.2182006537914276, "learning_rate": 6.0909179121645924e-05, "loss": 0.0256, "step": 36680 }, { "grad_norm": 0.23499388992786407, "learning_rate": 6.0889007765190576e-05, "loss": 0.0202, "step": 36690 }, { "grad_norm": 0.14416444301605225, "learning_rate": 6.086883454810162e-05, "loss": 0.0151, "step": 36700 }, { "grad_norm": 0.26586854457855225, "learning_rate": 6.0848659473826084e-05, "loss": 0.0168, "step": 36710 }, { "grad_norm": 0.29717230796813965, "learning_rate": 6.082848254581138e-05, "loss": 0.017, "step": 36720 }, { "grad_norm": 0.25152188539505005, "learning_rate": 6.080830376750517e-05, "loss": 0.0143, "step": 36730 }, { "grad_norm": 0.1943407505750656, "learning_rate": 6.0788123142355445e-05, "loss": 0.0149, "step": 36740 }, { "grad_norm": 0.31252726912498474, "learning_rate": 6.076794067381052e-05, "loss": 0.0164, "step": 36750 }, { "grad_norm": 0.22026091814041138, "learning_rate": 6.074775636531905e-05, "loss": 0.0171, "step": 36760 }, { "grad_norm": 0.3882738947868347, "learning_rate": 6.072757022032997e-05, "loss": 0.0177, "step": 36770 }, { "grad_norm": 0.19023337960243225, "learning_rate": 6.070738224229253e-05, "loss": 0.0197, "step": 36780 }, { "grad_norm": 0.15596459805965424, "learning_rate": 6.0687192434656314e-05, "loss": 0.0174, "step": 36790 }, { "grad_norm": 0.2506418526172638, "learning_rate": 6.066700080087121e-05, "loss": 0.016, "step": 36800 }, { "grad_norm": 0.22231334447860718, "learning_rate": 6.0646807344387424e-05, "loss": 0.0165, "step": 36810 }, { "grad_norm": 0.22055254876613617, "learning_rate": 6.062661206865543e-05, "loss": 0.0165, "step": 36820 }, { "grad_norm": 0.25238171219825745, "learning_rate": 6.06064149771261e-05, "loss": 0.0175, "step": 36830 }, { "grad_norm": 0.22532254457473755, "learning_rate": 6.058621607325051e-05, "loss": 0.0153, "step": 36840 }, { "grad_norm": 0.27412763237953186, "learning_rate": 6.056601536048014e-05, "loss": 0.016, "step": 36850 }, { "grad_norm": 0.3886145055294037, "learning_rate": 6.0545812842266725e-05, "loss": 0.0228, "step": 36860 }, { "grad_norm": 0.12466207891702652, "learning_rate": 6.052560852206232e-05, "loss": 0.0164, "step": 36870 }, { "grad_norm": 0.19385507702827454, "learning_rate": 6.05054024033193e-05, "loss": 0.0276, "step": 36880 }, { "grad_norm": 0.26235976815223694, "learning_rate": 6.048519448949032e-05, "loss": 0.0156, "step": 36890 }, { "grad_norm": 0.11603844165802002, "learning_rate": 6.046498478402839e-05, "loss": 0.0173, "step": 36900 }, { "grad_norm": 0.19152288138866425, "learning_rate": 6.044477329038677e-05, "loss": 0.0158, "step": 36910 }, { "grad_norm": 0.24799437820911407, "learning_rate": 6.042456001201906e-05, "loss": 0.0167, "step": 36920 }, { "grad_norm": 0.25466009974479675, "learning_rate": 6.040434495237917e-05, "loss": 0.0185, "step": 36930 }, { "grad_norm": 0.29203760623931885, "learning_rate": 6.0384128114921256e-05, "loss": 0.0191, "step": 36940 }, { "grad_norm": 0.19478438794612885, "learning_rate": 6.036390950309987e-05, "loss": 0.0124, "step": 36950 }, { "grad_norm": 0.21326379477977753, "learning_rate": 6.0343689120369805e-05, "loss": 0.0172, "step": 36960 }, { "grad_norm": 0.2618653178215027, "learning_rate": 6.032346697018616e-05, "loss": 0.0132, "step": 36970 }, { "grad_norm": 0.23883144557476044, "learning_rate": 6.0303243056004375e-05, "loss": 0.0239, "step": 36980 }, { "grad_norm": 0.30554598569869995, "learning_rate": 6.0283017381280136e-05, "loss": 0.0159, "step": 36990 }, { "grad_norm": 0.2420797199010849, "learning_rate": 6.026278994946948e-05, "loss": 0.0134, "step": 37000 }, { "grad_norm": 0.17964735627174377, "learning_rate": 6.02425607640287e-05, "loss": 0.0185, "step": 37010 }, { "grad_norm": 0.17754127085208893, "learning_rate": 6.022232982841441e-05, "loss": 0.0151, "step": 37020 }, { "grad_norm": 0.20900069177150726, "learning_rate": 6.020209714608355e-05, "loss": 0.0172, "step": 37030 }, { "grad_norm": 0.35824811458587646, "learning_rate": 6.018186272049331e-05, "loss": 0.0206, "step": 37040 }, { "grad_norm": 0.25043386220932007, "learning_rate": 6.01616265551012e-05, "loss": 0.018, "step": 37050 }, { "grad_norm": 0.1664978563785553, "learning_rate": 6.014138865336503e-05, "loss": 0.0173, "step": 37060 }, { "grad_norm": 0.26895156502723694, "learning_rate": 6.0121149018742905e-05, "loss": 0.0196, "step": 37070 }, { "grad_norm": 0.2604411840438843, "learning_rate": 6.010090765469325e-05, "loss": 0.0165, "step": 37080 }, { "grad_norm": 0.2316180318593979, "learning_rate": 6.008066456467473e-05, "loss": 0.0164, "step": 37090 }, { "grad_norm": 0.26230964064598083, "learning_rate": 6.0060419752146335e-05, "loss": 0.016, "step": 37100 }, { "grad_norm": 0.23101426661014557, "learning_rate": 6.0040173220567353e-05, "loss": 0.0133, "step": 37110 }, { "grad_norm": 0.20288203656673431, "learning_rate": 6.001992497339737e-05, "loss": 0.0154, "step": 37120 }, { "grad_norm": 0.2816343903541565, "learning_rate": 5.999967501409626e-05, "loss": 0.0209, "step": 37130 }, { "grad_norm": 0.1751715987920761, "learning_rate": 5.997942334612418e-05, "loss": 0.0171, "step": 37140 }, { "grad_norm": 0.19020847976207733, "learning_rate": 5.995916997294158e-05, "loss": 0.0119, "step": 37150 }, { "grad_norm": 0.20585359632968903, "learning_rate": 5.9938914898009214e-05, "loss": 0.0195, "step": 37160 }, { "grad_norm": 0.16738905012607574, "learning_rate": 5.991865812478813e-05, "loss": 0.0144, "step": 37170 }, { "grad_norm": 0.29982316493988037, "learning_rate": 5.989839965673964e-05, "loss": 0.0292, "step": 37180 }, { "grad_norm": 0.24457429349422455, "learning_rate": 5.987813949732539e-05, "loss": 0.0166, "step": 37190 }, { "grad_norm": 0.17231281101703644, "learning_rate": 5.9857877650007255e-05, "loss": 0.0173, "step": 37200 }, { "grad_norm": 0.18114633858203888, "learning_rate": 5.983761411824744e-05, "loss": 0.0151, "step": 37210 }, { "grad_norm": 0.29283174872398376, "learning_rate": 5.981734890550844e-05, "loss": 0.0149, "step": 37220 }, { "grad_norm": 0.18281865119934082, "learning_rate": 5.979708201525301e-05, "loss": 0.0154, "step": 37230 }, { "grad_norm": 0.2738310992717743, "learning_rate": 5.977681345094422e-05, "loss": 0.0177, "step": 37240 }, { "grad_norm": 0.29481837153434753, "learning_rate": 5.97565432160454e-05, "loss": 0.0166, "step": 37250 }, { "grad_norm": 0.23694215714931488, "learning_rate": 5.9736271314020186e-05, "loss": 0.0225, "step": 37260 }, { "grad_norm": 0.27050188183784485, "learning_rate": 5.971599774833251e-05, "loss": 0.0182, "step": 37270 }, { "grad_norm": 0.21427835524082184, "learning_rate": 5.9695722522446525e-05, "loss": 0.0176, "step": 37280 }, { "grad_norm": 0.2568056285381317, "learning_rate": 5.9675445639826765e-05, "loss": 0.0148, "step": 37290 }, { "grad_norm": 0.13818253576755524, "learning_rate": 5.965516710393796e-05, "loss": 0.0154, "step": 37300 }, { "grad_norm": 0.30709829926490784, "learning_rate": 5.963488691824516e-05, "loss": 0.0219, "step": 37310 }, { "grad_norm": 0.23755644261837006, "learning_rate": 5.96146050862137e-05, "loss": 0.0219, "step": 37320 }, { "grad_norm": 0.2726256847381592, "learning_rate": 5.959432161130919e-05, "loss": 0.0158, "step": 37330 }, { "grad_norm": 0.24511946737766266, "learning_rate": 5.9574036496997545e-05, "loss": 0.0197, "step": 37340 }, { "grad_norm": 0.24802206456661224, "learning_rate": 5.955374974674488e-05, "loss": 0.0137, "step": 37350 }, { "grad_norm": 0.18481110036373138, "learning_rate": 5.9533461364017696e-05, "loss": 0.015, "step": 37360 }, { "grad_norm": 0.18891778588294983, "learning_rate": 5.9513171352282716e-05, "loss": 0.0153, "step": 37370 }, { "grad_norm": 0.1476997435092926, "learning_rate": 5.949287971500692e-05, "loss": 0.0127, "step": 37380 }, { "grad_norm": 0.11193474382162094, "learning_rate": 5.947258645565762e-05, "loss": 0.012, "step": 37390 }, { "grad_norm": 0.23014582693576813, "learning_rate": 5.945229157770237e-05, "loss": 0.0145, "step": 37400 }, { "grad_norm": 0.17288559675216675, "learning_rate": 5.9431995084609006e-05, "loss": 0.0161, "step": 37410 }, { "grad_norm": 0.28653016686439514, "learning_rate": 5.941169697984564e-05, "loss": 0.0146, "step": 37420 }, { "grad_norm": 0.21916553378105164, "learning_rate": 5.9391397266880675e-05, "loss": 0.0116, "step": 37430 }, { "grad_norm": 0.30886080861091614, "learning_rate": 5.937109594918279e-05, "loss": 0.0241, "step": 37440 }, { "grad_norm": 0.23318366706371307, "learning_rate": 5.9350793030220884e-05, "loss": 0.0308, "step": 37450 }, { "grad_norm": 0.3177896738052368, "learning_rate": 5.933048851346421e-05, "loss": 0.0135, "step": 37460 }, { "grad_norm": 0.2602175176143646, "learning_rate": 5.931018240238224e-05, "loss": 0.027, "step": 37470 }, { "grad_norm": 0.17763552069664001, "learning_rate": 5.928987470044471e-05, "loss": 0.0171, "step": 37480 }, { "grad_norm": 0.27266085147857666, "learning_rate": 5.9269565411121695e-05, "loss": 0.0161, "step": 37490 }, { "grad_norm": 0.26743626594543457, "learning_rate": 5.924925453788347e-05, "loss": 0.0191, "step": 37500 }, { "grad_norm": 0.23404888808727264, "learning_rate": 5.92289420842006e-05, "loss": 0.0187, "step": 37510 }, { "grad_norm": 0.2640441358089447, "learning_rate": 5.9208628053543945e-05, "loss": 0.0268, "step": 37520 }, { "grad_norm": 0.3481324315071106, "learning_rate": 5.918831244938462e-05, "loss": 0.0188, "step": 37530 }, { "grad_norm": 0.26039254665374756, "learning_rate": 5.9167995275194e-05, "loss": 0.0151, "step": 37540 }, { "grad_norm": 0.2864885628223419, "learning_rate": 5.914767653444373e-05, "loss": 0.0206, "step": 37550 }, { "grad_norm": 0.39628273248672485, "learning_rate": 5.912735623060572e-05, "loss": 0.0231, "step": 37560 }, { "grad_norm": 0.15550842881202698, "learning_rate": 5.910703436715217e-05, "loss": 0.0169, "step": 37570 }, { "grad_norm": 0.18091130256652832, "learning_rate": 5.908671094755552e-05, "loss": 0.0148, "step": 37580 }, { "grad_norm": 0.3743480443954468, "learning_rate": 5.906638597528851e-05, "loss": 0.0187, "step": 37590 }, { "grad_norm": 0.2235632687807083, "learning_rate": 5.9046059453824076e-05, "loss": 0.015, "step": 37600 }, { "grad_norm": 0.28075966238975525, "learning_rate": 5.9025731386635505e-05, "loss": 0.0142, "step": 37610 }, { "grad_norm": 0.24832946062088013, "learning_rate": 5.900540177719629e-05, "loss": 0.0175, "step": 37620 }, { "grad_norm": 0.26128557324409485, "learning_rate": 5.898507062898021e-05, "loss": 0.0192, "step": 37630 }, { "grad_norm": 0.4982539415359497, "learning_rate": 5.8964737945461316e-05, "loss": 0.0171, "step": 37640 }, { "grad_norm": 0.18379400670528412, "learning_rate": 5.8944403730113885e-05, "loss": 0.0211, "step": 37650 }, { "grad_norm": 0.33711811900138855, "learning_rate": 5.892406798641248e-05, "loss": 0.0186, "step": 37660 }, { "grad_norm": 0.4026443362236023, "learning_rate": 5.890373071783193e-05, "loss": 0.0154, "step": 37670 }, { "grad_norm": 0.3227899372577667, "learning_rate": 5.888339192784732e-05, "loss": 0.0194, "step": 37680 }, { "grad_norm": 0.22637610137462616, "learning_rate": 5.8863051619934003e-05, "loss": 0.018, "step": 37690 }, { "grad_norm": 0.20111660659313202, "learning_rate": 5.8842709797567554e-05, "loss": 0.0191, "step": 37700 }, { "grad_norm": 0.24931050837039948, "learning_rate": 5.8822366464223855e-05, "loss": 0.0161, "step": 37710 }, { "grad_norm": 0.2202039510011673, "learning_rate": 5.880202162337901e-05, "loss": 0.0181, "step": 37720 }, { "grad_norm": 0.21366667747497559, "learning_rate": 5.8781675278509405e-05, "loss": 0.0157, "step": 37730 }, { "grad_norm": 0.18878184258937836, "learning_rate": 5.8761327433091696e-05, "loss": 0.0149, "step": 37740 }, { "grad_norm": 0.2828807830810547, "learning_rate": 5.874097809060275e-05, "loss": 0.019, "step": 37750 }, { "grad_norm": 0.24751105904579163, "learning_rate": 5.87206272545197e-05, "loss": 0.0274, "step": 37760 }, { "grad_norm": 0.27973851561546326, "learning_rate": 5.8700274928319955e-05, "loss": 0.0165, "step": 37770 }, { "grad_norm": 0.2712504267692566, "learning_rate": 5.867992111548118e-05, "loss": 0.0143, "step": 37780 }, { "grad_norm": 0.35530513525009155, "learning_rate": 5.865956581948131e-05, "loss": 0.017, "step": 37790 }, { "grad_norm": 0.2317829430103302, "learning_rate": 5.863920904379845e-05, "loss": 0.015, "step": 37800 }, { "grad_norm": 0.32425665855407715, "learning_rate": 5.8618850791911064e-05, "loss": 0.022, "step": 37810 }, { "grad_norm": 0.21326886117458344, "learning_rate": 5.859849106729779e-05, "loss": 0.0164, "step": 37820 }, { "grad_norm": 0.24019722640514374, "learning_rate": 5.857812987343758e-05, "loss": 0.0169, "step": 37830 }, { "grad_norm": 0.2754792869091034, "learning_rate": 5.855776721380957e-05, "loss": 0.0197, "step": 37840 }, { "grad_norm": 0.21829773485660553, "learning_rate": 5.8537403091893217e-05, "loss": 0.0173, "step": 37850 }, { "grad_norm": 0.22635231912136078, "learning_rate": 5.851703751116816e-05, "loss": 0.0153, "step": 37860 }, { "grad_norm": 0.21313989162445068, "learning_rate": 5.8496670475114336e-05, "loss": 0.0158, "step": 37870 }, { "grad_norm": 0.3249996602535248, "learning_rate": 5.84763019872119e-05, "loss": 0.0202, "step": 37880 }, { "grad_norm": 0.28541144728660583, "learning_rate": 5.845593205094131e-05, "loss": 0.0172, "step": 37890 }, { "grad_norm": 0.2187160849571228, "learning_rate": 5.843556066978318e-05, "loss": 0.0135, "step": 37900 }, { "grad_norm": 0.16788586974143982, "learning_rate": 5.8415187847218455e-05, "loss": 0.018, "step": 37910 }, { "grad_norm": 0.4787147343158722, "learning_rate": 5.839481358672827e-05, "loss": 0.0149, "step": 37920 }, { "grad_norm": 0.29573920369148254, "learning_rate": 5.837443789179407e-05, "loss": 0.022, "step": 37930 }, { "grad_norm": 0.2872985601425171, "learning_rate": 5.8354060765897445e-05, "loss": 0.0173, "step": 37940 }, { "grad_norm": 0.4356560707092285, "learning_rate": 5.8333682212520334e-05, "loss": 0.0292, "step": 37950 }, { "grad_norm": 0.30452361702919006, "learning_rate": 5.831330223514486e-05, "loss": 0.0175, "step": 37960 }, { "grad_norm": 0.24836303293704987, "learning_rate": 5.8292920837253396e-05, "loss": 0.0143, "step": 37970 }, { "grad_norm": 0.262770414352417, "learning_rate": 5.827253802232857e-05, "loss": 0.014, "step": 37980 }, { "grad_norm": 0.41723841428756714, "learning_rate": 5.825215379385327e-05, "loss": 0.0127, "step": 37990 }, { "grad_norm": 0.21902072429656982, "learning_rate": 5.823176815531057e-05, "loss": 0.0284, "step": 38000 }, { "grad_norm": 0.19067713618278503, "learning_rate": 5.8211381110183826e-05, "loss": 0.0256, "step": 38010 }, { "grad_norm": 0.18991351127624512, "learning_rate": 5.8190992661956645e-05, "loss": 0.0136, "step": 38020 }, { "grad_norm": 0.1558532565832138, "learning_rate": 5.817060281411284e-05, "loss": 0.0195, "step": 38030 }, { "grad_norm": 0.24784663319587708, "learning_rate": 5.815021157013647e-05, "loss": 0.0156, "step": 38040 }, { "grad_norm": 0.24295970797538757, "learning_rate": 5.8129818933511856e-05, "loss": 0.0248, "step": 38050 }, { "grad_norm": 0.4231337308883667, "learning_rate": 5.8109424907723544e-05, "loss": 0.0178, "step": 38060 }, { "grad_norm": 0.19206082820892334, "learning_rate": 5.80890294962563e-05, "loss": 0.0119, "step": 38070 }, { "grad_norm": 0.24495892226696014, "learning_rate": 5.806863270259515e-05, "loss": 0.0184, "step": 38080 }, { "grad_norm": 0.2632710337638855, "learning_rate": 5.804823453022536e-05, "loss": 0.0157, "step": 38090 }, { "grad_norm": 0.2598998546600342, "learning_rate": 5.80278349826324e-05, "loss": 0.0159, "step": 38100 }, { "grad_norm": 0.2792172431945801, "learning_rate": 5.8007434063302014e-05, "loss": 0.0157, "step": 38110 }, { "grad_norm": 0.2041218876838684, "learning_rate": 5.7987031775720136e-05, "loss": 0.0179, "step": 38120 }, { "grad_norm": 0.16748234629631042, "learning_rate": 5.7966628123372976e-05, "loss": 0.021, "step": 38130 }, { "grad_norm": 0.33942270278930664, "learning_rate": 5.7946223109746956e-05, "loss": 0.0263, "step": 38140 }, { "grad_norm": 0.25979384779930115, "learning_rate": 5.7925816738328754e-05, "loss": 0.0125, "step": 38150 }, { "grad_norm": 0.25340792536735535, "learning_rate": 5.790540901260521e-05, "loss": 0.0153, "step": 38160 }, { "grad_norm": 0.21249596774578094, "learning_rate": 5.788499993606351e-05, "loss": 0.0176, "step": 38170 }, { "grad_norm": 0.21322765946388245, "learning_rate": 5.786458951219096e-05, "loss": 0.0176, "step": 38180 }, { "grad_norm": 0.17089000344276428, "learning_rate": 5.784417774447517e-05, "loss": 0.0189, "step": 38190 }, { "grad_norm": 0.19226892292499542, "learning_rate": 5.782376463640393e-05, "loss": 0.0168, "step": 38200 }, { "grad_norm": 0.24284961819648743, "learning_rate": 5.780335019146531e-05, "loss": 0.0157, "step": 38210 }, { "grad_norm": 0.37989774346351624, "learning_rate": 5.778293441314755e-05, "loss": 0.0196, "step": 38220 }, { "grad_norm": 0.25798776745796204, "learning_rate": 5.776251730493917e-05, "loss": 0.0239, "step": 38230 }, { "grad_norm": 0.40315955877304077, "learning_rate": 5.774209887032887e-05, "loss": 0.0139, "step": 38240 }, { "grad_norm": 0.1875094622373581, "learning_rate": 5.772167911280565e-05, "loss": 0.0143, "step": 38250 }, { "grad_norm": 0.23882395029067993, "learning_rate": 5.770125803585864e-05, "loss": 0.0147, "step": 38260 }, { "grad_norm": 0.22958657145500183, "learning_rate": 5.768083564297726e-05, "loss": 0.0207, "step": 38270 }, { "grad_norm": 0.23560386896133423, "learning_rate": 5.766041193765114e-05, "loss": 0.0174, "step": 38280 }, { "grad_norm": 0.30234357714653015, "learning_rate": 5.763998692337015e-05, "loss": 0.0126, "step": 38290 }, { "grad_norm": 0.20923906564712524, "learning_rate": 5.761956060362433e-05, "loss": 0.0172, "step": 38300 }, { "grad_norm": 0.24091528356075287, "learning_rate": 5.7599132981904e-05, "loss": 0.0168, "step": 38310 }, { "grad_norm": 0.22849078476428986, "learning_rate": 5.75787040616997e-05, "loss": 0.0218, "step": 38320 }, { "grad_norm": 0.24398328363895416, "learning_rate": 5.755827384650212e-05, "loss": 0.0162, "step": 38330 }, { "grad_norm": 0.24208980798721313, "learning_rate": 5.753784233980228e-05, "loss": 0.0153, "step": 38340 }, { "grad_norm": 0.24997539818286896, "learning_rate": 5.751740954509135e-05, "loss": 0.0125, "step": 38350 }, { "grad_norm": 0.17294757068157196, "learning_rate": 5.7496975465860715e-05, "loss": 0.0121, "step": 38360 }, { "grad_norm": 0.15108424425125122, "learning_rate": 5.747654010560202e-05, "loss": 0.0129, "step": 38370 }, { "grad_norm": 0.15707695484161377, "learning_rate": 5.7456103467807097e-05, "loss": 0.0165, "step": 38380 }, { "grad_norm": 0.23238380253314972, "learning_rate": 5.7435665555968046e-05, "loss": 0.0125, "step": 38390 }, { "grad_norm": 0.20108994841575623, "learning_rate": 5.74152263735771e-05, "loss": 0.0201, "step": 38400 }, { "grad_norm": 0.2563808858394623, "learning_rate": 5.739478592412677e-05, "loss": 0.016, "step": 38410 }, { "grad_norm": 0.32052311301231384, "learning_rate": 5.7374344211109766e-05, "loss": 0.021, "step": 38420 }, { "grad_norm": 0.22139303386211395, "learning_rate": 5.735390123801904e-05, "loss": 0.0205, "step": 38430 }, { "grad_norm": 0.3926257789134979, "learning_rate": 5.7333457008347704e-05, "loss": 0.0225, "step": 38440 }, { "grad_norm": 0.2714630961418152, "learning_rate": 5.7313011525589156e-05, "loss": 0.0193, "step": 38450 }, { "grad_norm": 0.4015464186668396, "learning_rate": 5.729256479323694e-05, "loss": 0.0177, "step": 38460 }, { "grad_norm": 0.18202830851078033, "learning_rate": 5.727211681478485e-05, "loss": 0.0172, "step": 38470 }, { "grad_norm": 0.19924695789813995, "learning_rate": 5.7251667593726886e-05, "loss": 0.0223, "step": 38480 }, { "grad_norm": 0.32443416118621826, "learning_rate": 5.723121713355728e-05, "loss": 0.0259, "step": 38490 }, { "grad_norm": 0.20766153931617737, "learning_rate": 5.721076543777044e-05, "loss": 0.0161, "step": 38500 }, { "grad_norm": 0.2226724773645401, "learning_rate": 5.7190312509860986e-05, "loss": 0.0154, "step": 38510 }, { "grad_norm": 0.6551321148872375, "learning_rate": 5.716985835332379e-05, "loss": 0.0172, "step": 38520 }, { "grad_norm": 0.21460884809494019, "learning_rate": 5.714940297165389e-05, "loss": 0.0194, "step": 38530 }, { "grad_norm": 0.2745901048183441, "learning_rate": 5.712894636834656e-05, "loss": 0.0146, "step": 38540 }, { "grad_norm": 0.17887839674949646, "learning_rate": 5.7108488546897276e-05, "loss": 0.0145, "step": 38550 }, { "grad_norm": 0.15506990253925323, "learning_rate": 5.708802951080172e-05, "loss": 0.01, "step": 38560 }, { "grad_norm": 0.2688819169998169, "learning_rate": 5.706756926355576e-05, "loss": 0.0207, "step": 38570 }, { "grad_norm": 0.18897639214992523, "learning_rate": 5.704710780865554e-05, "loss": 0.0139, "step": 38580 }, { "grad_norm": 0.5459610819816589, "learning_rate": 5.7026645149597325e-05, "loss": 0.0186, "step": 38590 }, { "grad_norm": 0.24503982067108154, "learning_rate": 5.700618128987764e-05, "loss": 0.0185, "step": 38600 }, { "grad_norm": 0.14155232906341553, "learning_rate": 5.698571623299317e-05, "loss": 0.015, "step": 38610 }, { "grad_norm": 0.22633568942546844, "learning_rate": 5.696524998244086e-05, "loss": 0.0148, "step": 38620 }, { "grad_norm": 0.22254931926727295, "learning_rate": 5.6944782541717836e-05, "loss": 0.021, "step": 38630 }, { "grad_norm": 0.17776253819465637, "learning_rate": 5.69243139143214e-05, "loss": 0.0172, "step": 38640 }, { "grad_norm": 0.2741353213787079, "learning_rate": 5.6903844103749125e-05, "loss": 0.0175, "step": 38650 }, { "grad_norm": 0.1871844381093979, "learning_rate": 5.688337311349869e-05, "loss": 0.0136, "step": 38660 }, { "grad_norm": 0.15208882093429565, "learning_rate": 5.6862900947068074e-05, "loss": 0.0136, "step": 38670 }, { "grad_norm": 0.31823986768722534, "learning_rate": 5.6842427607955374e-05, "loss": 0.0211, "step": 38680 }, { "grad_norm": 0.39476537704467773, "learning_rate": 5.682195309965893e-05, "loss": 0.0147, "step": 38690 }, { "grad_norm": 0.2873660922050476, "learning_rate": 5.6801477425677294e-05, "loss": 0.0212, "step": 38700 }, { "grad_norm": 0.242892786860466, "learning_rate": 5.678100058950917e-05, "loss": 0.0177, "step": 38710 }, { "grad_norm": 0.5704823732376099, "learning_rate": 5.676052259465352e-05, "loss": 0.0272, "step": 38720 }, { "grad_norm": 0.2743457853794098, "learning_rate": 5.674004344460945e-05, "loss": 0.0131, "step": 38730 }, { "grad_norm": 0.33335086703300476, "learning_rate": 5.6719563142876295e-05, "loss": 0.0155, "step": 38740 }, { "grad_norm": 0.19584526121616364, "learning_rate": 5.669908169295359e-05, "loss": 0.0138, "step": 38750 }, { "grad_norm": 0.18633168935775757, "learning_rate": 5.667859909834105e-05, "loss": 0.0178, "step": 38760 }, { "grad_norm": 0.2858293056488037, "learning_rate": 5.6658115362538565e-05, "loss": 0.0187, "step": 38770 }, { "grad_norm": 0.20574478805065155, "learning_rate": 5.663763048904628e-05, "loss": 0.015, "step": 38780 }, { "grad_norm": 0.3360702097415924, "learning_rate": 5.661714448136447e-05, "loss": 0.0133, "step": 38790 }, { "grad_norm": 0.24170450866222382, "learning_rate": 5.659665734299366e-05, "loss": 0.013, "step": 38800 }, { "grad_norm": 0.26595520973205566, "learning_rate": 5.6576169077434516e-05, "loss": 0.0196, "step": 38810 }, { "grad_norm": 0.14635466039180756, "learning_rate": 5.6555679688187944e-05, "loss": 0.0149, "step": 38820 }, { "grad_norm": 0.30663737654685974, "learning_rate": 5.6535189178755e-05, "loss": 0.0151, "step": 38830 }, { "grad_norm": 0.26241186261177063, "learning_rate": 5.651469755263698e-05, "loss": 0.0143, "step": 38840 }, { "grad_norm": 0.1339903622865677, "learning_rate": 5.6494204813335316e-05, "loss": 0.0257, "step": 38850 }, { "grad_norm": 0.17837649583816528, "learning_rate": 5.647371096435168e-05, "loss": 0.015, "step": 38860 }, { "grad_norm": 0.23661892116069794, "learning_rate": 5.645321600918788e-05, "loss": 0.0145, "step": 38870 }, { "grad_norm": 0.325788289308548, "learning_rate": 5.643271995134597e-05, "loss": 0.0186, "step": 38880 }, { "grad_norm": 0.18016144633293152, "learning_rate": 5.641222279432814e-05, "loss": 0.0164, "step": 38890 }, { "grad_norm": 0.22705672681331635, "learning_rate": 5.6391724541636834e-05, "loss": 0.0141, "step": 38900 }, { "grad_norm": 0.4390092194080353, "learning_rate": 5.6371225196774605e-05, "loss": 0.0176, "step": 38910 }, { "grad_norm": 0.24088945984840393, "learning_rate": 5.635072476324423e-05, "loss": 0.0203, "step": 38920 }, { "grad_norm": 0.3086429238319397, "learning_rate": 5.63302232445487e-05, "loss": 0.0158, "step": 38930 }, { "grad_norm": 0.15760153532028198, "learning_rate": 5.6309720644191144e-05, "loss": 0.0168, "step": 38940 }, { "grad_norm": 0.20424172282218933, "learning_rate": 5.628921696567491e-05, "loss": 0.0154, "step": 38950 }, { "grad_norm": 0.42960476875305176, "learning_rate": 5.62687122125035e-05, "loss": 0.0155, "step": 38960 }, { "grad_norm": 0.2338401824235916, "learning_rate": 5.624820638818062e-05, "loss": 0.0149, "step": 38970 }, { "grad_norm": 0.2587290406227112, "learning_rate": 5.6227699496210164e-05, "loss": 0.0256, "step": 38980 }, { "grad_norm": 0.19036948680877686, "learning_rate": 5.6207191540096195e-05, "loss": 0.0177, "step": 38990 }, { "grad_norm": 0.3083095848560333, "learning_rate": 5.618668252334296e-05, "loss": 0.0116, "step": 39000 }, { "grad_norm": 0.1507415920495987, "learning_rate": 5.616617244945488e-05, "loss": 0.0137, "step": 39010 }, { "grad_norm": 0.19341732561588287, "learning_rate": 5.614566132193656e-05, "loss": 0.015, "step": 39020 }, { "grad_norm": 0.28040385246276855, "learning_rate": 5.612514914429282e-05, "loss": 0.016, "step": 39030 }, { "grad_norm": 0.29001739621162415, "learning_rate": 5.610463592002863e-05, "loss": 0.0151, "step": 39040 }, { "grad_norm": 0.27354755997657776, "learning_rate": 5.608412165264909e-05, "loss": 0.0149, "step": 39050 }, { "grad_norm": 0.3442119359970093, "learning_rate": 5.606360634565959e-05, "loss": 0.0126, "step": 39060 }, { "grad_norm": 0.12544219195842743, "learning_rate": 5.604309000256559e-05, "loss": 0.0145, "step": 39070 }, { "grad_norm": 0.2409738451242447, "learning_rate": 5.6022572626872785e-05, "loss": 0.0121, "step": 39080 }, { "grad_norm": 0.3023698031902313, "learning_rate": 5.600205422208704e-05, "loss": 0.0182, "step": 39090 }, { "grad_norm": 0.17871630191802979, "learning_rate": 5.59815347917144e-05, "loss": 0.0167, "step": 39100 }, { "grad_norm": 0.41978588700294495, "learning_rate": 5.596101433926103e-05, "loss": 0.0182, "step": 39110 }, { "grad_norm": 0.30455318093299866, "learning_rate": 5.5940492868233364e-05, "loss": 0.0157, "step": 39120 }, { "grad_norm": 0.5755547285079956, "learning_rate": 5.591997038213793e-05, "loss": 0.0197, "step": 39130 }, { "grad_norm": 0.20004642009735107, "learning_rate": 5.5899446884481475e-05, "loss": 0.012, "step": 39140 }, { "grad_norm": 0.2790363132953644, "learning_rate": 5.5878922378770906e-05, "loss": 0.0254, "step": 39150 }, { "grad_norm": 0.2583259344100952, "learning_rate": 5.5858396868513285e-05, "loss": 0.0174, "step": 39160 }, { "grad_norm": 0.23973625898361206, "learning_rate": 5.583787035721586e-05, "loss": 0.0128, "step": 39170 }, { "grad_norm": 0.25642523169517517, "learning_rate": 5.581734284838606e-05, "loss": 0.0146, "step": 39180 }, { "grad_norm": 0.3218405544757843, "learning_rate": 5.579681434553147e-05, "loss": 0.0153, "step": 39190 }, { "grad_norm": 0.19049657881259918, "learning_rate": 5.5776284852159854e-05, "loss": 0.0134, "step": 39200 }, { "grad_norm": 0.23673969507217407, "learning_rate": 5.575575437177913e-05, "loss": 0.022, "step": 39210 }, { "grad_norm": 0.2119424045085907, "learning_rate": 5.573522290789742e-05, "loss": 0.0147, "step": 39220 }, { "grad_norm": 0.14003200829029083, "learning_rate": 5.571469046402298e-05, "loss": 0.0172, "step": 39230 }, { "grad_norm": 0.19706659018993378, "learning_rate": 5.5694157043664205e-05, "loss": 0.0188, "step": 39240 }, { "grad_norm": 0.37323930859565735, "learning_rate": 5.567362265032975e-05, "loss": 0.0126, "step": 39250 }, { "grad_norm": 0.19581767916679382, "learning_rate": 5.565308728752836e-05, "loss": 0.0148, "step": 39260 }, { "grad_norm": 0.23685459792613983, "learning_rate": 5.5632550958768956e-05, "loss": 0.0162, "step": 39270 }, { "grad_norm": 0.23823615908622742, "learning_rate": 5.5612013667560636e-05, "loss": 0.0146, "step": 39280 }, { "grad_norm": 0.26199689507484436, "learning_rate": 5.5591475417412676e-05, "loss": 0.0298, "step": 39290 }, { "grad_norm": 0.3110692501068115, "learning_rate": 5.557093621183451e-05, "loss": 0.0169, "step": 39300 }, { "grad_norm": 0.19582730531692505, "learning_rate": 5.55503960543357e-05, "loss": 0.016, "step": 39310 }, { "grad_norm": 0.2919884920120239, "learning_rate": 5.552985494842601e-05, "loss": 0.0151, "step": 39320 }, { "grad_norm": 0.16759729385375977, "learning_rate": 5.550931289761534e-05, "loss": 0.0118, "step": 39330 }, { "grad_norm": 0.26033779978752136, "learning_rate": 5.548876990541378e-05, "loss": 0.0151, "step": 39340 }, { "grad_norm": 0.41967371106147766, "learning_rate": 5.5468225975331564e-05, "loss": 0.0238, "step": 39350 }, { "grad_norm": 0.33286380767822266, "learning_rate": 5.544768111087909e-05, "loss": 0.0168, "step": 39360 }, { "grad_norm": 0.2562794089317322, "learning_rate": 5.54271353155669e-05, "loss": 0.0172, "step": 39370 }, { "grad_norm": 0.23474550247192383, "learning_rate": 5.5406588592905715e-05, "loss": 0.0161, "step": 39380 }, { "grad_norm": 0.2800871431827545, "learning_rate": 5.5386040946406416e-05, "loss": 0.0174, "step": 39390 }, { "grad_norm": 0.3675166666507721, "learning_rate": 5.536549237958004e-05, "loss": 0.0236, "step": 39400 }, { "grad_norm": 0.19452114403247833, "learning_rate": 5.5344942895937744e-05, "loss": 0.023, "step": 39410 }, { "grad_norm": 0.22976045310497284, "learning_rate": 5.5324392498990904e-05, "loss": 0.0134, "step": 39420 }, { "grad_norm": 0.34780353307724, "learning_rate": 5.5303841192251e-05, "loss": 0.0166, "step": 39430 }, { "grad_norm": 0.25062838196754456, "learning_rate": 5.52832889792297e-05, "loss": 0.0177, "step": 39440 }, { "grad_norm": 0.1839950680732727, "learning_rate": 5.526273586343881e-05, "loss": 0.0124, "step": 39450 }, { "grad_norm": 0.21436311304569244, "learning_rate": 5.5242181848390306e-05, "loss": 0.0121, "step": 39460 }, { "grad_norm": 0.22847899794578552, "learning_rate": 5.5221626937596285e-05, "loss": 0.0128, "step": 39470 }, { "grad_norm": 0.20952364802360535, "learning_rate": 5.520107113456903e-05, "loss": 0.0203, "step": 39480 }, { "grad_norm": 0.3817097544670105, "learning_rate": 5.5180514442820974e-05, "loss": 0.0214, "step": 39490 }, { "grad_norm": 0.18396472930908203, "learning_rate": 5.515995686586469e-05, "loss": 0.0123, "step": 39500 }, { "grad_norm": 0.23627188801765442, "learning_rate": 5.5139398407212916e-05, "loss": 0.013, "step": 39510 }, { "grad_norm": 0.2626529932022095, "learning_rate": 5.511883907037849e-05, "loss": 0.0171, "step": 39520 }, { "grad_norm": 0.2754686772823334, "learning_rate": 5.509827885887449e-05, "loss": 0.0152, "step": 39530 }, { "grad_norm": 0.21912412345409393, "learning_rate": 5.507771777621406e-05, "loss": 0.0153, "step": 39540 }, { "grad_norm": 0.23583540320396423, "learning_rate": 5.505715582591052e-05, "loss": 0.0242, "step": 39550 }, { "grad_norm": 0.38808709383010864, "learning_rate": 5.50365930114774e-05, "loss": 0.0185, "step": 39560 }, { "grad_norm": 0.1630745828151703, "learning_rate": 5.5016029336428255e-05, "loss": 0.0172, "step": 39570 }, { "grad_norm": 0.3363989293575287, "learning_rate": 5.49954648042769e-05, "loss": 0.0164, "step": 39580 }, { "grad_norm": 0.2645321786403656, "learning_rate": 5.4974899418537226e-05, "loss": 0.0183, "step": 39590 }, { "grad_norm": 0.2317321002483368, "learning_rate": 5.4954333182723316e-05, "loss": 0.015, "step": 39600 }, { "grad_norm": 0.2537510097026825, "learning_rate": 5.493376610034937e-05, "loss": 0.015, "step": 39610 }, { "grad_norm": 0.22908183932304382, "learning_rate": 5.4913198174929735e-05, "loss": 0.0165, "step": 39620 }, { "grad_norm": 0.17430317401885986, "learning_rate": 5.489262940997889e-05, "loss": 0.0199, "step": 39630 }, { "grad_norm": 0.12768948078155518, "learning_rate": 5.487205980901151e-05, "loss": 0.0189, "step": 39640 }, { "grad_norm": 0.24923577904701233, "learning_rate": 5.485148937554234e-05, "loss": 0.0363, "step": 39650 }, { "grad_norm": 0.2608599364757538, "learning_rate": 5.483091811308635e-05, "loss": 0.0131, "step": 39660 }, { "grad_norm": 0.2533006966114044, "learning_rate": 5.4810346025158564e-05, "loss": 0.016, "step": 39670 }, { "grad_norm": 0.23687292635440826, "learning_rate": 5.478977311527421e-05, "loss": 0.0147, "step": 39680 }, { "grad_norm": 0.16069304943084717, "learning_rate": 5.476919938694863e-05, "loss": 0.0136, "step": 39690 }, { "grad_norm": 0.22156435251235962, "learning_rate": 5.474862484369733e-05, "loss": 0.0102, "step": 39700 }, { "grad_norm": 0.26348263025283813, "learning_rate": 5.472804948903589e-05, "loss": 0.0169, "step": 39710 }, { "grad_norm": 0.24263930320739746, "learning_rate": 5.470747332648013e-05, "loss": 0.0099, "step": 39720 }, { "grad_norm": 0.23808135092258453, "learning_rate": 5.468689635954591e-05, "loss": 0.0287, "step": 39730 }, { "grad_norm": 0.15807540714740753, "learning_rate": 5.46663185917493e-05, "loss": 0.0168, "step": 39740 }, { "grad_norm": 0.32703566551208496, "learning_rate": 5.464574002660645e-05, "loss": 0.0178, "step": 39750 }, { "grad_norm": 0.16176749765872955, "learning_rate": 5.4625160667633724e-05, "loss": 0.0154, "step": 39760 }, { "grad_norm": 0.2057710736989975, "learning_rate": 5.4604580518347505e-05, "loss": 0.0189, "step": 39770 }, { "grad_norm": 0.22278869152069092, "learning_rate": 5.458399958226442e-05, "loss": 0.0188, "step": 39780 }, { "grad_norm": 0.31496870517730713, "learning_rate": 5.456341786290119e-05, "loss": 0.0157, "step": 39790 }, { "grad_norm": 0.332668662071228, "learning_rate": 5.454283536377465e-05, "loss": 0.0183, "step": 39800 }, { "grad_norm": 0.1694195717573166, "learning_rate": 5.452225208840179e-05, "loss": 0.015, "step": 39810 }, { "grad_norm": 0.21883566677570343, "learning_rate": 5.450166804029975e-05, "loss": 0.0192, "step": 39820 }, { "grad_norm": 0.32711225748062134, "learning_rate": 5.448108322298574e-05, "loss": 0.014, "step": 39830 }, { "grad_norm": 0.2058287411928177, "learning_rate": 5.446049763997717e-05, "loss": 0.0131, "step": 39840 }, { "grad_norm": 0.22535523772239685, "learning_rate": 5.4439911294791546e-05, "loss": 0.0152, "step": 39850 }, { "grad_norm": 0.2536124587059021, "learning_rate": 5.441932419094652e-05, "loss": 0.0116, "step": 39860 }, { "grad_norm": 0.20057381689548492, "learning_rate": 5.439873633195985e-05, "loss": 0.0137, "step": 39870 }, { "grad_norm": 0.20593158900737762, "learning_rate": 5.437814772134947e-05, "loss": 0.0192, "step": 39880 }, { "grad_norm": 0.23568107187747955, "learning_rate": 5.4357558362633366e-05, "loss": 0.0184, "step": 39890 }, { "grad_norm": 0.1907488852739334, "learning_rate": 5.4336968259329726e-05, "loss": 0.0121, "step": 39900 }, { "grad_norm": 0.32653120160102844, "learning_rate": 5.431637741495681e-05, "loss": 0.0171, "step": 39910 }, { "grad_norm": 0.2529616355895996, "learning_rate": 5.429578583303307e-05, "loss": 0.0202, "step": 39920 }, { "grad_norm": 0.29426926374435425, "learning_rate": 5.427519351707701e-05, "loss": 0.0209, "step": 39930 }, { "grad_norm": 0.2014048546552658, "learning_rate": 5.42546004706073e-05, "loss": 0.0152, "step": 39940 }, { "grad_norm": 0.23047791421413422, "learning_rate": 5.4234006697142735e-05, "loss": 0.0135, "step": 39950 }, { "grad_norm": 0.21113355457782745, "learning_rate": 5.421341220020224e-05, "loss": 0.017, "step": 39960 }, { "grad_norm": 0.19454048573970795, "learning_rate": 5.419281698330482e-05, "loss": 0.0212, "step": 39970 }, { "grad_norm": 0.17272046208381653, "learning_rate": 5.4172221049969665e-05, "loss": 0.0172, "step": 39980 }, { "grad_norm": 0.2111842930316925, "learning_rate": 5.415162440371604e-05, "loss": 0.0122, "step": 39990 }, { "grad_norm": 0.22576290369033813, "learning_rate": 5.413102704806334e-05, "loss": 0.0238, "step": 40000 }, { "grad_norm": 0.28139564394950867, "learning_rate": 5.41104289865311e-05, "loss": 0.0181, "step": 40010 }, { "grad_norm": 0.23316654562950134, "learning_rate": 5.408983022263898e-05, "loss": 0.0175, "step": 40020 }, { "grad_norm": 0.3360360264778137, "learning_rate": 5.406923075990673e-05, "loss": 0.0224, "step": 40030 }, { "grad_norm": 0.25898048281669617, "learning_rate": 5.404863060185423e-05, "loss": 0.0164, "step": 40040 }, { "grad_norm": 0.28135499358177185, "learning_rate": 5.40280297520015e-05, "loss": 0.0185, "step": 40050 }, { "grad_norm": 0.25101587176322937, "learning_rate": 5.400742821386865e-05, "loss": 0.0183, "step": 40060 }, { "grad_norm": 0.17259402573108673, "learning_rate": 5.398682599097592e-05, "loss": 0.0171, "step": 40070 }, { "grad_norm": 0.15974010527133942, "learning_rate": 5.396622308684367e-05, "loss": 0.0142, "step": 40080 }, { "grad_norm": 0.3635077476501465, "learning_rate": 5.394561950499236e-05, "loss": 0.0206, "step": 40090 }, { "grad_norm": 0.13094107806682587, "learning_rate": 5.39250152489426e-05, "loss": 0.0154, "step": 40100 }, { "grad_norm": 0.2983614206314087, "learning_rate": 5.390441032221507e-05, "loss": 0.0138, "step": 40110 }, { "grad_norm": 0.2141474485397339, "learning_rate": 5.388380472833062e-05, "loss": 0.0154, "step": 40120 }, { "grad_norm": 0.16161063313484192, "learning_rate": 5.386319847081016e-05, "loss": 0.0176, "step": 40130 }, { "grad_norm": 0.4112052619457245, "learning_rate": 5.384259155317473e-05, "loss": 0.0179, "step": 40140 }, { "grad_norm": 0.35690081119537354, "learning_rate": 5.38219839789455e-05, "loss": 0.0177, "step": 40150 }, { "grad_norm": 0.18252861499786377, "learning_rate": 5.380137575164376e-05, "loss": 0.0235, "step": 40160 }, { "grad_norm": 0.3301919996738434, "learning_rate": 5.378076687479085e-05, "loss": 0.0158, "step": 40170 }, { "grad_norm": 0.18434743583202362, "learning_rate": 5.3760157351908305e-05, "loss": 0.015, "step": 40180 }, { "grad_norm": 0.1376824826002121, "learning_rate": 5.373954718651768e-05, "loss": 0.0164, "step": 40190 }, { "grad_norm": 0.19955523312091827, "learning_rate": 5.371893638214074e-05, "loss": 0.0132, "step": 40200 }, { "grad_norm": 0.20914845168590546, "learning_rate": 5.369832494229927e-05, "loss": 0.0138, "step": 40210 }, { "grad_norm": 0.30663368105888367, "learning_rate": 5.367771287051524e-05, "loss": 0.0215, "step": 40220 }, { "grad_norm": 0.2041199952363968, "learning_rate": 5.365710017031065e-05, "loss": 0.016, "step": 40230 }, { "grad_norm": 0.22978542745113373, "learning_rate": 5.3636486845207654e-05, "loss": 0.0164, "step": 40240 }, { "grad_norm": 0.2016407549381256, "learning_rate": 5.3615872898728526e-05, "loss": 0.0124, "step": 40250 }, { "grad_norm": 0.3461759388446808, "learning_rate": 5.3595258334395614e-05, "loss": 0.0186, "step": 40260 }, { "grad_norm": 0.25690504908561707, "learning_rate": 5.3574643155731396e-05, "loss": 0.0204, "step": 40270 }, { "grad_norm": 0.18031355738639832, "learning_rate": 5.3554027366258405e-05, "loss": 0.0152, "step": 40280 }, { "grad_norm": 0.20975136756896973, "learning_rate": 5.3533410969499355e-05, "loss": 0.016, "step": 40290 }, { "grad_norm": 0.2078825980424881, "learning_rate": 5.3512793968977e-05, "loss": 0.0193, "step": 40300 }, { "grad_norm": 0.35431933403015137, "learning_rate": 5.349217636821423e-05, "loss": 0.0196, "step": 40310 }, { "grad_norm": 0.19350315630435944, "learning_rate": 5.347155817073405e-05, "loss": 0.0114, "step": 40320 }, { "grad_norm": 0.23985512554645538, "learning_rate": 5.34509393800595e-05, "loss": 0.0153, "step": 40330 }, { "grad_norm": 0.29753538966178894, "learning_rate": 5.343031999971381e-05, "loss": 0.0159, "step": 40340 }, { "grad_norm": 0.19219154119491577, "learning_rate": 5.340970003322026e-05, "loss": 0.013, "step": 40350 }, { "grad_norm": 0.523556113243103, "learning_rate": 5.33890794841022e-05, "loss": 0.0179, "step": 40360 }, { "grad_norm": 0.1713045835494995, "learning_rate": 5.336845835588318e-05, "loss": 0.0152, "step": 40370 }, { "grad_norm": 0.2208050787448883, "learning_rate": 5.334783665208674e-05, "loss": 0.0147, "step": 40380 }, { "grad_norm": 0.2851036787033081, "learning_rate": 5.332721437623657e-05, "loss": 0.013, "step": 40390 }, { "grad_norm": 0.28118911385536194, "learning_rate": 5.3306591531856464e-05, "loss": 0.0183, "step": 40400 }, { "grad_norm": 0.12350267171859741, "learning_rate": 5.3285968122470295e-05, "loss": 0.0129, "step": 40410 }, { "grad_norm": 0.48025619983673096, "learning_rate": 5.326534415160207e-05, "loss": 0.0172, "step": 40420 }, { "grad_norm": 0.18346437811851501, "learning_rate": 5.324471962277582e-05, "loss": 0.0249, "step": 40430 }, { "grad_norm": 0.21719132363796234, "learning_rate": 5.3224094539515746e-05, "loss": 0.0143, "step": 40440 }, { "grad_norm": 0.19192416965961456, "learning_rate": 5.3203468905346076e-05, "loss": 0.0126, "step": 40450 }, { "grad_norm": 0.14917120337486267, "learning_rate": 5.3182842723791195e-05, "loss": 0.0173, "step": 40460 }, { "grad_norm": 0.15753614902496338, "learning_rate": 5.316221599837554e-05, "loss": 0.0136, "step": 40470 }, { "grad_norm": 0.34221696853637695, "learning_rate": 5.314158873262366e-05, "loss": 0.0204, "step": 40480 }, { "grad_norm": 0.14592570066452026, "learning_rate": 5.312096093006018e-05, "loss": 0.0136, "step": 40490 }, { "grad_norm": 0.23944951593875885, "learning_rate": 5.3100332594209846e-05, "loss": 0.0182, "step": 40500 }, { "grad_norm": 0.23583605885505676, "learning_rate": 5.3079703728597454e-05, "loss": 0.0156, "step": 40510 }, { "grad_norm": 0.3233206868171692, "learning_rate": 5.305907433674794e-05, "loss": 0.0149, "step": 40520 }, { "grad_norm": 0.3537280857563019, "learning_rate": 5.3038444422186284e-05, "loss": 0.0125, "step": 40530 }, { "grad_norm": 0.2317022979259491, "learning_rate": 5.301781398843757e-05, "loss": 0.0141, "step": 40540 }, { "grad_norm": 0.12394317239522934, "learning_rate": 5.299718303902699e-05, "loss": 0.012, "step": 40550 }, { "grad_norm": 0.21713998913764954, "learning_rate": 5.297655157747979e-05, "loss": 0.0134, "step": 40560 }, { "grad_norm": 0.2856362760066986, "learning_rate": 5.295591960732136e-05, "loss": 0.015, "step": 40570 }, { "grad_norm": 0.2068920135498047, "learning_rate": 5.293528713207708e-05, "loss": 0.0173, "step": 40580 }, { "grad_norm": 0.2429693043231964, "learning_rate": 5.291465415527253e-05, "loss": 0.0186, "step": 40590 }, { "grad_norm": 0.1460663229227066, "learning_rate": 5.2894020680433296e-05, "loss": 0.0121, "step": 40600 }, { "grad_norm": 0.2118416130542755, "learning_rate": 5.287338671108507e-05, "loss": 0.0169, "step": 40610 }, { "grad_norm": 0.31930896639823914, "learning_rate": 5.285275225075367e-05, "loss": 0.016, "step": 40620 }, { "grad_norm": 0.20770618319511414, "learning_rate": 5.283211730296492e-05, "loss": 0.0147, "step": 40630 }, { "grad_norm": 0.1614687591791153, "learning_rate": 5.281148187124477e-05, "loss": 0.0147, "step": 40640 }, { "grad_norm": 0.27003994584083557, "learning_rate": 5.279084595911927e-05, "loss": 0.0153, "step": 40650 }, { "grad_norm": 0.33937868475914, "learning_rate": 5.277020957011453e-05, "loss": 0.0196, "step": 40660 }, { "grad_norm": 0.22932209074497223, "learning_rate": 5.274957270775673e-05, "loss": 0.0197, "step": 40670 }, { "grad_norm": 0.18652568757534027, "learning_rate": 5.2728935375572164e-05, "loss": 0.0131, "step": 40680 }, { "grad_norm": 0.15339897572994232, "learning_rate": 5.2708297577087176e-05, "loss": 0.0127, "step": 40690 }, { "grad_norm": 0.19829674065113068, "learning_rate": 5.2687659315828196e-05, "loss": 0.0106, "step": 40700 }, { "grad_norm": 0.23877935111522675, "learning_rate": 5.266702059532175e-05, "loss": 0.0155, "step": 40710 }, { "grad_norm": 0.27495527267456055, "learning_rate": 5.264638141909444e-05, "loss": 0.018, "step": 40720 }, { "grad_norm": 0.22497254610061646, "learning_rate": 5.2625741790672925e-05, "loss": 0.0171, "step": 40730 }, { "grad_norm": 0.24621696770191193, "learning_rate": 5.2605101713583935e-05, "loss": 0.0125, "step": 40740 }, { "grad_norm": 0.2630653977394104, "learning_rate": 5.2584461191354315e-05, "loss": 0.0229, "step": 40750 }, { "grad_norm": 0.1547364592552185, "learning_rate": 5.2563820227510964e-05, "loss": 0.0128, "step": 40760 }, { "grad_norm": 0.21003347635269165, "learning_rate": 5.2543178825580864e-05, "loss": 0.0126, "step": 40770 }, { "grad_norm": 0.3085794150829315, "learning_rate": 5.2522536989091054e-05, "loss": 0.0169, "step": 40780 }, { "grad_norm": 0.25406116247177124, "learning_rate": 5.2501894721568655e-05, "loss": 0.018, "step": 40790 }, { "grad_norm": 0.2518855631351471, "learning_rate": 5.248125202654089e-05, "loss": 0.0155, "step": 40800 }, { "grad_norm": 0.32310545444488525, "learning_rate": 5.246060890753501e-05, "loss": 0.0215, "step": 40810 }, { "grad_norm": 0.16474103927612305, "learning_rate": 5.243996536807837e-05, "loss": 0.0198, "step": 40820 }, { "grad_norm": 0.22698244452476501, "learning_rate": 5.2419321411698384e-05, "loss": 0.0146, "step": 40830 }, { "grad_norm": 0.48051807284355164, "learning_rate": 5.239867704192253e-05, "loss": 0.017, "step": 40840 }, { "grad_norm": 0.1396094709634781, "learning_rate": 5.237803226227838e-05, "loss": 0.0119, "step": 40850 }, { "grad_norm": 0.14240078628063202, "learning_rate": 5.235738707629354e-05, "loss": 0.0151, "step": 40860 }, { "grad_norm": 0.21031054854393005, "learning_rate": 5.233674148749575e-05, "loss": 0.0154, "step": 40870 }, { "grad_norm": 0.24221092462539673, "learning_rate": 5.231609549941272e-05, "loss": 0.0197, "step": 40880 }, { "grad_norm": 0.1743067055940628, "learning_rate": 5.229544911557233e-05, "loss": 0.0202, "step": 40890 }, { "grad_norm": 0.1787731945514679, "learning_rate": 5.227480233950246e-05, "loss": 0.0179, "step": 40900 }, { "grad_norm": 0.2614964246749878, "learning_rate": 5.22541551747311e-05, "loss": 0.0184, "step": 40910 }, { "grad_norm": 0.28397971391677856, "learning_rate": 5.223350762478626e-05, "loss": 0.0122, "step": 40920 }, { "grad_norm": 0.19037507474422455, "learning_rate": 5.221285969319606e-05, "loss": 0.0158, "step": 40930 }, { "grad_norm": 0.36536288261413574, "learning_rate": 5.219221138348865e-05, "loss": 0.0149, "step": 40940 }, { "grad_norm": 0.20555627346038818, "learning_rate": 5.217156269919228e-05, "loss": 0.0152, "step": 40950 }, { "grad_norm": 0.17155134677886963, "learning_rate": 5.215091364383523e-05, "loss": 0.012, "step": 40960 }, { "grad_norm": 0.30636870861053467, "learning_rate": 5.213026422094588e-05, "loss": 0.0162, "step": 40970 }, { "grad_norm": 0.18256136775016785, "learning_rate": 5.210961443405262e-05, "loss": 0.0114, "step": 40980 }, { "grad_norm": 0.16789284348487854, "learning_rate": 5.208896428668396e-05, "loss": 0.012, "step": 40990 }, { "grad_norm": 0.2191103994846344, "learning_rate": 5.206831378236845e-05, "loss": 0.0163, "step": 41000 }, { "grad_norm": 0.1537189930677414, "learning_rate": 5.2047662924634666e-05, "loss": 0.0123, "step": 41010 }, { "grad_norm": 0.26200205087661743, "learning_rate": 5.20270117170113e-05, "loss": 0.0104, "step": 41020 }, { "grad_norm": 0.3450767993927002, "learning_rate": 5.200636016302707e-05, "loss": 0.0215, "step": 41030 }, { "grad_norm": 0.3568514287471771, "learning_rate": 5.1985708266210754e-05, "loss": 0.0204, "step": 41040 }, { "grad_norm": 0.15589575469493866, "learning_rate": 5.1965056030091206e-05, "loss": 0.0191, "step": 41050 }, { "grad_norm": 0.20447427034378052, "learning_rate": 5.194440345819732e-05, "loss": 0.0139, "step": 41060 }, { "grad_norm": 0.28027912974357605, "learning_rate": 5.1923750554058084e-05, "loss": 0.0209, "step": 41070 }, { "grad_norm": 0.1859513372182846, "learning_rate": 5.1903097321202476e-05, "loss": 0.0127, "step": 41080 }, { "grad_norm": 0.21191582083702087, "learning_rate": 5.18824437631596e-05, "loss": 0.0225, "step": 41090 }, { "grad_norm": 0.22049148380756378, "learning_rate": 5.186178988345856e-05, "loss": 0.021, "step": 41100 }, { "grad_norm": 0.29802075028419495, "learning_rate": 5.184113568562855e-05, "loss": 0.0202, "step": 41110 }, { "grad_norm": 0.27118220925331116, "learning_rate": 5.18204811731988e-05, "loss": 0.0205, "step": 41120 }, { "grad_norm": 0.3101358413696289, "learning_rate": 5.179982634969862e-05, "loss": 0.0179, "step": 41130 }, { "grad_norm": 0.26635628938674927, "learning_rate": 5.177917121865734e-05, "loss": 0.0206, "step": 41140 }, { "grad_norm": 0.21514946222305298, "learning_rate": 5.1758515783604346e-05, "loss": 0.0126, "step": 41150 }, { "grad_norm": 0.22193773090839386, "learning_rate": 5.173786004806911e-05, "loss": 0.0206, "step": 41160 }, { "grad_norm": 0.4069126546382904, "learning_rate": 5.1717204015581135e-05, "loss": 0.0159, "step": 41170 }, { "grad_norm": 0.1843067854642868, "learning_rate": 5.169654768966994e-05, "loss": 0.0146, "step": 41180 }, { "grad_norm": 0.3211350739002228, "learning_rate": 5.1675891073865156e-05, "loss": 0.0189, "step": 41190 }, { "grad_norm": 0.17127728462219238, "learning_rate": 5.1655234171696424e-05, "loss": 0.0109, "step": 41200 }, { "grad_norm": 0.21570108830928802, "learning_rate": 5.163457698669343e-05, "loss": 0.0119, "step": 41210 }, { "grad_norm": 0.2017252892255783, "learning_rate": 5.1613919522385936e-05, "loss": 0.0131, "step": 41220 }, { "grad_norm": 0.31073203682899475, "learning_rate": 5.1593261782303746e-05, "loss": 0.0203, "step": 41230 }, { "grad_norm": 0.22550813853740692, "learning_rate": 5.157260376997669e-05, "loss": 0.0152, "step": 41240 }, { "grad_norm": 0.2438179850578308, "learning_rate": 5.155194548893466e-05, "loss": 0.0192, "step": 41250 }, { "grad_norm": 0.2607491910457611, "learning_rate": 5.153128694270759e-05, "loss": 0.0233, "step": 41260 }, { "grad_norm": 0.3057344853878021, "learning_rate": 5.151062813482548e-05, "loss": 0.0212, "step": 41270 }, { "grad_norm": 0.26549193263053894, "learning_rate": 5.148996906881832e-05, "loss": 0.014, "step": 41280 }, { "grad_norm": 0.34326061606407166, "learning_rate": 5.1469309748216196e-05, "loss": 0.0172, "step": 41290 }, { "grad_norm": 0.2504939138889313, "learning_rate": 5.144865017654923e-05, "loss": 0.0119, "step": 41300 }, { "grad_norm": 0.2737393081188202, "learning_rate": 5.1427990357347556e-05, "loss": 0.0169, "step": 41310 }, { "grad_norm": 0.19028988480567932, "learning_rate": 5.140733029414139e-05, "loss": 0.0167, "step": 41320 }, { "grad_norm": 0.26960545778274536, "learning_rate": 5.138666999046098e-05, "loss": 0.021, "step": 41330 }, { "grad_norm": 0.6377607583999634, "learning_rate": 5.136600944983658e-05, "loss": 0.0158, "step": 41340 }, { "grad_norm": 0.2061631828546524, "learning_rate": 5.134534867579853e-05, "loss": 0.014, "step": 41350 }, { "grad_norm": 0.3359750509262085, "learning_rate": 5.1324687671877184e-05, "loss": 0.0158, "step": 41360 }, { "grad_norm": 0.2820887863636017, "learning_rate": 5.130402644160296e-05, "loss": 0.0191, "step": 41370 }, { "grad_norm": 0.2140449434518814, "learning_rate": 5.128336498850628e-05, "loss": 0.0115, "step": 41380 }, { "grad_norm": 0.2677750587463379, "learning_rate": 5.126270331611761e-05, "loss": 0.0165, "step": 41390 }, { "grad_norm": 0.30323025584220886, "learning_rate": 5.124204142796748e-05, "loss": 0.0139, "step": 41400 }, { "grad_norm": 0.25469037890434265, "learning_rate": 5.122137932758644e-05, "loss": 0.0121, "step": 41410 }, { "grad_norm": 0.18385018408298492, "learning_rate": 5.1200717018505074e-05, "loss": 0.0138, "step": 41420 }, { "grad_norm": 0.17966902256011963, "learning_rate": 5.118005450425403e-05, "loss": 0.0147, "step": 41430 }, { "grad_norm": 0.33193206787109375, "learning_rate": 5.115939178836392e-05, "loss": 0.0125, "step": 41440 }, { "grad_norm": 0.2250811904668808, "learning_rate": 5.113872887436547e-05, "loss": 0.0125, "step": 41450 }, { "grad_norm": 0.13759498298168182, "learning_rate": 5.11180657657894e-05, "loss": 0.017, "step": 41460 }, { "grad_norm": 0.3059917688369751, "learning_rate": 5.109740246616649e-05, "loss": 0.0127, "step": 41470 }, { "grad_norm": 0.3620196580886841, "learning_rate": 5.10767389790275e-05, "loss": 0.022, "step": 41480 }, { "grad_norm": 0.2736525237560272, "learning_rate": 5.1056075307903265e-05, "loss": 0.0129, "step": 41490 }, { "grad_norm": 0.27500975131988525, "learning_rate": 5.1035411456324634e-05, "loss": 0.0189, "step": 41500 }, { "grad_norm": 0.20169910788536072, "learning_rate": 5.1014747427822504e-05, "loss": 0.0198, "step": 41510 }, { "grad_norm": 0.34763193130493164, "learning_rate": 5.0994083225927804e-05, "loss": 0.0184, "step": 41520 }, { "grad_norm": 0.21633276343345642, "learning_rate": 5.0973418854171475e-05, "loss": 0.0174, "step": 41530 }, { "grad_norm": 0.22619684040546417, "learning_rate": 5.095275431608447e-05, "loss": 0.014, "step": 41540 }, { "grad_norm": 0.17298033833503723, "learning_rate": 5.093208961519782e-05, "loss": 0.0135, "step": 41550 }, { "grad_norm": 0.49250465631484985, "learning_rate": 5.091142475504255e-05, "loss": 0.0128, "step": 41560 }, { "grad_norm": 0.330530047416687, "learning_rate": 5.089075973914971e-05, "loss": 0.0166, "step": 41570 }, { "grad_norm": 0.26790425181388855, "learning_rate": 5.08700945710504e-05, "loss": 0.021, "step": 41580 }, { "grad_norm": 0.2090110033750534, "learning_rate": 5.0849429254275714e-05, "loss": 0.0139, "step": 41590 }, { "grad_norm": 0.1275169551372528, "learning_rate": 5.0828763792356804e-05, "loss": 0.0133, "step": 41600 }, { "grad_norm": 0.2671986520290375, "learning_rate": 5.080809818882483e-05, "loss": 0.0152, "step": 41610 }, { "grad_norm": 0.24281634390354156, "learning_rate": 5.078743244721097e-05, "loss": 0.0208, "step": 41620 }, { "grad_norm": 0.24281680583953857, "learning_rate": 5.0766766571046455e-05, "loss": 0.0183, "step": 41630 }, { "grad_norm": 0.17754724621772766, "learning_rate": 5.07461005638625e-05, "loss": 0.0176, "step": 41640 }, { "grad_norm": 0.22817344963550568, "learning_rate": 5.072543442919037e-05, "loss": 0.0135, "step": 41650 }, { "grad_norm": 0.31260037422180176, "learning_rate": 5.070476817056132e-05, "loss": 0.0208, "step": 41660 }, { "grad_norm": 0.3435073792934418, "learning_rate": 5.068410179150668e-05, "loss": 0.0112, "step": 41670 }, { "grad_norm": 0.16629619896411896, "learning_rate": 5.066343529555775e-05, "loss": 0.0188, "step": 41680 }, { "grad_norm": 0.21108755469322205, "learning_rate": 5.064276868624587e-05, "loss": 0.0134, "step": 41690 }, { "grad_norm": 0.4245493412017822, "learning_rate": 5.0622101967102396e-05, "loss": 0.0111, "step": 41700 }, { "grad_norm": 0.2852708399295807, "learning_rate": 5.0601435141658714e-05, "loss": 0.013, "step": 41710 }, { "grad_norm": 0.3087567090988159, "learning_rate": 5.058076821344622e-05, "loss": 0.0163, "step": 41720 }, { "grad_norm": 0.28932058811187744, "learning_rate": 5.056010118599632e-05, "loss": 0.0202, "step": 41730 }, { "grad_norm": 0.24866123497486115, "learning_rate": 5.053943406284044e-05, "loss": 0.0179, "step": 41740 }, { "grad_norm": 0.31131696701049805, "learning_rate": 5.051876684751006e-05, "loss": 0.0182, "step": 41750 }, { "grad_norm": 0.20995360612869263, "learning_rate": 5.0498099543536584e-05, "loss": 0.0151, "step": 41760 }, { "grad_norm": 0.21254661679267883, "learning_rate": 5.047743215445152e-05, "loss": 0.0152, "step": 41770 }, { "grad_norm": 0.21149842441082, "learning_rate": 5.045676468378637e-05, "loss": 0.0168, "step": 41780 }, { "grad_norm": 0.29592394828796387, "learning_rate": 5.0436097135072626e-05, "loss": 0.0183, "step": 41790 }, { "grad_norm": 0.23236849904060364, "learning_rate": 5.041542951184181e-05, "loss": 0.0148, "step": 41800 }, { "grad_norm": 0.16642412543296814, "learning_rate": 5.039476181762545e-05, "loss": 0.015, "step": 41810 }, { "grad_norm": 0.23521366715431213, "learning_rate": 5.037409405595508e-05, "loss": 0.011, "step": 41820 }, { "grad_norm": 0.2711583375930786, "learning_rate": 5.035342623036229e-05, "loss": 0.0144, "step": 41830 }, { "grad_norm": 0.2066992074251175, "learning_rate": 5.033275834437862e-05, "loss": 0.015, "step": 41840 }, { "grad_norm": 0.48490625619888306, "learning_rate": 5.031209040153564e-05, "loss": 0.0122, "step": 41850 }, { "grad_norm": 0.21493297815322876, "learning_rate": 5.0291422405364955e-05, "loss": 0.0233, "step": 41860 }, { "grad_norm": 0.33453860878944397, "learning_rate": 5.0270754359398133e-05, "loss": 0.0163, "step": 41870 }, { "grad_norm": 0.2515479326248169, "learning_rate": 5.025008626716682e-05, "loss": 0.0148, "step": 41880 }, { "grad_norm": 0.18086087703704834, "learning_rate": 5.0229418132202585e-05, "loss": 0.0135, "step": 41890 }, { "grad_norm": 0.21961039304733276, "learning_rate": 5.020874995803707e-05, "loss": 0.0219, "step": 41900 }, { "grad_norm": 0.23343826830387115, "learning_rate": 5.01880817482019e-05, "loss": 0.0198, "step": 41910 }, { "grad_norm": 0.29910406470298767, "learning_rate": 5.01674135062287e-05, "loss": 0.0146, "step": 41920 }, { "grad_norm": 0.3023291528224945, "learning_rate": 5.0146745235649115e-05, "loss": 0.0178, "step": 41930 }, { "grad_norm": 0.23259906470775604, "learning_rate": 5.0126076939994795e-05, "loss": 0.012, "step": 41940 }, { "grad_norm": 0.23206785321235657, "learning_rate": 5.010540862279736e-05, "loss": 0.0128, "step": 41950 }, { "grad_norm": 0.15427657961845398, "learning_rate": 5.008474028758846e-05, "loss": 0.0117, "step": 41960 }, { "grad_norm": 0.19362232089042664, "learning_rate": 5.0064071937899784e-05, "loss": 0.0161, "step": 41970 }, { "grad_norm": 0.6101958751678467, "learning_rate": 5.004340357726296e-05, "loss": 0.0155, "step": 41980 }, { "grad_norm": 0.26373738050460815, "learning_rate": 5.002273520920965e-05, "loss": 0.0126, "step": 41990 }, { "grad_norm": 0.16148504614830017, "learning_rate": 5.000206683727151e-05, "loss": 0.013, "step": 42000 }, { "grad_norm": 0.1852637529373169, "learning_rate": 4.998139846498021e-05, "loss": 0.0118, "step": 42010 }, { "grad_norm": 0.21160291135311127, "learning_rate": 4.9960730095867405e-05, "loss": 0.0169, "step": 42020 }, { "grad_norm": 0.31675606966018677, "learning_rate": 4.9940061733464755e-05, "loss": 0.0176, "step": 42030 }, { "grad_norm": 0.15663760900497437, "learning_rate": 4.991939338130392e-05, "loss": 0.0221, "step": 42040 }, { "grad_norm": 0.36672088503837585, "learning_rate": 4.989872504291653e-05, "loss": 0.0205, "step": 42050 }, { "grad_norm": 0.276825875043869, "learning_rate": 4.9878056721834273e-05, "loss": 0.0185, "step": 42060 }, { "grad_norm": 0.17907243967056274, "learning_rate": 4.98573884215888e-05, "loss": 0.0162, "step": 42070 }, { "grad_norm": 0.21741105616092682, "learning_rate": 4.9836720145711715e-05, "loss": 0.0162, "step": 42080 }, { "grad_norm": 0.18393026292324066, "learning_rate": 4.9816051897734725e-05, "loss": 0.0196, "step": 42090 }, { "grad_norm": 0.2778673470020294, "learning_rate": 4.979538368118941e-05, "loss": 0.0183, "step": 42100 }, { "grad_norm": 0.19499333202838898, "learning_rate": 4.9774715499607446e-05, "loss": 0.0131, "step": 42110 }, { "grad_norm": 0.30698493123054504, "learning_rate": 4.9754047356520416e-05, "loss": 0.0193, "step": 42120 }, { "grad_norm": 0.2626567482948303, "learning_rate": 4.973337925545999e-05, "loss": 0.0179, "step": 42130 }, { "grad_norm": 0.28383180499076843, "learning_rate": 4.9712711199957746e-05, "loss": 0.0112, "step": 42140 }, { "grad_norm": 0.26981303095817566, "learning_rate": 4.969204319354531e-05, "loss": 0.0156, "step": 42150 }, { "grad_norm": 0.5178611874580383, "learning_rate": 4.9671375239754267e-05, "loss": 0.0168, "step": 42160 }, { "grad_norm": 0.13416387140750885, "learning_rate": 4.9650707342116195e-05, "loss": 0.011, "step": 42170 }, { "grad_norm": 0.16855083405971527, "learning_rate": 4.96300395041627e-05, "loss": 0.0142, "step": 42180 }, { "grad_norm": 0.28035956621170044, "learning_rate": 4.960937172942532e-05, "loss": 0.0176, "step": 42190 }, { "grad_norm": 0.1944957673549652, "learning_rate": 4.958870402143565e-05, "loss": 0.0161, "step": 42200 }, { "grad_norm": 0.1487770825624466, "learning_rate": 4.9568036383725186e-05, "loss": 0.018, "step": 42210 }, { "grad_norm": 0.19761785864830017, "learning_rate": 4.9547368819825516e-05, "loss": 0.0119, "step": 42220 }, { "grad_norm": 0.1940612941980362, "learning_rate": 4.952670133326812e-05, "loss": 0.0133, "step": 42230 }, { "grad_norm": 0.19421719014644623, "learning_rate": 4.950603392758453e-05, "loss": 0.0162, "step": 42240 }, { "grad_norm": 0.40135657787323, "learning_rate": 4.948536660630621e-05, "loss": 0.0143, "step": 42250 }, { "grad_norm": 0.3007216155529022, "learning_rate": 4.9464699372964685e-05, "loss": 0.0175, "step": 42260 }, { "grad_norm": 0.2478901594877243, "learning_rate": 4.9444032231091395e-05, "loss": 0.0159, "step": 42270 }, { "grad_norm": 0.20469588041305542, "learning_rate": 4.9423365184217776e-05, "loss": 0.0151, "step": 42280 }, { "grad_norm": 0.2847445011138916, "learning_rate": 4.940269823587529e-05, "loss": 0.0139, "step": 42290 }, { "grad_norm": 0.1650535762310028, "learning_rate": 4.938203138959533e-05, "loss": 0.0117, "step": 42300 }, { "grad_norm": 0.21497465670108795, "learning_rate": 4.9361364648909325e-05, "loss": 0.0133, "step": 42310 }, { "grad_norm": 0.2469409704208374, "learning_rate": 4.9340698017348605e-05, "loss": 0.0128, "step": 42320 }, { "grad_norm": 0.358266144990921, "learning_rate": 4.932003149844458e-05, "loss": 0.0144, "step": 42330 }, { "grad_norm": 0.2951754331588745, "learning_rate": 4.929936509572857e-05, "loss": 0.0189, "step": 42340 }, { "grad_norm": 0.26644864678382874, "learning_rate": 4.927869881273191e-05, "loss": 0.0161, "step": 42350 }, { "grad_norm": 0.19534450769424438, "learning_rate": 4.9258032652985894e-05, "loss": 0.0143, "step": 42360 }, { "grad_norm": 0.2173694521188736, "learning_rate": 4.9237366620021786e-05, "loss": 0.0185, "step": 42370 }, { "grad_norm": 0.29221513867378235, "learning_rate": 4.921670071737089e-05, "loss": 0.0113, "step": 42380 }, { "grad_norm": 0.2432335466146469, "learning_rate": 4.91960349485644e-05, "loss": 0.016, "step": 42390 }, { "grad_norm": 0.2287754863500595, "learning_rate": 4.9175369317133566e-05, "loss": 0.0178, "step": 42400 }, { "grad_norm": 0.1956203430891037, "learning_rate": 4.915470382660953e-05, "loss": 0.0155, "step": 42410 }, { "grad_norm": 0.29402056336402893, "learning_rate": 4.9134038480523524e-05, "loss": 0.0146, "step": 42420 }, { "grad_norm": 0.16506101191043854, "learning_rate": 4.911337328240664e-05, "loss": 0.0179, "step": 42430 }, { "grad_norm": 0.2693036198616028, "learning_rate": 4.909270823579003e-05, "loss": 0.0159, "step": 42440 }, { "grad_norm": 0.24576544761657715, "learning_rate": 4.907204334420476e-05, "loss": 0.0137, "step": 42450 }, { "grad_norm": 0.2370174080133438, "learning_rate": 4.9051378611181884e-05, "loss": 0.0166, "step": 42460 }, { "grad_norm": 0.1532091200351715, "learning_rate": 4.9030714040252486e-05, "loss": 0.0144, "step": 42470 }, { "grad_norm": 0.17788107693195343, "learning_rate": 4.901004963494752e-05, "loss": 0.0129, "step": 42480 }, { "grad_norm": 0.2859058678150177, "learning_rate": 4.898938539879802e-05, "loss": 0.0176, "step": 42490 }, { "grad_norm": 0.2792832553386688, "learning_rate": 4.8968721335334904e-05, "loss": 0.0126, "step": 42500 }, { "grad_norm": 0.22284740209579468, "learning_rate": 4.894805744808912e-05, "loss": 0.014, "step": 42510 }, { "grad_norm": 0.1870003491640091, "learning_rate": 4.8927393740591524e-05, "loss": 0.0155, "step": 42520 }, { "grad_norm": 0.2791217863559723, "learning_rate": 4.890673021637302e-05, "loss": 0.0147, "step": 42530 }, { "grad_norm": 0.26826077699661255, "learning_rate": 4.888606687896442e-05, "loss": 0.0169, "step": 42540 }, { "grad_norm": 0.2946031391620636, "learning_rate": 4.886540373189652e-05, "loss": 0.0176, "step": 42550 }, { "grad_norm": 0.23500177264213562, "learning_rate": 4.88447407787001e-05, "loss": 0.0181, "step": 42560 }, { "grad_norm": 0.2173333764076233, "learning_rate": 4.8824078022905864e-05, "loss": 0.014, "step": 42570 }, { "grad_norm": 0.19037017226219177, "learning_rate": 4.880341546804455e-05, "loss": 0.0101, "step": 42580 }, { "grad_norm": 0.2010810673236847, "learning_rate": 4.878275311764679e-05, "loss": 0.0153, "step": 42590 }, { "grad_norm": 0.29579561948776245, "learning_rate": 4.8762090975243254e-05, "loss": 0.013, "step": 42600 }, { "grad_norm": 0.17560485005378723, "learning_rate": 4.874142904436448e-05, "loss": 0.0116, "step": 42610 }, { "grad_norm": 0.21189923584461212, "learning_rate": 4.8720767328541094e-05, "loss": 0.0239, "step": 42620 }, { "grad_norm": 0.2956711947917938, "learning_rate": 4.870010583130358e-05, "loss": 0.0116, "step": 42630 }, { "grad_norm": 0.15139663219451904, "learning_rate": 4.867944455618241e-05, "loss": 0.0161, "step": 42640 }, { "grad_norm": 0.24251003563404083, "learning_rate": 4.865878350670807e-05, "loss": 0.0158, "step": 42650 }, { "grad_norm": 0.3504469692707062, "learning_rate": 4.8638122686410914e-05, "loss": 0.0137, "step": 42660 }, { "grad_norm": 0.303101509809494, "learning_rate": 4.861746209882137e-05, "loss": 0.0116, "step": 42670 }, { "grad_norm": 0.2301385998725891, "learning_rate": 4.859680174746972e-05, "loss": 0.0188, "step": 42680 }, { "grad_norm": 0.21315652132034302, "learning_rate": 4.857614163588629e-05, "loss": 0.0184, "step": 42690 }, { "grad_norm": 0.1928529590368271, "learning_rate": 4.855548176760131e-05, "loss": 0.0128, "step": 42700 }, { "grad_norm": 0.2051728218793869, "learning_rate": 4.8534822146144986e-05, "loss": 0.0152, "step": 42710 }, { "grad_norm": 0.22035309672355652, "learning_rate": 4.851416277504749e-05, "loss": 0.022, "step": 42720 }, { "grad_norm": 0.2154427468776703, "learning_rate": 4.8493503657838923e-05, "loss": 0.012, "step": 42730 }, { "grad_norm": 0.14649824798107147, "learning_rate": 4.84728447980494e-05, "loss": 0.0181, "step": 42740 }, { "grad_norm": 0.7276272773742676, "learning_rate": 4.8452186199208914e-05, "loss": 0.0165, "step": 42750 }, { "grad_norm": 0.3627438545227051, "learning_rate": 4.843152786484749e-05, "loss": 0.0164, "step": 42760 }, { "grad_norm": 0.13018247485160828, "learning_rate": 4.8410869798495035e-05, "loss": 0.0146, "step": 42770 }, { "grad_norm": 0.16436690092086792, "learning_rate": 4.8390212003681486e-05, "loss": 0.0139, "step": 42780 }, { "grad_norm": 0.3201057016849518, "learning_rate": 4.836955448393667e-05, "loss": 0.0201, "step": 42790 }, { "grad_norm": 0.2550129294395447, "learning_rate": 4.8348897242790406e-05, "loss": 0.0148, "step": 42800 }, { "grad_norm": 0.1528712511062622, "learning_rate": 4.832824028377243e-05, "loss": 0.0167, "step": 42810 }, { "grad_norm": 0.29285919666290283, "learning_rate": 4.830758361041249e-05, "loss": 0.0185, "step": 42820 }, { "grad_norm": 0.3014391362667084, "learning_rate": 4.828692722624022e-05, "loss": 0.0164, "step": 42830 }, { "grad_norm": 0.18032103776931763, "learning_rate": 4.826627113478522e-05, "loss": 0.0267, "step": 42840 }, { "grad_norm": 0.4220387637615204, "learning_rate": 4.824561533957708e-05, "loss": 0.0143, "step": 42850 }, { "grad_norm": 0.25856223702430725, "learning_rate": 4.822495984414527e-05, "loss": 0.0168, "step": 42860 }, { "grad_norm": 0.19848233461380005, "learning_rate": 4.8204304652019304e-05, "loss": 0.012, "step": 42870 }, { "grad_norm": 0.4397195875644684, "learning_rate": 4.8183649766728537e-05, "loss": 0.0164, "step": 42880 }, { "grad_norm": 0.3521536886692047, "learning_rate": 4.816299519180237e-05, "loss": 0.0156, "step": 42890 }, { "grad_norm": 0.13453279435634613, "learning_rate": 4.8142340930770066e-05, "loss": 0.0121, "step": 42900 }, { "grad_norm": 0.276167094707489, "learning_rate": 4.81216869871609e-05, "loss": 0.0123, "step": 42910 }, { "grad_norm": 0.2016342580318451, "learning_rate": 4.8101033364504064e-05, "loss": 0.0152, "step": 42920 }, { "grad_norm": 0.2461564689874649, "learning_rate": 4.808038006632866e-05, "loss": 0.0185, "step": 42930 }, { "grad_norm": 0.3112727403640747, "learning_rate": 4.805972709616383e-05, "loss": 0.0139, "step": 42940 }, { "grad_norm": 0.16456003487110138, "learning_rate": 4.8039074457538556e-05, "loss": 0.0101, "step": 42950 }, { "grad_norm": 0.16363884508609772, "learning_rate": 4.801842215398184e-05, "loss": 0.016, "step": 42960 }, { "grad_norm": 0.27560555934906006, "learning_rate": 4.799777018902256e-05, "loss": 0.0171, "step": 42970 }, { "grad_norm": 0.23224827647209167, "learning_rate": 4.797711856618961e-05, "loss": 0.0176, "step": 42980 }, { "grad_norm": 0.16502107679843903, "learning_rate": 4.7956467289011765e-05, "loss": 0.0131, "step": 42990 }, { "grad_norm": 0.2628331482410431, "learning_rate": 4.793581636101778e-05, "loss": 0.0136, "step": 43000 }, { "grad_norm": 0.3356396555900574, "learning_rate": 4.7915165785736326e-05, "loss": 0.0132, "step": 43010 }, { "grad_norm": 0.30747517943382263, "learning_rate": 4.789451556669599e-05, "loss": 0.0136, "step": 43020 }, { "grad_norm": 0.2414349764585495, "learning_rate": 4.7873865707425384e-05, "loss": 0.0195, "step": 43030 }, { "grad_norm": 0.20770230889320374, "learning_rate": 4.7853216211452975e-05, "loss": 0.014, "step": 43040 }, { "grad_norm": 0.3249601423740387, "learning_rate": 4.78325670823072e-05, "loss": 0.0161, "step": 43050 }, { "grad_norm": 0.19901567697525024, "learning_rate": 4.781191832351641e-05, "loss": 0.016, "step": 43060 }, { "grad_norm": 0.3761148154735565, "learning_rate": 4.7791269938608965e-05, "loss": 0.0132, "step": 43070 }, { "grad_norm": 0.22415146231651306, "learning_rate": 4.777062193111305e-05, "loss": 0.0127, "step": 43080 }, { "grad_norm": 0.23141881823539734, "learning_rate": 4.77499743045569e-05, "loss": 0.0119, "step": 43090 }, { "grad_norm": 0.2061791718006134, "learning_rate": 4.7729327062468596e-05, "loss": 0.0212, "step": 43100 }, { "grad_norm": 0.33445438742637634, "learning_rate": 4.770868020837619e-05, "loss": 0.0191, "step": 43110 }, { "grad_norm": 0.2424192577600479, "learning_rate": 4.768803374580768e-05, "loss": 0.0102, "step": 43120 }, { "grad_norm": 0.26909512281417847, "learning_rate": 4.766738767829094e-05, "loss": 0.015, "step": 43130 }, { "grad_norm": 0.1880640834569931, "learning_rate": 4.764674200935388e-05, "loss": 0.0123, "step": 43140 }, { "grad_norm": 0.1298830360174179, "learning_rate": 4.762609674252424e-05, "loss": 0.0118, "step": 43150 }, { "grad_norm": 0.15398553013801575, "learning_rate": 4.760545188132974e-05, "loss": 0.0244, "step": 43160 }, { "grad_norm": 0.22846364974975586, "learning_rate": 4.758480742929801e-05, "loss": 0.0161, "step": 43170 }, { "grad_norm": 0.2921774983406067, "learning_rate": 4.756416338995664e-05, "loss": 0.0176, "step": 43180 }, { "grad_norm": 0.33859971165657043, "learning_rate": 4.7543519766833145e-05, "loss": 0.0145, "step": 43190 }, { "grad_norm": 0.21017898619174957, "learning_rate": 4.752287656345492e-05, "loss": 0.0172, "step": 43200 }, { "grad_norm": 0.2343364655971527, "learning_rate": 4.7502233783349355e-05, "loss": 0.0145, "step": 43210 }, { "grad_norm": 0.1862674355506897, "learning_rate": 4.7481591430043694e-05, "loss": 0.0125, "step": 43220 }, { "grad_norm": 0.16948352754116058, "learning_rate": 4.746094950706521e-05, "loss": 0.0142, "step": 43230 }, { "grad_norm": 0.21485553681850433, "learning_rate": 4.744030801794099e-05, "loss": 0.0175, "step": 43240 }, { "grad_norm": 0.23704387247562408, "learning_rate": 4.741966696619813e-05, "loss": 0.0193, "step": 43250 }, { "grad_norm": 0.15107658505439758, "learning_rate": 4.739902635536361e-05, "loss": 0.0165, "step": 43260 }, { "grad_norm": 0.24514390528202057, "learning_rate": 4.737838618896436e-05, "loss": 0.0125, "step": 43270 }, { "grad_norm": 0.1994747519493103, "learning_rate": 4.7357746470527203e-05, "loss": 0.0158, "step": 43280 }, { "grad_norm": 0.3225542902946472, "learning_rate": 4.73371072035789e-05, "loss": 0.0184, "step": 43290 }, { "grad_norm": 0.34286239743232727, "learning_rate": 4.731646839164616e-05, "loss": 0.0192, "step": 43300 }, { "grad_norm": 0.25676944851875305, "learning_rate": 4.729583003825556e-05, "loss": 0.0168, "step": 43310 }, { "grad_norm": 0.15220151841640472, "learning_rate": 4.727519214693367e-05, "loss": 0.0172, "step": 43320 }, { "grad_norm": 0.7086071968078613, "learning_rate": 4.725455472120689e-05, "loss": 0.0196, "step": 43330 }, { "grad_norm": 0.15334513783454895, "learning_rate": 4.723391776460164e-05, "loss": 0.0165, "step": 43340 }, { "grad_norm": 0.5681425929069519, "learning_rate": 4.7213281280644186e-05, "loss": 0.0189, "step": 43350 }, { "grad_norm": 0.18913879990577698, "learning_rate": 4.719264527286075e-05, "loss": 0.0157, "step": 43360 }, { "grad_norm": 0.16483929753303528, "learning_rate": 4.717200974477744e-05, "loss": 0.0177, "step": 43370 }, { "grad_norm": 0.19458912312984467, "learning_rate": 4.715137469992034e-05, "loss": 0.021, "step": 43380 }, { "grad_norm": 0.7167888283729553, "learning_rate": 4.713074014181539e-05, "loss": 0.0201, "step": 43390 }, { "grad_norm": 0.2184685617685318, "learning_rate": 4.711010607398846e-05, "loss": 0.013, "step": 43400 }, { "grad_norm": 0.4009825885295868, "learning_rate": 4.708947249996538e-05, "loss": 0.0131, "step": 43410 }, { "grad_norm": 0.19036346673965454, "learning_rate": 4.706883942327183e-05, "loss": 0.0125, "step": 43420 }, { "grad_norm": 0.21471072733402252, "learning_rate": 4.704820684743347e-05, "loss": 0.0177, "step": 43430 }, { "grad_norm": 0.2637695074081421, "learning_rate": 4.702757477597581e-05, "loss": 0.0122, "step": 43440 }, { "grad_norm": 0.19177453219890594, "learning_rate": 4.700694321242435e-05, "loss": 0.0153, "step": 43450 }, { "grad_norm": 0.39404934644699097, "learning_rate": 4.698631216030441e-05, "loss": 0.0153, "step": 43460 }, { "grad_norm": 0.11816108226776123, "learning_rate": 4.6965681623141314e-05, "loss": 0.0113, "step": 43470 }, { "grad_norm": 0.1669534146785736, "learning_rate": 4.694505160446024e-05, "loss": 0.0105, "step": 43480 }, { "grad_norm": 0.17652527987957, "learning_rate": 4.692442210778628e-05, "loss": 0.0147, "step": 43490 }, { "grad_norm": 0.3692210614681244, "learning_rate": 4.690379313664448e-05, "loss": 0.016, "step": 43500 }, { "grad_norm": 0.2603709399700165, "learning_rate": 4.688316469455973e-05, "loss": 0.0173, "step": 43510 }, { "grad_norm": 0.13846810162067413, "learning_rate": 4.6862536785056914e-05, "loss": 0.0145, "step": 43520 }, { "grad_norm": 0.35643038153648376, "learning_rate": 4.684190941166073e-05, "loss": 0.0175, "step": 43530 }, { "grad_norm": 0.2887943983078003, "learning_rate": 4.682128257789587e-05, "loss": 0.0135, "step": 43540 }, { "grad_norm": 0.3522639572620392, "learning_rate": 4.680065628728687e-05, "loss": 0.011, "step": 43550 }, { "grad_norm": 0.2639501094818115, "learning_rate": 4.678003054335822e-05, "loss": 0.015, "step": 43560 }, { "grad_norm": 0.27531999349594116, "learning_rate": 4.675940534963428e-05, "loss": 0.0174, "step": 43570 }, { "grad_norm": 0.2036457508802414, "learning_rate": 4.673878070963931e-05, "loss": 0.0148, "step": 43580 }, { "grad_norm": 0.20174327492713928, "learning_rate": 4.671815662689756e-05, "loss": 0.0131, "step": 43590 }, { "grad_norm": 0.26134824752807617, "learning_rate": 4.669753310493306e-05, "loss": 0.0149, "step": 43600 }, { "grad_norm": 0.19067835807800293, "learning_rate": 4.6676910147269845e-05, "loss": 0.0174, "step": 43610 }, { "grad_norm": 0.2684677541255951, "learning_rate": 4.665628775743177e-05, "loss": 0.0162, "step": 43620 }, { "grad_norm": 0.39737364649772644, "learning_rate": 4.66356659389427e-05, "loss": 0.0183, "step": 43630 }, { "grad_norm": 0.21554571390151978, "learning_rate": 4.661504469532629e-05, "loss": 0.0125, "step": 43640 }, { "grad_norm": 0.3171975612640381, "learning_rate": 4.659442403010617e-05, "loss": 0.0141, "step": 43650 }, { "grad_norm": 0.23981012403964996, "learning_rate": 4.6573803946805845e-05, "loss": 0.0147, "step": 43660 }, { "grad_norm": 0.2478531152009964, "learning_rate": 4.6553184448948695e-05, "loss": 0.0142, "step": 43670 }, { "grad_norm": 0.2129640132188797, "learning_rate": 4.653256554005807e-05, "loss": 0.0125, "step": 43680 }, { "grad_norm": 0.2200545072555542, "learning_rate": 4.6511947223657145e-05, "loss": 0.0152, "step": 43690 }, { "grad_norm": 0.28651320934295654, "learning_rate": 4.649132950326906e-05, "loss": 0.0163, "step": 43700 }, { "grad_norm": 0.35661178827285767, "learning_rate": 4.647071238241679e-05, "loss": 0.0182, "step": 43710 }, { "grad_norm": 0.22940941154956818, "learning_rate": 4.6450095864623264e-05, "loss": 0.0121, "step": 43720 }, { "grad_norm": 0.1955365389585495, "learning_rate": 4.642947995341123e-05, "loss": 0.0093, "step": 43730 }, { "grad_norm": 0.14363577961921692, "learning_rate": 4.640886465230345e-05, "loss": 0.0164, "step": 43740 }, { "grad_norm": 0.19104284048080444, "learning_rate": 4.6388249964822485e-05, "loss": 0.0186, "step": 43750 }, { "grad_norm": 0.19896692037582397, "learning_rate": 4.63676358944908e-05, "loss": 0.0127, "step": 43760 }, { "grad_norm": 0.3427293002605438, "learning_rate": 4.634702244483081e-05, "loss": 0.0197, "step": 43770 }, { "grad_norm": 0.1982497274875641, "learning_rate": 4.632640961936475e-05, "loss": 0.015, "step": 43780 }, { "grad_norm": 0.26634371280670166, "learning_rate": 4.6305797421614835e-05, "loss": 0.0139, "step": 43790 }, { "grad_norm": 0.16071000695228577, "learning_rate": 4.6285185855103105e-05, "loss": 0.0129, "step": 43800 }, { "grad_norm": 0.23089614510536194, "learning_rate": 4.626457492335151e-05, "loss": 0.0137, "step": 43810 }, { "grad_norm": 0.15994171798229218, "learning_rate": 4.624396462988188e-05, "loss": 0.0108, "step": 43820 }, { "grad_norm": 0.16461560130119324, "learning_rate": 4.6223354978215985e-05, "loss": 0.0143, "step": 43830 }, { "grad_norm": 0.23865775763988495, "learning_rate": 4.620274597187544e-05, "loss": 0.0163, "step": 43840 }, { "grad_norm": 0.16446366906166077, "learning_rate": 4.6182137614381736e-05, "loss": 0.0167, "step": 43850 }, { "grad_norm": 0.30385664105415344, "learning_rate": 4.616152990925631e-05, "loss": 0.012, "step": 43860 }, { "grad_norm": 0.13003812730312347, "learning_rate": 4.614092286002041e-05, "loss": 0.0128, "step": 43870 }, { "grad_norm": 0.36197951436042786, "learning_rate": 4.612031647019528e-05, "loss": 0.0144, "step": 43880 }, { "grad_norm": 0.2631435990333557, "learning_rate": 4.609971074330193e-05, "loss": 0.016, "step": 43890 }, { "grad_norm": 0.310317724943161, "learning_rate": 4.607910568286136e-05, "loss": 0.0148, "step": 43900 }, { "grad_norm": 0.3080042004585266, "learning_rate": 4.605850129239438e-05, "loss": 0.015, "step": 43910 }, { "grad_norm": 0.22307844460010529, "learning_rate": 4.603789757542174e-05, "loss": 0.0124, "step": 43920 }, { "grad_norm": 0.2670902609825134, "learning_rate": 4.601729453546404e-05, "loss": 0.0171, "step": 43930 }, { "grad_norm": 0.22697021067142487, "learning_rate": 4.599669217604177e-05, "loss": 0.0141, "step": 43940 }, { "grad_norm": 0.5602918267250061, "learning_rate": 4.597609050067532e-05, "loss": 0.0239, "step": 43950 }, { "grad_norm": 0.25818318128585815, "learning_rate": 4.595548951288495e-05, "loss": 0.0166, "step": 43960 }, { "grad_norm": 0.22585321962833405, "learning_rate": 4.593488921619081e-05, "loss": 0.0182, "step": 43970 }, { "grad_norm": 0.37948864698410034, "learning_rate": 4.591428961411289e-05, "loss": 0.0177, "step": 43980 }, { "grad_norm": 0.27548569440841675, "learning_rate": 4.589369071017117e-05, "loss": 0.0184, "step": 43990 }, { "grad_norm": 0.13693726062774658, "learning_rate": 4.587309250788538e-05, "loss": 0.0164, "step": 44000 }, { "grad_norm": 0.18347689509391785, "learning_rate": 4.585249501077522e-05, "loss": 0.0097, "step": 44010 }, { "grad_norm": 0.12168547511100769, "learning_rate": 4.583189822236021e-05, "loss": 0.0122, "step": 44020 }, { "grad_norm": 0.2289748191833496, "learning_rate": 4.5811302146159816e-05, "loss": 0.0123, "step": 44030 }, { "grad_norm": 0.15618300437927246, "learning_rate": 4.579070678569332e-05, "loss": 0.013, "step": 44040 }, { "grad_norm": 0.25414764881134033, "learning_rate": 4.5770112144479904e-05, "loss": 0.0139, "step": 44050 }, { "grad_norm": 0.30350661277770996, "learning_rate": 4.5749518226038645e-05, "loss": 0.0131, "step": 44060 }, { "grad_norm": 0.31950563192367554, "learning_rate": 4.572892503388845e-05, "loss": 0.0115, "step": 44070 }, { "grad_norm": 0.6062046885490417, "learning_rate": 4.570833257154817e-05, "loss": 0.0161, "step": 44080 }, { "grad_norm": 0.39880090951919556, "learning_rate": 4.568774084253646e-05, "loss": 0.0183, "step": 44090 }, { "grad_norm": 0.22287611663341522, "learning_rate": 4.566714985037191e-05, "loss": 0.0205, "step": 44100 }, { "grad_norm": 0.23656371235847473, "learning_rate": 4.564655959857295e-05, "loss": 0.0153, "step": 44110 }, { "grad_norm": 0.23902830481529236, "learning_rate": 4.5625970090657893e-05, "loss": 0.0155, "step": 44120 }, { "grad_norm": 0.17019107937812805, "learning_rate": 4.560538133014491e-05, "loss": 0.0119, "step": 44130 }, { "grad_norm": 0.22310452163219452, "learning_rate": 4.5584793320552055e-05, "loss": 0.017, "step": 44140 }, { "grad_norm": 0.15475566685199738, "learning_rate": 4.556420606539728e-05, "loss": 0.0165, "step": 44150 }, { "grad_norm": 0.14794370532035828, "learning_rate": 4.554361956819836e-05, "loss": 0.0142, "step": 44160 }, { "grad_norm": 0.15264026820659637, "learning_rate": 4.552303383247299e-05, "loss": 0.0133, "step": 44170 }, { "grad_norm": 0.6552991271018982, "learning_rate": 4.5502448861738664e-05, "loss": 0.0159, "step": 44180 }, { "grad_norm": 0.3570263385772705, "learning_rate": 4.548186465951284e-05, "loss": 0.0206, "step": 44190 }, { "grad_norm": 0.26092901825904846, "learning_rate": 4.546128122931277e-05, "loss": 0.0191, "step": 44200 }, { "grad_norm": 0.21764449775218964, "learning_rate": 4.544069857465561e-05, "loss": 0.0142, "step": 44210 }, { "grad_norm": 0.35105258226394653, "learning_rate": 4.542011669905837e-05, "loss": 0.0119, "step": 44220 }, { "grad_norm": 0.23070384562015533, "learning_rate": 4.53995356060379e-05, "loss": 0.0116, "step": 44230 }, { "grad_norm": 0.10563096404075623, "learning_rate": 4.5378955299110994e-05, "loss": 0.0131, "step": 44240 }, { "grad_norm": 0.30786290764808655, "learning_rate": 4.5358375781794233e-05, "loss": 0.0149, "step": 44250 }, { "grad_norm": 0.18565572798252106, "learning_rate": 4.533779705760411e-05, "loss": 0.0169, "step": 44260 }, { "grad_norm": 0.15716759860515594, "learning_rate": 4.5317219130056934e-05, "loss": 0.0122, "step": 44270 }, { "grad_norm": 0.239427849650383, "learning_rate": 4.5296642002668964e-05, "loss": 0.0156, "step": 44280 }, { "grad_norm": 0.2889481484889984, "learning_rate": 4.5276065678956216e-05, "loss": 0.0154, "step": 44290 }, { "grad_norm": 0.19725452363491058, "learning_rate": 4.525549016243466e-05, "loss": 0.013, "step": 44300 }, { "grad_norm": 0.1729307621717453, "learning_rate": 4.523491545662008e-05, "loss": 0.0152, "step": 44310 }, { "grad_norm": 0.19012433290481567, "learning_rate": 4.5214341565028104e-05, "loss": 0.0158, "step": 44320 }, { "grad_norm": 0.2879362404346466, "learning_rate": 4.519376849117428e-05, "loss": 0.0144, "step": 44330 }, { "grad_norm": 0.18723152577877045, "learning_rate": 4.517319623857395e-05, "loss": 0.0216, "step": 44340 }, { "grad_norm": 0.22792065143585205, "learning_rate": 4.5152624810742394e-05, "loss": 0.0145, "step": 44350 }, { "grad_norm": 0.15714086592197418, "learning_rate": 4.5132054211194664e-05, "loss": 0.014, "step": 44360 }, { "grad_norm": 0.24785038828849792, "learning_rate": 4.511148444344574e-05, "loss": 0.013, "step": 44370 }, { "grad_norm": 0.2208065390586853, "learning_rate": 4.509091551101041e-05, "loss": 0.016, "step": 44380 }, { "grad_norm": 0.20615467429161072, "learning_rate": 4.507034741740338e-05, "loss": 0.0113, "step": 44390 }, { "grad_norm": 0.29908230900764465, "learning_rate": 4.5049780166139145e-05, "loss": 0.0174, "step": 44400 }, { "grad_norm": 0.26384294033050537, "learning_rate": 4.5029213760732075e-05, "loss": 0.0242, "step": 44410 }, { "grad_norm": 0.2045619785785675, "learning_rate": 4.5008648204696434e-05, "loss": 0.0182, "step": 44420 }, { "grad_norm": 0.24262697994709015, "learning_rate": 4.4988083501546284e-05, "loss": 0.0106, "step": 44430 }, { "grad_norm": 0.1737620234489441, "learning_rate": 4.4967519654795606e-05, "loss": 0.0122, "step": 44440 }, { "grad_norm": 0.2243032157421112, "learning_rate": 4.494695666795816e-05, "loss": 0.0147, "step": 44450 }, { "grad_norm": 0.2029704451560974, "learning_rate": 4.4926394544547644e-05, "loss": 0.0136, "step": 44460 }, { "grad_norm": 0.17211918532848358, "learning_rate": 4.490583328807752e-05, "loss": 0.0154, "step": 44470 }, { "grad_norm": 0.23511509597301483, "learning_rate": 4.488527290206117e-05, "loss": 0.0151, "step": 44480 }, { "grad_norm": 0.24928702414035797, "learning_rate": 4.48647133900118e-05, "loss": 0.0122, "step": 44490 }, { "grad_norm": 0.184550940990448, "learning_rate": 4.484415475544244e-05, "loss": 0.0195, "step": 44500 }, { "grad_norm": 0.1658187359571457, "learning_rate": 4.4823597001866035e-05, "loss": 0.0123, "step": 44510 }, { "grad_norm": 0.16029325127601624, "learning_rate": 4.480304013279532e-05, "loss": 0.013, "step": 44520 }, { "grad_norm": 0.16948221623897552, "learning_rate": 4.478248415174292e-05, "loss": 0.0107, "step": 44530 }, { "grad_norm": 0.25264155864715576, "learning_rate": 4.476192906222126e-05, "loss": 0.0182, "step": 44540 }, { "grad_norm": 0.46504834294319153, "learning_rate": 4.474137486774268e-05, "loss": 0.027, "step": 44550 }, { "grad_norm": 0.16032925248146057, "learning_rate": 4.4720821571819296e-05, "loss": 0.0148, "step": 44560 }, { "grad_norm": 0.18560844659805298, "learning_rate": 4.470026917796314e-05, "loss": 0.0197, "step": 44570 }, { "grad_norm": 0.2565123438835144, "learning_rate": 4.4679717689686005e-05, "loss": 0.0255, "step": 44580 }, { "grad_norm": 0.25516825914382935, "learning_rate": 4.465916711049963e-05, "loss": 0.0147, "step": 44590 }, { "grad_norm": 0.23346926271915436, "learning_rate": 4.4638617443915524e-05, "loss": 0.0182, "step": 44600 }, { "grad_norm": 1.036013126373291, "learning_rate": 4.4618068693445055e-05, "loss": 0.0208, "step": 44610 }, { "grad_norm": 0.18846525251865387, "learning_rate": 4.459752086259946e-05, "loss": 0.0134, "step": 44620 }, { "grad_norm": 0.4737779200077057, "learning_rate": 4.457697395488977e-05, "loss": 0.0251, "step": 44630 }, { "grad_norm": 0.23865847289562225, "learning_rate": 4.455642797382693e-05, "loss": 0.0198, "step": 44640 }, { "grad_norm": 0.29129189252853394, "learning_rate": 4.4535882922921643e-05, "loss": 0.0193, "step": 44650 }, { "grad_norm": 0.2554002106189728, "learning_rate": 4.451533880568455e-05, "loss": 0.0119, "step": 44660 }, { "grad_norm": 0.19731907546520233, "learning_rate": 4.449479562562603e-05, "loss": 0.0206, "step": 44670 }, { "grad_norm": 0.2264396846294403, "learning_rate": 4.447425338625639e-05, "loss": 0.0156, "step": 44680 }, { "grad_norm": 0.2343824952840805, "learning_rate": 4.4453712091085705e-05, "loss": 0.0119, "step": 44690 }, { "grad_norm": 0.22951456904411316, "learning_rate": 4.443317174362392e-05, "loss": 0.0164, "step": 44700 }, { "grad_norm": 0.31648412346839905, "learning_rate": 4.441263234738085e-05, "loss": 0.0158, "step": 44710 }, { "grad_norm": 0.22106021642684937, "learning_rate": 4.4392093905866086e-05, "loss": 0.0117, "step": 44720 }, { "grad_norm": 0.2874446213245392, "learning_rate": 4.437155642258911e-05, "loss": 0.0158, "step": 44730 }, { "grad_norm": 0.2008458375930786, "learning_rate": 4.4351019901059177e-05, "loss": 0.0189, "step": 44740 }, { "grad_norm": 0.29469311237335205, "learning_rate": 4.433048434478547e-05, "loss": 0.0243, "step": 44750 }, { "grad_norm": 0.26234930753707886, "learning_rate": 4.4309949757276906e-05, "loss": 0.0186, "step": 44760 }, { "grad_norm": 0.1969006210565567, "learning_rate": 4.428941614204233e-05, "loss": 0.0189, "step": 44770 }, { "grad_norm": 0.19008439779281616, "learning_rate": 4.426888350259034e-05, "loss": 0.0159, "step": 44780 }, { "grad_norm": 0.255880743265152, "learning_rate": 4.424835184242939e-05, "loss": 0.027, "step": 44790 }, { "grad_norm": 0.16885297000408173, "learning_rate": 4.422782116506784e-05, "loss": 0.0107, "step": 44800 }, { "grad_norm": 0.2881675362586975, "learning_rate": 4.4207291474013754e-05, "loss": 0.0177, "step": 44810 }, { "grad_norm": 0.21206453442573547, "learning_rate": 4.418676277277514e-05, "loss": 0.0219, "step": 44820 }, { "grad_norm": 0.22935649752616882, "learning_rate": 4.416623506485976e-05, "loss": 0.0179, "step": 44830 }, { "grad_norm": 0.19582118093967438, "learning_rate": 4.414570835377527e-05, "loss": 0.0152, "step": 44840 }, { "grad_norm": 0.1890067458152771, "learning_rate": 4.412518264302909e-05, "loss": 0.0167, "step": 44850 }, { "grad_norm": 0.3181256651878357, "learning_rate": 4.4104657936128535e-05, "loss": 0.0143, "step": 44860 }, { "grad_norm": 0.2776077091693878, "learning_rate": 4.408413423658071e-05, "loss": 0.0153, "step": 44870 }, { "grad_norm": 0.23518474400043488, "learning_rate": 4.406361154789253e-05, "loss": 0.0166, "step": 44880 }, { "grad_norm": 0.32405343651771545, "learning_rate": 4.4043089873570776e-05, "loss": 0.0181, "step": 44890 }, { "grad_norm": 0.35418134927749634, "learning_rate": 4.4022569217122025e-05, "loss": 0.0155, "step": 44900 }, { "grad_norm": 0.1605292111635208, "learning_rate": 4.400204958205274e-05, "loss": 0.0139, "step": 44910 }, { "grad_norm": 0.26002585887908936, "learning_rate": 4.3981530971869125e-05, "loss": 0.0192, "step": 44920 }, { "grad_norm": 0.2447129786014557, "learning_rate": 4.396101339007727e-05, "loss": 0.0131, "step": 44930 }, { "grad_norm": 0.5377321839332581, "learning_rate": 4.394049684018304e-05, "loss": 0.0186, "step": 44940 }, { "grad_norm": 0.2698923647403717, "learning_rate": 4.3919981325692186e-05, "loss": 0.0159, "step": 44950 }, { "grad_norm": 0.12822797894477844, "learning_rate": 4.389946685011024e-05, "loss": 0.0108, "step": 44960 }, { "grad_norm": 0.2053861916065216, "learning_rate": 4.387895341694255e-05, "loss": 0.0205, "step": 44970 }, { "grad_norm": 0.14764700829982758, "learning_rate": 4.3858441029694324e-05, "loss": 0.0193, "step": 44980 }, { "grad_norm": 0.3278878331184387, "learning_rate": 4.3837929691870527e-05, "loss": 0.0137, "step": 44990 }, { "grad_norm": 0.24319900572299957, "learning_rate": 4.381741940697604e-05, "loss": 0.0193, "step": 45000 }, { "grad_norm": 0.2797369658946991, "learning_rate": 4.379691017851547e-05, "loss": 0.0132, "step": 45010 }, { "grad_norm": 0.2901271879673004, "learning_rate": 4.3776402009993304e-05, "loss": 0.0216, "step": 45020 }, { "grad_norm": 0.1598609983921051, "learning_rate": 4.3755894904913794e-05, "loss": 0.0143, "step": 45030 }, { "grad_norm": 0.23079656064510345, "learning_rate": 4.373538886678109e-05, "loss": 0.0136, "step": 45040 }, { "grad_norm": 0.1379605084657669, "learning_rate": 4.371488389909909e-05, "loss": 0.0117, "step": 45050 }, { "grad_norm": 0.17582732439041138, "learning_rate": 4.3694380005371515e-05, "loss": 0.0181, "step": 45060 }, { "grad_norm": 0.2097046673297882, "learning_rate": 4.367387718910196e-05, "loss": 0.0123, "step": 45070 }, { "grad_norm": 0.2554190754890442, "learning_rate": 4.3653375453793764e-05, "loss": 0.0162, "step": 45080 }, { "grad_norm": 0.19550283253192902, "learning_rate": 4.3632874802950136e-05, "loss": 0.0125, "step": 45090 }, { "grad_norm": 0.262952983379364, "learning_rate": 4.3612375240074034e-05, "loss": 0.0155, "step": 45100 }, { "grad_norm": 0.1900954395532608, "learning_rate": 4.3591876768668325e-05, "loss": 0.0158, "step": 45110 }, { "grad_norm": 0.1893354207277298, "learning_rate": 4.3571379392235605e-05, "loss": 0.0185, "step": 45120 }, { "grad_norm": 0.3042406439781189, "learning_rate": 4.3550883114278335e-05, "loss": 0.0133, "step": 45130 }, { "grad_norm": 0.19670146703720093, "learning_rate": 4.353038793829876e-05, "loss": 0.0131, "step": 45140 }, { "grad_norm": 0.2326669991016388, "learning_rate": 4.350989386779891e-05, "loss": 0.0141, "step": 45150 }, { "grad_norm": 0.20006781816482544, "learning_rate": 4.3489400906280724e-05, "loss": 0.0223, "step": 45160 }, { "grad_norm": 0.2105630487203598, "learning_rate": 4.3468909057245845e-05, "loss": 0.0188, "step": 45170 }, { "grad_norm": 0.12898200750350952, "learning_rate": 4.3448418324195794e-05, "loss": 0.018, "step": 45180 }, { "grad_norm": 0.16207224130630493, "learning_rate": 4.342792871063184e-05, "loss": 0.0204, "step": 45190 }, { "grad_norm": 0.29474204778671265, "learning_rate": 4.3407440220055145e-05, "loss": 0.016, "step": 45200 }, { "grad_norm": 0.18869169056415558, "learning_rate": 4.33869528559666e-05, "loss": 0.0115, "step": 45210 }, { "grad_norm": 0.1617625504732132, "learning_rate": 4.336646662186696e-05, "loss": 0.0101, "step": 45220 }, { "grad_norm": 0.14183443784713745, "learning_rate": 4.334598152125672e-05, "loss": 0.0179, "step": 45230 }, { "grad_norm": 0.2962068021297455, "learning_rate": 4.3325497557636276e-05, "loss": 0.0159, "step": 45240 }, { "grad_norm": 0.1514901965856552, "learning_rate": 4.330501473450574e-05, "loss": 0.0157, "step": 45250 }, { "grad_norm": 0.23221610486507416, "learning_rate": 4.328453305536507e-05, "loss": 0.0122, "step": 45260 }, { "grad_norm": 0.20990820229053497, "learning_rate": 4.326405252371404e-05, "loss": 0.0147, "step": 45270 }, { "grad_norm": 0.21568378806114197, "learning_rate": 4.324357314305221e-05, "loss": 0.0142, "step": 45280 }, { "grad_norm": 0.33564385771751404, "learning_rate": 4.3223094916878945e-05, "loss": 0.0128, "step": 45290 }, { "grad_norm": 0.15653188526630402, "learning_rate": 4.320261784869338e-05, "loss": 0.023, "step": 45300 }, { "grad_norm": 0.10526378452777863, "learning_rate": 4.318214194199455e-05, "loss": 0.0149, "step": 45310 }, { "grad_norm": 0.1073727011680603, "learning_rate": 4.316166720028118e-05, "loss": 0.0113, "step": 45320 }, { "grad_norm": 0.1558266580104828, "learning_rate": 4.3141193627051864e-05, "loss": 0.0128, "step": 45330 }, { "grad_norm": 0.2564347982406616, "learning_rate": 4.312072122580496e-05, "loss": 0.0147, "step": 45340 }, { "grad_norm": 0.5587430596351624, "learning_rate": 4.3100250000038646e-05, "loss": 0.0183, "step": 45350 }, { "grad_norm": 0.4610910415649414, "learning_rate": 4.307977995325091e-05, "loss": 0.0244, "step": 45360 }, { "grad_norm": 0.21610422432422638, "learning_rate": 4.30593110889395e-05, "loss": 0.0169, "step": 45370 }, { "grad_norm": 0.2797432243824005, "learning_rate": 4.3038843410602016e-05, "loss": 0.0108, "step": 45380 }, { "grad_norm": 0.17663753032684326, "learning_rate": 4.3018376921735774e-05, "loss": 0.0153, "step": 45390 }, { "grad_norm": 0.15154463052749634, "learning_rate": 4.299791162583799e-05, "loss": 0.0147, "step": 45400 }, { "grad_norm": 0.30715784430503845, "learning_rate": 4.29774475264056e-05, "loss": 0.0197, "step": 45410 }, { "grad_norm": 0.21253709495067596, "learning_rate": 4.2956984626935365e-05, "loss": 0.0111, "step": 45420 }, { "grad_norm": 0.148468479514122, "learning_rate": 4.293652293092383e-05, "loss": 0.0164, "step": 45430 }, { "grad_norm": 0.20471632480621338, "learning_rate": 4.2916062441867324e-05, "loss": 0.0154, "step": 45440 }, { "grad_norm": 0.24206486344337463, "learning_rate": 4.289560316326201e-05, "loss": 0.0191, "step": 45450 }, { "grad_norm": 0.3053034245967865, "learning_rate": 4.28751450986038e-05, "loss": 0.0152, "step": 45460 }, { "grad_norm": 0.23390421271324158, "learning_rate": 4.2854688251388444e-05, "loss": 0.0159, "step": 45470 }, { "grad_norm": 0.20276598632335663, "learning_rate": 4.2834232625111425e-05, "loss": 0.0193, "step": 45480 }, { "grad_norm": 0.2695057690143585, "learning_rate": 4.2813778223268086e-05, "loss": 0.0115, "step": 45490 }, { "grad_norm": 0.46368712186813354, "learning_rate": 4.2793325049353477e-05, "loss": 0.0143, "step": 45500 }, { "grad_norm": 0.2990853190422058, "learning_rate": 4.2772873106862535e-05, "loss": 0.0145, "step": 45510 }, { "grad_norm": 0.15124103426933289, "learning_rate": 4.275242239928993e-05, "loss": 0.0147, "step": 45520 }, { "grad_norm": 0.2111523598432541, "learning_rate": 4.273197293013009e-05, "loss": 0.0106, "step": 45530 }, { "grad_norm": 0.22971661388874054, "learning_rate": 4.271152470287731e-05, "loss": 0.0093, "step": 45540 }, { "grad_norm": 0.11658167839050293, "learning_rate": 4.2691077721025594e-05, "loss": 0.0143, "step": 45550 }, { "grad_norm": 0.22347992658615112, "learning_rate": 4.267063198806883e-05, "loss": 0.0129, "step": 45560 }, { "grad_norm": 0.190806046128273, "learning_rate": 4.2650187507500574e-05, "loss": 0.0135, "step": 45570 }, { "grad_norm": 0.20925356447696686, "learning_rate": 4.2629744282814275e-05, "loss": 0.0092, "step": 45580 }, { "grad_norm": 0.1603039652109146, "learning_rate": 4.2609302317503074e-05, "loss": 0.0123, "step": 45590 }, { "grad_norm": 0.12703412771224976, "learning_rate": 4.258886161505999e-05, "loss": 0.012, "step": 45600 }, { "grad_norm": 0.17997460067272186, "learning_rate": 4.2568422178977775e-05, "loss": 0.0117, "step": 45610 }, { "grad_norm": 0.2584209144115448, "learning_rate": 4.254798401274894e-05, "loss": 0.0184, "step": 45620 }, { "grad_norm": 0.28565436601638794, "learning_rate": 4.252754711986583e-05, "loss": 0.0144, "step": 45630 }, { "grad_norm": 0.16272865235805511, "learning_rate": 4.250711150382052e-05, "loss": 0.0133, "step": 45640 }, { "grad_norm": 0.15969334542751312, "learning_rate": 4.248667716810495e-05, "loss": 0.0166, "step": 45650 }, { "grad_norm": 0.16480770707130432, "learning_rate": 4.246624411621074e-05, "loss": 0.0106, "step": 45660 }, { "grad_norm": 0.3237471580505371, "learning_rate": 4.244581235162938e-05, "loss": 0.0142, "step": 45670 }, { "grad_norm": 0.13496167957782745, "learning_rate": 4.2425381877852075e-05, "loss": 0.0168, "step": 45680 }, { "grad_norm": 0.19055071473121643, "learning_rate": 4.2404952698369856e-05, "loss": 0.0126, "step": 45690 }, { "grad_norm": 0.10516860336065292, "learning_rate": 4.238452481667349e-05, "loss": 0.013, "step": 45700 }, { "grad_norm": 0.24274364113807678, "learning_rate": 4.2364098236253526e-05, "loss": 0.0182, "step": 45710 }, { "grad_norm": 0.2444211095571518, "learning_rate": 4.2343672960600356e-05, "loss": 0.0133, "step": 45720 }, { "grad_norm": 0.18093493580818176, "learning_rate": 4.232324899320406e-05, "loss": 0.0101, "step": 45730 }, { "grad_norm": 0.17799049615859985, "learning_rate": 4.230282633755457e-05, "loss": 0.0119, "step": 45740 }, { "grad_norm": 0.1414921134710312, "learning_rate": 4.2282404997141515e-05, "loss": 0.0132, "step": 45750 }, { "grad_norm": 0.22706079483032227, "learning_rate": 4.2261984975454397e-05, "loss": 0.0169, "step": 45760 }, { "grad_norm": 0.2568851709365845, "learning_rate": 4.224156627598239e-05, "loss": 0.0109, "step": 45770 }, { "grad_norm": 0.17944960296154022, "learning_rate": 4.222114890221453e-05, "loss": 0.0155, "step": 45780 }, { "grad_norm": 0.24358808994293213, "learning_rate": 4.2200732857639546e-05, "loss": 0.0167, "step": 45790 }, { "grad_norm": 0.6316497921943665, "learning_rate": 4.2180318145746035e-05, "loss": 0.0158, "step": 45800 }, { "grad_norm": 0.22316011786460876, "learning_rate": 4.215990477002227e-05, "loss": 0.0183, "step": 45810 }, { "grad_norm": 0.23614688217639923, "learning_rate": 4.2139492733956356e-05, "loss": 0.0119, "step": 45820 }, { "grad_norm": 0.5676702857017517, "learning_rate": 4.211908204103615e-05, "loss": 0.0288, "step": 45830 }, { "grad_norm": 0.34180667996406555, "learning_rate": 4.2098672694749265e-05, "loss": 0.0253, "step": 45840 }, { "grad_norm": 0.5200327634811401, "learning_rate": 4.2078264698583133e-05, "loss": 0.0181, "step": 45850 }, { "grad_norm": 0.19377821683883667, "learning_rate": 4.205785805602488e-05, "loss": 0.0142, "step": 45860 }, { "grad_norm": 0.5740883946418762, "learning_rate": 4.203745277056149e-05, "loss": 0.0156, "step": 45870 }, { "grad_norm": 0.17947068810462952, "learning_rate": 4.201704884567964e-05, "loss": 0.0136, "step": 45880 }, { "grad_norm": 0.19991521537303925, "learning_rate": 4.1996646284865816e-05, "loss": 0.0214, "step": 45890 }, { "grad_norm": 0.29546359181404114, "learning_rate": 4.197624509160625e-05, "loss": 0.0153, "step": 45900 }, { "grad_norm": 0.191972553730011, "learning_rate": 4.195584526938692e-05, "loss": 0.0109, "step": 45910 }, { "grad_norm": 0.27275359630584717, "learning_rate": 4.193544682169365e-05, "loss": 0.0154, "step": 45920 }, { "grad_norm": 0.265752375125885, "learning_rate": 4.1915049752011946e-05, "loss": 0.0144, "step": 45930 }, { "grad_norm": 0.4060683250427246, "learning_rate": 4.189465406382712e-05, "loss": 0.013, "step": 45940 }, { "grad_norm": 0.32918936014175415, "learning_rate": 4.187425976062422e-05, "loss": 0.0182, "step": 45950 }, { "grad_norm": 0.25954142212867737, "learning_rate": 4.18538668458881e-05, "loss": 0.0198, "step": 45960 }, { "grad_norm": 0.2001768946647644, "learning_rate": 4.183347532310333e-05, "loss": 0.0131, "step": 45970 }, { "grad_norm": 0.19756782054901123, "learning_rate": 4.181308519575429e-05, "loss": 0.0166, "step": 45980 }, { "grad_norm": 0.22613175213336945, "learning_rate": 4.179269646732507e-05, "loss": 0.0215, "step": 45990 }, { "grad_norm": 0.20150983333587646, "learning_rate": 4.177230914129954e-05, "loss": 0.0114, "step": 46000 }, { "grad_norm": 0.2045871615409851, "learning_rate": 4.175192322116136e-05, "loss": 0.0129, "step": 46010 }, { "grad_norm": 0.18589657545089722, "learning_rate": 4.173153871039391e-05, "loss": 0.0097, "step": 46020 }, { "grad_norm": 0.2338852435350418, "learning_rate": 4.171115561248036e-05, "loss": 0.0148, "step": 46030 }, { "grad_norm": 0.16840782761573792, "learning_rate": 4.16907739309036e-05, "loss": 0.0129, "step": 46040 }, { "grad_norm": 0.22329865396022797, "learning_rate": 4.167039366914633e-05, "loss": 0.012, "step": 46050 }, { "grad_norm": 0.22182342410087585, "learning_rate": 4.165001483069096e-05, "loss": 0.0135, "step": 46060 }, { "grad_norm": 0.19955548644065857, "learning_rate": 4.16296374190197e-05, "loss": 0.0202, "step": 46070 }, { "grad_norm": 0.23402263224124908, "learning_rate": 4.1609261437614464e-05, "loss": 0.0134, "step": 46080 }, { "grad_norm": 0.3637716472148895, "learning_rate": 4.158888688995696e-05, "loss": 0.0157, "step": 46090 }, { "grad_norm": 0.5737534165382385, "learning_rate": 4.1568513779528645e-05, "loss": 0.0126, "step": 46100 }, { "grad_norm": 0.20439954102039337, "learning_rate": 4.1548142109810704e-05, "loss": 0.0183, "step": 46110 }, { "grad_norm": 0.1342868208885193, "learning_rate": 4.152777188428414e-05, "loss": 0.0131, "step": 46120 }, { "grad_norm": 0.21444065868854523, "learning_rate": 4.1507403106429646e-05, "loss": 0.0112, "step": 46130 }, { "grad_norm": 0.28993648290634155, "learning_rate": 4.148703577972768e-05, "loss": 0.0147, "step": 46140 }, { "grad_norm": 0.33610352873802185, "learning_rate": 4.146666990765846e-05, "loss": 0.0154, "step": 46150 }, { "grad_norm": 0.15839402377605438, "learning_rate": 4.1446305493701986e-05, "loss": 0.0138, "step": 46160 }, { "grad_norm": 0.2598436772823334, "learning_rate": 4.142594254133796e-05, "loss": 0.0151, "step": 46170 }, { "grad_norm": 0.2412457913160324, "learning_rate": 4.1405581054045835e-05, "loss": 0.0165, "step": 46180 }, { "grad_norm": 0.2545396387577057, "learning_rate": 4.1385221035304864e-05, "loss": 0.0133, "step": 46190 }, { "grad_norm": 0.2784145474433899, "learning_rate": 4.136486248859398e-05, "loss": 0.0183, "step": 46200 }, { "grad_norm": 0.37625426054000854, "learning_rate": 4.1344505417391955e-05, "loss": 0.0159, "step": 46210 }, { "grad_norm": 0.18658031523227692, "learning_rate": 4.132414982517721e-05, "loss": 0.0142, "step": 46220 }, { "grad_norm": 0.11049514263868332, "learning_rate": 4.130379571542798e-05, "loss": 0.0124, "step": 46230 }, { "grad_norm": 0.09239871799945831, "learning_rate": 4.12834430916222e-05, "loss": 0.0143, "step": 46240 }, { "grad_norm": 0.22735144197940826, "learning_rate": 4.126309195723763e-05, "loss": 0.0104, "step": 46250 }, { "grad_norm": 0.25378039479255676, "learning_rate": 4.124274231575168e-05, "loss": 0.0152, "step": 46260 }, { "grad_norm": 0.1355726569890976, "learning_rate": 4.122239417064154e-05, "loss": 0.0124, "step": 46270 }, { "grad_norm": 0.13027046620845795, "learning_rate": 4.1202047525384184e-05, "loss": 0.0141, "step": 46280 }, { "grad_norm": 0.22590190172195435, "learning_rate": 4.118170238345627e-05, "loss": 0.0131, "step": 46290 }, { "grad_norm": 0.21394020318984985, "learning_rate": 4.1161358748334256e-05, "loss": 0.0125, "step": 46300 }, { "grad_norm": 0.2280002236366272, "learning_rate": 4.1141016623494266e-05, "loss": 0.014, "step": 46310 }, { "grad_norm": 0.25948214530944824, "learning_rate": 4.112067601241227e-05, "loss": 0.0122, "step": 46320 }, { "grad_norm": 0.3254707157611847, "learning_rate": 4.110033691856387e-05, "loss": 0.0169, "step": 46330 }, { "grad_norm": 0.3294628858566284, "learning_rate": 4.107999934542451e-05, "loss": 0.0208, "step": 46340 }, { "grad_norm": 0.16228047013282776, "learning_rate": 4.105966329646928e-05, "loss": 0.0166, "step": 46350 }, { "grad_norm": 0.7300127744674683, "learning_rate": 4.103932877517308e-05, "loss": 0.019, "step": 46360 }, { "grad_norm": 0.18055187165737152, "learning_rate": 4.101899578501052e-05, "loss": 0.0129, "step": 46370 }, { "grad_norm": 0.1760624349117279, "learning_rate": 4.099866432945595e-05, "loss": 0.0148, "step": 46380 }, { "grad_norm": 0.7090095281600952, "learning_rate": 4.097833441198346e-05, "loss": 0.0186, "step": 46390 }, { "grad_norm": 0.3086065948009491, "learning_rate": 4.095800603606685e-05, "loss": 0.0159, "step": 46400 }, { "grad_norm": 0.190058633685112, "learning_rate": 4.093767920517975e-05, "loss": 0.013, "step": 46410 }, { "grad_norm": 0.21547172963619232, "learning_rate": 4.091735392279539e-05, "loss": 0.0141, "step": 46420 }, { "grad_norm": 0.269855260848999, "learning_rate": 4.089703019238685e-05, "loss": 0.0127, "step": 46430 }, { "grad_norm": 0.2469334453344345, "learning_rate": 4.0876708017426866e-05, "loss": 0.0121, "step": 46440 }, { "grad_norm": 0.4320361316204071, "learning_rate": 4.085638740138798e-05, "loss": 0.0171, "step": 46450 }, { "grad_norm": 0.24613133072853088, "learning_rate": 4.08360683477424e-05, "loss": 0.0158, "step": 46460 }, { "grad_norm": 0.24648268520832062, "learning_rate": 4.0815750859962085e-05, "loss": 0.0137, "step": 46470 }, { "grad_norm": 0.17508740723133087, "learning_rate": 4.079543494151879e-05, "loss": 0.0203, "step": 46480 }, { "grad_norm": 0.20672887563705444, "learning_rate": 4.07751205958839e-05, "loss": 0.0136, "step": 46490 }, { "grad_norm": 0.2242881953716278, "learning_rate": 4.0754807826528615e-05, "loss": 0.0102, "step": 46500 }, { "grad_norm": 0.1498345285654068, "learning_rate": 4.073449663692379e-05, "loss": 0.0152, "step": 46510 }, { "grad_norm": 0.20081005990505219, "learning_rate": 4.07141870305401e-05, "loss": 0.0107, "step": 46520 }, { "grad_norm": 0.32566359639167786, "learning_rate": 4.0693879010847866e-05, "loss": 0.021, "step": 46530 }, { "grad_norm": 0.3466605246067047, "learning_rate": 4.06735725813172e-05, "loss": 0.013, "step": 46540 }, { "grad_norm": 0.21337640285491943, "learning_rate": 4.0653267745417903e-05, "loss": 0.0103, "step": 46550 }, { "grad_norm": 0.14155155420303345, "learning_rate": 4.063296450661949e-05, "loss": 0.0135, "step": 46560 }, { "grad_norm": 0.12865231931209564, "learning_rate": 4.061266286839128e-05, "loss": 0.0128, "step": 46570 }, { "grad_norm": 0.2717965245246887, "learning_rate": 4.0592362834202225e-05, "loss": 0.0162, "step": 46580 }, { "grad_norm": 0.3148692846298218, "learning_rate": 4.057206440752107e-05, "loss": 0.013, "step": 46590 }, { "grad_norm": 0.19184158742427826, "learning_rate": 4.0551767591816245e-05, "loss": 0.0126, "step": 46600 }, { "grad_norm": 0.15708136558532715, "learning_rate": 4.0531472390555935e-05, "loss": 0.0165, "step": 46610 }, { "grad_norm": 0.20935463905334473, "learning_rate": 4.051117880720802e-05, "loss": 0.0139, "step": 46620 }, { "grad_norm": 0.1826874166727066, "learning_rate": 4.049088684524015e-05, "loss": 0.0121, "step": 46630 }, { "grad_norm": 0.18430225551128387, "learning_rate": 4.0470596508119636e-05, "loss": 0.0105, "step": 46640 }, { "grad_norm": 0.18551886081695557, "learning_rate": 4.0450307799313524e-05, "loss": 0.0144, "step": 46650 }, { "grad_norm": 0.2994365096092224, "learning_rate": 4.0430020722288656e-05, "loss": 0.0161, "step": 46660 }, { "grad_norm": 0.33318740129470825, "learning_rate": 4.040973528051148e-05, "loss": 0.0154, "step": 46670 }, { "grad_norm": 0.3078770935535431, "learning_rate": 4.038945147744827e-05, "loss": 0.015, "step": 46680 }, { "grad_norm": 0.160889133810997, "learning_rate": 4.0369169316564945e-05, "loss": 0.0126, "step": 46690 }, { "grad_norm": 1.2044885158538818, "learning_rate": 4.034888880132718e-05, "loss": 0.0185, "step": 46700 }, { "grad_norm": 0.37138065695762634, "learning_rate": 4.032860993520035e-05, "loss": 0.0134, "step": 46710 }, { "grad_norm": 0.18032272160053253, "learning_rate": 4.030833272164959e-05, "loss": 0.0129, "step": 46720 }, { "grad_norm": 0.2084587961435318, "learning_rate": 4.0288057164139705e-05, "loss": 0.0168, "step": 46730 }, { "grad_norm": 0.1822344809770584, "learning_rate": 4.0267783266135205e-05, "loss": 0.0102, "step": 46740 }, { "grad_norm": 0.12473985552787781, "learning_rate": 4.024751103110039e-05, "loss": 0.0143, "step": 46750 }, { "grad_norm": 0.19618777930736542, "learning_rate": 4.0227240462499176e-05, "loss": 0.0142, "step": 46760 }, { "grad_norm": 0.14226724207401276, "learning_rate": 4.020697156379531e-05, "loss": 0.0126, "step": 46770 }, { "grad_norm": 0.2690681517124176, "learning_rate": 4.018670433845215e-05, "loss": 0.0212, "step": 46780 }, { "grad_norm": 0.17244240641593933, "learning_rate": 4.016643878993284e-05, "loss": 0.0109, "step": 46790 }, { "grad_norm": 0.24640774726867676, "learning_rate": 4.014617492170017e-05, "loss": 0.0148, "step": 46800 }, { "grad_norm": 0.22227799892425537, "learning_rate": 4.0125912737216726e-05, "loss": 0.0242, "step": 46810 }, { "grad_norm": 0.14882807433605194, "learning_rate": 4.0105652239944735e-05, "loss": 0.0152, "step": 46820 }, { "grad_norm": 0.13833189010620117, "learning_rate": 4.0085393433346144e-05, "loss": 0.0127, "step": 46830 }, { "grad_norm": 0.1509796679019928, "learning_rate": 4.006513632088268e-05, "loss": 0.0111, "step": 46840 }, { "grad_norm": 0.16270636022090912, "learning_rate": 4.004488090601567e-05, "loss": 0.01, "step": 46850 }, { "grad_norm": 0.2181973159313202, "learning_rate": 4.002462719220626e-05, "loss": 0.0115, "step": 46860 }, { "grad_norm": 0.1389290988445282, "learning_rate": 4.000437518291522e-05, "loss": 0.0142, "step": 46870 }, { "grad_norm": 0.12590008974075317, "learning_rate": 3.9984124881603094e-05, "loss": 0.0118, "step": 46880 }, { "grad_norm": 0.1914697140455246, "learning_rate": 3.9963876291730086e-05, "loss": 0.0126, "step": 46890 }, { "grad_norm": 0.2068258821964264, "learning_rate": 3.994362941675614e-05, "loss": 0.0146, "step": 46900 }, { "grad_norm": 0.20370328426361084, "learning_rate": 3.992338426014088e-05, "loss": 0.0162, "step": 46910 }, { "grad_norm": 0.49012064933776855, "learning_rate": 3.9903140825343636e-05, "loss": 0.0153, "step": 46920 }, { "grad_norm": 0.18064309656620026, "learning_rate": 3.98828991158235e-05, "loss": 0.0144, "step": 46930 }, { "grad_norm": 0.24249622225761414, "learning_rate": 3.9862659135039185e-05, "loss": 0.0137, "step": 46940 }, { "grad_norm": 0.3178917467594147, "learning_rate": 3.984242088644918e-05, "loss": 0.0163, "step": 46950 }, { "grad_norm": 0.2177097350358963, "learning_rate": 3.9822184373511615e-05, "loss": 0.0116, "step": 46960 }, { "grad_norm": 0.13339117169380188, "learning_rate": 3.98019495996844e-05, "loss": 0.0139, "step": 46970 }, { "grad_norm": 0.17196311056613922, "learning_rate": 3.978171656842507e-05, "loss": 0.0146, "step": 46980 }, { "grad_norm": 0.22863546013832092, "learning_rate": 3.976148528319091e-05, "loss": 0.0133, "step": 46990 }, { "grad_norm": 0.47603729367256165, "learning_rate": 3.974125574743888e-05, "loss": 0.0138, "step": 47000 }, { "grad_norm": 0.2814715504646301, "learning_rate": 3.9721027964625686e-05, "loss": 0.0176, "step": 47010 }, { "grad_norm": 0.29446542263031006, "learning_rate": 3.9700801938207676e-05, "loss": 0.0131, "step": 47020 }, { "grad_norm": 0.17760322988033295, "learning_rate": 3.9680577671640916e-05, "loss": 0.0134, "step": 47030 }, { "grad_norm": 0.204605370759964, "learning_rate": 3.966035516838121e-05, "loss": 0.0202, "step": 47040 }, { "grad_norm": 0.3242255449295044, "learning_rate": 3.9640134431884014e-05, "loss": 0.0129, "step": 47050 }, { "grad_norm": 0.4020192623138428, "learning_rate": 3.961991546560451e-05, "loss": 0.0116, "step": 47060 }, { "grad_norm": 0.1744835525751114, "learning_rate": 3.959969827299753e-05, "loss": 0.014, "step": 47070 }, { "grad_norm": 0.24463410675525665, "learning_rate": 3.9579482857517684e-05, "loss": 0.0141, "step": 47080 }, { "grad_norm": 0.2690531015396118, "learning_rate": 3.955926922261921e-05, "loss": 0.0118, "step": 47090 }, { "grad_norm": 0.5162926912307739, "learning_rate": 3.9539057371756084e-05, "loss": 0.0162, "step": 47100 }, { "grad_norm": 0.27904415130615234, "learning_rate": 3.951884730838195e-05, "loss": 0.0091, "step": 47110 }, { "grad_norm": 0.2666524052619934, "learning_rate": 3.949863903595012e-05, "loss": 0.0101, "step": 47120 }, { "grad_norm": 0.40620487928390503, "learning_rate": 3.947843255791369e-05, "loss": 0.0126, "step": 47130 }, { "grad_norm": 0.3819906711578369, "learning_rate": 3.9458227877725364e-05, "loss": 0.0155, "step": 47140 }, { "grad_norm": 0.3419840931892395, "learning_rate": 3.943802499883758e-05, "loss": 0.0092, "step": 47150 }, { "grad_norm": 0.28916290402412415, "learning_rate": 3.9417823924702437e-05, "loss": 0.0197, "step": 47160 }, { "grad_norm": 0.45545831322669983, "learning_rate": 3.939762465877178e-05, "loss": 0.0158, "step": 47170 }, { "grad_norm": 0.29654455184936523, "learning_rate": 3.937742720449708e-05, "loss": 0.0155, "step": 47180 }, { "grad_norm": 0.2884349822998047, "learning_rate": 3.9357231565329563e-05, "loss": 0.0124, "step": 47190 }, { "grad_norm": 0.16459283232688904, "learning_rate": 3.933703774472008e-05, "loss": 0.0173, "step": 47200 }, { "grad_norm": 0.3452872037887573, "learning_rate": 3.93168457461192e-05, "loss": 0.0101, "step": 47210 }, { "grad_norm": 0.19553984701633453, "learning_rate": 3.9296655572977216e-05, "loss": 0.013, "step": 47220 }, { "grad_norm": 0.28747087717056274, "learning_rate": 3.927646722874404e-05, "loss": 0.016, "step": 47230 }, { "grad_norm": 0.1612507700920105, "learning_rate": 3.925628071686934e-05, "loss": 0.0129, "step": 47240 }, { "grad_norm": 0.17603561282157898, "learning_rate": 3.9236096040802415e-05, "loss": 0.0146, "step": 47250 }, { "grad_norm": 0.36534491181373596, "learning_rate": 3.9215913203992294e-05, "loss": 0.0197, "step": 47260 }, { "grad_norm": 0.18278631567955017, "learning_rate": 3.9195732209887645e-05, "loss": 0.0119, "step": 47270 }, { "grad_norm": 0.18376219272613525, "learning_rate": 3.9175553061936875e-05, "loss": 0.0145, "step": 47280 }, { "grad_norm": 0.38106533885002136, "learning_rate": 3.9155375763588045e-05, "loss": 0.0196, "step": 47290 }, { "grad_norm": 0.1861959844827652, "learning_rate": 3.913520031828889e-05, "loss": 0.0119, "step": 47300 }, { "grad_norm": 0.22012895345687866, "learning_rate": 3.911502672948685e-05, "loss": 0.0143, "step": 47310 }, { "grad_norm": 0.17252863943576813, "learning_rate": 3.9094855000629014e-05, "loss": 0.0145, "step": 47320 }, { "grad_norm": 0.19311818480491638, "learning_rate": 3.907468513516223e-05, "loss": 0.0124, "step": 47330 }, { "grad_norm": 0.1838878095149994, "learning_rate": 3.905451713653294e-05, "loss": 0.0137, "step": 47340 }, { "grad_norm": 0.21259960532188416, "learning_rate": 3.903435100818731e-05, "loss": 0.0125, "step": 47350 }, { "grad_norm": 0.23361395299434662, "learning_rate": 3.901418675357117e-05, "loss": 0.0145, "step": 47360 }, { "grad_norm": 0.13601166009902954, "learning_rate": 3.8994024376130075e-05, "loss": 0.0178, "step": 47370 }, { "grad_norm": 0.1846788078546524, "learning_rate": 3.8973863879309194e-05, "loss": 0.0136, "step": 47380 }, { "grad_norm": 0.3735794126987457, "learning_rate": 3.8953705266553394e-05, "loss": 0.017, "step": 47390 }, { "grad_norm": 0.2963937222957611, "learning_rate": 3.893354854130727e-05, "loss": 0.0197, "step": 47400 }, { "grad_norm": 0.23247885704040527, "learning_rate": 3.8913393707015006e-05, "loss": 0.0131, "step": 47410 }, { "grad_norm": 0.22320912778377533, "learning_rate": 3.889324076712056e-05, "loss": 0.0169, "step": 47420 }, { "grad_norm": 0.1307169795036316, "learning_rate": 3.8873089725067476e-05, "loss": 0.0104, "step": 47430 }, { "grad_norm": 0.2120485156774521, "learning_rate": 3.885294058429905e-05, "loss": 0.0196, "step": 47440 }, { "grad_norm": 0.10827440023422241, "learning_rate": 3.8832793348258206e-05, "loss": 0.0121, "step": 47450 }, { "grad_norm": 0.2558148205280304, "learning_rate": 3.881264802038756e-05, "loss": 0.0227, "step": 47460 }, { "grad_norm": 0.2483053207397461, "learning_rate": 3.879250460412939e-05, "loss": 0.017, "step": 47470 }, { "grad_norm": 0.16849125921726227, "learning_rate": 3.8772363102925644e-05, "loss": 0.0133, "step": 47480 }, { "grad_norm": 0.18664292991161346, "learning_rate": 3.875222352021798e-05, "loss": 0.0113, "step": 47490 }, { "grad_norm": 0.16046392917633057, "learning_rate": 3.8732085859447686e-05, "loss": 0.015, "step": 47500 }, { "grad_norm": 0.22038546204566956, "learning_rate": 3.871195012405575e-05, "loss": 0.0128, "step": 47510 }, { "grad_norm": 0.19903036952018738, "learning_rate": 3.869181631748278e-05, "loss": 0.0158, "step": 47520 }, { "grad_norm": 0.6853519678115845, "learning_rate": 3.867168444316915e-05, "loss": 0.013, "step": 47530 }, { "grad_norm": 0.15684445202350616, "learning_rate": 3.865155450455481e-05, "loss": 0.0127, "step": 47540 }, { "grad_norm": 0.21464912593364716, "learning_rate": 3.8631426505079426e-05, "loss": 0.0188, "step": 47550 }, { "grad_norm": 0.3415186107158661, "learning_rate": 3.8611300448182304e-05, "loss": 0.0165, "step": 47560 }, { "grad_norm": 0.16390439867973328, "learning_rate": 3.859117633730248e-05, "loss": 0.0122, "step": 47570 }, { "grad_norm": 0.30140718817710876, "learning_rate": 3.857105417587858e-05, "loss": 0.0142, "step": 47580 }, { "grad_norm": 0.213969424366951, "learning_rate": 3.855093396734894e-05, "loss": 0.0119, "step": 47590 }, { "grad_norm": 0.15301570296287537, "learning_rate": 3.8530815715151545e-05, "loss": 0.0232, "step": 47600 }, { "grad_norm": 0.17777018249034882, "learning_rate": 3.851069942272405e-05, "loss": 0.0106, "step": 47610 }, { "grad_norm": 0.23689766228199005, "learning_rate": 3.849058509350382e-05, "loss": 0.011, "step": 47620 }, { "grad_norm": 0.1917024403810501, "learning_rate": 3.8470472730927783e-05, "loss": 0.0166, "step": 47630 }, { "grad_norm": 0.3407021462917328, "learning_rate": 3.845036233843264e-05, "loss": 0.0135, "step": 47640 }, { "grad_norm": 0.15328460931777954, "learning_rate": 3.843025391945469e-05, "loss": 0.0106, "step": 47650 }, { "grad_norm": 0.16173148155212402, "learning_rate": 3.841014747742992e-05, "loss": 0.0103, "step": 47660 }, { "grad_norm": 0.20708800852298737, "learning_rate": 3.839004301579397e-05, "loss": 0.011, "step": 47670 }, { "grad_norm": 0.23710133135318756, "learning_rate": 3.8369940537982097e-05, "loss": 0.0141, "step": 47680 }, { "grad_norm": 0.11058596521615982, "learning_rate": 3.834984004742933e-05, "loss": 0.0095, "step": 47690 }, { "grad_norm": 0.20055684447288513, "learning_rate": 3.832974154757026e-05, "loss": 0.0131, "step": 47700 }, { "grad_norm": 0.19871263206005096, "learning_rate": 3.830964504183919e-05, "loss": 0.0117, "step": 47710 }, { "grad_norm": 0.1665843278169632, "learning_rate": 3.828955053367003e-05, "loss": 0.012, "step": 47720 }, { "grad_norm": 0.3645382523536682, "learning_rate": 3.826945802649642e-05, "loss": 0.0152, "step": 47730 }, { "grad_norm": 0.17486530542373657, "learning_rate": 3.824936752375159e-05, "loss": 0.019, "step": 47740 }, { "grad_norm": 0.21232303977012634, "learning_rate": 3.822927902886848e-05, "loss": 0.0143, "step": 47750 }, { "grad_norm": 0.12490136921405792, "learning_rate": 3.8209192545279653e-05, "loss": 0.0116, "step": 47760 }, { "grad_norm": 0.12380499392747879, "learning_rate": 3.8189108076417326e-05, "loss": 0.0122, "step": 47770 }, { "grad_norm": 0.1897064596414566, "learning_rate": 3.816902562571342e-05, "loss": 0.022, "step": 47780 }, { "grad_norm": 0.29201605916023254, "learning_rate": 3.814894519659944e-05, "loss": 0.0136, "step": 47790 }, { "grad_norm": 0.2838258147239685, "learning_rate": 3.812886679250661e-05, "loss": 0.0144, "step": 47800 }, { "grad_norm": 0.15027254819869995, "learning_rate": 3.810879041686575e-05, "loss": 0.0139, "step": 47810 }, { "grad_norm": 0.36696821451187134, "learning_rate": 3.808871607310741e-05, "loss": 0.0202, "step": 47820 }, { "grad_norm": 0.2573253810405731, "learning_rate": 3.806864376466169e-05, "loss": 0.0153, "step": 47830 }, { "grad_norm": 0.23001189529895782, "learning_rate": 3.804857349495845e-05, "loss": 0.0131, "step": 47840 }, { "grad_norm": 0.32217323780059814, "learning_rate": 3.802850526742713e-05, "loss": 0.0293, "step": 47850 }, { "grad_norm": 0.1731930375099182, "learning_rate": 3.800843908549683e-05, "loss": 0.0097, "step": 47860 }, { "grad_norm": 0.16308967769145966, "learning_rate": 3.7988374952596325e-05, "loss": 0.0099, "step": 47870 }, { "grad_norm": 0.16724763810634613, "learning_rate": 3.7968312872154e-05, "loss": 0.0116, "step": 47880 }, { "grad_norm": 0.32844921946525574, "learning_rate": 3.7948252847597965e-05, "loss": 0.0117, "step": 47890 }, { "grad_norm": 0.18898309767246246, "learning_rate": 3.7928194882355885e-05, "loss": 0.0139, "step": 47900 }, { "grad_norm": 0.1693504899740219, "learning_rate": 3.790813897985515e-05, "loss": 0.0107, "step": 47910 }, { "grad_norm": 0.42022010684013367, "learning_rate": 3.7888085143522726e-05, "loss": 0.0166, "step": 47920 }, { "grad_norm": 0.2629298269748688, "learning_rate": 3.7868033376785314e-05, "loss": 0.0122, "step": 47930 }, { "grad_norm": 0.32185161113739014, "learning_rate": 3.784798368306919e-05, "loss": 0.0144, "step": 47940 }, { "grad_norm": 0.23059864342212677, "learning_rate": 3.782793606580029e-05, "loss": 0.0095, "step": 47950 }, { "grad_norm": 0.11846352368593216, "learning_rate": 3.7807890528404205e-05, "loss": 0.0151, "step": 47960 }, { "grad_norm": 0.17610736191272736, "learning_rate": 3.778784707430616e-05, "loss": 0.0139, "step": 47970 }, { "grad_norm": 0.17968086898326874, "learning_rate": 3.776780570693107e-05, "loss": 0.0096, "step": 47980 }, { "grad_norm": 0.2212066948413849, "learning_rate": 3.774776642970342e-05, "loss": 0.0171, "step": 47990 }, { "grad_norm": 0.15509916841983795, "learning_rate": 3.77277292460474e-05, "loss": 0.0134, "step": 48000 }, { "grad_norm": 0.17947930097579956, "learning_rate": 3.770769415938678e-05, "loss": 0.0106, "step": 48010 }, { "grad_norm": 0.17010073363780975, "learning_rate": 3.768766117314506e-05, "loss": 0.0165, "step": 48020 }, { "grad_norm": 0.22926540672779083, "learning_rate": 3.76676302907453e-05, "loss": 0.0137, "step": 48030 }, { "grad_norm": 0.2527560293674469, "learning_rate": 3.764760151561021e-05, "loss": 0.0111, "step": 48040 }, { "grad_norm": 0.18108122050762177, "learning_rate": 3.76275748511622e-05, "loss": 0.0105, "step": 48050 }, { "grad_norm": 0.3731929361820221, "learning_rate": 3.7607550300823255e-05, "loss": 0.0135, "step": 48060 }, { "grad_norm": 0.2580423951148987, "learning_rate": 3.7587527868015044e-05, "loss": 0.0225, "step": 48070 }, { "grad_norm": 0.2633064091205597, "learning_rate": 3.756750755615881e-05, "loss": 0.0154, "step": 48080 }, { "grad_norm": 0.12633965909481049, "learning_rate": 3.754748936867553e-05, "loss": 0.0133, "step": 48090 }, { "grad_norm": 0.21305660903453827, "learning_rate": 3.752747330898573e-05, "loss": 0.0116, "step": 48100 }, { "grad_norm": 0.2606160640716553, "learning_rate": 3.750745938050962e-05, "loss": 0.0126, "step": 48110 }, { "grad_norm": 0.32138368487358093, "learning_rate": 3.7487447586667025e-05, "loss": 0.0139, "step": 48120 }, { "grad_norm": 0.15870262682437897, "learning_rate": 3.7467437930877426e-05, "loss": 0.0117, "step": 48130 }, { "grad_norm": 0.2003011256456375, "learning_rate": 3.744743041655992e-05, "loss": 0.014, "step": 48140 }, { "grad_norm": 0.24057745933532715, "learning_rate": 3.742742504713324e-05, "loss": 0.0111, "step": 48150 }, { "grad_norm": 0.11869821697473526, "learning_rate": 3.740742182601576e-05, "loss": 0.0116, "step": 48160 }, { "grad_norm": 0.17808260023593903, "learning_rate": 3.7387420756625464e-05, "loss": 0.0148, "step": 48170 }, { "grad_norm": 0.19996698200702667, "learning_rate": 3.736742184238002e-05, "loss": 0.014, "step": 48180 }, { "grad_norm": 0.12294289469718933, "learning_rate": 3.7347425086696684e-05, "loss": 0.0137, "step": 48190 }, { "grad_norm": 0.31536149978637695, "learning_rate": 3.732743049299235e-05, "loss": 0.0137, "step": 48200 }, { "grad_norm": 0.16346606612205505, "learning_rate": 3.730743806468354e-05, "loss": 0.0149, "step": 48210 }, { "grad_norm": 0.1590852290391922, "learning_rate": 3.7287447805186436e-05, "loss": 0.0115, "step": 48220 }, { "grad_norm": 0.1976201832294464, "learning_rate": 3.726745971791682e-05, "loss": 0.0154, "step": 48230 }, { "grad_norm": 0.30632638931274414, "learning_rate": 3.724747380629008e-05, "loss": 0.022, "step": 48240 }, { "grad_norm": 0.13384120166301727, "learning_rate": 3.72274900737213e-05, "loss": 0.0175, "step": 48250 }, { "grad_norm": 0.17811249196529388, "learning_rate": 3.7207508523625123e-05, "loss": 0.0116, "step": 48260 }, { "grad_norm": 0.30729952454566956, "learning_rate": 3.718752915941588e-05, "loss": 0.0139, "step": 48270 }, { "grad_norm": 0.2997639775276184, "learning_rate": 3.7167551984507464e-05, "loss": 0.0115, "step": 48280 }, { "grad_norm": 0.23609384894371033, "learning_rate": 3.714757700231346e-05, "loss": 0.014, "step": 48290 }, { "grad_norm": 0.13057507574558258, "learning_rate": 3.712760421624703e-05, "loss": 0.0114, "step": 48300 }, { "grad_norm": 0.1728297919034958, "learning_rate": 3.710763362972099e-05, "loss": 0.0134, "step": 48310 }, { "grad_norm": 0.20788222551345825, "learning_rate": 3.708766524614774e-05, "loss": 0.0179, "step": 48320 }, { "grad_norm": 0.18733681738376617, "learning_rate": 3.7067699068939335e-05, "loss": 0.0119, "step": 48330 }, { "grad_norm": 0.1483364850282669, "learning_rate": 3.704773510150748e-05, "loss": 0.0137, "step": 48340 }, { "grad_norm": 0.2773510813713074, "learning_rate": 3.702777334726344e-05, "loss": 0.0142, "step": 48350 }, { "grad_norm": 0.23586714267730713, "learning_rate": 3.7007813809618164e-05, "loss": 0.0103, "step": 48360 }, { "grad_norm": 0.27529987692832947, "learning_rate": 3.6987856491982145e-05, "loss": 0.0159, "step": 48370 }, { "grad_norm": 0.21012724936008453, "learning_rate": 3.696790139776558e-05, "loss": 0.0154, "step": 48380 }, { "grad_norm": 0.2003803700208664, "learning_rate": 3.6947948530378235e-05, "loss": 0.0156, "step": 48390 }, { "grad_norm": 0.14361540973186493, "learning_rate": 3.6927997893229516e-05, "loss": 0.0131, "step": 48400 }, { "grad_norm": 0.1340837925672531, "learning_rate": 3.690804948972845e-05, "loss": 0.0109, "step": 48410 }, { "grad_norm": 0.18623696267604828, "learning_rate": 3.688810332328363e-05, "loss": 0.0126, "step": 48420 }, { "grad_norm": 0.2569711208343506, "learning_rate": 3.686815939730336e-05, "loss": 0.0134, "step": 48430 }, { "grad_norm": 0.21835607290267944, "learning_rate": 3.684821771519548e-05, "loss": 0.0137, "step": 48440 }, { "grad_norm": 0.5891335606575012, "learning_rate": 3.68282782803675e-05, "loss": 0.0148, "step": 48450 }, { "grad_norm": 0.552564263343811, "learning_rate": 3.6808341096226504e-05, "loss": 0.0192, "step": 48460 }, { "grad_norm": 0.36520621180534363, "learning_rate": 3.678840616617924e-05, "loss": 0.0156, "step": 48470 }, { "grad_norm": 0.2704387605190277, "learning_rate": 3.6768473493632e-05, "loss": 0.0156, "step": 48480 }, { "grad_norm": 0.15984074771404266, "learning_rate": 3.6748543081990783e-05, "loss": 0.0151, "step": 48490 }, { "grad_norm": 0.19269423186779022, "learning_rate": 3.672861493466112e-05, "loss": 0.0112, "step": 48500 }, { "grad_norm": 0.26116758584976196, "learning_rate": 3.670868905504818e-05, "loss": 0.0135, "step": 48510 }, { "grad_norm": 0.3392460346221924, "learning_rate": 3.6688765446556784e-05, "loss": 0.0114, "step": 48520 }, { "grad_norm": 0.2901696264743805, "learning_rate": 3.6668844112591276e-05, "loss": 0.0133, "step": 48530 }, { "grad_norm": 0.31664708256721497, "learning_rate": 3.664892505655573e-05, "loss": 0.0137, "step": 48540 }, { "grad_norm": 0.258571594953537, "learning_rate": 3.662900828185373e-05, "loss": 0.0179, "step": 48550 }, { "grad_norm": 0.18263408541679382, "learning_rate": 3.6609093791888516e-05, "loss": 0.0124, "step": 48560 }, { "grad_norm": 0.33407774567604065, "learning_rate": 3.658918159006292e-05, "loss": 0.0192, "step": 48570 }, { "grad_norm": 0.26629602909088135, "learning_rate": 3.656927167977942e-05, "loss": 0.017, "step": 48580 }, { "grad_norm": 0.17586471140384674, "learning_rate": 3.654936406444006e-05, "loss": 0.0141, "step": 48590 }, { "grad_norm": 0.1949443519115448, "learning_rate": 3.65294587474465e-05, "loss": 0.01, "step": 48600 }, { "grad_norm": 0.18039566278457642, "learning_rate": 3.650955573220002e-05, "loss": 0.0392, "step": 48610 }, { "grad_norm": 0.21981260180473328, "learning_rate": 3.648965502210149e-05, "loss": 0.0103, "step": 48620 }, { "grad_norm": 0.2548348903656006, "learning_rate": 3.646975662055142e-05, "loss": 0.013, "step": 48630 }, { "grad_norm": 0.192405104637146, "learning_rate": 3.644986053094987e-05, "loss": 0.0119, "step": 48640 }, { "grad_norm": 0.24932323396205902, "learning_rate": 3.642996675669659e-05, "loss": 0.0138, "step": 48650 }, { "grad_norm": 0.15600326657295227, "learning_rate": 3.641007530119083e-05, "loss": 0.0093, "step": 48660 }, { "grad_norm": 0.2042674422264099, "learning_rate": 3.639018616783153e-05, "loss": 0.0171, "step": 48670 }, { "grad_norm": 0.19372043013572693, "learning_rate": 3.637029936001719e-05, "loss": 0.0114, "step": 48680 }, { "grad_norm": 0.19188924133777618, "learning_rate": 3.6350414881145886e-05, "loss": 0.0157, "step": 48690 }, { "grad_norm": 0.1322959065437317, "learning_rate": 3.6330532734615386e-05, "loss": 0.015, "step": 48700 }, { "grad_norm": 0.15531431138515472, "learning_rate": 3.6310652923822975e-05, "loss": 0.0143, "step": 48710 }, { "grad_norm": 0.21641187369823456, "learning_rate": 3.629077545216558e-05, "loss": 0.0151, "step": 48720 }, { "grad_norm": 0.27478310465812683, "learning_rate": 3.627090032303969e-05, "loss": 0.0175, "step": 48730 }, { "grad_norm": 0.14077745378017426, "learning_rate": 3.625102753984146e-05, "loss": 0.0141, "step": 48740 }, { "grad_norm": 0.1927359700202942, "learning_rate": 3.623115710596659e-05, "loss": 0.0112, "step": 48750 }, { "grad_norm": 0.28637221455574036, "learning_rate": 3.6211289024810395e-05, "loss": 0.0158, "step": 48760 }, { "grad_norm": 0.157927468419075, "learning_rate": 3.619142329976777e-05, "loss": 0.0115, "step": 48770 }, { "grad_norm": 0.20783063769340515, "learning_rate": 3.6171559934233247e-05, "loss": 0.0121, "step": 48780 }, { "grad_norm": 0.14028604328632355, "learning_rate": 3.615169893160093e-05, "loss": 0.0117, "step": 48790 }, { "grad_norm": 0.22088052332401276, "learning_rate": 3.61318402952645e-05, "loss": 0.0109, "step": 48800 }, { "grad_norm": 0.27753746509552, "learning_rate": 3.6111984028617285e-05, "loss": 0.014, "step": 48810 }, { "grad_norm": 0.15080060064792633, "learning_rate": 3.6092130135052134e-05, "loss": 0.0141, "step": 48820 }, { "grad_norm": 0.17896172404289246, "learning_rate": 3.6072278617961584e-05, "loss": 0.018, "step": 48830 }, { "grad_norm": 0.2849087417125702, "learning_rate": 3.605242948073767e-05, "loss": 0.0177, "step": 48840 }, { "grad_norm": 0.1826014518737793, "learning_rate": 3.603258272677212e-05, "loss": 0.0161, "step": 48850 }, { "grad_norm": 0.20694565773010254, "learning_rate": 3.601273835945616e-05, "loss": 0.0138, "step": 48860 }, { "grad_norm": 0.20149767398834229, "learning_rate": 3.5992896382180664e-05, "loss": 0.0116, "step": 48870 }, { "grad_norm": 0.07950137555599213, "learning_rate": 3.597305679833609e-05, "loss": 0.0183, "step": 48880 }, { "grad_norm": 0.3038271963596344, "learning_rate": 3.595321961131245e-05, "loss": 0.016, "step": 48890 }, { "grad_norm": 0.29804539680480957, "learning_rate": 3.593338482449942e-05, "loss": 0.0178, "step": 48900 }, { "grad_norm": 0.24198967218399048, "learning_rate": 3.591355244128618e-05, "loss": 0.0199, "step": 48910 }, { "grad_norm": 0.20252883434295654, "learning_rate": 3.589372246506158e-05, "loss": 0.0189, "step": 48920 }, { "grad_norm": 0.17720966041088104, "learning_rate": 3.5873894899213984e-05, "loss": 0.0086, "step": 48930 }, { "grad_norm": 0.221820667386055, "learning_rate": 3.5854069747131416e-05, "loss": 0.0129, "step": 48940 }, { "grad_norm": 0.2581109404563904, "learning_rate": 3.583424701220143e-05, "loss": 0.017, "step": 48950 }, { "grad_norm": 0.25446686148643494, "learning_rate": 3.581442669781121e-05, "loss": 0.0156, "step": 48960 }, { "grad_norm": 0.18530991673469543, "learning_rate": 3.579460880734749e-05, "loss": 0.0181, "step": 48970 }, { "grad_norm": 0.17514953017234802, "learning_rate": 3.577479334419657e-05, "loss": 0.0185, "step": 48980 }, { "grad_norm": 0.2187737673521042, "learning_rate": 3.575498031174444e-05, "loss": 0.0116, "step": 48990 }, { "grad_norm": 0.19534297287464142, "learning_rate": 3.573516971337657e-05, "loss": 0.0147, "step": 49000 }, { "grad_norm": 0.3309464156627655, "learning_rate": 3.5715361552478046e-05, "loss": 0.0183, "step": 49010 }, { "grad_norm": 0.30239537358283997, "learning_rate": 3.5695555832433536e-05, "loss": 0.0135, "step": 49020 }, { "grad_norm": 0.29330962896347046, "learning_rate": 3.5675752556627325e-05, "loss": 0.0127, "step": 49030 }, { "grad_norm": 0.2654162645339966, "learning_rate": 3.565595172844322e-05, "loss": 0.0117, "step": 49040 }, { "grad_norm": 0.32838237285614014, "learning_rate": 3.5636153351264666e-05, "loss": 0.0135, "step": 49050 }, { "grad_norm": 0.25447142124176025, "learning_rate": 3.5616357428474655e-05, "loss": 0.0133, "step": 49060 }, { "grad_norm": 0.17820699512958527, "learning_rate": 3.559656396345575e-05, "loss": 0.0161, "step": 49070 }, { "grad_norm": 0.17808762192726135, "learning_rate": 3.5576772959590146e-05, "loss": 0.0177, "step": 49080 }, { "grad_norm": 0.16160637140274048, "learning_rate": 3.5556984420259545e-05, "loss": 0.0123, "step": 49090 }, { "grad_norm": 0.18662205338478088, "learning_rate": 3.5537198348845305e-05, "loss": 0.0147, "step": 49100 }, { "grad_norm": 0.24623729288578033, "learning_rate": 3.551741474872831e-05, "loss": 0.0157, "step": 49110 }, { "grad_norm": 0.2073897123336792, "learning_rate": 3.549763362328903e-05, "loss": 0.0172, "step": 49120 }, { "grad_norm": 0.2569985091686249, "learning_rate": 3.5477854975907515e-05, "loss": 0.0114, "step": 49130 }, { "grad_norm": 0.20185811817646027, "learning_rate": 3.5458078809963416e-05, "loss": 0.0166, "step": 49140 }, { "grad_norm": 0.20710185170173645, "learning_rate": 3.543830512883594e-05, "loss": 0.0113, "step": 49150 }, { "grad_norm": 0.2193596363067627, "learning_rate": 3.5418533935903824e-05, "loss": 0.0138, "step": 49160 }, { "grad_norm": 0.2759494483470917, "learning_rate": 3.539876523454547e-05, "loss": 0.0137, "step": 49170 }, { "grad_norm": 0.20785385370254517, "learning_rate": 3.537899902813878e-05, "loss": 0.0128, "step": 49180 }, { "grad_norm": 0.18897169828414917, "learning_rate": 3.5359235320061293e-05, "loss": 0.0109, "step": 49190 }, { "grad_norm": 0.19427837431430817, "learning_rate": 3.533947411369003e-05, "loss": 0.0101, "step": 49200 }, { "grad_norm": 0.19614702463150024, "learning_rate": 3.53197154124017e-05, "loss": 0.0123, "step": 49210 }, { "grad_norm": 0.28283289074897766, "learning_rate": 3.52999592195725e-05, "loss": 0.014, "step": 49220 }, { "grad_norm": 0.23431402444839478, "learning_rate": 3.5280205538578224e-05, "loss": 0.0123, "step": 49230 }, { "grad_norm": 0.2949119210243225, "learning_rate": 3.5260454372794236e-05, "loss": 0.0107, "step": 49240 }, { "grad_norm": 0.21920987963676453, "learning_rate": 3.524070572559545e-05, "loss": 0.0136, "step": 49250 }, { "grad_norm": 0.15838943421840668, "learning_rate": 3.5220959600356395e-05, "loss": 0.0113, "step": 49260 }, { "grad_norm": 0.175127312541008, "learning_rate": 3.5201216000451145e-05, "loss": 0.0132, "step": 49270 }, { "grad_norm": 0.22050976753234863, "learning_rate": 3.5181474929253335e-05, "loss": 0.0112, "step": 49280 }, { "grad_norm": 0.14555081725120544, "learning_rate": 3.516173639013615e-05, "loss": 0.0125, "step": 49290 }, { "grad_norm": 0.12013810873031616, "learning_rate": 3.5142000386472406e-05, "loss": 0.0117, "step": 49300 }, { "grad_norm": 0.20064927637577057, "learning_rate": 3.5122266921634427e-05, "loss": 0.0102, "step": 49310 }, { "grad_norm": 0.24735859036445618, "learning_rate": 3.510253599899413e-05, "loss": 0.0131, "step": 49320 }, { "grad_norm": 0.16107489168643951, "learning_rate": 3.5082807621922965e-05, "loss": 0.0112, "step": 49330 }, { "grad_norm": 0.2273879200220108, "learning_rate": 3.506308179379201e-05, "loss": 0.0125, "step": 49340 }, { "grad_norm": 0.15760965645313263, "learning_rate": 3.5043358517971844e-05, "loss": 0.0116, "step": 49350 }, { "grad_norm": 0.16341851651668549, "learning_rate": 3.502363779783264e-05, "loss": 0.0165, "step": 49360 }, { "grad_norm": 0.1812717467546463, "learning_rate": 3.500391963674415e-05, "loss": 0.0122, "step": 49370 }, { "grad_norm": 0.22582779824733734, "learning_rate": 3.4984204038075615e-05, "loss": 0.0194, "step": 49380 }, { "grad_norm": 0.2194373607635498, "learning_rate": 3.496449100519595e-05, "loss": 0.0194, "step": 49390 }, { "grad_norm": 0.23992745578289032, "learning_rate": 3.494478054147354e-05, "loss": 0.0125, "step": 49400 }, { "grad_norm": 0.27345314621925354, "learning_rate": 3.4925072650276395e-05, "loss": 0.0119, "step": 49410 }, { "grad_norm": 0.239719957113266, "learning_rate": 3.4905367334972016e-05, "loss": 0.0139, "step": 49420 }, { "grad_norm": 0.3086252510547638, "learning_rate": 3.488566459892752e-05, "loss": 0.0156, "step": 49430 }, { "grad_norm": 0.2399258315563202, "learning_rate": 3.4865964445509585e-05, "loss": 0.011, "step": 49440 }, { "grad_norm": 0.21098652482032776, "learning_rate": 3.484626687808438e-05, "loss": 0.0124, "step": 49450 }, { "grad_norm": 0.1863051950931549, "learning_rate": 3.4826571900017735e-05, "loss": 0.0086, "step": 49460 }, { "grad_norm": 0.1690545231103897, "learning_rate": 3.480687951467495e-05, "loss": 0.012, "step": 49470 }, { "grad_norm": 0.09860427677631378, "learning_rate": 3.4787189725420925e-05, "loss": 0.0117, "step": 49480 }, { "grad_norm": 0.15689876675605774, "learning_rate": 3.4767502535620086e-05, "loss": 0.011, "step": 49490 }, { "grad_norm": 0.20153765380382538, "learning_rate": 3.474781794863648e-05, "loss": 0.0153, "step": 49500 }, { "grad_norm": 0.2207801789045334, "learning_rate": 3.472813596783363e-05, "loss": 0.0183, "step": 49510 }, { "grad_norm": 0.2022220939397812, "learning_rate": 3.470845659657466e-05, "loss": 0.0148, "step": 49520 }, { "grad_norm": 0.1483813226222992, "learning_rate": 3.468877983822223e-05, "loss": 0.0096, "step": 49530 }, { "grad_norm": 0.16239984333515167, "learning_rate": 3.466910569613855e-05, "loss": 0.0113, "step": 49540 }, { "grad_norm": 0.3809661269187927, "learning_rate": 3.464943417368542e-05, "loss": 0.0138, "step": 49550 }, { "grad_norm": 0.19684311747550964, "learning_rate": 3.462976527422415e-05, "loss": 0.0123, "step": 49560 }, { "grad_norm": 0.26630130410194397, "learning_rate": 3.461009900111562e-05, "loss": 0.0125, "step": 49570 }, { "grad_norm": 0.2261451780796051, "learning_rate": 3.459043535772023e-05, "loss": 0.013, "step": 49580 }, { "grad_norm": 0.15672451257705688, "learning_rate": 3.4570774347398014e-05, "loss": 0.014, "step": 49590 }, { "grad_norm": 0.2939058840274811, "learning_rate": 3.4551115973508454e-05, "loss": 0.0161, "step": 49600 }, { "grad_norm": 0.2839631736278534, "learning_rate": 3.453146023941066e-05, "loss": 0.02, "step": 49610 }, { "grad_norm": 0.24620310962200165, "learning_rate": 3.451180714846325e-05, "loss": 0.0099, "step": 49620 }, { "grad_norm": 0.30224156379699707, "learning_rate": 3.449215670402438e-05, "loss": 0.0122, "step": 49630 }, { "grad_norm": 0.2975689172744751, "learning_rate": 3.447250890945181e-05, "loss": 0.0161, "step": 49640 }, { "grad_norm": 0.16302081942558289, "learning_rate": 3.4452863768102754e-05, "loss": 0.0118, "step": 49650 }, { "grad_norm": 0.14225047826766968, "learning_rate": 3.443322128333409e-05, "loss": 0.0123, "step": 49660 }, { "grad_norm": 0.3519029915332794, "learning_rate": 3.441358145850215e-05, "loss": 0.0139, "step": 49670 }, { "grad_norm": 0.2362835556268692, "learning_rate": 3.439394429696286e-05, "loss": 0.0185, "step": 49680 }, { "grad_norm": 0.16574305295944214, "learning_rate": 3.4374309802071644e-05, "loss": 0.0135, "step": 49690 }, { "grad_norm": 0.32980579137802124, "learning_rate": 3.435467797718353e-05, "loss": 0.0094, "step": 49700 }, { "grad_norm": 0.17014147341251373, "learning_rate": 3.433504882565306e-05, "loss": 0.0107, "step": 49710 }, { "grad_norm": 0.16982221603393555, "learning_rate": 3.43154223508343e-05, "loss": 0.0138, "step": 49720 }, { "grad_norm": 0.1797260195016861, "learning_rate": 3.429579855608089e-05, "loss": 0.0113, "step": 49730 }, { "grad_norm": 0.22084130346775055, "learning_rate": 3.427617744474597e-05, "loss": 0.0148, "step": 49740 }, { "grad_norm": 0.17843614518642426, "learning_rate": 3.425655902018231e-05, "loss": 0.0138, "step": 49750 }, { "grad_norm": 0.1399189531803131, "learning_rate": 3.423694328574211e-05, "loss": 0.0147, "step": 49760 }, { "grad_norm": 0.20714916288852692, "learning_rate": 3.42173302447772e-05, "loss": 0.0118, "step": 49770 }, { "grad_norm": 0.14276830852031708, "learning_rate": 3.419771990063886e-05, "loss": 0.0092, "step": 49780 }, { "grad_norm": 0.1639075130224228, "learning_rate": 3.417811225667803e-05, "loss": 0.0076, "step": 49790 }, { "grad_norm": 0.21058735251426697, "learning_rate": 3.415850731624508e-05, "loss": 0.0155, "step": 49800 }, { "grad_norm": 0.20046748220920563, "learning_rate": 3.4138905082689945e-05, "loss": 0.017, "step": 49810 }, { "grad_norm": 0.18423305451869965, "learning_rate": 3.4119305559362145e-05, "loss": 0.0132, "step": 49820 }, { "grad_norm": 0.2318919152021408, "learning_rate": 3.4099708749610684e-05, "loss": 0.0094, "step": 49830 }, { "grad_norm": 0.3032605051994324, "learning_rate": 3.408011465678413e-05, "loss": 0.0163, "step": 49840 }, { "grad_norm": 0.16530416905879974, "learning_rate": 3.406052328423055e-05, "loss": 0.0169, "step": 49850 }, { "grad_norm": 0.16544388234615326, "learning_rate": 3.4040934635297615e-05, "loss": 0.0105, "step": 49860 }, { "grad_norm": 0.2128492146730423, "learning_rate": 3.4021348713332466e-05, "loss": 0.0121, "step": 49870 }, { "grad_norm": 0.2344541698694229, "learning_rate": 3.4001765521681807e-05, "loss": 0.0136, "step": 49880 }, { "grad_norm": 0.17481942474842072, "learning_rate": 3.398218506369188e-05, "loss": 0.0207, "step": 49890 }, { "grad_norm": 0.14626416563987732, "learning_rate": 3.3962607342708404e-05, "loss": 0.0125, "step": 49900 }, { "grad_norm": 0.23878400027751923, "learning_rate": 3.394303236207673e-05, "loss": 0.01, "step": 49910 }, { "grad_norm": 0.19334641098976135, "learning_rate": 3.392346012514166e-05, "loss": 0.0141, "step": 49920 }, { "grad_norm": 0.28526535630226135, "learning_rate": 3.390389063524757e-05, "loss": 0.0115, "step": 49930 }, { "grad_norm": 0.22203122079372406, "learning_rate": 3.3884323895738324e-05, "loss": 0.0108, "step": 49940 }, { "grad_norm": 0.29209572076797485, "learning_rate": 3.386475990995738e-05, "loss": 0.0162, "step": 49950 }, { "grad_norm": 0.18967612087726593, "learning_rate": 3.384519868124765e-05, "loss": 0.0115, "step": 49960 }, { "grad_norm": 0.18734222650527954, "learning_rate": 3.3825640212951645e-05, "loss": 0.0114, "step": 49970 }, { "grad_norm": 0.30545520782470703, "learning_rate": 3.380608450841134e-05, "loss": 0.0133, "step": 49980 }, { "grad_norm": 0.20806057751178741, "learning_rate": 3.3786531570968305e-05, "loss": 0.0117, "step": 49990 }, { "grad_norm": 0.19497455656528473, "learning_rate": 3.3766981403963584e-05, "loss": 0.0161, "step": 50000 }, { "grad_norm": 0.16303792595863342, "learning_rate": 3.374743401073775e-05, "loss": 0.0115, "step": 50010 }, { "grad_norm": 0.5920798182487488, "learning_rate": 3.372788939463095e-05, "loss": 0.01, "step": 50020 }, { "grad_norm": 0.1852784901857376, "learning_rate": 3.370834755898281e-05, "loss": 0.0084, "step": 50030 }, { "grad_norm": 0.27342671155929565, "learning_rate": 3.3688808507132493e-05, "loss": 0.0145, "step": 50040 }, { "grad_norm": 0.2460198849439621, "learning_rate": 3.3669272242418685e-05, "loss": 0.0178, "step": 50050 }, { "grad_norm": 0.19499477744102478, "learning_rate": 3.364973876817961e-05, "loss": 0.011, "step": 50060 }, { "grad_norm": 0.18854601681232452, "learning_rate": 3.363020808775299e-05, "loss": 0.0117, "step": 50070 }, { "grad_norm": 0.11259226500988007, "learning_rate": 3.361068020447611e-05, "loss": 0.0109, "step": 50080 }, { "grad_norm": 0.15898704528808594, "learning_rate": 3.3591155121685724e-05, "loss": 0.009, "step": 50090 }, { "grad_norm": 0.2505720555782318, "learning_rate": 3.357163284271814e-05, "loss": 0.0122, "step": 50100 }, { "grad_norm": 0.1816575676202774, "learning_rate": 3.355211337090919e-05, "loss": 0.0088, "step": 50110 }, { "grad_norm": 0.45298171043395996, "learning_rate": 3.353259670959421e-05, "loss": 0.0171, "step": 50120 }, { "grad_norm": 0.24397534132003784, "learning_rate": 3.351308286210808e-05, "loss": 0.0151, "step": 50130 }, { "grad_norm": 1.0501302480697632, "learning_rate": 3.3493571831785156e-05, "loss": 0.0132, "step": 50140 }, { "grad_norm": 0.21700139343738556, "learning_rate": 3.347406362195936e-05, "loss": 0.0095, "step": 50150 }, { "grad_norm": 0.18668420612812042, "learning_rate": 3.345455823596411e-05, "loss": 0.0122, "step": 50160 }, { "grad_norm": 0.16674242913722992, "learning_rate": 3.3435055677132346e-05, "loss": 0.0126, "step": 50170 }, { "grad_norm": 0.2486514151096344, "learning_rate": 3.3415555948796505e-05, "loss": 0.0119, "step": 50180 }, { "grad_norm": 0.25799688696861267, "learning_rate": 3.3396059054288556e-05, "loss": 0.0136, "step": 50190 }, { "grad_norm": 0.30545470118522644, "learning_rate": 3.3376564996940015e-05, "loss": 0.0219, "step": 50200 }, { "grad_norm": 0.21333126723766327, "learning_rate": 3.3357073780081836e-05, "loss": 0.0138, "step": 50210 }, { "grad_norm": 0.14687763154506683, "learning_rate": 3.333758540704459e-05, "loss": 0.0112, "step": 50220 }, { "grad_norm": 0.2008771151304245, "learning_rate": 3.331809988115827e-05, "loss": 0.0106, "step": 50230 }, { "grad_norm": 0.21848762035369873, "learning_rate": 3.329861720575244e-05, "loss": 0.0137, "step": 50240 }, { "grad_norm": 0.21338142454624176, "learning_rate": 3.3279137384156126e-05, "loss": 0.0101, "step": 50250 }, { "grad_norm": 0.2340581715106964, "learning_rate": 3.3259660419697934e-05, "loss": 0.0171, "step": 50260 }, { "grad_norm": 0.16629011929035187, "learning_rate": 3.3240186315705926e-05, "loss": 0.0159, "step": 50270 }, { "grad_norm": 0.2656824588775635, "learning_rate": 3.322071507550769e-05, "loss": 0.0176, "step": 50280 }, { "grad_norm": 0.24535709619522095, "learning_rate": 3.320124670243033e-05, "loss": 0.0104, "step": 50290 }, { "grad_norm": 0.27439191937446594, "learning_rate": 3.318178119980045e-05, "loss": 0.0115, "step": 50300 }, { "grad_norm": 0.26728102564811707, "learning_rate": 3.316231857094421e-05, "loss": 0.0115, "step": 50310 }, { "grad_norm": 0.32575416564941406, "learning_rate": 3.31428588191872e-05, "loss": 0.0186, "step": 50320 }, { "grad_norm": 0.2354484349489212, "learning_rate": 3.312340194785458e-05, "loss": 0.0121, "step": 50330 }, { "grad_norm": 0.29534071683883667, "learning_rate": 3.3103947960270975e-05, "loss": 0.0131, "step": 50340 }, { "grad_norm": 0.23149257898330688, "learning_rate": 3.308449685976058e-05, "loss": 0.0099, "step": 50350 }, { "grad_norm": 0.21832287311553955, "learning_rate": 3.3065048649647024e-05, "loss": 0.014, "step": 50360 }, { "grad_norm": 0.23527970910072327, "learning_rate": 3.304560333325348e-05, "loss": 0.011, "step": 50370 }, { "grad_norm": 0.1886683702468872, "learning_rate": 3.3026160913902624e-05, "loss": 0.0155, "step": 50380 }, { "grad_norm": 0.3058703541755676, "learning_rate": 3.300672139491662e-05, "loss": 0.0156, "step": 50390 }, { "grad_norm": 0.22524958848953247, "learning_rate": 3.298728477961717e-05, "loss": 0.0138, "step": 50400 }, { "grad_norm": 0.21532903611660004, "learning_rate": 3.296785107132544e-05, "loss": 0.0167, "step": 50410 }, { "grad_norm": 0.5857471823692322, "learning_rate": 3.2948420273362156e-05, "loss": 0.0129, "step": 50420 }, { "grad_norm": 0.085137739777565, "learning_rate": 3.292899238904747e-05, "loss": 0.0133, "step": 50430 }, { "grad_norm": 0.19710348546504974, "learning_rate": 3.29095674217011e-05, "loss": 0.013, "step": 50440 }, { "grad_norm": 0.19615770876407623, "learning_rate": 3.289014537464224e-05, "loss": 0.0137, "step": 50450 }, { "grad_norm": 0.18384677171707153, "learning_rate": 3.287072625118955e-05, "loss": 0.0132, "step": 50460 }, { "grad_norm": 0.19633488357067108, "learning_rate": 3.285131005466129e-05, "loss": 0.0119, "step": 50470 }, { "grad_norm": 0.20425792038440704, "learning_rate": 3.2831896788375105e-05, "loss": 0.014, "step": 50480 }, { "grad_norm": 0.3437815308570862, "learning_rate": 3.281248645564822e-05, "loss": 0.0124, "step": 50490 }, { "grad_norm": 0.17571334540843964, "learning_rate": 3.2793079059797306e-05, "loss": 0.0202, "step": 50500 }, { "grad_norm": 0.1676902025938034, "learning_rate": 3.277367460413859e-05, "loss": 0.0119, "step": 50510 }, { "grad_norm": 0.15573488175868988, "learning_rate": 3.275427309198773e-05, "loss": 0.0178, "step": 50520 }, { "grad_norm": 0.15814897418022156, "learning_rate": 3.273487452665993e-05, "loss": 0.0105, "step": 50530 }, { "grad_norm": 0.21934516727924347, "learning_rate": 3.271547891146986e-05, "loss": 0.0119, "step": 50540 }, { "grad_norm": 0.2814745604991913, "learning_rate": 3.269608624973173e-05, "loss": 0.013, "step": 50550 }, { "grad_norm": 0.25741997361183167, "learning_rate": 3.26766965447592e-05, "loss": 0.0139, "step": 50560 }, { "grad_norm": 0.2630132734775543, "learning_rate": 3.2657309799865424e-05, "loss": 0.0165, "step": 50570 }, { "grad_norm": 0.17052684724330902, "learning_rate": 3.2637926018363084e-05, "loss": 0.0116, "step": 50580 }, { "grad_norm": 0.28928953409194946, "learning_rate": 3.261854520356432e-05, "loss": 0.0153, "step": 50590 }, { "grad_norm": 0.23930582404136658, "learning_rate": 3.2599167358780816e-05, "loss": 0.0149, "step": 50600 }, { "grad_norm": 0.1831030696630478, "learning_rate": 3.2579792487323676e-05, "loss": 0.0093, "step": 50610 }, { "grad_norm": 0.22382241487503052, "learning_rate": 3.256042059250358e-05, "loss": 0.0172, "step": 50620 }, { "grad_norm": 0.36588412523269653, "learning_rate": 3.254105167763062e-05, "loss": 0.018, "step": 50630 }, { "grad_norm": 0.1424439698457718, "learning_rate": 3.252168574601443e-05, "loss": 0.012, "step": 50640 }, { "grad_norm": 0.25440001487731934, "learning_rate": 3.2502322800964115e-05, "loss": 0.012, "step": 50650 }, { "grad_norm": 0.3217676281929016, "learning_rate": 3.248296284578825e-05, "loss": 0.0124, "step": 50660 }, { "grad_norm": 0.4307076632976532, "learning_rate": 3.246360588379497e-05, "loss": 0.0163, "step": 50670 }, { "grad_norm": 0.15490363538265228, "learning_rate": 3.24442519182918e-05, "loss": 0.0196, "step": 50680 }, { "grad_norm": 0.15092864632606506, "learning_rate": 3.2424900952585845e-05, "loss": 0.0126, "step": 50690 }, { "grad_norm": 0.23906034231185913, "learning_rate": 3.240555298998362e-05, "loss": 0.0167, "step": 50700 }, { "grad_norm": 0.1949174702167511, "learning_rate": 3.2386208033791194e-05, "loss": 0.013, "step": 50710 }, { "grad_norm": 0.23780441284179688, "learning_rate": 3.2366866087314064e-05, "loss": 0.0141, "step": 50720 }, { "grad_norm": 0.15956179797649384, "learning_rate": 3.234752715385727e-05, "loss": 0.0158, "step": 50730 }, { "grad_norm": 0.26326701045036316, "learning_rate": 3.232819123672529e-05, "loss": 0.0179, "step": 50740 }, { "grad_norm": 0.4188988506793976, "learning_rate": 3.2308858339222076e-05, "loss": 0.0198, "step": 50750 }, { "grad_norm": 0.24768050014972687, "learning_rate": 3.2289528464651144e-05, "loss": 0.0128, "step": 50760 }, { "grad_norm": 0.22861607372760773, "learning_rate": 3.22702016163154e-05, "loss": 0.0142, "step": 50770 }, { "grad_norm": 0.1974707692861557, "learning_rate": 3.2250877797517295e-05, "loss": 0.017, "step": 50780 }, { "grad_norm": 0.11301730573177338, "learning_rate": 3.223155701155872e-05, "loss": 0.0105, "step": 50790 }, { "grad_norm": 0.19848957657814026, "learning_rate": 3.2212239261741105e-05, "loss": 0.0128, "step": 50800 }, { "grad_norm": 0.148850217461586, "learning_rate": 3.219292455136528e-05, "loss": 0.011, "step": 50810 }, { "grad_norm": 0.1448037475347519, "learning_rate": 3.217361288373165e-05, "loss": 0.0118, "step": 50820 }, { "grad_norm": 0.1488317847251892, "learning_rate": 3.215430426214002e-05, "loss": 0.0102, "step": 50830 }, { "grad_norm": 0.30506467819213867, "learning_rate": 3.2134998689889705e-05, "loss": 0.0144, "step": 50840 }, { "grad_norm": 0.17862717807292938, "learning_rate": 3.211569617027952e-05, "loss": 0.0138, "step": 50850 }, { "grad_norm": 0.19046518206596375, "learning_rate": 3.209639670660769e-05, "loss": 0.0146, "step": 50860 }, { "grad_norm": 0.45700928568840027, "learning_rate": 3.207710030217202e-05, "loss": 0.0128, "step": 50870 }, { "grad_norm": 0.3603476285934448, "learning_rate": 3.205780696026972e-05, "loss": 0.0119, "step": 50880 }, { "grad_norm": 0.274362713098526, "learning_rate": 3.203851668419749e-05, "loss": 0.0104, "step": 50890 }, { "grad_norm": 0.18941041827201843, "learning_rate": 3.201922947725149e-05, "loss": 0.0126, "step": 50900 }, { "grad_norm": 0.2990725338459015, "learning_rate": 3.199994534272742e-05, "loss": 0.0117, "step": 50910 }, { "grad_norm": 0.1931208223104477, "learning_rate": 3.1980664283920394e-05, "loss": 0.015, "step": 50920 }, { "grad_norm": 0.24442023038864136, "learning_rate": 3.196138630412499e-05, "loss": 0.0153, "step": 50930 }, { "grad_norm": 0.25457027554512024, "learning_rate": 3.194211140663532e-05, "loss": 0.0138, "step": 50940 }, { "grad_norm": 0.22372858226299286, "learning_rate": 3.1922839594744914e-05, "loss": 0.0182, "step": 50950 }, { "grad_norm": 0.16776339709758759, "learning_rate": 3.190357087174683e-05, "loss": 0.0149, "step": 50960 }, { "grad_norm": 0.12485917657613754, "learning_rate": 3.188430524093353e-05, "loss": 0.011, "step": 50970 }, { "grad_norm": 0.3633487820625305, "learning_rate": 3.1865042705597014e-05, "loss": 0.0135, "step": 50980 }, { "grad_norm": 0.27933716773986816, "learning_rate": 3.184578326902868e-05, "loss": 0.0106, "step": 50990 }, { "grad_norm": 0.38950708508491516, "learning_rate": 3.182652693451949e-05, "loss": 0.0123, "step": 51000 }, { "grad_norm": 0.1979282945394516, "learning_rate": 3.180727370535981e-05, "loss": 0.0124, "step": 51010 }, { "grad_norm": 0.18932880461215973, "learning_rate": 3.178802358483945e-05, "loss": 0.0113, "step": 51020 }, { "grad_norm": 0.22463174164295197, "learning_rate": 3.17687765762478e-05, "loss": 0.0117, "step": 51030 }, { "grad_norm": 0.23521782457828522, "learning_rate": 3.174953268287357e-05, "loss": 0.0158, "step": 51040 }, { "grad_norm": 0.49123114347457886, "learning_rate": 3.173029190800508e-05, "loss": 0.0108, "step": 51050 }, { "grad_norm": 0.25875839591026306, "learning_rate": 3.171105425493e-05, "loss": 0.0141, "step": 51060 }, { "grad_norm": 0.1642650067806244, "learning_rate": 3.169181972693557e-05, "loss": 0.008, "step": 51070 }, { "grad_norm": 0.18906660377979279, "learning_rate": 3.1672588327308404e-05, "loss": 0.0097, "step": 51080 }, { "grad_norm": 0.32284656167030334, "learning_rate": 3.1653360059334646e-05, "loss": 0.0149, "step": 51090 }, { "grad_norm": 0.3690264821052551, "learning_rate": 3.163413492629985e-05, "loss": 0.0259, "step": 51100 }, { "grad_norm": 0.762343168258667, "learning_rate": 3.161491293148912e-05, "loss": 0.0176, "step": 51110 }, { "grad_norm": 0.21282295882701874, "learning_rate": 3.1595694078186925e-05, "loss": 0.0135, "step": 51120 }, { "grad_norm": 0.20108383893966675, "learning_rate": 3.157647836967724e-05, "loss": 0.0136, "step": 51130 }, { "grad_norm": 0.37834250926971436, "learning_rate": 3.155726580924354e-05, "loss": 0.0136, "step": 51140 }, { "grad_norm": 0.24130988121032715, "learning_rate": 3.1538056400168676e-05, "loss": 0.0195, "step": 51150 }, { "grad_norm": 0.15873591601848602, "learning_rate": 3.151885014573506e-05, "loss": 0.0155, "step": 51160 }, { "grad_norm": 0.15212544798851013, "learning_rate": 3.149964704922448e-05, "loss": 0.0157, "step": 51170 }, { "grad_norm": 0.12401442229747772, "learning_rate": 3.148044711391825e-05, "loss": 0.012, "step": 51180 }, { "grad_norm": 0.2930377721786499, "learning_rate": 3.146125034309707e-05, "loss": 0.0149, "step": 51190 }, { "grad_norm": 0.11556033045053482, "learning_rate": 3.1442056740041195e-05, "loss": 0.0154, "step": 51200 }, { "grad_norm": 0.1644286960363388, "learning_rate": 3.142286630803026e-05, "loss": 0.0198, "step": 51210 }, { "grad_norm": 0.21439261734485626, "learning_rate": 3.140367905034337e-05, "loss": 0.0104, "step": 51220 }, { "grad_norm": 0.21661190688610077, "learning_rate": 3.138449497025914e-05, "loss": 0.0146, "step": 51230 }, { "grad_norm": 0.2353508621454239, "learning_rate": 3.136531407105557e-05, "loss": 0.0129, "step": 51240 }, { "grad_norm": 0.20774270594120026, "learning_rate": 3.1346136356010184e-05, "loss": 0.0146, "step": 51250 }, { "grad_norm": 0.15568187832832336, "learning_rate": 3.132696182839988e-05, "loss": 0.0098, "step": 51260 }, { "grad_norm": 0.26932254433631897, "learning_rate": 3.130779049150111e-05, "loss": 0.0165, "step": 51270 }, { "grad_norm": 0.3340120315551758, "learning_rate": 3.128862234858971e-05, "loss": 0.0134, "step": 51280 }, { "grad_norm": 0.20765982568264008, "learning_rate": 3.1269457402941005e-05, "loss": 0.0121, "step": 51290 }, { "grad_norm": 0.21482408046722412, "learning_rate": 3.125029565782974e-05, "loss": 0.0135, "step": 51300 }, { "grad_norm": 0.39209800958633423, "learning_rate": 3.123113711653013e-05, "loss": 0.0133, "step": 51310 }, { "grad_norm": 0.24820880591869354, "learning_rate": 3.121198178231587e-05, "loss": 0.0144, "step": 51320 }, { "grad_norm": 0.08737233281135559, "learning_rate": 3.119282965846006e-05, "loss": 0.0155, "step": 51330 }, { "grad_norm": 0.286571741104126, "learning_rate": 3.1173680748235303e-05, "loss": 0.0134, "step": 51340 }, { "grad_norm": 0.22947178781032562, "learning_rate": 3.115453505491358e-05, "loss": 0.0145, "step": 51350 }, { "grad_norm": 0.171741783618927, "learning_rate": 3.1135392581766404e-05, "loss": 0.0125, "step": 51360 }, { "grad_norm": 0.1623883843421936, "learning_rate": 3.111625333206467e-05, "loss": 0.0144, "step": 51370 }, { "grad_norm": 0.15262554585933685, "learning_rate": 3.109711730907881e-05, "loss": 0.0083, "step": 51380 }, { "grad_norm": 0.326744019985199, "learning_rate": 3.107798451607859e-05, "loss": 0.0113, "step": 51390 }, { "grad_norm": 0.2271246612071991, "learning_rate": 3.105885495633328e-05, "loss": 0.0139, "step": 51400 }, { "grad_norm": 0.11579545587301254, "learning_rate": 3.103972863311163e-05, "loss": 0.0082, "step": 51410 }, { "grad_norm": 0.15060018002986908, "learning_rate": 3.102060554968178e-05, "loss": 0.0128, "step": 51420 }, { "grad_norm": 0.19380344450473785, "learning_rate": 3.100148570931137e-05, "loss": 0.0096, "step": 51430 }, { "grad_norm": 0.23274950683116913, "learning_rate": 3.098236911526744e-05, "loss": 0.0159, "step": 51440 }, { "grad_norm": 0.17680753767490387, "learning_rate": 3.0963255770816495e-05, "loss": 0.0087, "step": 51450 }, { "grad_norm": 0.16270600259304047, "learning_rate": 3.094414567922447e-05, "loss": 0.0108, "step": 51460 }, { "grad_norm": 0.2544630467891693, "learning_rate": 3.092503884375678e-05, "loss": 0.0147, "step": 51470 }, { "grad_norm": 0.2804712951183319, "learning_rate": 3.0905935267678255e-05, "loss": 0.0158, "step": 51480 }, { "grad_norm": 0.18309541046619415, "learning_rate": 3.0886834954253166e-05, "loss": 0.0114, "step": 51490 }, { "grad_norm": 0.2371545135974884, "learning_rate": 3.086773790674524e-05, "loss": 0.0126, "step": 51500 }, { "grad_norm": 0.3323509395122528, "learning_rate": 3.0848644128417617e-05, "loss": 0.0137, "step": 51510 }, { "grad_norm": 0.30202963948249817, "learning_rate": 3.082955362253294e-05, "loss": 0.012, "step": 51520 }, { "grad_norm": 0.17932870984077454, "learning_rate": 3.081046639235323e-05, "loss": 0.0183, "step": 51530 }, { "grad_norm": 0.16803193092346191, "learning_rate": 3.0791382441139986e-05, "loss": 0.0104, "step": 51540 }, { "grad_norm": 0.2181318998336792, "learning_rate": 3.077230177215411e-05, "loss": 0.0131, "step": 51550 }, { "grad_norm": 0.114869125187397, "learning_rate": 3.0753224388656e-05, "loss": 0.0095, "step": 51560 }, { "grad_norm": 0.1629619598388672, "learning_rate": 3.073415029390544e-05, "loss": 0.0167, "step": 51570 }, { "grad_norm": 0.22522225975990295, "learning_rate": 3.071507949116166e-05, "loss": 0.0124, "step": 51580 }, { "grad_norm": 0.2692004442214966, "learning_rate": 3.069601198368337e-05, "loss": 0.0111, "step": 51590 }, { "grad_norm": 0.08978539705276489, "learning_rate": 3.067694777472864e-05, "loss": 0.0156, "step": 51600 }, { "grad_norm": 0.1633298397064209, "learning_rate": 3.065788686755508e-05, "loss": 0.0118, "step": 51610 }, { "grad_norm": 0.19392681121826172, "learning_rate": 3.063882926541961e-05, "loss": 0.0137, "step": 51620 }, { "grad_norm": 0.3620128333568573, "learning_rate": 3.061977497157872e-05, "loss": 0.0106, "step": 51630 }, { "grad_norm": 0.151688814163208, "learning_rate": 3.0600723989288224e-05, "loss": 0.0113, "step": 51640 }, { "grad_norm": 0.17019058763980865, "learning_rate": 3.058167632180343e-05, "loss": 0.0165, "step": 51650 }, { "grad_norm": 0.18998980522155762, "learning_rate": 3.0562631972379074e-05, "loss": 0.0146, "step": 51660 }, { "grad_norm": 0.20822378993034363, "learning_rate": 3.0543590944269276e-05, "loss": 0.0127, "step": 51670 }, { "grad_norm": 0.2623552083969116, "learning_rate": 3.052455324072766e-05, "loss": 0.0179, "step": 51680 }, { "grad_norm": 0.30851656198501587, "learning_rate": 3.0505518865007244e-05, "loss": 0.0127, "step": 51690 }, { "grad_norm": 0.16632458567619324, "learning_rate": 3.048648782036048e-05, "loss": 0.0105, "step": 51700 }, { "grad_norm": 0.21120326220989227, "learning_rate": 3.0467460110039236e-05, "loss": 0.0117, "step": 51710 }, { "grad_norm": 0.26257798075675964, "learning_rate": 3.0448435737294866e-05, "loss": 0.0133, "step": 51720 }, { "grad_norm": 0.12777842581272125, "learning_rate": 3.042941470537808e-05, "loss": 0.0109, "step": 51730 }, { "grad_norm": 0.1370861977338791, "learning_rate": 3.0410397017539072e-05, "loss": 0.0107, "step": 51740 }, { "grad_norm": 0.29552143812179565, "learning_rate": 3.039138267702742e-05, "loss": 0.0139, "step": 51750 }, { "grad_norm": 0.11979028582572937, "learning_rate": 3.0372371687092193e-05, "loss": 0.0091, "step": 51760 }, { "grad_norm": 0.29893937706947327, "learning_rate": 3.0353364050981823e-05, "loss": 0.0131, "step": 51770 }, { "grad_norm": 0.26240435242652893, "learning_rate": 3.033435977194418e-05, "loss": 0.0137, "step": 51780 }, { "grad_norm": 0.2409905642271042, "learning_rate": 3.0315358853226618e-05, "loss": 0.0154, "step": 51790 }, { "grad_norm": 0.10889942944049835, "learning_rate": 3.0296361298075844e-05, "loss": 0.0123, "step": 51800 }, { "grad_norm": 0.19180338084697723, "learning_rate": 3.027736710973803e-05, "loss": 0.0147, "step": 51810 }, { "grad_norm": 0.13987238705158234, "learning_rate": 3.025837629145875e-05, "loss": 0.0121, "step": 51820 }, { "grad_norm": 0.17057234048843384, "learning_rate": 3.0239388846483048e-05, "loss": 0.0113, "step": 51830 }, { "grad_norm": 0.23060175776481628, "learning_rate": 3.022040477805532e-05, "loss": 0.0135, "step": 51840 }, { "grad_norm": 0.3080725371837616, "learning_rate": 3.020142408941946e-05, "loss": 0.0166, "step": 51850 }, { "grad_norm": 0.20486296713352203, "learning_rate": 3.018244678381873e-05, "loss": 0.0095, "step": 51860 }, { "grad_norm": 0.3383382260799408, "learning_rate": 3.0163472864495812e-05, "loss": 0.0163, "step": 51870 }, { "grad_norm": 0.23567555844783783, "learning_rate": 3.014450233469287e-05, "loss": 0.0139, "step": 51880 }, { "grad_norm": 0.4278244078159332, "learning_rate": 3.0125535197651422e-05, "loss": 0.0195, "step": 51890 }, { "grad_norm": 0.1976940631866455, "learning_rate": 3.0106571456612448e-05, "loss": 0.0105, "step": 51900 }, { "grad_norm": 0.2991487979888916, "learning_rate": 3.0087611114816305e-05, "loss": 0.0158, "step": 51910 }, { "grad_norm": 0.20062053203582764, "learning_rate": 3.006865417550284e-05, "loss": 0.0119, "step": 51920 }, { "grad_norm": 0.2566394805908203, "learning_rate": 3.0049700641911242e-05, "loss": 0.015, "step": 51930 }, { "grad_norm": 0.21972432732582092, "learning_rate": 3.0030750517280175e-05, "loss": 0.0118, "step": 51940 }, { "grad_norm": 0.14696134626865387, "learning_rate": 3.0011803804847682e-05, "loss": 0.017, "step": 51950 }, { "grad_norm": 0.6708967685699463, "learning_rate": 2.9992860507851224e-05, "loss": 0.0158, "step": 51960 }, { "grad_norm": 0.14542295038700104, "learning_rate": 2.9973920629527723e-05, "loss": 0.0083, "step": 51970 }, { "grad_norm": 0.14472199976444244, "learning_rate": 2.9954984173113453e-05, "loss": 0.0116, "step": 51980 }, { "grad_norm": 0.28691333532333374, "learning_rate": 2.993605114184418e-05, "loss": 0.0128, "step": 51990 }, { "grad_norm": 0.3260575532913208, "learning_rate": 2.9917121538955005e-05, "loss": 0.0152, "step": 52000 }, { "grad_norm": 0.2214534431695938, "learning_rate": 2.98981953676805e-05, "loss": 0.0173, "step": 52010 }, { "grad_norm": 0.18422965705394745, "learning_rate": 2.9879272631254594e-05, "loss": 0.013, "step": 52020 }, { "grad_norm": 0.08953439444303513, "learning_rate": 2.986035333291073e-05, "loss": 0.0107, "step": 52030 }, { "grad_norm": 0.20562712848186493, "learning_rate": 2.9841437475881646e-05, "loss": 0.0143, "step": 52040 }, { "grad_norm": 0.14490137994289398, "learning_rate": 2.9822525063399555e-05, "loss": 0.012, "step": 52050 }, { "grad_norm": 0.14991235733032227, "learning_rate": 2.9803616098696087e-05, "loss": 0.0174, "step": 52060 }, { "grad_norm": 0.22618570923805237, "learning_rate": 2.9784710585002228e-05, "loss": 0.0125, "step": 52070 }, { "grad_norm": 0.27286189794540405, "learning_rate": 2.9765808525548467e-05, "loss": 0.012, "step": 52080 }, { "grad_norm": 0.19679777324199677, "learning_rate": 2.974690992356461e-05, "loss": 0.0141, "step": 52090 }, { "grad_norm": 0.1524830311536789, "learning_rate": 2.9728014782279934e-05, "loss": 0.0135, "step": 52100 }, { "grad_norm": 0.1934991329908371, "learning_rate": 2.970912310492307e-05, "loss": 0.0173, "step": 52110 }, { "grad_norm": 0.24974898993968964, "learning_rate": 2.9690234894722134e-05, "loss": 0.0106, "step": 52120 }, { "grad_norm": 0.128055140376091, "learning_rate": 2.9671350154904577e-05, "loss": 0.0163, "step": 52130 }, { "grad_norm": 0.5031744837760925, "learning_rate": 2.9652468888697282e-05, "loss": 0.0133, "step": 52140 }, { "grad_norm": 0.20552019774913788, "learning_rate": 2.9633591099326562e-05, "loss": 0.012, "step": 52150 }, { "grad_norm": 0.1743410974740982, "learning_rate": 2.9614716790018072e-05, "loss": 0.0119, "step": 52160 }, { "grad_norm": 0.15810680389404297, "learning_rate": 2.959584596399697e-05, "loss": 0.0137, "step": 52170 }, { "grad_norm": 0.235332190990448, "learning_rate": 2.9576978624487717e-05, "loss": 0.0144, "step": 52180 }, { "grad_norm": 0.23973405361175537, "learning_rate": 2.9558114774714263e-05, "loss": 0.0114, "step": 52190 }, { "grad_norm": 0.26199060678482056, "learning_rate": 2.9539254417899897e-05, "loss": 0.0176, "step": 52200 }, { "grad_norm": 0.19351769983768463, "learning_rate": 2.9520397557267365e-05, "loss": 0.01, "step": 52210 }, { "grad_norm": 0.21970300376415253, "learning_rate": 2.9501544196038765e-05, "loss": 0.0154, "step": 52220 }, { "grad_norm": 0.1682150959968567, "learning_rate": 2.9482694337435614e-05, "loss": 0.0176, "step": 52230 }, { "grad_norm": 0.15470623970031738, "learning_rate": 2.946384798467887e-05, "loss": 0.0114, "step": 52240 }, { "grad_norm": 0.1446777880191803, "learning_rate": 2.9445005140988824e-05, "loss": 0.0117, "step": 52250 }, { "grad_norm": 0.2656604051589966, "learning_rate": 2.942616580958524e-05, "loss": 0.013, "step": 52260 }, { "grad_norm": 0.20999810099601746, "learning_rate": 2.9407329993687193e-05, "loss": 0.0097, "step": 52270 }, { "grad_norm": 0.24691204726696014, "learning_rate": 2.938849769651326e-05, "loss": 0.012, "step": 52280 }, { "grad_norm": 0.35194212198257446, "learning_rate": 2.9369668921281345e-05, "loss": 0.0123, "step": 52290 }, { "grad_norm": 0.10719018429517746, "learning_rate": 2.9350843671208773e-05, "loss": 0.0094, "step": 52300 }, { "grad_norm": 0.3073351979255676, "learning_rate": 2.9332021949512245e-05, "loss": 0.0104, "step": 52310 }, { "grad_norm": 0.25325149297714233, "learning_rate": 2.9313203759407908e-05, "loss": 0.0113, "step": 52320 }, { "grad_norm": 0.20074474811553955, "learning_rate": 2.929438910411127e-05, "loss": 0.0221, "step": 52330 }, { "grad_norm": 0.3631734848022461, "learning_rate": 2.927557798683722e-05, "loss": 0.015, "step": 52340 }, { "grad_norm": 0.2745264768600464, "learning_rate": 2.9256770410800095e-05, "loss": 0.0124, "step": 52350 }, { "grad_norm": 0.2681700885295868, "learning_rate": 2.9237966379213554e-05, "loss": 0.0121, "step": 52360 }, { "grad_norm": 0.12300710380077362, "learning_rate": 2.9219165895290736e-05, "loss": 0.0153, "step": 52370 }, { "grad_norm": 0.15926648676395416, "learning_rate": 2.9200368962244083e-05, "loss": 0.0159, "step": 52380 }, { "grad_norm": 0.18529248237609863, "learning_rate": 2.9181575583285536e-05, "loss": 0.0118, "step": 52390 }, { "grad_norm": 0.1537470519542694, "learning_rate": 2.916278576162632e-05, "loss": 0.014, "step": 52400 }, { "grad_norm": 0.17289651930332184, "learning_rate": 2.9143999500477116e-05, "loss": 0.0133, "step": 52410 }, { "grad_norm": 0.2663263976573944, "learning_rate": 2.9125216803048004e-05, "loss": 0.0173, "step": 52420 }, { "grad_norm": 0.19047856330871582, "learning_rate": 2.9106437672548403e-05, "loss": 0.0115, "step": 52430 }, { "grad_norm": 0.1480434238910675, "learning_rate": 2.9087662112187154e-05, "loss": 0.0106, "step": 52440 }, { "grad_norm": 0.2513265907764435, "learning_rate": 2.9068890125172498e-05, "loss": 0.0111, "step": 52450 }, { "grad_norm": 0.3708904981613159, "learning_rate": 2.9050121714712058e-05, "loss": 0.0143, "step": 52460 }, { "grad_norm": 0.3832773268222809, "learning_rate": 2.9031356884012835e-05, "loss": 0.0148, "step": 52470 }, { "grad_norm": 0.13886819779872894, "learning_rate": 2.9012595636281247e-05, "loss": 0.0162, "step": 52480 }, { "grad_norm": 0.18923312425613403, "learning_rate": 2.8993837974723016e-05, "loss": 0.0215, "step": 52490 }, { "grad_norm": 0.1306181401014328, "learning_rate": 2.8975083902543394e-05, "loss": 0.011, "step": 52500 }, { "grad_norm": 0.18658025562763214, "learning_rate": 2.895633342294688e-05, "loss": 0.0137, "step": 52510 }, { "grad_norm": 0.25831976532936096, "learning_rate": 2.893758653913744e-05, "loss": 0.0111, "step": 52520 }, { "grad_norm": 0.1984270066022873, "learning_rate": 2.8918843254318423e-05, "loss": 0.0124, "step": 52530 }, { "grad_norm": 0.27534809708595276, "learning_rate": 2.8900103571692483e-05, "loss": 0.0145, "step": 52540 }, { "grad_norm": 0.2873865067958832, "learning_rate": 2.8881367494461796e-05, "loss": 0.0119, "step": 52550 }, { "grad_norm": 0.17576910555362701, "learning_rate": 2.886263502582779e-05, "loss": 0.0136, "step": 52560 }, { "grad_norm": 0.16817189753055573, "learning_rate": 2.8843906168991353e-05, "loss": 0.0116, "step": 52570 }, { "grad_norm": 0.269283652305603, "learning_rate": 2.8825180927152728e-05, "loss": 0.0134, "step": 52580 }, { "grad_norm": 0.290396511554718, "learning_rate": 2.8806459303511547e-05, "loss": 0.0099, "step": 52590 }, { "grad_norm": 0.25437504053115845, "learning_rate": 2.8787741301266852e-05, "loss": 0.0112, "step": 52600 }, { "grad_norm": 0.19642600417137146, "learning_rate": 2.876902692361697e-05, "loss": 0.0133, "step": 52610 }, { "grad_norm": 0.2708193361759186, "learning_rate": 2.875031617375976e-05, "loss": 0.0122, "step": 52620 }, { "grad_norm": 0.09415002912282944, "learning_rate": 2.8731609054892318e-05, "loss": 0.0099, "step": 52630 }, { "grad_norm": 0.14612989127635956, "learning_rate": 2.871290557021119e-05, "loss": 0.0121, "step": 52640 }, { "grad_norm": 0.23137961328029633, "learning_rate": 2.8694205722912298e-05, "loss": 0.0119, "step": 52650 }, { "grad_norm": 0.38564857840538025, "learning_rate": 2.8675509516190936e-05, "loss": 0.0139, "step": 52660 }, { "grad_norm": 0.20645703375339508, "learning_rate": 2.8656816953241772e-05, "loss": 0.0203, "step": 52670 }, { "grad_norm": 0.5520851612091064, "learning_rate": 2.863812803725887e-05, "loss": 0.0115, "step": 52680 }, { "grad_norm": 0.32163146138191223, "learning_rate": 2.8619442771435623e-05, "loss": 0.0121, "step": 52690 }, { "grad_norm": 0.15626239776611328, "learning_rate": 2.860076115896484e-05, "loss": 0.0158, "step": 52700 }, { "grad_norm": 0.22562463581562042, "learning_rate": 2.8582083203038712e-05, "loss": 0.0136, "step": 52710 }, { "grad_norm": 0.2418198585510254, "learning_rate": 2.8563408906848778e-05, "loss": 0.0106, "step": 52720 }, { "grad_norm": 0.23704105615615845, "learning_rate": 2.8544738273585993e-05, "loss": 0.0129, "step": 52730 }, { "grad_norm": 0.3579631745815277, "learning_rate": 2.8526071306440595e-05, "loss": 0.0087, "step": 52740 }, { "grad_norm": 0.42717936635017395, "learning_rate": 2.850740800860233e-05, "loss": 0.0139, "step": 52750 }, { "grad_norm": 0.16918060183525085, "learning_rate": 2.8488748383260178e-05, "loss": 0.0103, "step": 52760 }, { "grad_norm": 0.5023888945579529, "learning_rate": 2.8470092433602635e-05, "loss": 0.0151, "step": 52770 }, { "grad_norm": 0.1471358835697174, "learning_rate": 2.845144016281745e-05, "loss": 0.0106, "step": 52780 }, { "grad_norm": 0.2274077981710434, "learning_rate": 2.8432791574091744e-05, "loss": 0.0212, "step": 52790 }, { "grad_norm": 0.1593160778284073, "learning_rate": 2.8414146670612134e-05, "loss": 0.0136, "step": 52800 }, { "grad_norm": 0.11608046293258667, "learning_rate": 2.8395505455564446e-05, "loss": 0.0143, "step": 52810 }, { "grad_norm": 0.5815920829772949, "learning_rate": 2.837686793213403e-05, "loss": 0.012, "step": 52820 }, { "grad_norm": 0.2756637632846832, "learning_rate": 2.8358234103505477e-05, "loss": 0.0151, "step": 52830 }, { "grad_norm": 0.2658093571662903, "learning_rate": 2.8339603972862806e-05, "loss": 0.0161, "step": 52840 }, { "grad_norm": 0.39368996024131775, "learning_rate": 2.83209775433894e-05, "loss": 0.0156, "step": 52850 }, { "grad_norm": 0.2727554142475128, "learning_rate": 2.8302354818268013e-05, "loss": 0.0118, "step": 52860 }, { "grad_norm": 0.2240845263004303, "learning_rate": 2.8283735800680754e-05, "loss": 0.0192, "step": 52870 }, { "grad_norm": 0.16577932238578796, "learning_rate": 2.826512049380913e-05, "loss": 0.0093, "step": 52880 }, { "grad_norm": 0.1512925773859024, "learning_rate": 2.8246508900833935e-05, "loss": 0.0125, "step": 52890 }, { "grad_norm": 0.24569152295589447, "learning_rate": 2.822790102493541e-05, "loss": 0.0155, "step": 52900 }, { "grad_norm": 0.17383016645908356, "learning_rate": 2.820929686929314e-05, "loss": 0.0131, "step": 52910 }, { "grad_norm": 0.17283539474010468, "learning_rate": 2.819069643708605e-05, "loss": 0.0102, "step": 52920 }, { "grad_norm": 0.32651597261428833, "learning_rate": 2.8172099731492484e-05, "loss": 0.0138, "step": 52930 }, { "grad_norm": 0.18481974303722382, "learning_rate": 2.815350675569003e-05, "loss": 0.019, "step": 52940 }, { "grad_norm": 0.1802951693534851, "learning_rate": 2.8134917512855807e-05, "loss": 0.0188, "step": 52950 }, { "grad_norm": 0.34632015228271484, "learning_rate": 2.8116332006166134e-05, "loss": 0.0091, "step": 52960 }, { "grad_norm": 0.26846370100975037, "learning_rate": 2.809775023879685e-05, "loss": 0.0252, "step": 52970 }, { "grad_norm": 0.1972486674785614, "learning_rate": 2.807917221392299e-05, "loss": 0.0137, "step": 52980 }, { "grad_norm": 0.19446252286434174, "learning_rate": 2.8060597934719067e-05, "loss": 0.0126, "step": 52990 }, { "grad_norm": 0.10657001286745071, "learning_rate": 2.8042027404358927e-05, "loss": 0.0115, "step": 53000 }, { "grad_norm": 0.1573055535554886, "learning_rate": 2.802346062601572e-05, "loss": 0.0124, "step": 53010 }, { "grad_norm": 0.3268216550350189, "learning_rate": 2.800489760286207e-05, "loss": 0.0234, "step": 53020 }, { "grad_norm": 0.28457677364349365, "learning_rate": 2.7986338338069834e-05, "loss": 0.012, "step": 53030 }, { "grad_norm": 0.5333677530288696, "learning_rate": 2.7967782834810297e-05, "loss": 0.0186, "step": 53040 }, { "grad_norm": 0.29398226737976074, "learning_rate": 2.794923109625409e-05, "loss": 0.0104, "step": 53050 }, { "grad_norm": 0.22012759745121002, "learning_rate": 2.7930683125571205e-05, "loss": 0.0123, "step": 53060 }, { "grad_norm": 0.3608597218990326, "learning_rate": 2.7912138925930997e-05, "loss": 0.0185, "step": 53070 }, { "grad_norm": 0.14089132845401764, "learning_rate": 2.7893598500502117e-05, "loss": 0.0201, "step": 53080 }, { "grad_norm": 0.2922230362892151, "learning_rate": 2.7875061852452644e-05, "loss": 0.0116, "step": 53090 }, { "grad_norm": 0.29840990900993347, "learning_rate": 2.7856528984949982e-05, "loss": 0.0141, "step": 53100 }, { "grad_norm": 0.17898131906986237, "learning_rate": 2.7837999901160888e-05, "loss": 0.011, "step": 53110 }, { "grad_norm": 0.22009223699569702, "learning_rate": 2.7819474604251484e-05, "loss": 0.0099, "step": 53120 }, { "grad_norm": 0.18365339934825897, "learning_rate": 2.780095309738725e-05, "loss": 0.0129, "step": 53130 }, { "grad_norm": 0.3431019186973572, "learning_rate": 2.778243538373294e-05, "loss": 0.0116, "step": 53140 }, { "grad_norm": 0.13269945979118347, "learning_rate": 2.7763921466452826e-05, "loss": 0.0126, "step": 53150 }, { "grad_norm": 0.26654285192489624, "learning_rate": 2.7745411348710336e-05, "loss": 0.0182, "step": 53160 }, { "grad_norm": 0.288775235414505, "learning_rate": 2.7726905033668395e-05, "loss": 0.0163, "step": 53170 }, { "grad_norm": 0.24861672520637512, "learning_rate": 2.7708402524489214e-05, "loss": 0.017, "step": 53180 }, { "grad_norm": 0.21734943985939026, "learning_rate": 2.7689903824334364e-05, "loss": 0.016, "step": 53190 }, { "grad_norm": 0.18308980762958527, "learning_rate": 2.7671408936364785e-05, "loss": 0.013, "step": 53200 }, { "grad_norm": 0.1993204653263092, "learning_rate": 2.765291786374069e-05, "loss": 0.0104, "step": 53210 }, { "grad_norm": 0.2620111107826233, "learning_rate": 2.7634430609621775e-05, "loss": 0.0101, "step": 53220 }, { "grad_norm": 0.4955994784832001, "learning_rate": 2.7615947177166956e-05, "loss": 0.0139, "step": 53230 }, { "grad_norm": 0.2422933131456375, "learning_rate": 2.7597467569534553e-05, "loss": 0.0107, "step": 53240 }, { "grad_norm": 0.12134549021720886, "learning_rate": 2.757899178988226e-05, "loss": 0.0118, "step": 53250 }, { "grad_norm": 0.13805362582206726, "learning_rate": 2.7560519841367005e-05, "loss": 0.0133, "step": 53260 }, { "grad_norm": 0.12659162282943726, "learning_rate": 2.7542051727145237e-05, "loss": 0.0127, "step": 53270 }, { "grad_norm": 0.19441309571266174, "learning_rate": 2.7523587450372578e-05, "loss": 0.0133, "step": 53280 }, { "grad_norm": 0.28433501720428467, "learning_rate": 2.750512701420409e-05, "loss": 0.0112, "step": 53290 }, { "grad_norm": 0.14335471391677856, "learning_rate": 2.7486670421794158e-05, "loss": 0.0143, "step": 53300 }, { "grad_norm": 0.1806081384420395, "learning_rate": 2.7468217676296515e-05, "loss": 0.0188, "step": 53310 }, { "grad_norm": 0.2681228518486023, "learning_rate": 2.7449768780864226e-05, "loss": 0.0164, "step": 53320 }, { "grad_norm": 0.1962004005908966, "learning_rate": 2.7431323738649724e-05, "loss": 0.0124, "step": 53330 }, { "grad_norm": 0.3307831287384033, "learning_rate": 2.7412882552804713e-05, "loss": 0.0131, "step": 53340 }, { "grad_norm": 0.15082170069217682, "learning_rate": 2.739444522648032e-05, "loss": 0.0092, "step": 53350 }, { "grad_norm": 0.12022761255502701, "learning_rate": 2.737601176282697e-05, "loss": 0.0127, "step": 53360 }, { "grad_norm": 0.2851737439632416, "learning_rate": 2.735758216499445e-05, "loss": 0.0113, "step": 53370 }, { "grad_norm": 0.2208307832479477, "learning_rate": 2.7339156436131864e-05, "loss": 0.0096, "step": 53380 }, { "grad_norm": 0.1874697357416153, "learning_rate": 2.7320734579387663e-05, "loss": 0.0103, "step": 53390 }, { "grad_norm": 0.23000766336917877, "learning_rate": 2.730231659790966e-05, "loss": 0.0097, "step": 53400 }, { "grad_norm": 0.4640837609767914, "learning_rate": 2.728390249484494e-05, "loss": 0.0173, "step": 53410 }, { "grad_norm": 0.1352149099111557, "learning_rate": 2.7265492273340032e-05, "loss": 0.0125, "step": 53420 }, { "grad_norm": 0.29598578810691833, "learning_rate": 2.7247085936540678e-05, "loss": 0.0115, "step": 53430 }, { "grad_norm": 0.16179901361465454, "learning_rate": 2.722868348759205e-05, "loss": 0.0151, "step": 53440 }, { "grad_norm": 0.1778426617383957, "learning_rate": 2.7210284929638635e-05, "loss": 0.0136, "step": 53450 }, { "grad_norm": 0.1074623093008995, "learning_rate": 2.7191890265824183e-05, "loss": 0.012, "step": 53460 }, { "grad_norm": 0.16726192831993103, "learning_rate": 2.7173499499291926e-05, "loss": 0.0119, "step": 53470 }, { "grad_norm": 0.272187739610672, "learning_rate": 2.7155112633184277e-05, "loss": 0.0145, "step": 53480 }, { "grad_norm": 0.3785352110862732, "learning_rate": 2.713672967064307e-05, "loss": 0.015, "step": 53490 }, { "grad_norm": 0.13211874663829803, "learning_rate": 2.711835061480945e-05, "loss": 0.0108, "step": 53500 }, { "grad_norm": 0.31770452857017517, "learning_rate": 2.7099975468823896e-05, "loss": 0.0142, "step": 53510 }, { "grad_norm": 0.3157845735549927, "learning_rate": 2.708160423582622e-05, "loss": 0.0154, "step": 53520 }, { "grad_norm": 0.40443456172943115, "learning_rate": 2.706323691895557e-05, "loss": 0.0154, "step": 53530 }, { "grad_norm": 0.5177525877952576, "learning_rate": 2.70448735213504e-05, "loss": 0.0154, "step": 53540 }, { "grad_norm": 0.2337908148765564, "learning_rate": 2.702651404614852e-05, "loss": 0.012, "step": 53550 }, { "grad_norm": 0.9085070490837097, "learning_rate": 2.7008158496487056e-05, "loss": 0.0118, "step": 53560 }, { "grad_norm": 0.12495724856853485, "learning_rate": 2.6989806875502487e-05, "loss": 0.0115, "step": 53570 }, { "grad_norm": 0.2608342170715332, "learning_rate": 2.6971459186330584e-05, "loss": 0.0102, "step": 53580 }, { "grad_norm": 0.29201915860176086, "learning_rate": 2.695311543210648e-05, "loss": 0.0133, "step": 53590 }, { "grad_norm": 0.27182385325431824, "learning_rate": 2.693477561596463e-05, "loss": 0.0102, "step": 53600 }, { "grad_norm": 0.28533512353897095, "learning_rate": 2.6916439741038756e-05, "loss": 0.011, "step": 53610 }, { "grad_norm": 0.20948083698749542, "learning_rate": 2.689810781046203e-05, "loss": 0.0103, "step": 53620 }, { "grad_norm": 0.1554373800754547, "learning_rate": 2.6879779827366823e-05, "loss": 0.0108, "step": 53630 }, { "grad_norm": 0.1589299738407135, "learning_rate": 2.6861455794884904e-05, "loss": 0.0117, "step": 53640 }, { "grad_norm": 0.11031945049762726, "learning_rate": 2.6843135716147373e-05, "loss": 0.0113, "step": 53650 }, { "grad_norm": 0.1999182552099228, "learning_rate": 2.6824819594284556e-05, "loss": 0.0139, "step": 53660 }, { "grad_norm": 0.2752108871936798, "learning_rate": 2.6806507432426275e-05, "loss": 0.0144, "step": 53670 }, { "grad_norm": 0.4778521656990051, "learning_rate": 2.6788199233701512e-05, "loss": 0.0131, "step": 53680 }, { "grad_norm": 0.16257651150226593, "learning_rate": 2.6769895001238652e-05, "loss": 0.0132, "step": 53690 }, { "grad_norm": 0.204511821269989, "learning_rate": 2.67515947381654e-05, "loss": 0.0145, "step": 53700 }, { "grad_norm": 0.16725707054138184, "learning_rate": 2.6733298447608745e-05, "loss": 0.0103, "step": 53710 }, { "grad_norm": 0.2950039505958557, "learning_rate": 2.6715006132695074e-05, "loss": 0.0144, "step": 53720 }, { "grad_norm": 0.19637525081634521, "learning_rate": 2.6696717796549985e-05, "loss": 0.0181, "step": 53730 }, { "grad_norm": 0.3109303116798401, "learning_rate": 2.667843344229848e-05, "loss": 0.0137, "step": 53740 }, { "grad_norm": 0.22222869098186493, "learning_rate": 2.6660153073064846e-05, "loss": 0.0124, "step": 53750 }, { "grad_norm": 0.36374014616012573, "learning_rate": 2.664187669197271e-05, "loss": 0.0174, "step": 53760 }, { "grad_norm": 0.20196174085140228, "learning_rate": 2.6623604302145005e-05, "loss": 0.0108, "step": 53770 }, { "grad_norm": 0.23920227587223053, "learning_rate": 2.6605335906703975e-05, "loss": 0.0102, "step": 53780 }, { "grad_norm": 0.13859425485134125, "learning_rate": 2.65870715087712e-05, "loss": 0.0114, "step": 53790 }, { "grad_norm": 0.19946466386318207, "learning_rate": 2.6568811111467573e-05, "loss": 0.0097, "step": 53800 }, { "grad_norm": 0.19494488835334778, "learning_rate": 2.6550554717913258e-05, "loss": 0.0131, "step": 53810 }, { "grad_norm": 0.169777512550354, "learning_rate": 2.6532302331227805e-05, "loss": 0.0129, "step": 53820 }, { "grad_norm": 0.22072415053844452, "learning_rate": 2.651405395453004e-05, "loss": 0.0156, "step": 53830 }, { "grad_norm": 0.7275663614273071, "learning_rate": 2.6495809590938115e-05, "loss": 0.0155, "step": 53840 }, { "grad_norm": 0.14305898547172546, "learning_rate": 2.64775692435695e-05, "loss": 0.0118, "step": 53850 }, { "grad_norm": 0.22640138864517212, "learning_rate": 2.6459332915540928e-05, "loss": 0.0126, "step": 53860 }, { "grad_norm": 0.17120711505413055, "learning_rate": 2.644110060996856e-05, "loss": 0.0147, "step": 53870 }, { "grad_norm": 0.2030773162841797, "learning_rate": 2.642287232996774e-05, "loss": 0.0102, "step": 53880 }, { "grad_norm": 0.47294706106185913, "learning_rate": 2.6404648078653205e-05, "loss": 0.0108, "step": 53890 }, { "grad_norm": 0.2038566768169403, "learning_rate": 2.6386427859139002e-05, "loss": 0.0104, "step": 53900 }, { "grad_norm": 0.3209647536277771, "learning_rate": 2.63682116745384e-05, "loss": 0.0119, "step": 53910 }, { "grad_norm": 0.1617012768983841, "learning_rate": 2.6349999527964138e-05, "loss": 0.0142, "step": 53920 }, { "grad_norm": 0.16962090134620667, "learning_rate": 2.63317914225281e-05, "loss": 0.0135, "step": 53930 }, { "grad_norm": 0.19772152602672577, "learning_rate": 2.6313587361341585e-05, "loss": 0.0145, "step": 53940 }, { "grad_norm": 0.3459087610244751, "learning_rate": 2.6295387347515165e-05, "loss": 0.0148, "step": 53950 }, { "grad_norm": 0.08794500678777695, "learning_rate": 2.6277191384158727e-05, "loss": 0.0182, "step": 53960 }, { "grad_norm": 0.307452529668808, "learning_rate": 2.6258999474381457e-05, "loss": 0.0138, "step": 53970 }, { "grad_norm": 0.24244801700115204, "learning_rate": 2.624081162129186e-05, "loss": 0.0118, "step": 53980 }, { "grad_norm": 0.17460592091083527, "learning_rate": 2.6222627827997765e-05, "loss": 0.0216, "step": 53990 }, { "grad_norm": 0.2301284372806549, "learning_rate": 2.6204448097606236e-05, "loss": 0.0143, "step": 54000 }, { "grad_norm": 0.14303404092788696, "learning_rate": 2.6186272433223726e-05, "loss": 0.0134, "step": 54010 }, { "grad_norm": 0.14533789455890656, "learning_rate": 2.6168100837955943e-05, "loss": 0.0113, "step": 54020 }, { "grad_norm": 0.16713017225265503, "learning_rate": 2.6149933314907926e-05, "loss": 0.0166, "step": 54030 }, { "grad_norm": 0.22747623920440674, "learning_rate": 2.6131769867184e-05, "loss": 0.0146, "step": 54040 }, { "grad_norm": 0.15225352346897125, "learning_rate": 2.611361049788783e-05, "loss": 0.0115, "step": 54050 }, { "grad_norm": 0.212117001414299, "learning_rate": 2.6095455210122292e-05, "loss": 0.0123, "step": 54060 }, { "grad_norm": 0.11946365237236023, "learning_rate": 2.6077304006989712e-05, "loss": 0.0086, "step": 54070 }, { "grad_norm": 0.23676805198192596, "learning_rate": 2.6059156891591562e-05, "loss": 0.0105, "step": 54080 }, { "grad_norm": 0.17178240418434143, "learning_rate": 2.6041013867028718e-05, "loss": 0.0121, "step": 54090 }, { "grad_norm": 0.18862789869308472, "learning_rate": 2.6022874936401347e-05, "loss": 0.0105, "step": 54100 }, { "grad_norm": 0.18076810240745544, "learning_rate": 2.6004740102808832e-05, "loss": 0.0151, "step": 54110 }, { "grad_norm": 0.2023645043373108, "learning_rate": 2.598660936935e-05, "loss": 0.0108, "step": 54120 }, { "grad_norm": 0.612035870552063, "learning_rate": 2.5968482739122845e-05, "loss": 0.0136, "step": 54130 }, { "grad_norm": 0.21984648704528809, "learning_rate": 2.595036021522472e-05, "loss": 0.0105, "step": 54140 }, { "grad_norm": 0.43704044818878174, "learning_rate": 2.5932241800752278e-05, "loss": 0.014, "step": 54150 }, { "grad_norm": 0.4634752869606018, "learning_rate": 2.5914127498801453e-05, "loss": 0.0236, "step": 54160 }, { "grad_norm": 0.1923566609621048, "learning_rate": 2.5896017312467497e-05, "loss": 0.0117, "step": 54170 }, { "grad_norm": 0.26492759585380554, "learning_rate": 2.587791124484493e-05, "loss": 0.0103, "step": 54180 }, { "grad_norm": 0.1463291347026825, "learning_rate": 2.5859809299027615e-05, "loss": 0.0136, "step": 54190 }, { "grad_norm": 0.17785386741161346, "learning_rate": 2.5841711478108632e-05, "loss": 0.0205, "step": 54200 }, { "grad_norm": 0.27036812901496887, "learning_rate": 2.582361778518043e-05, "loss": 0.0119, "step": 54210 }, { "grad_norm": 0.24210526049137115, "learning_rate": 2.580552822333472e-05, "loss": 0.0145, "step": 54220 }, { "grad_norm": 0.30370190739631653, "learning_rate": 2.578744279566252e-05, "loss": 0.0151, "step": 54230 }, { "grad_norm": 0.11826736479997635, "learning_rate": 2.576936150525413e-05, "loss": 0.0132, "step": 54240 }, { "grad_norm": 0.2132733315229416, "learning_rate": 2.5751284355199168e-05, "loss": 0.0127, "step": 54250 }, { "grad_norm": 0.16121834516525269, "learning_rate": 2.573321134858646e-05, "loss": 0.0083, "step": 54260 }, { "grad_norm": 0.14052073657512665, "learning_rate": 2.5715142488504286e-05, "loss": 0.0091, "step": 54270 }, { "grad_norm": 0.11805948615074158, "learning_rate": 2.5697077778040042e-05, "loss": 0.0132, "step": 54280 }, { "grad_norm": 0.16686825454235077, "learning_rate": 2.5679017220280522e-05, "loss": 0.0211, "step": 54290 }, { "grad_norm": 0.49887895584106445, "learning_rate": 2.5660960818311796e-05, "loss": 0.0147, "step": 54300 }, { "grad_norm": 0.28604215383529663, "learning_rate": 2.564290857521915e-05, "loss": 0.0134, "step": 54310 }, { "grad_norm": 0.5170040130615234, "learning_rate": 2.5624860494087298e-05, "loss": 0.0154, "step": 54320 }, { "grad_norm": 0.10711702704429626, "learning_rate": 2.5606816578000115e-05, "loss": 0.0163, "step": 54330 }, { "grad_norm": 0.2594093382358551, "learning_rate": 2.558877683004082e-05, "loss": 0.0124, "step": 54340 }, { "grad_norm": 0.21337401866912842, "learning_rate": 2.557074125329192e-05, "loss": 0.0096, "step": 54350 }, { "grad_norm": 0.1468411237001419, "learning_rate": 2.5552709850835195e-05, "loss": 0.0124, "step": 54360 }, { "grad_norm": 0.1617211550474167, "learning_rate": 2.5534682625751738e-05, "loss": 0.0082, "step": 54370 }, { "grad_norm": 0.21249645948410034, "learning_rate": 2.551665958112186e-05, "loss": 0.0171, "step": 54380 }, { "grad_norm": 0.16010677814483643, "learning_rate": 2.549864072002527e-05, "loss": 0.0101, "step": 54390 }, { "grad_norm": 0.5576587319374084, "learning_rate": 2.5480626045540858e-05, "loss": 0.0161, "step": 54400 }, { "grad_norm": 0.15892812609672546, "learning_rate": 2.546261556074684e-05, "loss": 0.0089, "step": 54410 }, { "grad_norm": 0.241409033536911, "learning_rate": 2.5444609268720726e-05, "loss": 0.0139, "step": 54420 }, { "grad_norm": 0.24926263093948364, "learning_rate": 2.5426607172539297e-05, "loss": 0.0153, "step": 54430 }, { "grad_norm": 0.10013938695192337, "learning_rate": 2.5408609275278617e-05, "loss": 0.0153, "step": 54440 }, { "grad_norm": 0.12277664989233017, "learning_rate": 2.5390615580014055e-05, "loss": 0.0135, "step": 54450 }, { "grad_norm": 0.1910189539194107, "learning_rate": 2.5372626089820207e-05, "loss": 0.015, "step": 54460 }, { "grad_norm": 0.24891692399978638, "learning_rate": 2.5354640807770997e-05, "loss": 0.0116, "step": 54470 }, { "grad_norm": 0.30448541045188904, "learning_rate": 2.5336659736939622e-05, "loss": 0.0117, "step": 54480 }, { "grad_norm": 0.16859519481658936, "learning_rate": 2.5318682880398554e-05, "loss": 0.0149, "step": 54490 }, { "grad_norm": 0.16212475299835205, "learning_rate": 2.530071024121956e-05, "loss": 0.0152, "step": 54500 }, { "grad_norm": 0.2085094451904297, "learning_rate": 2.5282741822473627e-05, "loss": 0.0097, "step": 54510 }, { "grad_norm": 0.2452608197927475, "learning_rate": 2.526477762723114e-05, "loss": 0.0109, "step": 54520 }, { "grad_norm": 0.17867811024188995, "learning_rate": 2.5246817658561618e-05, "loss": 0.0147, "step": 54530 }, { "grad_norm": 0.12711495161056519, "learning_rate": 2.5228861919533965e-05, "loss": 0.0148, "step": 54540 }, { "grad_norm": 0.1578676849603653, "learning_rate": 2.5210910413216326e-05, "loss": 0.022, "step": 54550 }, { "grad_norm": 0.13789169490337372, "learning_rate": 2.5192963142676086e-05, "loss": 0.0097, "step": 54560 }, { "grad_norm": 0.22240205109119415, "learning_rate": 2.517502011098001e-05, "loss": 0.0118, "step": 54570 }, { "grad_norm": 0.1760321706533432, "learning_rate": 2.5157081321193987e-05, "loss": 0.0118, "step": 54580 }, { "grad_norm": 0.24772189557552338, "learning_rate": 2.5139146776383356e-05, "loss": 0.012, "step": 54590 }, { "grad_norm": 0.18001671135425568, "learning_rate": 2.5121216479612575e-05, "loss": 0.009, "step": 54600 }, { "grad_norm": 0.1945400834083557, "learning_rate": 2.510329043394546e-05, "loss": 0.0174, "step": 54610 }, { "grad_norm": 0.22866033017635345, "learning_rate": 2.508536864244508e-05, "loss": 0.0142, "step": 54620 }, { "grad_norm": 0.23247003555297852, "learning_rate": 2.5067451108173778e-05, "loss": 0.0117, "step": 54630 }, { "grad_norm": 0.2316293716430664, "learning_rate": 2.5049537834193204e-05, "loss": 0.0116, "step": 54640 }, { "grad_norm": 0.17393483221530914, "learning_rate": 2.5031628823564194e-05, "loss": 0.0113, "step": 54650 }, { "grad_norm": 0.18447616696357727, "learning_rate": 2.5013724079346933e-05, "loss": 0.0126, "step": 54660 }, { "grad_norm": 0.17303383350372314, "learning_rate": 2.4995823604600854e-05, "loss": 0.0107, "step": 54670 }, { "grad_norm": 0.2092570662498474, "learning_rate": 2.497792740238465e-05, "loss": 0.0093, "step": 54680 }, { "grad_norm": 0.2466132938861847, "learning_rate": 2.49600354757563e-05, "loss": 0.0143, "step": 54690 }, { "grad_norm": 0.16038274765014648, "learning_rate": 2.494214782777306e-05, "loss": 0.0119, "step": 54700 }, { "grad_norm": 0.2935824692249298, "learning_rate": 2.4924264461491386e-05, "loss": 0.0161, "step": 54710 }, { "grad_norm": 0.25934144854545593, "learning_rate": 2.4906385379967133e-05, "loss": 0.0133, "step": 54720 }, { "grad_norm": 0.19707868993282318, "learning_rate": 2.4888510586255285e-05, "loss": 0.0086, "step": 54730 }, { "grad_norm": 0.20551590621471405, "learning_rate": 2.487064008341018e-05, "loss": 0.0131, "step": 54740 }, { "grad_norm": 0.2907033860683441, "learning_rate": 2.4852773874485407e-05, "loss": 0.0129, "step": 54750 }, { "grad_norm": 0.244459331035614, "learning_rate": 2.483491196253377e-05, "loss": 0.0138, "step": 54760 }, { "grad_norm": 0.34764617681503296, "learning_rate": 2.4817054350607443e-05, "loss": 0.0144, "step": 54770 }, { "grad_norm": 0.27919068932533264, "learning_rate": 2.4799201041757742e-05, "loss": 0.013, "step": 54780 }, { "grad_norm": 0.3145888149738312, "learning_rate": 2.4781352039035373e-05, "loss": 0.0203, "step": 54790 }, { "grad_norm": 0.10365158319473267, "learning_rate": 2.4763507345490194e-05, "loss": 0.0079, "step": 54800 }, { "grad_norm": 0.21749813854694366, "learning_rate": 2.4745666964171386e-05, "loss": 0.0138, "step": 54810 }, { "grad_norm": 0.32274022698402405, "learning_rate": 2.4727830898127387e-05, "loss": 0.0118, "step": 54820 }, { "grad_norm": 0.16489621996879578, "learning_rate": 2.4709999150405895e-05, "loss": 0.0167, "step": 54830 }, { "grad_norm": 0.3222618103027344, "learning_rate": 2.469217172405388e-05, "loss": 0.0122, "step": 54840 }, { "grad_norm": 0.13486707210540771, "learning_rate": 2.4674348622117527e-05, "loss": 0.009, "step": 54850 }, { "grad_norm": 0.2145063877105713, "learning_rate": 2.465652984764234e-05, "loss": 0.0124, "step": 54860 }, { "grad_norm": 0.12495129555463791, "learning_rate": 2.4638715403673056e-05, "loss": 0.008, "step": 54870 }, { "grad_norm": 0.34647953510284424, "learning_rate": 2.4620905293253676e-05, "loss": 0.0117, "step": 54880 }, { "grad_norm": 0.19724857807159424, "learning_rate": 2.4603099519427463e-05, "loss": 0.0139, "step": 54890 }, { "grad_norm": 0.7928087115287781, "learning_rate": 2.458529808523695e-05, "loss": 0.0151, "step": 54900 }, { "grad_norm": 0.22980818152427673, "learning_rate": 2.456750099372387e-05, "loss": 0.0136, "step": 54910 }, { "grad_norm": 0.2739885747432709, "learning_rate": 2.4549708247929327e-05, "loss": 0.0163, "step": 54920 }, { "grad_norm": 0.17485840618610382, "learning_rate": 2.4531919850893554e-05, "loss": 0.0128, "step": 54930 }, { "grad_norm": 0.12938335537910461, "learning_rate": 2.4514135805656125e-05, "loss": 0.0094, "step": 54940 }, { "grad_norm": 0.20704978704452515, "learning_rate": 2.449635611525587e-05, "loss": 0.0148, "step": 54950 }, { "grad_norm": 0.18584871292114258, "learning_rate": 2.447858078273079e-05, "loss": 0.0127, "step": 54960 }, { "grad_norm": 0.24372711777687073, "learning_rate": 2.4460809811118275e-05, "loss": 0.018, "step": 54970 }, { "grad_norm": 0.17666377127170563, "learning_rate": 2.444304320345483e-05, "loss": 0.0102, "step": 54980 }, { "grad_norm": 0.11578820645809174, "learning_rate": 2.4425280962776354e-05, "loss": 0.0136, "step": 54990 }, { "grad_norm": 0.2684764564037323, "learning_rate": 2.4407523092117875e-05, "loss": 0.0109, "step": 55000 }, { "grad_norm": 0.30738046765327454, "learning_rate": 2.438976959451374e-05, "loss": 0.0115, "step": 55010 }, { "grad_norm": 0.17619509994983673, "learning_rate": 2.4372020472997565e-05, "loss": 0.009, "step": 55020 }, { "grad_norm": 0.18139098584651947, "learning_rate": 2.4354275730602122e-05, "loss": 0.011, "step": 55030 }, { "grad_norm": 0.1327485889196396, "learning_rate": 2.433653537035958e-05, "loss": 0.0131, "step": 55040 }, { "grad_norm": 0.1709400862455368, "learning_rate": 2.431879939530123e-05, "loss": 0.0129, "step": 55050 }, { "grad_norm": 0.2292070984840393, "learning_rate": 2.4301067808457684e-05, "loss": 0.0087, "step": 55060 }, { "grad_norm": 0.15451978147029877, "learning_rate": 2.428334061285878e-05, "loss": 0.0117, "step": 55070 }, { "grad_norm": 0.26345738768577576, "learning_rate": 2.426561781153361e-05, "loss": 0.0107, "step": 55080 }, { "grad_norm": 0.25380218029022217, "learning_rate": 2.424789940751052e-05, "loss": 0.0103, "step": 55090 }, { "grad_norm": 0.1920103281736374, "learning_rate": 2.423018540381712e-05, "loss": 0.0114, "step": 55100 }, { "grad_norm": 0.14756475389003754, "learning_rate": 2.421247580348021e-05, "loss": 0.0122, "step": 55110 }, { "grad_norm": 0.19886177778244019, "learning_rate": 2.4194770609525897e-05, "loss": 0.0104, "step": 55120 }, { "grad_norm": 0.22050327062606812, "learning_rate": 2.417706982497951e-05, "loss": 0.013, "step": 55130 }, { "grad_norm": 0.16119594871997833, "learning_rate": 2.415937345286563e-05, "loss": 0.0131, "step": 55140 }, { "grad_norm": 0.37099435925483704, "learning_rate": 2.4141681496208087e-05, "loss": 0.0122, "step": 55150 }, { "grad_norm": 0.19723761081695557, "learning_rate": 2.4123993958029946e-05, "loss": 0.009, "step": 55160 }, { "grad_norm": 0.2200501412153244, "learning_rate": 2.4106310841353548e-05, "loss": 0.0098, "step": 55170 }, { "grad_norm": 0.18410852551460266, "learning_rate": 2.4088632149200398e-05, "loss": 0.0154, "step": 55180 }, { "grad_norm": 0.135929137468338, "learning_rate": 2.4070957884591367e-05, "loss": 0.0124, "step": 55190 }, { "grad_norm": 0.4058077335357666, "learning_rate": 2.4053288050546464e-05, "loss": 0.0168, "step": 55200 }, { "grad_norm": 0.12664289772510529, "learning_rate": 2.403562265008498e-05, "loss": 0.0123, "step": 55210 }, { "grad_norm": 0.3476065993309021, "learning_rate": 2.4017961686225483e-05, "loss": 0.0138, "step": 55220 }, { "grad_norm": 0.19121354818344116, "learning_rate": 2.400030516198568e-05, "loss": 0.0161, "step": 55230 }, { "grad_norm": 0.19984111189842224, "learning_rate": 2.3982653080382673e-05, "loss": 0.0126, "step": 55240 }, { "grad_norm": 0.19423338770866394, "learning_rate": 2.396500544443266e-05, "loss": 0.0101, "step": 55250 }, { "grad_norm": 0.20210199058055878, "learning_rate": 2.3947362257151156e-05, "loss": 0.0114, "step": 55260 }, { "grad_norm": 0.2768977880477905, "learning_rate": 2.39297235215529e-05, "loss": 0.0156, "step": 55270 }, { "grad_norm": 0.17813080549240112, "learning_rate": 2.3912089240651873e-05, "loss": 0.0164, "step": 55280 }, { "grad_norm": 0.10815384984016418, "learning_rate": 2.389445941746129e-05, "loss": 0.0182, "step": 55290 }, { "grad_norm": 0.2627178430557251, "learning_rate": 2.3876834054993625e-05, "loss": 0.017, "step": 55300 }, { "grad_norm": 0.24097618460655212, "learning_rate": 2.3859213156260522e-05, "loss": 0.0139, "step": 55310 }, { "grad_norm": 0.16253399848937988, "learning_rate": 2.3841596724272948e-05, "loss": 0.0129, "step": 55320 }, { "grad_norm": 0.332727313041687, "learning_rate": 2.382398476204106e-05, "loss": 0.0126, "step": 55330 }, { "grad_norm": 0.18167901039123535, "learning_rate": 2.3806377272574254e-05, "loss": 0.0103, "step": 55340 }, { "grad_norm": 0.19635000824928284, "learning_rate": 2.3788774258881174e-05, "loss": 0.0166, "step": 55350 }, { "grad_norm": 0.5829538106918335, "learning_rate": 2.37711757239697e-05, "loss": 0.0161, "step": 55360 }, { "grad_norm": 0.25308430194854736, "learning_rate": 2.3753581670846954e-05, "loss": 0.007, "step": 55370 }, { "grad_norm": 0.11043558269739151, "learning_rate": 2.3735992102519216e-05, "loss": 0.0091, "step": 55380 }, { "grad_norm": 0.29509851336479187, "learning_rate": 2.371840702199215e-05, "loss": 0.0136, "step": 55390 }, { "grad_norm": 0.14804290235042572, "learning_rate": 2.3700826432270494e-05, "loss": 0.0091, "step": 55400 }, { "grad_norm": 0.22644536197185516, "learning_rate": 2.3683250336358326e-05, "loss": 0.0186, "step": 55410 }, { "grad_norm": 0.1633194088935852, "learning_rate": 2.3665678737258923e-05, "loss": 0.0183, "step": 55420 }, { "grad_norm": 0.17259006202220917, "learning_rate": 2.3648111637974745e-05, "loss": 0.013, "step": 55430 }, { "grad_norm": 0.18733102083206177, "learning_rate": 2.36305490415076e-05, "loss": 0.012, "step": 55440 }, { "grad_norm": 0.17693662643432617, "learning_rate": 2.3612990950858392e-05, "loss": 0.0118, "step": 55450 }, { "grad_norm": 0.12332721799612045, "learning_rate": 2.359543736902735e-05, "loss": 0.0126, "step": 55460 }, { "grad_norm": 0.1864248514175415, "learning_rate": 2.3577888299013896e-05, "loss": 0.0094, "step": 55470 }, { "grad_norm": 0.17053353786468506, "learning_rate": 2.3560343743816683e-05, "loss": 0.0109, "step": 55480 }, { "grad_norm": 0.2052556425333023, "learning_rate": 2.354280370643362e-05, "loss": 0.0146, "step": 55490 }, { "grad_norm": 0.2865775525569916, "learning_rate": 2.3525268189861777e-05, "loss": 0.0116, "step": 55500 }, { "grad_norm": 0.34481075406074524, "learning_rate": 2.3507737197097513e-05, "loss": 0.011, "step": 55510 }, { "grad_norm": 0.14159363508224487, "learning_rate": 2.34902107311364e-05, "loss": 0.0117, "step": 55520 }, { "grad_norm": 0.15019555389881134, "learning_rate": 2.3472688794973225e-05, "loss": 0.013, "step": 55530 }, { "grad_norm": 0.21518410742282867, "learning_rate": 2.3455171391602016e-05, "loss": 0.0143, "step": 55540 }, { "grad_norm": 0.20144368708133698, "learning_rate": 2.343765852401601e-05, "loss": 0.008, "step": 55550 }, { "grad_norm": 0.1191081553697586, "learning_rate": 2.342015019520768e-05, "loss": 0.0107, "step": 55560 }, { "grad_norm": 0.15766547620296478, "learning_rate": 2.3402646408168742e-05, "loss": 0.0106, "step": 55570 }, { "grad_norm": 0.2458823323249817, "learning_rate": 2.3385147165890074e-05, "loss": 0.0147, "step": 55580 }, { "grad_norm": 0.1356942504644394, "learning_rate": 2.336765247136184e-05, "loss": 0.0108, "step": 55590 }, { "grad_norm": 0.3072143793106079, "learning_rate": 2.33501623275734e-05, "loss": 0.0185, "step": 55600 }, { "grad_norm": 0.18102659285068512, "learning_rate": 2.333267673751334e-05, "loss": 0.0105, "step": 55610 }, { "grad_norm": 0.16950587928295135, "learning_rate": 2.33151957041695e-05, "loss": 0.0096, "step": 55620 }, { "grad_norm": 0.2490888088941574, "learning_rate": 2.329771923052884e-05, "loss": 0.013, "step": 55630 }, { "grad_norm": 0.2822636067867279, "learning_rate": 2.3280247319577697e-05, "loss": 0.013, "step": 55640 }, { "grad_norm": 0.1637926995754242, "learning_rate": 2.3262779974301473e-05, "loss": 0.0165, "step": 55650 }, { "grad_norm": 0.16484908759593964, "learning_rate": 2.3245317197684895e-05, "loss": 0.0109, "step": 55660 }, { "grad_norm": 0.18371465802192688, "learning_rate": 2.322785899271188e-05, "loss": 0.0127, "step": 55670 }, { "grad_norm": 0.16980235278606415, "learning_rate": 2.3210405362365507e-05, "loss": 0.0195, "step": 55680 }, { "grad_norm": 0.23531058430671692, "learning_rate": 2.31929563096282e-05, "loss": 0.0137, "step": 55690 }, { "grad_norm": 0.1287543624639511, "learning_rate": 2.317551183748146e-05, "loss": 0.0105, "step": 55700 }, { "grad_norm": 0.3015947639942169, "learning_rate": 2.3158071948906103e-05, "loss": 0.0097, "step": 55710 }, { "grad_norm": 0.2045837640762329, "learning_rate": 2.314063664688212e-05, "loss": 0.0105, "step": 55720 }, { "grad_norm": 0.23244978487491608, "learning_rate": 2.3123205934388725e-05, "loss": 0.0079, "step": 55730 }, { "grad_norm": 0.1357385516166687, "learning_rate": 2.310577981440436e-05, "loss": 0.0157, "step": 55740 }, { "grad_norm": 0.19677209854125977, "learning_rate": 2.3088358289906665e-05, "loss": 0.0103, "step": 55750 }, { "grad_norm": 0.1939065009355545, "learning_rate": 2.307094136387252e-05, "loss": 0.0109, "step": 55760 }, { "grad_norm": 0.21112753450870514, "learning_rate": 2.305352903927796e-05, "loss": 0.0155, "step": 55770 }, { "grad_norm": 0.19552744925022125, "learning_rate": 2.303612131909831e-05, "loss": 0.0091, "step": 55780 }, { "grad_norm": 0.29366397857666016, "learning_rate": 2.3018718206308054e-05, "loss": 0.0162, "step": 55790 }, { "grad_norm": 0.28221678733825684, "learning_rate": 2.3001319703880925e-05, "loss": 0.0135, "step": 55800 }, { "grad_norm": 0.19661885499954224, "learning_rate": 2.2983925814789835e-05, "loss": 0.011, "step": 55810 }, { "grad_norm": 0.18983159959316254, "learning_rate": 2.2966536542006957e-05, "loss": 0.0113, "step": 55820 }, { "grad_norm": 0.2031974047422409, "learning_rate": 2.294915188850358e-05, "loss": 0.0097, "step": 55830 }, { "grad_norm": 0.17731337249279022, "learning_rate": 2.2931771857250333e-05, "loss": 0.012, "step": 55840 }, { "grad_norm": 0.12974660098552704, "learning_rate": 2.2914396451216946e-05, "loss": 0.0091, "step": 55850 }, { "grad_norm": 0.33750051259994507, "learning_rate": 2.2897025673372412e-05, "loss": 0.0125, "step": 55860 }, { "grad_norm": 0.11648420244455338, "learning_rate": 2.287965952668494e-05, "loss": 0.0101, "step": 55870 }, { "grad_norm": 0.18073047697544098, "learning_rate": 2.2862298014121873e-05, "loss": 0.0126, "step": 55880 }, { "grad_norm": 0.2802848219871521, "learning_rate": 2.28449411386499e-05, "loss": 0.0114, "step": 55890 }, { "grad_norm": 0.3304993212223053, "learning_rate": 2.282758890323477e-05, "loss": 0.0147, "step": 55900 }, { "grad_norm": 0.09876184910535812, "learning_rate": 2.2810241310841528e-05, "loss": 0.0101, "step": 55910 }, { "grad_norm": 0.1728246510028839, "learning_rate": 2.2792898364434412e-05, "loss": 0.0083, "step": 55920 }, { "grad_norm": 0.28478676080703735, "learning_rate": 2.2775560066976843e-05, "loss": 0.0127, "step": 55930 }, { "grad_norm": 0.13325007259845734, "learning_rate": 2.275822642143147e-05, "loss": 0.0111, "step": 55940 }, { "grad_norm": 0.2377435564994812, "learning_rate": 2.274089743076014e-05, "loss": 0.0097, "step": 55950 }, { "grad_norm": 0.21387389302253723, "learning_rate": 2.2723573097923917e-05, "loss": 0.0098, "step": 55960 }, { "grad_norm": 0.1712123304605484, "learning_rate": 2.2706253425883024e-05, "loss": 0.0128, "step": 55970 }, { "grad_norm": 0.2347630262374878, "learning_rate": 2.2688938417596933e-05, "loss": 0.0112, "step": 55980 }, { "grad_norm": 0.10368086397647858, "learning_rate": 2.2671628076024305e-05, "loss": 0.0105, "step": 55990 }, { "grad_norm": 0.31907325983047485, "learning_rate": 2.2654322404123008e-05, "loss": 0.0107, "step": 56000 }, { "grad_norm": 0.15167205035686493, "learning_rate": 2.2637021404850105e-05, "loss": 0.0116, "step": 56010 }, { "grad_norm": 0.2904290556907654, "learning_rate": 2.2619725081161885e-05, "loss": 0.0132, "step": 56020 }, { "grad_norm": 0.14404673874378204, "learning_rate": 2.260243343601376e-05, "loss": 0.0104, "step": 56030 }, { "grad_norm": 0.1733635663986206, "learning_rate": 2.2585146472360473e-05, "loss": 0.0109, "step": 56040 }, { "grad_norm": 0.2800987958908081, "learning_rate": 2.2567864193155834e-05, "loss": 0.012, "step": 56050 }, { "grad_norm": 0.18163569271564484, "learning_rate": 2.255058660135294e-05, "loss": 0.0139, "step": 56060 }, { "grad_norm": 0.30761486291885376, "learning_rate": 2.253331369990407e-05, "loss": 0.0155, "step": 56070 }, { "grad_norm": 0.12399816513061523, "learning_rate": 2.251604549176063e-05, "loss": 0.0151, "step": 56080 }, { "grad_norm": 0.18196217715740204, "learning_rate": 2.249878197987337e-05, "loss": 0.0139, "step": 56090 }, { "grad_norm": 0.4017944633960724, "learning_rate": 2.2481523167192087e-05, "loss": 0.0133, "step": 56100 }, { "grad_norm": 0.194045290350914, "learning_rate": 2.2464269056665867e-05, "loss": 0.0121, "step": 56110 }, { "grad_norm": 0.17750456929206848, "learning_rate": 2.2447019651242958e-05, "loss": 0.0098, "step": 56120 }, { "grad_norm": 0.16970816254615784, "learning_rate": 2.2429774953870818e-05, "loss": 0.0099, "step": 56130 }, { "grad_norm": 0.2666793763637543, "learning_rate": 2.241253496749611e-05, "loss": 0.0117, "step": 56140 }, { "grad_norm": 0.25171518325805664, "learning_rate": 2.2395299695064614e-05, "loss": 0.0089, "step": 56150 }, { "grad_norm": 0.17585329711437225, "learning_rate": 2.237806913952145e-05, "loss": 0.0119, "step": 56160 }, { "grad_norm": 0.17490644752979279, "learning_rate": 2.2360843303810798e-05, "loss": 0.008, "step": 56170 }, { "grad_norm": 0.18781936168670654, "learning_rate": 2.2343622190876084e-05, "loss": 0.0118, "step": 56180 }, { "grad_norm": 0.21550993621349335, "learning_rate": 2.2326405803659935e-05, "loss": 0.0101, "step": 56190 }, { "grad_norm": 0.22325244545936584, "learning_rate": 2.230919414510416e-05, "loss": 0.0102, "step": 56200 }, { "grad_norm": 0.3080124258995056, "learning_rate": 2.229198721814976e-05, "loss": 0.0158, "step": 56210 }, { "grad_norm": 0.1816055029630661, "learning_rate": 2.2274785025736948e-05, "loss": 0.0109, "step": 56220 }, { "grad_norm": 0.1657799780368805, "learning_rate": 2.225758757080507e-05, "loss": 0.0103, "step": 56230 }, { "grad_norm": 0.29552924633026123, "learning_rate": 2.2240394856292723e-05, "loss": 0.0119, "step": 56240 }, { "grad_norm": 0.5262256264686584, "learning_rate": 2.2223206885137664e-05, "loss": 0.01, "step": 56250 }, { "grad_norm": 0.39895182847976685, "learning_rate": 2.2206023660276853e-05, "loss": 0.0104, "step": 56260 }, { "grad_norm": 0.24569445848464966, "learning_rate": 2.218884518464645e-05, "loss": 0.0124, "step": 56270 }, { "grad_norm": 0.3454674780368805, "learning_rate": 2.2171671461181732e-05, "loss": 0.0121, "step": 56280 }, { "grad_norm": 0.3176354169845581, "learning_rate": 2.2154502492817292e-05, "loss": 0.0144, "step": 56290 }, { "grad_norm": 0.21969634294509888, "learning_rate": 2.2137338282486782e-05, "loss": 0.0088, "step": 56300 }, { "grad_norm": 0.11436370015144348, "learning_rate": 2.2120178833123113e-05, "loss": 0.01, "step": 56310 }, { "grad_norm": 0.17719610035419464, "learning_rate": 2.210302414765838e-05, "loss": 0.0109, "step": 56320 }, { "grad_norm": 0.16115126013755798, "learning_rate": 2.2085874229023808e-05, "loss": 0.008, "step": 56330 }, { "grad_norm": 0.30056869983673096, "learning_rate": 2.2068729080149907e-05, "loss": 0.0116, "step": 56340 }, { "grad_norm": 0.27886486053466797, "learning_rate": 2.205158870396625e-05, "loss": 0.0122, "step": 56350 }, { "grad_norm": 0.16887974739074707, "learning_rate": 2.2034453103401732e-05, "loss": 0.0128, "step": 56360 }, { "grad_norm": 0.11844579130411148, "learning_rate": 2.20173222813843e-05, "loss": 0.0076, "step": 56370 }, { "grad_norm": 0.20926624536514282, "learning_rate": 2.200019624084116e-05, "loss": 0.0125, "step": 56380 }, { "grad_norm": 0.23380884528160095, "learning_rate": 2.1983074984698687e-05, "loss": 0.0104, "step": 56390 }, { "grad_norm": 0.2835746705532074, "learning_rate": 2.1965958515882433e-05, "loss": 0.0189, "step": 56400 }, { "grad_norm": 0.3737744688987732, "learning_rate": 2.1948846837317162e-05, "loss": 0.0138, "step": 56410 }, { "grad_norm": 0.19214215874671936, "learning_rate": 2.1931739951926738e-05, "loss": 0.0095, "step": 56420 }, { "grad_norm": 0.24157726764678955, "learning_rate": 2.191463786263429e-05, "loss": 0.0095, "step": 56430 }, { "grad_norm": 0.1915651112794876, "learning_rate": 2.1897540572362095e-05, "loss": 0.0104, "step": 56440 }, { "grad_norm": 0.3020951747894287, "learning_rate": 2.1880448084031614e-05, "loss": 0.0082, "step": 56450 }, { "grad_norm": 0.12044341117143631, "learning_rate": 2.1863360400563482e-05, "loss": 0.0114, "step": 56460 }, { "grad_norm": 0.20936435461044312, "learning_rate": 2.184627752487754e-05, "loss": 0.0117, "step": 56470 }, { "grad_norm": 0.19574858248233795, "learning_rate": 2.1829199459892725e-05, "loss": 0.0112, "step": 56480 }, { "grad_norm": 0.17122812569141388, "learning_rate": 2.1812126208527282e-05, "loss": 0.0123, "step": 56490 }, { "grad_norm": 0.30987682938575745, "learning_rate": 2.179505777369852e-05, "loss": 0.0145, "step": 56500 }, { "grad_norm": 0.20221979916095734, "learning_rate": 2.1777994158322974e-05, "loss": 0.0145, "step": 56510 }, { "grad_norm": 0.29321229457855225, "learning_rate": 2.176093536531637e-05, "loss": 0.0156, "step": 56520 }, { "grad_norm": 0.1521918773651123, "learning_rate": 2.1743881397593537e-05, "loss": 0.0124, "step": 56530 }, { "grad_norm": 0.4025879502296448, "learning_rate": 2.1726832258068595e-05, "loss": 0.0182, "step": 56540 }, { "grad_norm": 0.4306763708591461, "learning_rate": 2.170978794965472e-05, "loss": 0.0114, "step": 56550 }, { "grad_norm": 0.36872631311416626, "learning_rate": 2.169274847526438e-05, "loss": 0.0091, "step": 56560 }, { "grad_norm": 0.2158876359462738, "learning_rate": 2.1675713837809103e-05, "loss": 0.0107, "step": 56570 }, { "grad_norm": 0.10949110239744186, "learning_rate": 2.1658684040199655e-05, "loss": 0.008, "step": 56580 }, { "grad_norm": 0.1874031126499176, "learning_rate": 2.1641659085345974e-05, "loss": 0.009, "step": 56590 }, { "grad_norm": 0.26400840282440186, "learning_rate": 2.1624638976157154e-05, "loss": 0.0138, "step": 56600 }, { "grad_norm": 0.19063863158226013, "learning_rate": 2.1607623715541476e-05, "loss": 0.018, "step": 56610 }, { "grad_norm": 0.2604871094226837, "learning_rate": 2.159061330640636e-05, "loss": 0.0153, "step": 56620 }, { "grad_norm": 0.21743182837963104, "learning_rate": 2.1573607751658425e-05, "loss": 0.007, "step": 56630 }, { "grad_norm": 0.2658405303955078, "learning_rate": 2.155660705420347e-05, "loss": 0.0132, "step": 56640 }, { "grad_norm": 0.28829991817474365, "learning_rate": 2.153961121694644e-05, "loss": 0.0116, "step": 56650 }, { "grad_norm": 0.21020396053791046, "learning_rate": 2.152262024279145e-05, "loss": 0.0189, "step": 56660 }, { "grad_norm": 0.4070349335670471, "learning_rate": 2.150563413464183e-05, "loss": 0.0119, "step": 56670 }, { "grad_norm": 0.1721438765525818, "learning_rate": 2.1488652895399974e-05, "loss": 0.0098, "step": 56680 }, { "grad_norm": 0.3421782851219177, "learning_rate": 2.1471676527967587e-05, "loss": 0.012, "step": 56690 }, { "grad_norm": 0.15530501306056976, "learning_rate": 2.1454705035245414e-05, "loss": 0.0139, "step": 56700 }, { "grad_norm": 0.218333438038826, "learning_rate": 2.143773842013343e-05, "loss": 0.0113, "step": 56710 }, { "grad_norm": 0.14242130517959595, "learning_rate": 2.1420776685530796e-05, "loss": 0.0127, "step": 56720 }, { "grad_norm": 0.18422959744930267, "learning_rate": 2.1403819834335742e-05, "loss": 0.007, "step": 56730 }, { "grad_norm": 0.14486223459243774, "learning_rate": 2.1386867869445808e-05, "loss": 0.0099, "step": 56740 }, { "grad_norm": 0.15576767921447754, "learning_rate": 2.1369920793757548e-05, "loss": 0.013, "step": 56750 }, { "grad_norm": 0.16660954058170319, "learning_rate": 2.1352978610166828e-05, "loss": 0.0095, "step": 56760 }, { "grad_norm": 0.9501526355743408, "learning_rate": 2.1336041321568546e-05, "loss": 0.0164, "step": 56770 }, { "grad_norm": 0.16798754036426544, "learning_rate": 2.131910893085684e-05, "loss": 0.0114, "step": 56780 }, { "grad_norm": 0.15862339735031128, "learning_rate": 2.1302181440925012e-05, "loss": 0.0103, "step": 56790 }, { "grad_norm": 0.17324215173721313, "learning_rate": 2.128525885466546e-05, "loss": 0.01, "step": 56800 }, { "grad_norm": 0.2492876946926117, "learning_rate": 2.1268341174969847e-05, "loss": 0.0132, "step": 56810 }, { "grad_norm": 0.18068067729473114, "learning_rate": 2.1251428404728894e-05, "loss": 0.0114, "step": 56820 }, { "grad_norm": 0.1773652285337448, "learning_rate": 2.123452054683256e-05, "loss": 0.0114, "step": 56830 }, { "grad_norm": 0.29481467604637146, "learning_rate": 2.1217617604169927e-05, "loss": 0.0147, "step": 56840 }, { "grad_norm": 0.25741058588027954, "learning_rate": 2.120071957962924e-05, "loss": 0.0104, "step": 56850 }, { "grad_norm": 0.17965975403785706, "learning_rate": 2.1183826476097917e-05, "loss": 0.0157, "step": 56860 }, { "grad_norm": 0.22375889122486115, "learning_rate": 2.116693829646254e-05, "loss": 0.0162, "step": 56870 }, { "grad_norm": 0.3390311598777771, "learning_rate": 2.1150055043608806e-05, "loss": 0.0131, "step": 56880 }, { "grad_norm": 0.2115849405527115, "learning_rate": 2.113317672042162e-05, "loss": 0.0182, "step": 56890 }, { "grad_norm": 0.17655624449253082, "learning_rate": 2.1116303329785025e-05, "loss": 0.0079, "step": 56900 }, { "grad_norm": 0.16179314255714417, "learning_rate": 2.109943487458222e-05, "loss": 0.0133, "step": 56910 }, { "grad_norm": 0.1261323243379593, "learning_rate": 2.1082571357695574e-05, "loss": 0.0081, "step": 56920 }, { "grad_norm": 0.1831449270248413, "learning_rate": 2.1065712782006557e-05, "loss": 0.0128, "step": 56930 }, { "grad_norm": 0.10517355054616928, "learning_rate": 2.104885915039591e-05, "loss": 0.0088, "step": 56940 }, { "grad_norm": 0.1989891529083252, "learning_rate": 2.103201046574338e-05, "loss": 0.0126, "step": 56950 }, { "grad_norm": 0.29338914155960083, "learning_rate": 2.1015166730928022e-05, "loss": 0.0132, "step": 56960 }, { "grad_norm": 0.25243404507637024, "learning_rate": 2.0998327948827912e-05, "loss": 0.0136, "step": 56970 }, { "grad_norm": 0.20437099039554596, "learning_rate": 2.0981494122320363e-05, "loss": 0.0121, "step": 56980 }, { "grad_norm": 0.2810509204864502, "learning_rate": 2.0964665254281822e-05, "loss": 0.0107, "step": 56990 }, { "grad_norm": 0.18525883555412292, "learning_rate": 2.094784134758784e-05, "loss": 0.0104, "step": 57000 }, { "grad_norm": 0.11421804130077362, "learning_rate": 2.0931022405113226e-05, "loss": 0.0205, "step": 57010 }, { "grad_norm": 0.1731182038784027, "learning_rate": 2.091420842973183e-05, "loss": 0.01, "step": 57020 }, { "grad_norm": 0.13368748128414154, "learning_rate": 2.0897399424316715e-05, "loss": 0.0081, "step": 57030 }, { "grad_norm": 0.23094698786735535, "learning_rate": 2.0880595391740078e-05, "loss": 0.013, "step": 57040 }, { "grad_norm": 0.19477911293506622, "learning_rate": 2.0863796334873277e-05, "loss": 0.0118, "step": 57050 }, { "grad_norm": 0.14279983937740326, "learning_rate": 2.08470022565868e-05, "loss": 0.0119, "step": 57060 }, { "grad_norm": 0.20084504783153534, "learning_rate": 2.0830213159750317e-05, "loss": 0.0098, "step": 57070 }, { "grad_norm": 0.32637062668800354, "learning_rate": 2.0813429047232596e-05, "loss": 0.0115, "step": 57080 }, { "grad_norm": 0.19224968552589417, "learning_rate": 2.0796649921901594e-05, "loss": 0.0095, "step": 57090 }, { "grad_norm": 0.16371674835681915, "learning_rate": 2.077987578662441e-05, "loss": 0.0094, "step": 57100 }, { "grad_norm": 0.2455969899892807, "learning_rate": 2.0763106644267277e-05, "loss": 0.0119, "step": 57110 }, { "grad_norm": 0.14238552749156952, "learning_rate": 2.0746342497695607e-05, "loss": 0.0054, "step": 57120 }, { "grad_norm": 0.13660509884357452, "learning_rate": 2.0729583349773886e-05, "loss": 0.0072, "step": 57130 }, { "grad_norm": 0.48407280445098877, "learning_rate": 2.0712829203365853e-05, "loss": 0.0135, "step": 57140 }, { "grad_norm": 0.26037153601646423, "learning_rate": 2.0696080061334267e-05, "loss": 0.0135, "step": 57150 }, { "grad_norm": 0.19567163288593292, "learning_rate": 2.067933592654117e-05, "loss": 0.0088, "step": 57160 }, { "grad_norm": 0.24268701672554016, "learning_rate": 2.066259680184763e-05, "loss": 0.011, "step": 57170 }, { "grad_norm": 0.1810804307460785, "learning_rate": 2.0645862690113908e-05, "loss": 0.0099, "step": 57180 }, { "grad_norm": 0.26522934436798096, "learning_rate": 2.0629133594199436e-05, "loss": 0.0177, "step": 57190 }, { "grad_norm": 0.15689115226268768, "learning_rate": 2.0612409516962704e-05, "loss": 0.0113, "step": 57200 }, { "grad_norm": 0.37304922938346863, "learning_rate": 2.0595690461261467e-05, "loss": 0.0119, "step": 57210 }, { "grad_norm": 0.27571630477905273, "learning_rate": 2.0578976429952503e-05, "loss": 0.0131, "step": 57220 }, { "grad_norm": 0.23020382225513458, "learning_rate": 2.0562267425891802e-05, "loss": 0.0106, "step": 57230 }, { "grad_norm": 0.15565523505210876, "learning_rate": 2.0545563451934467e-05, "loss": 0.0126, "step": 57240 }, { "grad_norm": 0.1617077738046646, "learning_rate": 2.0528864510934764e-05, "loss": 0.0118, "step": 57250 }, { "grad_norm": 0.1940421313047409, "learning_rate": 2.0512170605746096e-05, "loss": 0.0183, "step": 57260 }, { "grad_norm": 0.17326217889785767, "learning_rate": 2.0495481739220963e-05, "loss": 0.0129, "step": 57270 }, { "grad_norm": 0.16877515614032745, "learning_rate": 2.0478797914211045e-05, "loss": 0.0128, "step": 57280 }, { "grad_norm": 0.1643219292163849, "learning_rate": 2.046211913356716e-05, "loss": 0.0111, "step": 57290 }, { "grad_norm": 0.09536116570234299, "learning_rate": 2.0445445400139247e-05, "loss": 0.0118, "step": 57300 }, { "grad_norm": 0.3606466054916382, "learning_rate": 2.0428776716776405e-05, "loss": 0.012, "step": 57310 }, { "grad_norm": 0.1165480688214302, "learning_rate": 2.0412113086326856e-05, "loss": 0.0111, "step": 57320 }, { "grad_norm": 0.21180760860443115, "learning_rate": 2.0395454511637918e-05, "loss": 0.0093, "step": 57330 }, { "grad_norm": 0.18709877133369446, "learning_rate": 2.037880099555616e-05, "loss": 0.0124, "step": 57340 }, { "grad_norm": 0.36199310421943665, "learning_rate": 2.0362152540927144e-05, "loss": 0.01, "step": 57350 }, { "grad_norm": 0.16670088469982147, "learning_rate": 2.0345509150595666e-05, "loss": 0.0168, "step": 57360 }, { "grad_norm": 0.2285122573375702, "learning_rate": 2.0328870827405617e-05, "loss": 0.0138, "step": 57370 }, { "grad_norm": 0.163565531373024, "learning_rate": 2.0312237574200043e-05, "loss": 0.0103, "step": 57380 }, { "grad_norm": 0.22420501708984375, "learning_rate": 2.029560939382112e-05, "loss": 0.0109, "step": 57390 }, { "grad_norm": 0.1693551242351532, "learning_rate": 2.0278986289110097e-05, "loss": 0.0099, "step": 57400 }, { "grad_norm": 0.1138201579451561, "learning_rate": 2.0262368262907484e-05, "loss": 0.0113, "step": 57410 }, { "grad_norm": 0.27065616846084595, "learning_rate": 2.024575531805279e-05, "loss": 0.0152, "step": 57420 }, { "grad_norm": 0.18898427486419678, "learning_rate": 2.0229147457384735e-05, "loss": 0.0091, "step": 57430 }, { "grad_norm": 0.2500804364681244, "learning_rate": 2.0212544683741157e-05, "loss": 0.0197, "step": 57440 }, { "grad_norm": 0.19601695239543915, "learning_rate": 2.0195946999958976e-05, "loss": 0.0183, "step": 57450 }, { "grad_norm": 0.15498627722263336, "learning_rate": 2.017935440887434e-05, "loss": 0.0161, "step": 57460 }, { "grad_norm": 0.09227701276540756, "learning_rate": 2.0162766913322423e-05, "loss": 0.011, "step": 57470 }, { "grad_norm": 0.10539406538009644, "learning_rate": 2.0146184516137588e-05, "loss": 0.0111, "step": 57480 }, { "grad_norm": 0.09373365342617035, "learning_rate": 2.012960722015332e-05, "loss": 0.0128, "step": 57490 }, { "grad_norm": 0.21661090850830078, "learning_rate": 2.0113035028202214e-05, "loss": 0.0142, "step": 57500 }, { "grad_norm": 0.20483095943927765, "learning_rate": 2.009646794311602e-05, "loss": 0.0119, "step": 57510 }, { "grad_norm": 0.11617262661457062, "learning_rate": 2.007990596772559e-05, "loss": 0.013, "step": 57520 }, { "grad_norm": 0.13795988261699677, "learning_rate": 2.0063349104860923e-05, "loss": 0.0105, "step": 57530 }, { "grad_norm": 0.17474393546581268, "learning_rate": 2.0046797357351116e-05, "loss": 0.0168, "step": 57540 }, { "grad_norm": 0.16862227022647858, "learning_rate": 2.0030250728024412e-05, "loss": 0.0085, "step": 57550 }, { "grad_norm": 0.30665117502212524, "learning_rate": 2.001370921970819e-05, "loss": 0.0103, "step": 57560 }, { "grad_norm": 0.7893131971359253, "learning_rate": 1.9997172835228932e-05, "loss": 0.0139, "step": 57570 }, { "grad_norm": 0.2058829665184021, "learning_rate": 1.9980641577412262e-05, "loss": 0.0112, "step": 57580 }, { "grad_norm": 0.4216221272945404, "learning_rate": 1.9964115449082925e-05, "loss": 0.0155, "step": 57590 }, { "grad_norm": 0.1619613915681839, "learning_rate": 1.9947594453064742e-05, "loss": 0.0143, "step": 57600 }, { "grad_norm": 0.1926039308309555, "learning_rate": 1.9931078592180774e-05, "loss": 0.0122, "step": 57610 }, { "grad_norm": 0.24107962846755981, "learning_rate": 1.9914567869253065e-05, "loss": 0.0127, "step": 57620 }, { "grad_norm": 0.08979606628417969, "learning_rate": 1.989806228710287e-05, "loss": 0.0164, "step": 57630 }, { "grad_norm": 0.12726067006587982, "learning_rate": 1.9881561848550555e-05, "loss": 0.0097, "step": 57640 }, { "grad_norm": 0.1434669941663742, "learning_rate": 1.9865066556415544e-05, "loss": 0.0139, "step": 57650 }, { "grad_norm": 0.149070143699646, "learning_rate": 1.98485764135165e-05, "loss": 0.0106, "step": 57660 }, { "grad_norm": 0.16180282831192017, "learning_rate": 1.983209142267109e-05, "loss": 0.0084, "step": 57670 }, { "grad_norm": 0.19211798906326294, "learning_rate": 1.9815611586696165e-05, "loss": 0.012, "step": 57680 }, { "grad_norm": 0.18691907823085785, "learning_rate": 1.9799136908407667e-05, "loss": 0.0126, "step": 57690 }, { "grad_norm": 0.17946933209896088, "learning_rate": 1.9782667390620678e-05, "loss": 0.0155, "step": 57700 }, { "grad_norm": 0.278155118227005, "learning_rate": 1.976620303614939e-05, "loss": 0.0083, "step": 57710 }, { "grad_norm": 0.21515819430351257, "learning_rate": 1.9749743847807108e-05, "loss": 0.0128, "step": 57720 }, { "grad_norm": 0.2990531027317047, "learning_rate": 1.9733289828406272e-05, "loss": 0.0095, "step": 57730 }, { "grad_norm": 0.22384986281394958, "learning_rate": 1.9716840980758382e-05, "loss": 0.0094, "step": 57740 }, { "grad_norm": 0.2189284861087799, "learning_rate": 1.9700397307674134e-05, "loss": 0.0108, "step": 57750 }, { "grad_norm": 0.11423980444669724, "learning_rate": 1.968395881196328e-05, "loss": 0.0088, "step": 57760 }, { "grad_norm": 0.2797832190990448, "learning_rate": 1.966752549643473e-05, "loss": 0.0121, "step": 57770 }, { "grad_norm": 0.1264469027519226, "learning_rate": 1.965109736389647e-05, "loss": 0.0099, "step": 57780 }, { "grad_norm": 0.24363256990909576, "learning_rate": 1.9634674417155645e-05, "loss": 0.0136, "step": 57790 }, { "grad_norm": 0.2150445431470871, "learning_rate": 1.9618256659018434e-05, "loss": 0.0244, "step": 57800 }, { "grad_norm": 0.3178436756134033, "learning_rate": 1.9601844092290257e-05, "loss": 0.0096, "step": 57810 }, { "grad_norm": 0.3234167993068695, "learning_rate": 1.9585436719775512e-05, "loss": 0.0133, "step": 57820 }, { "grad_norm": 0.23223665356636047, "learning_rate": 1.9569034544277793e-05, "loss": 0.0122, "step": 57830 }, { "grad_norm": 0.17136500775814056, "learning_rate": 1.9552637568599798e-05, "loss": 0.0089, "step": 57840 }, { "grad_norm": 0.18323728442192078, "learning_rate": 1.953624579554327e-05, "loss": 0.0119, "step": 57850 }, { "grad_norm": 0.1690467745065689, "learning_rate": 1.951985922790918e-05, "loss": 0.0177, "step": 57860 }, { "grad_norm": 0.12569667398929596, "learning_rate": 1.9503477868497505e-05, "loss": 0.0069, "step": 57870 }, { "grad_norm": 0.3386094272136688, "learning_rate": 1.9487101720107375e-05, "loss": 0.0145, "step": 57880 }, { "grad_norm": 0.218462273478508, "learning_rate": 1.9470730785537032e-05, "loss": 0.0136, "step": 57890 }, { "grad_norm": 0.3390091061592102, "learning_rate": 1.9454365067583823e-05, "loss": 0.0118, "step": 57900 }, { "grad_norm": 0.14969313144683838, "learning_rate": 1.9438004569044215e-05, "loss": 0.0091, "step": 57910 }, { "grad_norm": 0.23940591514110565, "learning_rate": 1.9421649292713724e-05, "loss": 0.0255, "step": 57920 }, { "grad_norm": 0.14528261125087738, "learning_rate": 1.9405299241387076e-05, "loss": 0.0135, "step": 57930 }, { "grad_norm": 0.1567799299955368, "learning_rate": 1.9388954417858007e-05, "loss": 0.0101, "step": 57940 }, { "grad_norm": 0.22772426903247833, "learning_rate": 1.9372614824919417e-05, "loss": 0.0111, "step": 57950 }, { "grad_norm": 0.27681946754455566, "learning_rate": 1.9356280465363284e-05, "loss": 0.0131, "step": 57960 }, { "grad_norm": 0.23091086745262146, "learning_rate": 1.9339951341980723e-05, "loss": 0.0098, "step": 57970 }, { "grad_norm": 0.21748819947242737, "learning_rate": 1.9323627457561916e-05, "loss": 0.011, "step": 57980 }, { "grad_norm": 0.16159188747406006, "learning_rate": 1.9307308814896198e-05, "loss": 0.0076, "step": 57990 }, { "grad_norm": 0.11922535300254822, "learning_rate": 1.9290995416771935e-05, "loss": 0.0096, "step": 58000 }, { "grad_norm": 0.14351217448711395, "learning_rate": 1.9274687265976665e-05, "loss": 0.0097, "step": 58010 }, { "grad_norm": 0.300422728061676, "learning_rate": 1.9258384365297e-05, "loss": 0.0112, "step": 58020 }, { "grad_norm": 0.277974009513855, "learning_rate": 1.924208671751866e-05, "loss": 0.0086, "step": 58030 }, { "grad_norm": 0.21405543386936188, "learning_rate": 1.9225794325426492e-05, "loss": 0.0098, "step": 58040 }, { "grad_norm": 0.11109678447246552, "learning_rate": 1.920950719180436e-05, "loss": 0.0081, "step": 58050 }, { "grad_norm": 0.18050090968608856, "learning_rate": 1.919322531943536e-05, "loss": 0.0111, "step": 58060 }, { "grad_norm": 0.2682870626449585, "learning_rate": 1.917694871110157e-05, "loss": 0.0093, "step": 58070 }, { "grad_norm": 0.20584198832511902, "learning_rate": 1.9160677369584234e-05, "loss": 0.0074, "step": 58080 }, { "grad_norm": 0.28865140676498413, "learning_rate": 1.9144411297663694e-05, "loss": 0.0118, "step": 58090 }, { "grad_norm": 0.1276620328426361, "learning_rate": 1.9128150498119328e-05, "loss": 0.0137, "step": 58100 }, { "grad_norm": 0.17709024250507355, "learning_rate": 1.9111894973729726e-05, "loss": 0.0088, "step": 58110 }, { "grad_norm": 0.25028306245803833, "learning_rate": 1.9095644727272454e-05, "loss": 0.0101, "step": 58120 }, { "grad_norm": 0.2480059564113617, "learning_rate": 1.907939976152429e-05, "loss": 0.0135, "step": 58130 }, { "grad_norm": 0.2236786037683487, "learning_rate": 1.906316007926101e-05, "loss": 0.0147, "step": 58140 }, { "grad_norm": 0.14839045703411102, "learning_rate": 1.904692568325755e-05, "loss": 0.0123, "step": 58150 }, { "grad_norm": 0.25779590010643005, "learning_rate": 1.9030696576287925e-05, "loss": 0.0102, "step": 58160 }, { "grad_norm": 0.24591362476348877, "learning_rate": 1.9014472761125242e-05, "loss": 0.0103, "step": 58170 }, { "grad_norm": 0.18338635563850403, "learning_rate": 1.899825424054172e-05, "loss": 0.0095, "step": 58180 }, { "grad_norm": 0.15120866894721985, "learning_rate": 1.898204101730863e-05, "loss": 0.0092, "step": 58190 }, { "grad_norm": 0.15065224468708038, "learning_rate": 1.8965833094196394e-05, "loss": 0.0202, "step": 58200 }, { "grad_norm": 0.15856945514678955, "learning_rate": 1.8949630473974495e-05, "loss": 0.0121, "step": 58210 }, { "grad_norm": 0.15729036927223206, "learning_rate": 1.8933433159411517e-05, "loss": 0.0112, "step": 58220 }, { "grad_norm": 0.32211020588874817, "learning_rate": 1.891724115327514e-05, "loss": 0.0102, "step": 58230 }, { "grad_norm": 0.2370353639125824, "learning_rate": 1.8901054458332152e-05, "loss": 0.0137, "step": 58240 }, { "grad_norm": 0.13425810635089874, "learning_rate": 1.8884873077348364e-05, "loss": 0.0123, "step": 58250 }, { "grad_norm": 0.1583547741174698, "learning_rate": 1.88686970130888e-05, "loss": 0.0141, "step": 58260 }, { "grad_norm": 0.18549956381320953, "learning_rate": 1.8852526268317456e-05, "loss": 0.0127, "step": 58270 }, { "grad_norm": 0.29429128766059875, "learning_rate": 1.883636084579749e-05, "loss": 0.009, "step": 58280 }, { "grad_norm": 0.2445501685142517, "learning_rate": 1.8820200748291145e-05, "loss": 0.0108, "step": 58290 }, { "grad_norm": 0.14604845643043518, "learning_rate": 1.8804045978559686e-05, "loss": 0.0105, "step": 58300 }, { "grad_norm": 0.18490654230117798, "learning_rate": 1.8787896539363594e-05, "loss": 0.0107, "step": 58310 }, { "grad_norm": 0.1646895557641983, "learning_rate": 1.8771752433462298e-05, "loss": 0.0108, "step": 58320 }, { "grad_norm": 0.19661225378513336, "learning_rate": 1.8755613663614447e-05, "loss": 0.0153, "step": 58330 }, { "grad_norm": 0.3181649446487427, "learning_rate": 1.873948023257767e-05, "loss": 0.0124, "step": 58340 }, { "grad_norm": 0.15146875381469727, "learning_rate": 1.8723352143108747e-05, "loss": 0.007, "step": 58350 }, { "grad_norm": 0.2926870882511139, "learning_rate": 1.870722939796352e-05, "loss": 0.0127, "step": 58360 }, { "grad_norm": 0.2142648547887802, "learning_rate": 1.869111199989693e-05, "loss": 0.0114, "step": 58370 }, { "grad_norm": 0.22357000410556793, "learning_rate": 1.8674999951663018e-05, "loss": 0.0096, "step": 58380 }, { "grad_norm": 0.36913934350013733, "learning_rate": 1.8658893256014857e-05, "loss": 0.0123, "step": 58390 }, { "grad_norm": 0.13480721414089203, "learning_rate": 1.8642791915704655e-05, "loss": 0.0086, "step": 58400 }, { "grad_norm": 0.3146457076072693, "learning_rate": 1.8626695933483697e-05, "loss": 0.0102, "step": 58410 }, { "grad_norm": 0.20136027038097382, "learning_rate": 1.8610605312102335e-05, "loss": 0.0107, "step": 58420 }, { "grad_norm": 0.2564440369606018, "learning_rate": 1.8594520054310032e-05, "loss": 0.0145, "step": 58430 }, { "grad_norm": 0.23077642917633057, "learning_rate": 1.8578440162855326e-05, "loss": 0.0099, "step": 58440 }, { "grad_norm": 0.24583901464939117, "learning_rate": 1.8562365640485774e-05, "loss": 0.0128, "step": 58450 }, { "grad_norm": 0.10903871804475784, "learning_rate": 1.854629648994815e-05, "loss": 0.0075, "step": 58460 }, { "grad_norm": 0.14899013936519623, "learning_rate": 1.8530232713988183e-05, "loss": 0.0112, "step": 58470 }, { "grad_norm": 0.29677313566207886, "learning_rate": 1.851417431535074e-05, "loss": 0.0113, "step": 58480 }, { "grad_norm": 0.3051103949546814, "learning_rate": 1.8498121296779785e-05, "loss": 0.0199, "step": 58490 }, { "grad_norm": 0.12163463234901428, "learning_rate": 1.848207366101829e-05, "loss": 0.0118, "step": 58500 }, { "grad_norm": 0.22825871407985687, "learning_rate": 1.8466031410808422e-05, "loss": 0.0127, "step": 58510 }, { "grad_norm": 0.24541105329990387, "learning_rate": 1.844999454889129e-05, "loss": 0.0112, "step": 58520 }, { "grad_norm": 0.2326086163520813, "learning_rate": 1.8433963078007237e-05, "loss": 0.012, "step": 58530 }, { "grad_norm": 0.19710443913936615, "learning_rate": 1.841793700089554e-05, "loss": 0.0097, "step": 58540 }, { "grad_norm": 0.21194329857826233, "learning_rate": 1.8401916320294644e-05, "loss": 0.0117, "step": 58550 }, { "grad_norm": 0.27100783586502075, "learning_rate": 1.838590103894205e-05, "loss": 0.0105, "step": 58560 }, { "grad_norm": 0.10308840870857239, "learning_rate": 1.8369891159574292e-05, "loss": 0.0119, "step": 58570 }, { "grad_norm": 0.24511989951133728, "learning_rate": 1.835388668492708e-05, "loss": 0.0125, "step": 58580 }, { "grad_norm": 0.22321771085262299, "learning_rate": 1.8337887617735095e-05, "loss": 0.0108, "step": 58590 }, { "grad_norm": 0.14713436365127563, "learning_rate": 1.8321893960732157e-05, "loss": 0.0089, "step": 58600 }, { "grad_norm": 0.2417806088924408, "learning_rate": 1.8305905716651138e-05, "loss": 0.0105, "step": 58610 }, { "grad_norm": 0.11138244718313217, "learning_rate": 1.8289922888224e-05, "loss": 0.0105, "step": 58620 }, { "grad_norm": 0.14038299024105072, "learning_rate": 1.8273945478181765e-05, "loss": 0.012, "step": 58630 }, { "grad_norm": 0.17001034319400787, "learning_rate": 1.8257973489254558e-05, "loss": 0.0136, "step": 58640 }, { "grad_norm": 0.1771504133939743, "learning_rate": 1.824200692417152e-05, "loss": 0.0094, "step": 58650 }, { "grad_norm": 0.1837775707244873, "learning_rate": 1.8226045785660912e-05, "loss": 0.0118, "step": 58660 }, { "grad_norm": 0.2634885311126709, "learning_rate": 1.821009007645006e-05, "loss": 0.013, "step": 58670 }, { "grad_norm": 0.20574937760829926, "learning_rate": 1.8194139799265357e-05, "loss": 0.0104, "step": 58680 }, { "grad_norm": 0.20556654036045074, "learning_rate": 1.8178194956832295e-05, "loss": 0.0138, "step": 58690 }, { "grad_norm": 0.28491392731666565, "learning_rate": 1.8162255551875346e-05, "loss": 0.0121, "step": 58700 }, { "grad_norm": 0.16890044510364532, "learning_rate": 1.8146321587118194e-05, "loss": 0.0164, "step": 58710 }, { "grad_norm": 0.44327759742736816, "learning_rate": 1.8130393065283448e-05, "loss": 0.0131, "step": 58720 }, { "grad_norm": 0.10765722393989563, "learning_rate": 1.8114469989092925e-05, "loss": 0.0092, "step": 58730 }, { "grad_norm": 0.25456374883651733, "learning_rate": 1.8098552361267397e-05, "loss": 0.0141, "step": 58740 }, { "grad_norm": 0.13132667541503906, "learning_rate": 1.8082640184526763e-05, "loss": 0.0099, "step": 58750 }, { "grad_norm": 0.20359423756599426, "learning_rate": 1.806673346158999e-05, "loss": 0.0128, "step": 58760 }, { "grad_norm": 0.13535697758197784, "learning_rate": 1.8050832195175067e-05, "loss": 0.0106, "step": 58770 }, { "grad_norm": 0.31192389130592346, "learning_rate": 1.8034936387999136e-05, "loss": 0.0142, "step": 58780 }, { "grad_norm": 0.10968531668186188, "learning_rate": 1.8019046042778315e-05, "loss": 0.0091, "step": 58790 }, { "grad_norm": 0.22912009060382843, "learning_rate": 1.800316116222785e-05, "loss": 0.0095, "step": 58800 }, { "grad_norm": 0.12899331748485565, "learning_rate": 1.7987281749062018e-05, "loss": 0.0265, "step": 58810 }, { "grad_norm": 0.15100878477096558, "learning_rate": 1.7971407805994195e-05, "loss": 0.0196, "step": 58820 }, { "grad_norm": 0.12786570191383362, "learning_rate": 1.7955539335736787e-05, "loss": 0.0112, "step": 58830 }, { "grad_norm": 0.1912384033203125, "learning_rate": 1.7939676341001304e-05, "loss": 0.0172, "step": 58840 }, { "grad_norm": 0.22106994688510895, "learning_rate": 1.7923818824498275e-05, "loss": 0.0097, "step": 58850 }, { "grad_norm": 0.13994593918323517, "learning_rate": 1.7907966788937315e-05, "loss": 0.0118, "step": 58860 }, { "grad_norm": 0.14984731376171112, "learning_rate": 1.7892120237027116e-05, "loss": 0.0148, "step": 58870 }, { "grad_norm": 0.15500950813293457, "learning_rate": 1.7876279171475413e-05, "loss": 0.0125, "step": 58880 }, { "grad_norm": 0.11854762583971024, "learning_rate": 1.7860443594989028e-05, "loss": 0.0095, "step": 58890 }, { "grad_norm": 0.10907375812530518, "learning_rate": 1.784461351027379e-05, "loss": 0.0072, "step": 58900 }, { "grad_norm": 0.3792335093021393, "learning_rate": 1.7828788920034677e-05, "loss": 0.0162, "step": 58910 }, { "grad_norm": 0.2301693707704544, "learning_rate": 1.7812969826975623e-05, "loss": 0.0114, "step": 58920 }, { "grad_norm": 0.17498721182346344, "learning_rate": 1.7797156233799738e-05, "loss": 0.0111, "step": 58930 }, { "grad_norm": 0.24747073650360107, "learning_rate": 1.7781348143209094e-05, "loss": 0.0138, "step": 58940 }, { "grad_norm": 0.15605859458446503, "learning_rate": 1.7765545557904862e-05, "loss": 0.0193, "step": 58950 }, { "grad_norm": 0.30495956540107727, "learning_rate": 1.7749748480587302e-05, "loss": 0.0136, "step": 58960 }, { "grad_norm": 0.1435622274875641, "learning_rate": 1.773395691395564e-05, "loss": 0.0112, "step": 58970 }, { "grad_norm": 0.12053395062685013, "learning_rate": 1.7718170860708305e-05, "loss": 0.0089, "step": 58980 }, { "grad_norm": 0.11395370960235596, "learning_rate": 1.770239032354264e-05, "loss": 0.0084, "step": 58990 }, { "grad_norm": 0.3236495852470398, "learning_rate": 1.7686615305155124e-05, "loss": 0.022, "step": 59000 }, { "grad_norm": 0.16569183766841888, "learning_rate": 1.767084580824128e-05, "loss": 0.0155, "step": 59010 }, { "grad_norm": 0.3492218852043152, "learning_rate": 1.765508183549569e-05, "loss": 0.0132, "step": 59020 }, { "grad_norm": 0.15081386268138885, "learning_rate": 1.763932338961199e-05, "loss": 0.0109, "step": 59030 }, { "grad_norm": 0.24680902063846588, "learning_rate": 1.762357047328284e-05, "loss": 0.0085, "step": 59040 }, { "grad_norm": 0.2446589320898056, "learning_rate": 1.7607823089199997e-05, "loss": 0.0136, "step": 59050 }, { "grad_norm": 0.16067621111869812, "learning_rate": 1.7592081240054265e-05, "loss": 0.0153, "step": 59060 }, { "grad_norm": 0.18438605964183807, "learning_rate": 1.7576344928535488e-05, "loss": 0.0112, "step": 59070 }, { "grad_norm": 0.18809784948825836, "learning_rate": 1.7560614157332573e-05, "loss": 0.0129, "step": 59080 }, { "grad_norm": 0.18326641619205475, "learning_rate": 1.7544888929133495e-05, "loss": 0.0121, "step": 59090 }, { "grad_norm": 0.13839448988437653, "learning_rate": 1.752916924662522e-05, "loss": 0.0091, "step": 59100 }, { "grad_norm": 0.11875376105308533, "learning_rate": 1.751345511249387e-05, "loss": 0.0127, "step": 59110 }, { "grad_norm": 0.13549208641052246, "learning_rate": 1.749774652942452e-05, "loss": 0.0122, "step": 59120 }, { "grad_norm": 0.2614946663379669, "learning_rate": 1.748204350010135e-05, "loss": 0.0122, "step": 59130 }, { "grad_norm": 0.2062717229127884, "learning_rate": 1.7466346027207574e-05, "loss": 0.0126, "step": 59140 }, { "grad_norm": 0.2691473662853241, "learning_rate": 1.745065411342547e-05, "loss": 0.0136, "step": 59150 }, { "grad_norm": 0.18590331077575684, "learning_rate": 1.7434967761436366e-05, "loss": 0.0161, "step": 59160 }, { "grad_norm": 0.2173483669757843, "learning_rate": 1.741928697392058e-05, "loss": 0.0124, "step": 59170 }, { "grad_norm": 0.22755372524261475, "learning_rate": 1.7403611753557597e-05, "loss": 0.0121, "step": 59180 }, { "grad_norm": 0.16885493695735931, "learning_rate": 1.738794210302584e-05, "loss": 0.0073, "step": 59190 }, { "grad_norm": 0.24550455808639526, "learning_rate": 1.7372278025002837e-05, "loss": 0.0106, "step": 59200 }, { "grad_norm": 0.24885240197181702, "learning_rate": 1.7356619522165164e-05, "loss": 0.0119, "step": 59210 }, { "grad_norm": 0.22218240797519684, "learning_rate": 1.7340966597188378e-05, "loss": 0.007, "step": 59220 }, { "grad_norm": 0.11990901827812195, "learning_rate": 1.732531925274722e-05, "loss": 0.0102, "step": 59230 }, { "grad_norm": 0.16313450038433075, "learning_rate": 1.7309677491515318e-05, "loss": 0.0102, "step": 59240 }, { "grad_norm": 0.17798323929309845, "learning_rate": 1.7294041316165455e-05, "loss": 0.0111, "step": 59250 }, { "grad_norm": 0.23858268558979034, "learning_rate": 1.727841072936942e-05, "loss": 0.0121, "step": 59260 }, { "grad_norm": 0.2460704892873764, "learning_rate": 1.7262785733798058e-05, "loss": 0.0104, "step": 59270 }, { "grad_norm": 0.28074485063552856, "learning_rate": 1.7247166332121246e-05, "loss": 0.0104, "step": 59280 }, { "grad_norm": 0.14384983479976654, "learning_rate": 1.7231552527007933e-05, "loss": 0.01, "step": 59290 }, { "grad_norm": 0.18547217547893524, "learning_rate": 1.721594432112606e-05, "loss": 0.0116, "step": 59300 }, { "grad_norm": 0.11292345076799393, "learning_rate": 1.7200341717142655e-05, "loss": 0.0113, "step": 59310 }, { "grad_norm": 0.26250746846199036, "learning_rate": 1.718474471772377e-05, "loss": 0.0119, "step": 59320 }, { "grad_norm": 0.19215470552444458, "learning_rate": 1.7169153325534528e-05, "loss": 0.0103, "step": 59330 }, { "grad_norm": 0.18642394244670868, "learning_rate": 1.7153567543239045e-05, "loss": 0.0139, "step": 59340 }, { "grad_norm": 0.20517556369304657, "learning_rate": 1.7137987373500525e-05, "loss": 0.0108, "step": 59350 }, { "grad_norm": 0.15199555456638336, "learning_rate": 1.7122412818981198e-05, "loss": 0.012, "step": 59360 }, { "grad_norm": 0.21592377126216888, "learning_rate": 1.7106843882342283e-05, "loss": 0.01, "step": 59370 }, { "grad_norm": 0.11511275917291641, "learning_rate": 1.709128056624415e-05, "loss": 0.0096, "step": 59380 }, { "grad_norm": 0.17475168406963348, "learning_rate": 1.70757228733461e-05, "loss": 0.0081, "step": 59390 }, { "grad_norm": 0.16869506239891052, "learning_rate": 1.706017080630653e-05, "loss": 0.0107, "step": 59400 }, { "grad_norm": 0.14987722039222717, "learning_rate": 1.7044624367782873e-05, "loss": 0.0081, "step": 59410 }, { "grad_norm": 0.29438477754592896, "learning_rate": 1.7029083560431553e-05, "loss": 0.012, "step": 59420 }, { "grad_norm": 0.25217485427856445, "learning_rate": 1.701354838690813e-05, "loss": 0.013, "step": 59430 }, { "grad_norm": 0.24133621156215668, "learning_rate": 1.6998018849867087e-05, "loss": 0.0095, "step": 59440 }, { "grad_norm": 0.10338625311851501, "learning_rate": 1.698249495196202e-05, "loss": 0.0105, "step": 59450 }, { "grad_norm": 0.2365940660238266, "learning_rate": 1.6966976695845528e-05, "loss": 0.0127, "step": 59460 }, { "grad_norm": 0.10120467096567154, "learning_rate": 1.6951464084169268e-05, "loss": 0.0104, "step": 59470 }, { "grad_norm": 0.18511690199375153, "learning_rate": 1.6935957119583916e-05, "loss": 0.016, "step": 59480 }, { "grad_norm": 0.1040879637002945, "learning_rate": 1.6920455804739205e-05, "loss": 0.0102, "step": 59490 }, { "grad_norm": 0.23249825835227966, "learning_rate": 1.6904960142283855e-05, "loss": 0.0166, "step": 59500 }, { "grad_norm": 0.13743866980075836, "learning_rate": 1.6889470134865666e-05, "loss": 0.0087, "step": 59510 }, { "grad_norm": 0.23354904353618622, "learning_rate": 1.687398578513145e-05, "loss": 0.0155, "step": 59520 }, { "grad_norm": 0.3175034523010254, "learning_rate": 1.6858507095727066e-05, "loss": 0.0122, "step": 59530 }, { "grad_norm": 0.15056994557380676, "learning_rate": 1.6843034069297403e-05, "loss": 0.0093, "step": 59540 }, { "grad_norm": 0.19717583060264587, "learning_rate": 1.682756670848637e-05, "loss": 0.0139, "step": 59550 }, { "grad_norm": 0.17442114651203156, "learning_rate": 1.6812105015936937e-05, "loss": 0.008, "step": 59560 }, { "grad_norm": 0.19081243872642517, "learning_rate": 1.6796648994291027e-05, "loss": 0.0112, "step": 59570 }, { "grad_norm": 0.18454307317733765, "learning_rate": 1.678119864618973e-05, "loss": 0.0094, "step": 59580 }, { "grad_norm": 0.1650925725698471, "learning_rate": 1.6765753974273023e-05, "loss": 0.0105, "step": 59590 }, { "grad_norm": 0.14467066526412964, "learning_rate": 1.675031498118001e-05, "loss": 0.0108, "step": 59600 }, { "grad_norm": 0.18589156866073608, "learning_rate": 1.6734881669548804e-05, "loss": 0.0098, "step": 59610 }, { "grad_norm": 0.12838716804981232, "learning_rate": 1.671945404201648e-05, "loss": 0.0121, "step": 59620 }, { "grad_norm": 0.16438215970993042, "learning_rate": 1.670403210121927e-05, "loss": 0.009, "step": 59630 }, { "grad_norm": 0.19108663499355316, "learning_rate": 1.6688615849792312e-05, "loss": 0.0077, "step": 59640 }, { "grad_norm": 0.29342764616012573, "learning_rate": 1.6673205290369832e-05, "loss": 0.0117, "step": 59650 }, { "grad_norm": 0.2532418966293335, "learning_rate": 1.6657800425585086e-05, "loss": 0.0132, "step": 59660 }, { "grad_norm": 0.20352467894554138, "learning_rate": 1.6642401258070327e-05, "loss": 0.0174, "step": 59670 }, { "grad_norm": 0.2553826570510864, "learning_rate": 1.6627007790456884e-05, "loss": 0.0115, "step": 59680 }, { "grad_norm": 0.09546016156673431, "learning_rate": 1.6611620025375037e-05, "loss": 0.007, "step": 59690 }, { "grad_norm": 0.20388393104076385, "learning_rate": 1.6596237965454154e-05, "loss": 0.0118, "step": 59700 }, { "grad_norm": 0.2639777958393097, "learning_rate": 1.6580861613322607e-05, "loss": 0.0099, "step": 59710 }, { "grad_norm": 0.11617949604988098, "learning_rate": 1.6565490971607796e-05, "loss": 0.008, "step": 59720 }, { "grad_norm": 0.11344219744205475, "learning_rate": 1.655012604293614e-05, "loss": 0.0125, "step": 59730 }, { "grad_norm": 0.11435549706220627, "learning_rate": 1.653476682993309e-05, "loss": 0.0115, "step": 59740 }, { "grad_norm": 0.3376692533493042, "learning_rate": 1.651941333522311e-05, "loss": 0.0131, "step": 59750 }, { "grad_norm": 0.19319625198841095, "learning_rate": 1.6504065561429715e-05, "loss": 0.0114, "step": 59760 }, { "grad_norm": 0.23813992738723755, "learning_rate": 1.6488723511175385e-05, "loss": 0.0169, "step": 59770 }, { "grad_norm": 0.2583061754703522, "learning_rate": 1.6473387187081668e-05, "loss": 0.0129, "step": 59780 }, { "grad_norm": 0.15234613418579102, "learning_rate": 1.6458056591769123e-05, "loss": 0.0103, "step": 59790 }, { "grad_norm": 0.29302969574928284, "learning_rate": 1.6442731727857335e-05, "loss": 0.0104, "step": 59800 }, { "grad_norm": 0.19112351536750793, "learning_rate": 1.642741259796492e-05, "loss": 0.0099, "step": 59810 }, { "grad_norm": 0.14416386187076569, "learning_rate": 1.641209920470944e-05, "loss": 0.0113, "step": 59820 }, { "grad_norm": 0.2949903607368469, "learning_rate": 1.639679155070762e-05, "loss": 0.0106, "step": 59830 }, { "grad_norm": 0.15486931800842285, "learning_rate": 1.6381489638575048e-05, "loss": 0.0108, "step": 59840 }, { "grad_norm": 0.23115143179893494, "learning_rate": 1.636619347092643e-05, "loss": 0.0086, "step": 59850 }, { "grad_norm": 0.27948465943336487, "learning_rate": 1.6350903050375476e-05, "loss": 0.0095, "step": 59860 }, { "grad_norm": 0.30351853370666504, "learning_rate": 1.6335618379534856e-05, "loss": 0.0101, "step": 59870 }, { "grad_norm": 0.2034374475479126, "learning_rate": 1.6320339461016364e-05, "loss": 0.0085, "step": 59880 }, { "grad_norm": 0.1730884164571762, "learning_rate": 1.6305066297430687e-05, "loss": 0.012, "step": 59890 }, { "grad_norm": 0.5479697585105896, "learning_rate": 1.6289798891387654e-05, "loss": 0.0152, "step": 59900 }, { "grad_norm": 0.1585340052843094, "learning_rate": 1.6274537245495995e-05, "loss": 0.01, "step": 59910 }, { "grad_norm": 0.15782856941223145, "learning_rate": 1.6259281362363527e-05, "loss": 0.0115, "step": 59920 }, { "grad_norm": 0.19080673158168793, "learning_rate": 1.6244031244597068e-05, "loss": 0.0158, "step": 59930 }, { "grad_norm": 0.23232552409172058, "learning_rate": 1.622878689480244e-05, "loss": 0.0097, "step": 59940 }, { "grad_norm": 0.21941156685352325, "learning_rate": 1.6213548315584498e-05, "loss": 0.0145, "step": 59950 }, { "grad_norm": 0.2772652506828308, "learning_rate": 1.6198315509547074e-05, "loss": 0.0106, "step": 59960 }, { "grad_norm": 0.13235586881637573, "learning_rate": 1.6183088479293056e-05, "loss": 0.0102, "step": 59970 }, { "grad_norm": 0.13754485547542572, "learning_rate": 1.6167867227424316e-05, "loss": 0.0091, "step": 59980 }, { "grad_norm": 0.1029188334941864, "learning_rate": 1.615265175654176e-05, "loss": 0.0144, "step": 59990 }, { "grad_norm": 0.2416509985923767, "learning_rate": 1.613744206924529e-05, "loss": 0.012, "step": 60000 }, { "grad_norm": 0.19411130249500275, "learning_rate": 1.6122238168133845e-05, "loss": 0.0108, "step": 60010 }, { "grad_norm": 0.16746662557125092, "learning_rate": 1.6107040055805305e-05, "loss": 0.0098, "step": 60020 }, { "grad_norm": 0.1781017929315567, "learning_rate": 1.6091847734856673e-05, "loss": 0.0122, "step": 60030 }, { "grad_norm": 0.24551819264888763, "learning_rate": 1.6076661207883865e-05, "loss": 0.0109, "step": 60040 }, { "grad_norm": 0.12442407011985779, "learning_rate": 1.6061480477481848e-05, "loss": 0.0099, "step": 60050 }, { "grad_norm": 0.19112500548362732, "learning_rate": 1.604630554624461e-05, "loss": 0.0108, "step": 60060 }, { "grad_norm": 0.25880566239356995, "learning_rate": 1.603113641676509e-05, "loss": 0.0093, "step": 60070 }, { "grad_norm": 0.15215660631656647, "learning_rate": 1.6015973091635338e-05, "loss": 0.0106, "step": 60080 }, { "grad_norm": 0.17408804595470428, "learning_rate": 1.6000815573446288e-05, "loss": 0.0097, "step": 60090 }, { "grad_norm": 0.39603501558303833, "learning_rate": 1.5985663864788002e-05, "loss": 0.0151, "step": 60100 }, { "grad_norm": 0.21255479753017426, "learning_rate": 1.597051796824946e-05, "loss": 0.0128, "step": 60110 }, { "grad_norm": 0.19635790586471558, "learning_rate": 1.5955377886418682e-05, "loss": 0.0117, "step": 60120 }, { "grad_norm": 0.27018222212791443, "learning_rate": 1.5940243621882704e-05, "loss": 0.01, "step": 60130 }, { "grad_norm": 0.24733784794807434, "learning_rate": 1.5925115177227555e-05, "loss": 0.0147, "step": 60140 }, { "grad_norm": 0.1911962628364563, "learning_rate": 1.5909992555038288e-05, "loss": 0.0133, "step": 60150 }, { "grad_norm": 0.19367574155330658, "learning_rate": 1.5894875757898912e-05, "loss": 0.0184, "step": 60160 }, { "grad_norm": 0.11538475751876831, "learning_rate": 1.58797647883925e-05, "loss": 0.0114, "step": 60170 }, { "grad_norm": 0.3562721014022827, "learning_rate": 1.586465964910109e-05, "loss": 0.0111, "step": 60180 }, { "grad_norm": 0.1615115851163864, "learning_rate": 1.5849560342605734e-05, "loss": 0.0105, "step": 60190 }, { "grad_norm": 0.2074180245399475, "learning_rate": 1.583446687148651e-05, "loss": 0.0119, "step": 60200 }, { "grad_norm": 0.21831418573856354, "learning_rate": 1.581937923832248e-05, "loss": 0.0108, "step": 60210 }, { "grad_norm": 0.17526133358478546, "learning_rate": 1.5804297445691663e-05, "loss": 0.0093, "step": 60220 }, { "grad_norm": 0.34244996309280396, "learning_rate": 1.578922149617119e-05, "loss": 0.0114, "step": 60230 }, { "grad_norm": 0.12318448722362518, "learning_rate": 1.5774151392337084e-05, "loss": 0.0069, "step": 60240 }, { "grad_norm": 0.29638147354125977, "learning_rate": 1.575908713676442e-05, "loss": 0.0099, "step": 60250 }, { "grad_norm": 0.27960434556007385, "learning_rate": 1.5744028732027293e-05, "loss": 0.0132, "step": 60260 }, { "grad_norm": 0.2622523307800293, "learning_rate": 1.5728976180698723e-05, "loss": 0.0145, "step": 60270 }, { "grad_norm": 0.12983481585979462, "learning_rate": 1.5713929485350836e-05, "loss": 0.0136, "step": 60280 }, { "grad_norm": 0.17114174365997314, "learning_rate": 1.569888864855465e-05, "loss": 0.0104, "step": 60290 }, { "grad_norm": 0.11765246838331223, "learning_rate": 1.568385367288028e-05, "loss": 0.0072, "step": 60300 }, { "grad_norm": 0.15975303947925568, "learning_rate": 1.5668824560896755e-05, "loss": 0.0125, "step": 60310 }, { "grad_norm": 0.2113587111234665, "learning_rate": 1.5653801315172156e-05, "loss": 0.0112, "step": 60320 }, { "grad_norm": 0.19994352757930756, "learning_rate": 1.5638783938273554e-05, "loss": 0.0104, "step": 60330 }, { "grad_norm": 0.2901158332824707, "learning_rate": 1.5623772432766966e-05, "loss": 0.0149, "step": 60340 }, { "grad_norm": 0.20105165243148804, "learning_rate": 1.5608766801217507e-05, "loss": 0.0126, "step": 60350 }, { "grad_norm": 0.08034859597682953, "learning_rate": 1.5593767046189184e-05, "loss": 0.0085, "step": 60360 }, { "grad_norm": 0.12619633972644806, "learning_rate": 1.5578773170245064e-05, "loss": 0.0092, "step": 60370 }, { "grad_norm": 0.3779497444629669, "learning_rate": 1.5563785175947182e-05, "loss": 0.0131, "step": 60380 }, { "grad_norm": 0.11116298288106918, "learning_rate": 1.5548803065856582e-05, "loss": 0.0088, "step": 60390 }, { "grad_norm": 0.3993551731109619, "learning_rate": 1.55338268425333e-05, "loss": 0.0106, "step": 60400 }, { "grad_norm": 0.15877188742160797, "learning_rate": 1.5518856508536373e-05, "loss": 0.0117, "step": 60410 }, { "grad_norm": 0.3696359097957611, "learning_rate": 1.5503892066423786e-05, "loss": 0.0123, "step": 60420 }, { "grad_norm": 0.33199912309646606, "learning_rate": 1.548893351875258e-05, "loss": 0.017, "step": 60430 }, { "grad_norm": 0.32414892315864563, "learning_rate": 1.547398086807875e-05, "loss": 0.0114, "step": 60440 }, { "grad_norm": 0.27479785680770874, "learning_rate": 1.5459034116957305e-05, "loss": 0.018, "step": 60450 }, { "grad_norm": 0.2022894024848938, "learning_rate": 1.544409326794225e-05, "loss": 0.0254, "step": 60460 }, { "grad_norm": 0.18090300261974335, "learning_rate": 1.542915832358651e-05, "loss": 0.0124, "step": 60470 }, { "grad_norm": 0.1773895025253296, "learning_rate": 1.541422928644214e-05, "loss": 0.0077, "step": 60480 }, { "grad_norm": 0.29521363973617554, "learning_rate": 1.5399306159060024e-05, "loss": 0.0158, "step": 60490 }, { "grad_norm": 0.12154003232717514, "learning_rate": 1.5384388943990185e-05, "loss": 0.0081, "step": 60500 }, { "grad_norm": 0.21516850590705872, "learning_rate": 1.5369477643781526e-05, "loss": 0.0136, "step": 60510 }, { "grad_norm": 0.14594684541225433, "learning_rate": 1.5354572260981985e-05, "loss": 0.0124, "step": 60520 }, { "grad_norm": 0.15501444041728973, "learning_rate": 1.533967279813851e-05, "loss": 0.0068, "step": 60530 }, { "grad_norm": 0.36942964792251587, "learning_rate": 1.5324779257796956e-05, "loss": 0.0111, "step": 60540 }, { "grad_norm": 0.20513014495372772, "learning_rate": 1.5309891642502293e-05, "loss": 0.0101, "step": 60550 }, { "grad_norm": 0.2568749785423279, "learning_rate": 1.5295009954798357e-05, "loss": 0.0094, "step": 60560 }, { "grad_norm": 0.14774736762046814, "learning_rate": 1.5280134197228036e-05, "loss": 0.0119, "step": 60570 }, { "grad_norm": 0.09675546735525131, "learning_rate": 1.526526437233319e-05, "loss": 0.0114, "step": 60580 }, { "grad_norm": 0.12961329519748688, "learning_rate": 1.525040048265466e-05, "loss": 0.0099, "step": 60590 }, { "grad_norm": 0.1649303436279297, "learning_rate": 1.5235542530732288e-05, "loss": 0.0097, "step": 60600 }, { "grad_norm": 0.16344742476940155, "learning_rate": 1.5220690519104901e-05, "loss": 0.014, "step": 60610 }, { "grad_norm": 0.11100573092699051, "learning_rate": 1.5205844450310275e-05, "loss": 0.0125, "step": 60620 }, { "grad_norm": 0.1537809818983078, "learning_rate": 1.519100432688521e-05, "loss": 0.0096, "step": 60630 }, { "grad_norm": 0.1994544267654419, "learning_rate": 1.5176170151365476e-05, "loss": 0.011, "step": 60640 }, { "grad_norm": 0.09231479465961456, "learning_rate": 1.5161341926285833e-05, "loss": 0.0084, "step": 60650 }, { "grad_norm": 0.15130497515201569, "learning_rate": 1.5146519654180025e-05, "loss": 0.0082, "step": 60660 }, { "grad_norm": 0.2558031976222992, "learning_rate": 1.5131703337580739e-05, "loss": 0.0151, "step": 60670 }, { "grad_norm": 0.10521981865167618, "learning_rate": 1.5116892979019731e-05, "loss": 0.0114, "step": 60680 }, { "grad_norm": 0.15158866345882416, "learning_rate": 1.5102088581027623e-05, "loss": 0.0103, "step": 60690 }, { "grad_norm": 0.3115658462047577, "learning_rate": 1.5087290146134154e-05, "loss": 0.0073, "step": 60700 }, { "grad_norm": 0.30079585313796997, "learning_rate": 1.5072497676867915e-05, "loss": 0.0119, "step": 60710 }, { "grad_norm": 0.1944517195224762, "learning_rate": 1.505771117575655e-05, "loss": 0.015, "step": 60720 }, { "grad_norm": 0.2040664553642273, "learning_rate": 1.5042930645326691e-05, "loss": 0.0125, "step": 60730 }, { "grad_norm": 0.189458966255188, "learning_rate": 1.5028156088103879e-05, "loss": 0.0128, "step": 60740 }, { "grad_norm": 0.3834591209888458, "learning_rate": 1.5013387506612735e-05, "loss": 0.0139, "step": 60750 }, { "grad_norm": 0.10493011772632599, "learning_rate": 1.4998624903376767e-05, "loss": 0.0127, "step": 60760 }, { "grad_norm": 0.22123479843139648, "learning_rate": 1.4983868280918517e-05, "loss": 0.0112, "step": 60770 }, { "grad_norm": 0.22215905785560608, "learning_rate": 1.4969117641759478e-05, "loss": 0.0137, "step": 60780 }, { "grad_norm": 0.10038489103317261, "learning_rate": 1.495437298842014e-05, "loss": 0.0091, "step": 60790 }, { "grad_norm": 0.21705788373947144, "learning_rate": 1.4939634323419976e-05, "loss": 0.0095, "step": 60800 }, { "grad_norm": 0.2243167757987976, "learning_rate": 1.4924901649277384e-05, "loss": 0.0148, "step": 60810 }, { "grad_norm": 0.18841859698295593, "learning_rate": 1.4910174968509793e-05, "loss": 0.0096, "step": 60820 }, { "grad_norm": 0.12743787467479706, "learning_rate": 1.4895454283633598e-05, "loss": 0.0087, "step": 60830 }, { "grad_norm": 0.45276838541030884, "learning_rate": 1.488073959716415e-05, "loss": 0.0103, "step": 60840 }, { "grad_norm": 0.2537676990032196, "learning_rate": 1.4866030911615791e-05, "loss": 0.0105, "step": 60850 }, { "grad_norm": 0.23147797584533691, "learning_rate": 1.4851328229501849e-05, "loss": 0.0095, "step": 60860 }, { "grad_norm": 0.14206019043922424, "learning_rate": 1.4836631553334562e-05, "loss": 0.009, "step": 60870 }, { "grad_norm": 0.1585252434015274, "learning_rate": 1.4821940885625251e-05, "loss": 0.0117, "step": 60880 }, { "grad_norm": 0.27763432264328003, "learning_rate": 1.4807256228884109e-05, "loss": 0.012, "step": 60890 }, { "grad_norm": 0.1706155240535736, "learning_rate": 1.4792577585620353e-05, "loss": 0.0064, "step": 60900 }, { "grad_norm": 0.17559781670570374, "learning_rate": 1.4777904958342164e-05, "loss": 0.0105, "step": 60910 }, { "grad_norm": 0.12961037456989288, "learning_rate": 1.4763238349556691e-05, "loss": 0.0092, "step": 60920 }, { "grad_norm": 0.23988474905490875, "learning_rate": 1.4748577761770072e-05, "loss": 0.012, "step": 60930 }, { "grad_norm": 0.13637515902519226, "learning_rate": 1.4733923197487354e-05, "loss": 0.0107, "step": 60940 }, { "grad_norm": 0.16023284196853638, "learning_rate": 1.4719274659212662e-05, "loss": 0.0129, "step": 60950 }, { "grad_norm": 0.1178862527012825, "learning_rate": 1.4704632149448983e-05, "loss": 0.0098, "step": 60960 }, { "grad_norm": 0.14011171460151672, "learning_rate": 1.4689995670698343e-05, "loss": 0.0121, "step": 60970 }, { "grad_norm": 0.16814415156841278, "learning_rate": 1.4675365225461728e-05, "loss": 0.0105, "step": 60980 }, { "grad_norm": 0.2283068597316742, "learning_rate": 1.4660740816239032e-05, "loss": 0.0127, "step": 60990 }, { "grad_norm": 0.24735864996910095, "learning_rate": 1.464612244552923e-05, "loss": 0.0095, "step": 61000 }, { "grad_norm": 0.12578453123569489, "learning_rate": 1.4631510115830161e-05, "loss": 0.008, "step": 61010 }, { "grad_norm": 0.20865552127361298, "learning_rate": 1.4616903829638679e-05, "loss": 0.0105, "step": 61020 }, { "grad_norm": 0.13102445006370544, "learning_rate": 1.46023035894506e-05, "loss": 0.0098, "step": 61030 }, { "grad_norm": 0.29085639119148254, "learning_rate": 1.4587709397760713e-05, "loss": 0.0133, "step": 61040 }, { "grad_norm": 0.19411826133728027, "learning_rate": 1.4573121257062755e-05, "loss": 0.0105, "step": 61050 }, { "grad_norm": 0.38235580921173096, "learning_rate": 1.4558539169849472e-05, "loss": 0.0162, "step": 61060 }, { "grad_norm": 0.3493616282939911, "learning_rate": 1.4543963138612499e-05, "loss": 0.0131, "step": 61070 }, { "grad_norm": 0.20432500541210175, "learning_rate": 1.4529393165842498e-05, "loss": 0.009, "step": 61080 }, { "grad_norm": 0.2659079134464264, "learning_rate": 1.4514829254029084e-05, "loss": 0.0104, "step": 61090 }, { "grad_norm": 0.11254163086414337, "learning_rate": 1.450027140566083e-05, "loss": 0.0093, "step": 61100 }, { "grad_norm": 0.14847010374069214, "learning_rate": 1.4485719623225268e-05, "loss": 0.0106, "step": 61110 }, { "grad_norm": 0.14267978072166443, "learning_rate": 1.4471173909208912e-05, "loss": 0.0105, "step": 61120 }, { "grad_norm": 0.14003105461597443, "learning_rate": 1.4456634266097236e-05, "loss": 0.0128, "step": 61130 }, { "grad_norm": 0.19455096125602722, "learning_rate": 1.444210069637461e-05, "loss": 0.0123, "step": 61140 }, { "grad_norm": 0.3101329803466797, "learning_rate": 1.4427573202524502e-05, "loss": 0.0108, "step": 61150 }, { "grad_norm": 0.15150444209575653, "learning_rate": 1.4413051787029208e-05, "loss": 0.0149, "step": 61160 }, { "grad_norm": 0.15350545942783356, "learning_rate": 1.439853645237006e-05, "loss": 0.0131, "step": 61170 }, { "grad_norm": 0.12033756822347641, "learning_rate": 1.4384027201027344e-05, "loss": 0.0096, "step": 61180 }, { "grad_norm": 0.12941478192806244, "learning_rate": 1.4369524035480253e-05, "loss": 0.0106, "step": 61190 }, { "grad_norm": 0.1760236918926239, "learning_rate": 1.4355026958207035e-05, "loss": 0.0071, "step": 61200 }, { "grad_norm": 0.15137510001659393, "learning_rate": 1.434053597168481e-05, "loss": 0.01, "step": 61210 }, { "grad_norm": 0.298265665769577, "learning_rate": 1.432605107838969e-05, "loss": 0.0144, "step": 61220 }, { "grad_norm": 0.24922417104244232, "learning_rate": 1.4311572280796759e-05, "loss": 0.01, "step": 61230 }, { "grad_norm": 0.2127961814403534, "learning_rate": 1.4297099581380047e-05, "loss": 0.0151, "step": 61240 }, { "grad_norm": 0.2173650711774826, "learning_rate": 1.4282632982612538e-05, "loss": 0.0127, "step": 61250 }, { "grad_norm": 0.1564674973487854, "learning_rate": 1.4268172486966197e-05, "loss": 0.0139, "step": 61260 }, { "grad_norm": 0.14804649353027344, "learning_rate": 1.4253718096911894e-05, "loss": 0.0136, "step": 61270 }, { "grad_norm": 0.2455967217683792, "learning_rate": 1.4239269814919504e-05, "loss": 0.0159, "step": 61280 }, { "grad_norm": 0.4474382996559143, "learning_rate": 1.4224827643457843e-05, "loss": 0.0151, "step": 61290 }, { "grad_norm": 0.13726212084293365, "learning_rate": 1.4210391584994686e-05, "loss": 0.0131, "step": 61300 }, { "grad_norm": 0.12316098809242249, "learning_rate": 1.4195961641996763e-05, "loss": 0.0104, "step": 61310 }, { "grad_norm": 0.13221316039562225, "learning_rate": 1.4181537816929751e-05, "loss": 0.0127, "step": 61320 }, { "grad_norm": 0.14704474806785583, "learning_rate": 1.4167120112258303e-05, "loss": 0.0141, "step": 61330 }, { "grad_norm": 0.10979632288217545, "learning_rate": 1.4152708530445963e-05, "loss": 0.0095, "step": 61340 }, { "grad_norm": 0.14879490435123444, "learning_rate": 1.4138303073955344e-05, "loss": 0.0104, "step": 61350 }, { "grad_norm": 0.23321302235126495, "learning_rate": 1.4123903745247897e-05, "loss": 0.0092, "step": 61360 }, { "grad_norm": 0.10056035965681076, "learning_rate": 1.4109510546784082e-05, "loss": 0.0063, "step": 61370 }, { "grad_norm": 0.345564603805542, "learning_rate": 1.4095123481023325e-05, "loss": 0.0152, "step": 61380 }, { "grad_norm": 0.1351022571325302, "learning_rate": 1.4080742550423936e-05, "loss": 0.01, "step": 61390 }, { "grad_norm": 0.09642086923122406, "learning_rate": 1.4066367757443278e-05, "loss": 0.0131, "step": 61400 }, { "grad_norm": 0.13296827673912048, "learning_rate": 1.4051999104537572e-05, "loss": 0.01, "step": 61410 }, { "grad_norm": 0.21491970121860504, "learning_rate": 1.4037636594162034e-05, "loss": 0.0113, "step": 61420 }, { "grad_norm": 0.14674989879131317, "learning_rate": 1.402328022877083e-05, "loss": 0.0067, "step": 61430 }, { "grad_norm": 0.136611670255661, "learning_rate": 1.4008930010817073e-05, "loss": 0.0146, "step": 61440 }, { "grad_norm": 0.20835629105567932, "learning_rate": 1.3994585942752831e-05, "loss": 0.0089, "step": 61450 }, { "grad_norm": 0.4674297571182251, "learning_rate": 1.398024802702909e-05, "loss": 0.0124, "step": 61460 }, { "grad_norm": 0.25154948234558105, "learning_rate": 1.3965916266095819e-05, "loss": 0.0106, "step": 61470 }, { "grad_norm": 0.2134029120206833, "learning_rate": 1.395159066240192e-05, "loss": 0.0101, "step": 61480 }, { "grad_norm": 0.25579124689102173, "learning_rate": 1.3937271218395248e-05, "loss": 0.0085, "step": 61490 }, { "grad_norm": 0.21669447422027588, "learning_rate": 1.3922957936522613e-05, "loss": 0.0132, "step": 61500 }, { "grad_norm": 0.20921199023723602, "learning_rate": 1.3908650819229758e-05, "loss": 0.0098, "step": 61510 }, { "grad_norm": 0.1736782044172287, "learning_rate": 1.389434986896137e-05, "loss": 0.0088, "step": 61520 }, { "grad_norm": 0.1995721161365509, "learning_rate": 1.388005508816112e-05, "loss": 0.0083, "step": 61530 }, { "grad_norm": 0.30025652050971985, "learning_rate": 1.3865766479271557e-05, "loss": 0.014, "step": 61540 }, { "grad_norm": 0.16512787342071533, "learning_rate": 1.3851484044734225e-05, "loss": 0.012, "step": 61550 }, { "grad_norm": 0.14923222362995148, "learning_rate": 1.3837207786989608e-05, "loss": 0.0194, "step": 61560 }, { "grad_norm": 0.13701167702674866, "learning_rate": 1.382293770847713e-05, "loss": 0.0095, "step": 61570 }, { "grad_norm": 0.18528573215007782, "learning_rate": 1.3808673811635159e-05, "loss": 0.0113, "step": 61580 }, { "grad_norm": 0.20809119939804077, "learning_rate": 1.3794416098900975e-05, "loss": 0.0135, "step": 61590 }, { "grad_norm": 0.11878707259893417, "learning_rate": 1.3780164572710886e-05, "loss": 0.0155, "step": 61600 }, { "grad_norm": 0.16535726189613342, "learning_rate": 1.3765919235500035e-05, "loss": 0.0132, "step": 61610 }, { "grad_norm": 0.2893899977207184, "learning_rate": 1.375168008970259e-05, "loss": 0.0101, "step": 61620 }, { "grad_norm": 0.32828423380851746, "learning_rate": 1.3737447137751635e-05, "loss": 0.0076, "step": 61630 }, { "grad_norm": 0.23539121448993683, "learning_rate": 1.372322038207915e-05, "loss": 0.0077, "step": 61640 }, { "grad_norm": 0.2799283266067505, "learning_rate": 1.3708999825116159e-05, "loss": 0.017, "step": 61650 }, { "grad_norm": 0.1849956512451172, "learning_rate": 1.3694785469292526e-05, "loss": 0.015, "step": 61660 }, { "grad_norm": 0.15100045502185822, "learning_rate": 1.3680577317037101e-05, "loss": 0.0114, "step": 61670 }, { "grad_norm": 0.2796342074871063, "learning_rate": 1.3666375370777678e-05, "loss": 0.0092, "step": 61680 }, { "grad_norm": 0.2126503735780716, "learning_rate": 1.3652179632940981e-05, "loss": 0.0132, "step": 61690 }, { "grad_norm": 0.0978236049413681, "learning_rate": 1.3637990105952664e-05, "loss": 0.0115, "step": 61700 }, { "grad_norm": 0.17987719178199768, "learning_rate": 1.3623806792237337e-05, "loss": 0.0104, "step": 61710 }, { "grad_norm": 0.17612381279468536, "learning_rate": 1.3609629694218551e-05, "loss": 0.0084, "step": 61720 }, { "grad_norm": 0.20729897916316986, "learning_rate": 1.3595458814318762e-05, "loss": 0.0077, "step": 61730 }, { "grad_norm": 0.19113312661647797, "learning_rate": 1.3581294154959389e-05, "loss": 0.0109, "step": 61740 }, { "grad_norm": 0.2091827243566513, "learning_rate": 1.3567135718560792e-05, "loss": 0.0107, "step": 61750 }, { "grad_norm": 0.09657852351665497, "learning_rate": 1.3552983507542261e-05, "loss": 0.0101, "step": 61760 }, { "grad_norm": 0.21857166290283203, "learning_rate": 1.3538837524322023e-05, "loss": 0.0141, "step": 61770 }, { "grad_norm": 0.32287272810935974, "learning_rate": 1.3524697771317251e-05, "loss": 0.0125, "step": 61780 }, { "grad_norm": 0.11649960279464722, "learning_rate": 1.3510564250943997e-05, "loss": 0.0106, "step": 61790 }, { "grad_norm": 0.16343021392822266, "learning_rate": 1.3496436965617353e-05, "loss": 0.0085, "step": 61800 }, { "grad_norm": 0.1637071818113327, "learning_rate": 1.3482315917751243e-05, "loss": 0.0122, "step": 61810 }, { "grad_norm": 0.3934263288974762, "learning_rate": 1.3468201109758582e-05, "loss": 0.0112, "step": 61820 }, { "grad_norm": 0.329091340303421, "learning_rate": 1.3454092544051222e-05, "loss": 0.0121, "step": 61830 }, { "grad_norm": 0.18033386766910553, "learning_rate": 1.3439990223039878e-05, "loss": 0.0103, "step": 61840 }, { "grad_norm": 0.15109050273895264, "learning_rate": 1.3425894149134321e-05, "loss": 0.01, "step": 61850 }, { "grad_norm": 0.25837066769599915, "learning_rate": 1.3411804324743132e-05, "loss": 0.0134, "step": 61860 }, { "grad_norm": 0.14659270644187927, "learning_rate": 1.3397720752273896e-05, "loss": 0.0125, "step": 61870 }, { "grad_norm": 0.1275964379310608, "learning_rate": 1.3383643434133108e-05, "loss": 0.0087, "step": 61880 }, { "grad_norm": 0.23173357546329498, "learning_rate": 1.3369572372726196e-05, "loss": 0.0119, "step": 61890 }, { "grad_norm": 0.10472103953361511, "learning_rate": 1.3355507570457525e-05, "loss": 0.0085, "step": 61900 }, { "grad_norm": 0.25926703214645386, "learning_rate": 1.334144902973038e-05, "loss": 0.009, "step": 61910 }, { "grad_norm": 0.35922232270240784, "learning_rate": 1.3327396752946997e-05, "loss": 0.0108, "step": 61920 }, { "grad_norm": 0.12606406211853027, "learning_rate": 1.3313350742508496e-05, "loss": 0.0129, "step": 61930 }, { "grad_norm": 0.2709902226924896, "learning_rate": 1.3299311000814973e-05, "loss": 0.0134, "step": 61940 }, { "grad_norm": 0.073287732899189, "learning_rate": 1.3285277530265432e-05, "loss": 0.0142, "step": 61950 }, { "grad_norm": 0.3525666296482086, "learning_rate": 1.3271250333257813e-05, "loss": 0.0119, "step": 61960 }, { "grad_norm": 0.11325156688690186, "learning_rate": 1.3257229412188977e-05, "loss": 0.0086, "step": 61970 }, { "grad_norm": 0.24395617842674255, "learning_rate": 1.3243214769454726e-05, "loss": 0.0104, "step": 61980 }, { "grad_norm": 0.1353428065776825, "learning_rate": 1.3229206407449751e-05, "loss": 0.0081, "step": 61990 }, { "grad_norm": 0.27335670590400696, "learning_rate": 1.3215204328567742e-05, "loss": 0.008, "step": 62000 }, { "grad_norm": 0.1816713809967041, "learning_rate": 1.3201208535201232e-05, "loss": 0.0115, "step": 62010 }, { "grad_norm": 0.13149648904800415, "learning_rate": 1.3187219029741737e-05, "loss": 0.0112, "step": 62020 }, { "grad_norm": 0.19639058411121368, "learning_rate": 1.3173235814579693e-05, "loss": 0.0077, "step": 62030 }, { "grad_norm": 0.19723519682884216, "learning_rate": 1.3159258892104398e-05, "loss": 0.0119, "step": 62040 }, { "grad_norm": 0.154967799782753, "learning_rate": 1.3145288264704198e-05, "loss": 0.0131, "step": 62050 }, { "grad_norm": 0.13220266997814178, "learning_rate": 1.3131323934766237e-05, "loss": 0.0089, "step": 62060 }, { "grad_norm": 0.1174975112080574, "learning_rate": 1.311736590467666e-05, "loss": 0.0066, "step": 62070 }, { "grad_norm": 0.10033011436462402, "learning_rate": 1.3103414176820505e-05, "loss": 0.011, "step": 62080 }, { "grad_norm": 0.24901768565177917, "learning_rate": 1.308946875358174e-05, "loss": 0.02, "step": 62090 }, { "grad_norm": 0.6408109068870544, "learning_rate": 1.3075529637343276e-05, "loss": 0.0165, "step": 62100 }, { "grad_norm": 0.21994154155254364, "learning_rate": 1.3061596830486883e-05, "loss": 0.0131, "step": 62110 }, { "grad_norm": 0.17531953752040863, "learning_rate": 1.3047670335393353e-05, "loss": 0.0072, "step": 62120 }, { "grad_norm": 0.289400577545166, "learning_rate": 1.3033750154442298e-05, "loss": 0.0126, "step": 62130 }, { "grad_norm": 0.13332287967205048, "learning_rate": 1.3019836290012316e-05, "loss": 0.0077, "step": 62140 }, { "grad_norm": 0.1413017064332962, "learning_rate": 1.30059287444809e-05, "loss": 0.0128, "step": 62150 }, { "grad_norm": 0.25031405687332153, "learning_rate": 1.299202752022447e-05, "loss": 0.018, "step": 62160 }, { "grad_norm": 0.21169409155845642, "learning_rate": 1.2978132619618371e-05, "loss": 0.0085, "step": 62170 }, { "grad_norm": 0.2436373382806778, "learning_rate": 1.2964244045036866e-05, "loss": 0.013, "step": 62180 }, { "grad_norm": 0.10944829136133194, "learning_rate": 1.295036179885311e-05, "loss": 0.0084, "step": 62190 }, { "grad_norm": 0.12621337175369263, "learning_rate": 1.293648588343922e-05, "loss": 0.0111, "step": 62200 }, { "grad_norm": 0.3242246210575104, "learning_rate": 1.2922616301166196e-05, "loss": 0.0108, "step": 62210 }, { "grad_norm": 0.20766514539718628, "learning_rate": 1.2908753054403976e-05, "loss": 0.0116, "step": 62220 }, { "grad_norm": 0.21314561367034912, "learning_rate": 1.2894896145521429e-05, "loss": 0.0105, "step": 62230 }, { "grad_norm": 0.21456141769886017, "learning_rate": 1.2881045576886275e-05, "loss": 0.0123, "step": 62240 }, { "grad_norm": 0.16844171285629272, "learning_rate": 1.2867201350865254e-05, "loss": 0.0104, "step": 62250 }, { "grad_norm": 0.1928277313709259, "learning_rate": 1.2853363469823914e-05, "loss": 0.012, "step": 62260 }, { "grad_norm": 0.2400137186050415, "learning_rate": 1.283953193612682e-05, "loss": 0.0088, "step": 62270 }, { "grad_norm": 0.22270241379737854, "learning_rate": 1.2825706752137372e-05, "loss": 0.0097, "step": 62280 }, { "grad_norm": 0.24178655445575714, "learning_rate": 1.2811887920217896e-05, "loss": 0.0137, "step": 62290 }, { "grad_norm": 0.16620545089244843, "learning_rate": 1.2798075442729707e-05, "loss": 0.0092, "step": 62300 }, { "grad_norm": 0.15533174574375153, "learning_rate": 1.2784269322032922e-05, "loss": 0.0099, "step": 62310 }, { "grad_norm": 0.12222574651241302, "learning_rate": 1.2770469560486686e-05, "loss": 0.0109, "step": 62320 }, { "grad_norm": 0.12679877877235413, "learning_rate": 1.2756676160448956e-05, "loss": 0.0107, "step": 62330 }, { "grad_norm": 0.2070256471633911, "learning_rate": 1.2742889124276663e-05, "loss": 0.0129, "step": 62340 }, { "grad_norm": 0.2811201214790344, "learning_rate": 1.2729108454325639e-05, "loss": 0.0103, "step": 62350 }, { "grad_norm": 0.2310275286436081, "learning_rate": 1.2715334152950614e-05, "loss": 0.0115, "step": 62360 }, { "grad_norm": 0.17919765412807465, "learning_rate": 1.2701566222505246e-05, "loss": 0.0102, "step": 62370 }, { "grad_norm": 0.2793988883495331, "learning_rate": 1.2687804665342107e-05, "loss": 0.0111, "step": 62380 }, { "grad_norm": 0.13859187066555023, "learning_rate": 1.267404948381265e-05, "loss": 0.009, "step": 62390 }, { "grad_norm": 0.18995213508605957, "learning_rate": 1.2660300680267267e-05, "loss": 0.0088, "step": 62400 }, { "grad_norm": 0.2645934522151947, "learning_rate": 1.2646558257055257e-05, "loss": 0.0111, "step": 62410 }, { "grad_norm": 0.1537768393754959, "learning_rate": 1.263282221652482e-05, "loss": 0.0071, "step": 62420 }, { "grad_norm": 0.12922902405261993, "learning_rate": 1.2619092561023088e-05, "loss": 0.0092, "step": 62430 }, { "grad_norm": 0.11500909924507141, "learning_rate": 1.2605369292896036e-05, "loss": 0.0106, "step": 62440 }, { "grad_norm": 0.10068861395120621, "learning_rate": 1.2591652414488658e-05, "loss": 0.0097, "step": 62450 }, { "grad_norm": 0.1365872472524643, "learning_rate": 1.2577941928144732e-05, "loss": 0.0103, "step": 62460 }, { "grad_norm": 0.1478886753320694, "learning_rate": 1.256423783620706e-05, "loss": 0.0107, "step": 62470 }, { "grad_norm": 0.2543528079986572, "learning_rate": 1.2550540141017264e-05, "loss": 0.0087, "step": 62480 }, { "grad_norm": 0.19077365100383759, "learning_rate": 1.253684884491591e-05, "loss": 0.007, "step": 62490 }, { "grad_norm": 0.21770097315311432, "learning_rate": 1.2523163950242483e-05, "loss": 0.008, "step": 62500 }, { "grad_norm": 0.17023742198944092, "learning_rate": 1.2509485459335313e-05, "loss": 0.0082, "step": 62510 }, { "grad_norm": 0.13544277846813202, "learning_rate": 1.2495813374531739e-05, "loss": 0.0092, "step": 62520 }, { "grad_norm": 0.31463631987571716, "learning_rate": 1.2482147698167907e-05, "loss": 0.0086, "step": 62530 }, { "grad_norm": 0.11207950115203857, "learning_rate": 1.2468488432578911e-05, "loss": 0.0102, "step": 62540 }, { "grad_norm": 0.2513562738895416, "learning_rate": 1.2454835580098761e-05, "loss": 0.0103, "step": 62550 }, { "grad_norm": 0.19024232029914856, "learning_rate": 1.2441189143060338e-05, "loss": 0.0117, "step": 62560 }, { "grad_norm": 0.28095829486846924, "learning_rate": 1.242754912379548e-05, "loss": 0.0104, "step": 62570 }, { "grad_norm": 0.2182208150625229, "learning_rate": 1.2413915524634844e-05, "loss": 0.023, "step": 62580 }, { "grad_norm": 0.20202668011188507, "learning_rate": 1.2400288347908073e-05, "loss": 0.0109, "step": 62590 }, { "grad_norm": 0.2829579710960388, "learning_rate": 1.2386667595943663e-05, "loss": 0.0102, "step": 62600 }, { "grad_norm": 0.2141416221857071, "learning_rate": 1.2373053271069035e-05, "loss": 0.0096, "step": 62610 }, { "grad_norm": 0.1296742707490921, "learning_rate": 1.2359445375610501e-05, "loss": 0.0093, "step": 62620 }, { "grad_norm": 0.6895612478256226, "learning_rate": 1.2345843911893301e-05, "loss": 0.0124, "step": 62630 }, { "grad_norm": 0.19089531898498535, "learning_rate": 1.2332248882241498e-05, "loss": 0.0218, "step": 62640 }, { "grad_norm": 0.16592182219028473, "learning_rate": 1.2318660288978178e-05, "loss": 0.0095, "step": 62650 }, { "grad_norm": 0.18918393552303314, "learning_rate": 1.2305078134425213e-05, "loss": 0.0106, "step": 62660 }, { "grad_norm": 0.11311361938714981, "learning_rate": 1.2291502420903434e-05, "loss": 0.0142, "step": 62670 }, { "grad_norm": 0.12543293833732605, "learning_rate": 1.2277933150732567e-05, "loss": 0.0109, "step": 62680 }, { "grad_norm": 0.08482871949672699, "learning_rate": 1.2264370326231216e-05, "loss": 0.0096, "step": 62690 }, { "grad_norm": 0.16868773102760315, "learning_rate": 1.2250813949716927e-05, "loss": 0.0108, "step": 62700 }, { "grad_norm": 0.12288319319486618, "learning_rate": 1.2237264023506063e-05, "loss": 0.0122, "step": 62710 }, { "grad_norm": 0.12306611984968185, "learning_rate": 1.2223720549913987e-05, "loss": 0.0096, "step": 62720 }, { "grad_norm": 0.14942824840545654, "learning_rate": 1.221018353125487e-05, "loss": 0.0113, "step": 62730 }, { "grad_norm": 0.15232136845588684, "learning_rate": 1.2196652969841837e-05, "loss": 0.0116, "step": 62740 }, { "grad_norm": 0.3349095284938812, "learning_rate": 1.2183128867986904e-05, "loss": 0.0121, "step": 62750 }, { "grad_norm": 0.14643387496471405, "learning_rate": 1.2169611228000927e-05, "loss": 0.0087, "step": 62760 }, { "grad_norm": 0.2055252194404602, "learning_rate": 1.2156100052193752e-05, "loss": 0.01, "step": 62770 }, { "grad_norm": 0.2163863331079483, "learning_rate": 1.214259534287403e-05, "loss": 0.0099, "step": 62780 }, { "grad_norm": 0.1895943582057953, "learning_rate": 1.2129097102349363e-05, "loss": 0.0092, "step": 62790 }, { "grad_norm": 0.13056951761245728, "learning_rate": 1.2115605332926227e-05, "loss": 0.0102, "step": 62800 }, { "grad_norm": 0.1507519781589508, "learning_rate": 1.2102120036910003e-05, "loss": 0.0183, "step": 62810 }, { "grad_norm": 0.11252538114786148, "learning_rate": 1.2088641216604956e-05, "loss": 0.0091, "step": 62820 }, { "grad_norm": 0.14433498680591583, "learning_rate": 1.2075168874314264e-05, "loss": 0.0092, "step": 62830 }, { "grad_norm": 0.2847782075405121, "learning_rate": 1.2061703012339942e-05, "loss": 0.0081, "step": 62840 }, { "grad_norm": 0.24799232184886932, "learning_rate": 1.204824363298297e-05, "loss": 0.011, "step": 62850 }, { "grad_norm": 0.20350900292396545, "learning_rate": 1.2034790738543173e-05, "loss": 0.0101, "step": 62860 }, { "grad_norm": 0.15153396129608154, "learning_rate": 1.2021344331319284e-05, "loss": 0.0114, "step": 62870 }, { "grad_norm": 0.3393104374408722, "learning_rate": 1.200790441360894e-05, "loss": 0.0104, "step": 62880 }, { "grad_norm": 0.2686619162559509, "learning_rate": 1.199447098770864e-05, "loss": 0.0116, "step": 62890 }, { "grad_norm": 0.19070219993591309, "learning_rate": 1.198104405591381e-05, "loss": 0.0164, "step": 62900 }, { "grad_norm": 0.12160696089267731, "learning_rate": 1.1967623620518697e-05, "loss": 0.0107, "step": 62910 }, { "grad_norm": 0.3692034184932709, "learning_rate": 1.1954209683816554e-05, "loss": 0.0101, "step": 62920 }, { "grad_norm": 0.1462906002998352, "learning_rate": 1.1940802248099402e-05, "loss": 0.0119, "step": 62930 }, { "grad_norm": 0.37317320704460144, "learning_rate": 1.1927401315658232e-05, "loss": 0.0111, "step": 62940 }, { "grad_norm": 0.26911911368370056, "learning_rate": 1.1914006888782898e-05, "loss": 0.0135, "step": 62950 }, { "grad_norm": 0.20134234428405762, "learning_rate": 1.1900618969762106e-05, "loss": 0.0111, "step": 62960 }, { "grad_norm": 0.1775774210691452, "learning_rate": 1.188723756088354e-05, "loss": 0.0114, "step": 62970 }, { "grad_norm": 0.13667842745780945, "learning_rate": 1.1873862664433672e-05, "loss": 0.0083, "step": 62980 }, { "grad_norm": 0.11541322618722916, "learning_rate": 1.1860494282697927e-05, "loss": 0.0097, "step": 62990 }, { "grad_norm": 0.11493848264217377, "learning_rate": 1.1847132417960588e-05, "loss": 0.0065, "step": 63000 }, { "grad_norm": 0.13130222260951996, "learning_rate": 1.1833777072504832e-05, "loss": 0.0096, "step": 63010 }, { "grad_norm": 0.17009875178337097, "learning_rate": 1.1820428248612731e-05, "loss": 0.0096, "step": 63020 }, { "grad_norm": 0.4076317548751831, "learning_rate": 1.1807085948565245e-05, "loss": 0.014, "step": 63030 }, { "grad_norm": 0.12287862598896027, "learning_rate": 1.1793750174642172e-05, "loss": 0.009, "step": 63040 }, { "grad_norm": 0.11869288980960846, "learning_rate": 1.1780420929122254e-05, "loss": 0.0123, "step": 63050 }, { "grad_norm": 0.22583025693893433, "learning_rate": 1.176709821428309e-05, "loss": 0.0105, "step": 63060 }, { "grad_norm": 0.1903325468301773, "learning_rate": 1.1753782032401173e-05, "loss": 0.0134, "step": 63070 }, { "grad_norm": 0.25544267892837524, "learning_rate": 1.1740472385751866e-05, "loss": 0.0109, "step": 63080 }, { "grad_norm": 0.14772401750087738, "learning_rate": 1.172716927660943e-05, "loss": 0.0083, "step": 63090 }, { "grad_norm": 0.22698695957660675, "learning_rate": 1.1713872707247015e-05, "loss": 0.013, "step": 63100 }, { "grad_norm": 0.14490050077438354, "learning_rate": 1.1700582679936595e-05, "loss": 0.0092, "step": 63110 }, { "grad_norm": 0.1805352419614792, "learning_rate": 1.1687299196949136e-05, "loss": 0.0165, "step": 63120 }, { "grad_norm": 0.28434866666793823, "learning_rate": 1.1674022260554374e-05, "loss": 0.0142, "step": 63130 }, { "grad_norm": 0.21055547893047333, "learning_rate": 1.1660751873020987e-05, "loss": 0.0106, "step": 63140 }, { "grad_norm": 0.47708025574684143, "learning_rate": 1.1647488036616538e-05, "loss": 0.0111, "step": 63150 }, { "grad_norm": 0.11632315814495087, "learning_rate": 1.1634230753607417e-05, "loss": 0.0165, "step": 63160 }, { "grad_norm": 0.14342118799686432, "learning_rate": 1.1620980026258982e-05, "loss": 0.0107, "step": 63170 }, { "grad_norm": 0.1205945760011673, "learning_rate": 1.1607735856835373e-05, "loss": 0.0073, "step": 63180 }, { "grad_norm": 0.11568859964609146, "learning_rate": 1.1594498247599677e-05, "loss": 0.0097, "step": 63190 }, { "grad_norm": 0.14078523218631744, "learning_rate": 1.158126720081384e-05, "loss": 0.0146, "step": 63200 }, { "grad_norm": 0.3151005804538727, "learning_rate": 1.156804271873868e-05, "loss": 0.0132, "step": 63210 }, { "grad_norm": 0.17312663793563843, "learning_rate": 1.1554824803633924e-05, "loss": 0.0096, "step": 63220 }, { "grad_norm": 0.3471582233905792, "learning_rate": 1.154161345775811e-05, "loss": 0.0101, "step": 63230 }, { "grad_norm": 0.1323496550321579, "learning_rate": 1.1528408683368724e-05, "loss": 0.0138, "step": 63240 }, { "grad_norm": 0.3722290098667145, "learning_rate": 1.1515210482722088e-05, "loss": 0.0103, "step": 63250 }, { "grad_norm": 0.09757682681083679, "learning_rate": 1.150201885807342e-05, "loss": 0.015, "step": 63260 }, { "grad_norm": 0.14929279685020447, "learning_rate": 1.1488833811676807e-05, "loss": 0.0091, "step": 63270 }, { "grad_norm": 0.1768253594636917, "learning_rate": 1.1475655345785213e-05, "loss": 0.011, "step": 63280 }, { "grad_norm": 0.18641623854637146, "learning_rate": 1.1462483462650481e-05, "loss": 0.0112, "step": 63290 }, { "grad_norm": 0.22281154990196228, "learning_rate": 1.1449318164523331e-05, "loss": 0.0144, "step": 63300 }, { "grad_norm": 0.19892820715904236, "learning_rate": 1.1436159453653334e-05, "loss": 0.0097, "step": 63310 }, { "grad_norm": 0.2963273227214813, "learning_rate": 1.1423007332288955e-05, "loss": 0.0127, "step": 63320 }, { "grad_norm": 0.31132397055625916, "learning_rate": 1.1409861802677546e-05, "loss": 0.0102, "step": 63330 }, { "grad_norm": 0.15135473012924194, "learning_rate": 1.1396722867065313e-05, "loss": 0.0083, "step": 63340 }, { "grad_norm": 0.20207644999027252, "learning_rate": 1.1383590527697352e-05, "loss": 0.0128, "step": 63350 }, { "grad_norm": 0.12985888123512268, "learning_rate": 1.137046478681758e-05, "loss": 0.0088, "step": 63360 }, { "grad_norm": 0.2443946897983551, "learning_rate": 1.1357345646668888e-05, "loss": 0.0208, "step": 63370 }, { "grad_norm": 0.3288727402687073, "learning_rate": 1.1344233109492924e-05, "loss": 0.0115, "step": 63380 }, { "grad_norm": 0.14514245092868805, "learning_rate": 1.1331127177530292e-05, "loss": 0.0102, "step": 63390 }, { "grad_norm": 0.27698835730552673, "learning_rate": 1.1318027853020441e-05, "loss": 0.0086, "step": 63400 }, { "grad_norm": 0.11535844206809998, "learning_rate": 1.1304935138201645e-05, "loss": 0.0093, "step": 63410 }, { "grad_norm": 0.15205687284469604, "learning_rate": 1.1291849035311153e-05, "loss": 0.0098, "step": 63420 }, { "grad_norm": 0.14705824851989746, "learning_rate": 1.1278769546584972e-05, "loss": 0.0086, "step": 63430 }, { "grad_norm": 0.26504018902778625, "learning_rate": 1.1265696674258052e-05, "loss": 0.0093, "step": 63440 }, { "grad_norm": 0.12436935305595398, "learning_rate": 1.1252630420564186e-05, "loss": 0.0104, "step": 63450 }, { "grad_norm": 0.23147830367088318, "learning_rate": 1.1239570787736036e-05, "loss": 0.0115, "step": 63460 }, { "grad_norm": 0.18480515480041504, "learning_rate": 1.1226517778005135e-05, "loss": 0.0104, "step": 63470 }, { "grad_norm": 0.668125569820404, "learning_rate": 1.1213471393601893e-05, "loss": 0.0077, "step": 63480 }, { "grad_norm": 0.2782374918460846, "learning_rate": 1.1200431636755587e-05, "loss": 0.0167, "step": 63490 }, { "grad_norm": 0.3587455749511719, "learning_rate": 1.1187398509694336e-05, "loss": 0.0123, "step": 63500 }, { "grad_norm": 0.14054913818836212, "learning_rate": 1.1174372014645146e-05, "loss": 0.0086, "step": 63510 }, { "grad_norm": 0.12378821521997452, "learning_rate": 1.1161352153833899e-05, "loss": 0.0142, "step": 63520 }, { "grad_norm": 0.1436253786087036, "learning_rate": 1.1148338929485325e-05, "loss": 0.0117, "step": 63530 }, { "grad_norm": 0.10765668004751205, "learning_rate": 1.113533234382304e-05, "loss": 0.0086, "step": 63540 }, { "grad_norm": 0.19403846561908722, "learning_rate": 1.1122332399069513e-05, "loss": 0.0095, "step": 63550 }, { "grad_norm": 0.22733496129512787, "learning_rate": 1.1109339097446047e-05, "loss": 0.0121, "step": 63560 }, { "grad_norm": 0.2600625455379486, "learning_rate": 1.1096352441172897e-05, "loss": 0.0122, "step": 63570 }, { "grad_norm": 0.18023164570331573, "learning_rate": 1.1083372432469086e-05, "loss": 0.0091, "step": 63580 }, { "grad_norm": 0.13297297060489655, "learning_rate": 1.107039907355255e-05, "loss": 0.0138, "step": 63590 }, { "grad_norm": 0.29443594813346863, "learning_rate": 1.1057432366640103e-05, "loss": 0.0119, "step": 63600 }, { "grad_norm": 0.2979905903339386, "learning_rate": 1.1044472313947352e-05, "loss": 0.0163, "step": 63610 }, { "grad_norm": 0.18097157776355743, "learning_rate": 1.1031518917688877e-05, "loss": 0.0105, "step": 63620 }, { "grad_norm": 0.35752683877944946, "learning_rate": 1.1018572180078007e-05, "loss": 0.02, "step": 63630 }, { "grad_norm": 0.11843522638082504, "learning_rate": 1.1005632103327018e-05, "loss": 0.0089, "step": 63640 }, { "grad_norm": 0.10129386186599731, "learning_rate": 1.0992698689646996e-05, "loss": 0.0116, "step": 63650 }, { "grad_norm": 0.1907045841217041, "learning_rate": 1.0979771941247919e-05, "loss": 0.0137, "step": 63660 }, { "grad_norm": 0.18332765996456146, "learning_rate": 1.0966851860338611e-05, "loss": 0.0093, "step": 63670 }, { "grad_norm": 0.20829126238822937, "learning_rate": 1.0953938449126766e-05, "loss": 0.0115, "step": 63680 }, { "grad_norm": 0.36556658148765564, "learning_rate": 1.0941031709818933e-05, "loss": 0.0205, "step": 63690 }, { "grad_norm": 0.2288895547389984, "learning_rate": 1.0928131644620509e-05, "loss": 0.0201, "step": 63700 }, { "grad_norm": 0.13253135979175568, "learning_rate": 1.0915238255735766e-05, "loss": 0.0117, "step": 63710 }, { "grad_norm": 0.14736804366111755, "learning_rate": 1.0902351545367833e-05, "loss": 0.0133, "step": 63720 }, { "grad_norm": 0.20778371393680573, "learning_rate": 1.0889471515718702e-05, "loss": 0.009, "step": 63730 }, { "grad_norm": 0.25557059049606323, "learning_rate": 1.087659816898921e-05, "loss": 0.008, "step": 63740 }, { "grad_norm": 0.17569220066070557, "learning_rate": 1.0863731507379082e-05, "loss": 0.0089, "step": 63750 }, { "grad_norm": 0.2482437938451767, "learning_rate": 1.0850871533086827e-05, "loss": 0.0113, "step": 63760 }, { "grad_norm": 0.21029502153396606, "learning_rate": 1.0838018248309927e-05, "loss": 0.0088, "step": 63770 }, { "grad_norm": 0.29706650972366333, "learning_rate": 1.0825171655244615e-05, "loss": 0.0139, "step": 63780 }, { "grad_norm": 0.22065547108650208, "learning_rate": 1.0812331756086025e-05, "loss": 0.0098, "step": 63790 }, { "grad_norm": 0.28196898102760315, "learning_rate": 1.079949855302817e-05, "loss": 0.0093, "step": 63800 }, { "grad_norm": 0.1713750958442688, "learning_rate": 1.0786672048263852e-05, "loss": 0.011, "step": 63810 }, { "grad_norm": 0.16056495904922485, "learning_rate": 1.0773852243984817e-05, "loss": 0.0133, "step": 63820 }, { "grad_norm": 0.19376705586910248, "learning_rate": 1.0761039142381586e-05, "loss": 0.0116, "step": 63830 }, { "grad_norm": 0.1824403554201126, "learning_rate": 1.0748232745643577e-05, "loss": 0.0122, "step": 63840 }, { "grad_norm": 0.14850127696990967, "learning_rate": 1.0735433055959055e-05, "loss": 0.0121, "step": 63850 }, { "grad_norm": 0.18545784056186676, "learning_rate": 1.0722640075515133e-05, "loss": 0.01, "step": 63860 }, { "grad_norm": 0.15119880437850952, "learning_rate": 1.0709853806497795e-05, "loss": 0.0078, "step": 63870 }, { "grad_norm": 0.5931840538978577, "learning_rate": 1.0697074251091831e-05, "loss": 0.0154, "step": 63880 }, { "grad_norm": 0.1542857438325882, "learning_rate": 1.0684301411480962e-05, "loss": 0.0098, "step": 63890 }, { "grad_norm": 0.15258750319480896, "learning_rate": 1.067153528984769e-05, "loss": 0.0117, "step": 63900 }, { "grad_norm": 0.17745329439640045, "learning_rate": 1.0658775888373395e-05, "loss": 0.0088, "step": 63910 }, { "grad_norm": 0.2704722583293915, "learning_rate": 1.0646023209238314e-05, "loss": 0.0107, "step": 63920 }, { "grad_norm": 0.10637284070253372, "learning_rate": 1.0633277254621537e-05, "loss": 0.0074, "step": 63930 }, { "grad_norm": 0.10439513623714447, "learning_rate": 1.0620538026700994e-05, "loss": 0.01, "step": 63940 }, { "grad_norm": 0.14493945240974426, "learning_rate": 1.0607805527653486e-05, "loss": 0.0099, "step": 63950 }, { "grad_norm": 0.31926363706588745, "learning_rate": 1.059507975965462e-05, "loss": 0.013, "step": 63960 }, { "grad_norm": 0.14840061962604523, "learning_rate": 1.0582360724878898e-05, "loss": 0.0064, "step": 63970 }, { "grad_norm": 0.3747979998588562, "learning_rate": 1.0569648425499651e-05, "loss": 0.0118, "step": 63980 }, { "grad_norm": 0.13612130284309387, "learning_rate": 1.0556942863689063e-05, "loss": 0.0086, "step": 63990 }, { "grad_norm": 0.4090120494365692, "learning_rate": 1.054424404161819e-05, "loss": 0.0081, "step": 64000 }, { "grad_norm": 0.11095970124006271, "learning_rate": 1.053155196145686e-05, "loss": 0.0119, "step": 64010 }, { "grad_norm": 0.28698331117630005, "learning_rate": 1.0518866625373863e-05, "loss": 0.0082, "step": 64020 }, { "grad_norm": 0.11460090428590775, "learning_rate": 1.0506188035536735e-05, "loss": 0.0111, "step": 64030 }, { "grad_norm": 0.171247661113739, "learning_rate": 1.0493516194111919e-05, "loss": 0.0111, "step": 64040 }, { "grad_norm": 0.16608422994613647, "learning_rate": 1.0480851103264688e-05, "loss": 0.0121, "step": 64050 }, { "grad_norm": 0.24284982681274414, "learning_rate": 1.046819276515913e-05, "loss": 0.0146, "step": 64060 }, { "grad_norm": 0.2789357006549835, "learning_rate": 1.0455541181958256e-05, "loss": 0.0103, "step": 64070 }, { "grad_norm": 0.20184271037578583, "learning_rate": 1.0442896355823822e-05, "loss": 0.0109, "step": 64080 }, { "grad_norm": 0.2732013165950775, "learning_rate": 1.0430258288916539e-05, "loss": 0.0096, "step": 64090 }, { "grad_norm": 0.11422541737556458, "learning_rate": 1.0417626983395868e-05, "loss": 0.0092, "step": 64100 }, { "grad_norm": 0.17706148326396942, "learning_rate": 1.0405002441420165e-05, "loss": 0.0079, "step": 64110 }, { "grad_norm": 0.27234703302383423, "learning_rate": 1.039238466514662e-05, "loss": 0.0101, "step": 64120 }, { "grad_norm": 0.19911621510982513, "learning_rate": 1.0379773656731262e-05, "loss": 0.0135, "step": 64130 }, { "grad_norm": 0.14499498903751373, "learning_rate": 1.0367169418328986e-05, "loss": 0.0096, "step": 64140 }, { "grad_norm": 0.33634495735168457, "learning_rate": 1.0354571952093484e-05, "loss": 0.0131, "step": 64150 }, { "grad_norm": 0.5169599652290344, "learning_rate": 1.034198126017733e-05, "loss": 0.0121, "step": 64160 }, { "grad_norm": 0.14163081347942352, "learning_rate": 1.0329397344731928e-05, "loss": 0.0123, "step": 64170 }, { "grad_norm": 0.21637773513793945, "learning_rate": 1.0316820207907524e-05, "loss": 0.0128, "step": 64180 }, { "grad_norm": 0.3215234875679016, "learning_rate": 1.0304249851853215e-05, "loss": 0.0101, "step": 64190 }, { "grad_norm": 0.17869777977466583, "learning_rate": 1.0291686278716933e-05, "loss": 0.0105, "step": 64200 }, { "grad_norm": 0.3924002945423126, "learning_rate": 1.0279129490645418e-05, "loss": 0.0098, "step": 64210 }, { "grad_norm": 0.1497737020254135, "learning_rate": 1.0266579489784328e-05, "loss": 0.0093, "step": 64220 }, { "grad_norm": 0.16120748221874237, "learning_rate": 1.0254036278278084e-05, "loss": 0.0111, "step": 64230 }, { "grad_norm": 0.10168781131505966, "learning_rate": 1.0241499858269982e-05, "loss": 0.0075, "step": 64240 }, { "grad_norm": 0.18174321949481964, "learning_rate": 1.0228970231902169e-05, "loss": 0.0095, "step": 64250 }, { "grad_norm": 0.21718698740005493, "learning_rate": 1.0216447401315582e-05, "loss": 0.0108, "step": 64260 }, { "grad_norm": 0.27487292885780334, "learning_rate": 1.020393136865007e-05, "loss": 0.0096, "step": 64270 }, { "grad_norm": 0.2666522264480591, "learning_rate": 1.0191422136044242e-05, "loss": 0.0169, "step": 64280 }, { "grad_norm": 0.25468358397483826, "learning_rate": 1.017891970563563e-05, "loss": 0.012, "step": 64290 }, { "grad_norm": 0.2245706170797348, "learning_rate": 1.0166424079560516e-05, "loss": 0.0114, "step": 64300 }, { "grad_norm": 0.21879631280899048, "learning_rate": 1.0153935259954078e-05, "loss": 0.0083, "step": 64310 }, { "grad_norm": 0.14592061936855316, "learning_rate": 1.0141453248950311e-05, "loss": 0.0107, "step": 64320 }, { "grad_norm": 0.2756948173046112, "learning_rate": 1.0128978048682054e-05, "loss": 0.0115, "step": 64330 }, { "grad_norm": 0.30991971492767334, "learning_rate": 1.0116509661280982e-05, "loss": 0.011, "step": 64340 }, { "grad_norm": 0.5608318448066711, "learning_rate": 1.0104048088877576e-05, "loss": 0.0079, "step": 64350 }, { "grad_norm": 0.24384748935699463, "learning_rate": 1.0091593333601201e-05, "loss": 0.0102, "step": 64360 }, { "grad_norm": 0.2298741191625595, "learning_rate": 1.0079145397580031e-05, "loss": 0.009, "step": 64370 }, { "grad_norm": 0.3450584411621094, "learning_rate": 1.006670428294107e-05, "loss": 0.0148, "step": 64380 }, { "grad_norm": 0.20978054404258728, "learning_rate": 1.0054269991810166e-05, "loss": 0.0079, "step": 64390 }, { "grad_norm": 0.1148085817694664, "learning_rate": 1.0041842526312024e-05, "loss": 0.0075, "step": 64400 }, { "grad_norm": 0.15955893695354462, "learning_rate": 1.0029421888570101e-05, "loss": 0.0072, "step": 64410 }, { "grad_norm": 0.1936369389295578, "learning_rate": 1.0017008080706813e-05, "loss": 0.0144, "step": 64420 }, { "grad_norm": 0.4063687026500702, "learning_rate": 1.0004601104843287e-05, "loss": 0.0114, "step": 64430 }, { "grad_norm": 0.31068724393844604, "learning_rate": 9.992200963099562e-06, "loss": 0.0127, "step": 64440 }, { "grad_norm": 0.11613018810749054, "learning_rate": 9.979807657594486e-06, "loss": 0.0098, "step": 64450 }, { "grad_norm": 0.1702340692281723, "learning_rate": 9.967421190445703e-06, "loss": 0.0097, "step": 64460 }, { "grad_norm": 0.17850811779499054, "learning_rate": 9.955041563769769e-06, "loss": 0.0106, "step": 64470 }, { "grad_norm": 0.43652549386024475, "learning_rate": 9.942668779681974e-06, "loss": 0.0112, "step": 64480 }, { "grad_norm": 0.18235467374324799, "learning_rate": 9.930302840296541e-06, "loss": 0.0102, "step": 64490 }, { "grad_norm": 0.2052765041589737, "learning_rate": 9.917943747726426e-06, "loss": 0.0096, "step": 64500 }, { "grad_norm": 0.2930937707424164, "learning_rate": 9.905591504083484e-06, "loss": 0.0103, "step": 64510 }, { "grad_norm": 0.15374338626861572, "learning_rate": 9.893246111478382e-06, "loss": 0.0086, "step": 64520 }, { "grad_norm": 0.41804611682891846, "learning_rate": 9.88090757202057e-06, "loss": 0.0117, "step": 64530 }, { "grad_norm": 0.26083633303642273, "learning_rate": 9.868575887818421e-06, "loss": 0.017, "step": 64540 }, { "grad_norm": 0.10489602386951447, "learning_rate": 9.856251060979044e-06, "loss": 0.007, "step": 64550 }, { "grad_norm": 0.13878251612186432, "learning_rate": 9.843933093608426e-06, "loss": 0.0129, "step": 64560 }, { "grad_norm": 0.16344168782234192, "learning_rate": 9.831621987811368e-06, "loss": 0.0146, "step": 64570 }, { "grad_norm": 0.13925576210021973, "learning_rate": 9.819317745691509e-06, "loss": 0.0094, "step": 64580 }, { "grad_norm": 0.16573067009449005, "learning_rate": 9.8070203693513e-06, "loss": 0.0077, "step": 64590 }, { "grad_norm": 0.23282977938652039, "learning_rate": 9.794729860892048e-06, "loss": 0.0107, "step": 64600 }, { "grad_norm": 0.23320449888706207, "learning_rate": 9.782446222413827e-06, "loss": 0.0135, "step": 64610 }, { "grad_norm": 0.24179677665233612, "learning_rate": 9.770169456015598e-06, "loss": 0.0114, "step": 64620 }, { "grad_norm": 0.10544569045305252, "learning_rate": 9.75789956379512e-06, "loss": 0.011, "step": 64630 }, { "grad_norm": 0.2198694944381714, "learning_rate": 9.74563654784898e-06, "loss": 0.0112, "step": 64640 }, { "grad_norm": 0.20693303644657135, "learning_rate": 9.733380410272596e-06, "loss": 0.0096, "step": 64650 }, { "grad_norm": 0.18994203209877014, "learning_rate": 9.721131153160207e-06, "loss": 0.01, "step": 64660 }, { "grad_norm": 0.11939526349306107, "learning_rate": 9.708888778604886e-06, "loss": 0.0073, "step": 64670 }, { "grad_norm": 0.12200219929218292, "learning_rate": 9.696653288698488e-06, "loss": 0.0105, "step": 64680 }, { "grad_norm": 0.21972185373306274, "learning_rate": 9.684424685531763e-06, "loss": 0.013, "step": 64690 }, { "grad_norm": 0.22757892310619354, "learning_rate": 9.672202971194216e-06, "loss": 0.015, "step": 64700 }, { "grad_norm": 0.1769404262304306, "learning_rate": 9.659988147774213e-06, "loss": 0.0124, "step": 64710 }, { "grad_norm": 0.15937970578670502, "learning_rate": 9.647780217358942e-06, "loss": 0.011, "step": 64720 }, { "grad_norm": 0.20370326936244965, "learning_rate": 9.635579182034376e-06, "loss": 0.009, "step": 64730 }, { "grad_norm": 0.20075024664402008, "learning_rate": 9.623385043885386e-06, "loss": 0.0131, "step": 64740 }, { "grad_norm": 0.26142022013664246, "learning_rate": 9.61119780499557e-06, "loss": 0.0131, "step": 64750 }, { "grad_norm": 0.11011748015880585, "learning_rate": 9.599017467447418e-06, "loss": 0.0081, "step": 64760 }, { "grad_norm": 0.10968285799026489, "learning_rate": 9.586844033322206e-06, "loss": 0.0075, "step": 64770 }, { "grad_norm": 0.2654668986797333, "learning_rate": 9.574677504700052e-06, "loss": 0.0122, "step": 64780 }, { "grad_norm": 0.16260869801044464, "learning_rate": 9.562517883659877e-06, "loss": 0.0139, "step": 64790 }, { "grad_norm": 0.13427025079727173, "learning_rate": 9.55036517227944e-06, "loss": 0.0096, "step": 64800 }, { "grad_norm": 0.1297062486410141, "learning_rate": 9.538219372635282e-06, "loss": 0.0097, "step": 64810 }, { "grad_norm": 0.3882738947868347, "learning_rate": 9.526080486802802e-06, "loss": 0.0135, "step": 64820 }, { "grad_norm": 0.20763035118579865, "learning_rate": 9.513948516856203e-06, "loss": 0.0093, "step": 64830 }, { "grad_norm": 0.21419836580753326, "learning_rate": 9.501823464868504e-06, "loss": 0.0086, "step": 64840 }, { "grad_norm": 0.12999947369098663, "learning_rate": 9.489705332911547e-06, "loss": 0.0092, "step": 64850 }, { "grad_norm": 0.24898949265480042, "learning_rate": 9.477594123055994e-06, "loss": 0.0108, "step": 64860 }, { "grad_norm": 0.15546947717666626, "learning_rate": 9.465489837371321e-06, "loss": 0.0143, "step": 64870 }, { "grad_norm": 0.23097820580005646, "learning_rate": 9.453392477925794e-06, "loss": 0.0095, "step": 64880 }, { "grad_norm": 0.1970389187335968, "learning_rate": 9.441302046786566e-06, "loss": 0.0089, "step": 64890 }, { "grad_norm": 0.11908315122127533, "learning_rate": 9.429218546019519e-06, "loss": 0.0087, "step": 64900 }, { "grad_norm": 0.5995833873748779, "learning_rate": 9.41714197768941e-06, "loss": 0.0161, "step": 64910 }, { "grad_norm": 0.14457926154136658, "learning_rate": 9.405072343859805e-06, "loss": 0.009, "step": 64920 }, { "grad_norm": 0.17325638234615326, "learning_rate": 9.393009646593043e-06, "loss": 0.0088, "step": 64930 }, { "grad_norm": 0.1940944492816925, "learning_rate": 9.38095388795035e-06, "loss": 0.0112, "step": 64940 }, { "grad_norm": 0.33465275168418884, "learning_rate": 9.36890506999169e-06, "loss": 0.0097, "step": 64950 }, { "grad_norm": 0.13375809788703918, "learning_rate": 9.356863194775894e-06, "loss": 0.0152, "step": 64960 }, { "grad_norm": 0.26212042570114136, "learning_rate": 9.344828264360583e-06, "loss": 0.0106, "step": 64970 }, { "grad_norm": 0.2286486029624939, "learning_rate": 9.332800280802201e-06, "loss": 0.011, "step": 64980 }, { "grad_norm": 0.11892612278461456, "learning_rate": 9.32077924615602e-06, "loss": 0.0089, "step": 64990 }, { "grad_norm": 0.174282968044281, "learning_rate": 9.308765162476063e-06, "loss": 0.0106, "step": 65000 }, { "grad_norm": 0.19593027234077454, "learning_rate": 9.296758031815239e-06, "loss": 0.0128, "step": 65010 }, { "grad_norm": 0.24060207605361938, "learning_rate": 9.284757856225229e-06, "loss": 0.009, "step": 65020 }, { "grad_norm": 0.1287260204553604, "learning_rate": 9.272764637756538e-06, "loss": 0.0112, "step": 65030 }, { "grad_norm": 0.1726444661617279, "learning_rate": 9.260778378458479e-06, "loss": 0.013, "step": 65040 }, { "grad_norm": 0.09738026559352875, "learning_rate": 9.248799080379172e-06, "loss": 0.012, "step": 65050 }, { "grad_norm": 0.1540297418832779, "learning_rate": 9.236826745565558e-06, "loss": 0.0149, "step": 65060 }, { "grad_norm": 0.1594027727842331, "learning_rate": 9.224861376063388e-06, "loss": 0.0105, "step": 65070 }, { "grad_norm": 0.11001010984182358, "learning_rate": 9.212902973917192e-06, "loss": 0.0097, "step": 65080 }, { "grad_norm": 0.19003261625766754, "learning_rate": 9.20095154117035e-06, "loss": 0.011, "step": 65090 }, { "grad_norm": 0.44911837577819824, "learning_rate": 9.189007079865036e-06, "loss": 0.0076, "step": 65100 }, { "grad_norm": 0.20522615313529968, "learning_rate": 9.177069592042226e-06, "loss": 0.0083, "step": 65110 }, { "grad_norm": 0.15964889526367188, "learning_rate": 9.165139079741724e-06, "loss": 0.0133, "step": 65120 }, { "grad_norm": 0.2631320357322693, "learning_rate": 9.153215545002098e-06, "loss": 0.0142, "step": 65130 }, { "grad_norm": 0.18498404324054718, "learning_rate": 9.141298989860798e-06, "loss": 0.0148, "step": 65140 }, { "grad_norm": 0.05547394976019859, "learning_rate": 9.129389416353994e-06, "loss": 0.0077, "step": 65150 }, { "grad_norm": 0.1482437551021576, "learning_rate": 9.11748682651673e-06, "loss": 0.0081, "step": 65160 }, { "grad_norm": 0.1735645979642868, "learning_rate": 9.105591222382837e-06, "loss": 0.0102, "step": 65170 }, { "grad_norm": 0.13669177889823914, "learning_rate": 9.093702605984915e-06, "loss": 0.01, "step": 65180 }, { "grad_norm": 0.1972985863685608, "learning_rate": 9.081820979354455e-06, "loss": 0.0098, "step": 65190 }, { "grad_norm": 0.27951064705848694, "learning_rate": 9.069946344521663e-06, "loss": 0.0093, "step": 65200 }, { "grad_norm": 0.14692337810993195, "learning_rate": 9.058078703515598e-06, "loss": 0.0094, "step": 65210 }, { "grad_norm": 0.18886131048202515, "learning_rate": 9.046218058364125e-06, "loss": 0.0166, "step": 65220 }, { "grad_norm": 0.26271042227745056, "learning_rate": 9.034364411093893e-06, "loss": 0.01, "step": 65230 }, { "grad_norm": 0.13363832235336304, "learning_rate": 9.022517763730371e-06, "loss": 0.0113, "step": 65240 }, { "grad_norm": 0.22875158488750458, "learning_rate": 9.010678118297827e-06, "loss": 0.0099, "step": 65250 }, { "grad_norm": 0.20801162719726562, "learning_rate": 8.998845476819345e-06, "loss": 0.0122, "step": 65260 }, { "grad_norm": 0.20408698916435242, "learning_rate": 8.987019841316773e-06, "loss": 0.011, "step": 65270 }, { "grad_norm": 0.11326805502176285, "learning_rate": 8.975201213810802e-06, "loss": 0.0072, "step": 65280 }, { "grad_norm": 0.24173174798488617, "learning_rate": 8.963389596320915e-06, "loss": 0.0095, "step": 65290 }, { "grad_norm": 0.1263457089662552, "learning_rate": 8.951584990865391e-06, "loss": 0.0097, "step": 65300 }, { "grad_norm": 0.19641773402690887, "learning_rate": 8.939787399461319e-06, "loss": 0.0122, "step": 65310 }, { "grad_norm": 0.3713102638721466, "learning_rate": 8.927996824124591e-06, "loss": 0.0122, "step": 65320 }, { "grad_norm": 0.27489107847213745, "learning_rate": 8.916213266869854e-06, "loss": 0.0113, "step": 65330 }, { "grad_norm": 0.23821206390857697, "learning_rate": 8.904436729710658e-06, "loss": 0.0094, "step": 65340 }, { "grad_norm": 0.22463376820087433, "learning_rate": 8.892667214659245e-06, "loss": 0.0082, "step": 65350 }, { "grad_norm": 0.20189730823040009, "learning_rate": 8.880904723726713e-06, "loss": 0.0098, "step": 65360 }, { "grad_norm": 0.19449388980865479, "learning_rate": 8.869149258922971e-06, "loss": 0.0082, "step": 65370 }, { "grad_norm": 0.14769141376018524, "learning_rate": 8.857400822256662e-06, "loss": 0.0089, "step": 65380 }, { "grad_norm": 0.10463524609804153, "learning_rate": 8.845659415735324e-06, "loss": 0.0125, "step": 65390 }, { "grad_norm": 0.31120002269744873, "learning_rate": 8.83392504136521e-06, "loss": 0.0111, "step": 65400 }, { "grad_norm": 0.06277536600828171, "learning_rate": 8.822197701151407e-06, "loss": 0.0078, "step": 65410 }, { "grad_norm": 0.16532212495803833, "learning_rate": 8.810477397097804e-06, "loss": 0.0079, "step": 65420 }, { "grad_norm": 0.21462012827396393, "learning_rate": 8.798764131207077e-06, "loss": 0.011, "step": 65430 }, { "grad_norm": 0.244929701089859, "learning_rate": 8.787057905480706e-06, "loss": 0.0124, "step": 65440 }, { "grad_norm": 0.17652486264705658, "learning_rate": 8.775358721918958e-06, "loss": 0.0092, "step": 65450 }, { "grad_norm": 0.2526981234550476, "learning_rate": 8.763666582520923e-06, "loss": 0.0102, "step": 65460 }, { "grad_norm": 0.19834588468074799, "learning_rate": 8.751981489284445e-06, "loss": 0.0085, "step": 65470 }, { "grad_norm": 0.21344949305057526, "learning_rate": 8.740303444206188e-06, "loss": 0.0088, "step": 65480 }, { "grad_norm": 0.07879664748907089, "learning_rate": 8.72863244928162e-06, "loss": 0.008, "step": 65490 }, { "grad_norm": 0.12287349998950958, "learning_rate": 8.716968506504991e-06, "loss": 0.0125, "step": 65500 }, { "grad_norm": 0.15765917301177979, "learning_rate": 8.70531161786935e-06, "loss": 0.0078, "step": 65510 }, { "grad_norm": 0.19279997050762177, "learning_rate": 8.693661785366558e-06, "loss": 0.0167, "step": 65520 }, { "grad_norm": 0.11268260329961777, "learning_rate": 8.682019010987208e-06, "loss": 0.0132, "step": 65530 }, { "grad_norm": 0.13536569476127625, "learning_rate": 8.670383296720786e-06, "loss": 0.0079, "step": 65540 }, { "grad_norm": 0.2495325803756714, "learning_rate": 8.658754644555478e-06, "loss": 0.0074, "step": 65550 }, { "grad_norm": 0.2549690902233124, "learning_rate": 8.647133056478313e-06, "loss": 0.0128, "step": 65560 }, { "grad_norm": 0.15449224412441254, "learning_rate": 8.635518534475123e-06, "loss": 0.0117, "step": 65570 }, { "grad_norm": 0.15744496881961823, "learning_rate": 8.623911080530467e-06, "loss": 0.013, "step": 65580 }, { "grad_norm": 0.1790025383234024, "learning_rate": 8.61231069662779e-06, "loss": 0.0085, "step": 65590 }, { "grad_norm": 0.1800486147403717, "learning_rate": 8.600717384749252e-06, "loss": 0.0099, "step": 65600 }, { "grad_norm": 0.5739113092422485, "learning_rate": 8.58913114687584e-06, "loss": 0.0091, "step": 65610 }, { "grad_norm": 0.3847849369049072, "learning_rate": 8.57755198498732e-06, "loss": 0.0119, "step": 65620 }, { "grad_norm": 0.1422872692346573, "learning_rate": 8.565979901062265e-06, "loss": 0.0188, "step": 65630 }, { "grad_norm": 0.12550859153270721, "learning_rate": 8.554414897078033e-06, "loss": 0.0094, "step": 65640 }, { "grad_norm": 0.08329589664936066, "learning_rate": 8.542856975010727e-06, "loss": 0.0064, "step": 65650 }, { "grad_norm": 0.10851424932479858, "learning_rate": 8.531306136835337e-06, "loss": 0.0078, "step": 65660 }, { "grad_norm": 0.218182772397995, "learning_rate": 8.519762384525548e-06, "loss": 0.0118, "step": 65670 }, { "grad_norm": 0.21545854210853577, "learning_rate": 8.508225720053875e-06, "loss": 0.0084, "step": 65680 }, { "grad_norm": 0.1758679896593094, "learning_rate": 8.496696145391625e-06, "loss": 0.01, "step": 65690 }, { "grad_norm": 0.20880265533924103, "learning_rate": 8.485173662508889e-06, "loss": 0.0084, "step": 65700 }, { "grad_norm": 0.13321030139923096, "learning_rate": 8.473658273374536e-06, "loss": 0.0089, "step": 65710 }, { "grad_norm": 0.22359174489974976, "learning_rate": 8.462149979956253e-06, "loss": 0.0084, "step": 65720 }, { "grad_norm": 0.17169654369354248, "learning_rate": 8.450648784220461e-06, "loss": 0.0127, "step": 65730 }, { "grad_norm": 0.47103363275527954, "learning_rate": 8.439154688132417e-06, "loss": 0.0102, "step": 65740 }, { "grad_norm": 0.1953742504119873, "learning_rate": 8.427667693656143e-06, "loss": 0.0118, "step": 65750 }, { "grad_norm": 0.10928498208522797, "learning_rate": 8.416187802754454e-06, "loss": 0.0098, "step": 65760 }, { "grad_norm": 0.11307282745838165, "learning_rate": 8.404715017388965e-06, "loss": 0.0098, "step": 65770 }, { "grad_norm": 0.21188178658485413, "learning_rate": 8.393249339520015e-06, "loss": 0.0134, "step": 65780 }, { "grad_norm": 0.27886223793029785, "learning_rate": 8.381790771106834e-06, "loss": 0.0104, "step": 65790 }, { "grad_norm": 0.1770082712173462, "learning_rate": 8.370339314107339e-06, "loss": 0.0079, "step": 65800 }, { "grad_norm": 0.17537818849086761, "learning_rate": 8.358894970478281e-06, "loss": 0.0084, "step": 65810 }, { "grad_norm": 0.09760847687721252, "learning_rate": 8.347457742175196e-06, "loss": 0.0091, "step": 65820 }, { "grad_norm": 0.18342389166355133, "learning_rate": 8.33602763115236e-06, "loss": 0.0091, "step": 65830 }, { "grad_norm": 0.24217627942562103, "learning_rate": 8.324604639362916e-06, "loss": 0.0138, "step": 65840 }, { "grad_norm": 0.24469047784805298, "learning_rate": 8.31318876875869e-06, "loss": 0.0094, "step": 65850 }, { "grad_norm": 0.16955208778381348, "learning_rate": 8.30178002129039e-06, "loss": 0.0069, "step": 65860 }, { "grad_norm": 0.1915641576051712, "learning_rate": 8.290378398907423e-06, "loss": 0.0097, "step": 65870 }, { "grad_norm": 0.5933502316474915, "learning_rate": 8.278983903558029e-06, "loss": 0.0183, "step": 65880 }, { "grad_norm": 0.18310071527957916, "learning_rate": 8.26759653718921e-06, "loss": 0.0093, "step": 65890 }, { "grad_norm": 0.23419642448425293, "learning_rate": 8.25621630174676e-06, "loss": 0.0089, "step": 65900 }, { "grad_norm": 0.19848179817199707, "learning_rate": 8.244843199175261e-06, "loss": 0.0134, "step": 65910 }, { "grad_norm": 0.12621039152145386, "learning_rate": 8.233477231418046e-06, "loss": 0.007, "step": 65920 }, { "grad_norm": 0.2509082555770874, "learning_rate": 8.22211840041725e-06, "loss": 0.0123, "step": 65930 }, { "grad_norm": 0.33603471517562866, "learning_rate": 8.210766708113792e-06, "loss": 0.0142, "step": 65940 }, { "grad_norm": 0.15542864799499512, "learning_rate": 8.199422156447367e-06, "loss": 0.0099, "step": 65950 }, { "grad_norm": 0.22922329604625702, "learning_rate": 8.188084747356451e-06, "loss": 0.0124, "step": 65960 }, { "grad_norm": 0.15893273055553436, "learning_rate": 8.176754482778299e-06, "loss": 0.0081, "step": 65970 }, { "grad_norm": 0.17817513644695282, "learning_rate": 8.165431364648918e-06, "loss": 0.0142, "step": 65980 }, { "grad_norm": 0.2757965326309204, "learning_rate": 8.154115394903162e-06, "loss": 0.0116, "step": 65990 }, { "grad_norm": 0.2330307811498642, "learning_rate": 8.142806575474582e-06, "loss": 0.01, "step": 66000 }, { "grad_norm": 0.19968505203723907, "learning_rate": 8.131504908295562e-06, "loss": 0.0083, "step": 66010 }, { "grad_norm": 0.2360147386789322, "learning_rate": 8.120210395297262e-06, "loss": 0.0086, "step": 66020 }, { "grad_norm": 0.17215019464492798, "learning_rate": 8.108923038409565e-06, "loss": 0.0132, "step": 66030 }, { "grad_norm": 0.22954431176185608, "learning_rate": 8.097642839561226e-06, "loss": 0.0116, "step": 66040 }, { "grad_norm": 0.2853474020957947, "learning_rate": 8.08636980067966e-06, "loss": 0.0103, "step": 66050 }, { "grad_norm": 0.1647622287273407, "learning_rate": 8.075103923691186e-06, "loss": 0.0082, "step": 66060 }, { "grad_norm": 0.15058298408985138, "learning_rate": 8.063845210520793e-06, "loss": 0.0095, "step": 66070 }, { "grad_norm": 0.1699499487876892, "learning_rate": 8.052593663092295e-06, "loss": 0.0119, "step": 66080 }, { "grad_norm": 0.15617386996746063, "learning_rate": 8.04134928332827e-06, "loss": 0.0105, "step": 66090 }, { "grad_norm": 0.2249329388141632, "learning_rate": 8.030112073150086e-06, "loss": 0.0121, "step": 66100 }, { "grad_norm": 0.23382748663425446, "learning_rate": 8.018882034477881e-06, "loss": 0.009, "step": 66110 }, { "grad_norm": 0.11667381972074509, "learning_rate": 8.007659169230541e-06, "loss": 0.0079, "step": 66120 }, { "grad_norm": 0.25264066457748413, "learning_rate": 7.996443479325755e-06, "loss": 0.0112, "step": 66130 }, { "grad_norm": 0.06303641945123672, "learning_rate": 7.985234966679977e-06, "loss": 0.0094, "step": 66140 }, { "grad_norm": 0.1950313001871109, "learning_rate": 7.974033633208438e-06, "loss": 0.0078, "step": 66150 }, { "grad_norm": 0.11920412629842758, "learning_rate": 7.962839480825135e-06, "loss": 0.009, "step": 66160 }, { "grad_norm": 0.2384917140007019, "learning_rate": 7.951652511442858e-06, "loss": 0.0104, "step": 66170 }, { "grad_norm": 0.24958571791648865, "learning_rate": 7.940472726973125e-06, "loss": 0.0103, "step": 66180 }, { "grad_norm": 0.17489022016525269, "learning_rate": 7.929300129326289e-06, "loss": 0.011, "step": 66190 }, { "grad_norm": 0.23869752883911133, "learning_rate": 7.91813472041142e-06, "loss": 0.0097, "step": 66200 }, { "grad_norm": 0.1137799620628357, "learning_rate": 7.906976502136376e-06, "loss": 0.0126, "step": 66210 }, { "grad_norm": 0.15406231582164764, "learning_rate": 7.89582547640782e-06, "loss": 0.0122, "step": 66220 }, { "grad_norm": 0.22857597470283508, "learning_rate": 7.884681645131115e-06, "loss": 0.0082, "step": 66230 }, { "grad_norm": 0.24418173730373383, "learning_rate": 7.87354501021048e-06, "loss": 0.0093, "step": 66240 }, { "grad_norm": 0.22648842632770538, "learning_rate": 7.86241557354882e-06, "loss": 0.0095, "step": 66250 }, { "grad_norm": 0.20011919736862183, "learning_rate": 7.8512933370479e-06, "loss": 0.0102, "step": 66260 }, { "grad_norm": 0.16200508177280426, "learning_rate": 7.840178302608158e-06, "loss": 0.0088, "step": 66270 }, { "grad_norm": 0.11924420297145844, "learning_rate": 7.82907047212888e-06, "loss": 0.0092, "step": 66280 }, { "grad_norm": 0.17707285284996033, "learning_rate": 7.81796984750809e-06, "loss": 0.0093, "step": 66290 }, { "grad_norm": 0.13825708627700806, "learning_rate": 7.806876430642546e-06, "loss": 0.0096, "step": 66300 }, { "grad_norm": 0.4194372296333313, "learning_rate": 7.795790223427862e-06, "loss": 0.012, "step": 66310 }, { "grad_norm": 0.3067866265773773, "learning_rate": 7.784711227758324e-06, "loss": 0.0131, "step": 66320 }, { "grad_norm": 0.2893589735031128, "learning_rate": 7.773639445527053e-06, "loss": 0.0121, "step": 66330 }, { "grad_norm": 0.1133270189166069, "learning_rate": 7.762574878625905e-06, "loss": 0.0121, "step": 66340 }, { "grad_norm": 0.18113531172275543, "learning_rate": 7.751517528945513e-06, "loss": 0.0078, "step": 66350 }, { "grad_norm": 0.13886499404907227, "learning_rate": 7.740467398375278e-06, "loss": 0.009, "step": 66360 }, { "grad_norm": 0.17040108144283295, "learning_rate": 7.729424488803378e-06, "loss": 0.0109, "step": 66370 }, { "grad_norm": 0.34859922528266907, "learning_rate": 7.71838880211671e-06, "loss": 0.0138, "step": 66380 }, { "grad_norm": 0.28090211749076843, "learning_rate": 7.707360340200997e-06, "loss": 0.006, "step": 66390 }, { "grad_norm": 0.232291579246521, "learning_rate": 7.696339104940697e-06, "loss": 0.011, "step": 66400 }, { "grad_norm": 0.18284085392951965, "learning_rate": 7.685325098219038e-06, "loss": 0.0092, "step": 66410 }, { "grad_norm": 0.16206489503383636, "learning_rate": 7.674318321918017e-06, "loss": 0.01, "step": 66420 }, { "grad_norm": 0.14437220990657806, "learning_rate": 7.663318777918366e-06, "loss": 0.0112, "step": 66430 }, { "grad_norm": 0.16408319771289825, "learning_rate": 7.652326468099647e-06, "loss": 0.0132, "step": 66440 }, { "grad_norm": 0.17385345697402954, "learning_rate": 7.641341394340096e-06, "loss": 0.0079, "step": 66450 }, { "grad_norm": 0.35210272669792175, "learning_rate": 7.630363558516818e-06, "loss": 0.0095, "step": 66460 }, { "grad_norm": 0.20509088039398193, "learning_rate": 7.619392962505578e-06, "loss": 0.0117, "step": 66470 }, { "grad_norm": 0.14611288905143738, "learning_rate": 7.6084296081809725e-06, "loss": 0.0097, "step": 66480 }, { "grad_norm": 0.2276647537946701, "learning_rate": 7.597473497416347e-06, "loss": 0.01, "step": 66490 }, { "grad_norm": 0.2857561707496643, "learning_rate": 7.586524632083764e-06, "loss": 0.0072, "step": 66500 }, { "grad_norm": 0.146013543009758, "learning_rate": 7.5755830140541326e-06, "loss": 0.0077, "step": 66510 }, { "grad_norm": 0.11664427071809769, "learning_rate": 7.5646486451970475e-06, "loss": 0.0095, "step": 66520 }, { "grad_norm": 0.09156210720539093, "learning_rate": 7.553721527380897e-06, "loss": 0.0107, "step": 66530 }, { "grad_norm": 0.25167369842529297, "learning_rate": 7.542801662472826e-06, "loss": 0.0146, "step": 66540 }, { "grad_norm": 0.11104505509138107, "learning_rate": 7.5318890523387475e-06, "loss": 0.0084, "step": 66550 }, { "grad_norm": 0.19620086252689362, "learning_rate": 7.52098369884332e-06, "loss": 0.0086, "step": 66560 }, { "grad_norm": 0.24825817346572876, "learning_rate": 7.510085603849992e-06, "loss": 0.0121, "step": 66570 }, { "grad_norm": 0.26171228289604187, "learning_rate": 7.499194769220918e-06, "loss": 0.0102, "step": 66580 }, { "grad_norm": 0.4488421082496643, "learning_rate": 7.48831119681706e-06, "loss": 0.0171, "step": 66590 }, { "grad_norm": 0.2822832763195038, "learning_rate": 7.477434888498119e-06, "loss": 0.0099, "step": 66600 }, { "grad_norm": 0.17504772543907166, "learning_rate": 7.466565846122564e-06, "loss": 0.0135, "step": 66610 }, { "grad_norm": 0.2153870165348053, "learning_rate": 7.455704071547626e-06, "loss": 0.0083, "step": 66620 }, { "grad_norm": 0.11231555044651031, "learning_rate": 7.444849566629247e-06, "loss": 0.0094, "step": 66630 }, { "grad_norm": 0.28859761357307434, "learning_rate": 7.434002333222212e-06, "loss": 0.0071, "step": 66640 }, { "grad_norm": 0.26657232642173767, "learning_rate": 7.423162373179976e-06, "loss": 0.0114, "step": 66650 }, { "grad_norm": 0.23462149500846863, "learning_rate": 7.412329688354835e-06, "loss": 0.0123, "step": 66660 }, { "grad_norm": 0.16274943947792053, "learning_rate": 7.40150428059776e-06, "loss": 0.0083, "step": 66670 }, { "grad_norm": 0.2759654223918915, "learning_rate": 7.3906861517585354e-06, "loss": 0.0088, "step": 66680 }, { "grad_norm": 0.17179934680461884, "learning_rate": 7.37987530368569e-06, "loss": 0.0108, "step": 66690 }, { "grad_norm": 0.1518545299768448, "learning_rate": 7.369071738226474e-06, "loss": 0.0095, "step": 66700 }, { "grad_norm": 0.14738555252552032, "learning_rate": 7.358275457226954e-06, "loss": 0.0105, "step": 66710 }, { "grad_norm": 0.1777171641588211, "learning_rate": 7.347486462531899e-06, "loss": 0.015, "step": 66720 }, { "grad_norm": 0.13458633422851562, "learning_rate": 7.336704755984858e-06, "loss": 0.0089, "step": 66730 }, { "grad_norm": 0.09800738841295242, "learning_rate": 7.325930339428133e-06, "loss": 0.0116, "step": 66740 }, { "grad_norm": 0.23002105951309204, "learning_rate": 7.315163214702769e-06, "loss": 0.0095, "step": 66750 }, { "grad_norm": 0.531058669090271, "learning_rate": 7.304403383648595e-06, "loss": 0.0108, "step": 66760 }, { "grad_norm": 0.11208800971508026, "learning_rate": 7.293650848104139e-06, "loss": 0.0076, "step": 66770 }, { "grad_norm": 0.13861173391342163, "learning_rate": 7.2829056099067374e-06, "loss": 0.0092, "step": 66780 }, { "grad_norm": 0.20232145488262177, "learning_rate": 7.2721676708924494e-06, "loss": 0.0088, "step": 66790 }, { "grad_norm": 0.36130988597869873, "learning_rate": 7.261437032896096e-06, "loss": 0.0167, "step": 66800 }, { "grad_norm": 0.13090234994888306, "learning_rate": 7.250713697751255e-06, "loss": 0.0111, "step": 66810 }, { "grad_norm": 0.121758833527565, "learning_rate": 7.239997667290255e-06, "loss": 0.0099, "step": 66820 }, { "grad_norm": 0.0689171850681305, "learning_rate": 7.2292889433441425e-06, "loss": 0.0096, "step": 66830 }, { "grad_norm": 0.244666188955307, "learning_rate": 7.218587527742793e-06, "loss": 0.0092, "step": 66840 }, { "grad_norm": 0.21341648697853088, "learning_rate": 7.207893422314749e-06, "loss": 0.0092, "step": 66850 }, { "grad_norm": 0.19184690713882446, "learning_rate": 7.1972066288873545e-06, "loss": 0.0078, "step": 66860 }, { "grad_norm": 0.2567801773548126, "learning_rate": 7.186527149286687e-06, "loss": 0.0199, "step": 66870 }, { "grad_norm": 0.1870475709438324, "learning_rate": 7.175854985337576e-06, "loss": 0.0081, "step": 66880 }, { "grad_norm": 0.1296052485704422, "learning_rate": 7.16519013886362e-06, "loss": 0.0107, "step": 66890 }, { "grad_norm": 0.13238383829593658, "learning_rate": 7.154532611687109e-06, "loss": 0.0133, "step": 66900 }, { "grad_norm": 0.14777414500713348, "learning_rate": 7.143882405629177e-06, "loss": 0.0085, "step": 66910 }, { "grad_norm": 0.08679306507110596, "learning_rate": 7.133239522509605e-06, "loss": 0.0081, "step": 66920 }, { "grad_norm": 0.1125267967581749, "learning_rate": 7.1226039641469956e-06, "loss": 0.0076, "step": 66930 }, { "grad_norm": 0.2446165233850479, "learning_rate": 7.111975732358678e-06, "loss": 0.0084, "step": 66940 }, { "grad_norm": 0.09174378216266632, "learning_rate": 7.101354828960693e-06, "loss": 0.0077, "step": 66950 }, { "grad_norm": 0.2271941900253296, "learning_rate": 7.090741255767918e-06, "loss": 0.0097, "step": 66960 }, { "grad_norm": 0.19378235936164856, "learning_rate": 7.080135014593875e-06, "loss": 0.0089, "step": 66970 }, { "grad_norm": 0.11853880435228348, "learning_rate": 7.069536107250896e-06, "loss": 0.0069, "step": 66980 }, { "grad_norm": 0.24574178457260132, "learning_rate": 7.058944535550049e-06, "loss": 0.01, "step": 66990 }, { "grad_norm": 0.1660929024219513, "learning_rate": 7.048360301301138e-06, "loss": 0.0103, "step": 67000 }, { "grad_norm": 0.20169194042682648, "learning_rate": 7.03778340631272e-06, "loss": 0.0132, "step": 67010 }, { "grad_norm": 0.2998889088630676, "learning_rate": 7.0272138523921e-06, "loss": 0.013, "step": 67020 }, { "grad_norm": 0.11711385846138, "learning_rate": 7.016651641345334e-06, "loss": 0.0125, "step": 67030 }, { "grad_norm": 0.2500874996185303, "learning_rate": 7.0060967749772e-06, "loss": 0.0079, "step": 67040 }, { "grad_norm": 0.34209567308425903, "learning_rate": 6.995549255091238e-06, "loss": 0.0164, "step": 67050 }, { "grad_norm": 0.15857774019241333, "learning_rate": 6.98500908348973e-06, "loss": 0.0111, "step": 67060 }, { "grad_norm": 0.1526137888431549, "learning_rate": 6.974476261973711e-06, "loss": 0.0095, "step": 67070 }, { "grad_norm": 0.38483020663261414, "learning_rate": 6.963950792342949e-06, "loss": 0.01, "step": 67080 }, { "grad_norm": 0.11798062175512314, "learning_rate": 6.9534326763959715e-06, "loss": 0.0059, "step": 67090 }, { "grad_norm": 0.12707164883613586, "learning_rate": 6.942921915929995e-06, "loss": 0.009, "step": 67100 }, { "grad_norm": 0.09955660253763199, "learning_rate": 6.9324185127410735e-06, "loss": 0.0123, "step": 67110 }, { "grad_norm": 0.20745675265789032, "learning_rate": 6.921922468623921e-06, "loss": 0.0099, "step": 67120 }, { "grad_norm": 0.1739002764225006, "learning_rate": 6.911433785372023e-06, "loss": 0.0107, "step": 67130 }, { "grad_norm": 0.16688404977321625, "learning_rate": 6.900952464777632e-06, "loss": 0.0139, "step": 67140 }, { "grad_norm": 0.16052784025669098, "learning_rate": 6.8904785086316815e-06, "loss": 0.0071, "step": 67150 }, { "grad_norm": 0.22905759513378143, "learning_rate": 6.880011918723927e-06, "loss": 0.0082, "step": 67160 }, { "grad_norm": 0.2094099074602127, "learning_rate": 6.86955269684279e-06, "loss": 0.0095, "step": 67170 }, { "grad_norm": 0.19644950330257416, "learning_rate": 6.859100844775473e-06, "loss": 0.0202, "step": 67180 }, { "grad_norm": 0.13700024783611298, "learning_rate": 6.84865636430792e-06, "loss": 0.0102, "step": 67190 }, { "grad_norm": 0.24411426484584808, "learning_rate": 6.838219257224804e-06, "loss": 0.0118, "step": 67200 }, { "grad_norm": 0.38223379850387573, "learning_rate": 6.827789525309536e-06, "loss": 0.0174, "step": 67210 }, { "grad_norm": 0.2092248797416687, "learning_rate": 6.81736717034428e-06, "loss": 0.0106, "step": 67220 }, { "grad_norm": 0.2657434642314911, "learning_rate": 6.806952194109933e-06, "loss": 0.0166, "step": 67230 }, { "grad_norm": 0.09408833831548691, "learning_rate": 6.796544598386112e-06, "loss": 0.0065, "step": 67240 }, { "grad_norm": 0.16970482468605042, "learning_rate": 6.786144384951204e-06, "loss": 0.0113, "step": 67250 }, { "grad_norm": 0.12332461774349213, "learning_rate": 6.775751555582322e-06, "loss": 0.0092, "step": 67260 }, { "grad_norm": 0.1410437673330307, "learning_rate": 6.76536611205531e-06, "loss": 0.0106, "step": 67270 }, { "grad_norm": 0.25564247369766235, "learning_rate": 6.754988056144762e-06, "loss": 0.0171, "step": 67280 }, { "grad_norm": 0.10139190405607224, "learning_rate": 6.744617389624014e-06, "loss": 0.0081, "step": 67290 }, { "grad_norm": 0.34611594676971436, "learning_rate": 6.734254114265087e-06, "loss": 0.0095, "step": 67300 }, { "grad_norm": 0.16821689903736115, "learning_rate": 6.723898231838843e-06, "loss": 0.0072, "step": 67310 }, { "grad_norm": 0.16818849742412567, "learning_rate": 6.713549744114766e-06, "loss": 0.0132, "step": 67320 }, { "grad_norm": 0.15870340168476105, "learning_rate": 6.703208652861159e-06, "loss": 0.0117, "step": 67330 }, { "grad_norm": 0.3638278543949127, "learning_rate": 6.69287495984503e-06, "loss": 0.0135, "step": 67340 }, { "grad_norm": 0.11021564900875092, "learning_rate": 6.6825486668321e-06, "loss": 0.0072, "step": 67350 }, { "grad_norm": 0.19648905098438263, "learning_rate": 6.672229775586886e-06, "loss": 0.0114, "step": 67360 }, { "grad_norm": 0.1701938956975937, "learning_rate": 6.661918287872576e-06, "loss": 0.0081, "step": 67370 }, { "grad_norm": 0.14539088308811188, "learning_rate": 6.6516142054511346e-06, "loss": 0.0122, "step": 67380 }, { "grad_norm": 0.2628267705440521, "learning_rate": 6.641317530083241e-06, "loss": 0.0085, "step": 67390 }, { "grad_norm": 0.18944498896598816, "learning_rate": 6.631028263528322e-06, "loss": 0.0103, "step": 67400 }, { "grad_norm": 0.19418653845787048, "learning_rate": 6.620746407544537e-06, "loss": 0.0118, "step": 67410 }, { "grad_norm": 0.15772952139377594, "learning_rate": 6.610471963888742e-06, "loss": 0.0127, "step": 67420 }, { "grad_norm": 0.10359419137239456, "learning_rate": 6.600204934316606e-06, "loss": 0.0061, "step": 67430 }, { "grad_norm": 0.1792600005865097, "learning_rate": 6.589945320582452e-06, "loss": 0.0106, "step": 67440 }, { "grad_norm": 0.16080018877983093, "learning_rate": 6.579693124439374e-06, "loss": 0.0106, "step": 67450 }, { "grad_norm": 0.1851535439491272, "learning_rate": 6.56944834763919e-06, "loss": 0.0111, "step": 67460 }, { "grad_norm": 0.5236860513687134, "learning_rate": 6.5592109919324575e-06, "loss": 0.0092, "step": 67470 }, { "grad_norm": 0.23632386326789856, "learning_rate": 6.54898105906846e-06, "loss": 0.0143, "step": 67480 }, { "grad_norm": 0.10342004895210266, "learning_rate": 6.5387585507952155e-06, "loss": 0.0162, "step": 67490 }, { "grad_norm": 0.1533394753932953, "learning_rate": 6.528543468859461e-06, "loss": 0.009, "step": 67500 }, { "grad_norm": 0.19107544422149658, "learning_rate": 6.518335815006682e-06, "loss": 0.0119, "step": 67510 }, { "grad_norm": 0.25907909870147705, "learning_rate": 6.5081355909810845e-06, "loss": 0.0111, "step": 67520 }, { "grad_norm": 0.21910329163074493, "learning_rate": 6.497942798525608e-06, "loss": 0.0127, "step": 67530 }, { "grad_norm": 0.3376120328903198, "learning_rate": 6.487757439381936e-06, "loss": 0.0094, "step": 67540 }, { "grad_norm": 0.16400238871574402, "learning_rate": 6.477579515290433e-06, "loss": 0.0083, "step": 67550 }, { "grad_norm": 0.11746101081371307, "learning_rate": 6.467409027990273e-06, "loss": 0.014, "step": 67560 }, { "grad_norm": 0.1374082714319229, "learning_rate": 6.457245979219279e-06, "loss": 0.0106, "step": 67570 }, { "grad_norm": 0.3312512934207916, "learning_rate": 6.447090370714054e-06, "loss": 0.0161, "step": 67580 }, { "grad_norm": 0.11596120148897171, "learning_rate": 6.436942204209917e-06, "loss": 0.0127, "step": 67590 }, { "grad_norm": 0.29060134291648865, "learning_rate": 6.4268014814408804e-06, "loss": 0.0168, "step": 67600 }, { "grad_norm": 0.18942120671272278, "learning_rate": 6.41666820413977e-06, "loss": 0.0112, "step": 67610 }, { "grad_norm": 0.3307761251926422, "learning_rate": 6.406542374038033e-06, "loss": 0.0143, "step": 67620 }, { "grad_norm": 0.1530797928571701, "learning_rate": 6.396423992865935e-06, "loss": 0.0132, "step": 67630 }, { "grad_norm": 0.10860218852758408, "learning_rate": 6.386313062352412e-06, "loss": 0.0089, "step": 67640 }, { "grad_norm": 0.15471352636814117, "learning_rate": 6.376209584225152e-06, "loss": 0.0077, "step": 67650 }, { "grad_norm": 0.20167867839336395, "learning_rate": 6.366113560210557e-06, "loss": 0.0073, "step": 67660 }, { "grad_norm": 0.38401588797569275, "learning_rate": 6.356024992033766e-06, "loss": 0.015, "step": 67670 }, { "grad_norm": 0.1830083727836609, "learning_rate": 6.345943881418648e-06, "loss": 0.0076, "step": 67680 }, { "grad_norm": 0.15264473855495453, "learning_rate": 6.33587023008777e-06, "loss": 0.0098, "step": 67690 }, { "grad_norm": 0.17279303073883057, "learning_rate": 6.32580403976245e-06, "loss": 0.0127, "step": 67700 }, { "grad_norm": 0.21429041028022766, "learning_rate": 6.31574531216273e-06, "loss": 0.0111, "step": 67710 }, { "grad_norm": 0.24831083416938782, "learning_rate": 6.305694049007371e-06, "loss": 0.0143, "step": 67720 }, { "grad_norm": 0.18626607954502106, "learning_rate": 6.2956502520138575e-06, "loss": 0.0091, "step": 67730 }, { "grad_norm": 0.15185636281967163, "learning_rate": 6.285613922898409e-06, "loss": 0.0071, "step": 67740 }, { "grad_norm": 0.14709171652793884, "learning_rate": 6.275585063375927e-06, "loss": 0.0102, "step": 67750 }, { "grad_norm": 0.1459733247756958, "learning_rate": 6.265563675160113e-06, "loss": 0.0094, "step": 67760 }, { "grad_norm": 0.14737500250339508, "learning_rate": 6.25554975996332e-06, "loss": 0.0079, "step": 67770 }, { "grad_norm": 0.1509092003107071, "learning_rate": 6.245543319496661e-06, "loss": 0.0087, "step": 67780 }, { "grad_norm": 0.15665556490421295, "learning_rate": 6.2355443554699685e-06, "loss": 0.0095, "step": 67790 }, { "grad_norm": 0.08131066709756851, "learning_rate": 6.225552869591766e-06, "loss": 0.0073, "step": 67800 }, { "grad_norm": 0.15402187407016754, "learning_rate": 6.215568863569365e-06, "loss": 0.0125, "step": 67810 }, { "grad_norm": 0.3690434396266937, "learning_rate": 6.205592339108712e-06, "loss": 0.0148, "step": 67820 }, { "grad_norm": 0.16266728937625885, "learning_rate": 6.195623297914577e-06, "loss": 0.0108, "step": 67830 }, { "grad_norm": 0.2789618968963623, "learning_rate": 6.185661741690357e-06, "loss": 0.0107, "step": 67840 }, { "grad_norm": 0.16796371340751648, "learning_rate": 6.1757076721382145e-06, "loss": 0.0112, "step": 67850 }, { "grad_norm": 0.15225668251514435, "learning_rate": 6.165761090959038e-06, "loss": 0.0158, "step": 67860 }, { "grad_norm": 0.18910403549671173, "learning_rate": 6.155821999852424e-06, "loss": 0.0085, "step": 67870 }, { "grad_norm": 0.1420619636774063, "learning_rate": 6.145890400516696e-06, "loss": 0.0101, "step": 67880 }, { "grad_norm": 0.18985716998577118, "learning_rate": 6.1359662946488816e-06, "loss": 0.0087, "step": 67890 }, { "grad_norm": 0.20139016211032867, "learning_rate": 6.12604968394474e-06, "loss": 0.0085, "step": 67900 }, { "grad_norm": 0.21375876665115356, "learning_rate": 6.11614057009875e-06, "loss": 0.0103, "step": 67910 }, { "grad_norm": 0.25724878907203674, "learning_rate": 6.106238954804111e-06, "loss": 0.0126, "step": 67920 }, { "grad_norm": 0.16501425206661224, "learning_rate": 6.096344839752738e-06, "loss": 0.0108, "step": 67930 }, { "grad_norm": 0.10029324889183044, "learning_rate": 6.086458226635278e-06, "loss": 0.011, "step": 67940 }, { "grad_norm": 0.2515626847743988, "learning_rate": 6.076579117141046e-06, "loss": 0.019, "step": 67950 }, { "grad_norm": 0.305288702249527, "learning_rate": 6.066707512958153e-06, "loss": 0.0138, "step": 67960 }, { "grad_norm": 0.3055846691131592, "learning_rate": 6.056843415773361e-06, "loss": 0.0121, "step": 67970 }, { "grad_norm": 0.16154742240905762, "learning_rate": 6.0469868272721776e-06, "loss": 0.0101, "step": 67980 }, { "grad_norm": 0.15357668697834015, "learning_rate": 6.037137749138844e-06, "loss": 0.0125, "step": 67990 }, { "grad_norm": 0.2445686310529709, "learning_rate": 6.027296183056252e-06, "loss": 0.0134, "step": 68000 }, { "grad_norm": 0.16246360540390015, "learning_rate": 6.017462130706114e-06, "loss": 0.0082, "step": 68010 }, { "grad_norm": 0.14259129762649536, "learning_rate": 6.0076355937687516e-06, "loss": 0.0119, "step": 68020 }, { "grad_norm": 0.10939988493919373, "learning_rate": 5.9978165739232925e-06, "loss": 0.0086, "step": 68030 }, { "grad_norm": 0.46163442730903625, "learning_rate": 5.988005072847508e-06, "loss": 0.0078, "step": 68040 }, { "grad_norm": 0.16363824903964996, "learning_rate": 5.978201092217928e-06, "loss": 0.0191, "step": 68050 }, { "grad_norm": 0.24458050727844238, "learning_rate": 5.9684046337097895e-06, "loss": 0.0068, "step": 68060 }, { "grad_norm": 0.09113077074289322, "learning_rate": 5.958615698997016e-06, "loss": 0.012, "step": 68070 }, { "grad_norm": 0.19659318029880524, "learning_rate": 5.948834289752303e-06, "loss": 0.0079, "step": 68080 }, { "grad_norm": 0.13677255809307098, "learning_rate": 5.939060407646996e-06, "loss": 0.0132, "step": 68090 }, { "grad_norm": 0.19890642166137695, "learning_rate": 5.929294054351198e-06, "loss": 0.0132, "step": 68100 }, { "grad_norm": 0.2115347981452942, "learning_rate": 5.919535231533707e-06, "loss": 0.0099, "step": 68110 }, { "grad_norm": 0.18202556669712067, "learning_rate": 5.909783940862046e-06, "loss": 0.0076, "step": 68120 }, { "grad_norm": 0.1600828617811203, "learning_rate": 5.900040184002436e-06, "loss": 0.0104, "step": 68130 }, { "grad_norm": 0.18966978788375854, "learning_rate": 5.890303962619831e-06, "loss": 0.007, "step": 68140 }, { "grad_norm": 0.22749441862106323, "learning_rate": 5.88057527837787e-06, "loss": 0.0108, "step": 68150 }, { "grad_norm": 0.14892269670963287, "learning_rate": 5.870854132938924e-06, "loss": 0.0135, "step": 68160 }, { "grad_norm": 0.3083861172199249, "learning_rate": 5.861140527964071e-06, "loss": 0.0157, "step": 68170 }, { "grad_norm": 0.09785683453083038, "learning_rate": 5.851434465113098e-06, "loss": 0.0071, "step": 68180 }, { "grad_norm": 0.17886556684970856, "learning_rate": 5.841735946044524e-06, "loss": 0.0116, "step": 68190 }, { "grad_norm": 0.13093872368335724, "learning_rate": 5.832044972415523e-06, "loss": 0.0111, "step": 68200 }, { "grad_norm": 0.15546654164791107, "learning_rate": 5.8223615458820655e-06, "loss": 0.0094, "step": 68210 }, { "grad_norm": 0.0867006704211235, "learning_rate": 5.812685668098733e-06, "loss": 0.0102, "step": 68220 }, { "grad_norm": 0.16927890479564667, "learning_rate": 5.803017340718913e-06, "loss": 0.0113, "step": 68230 }, { "grad_norm": 0.14208753407001495, "learning_rate": 5.793356565394636e-06, "loss": 0.0095, "step": 68240 }, { "grad_norm": 0.2125440388917923, "learning_rate": 5.783703343776658e-06, "loss": 0.0109, "step": 68250 }, { "grad_norm": 0.18052679300308228, "learning_rate": 5.774057677514477e-06, "loss": 0.0076, "step": 68260 }, { "grad_norm": 0.24091975390911102, "learning_rate": 5.764419568256235e-06, "loss": 0.009, "step": 68270 }, { "grad_norm": 0.2933136522769928, "learning_rate": 5.754789017648865e-06, "loss": 0.0181, "step": 68280 }, { "grad_norm": 0.19007621705532074, "learning_rate": 5.745166027337934e-06, "loss": 0.0114, "step": 68290 }, { "grad_norm": 0.14790485799312592, "learning_rate": 5.735550598967754e-06, "loss": 0.0108, "step": 68300 }, { "grad_norm": 0.2833476662635803, "learning_rate": 5.725942734181339e-06, "loss": 0.0102, "step": 68310 }, { "grad_norm": 0.22917276620864868, "learning_rate": 5.7163424346204146e-06, "loss": 0.0088, "step": 68320 }, { "grad_norm": 0.18994635343551636, "learning_rate": 5.706749701925407e-06, "loss": 0.0096, "step": 68330 }, { "grad_norm": 0.14514461159706116, "learning_rate": 5.697164537735461e-06, "loss": 0.0083, "step": 68340 }, { "grad_norm": 0.20145010948181152, "learning_rate": 5.687586943688406e-06, "loss": 0.0094, "step": 68350 }, { "grad_norm": 0.15085536241531372, "learning_rate": 5.678016921420792e-06, "loss": 0.0102, "step": 68360 }, { "grad_norm": 0.1874011754989624, "learning_rate": 5.668454472567875e-06, "loss": 0.014, "step": 68370 }, { "grad_norm": 0.23746514320373535, "learning_rate": 5.658899598763617e-06, "loss": 0.0113, "step": 68380 }, { "grad_norm": 0.13963353633880615, "learning_rate": 5.6493523016407035e-06, "loss": 0.0083, "step": 68390 }, { "grad_norm": 0.12756465375423431, "learning_rate": 5.6398125828304615e-06, "loss": 0.0094, "step": 68400 }, { "grad_norm": 0.16948167979717255, "learning_rate": 5.630280443963015e-06, "loss": 0.011, "step": 68410 }, { "grad_norm": 0.2595258951187134, "learning_rate": 5.620755886667112e-06, "loss": 0.0117, "step": 68420 }, { "grad_norm": 0.2311277538537979, "learning_rate": 5.611238912570271e-06, "loss": 0.0114, "step": 68430 }, { "grad_norm": 0.27180778980255127, "learning_rate": 5.601729523298649e-06, "loss": 0.0084, "step": 68440 }, { "grad_norm": 0.1388825625181198, "learning_rate": 5.592227720477161e-06, "loss": 0.0115, "step": 68450 }, { "grad_norm": 0.25950729846954346, "learning_rate": 5.582733505729415e-06, "loss": 0.0113, "step": 68460 }, { "grad_norm": 0.25060614943504333, "learning_rate": 5.573246880677668e-06, "loss": 0.0123, "step": 68470 }, { "grad_norm": 0.30024904012680054, "learning_rate": 5.56376784694298e-06, "loss": 0.0127, "step": 68480 }, { "grad_norm": 0.3844282031059265, "learning_rate": 5.554296406145027e-06, "loss": 0.0155, "step": 68490 }, { "grad_norm": 0.23543043434619904, "learning_rate": 5.544832559902219e-06, "loss": 0.0103, "step": 68500 }, { "grad_norm": 0.24216613173484802, "learning_rate": 5.53537630983168e-06, "loss": 0.0096, "step": 68510 }, { "grad_norm": 0.20668719708919525, "learning_rate": 5.5259276575492125e-06, "loss": 0.015, "step": 68520 }, { "grad_norm": 0.13423657417297363, "learning_rate": 5.516486604669357e-06, "loss": 0.0135, "step": 68530 }, { "grad_norm": 0.152884379029274, "learning_rate": 5.507053152805303e-06, "loss": 0.0158, "step": 68540 }, { "grad_norm": 0.1637382209300995, "learning_rate": 5.497627303568975e-06, "loss": 0.0136, "step": 68550 }, { "grad_norm": 0.0935051292181015, "learning_rate": 5.488209058571003e-06, "loss": 0.0073, "step": 68560 }, { "grad_norm": 0.18620795011520386, "learning_rate": 5.4787984194207054e-06, "loss": 0.008, "step": 68570 }, { "grad_norm": 0.18440113961696625, "learning_rate": 5.469395387726095e-06, "loss": 0.0102, "step": 68580 }, { "grad_norm": 0.1800064742565155, "learning_rate": 5.45999996509391e-06, "loss": 0.0092, "step": 68590 }, { "grad_norm": 0.183782160282135, "learning_rate": 5.450612153129536e-06, "loss": 0.0109, "step": 68600 }, { "grad_norm": 0.16363373398780823, "learning_rate": 5.4412319534371426e-06, "loss": 0.0097, "step": 68610 }, { "grad_norm": 0.23168575763702393, "learning_rate": 5.431859367619513e-06, "loss": 0.0079, "step": 68620 }, { "grad_norm": 0.28278571367263794, "learning_rate": 5.422494397278172e-06, "loss": 0.0122, "step": 68630 }, { "grad_norm": 0.25736114382743835, "learning_rate": 5.413137044013344e-06, "loss": 0.0092, "step": 68640 }, { "grad_norm": 0.1398867964744568, "learning_rate": 5.403787309423941e-06, "loss": 0.0097, "step": 68650 }, { "grad_norm": 0.21458765864372253, "learning_rate": 5.3944451951075835e-06, "loss": 0.0091, "step": 68660 }, { "grad_norm": 0.13601434230804443, "learning_rate": 5.385110702660562e-06, "loss": 0.007, "step": 68670 }, { "grad_norm": 0.11557036638259888, "learning_rate": 5.375783833677922e-06, "loss": 0.0152, "step": 68680 }, { "grad_norm": 0.24982409179210663, "learning_rate": 5.36646458975334e-06, "loss": 0.0113, "step": 68690 }, { "grad_norm": 0.1531311571598053, "learning_rate": 5.3571529724792294e-06, "loss": 0.0062, "step": 68700 }, { "grad_norm": 0.9137638807296753, "learning_rate": 5.347848983446702e-06, "loss": 0.0097, "step": 68710 }, { "grad_norm": 0.2488725781440735, "learning_rate": 5.3385526242455185e-06, "loss": 0.0129, "step": 68720 }, { "grad_norm": 0.11539725214242935, "learning_rate": 5.329263896464226e-06, "loss": 0.0094, "step": 68730 }, { "grad_norm": 0.09833551198244095, "learning_rate": 5.3199828016899715e-06, "loss": 0.0081, "step": 68740 }, { "grad_norm": 0.17427197098731995, "learning_rate": 5.310709341508657e-06, "loss": 0.009, "step": 68750 }, { "grad_norm": 0.17041462659835815, "learning_rate": 5.301443517504861e-06, "loss": 0.0131, "step": 68760 }, { "grad_norm": 0.13465923070907593, "learning_rate": 5.292185331261862e-06, "loss": 0.0064, "step": 68770 }, { "grad_norm": 0.2286687046289444, "learning_rate": 5.28293478436162e-06, "loss": 0.0091, "step": 68780 }, { "grad_norm": 0.18849362432956696, "learning_rate": 5.273691878384829e-06, "loss": 0.013, "step": 68790 }, { "grad_norm": 0.1540256291627884, "learning_rate": 5.264456614910812e-06, "loss": 0.0102, "step": 68800 }, { "grad_norm": 0.130344957113266, "learning_rate": 5.255228995517647e-06, "loss": 0.0171, "step": 68810 }, { "grad_norm": 0.1650032252073288, "learning_rate": 5.246009021782067e-06, "loss": 0.0094, "step": 68820 }, { "grad_norm": 0.098884217441082, "learning_rate": 5.2367966952795225e-06, "loss": 0.0145, "step": 68830 }, { "grad_norm": 0.11084315180778503, "learning_rate": 5.2275920175841485e-06, "loss": 0.0092, "step": 68840 }, { "grad_norm": 0.07860121130943298, "learning_rate": 5.218394990268771e-06, "loss": 0.0093, "step": 68850 }, { "grad_norm": 0.3671659231185913, "learning_rate": 5.209205614904916e-06, "loss": 0.01, "step": 68860 }, { "grad_norm": 0.2753659784793854, "learning_rate": 5.200023893062777e-06, "loss": 0.0106, "step": 68870 }, { "grad_norm": 0.09972717612981796, "learning_rate": 5.190849826311289e-06, "loss": 0.0076, "step": 68880 }, { "grad_norm": 0.2704043984413147, "learning_rate": 5.181683416218025e-06, "loss": 0.0113, "step": 68890 }, { "grad_norm": 0.12543021142482758, "learning_rate": 5.172524664349276e-06, "loss": 0.008, "step": 68900 }, { "grad_norm": 0.41149717569351196, "learning_rate": 5.1633735722700416e-06, "loss": 0.0086, "step": 68910 }, { "grad_norm": 0.1190139502286911, "learning_rate": 5.154230141543958e-06, "loss": 0.0098, "step": 68920 }, { "grad_norm": 0.4243958592414856, "learning_rate": 5.145094373733433e-06, "loss": 0.0109, "step": 68930 }, { "grad_norm": 0.10653913021087646, "learning_rate": 5.135966270399478e-06, "loss": 0.0066, "step": 68940 }, { "grad_norm": 0.4383002519607544, "learning_rate": 5.126845833101857e-06, "loss": 0.0115, "step": 68950 }, { "grad_norm": 0.13476383686065674, "learning_rate": 5.1177330633989994e-06, "loss": 0.0074, "step": 68960 }, { "grad_norm": 0.3231612741947174, "learning_rate": 5.108627962848033e-06, "loss": 0.0208, "step": 68970 }, { "grad_norm": 0.17378386855125427, "learning_rate": 5.099530533004759e-06, "loss": 0.014, "step": 68980 }, { "grad_norm": 0.11892599612474442, "learning_rate": 5.090440775423699e-06, "loss": 0.0074, "step": 68990 }, { "grad_norm": 0.2209336757659912, "learning_rate": 5.081358691658022e-06, "loss": 0.0072, "step": 69000 }, { "grad_norm": 0.20767512917518616, "learning_rate": 5.072284283259621e-06, "loss": 0.0099, "step": 69010 }, { "grad_norm": 0.1136518195271492, "learning_rate": 5.063217551779053e-06, "loss": 0.0082, "step": 69020 }, { "grad_norm": 0.12174232304096222, "learning_rate": 5.054158498765588e-06, "loss": 0.0074, "step": 69030 }, { "grad_norm": 0.26353004574775696, "learning_rate": 5.0451071257671625e-06, "loss": 0.0106, "step": 69040 }, { "grad_norm": 0.2096206098794937, "learning_rate": 5.036063434330407e-06, "loss": 0.013, "step": 69050 }, { "grad_norm": 0.4405270516872406, "learning_rate": 5.027027426000652e-06, "loss": 0.0093, "step": 69060 }, { "grad_norm": 0.2145432084798813, "learning_rate": 5.017999102321886e-06, "loss": 0.0112, "step": 69070 }, { "grad_norm": 0.21122775971889496, "learning_rate": 5.0089784648368224e-06, "loss": 0.0128, "step": 69080 }, { "grad_norm": 0.3019452393054962, "learning_rate": 4.999965515086829e-06, "loss": 0.0081, "step": 69090 }, { "grad_norm": 0.14318829774856567, "learning_rate": 4.990960254611976e-06, "loss": 0.0072, "step": 69100 }, { "grad_norm": 0.35817399621009827, "learning_rate": 4.9819626849510194e-06, "loss": 0.0085, "step": 69110 }, { "grad_norm": 0.16873134672641754, "learning_rate": 4.97297280764138e-06, "loss": 0.0144, "step": 69120 }, { "grad_norm": 0.13388226926326752, "learning_rate": 4.963990624219211e-06, "loss": 0.0114, "step": 69130 }, { "grad_norm": 0.25544822216033936, "learning_rate": 4.955016136219298e-06, "loss": 0.0097, "step": 69140 }, { "grad_norm": 0.18757866322994232, "learning_rate": 4.946049345175146e-06, "loss": 0.0077, "step": 69150 }, { "grad_norm": 0.25970789790153503, "learning_rate": 4.937090252618937e-06, "loss": 0.0091, "step": 69160 }, { "grad_norm": 0.15019427239894867, "learning_rate": 4.928138860081521e-06, "loss": 0.0146, "step": 69170 }, { "grad_norm": 0.10628820210695267, "learning_rate": 4.919195169092472e-06, "loss": 0.0119, "step": 69180 }, { "grad_norm": 0.12372106313705444, "learning_rate": 4.910259181179994e-06, "loss": 0.0117, "step": 69190 }, { "grad_norm": 0.23949867486953735, "learning_rate": 4.901330897871015e-06, "loss": 0.0117, "step": 69200 }, { "grad_norm": 0.1298610419034958, "learning_rate": 4.8924103206911375e-06, "loss": 0.0073, "step": 69210 }, { "grad_norm": 0.1265319585800171, "learning_rate": 4.883497451164637e-06, "loss": 0.0116, "step": 69220 }, { "grad_norm": 0.23892323672771454, "learning_rate": 4.874592290814484e-06, "loss": 0.011, "step": 69230 }, { "grad_norm": 0.3162428140640259, "learning_rate": 4.865694841162327e-06, "loss": 0.0174, "step": 69240 }, { "grad_norm": 0.11760719120502472, "learning_rate": 4.856805103728496e-06, "loss": 0.0091, "step": 69250 }, { "grad_norm": 0.14676359295845032, "learning_rate": 4.847923080032007e-06, "loss": 0.0136, "step": 69260 }, { "grad_norm": 0.15625911951065063, "learning_rate": 4.8390487715905486e-06, "loss": 0.0097, "step": 69270 }, { "grad_norm": 0.18157973885536194, "learning_rate": 4.8301821799205e-06, "loss": 0.0084, "step": 69280 }, { "grad_norm": 0.09610479325056076, "learning_rate": 4.821323306536918e-06, "loss": 0.0068, "step": 69290 }, { "grad_norm": 0.11654523015022278, "learning_rate": 4.8124721529535455e-06, "loss": 0.0087, "step": 69300 }, { "grad_norm": 0.1836719661951065, "learning_rate": 4.803628720682807e-06, "loss": 0.0134, "step": 69310 }, { "grad_norm": 0.17315152287483215, "learning_rate": 4.794793011235776e-06, "loss": 0.009, "step": 69320 }, { "grad_norm": 0.08503145724534988, "learning_rate": 4.78596502612228e-06, "loss": 0.0093, "step": 69330 }, { "grad_norm": 0.2002987116575241, "learning_rate": 4.777144766850738e-06, "loss": 0.0138, "step": 69340 }, { "grad_norm": 0.14737704396247864, "learning_rate": 4.768332234928313e-06, "loss": 0.0134, "step": 69350 }, { "grad_norm": 0.12483392655849457, "learning_rate": 4.759527431860828e-06, "loss": 0.007, "step": 69360 }, { "grad_norm": 0.17757977545261383, "learning_rate": 4.750730359152755e-06, "loss": 0.0146, "step": 69370 }, { "grad_norm": 0.13857011497020721, "learning_rate": 4.741941018307311e-06, "loss": 0.0116, "step": 69380 }, { "grad_norm": 0.16740268468856812, "learning_rate": 4.733159410826321e-06, "loss": 0.0119, "step": 69390 }, { "grad_norm": 0.12362090498209, "learning_rate": 4.724385538210357e-06, "loss": 0.0079, "step": 69400 }, { "grad_norm": 0.21213513612747192, "learning_rate": 4.715619401958599e-06, "loss": 0.0104, "step": 69410 }, { "grad_norm": 0.17693379521369934, "learning_rate": 4.706861003568958e-06, "loss": 0.0117, "step": 69420 }, { "grad_norm": 0.24125051498413086, "learning_rate": 4.698110344538003e-06, "loss": 0.0095, "step": 69430 }, { "grad_norm": 0.23240429162979126, "learning_rate": 4.689367426360975e-06, "loss": 0.008, "step": 69440 }, { "grad_norm": 0.26034948229789734, "learning_rate": 4.680632250531819e-06, "loss": 0.0077, "step": 69450 }, { "grad_norm": 0.08684183657169342, "learning_rate": 4.671904818543115e-06, "loss": 0.0113, "step": 69460 }, { "grad_norm": 0.19684892892837524, "learning_rate": 4.66318513188615e-06, "loss": 0.0073, "step": 69470 }, { "grad_norm": 0.2651251554489136, "learning_rate": 4.654473192050884e-06, "loss": 0.0113, "step": 69480 }, { "grad_norm": 0.18752969801425934, "learning_rate": 4.64576900052594e-06, "loss": 0.0114, "step": 69490 }, { "grad_norm": 0.16211605072021484, "learning_rate": 4.637072558798638e-06, "loss": 0.0083, "step": 69500 }, { "grad_norm": 0.21639464795589447, "learning_rate": 4.628383868354969e-06, "loss": 0.0098, "step": 69510 }, { "grad_norm": 0.14596417546272278, "learning_rate": 4.6197029306795595e-06, "loss": 0.0074, "step": 69520 }, { "grad_norm": 0.13557098805904388, "learning_rate": 4.611029747255779e-06, "loss": 0.0096, "step": 69530 }, { "grad_norm": 0.18803563714027405, "learning_rate": 4.6023643195656164e-06, "loss": 0.0098, "step": 69540 }, { "grad_norm": 0.09488207846879959, "learning_rate": 4.593706649089768e-06, "loss": 0.0085, "step": 69550 }, { "grad_norm": 0.26903027296066284, "learning_rate": 4.585056737307597e-06, "loss": 0.0115, "step": 69560 }, { "grad_norm": 0.12487651407718658, "learning_rate": 4.576414585697103e-06, "loss": 0.007, "step": 69570 }, { "grad_norm": 0.3970640301704407, "learning_rate": 4.567780195735044e-06, "loss": 0.0109, "step": 69580 }, { "grad_norm": 0.17571866512298584, "learning_rate": 4.559153568896757e-06, "loss": 0.007, "step": 69590 }, { "grad_norm": 0.4141640365123749, "learning_rate": 4.550534706656329e-06, "loss": 0.0131, "step": 69600 }, { "grad_norm": 0.12384574115276337, "learning_rate": 4.541923610486465e-06, "loss": 0.0091, "step": 69610 }, { "grad_norm": 0.18168611824512482, "learning_rate": 4.533320281858578e-06, "loss": 0.0092, "step": 69620 }, { "grad_norm": 0.1397518813610077, "learning_rate": 4.52472472224274e-06, "loss": 0.0141, "step": 69630 }, { "grad_norm": 0.147012397646904, "learning_rate": 4.51613693310769e-06, "loss": 0.0113, "step": 69640 }, { "grad_norm": 0.12048559635877609, "learning_rate": 4.507556915920868e-06, "loss": 0.0101, "step": 69650 }, { "grad_norm": 0.1450357288122177, "learning_rate": 4.498984672148332e-06, "loss": 0.0092, "step": 69660 }, { "grad_norm": 0.13189448416233063, "learning_rate": 4.490420203254864e-06, "loss": 0.0104, "step": 69670 }, { "grad_norm": 0.35759949684143066, "learning_rate": 4.481863510703893e-06, "loss": 0.0131, "step": 69680 }, { "grad_norm": 0.4324384331703186, "learning_rate": 4.473314595957523e-06, "loss": 0.0089, "step": 69690 }, { "grad_norm": 0.21115633845329285, "learning_rate": 4.464773460476535e-06, "loss": 0.0101, "step": 69700 }, { "grad_norm": 0.16822107136249542, "learning_rate": 4.456240105720372e-06, "loss": 0.0099, "step": 69710 }, { "grad_norm": 0.12510442733764648, "learning_rate": 4.4477145331471405e-06, "loss": 0.0174, "step": 69720 }, { "grad_norm": 0.18473009765148163, "learning_rate": 4.439196744213653e-06, "loss": 0.0104, "step": 69730 }, { "grad_norm": 0.17473578453063965, "learning_rate": 4.430686740375339e-06, "loss": 0.0098, "step": 69740 }, { "grad_norm": 0.12532198429107666, "learning_rate": 4.422184523086342e-06, "loss": 0.0092, "step": 69750 }, { "grad_norm": 0.2653057277202606, "learning_rate": 4.41369009379946e-06, "loss": 0.011, "step": 69760 }, { "grad_norm": 0.27970778942108154, "learning_rate": 4.405203453966139e-06, "loss": 0.0088, "step": 69770 }, { "grad_norm": 0.31405922770500183, "learning_rate": 4.396724605036539e-06, "loss": 0.0095, "step": 69780 }, { "grad_norm": 0.16968552768230438, "learning_rate": 4.388253548459437e-06, "loss": 0.0082, "step": 69790 }, { "grad_norm": 0.11382703483104706, "learning_rate": 4.3797902856823395e-06, "loss": 0.0088, "step": 69800 }, { "grad_norm": 0.39186564087867737, "learning_rate": 4.371334818151357e-06, "loss": 0.0091, "step": 69810 }, { "grad_norm": 0.22002901136875153, "learning_rate": 4.362887147311306e-06, "loss": 0.0092, "step": 69820 }, { "grad_norm": 0.1084006279706955, "learning_rate": 4.354447274605672e-06, "loss": 0.0085, "step": 69830 }, { "grad_norm": 0.17824941873550415, "learning_rate": 4.346015201476572e-06, "loss": 0.0148, "step": 69840 }, { "grad_norm": 0.14364761114120483, "learning_rate": 4.337590929364855e-06, "loss": 0.0099, "step": 69850 }, { "grad_norm": 0.19836774468421936, "learning_rate": 4.329174459709973e-06, "loss": 0.0096, "step": 69860 }, { "grad_norm": 0.19374719262123108, "learning_rate": 4.320765793950071e-06, "loss": 0.0155, "step": 69870 }, { "grad_norm": 0.35126885771751404, "learning_rate": 4.312364933521962e-06, "loss": 0.0104, "step": 69880 }, { "grad_norm": 0.14103247225284576, "learning_rate": 4.303971879861129e-06, "loss": 0.0075, "step": 69890 }, { "grad_norm": 0.3654136657714844, "learning_rate": 4.295586634401716e-06, "loss": 0.0115, "step": 69900 }, { "grad_norm": 0.13427507877349854, "learning_rate": 4.287209198576536e-06, "loss": 0.0071, "step": 69910 }, { "grad_norm": 0.1448027789592743, "learning_rate": 4.278839573817045e-06, "loss": 0.0116, "step": 69920 }, { "grad_norm": 0.3074392080307007, "learning_rate": 4.270477761553399e-06, "loss": 0.0122, "step": 69930 }, { "grad_norm": 0.19457951188087463, "learning_rate": 4.262123763214393e-06, "loss": 0.0096, "step": 69940 }, { "grad_norm": 0.14252100884914398, "learning_rate": 4.253777580227508e-06, "loss": 0.0068, "step": 69950 }, { "grad_norm": 0.11461282521486282, "learning_rate": 4.2454392140188755e-06, "loss": 0.0099, "step": 69960 }, { "grad_norm": 0.46552401781082153, "learning_rate": 4.2371086660132785e-06, "loss": 0.0118, "step": 69970 }, { "grad_norm": 0.12279811501502991, "learning_rate": 4.228785937634205e-06, "loss": 0.0067, "step": 69980 }, { "grad_norm": 0.14569087326526642, "learning_rate": 4.220471030303758e-06, "loss": 0.0064, "step": 69990 }, { "grad_norm": 0.12837612628936768, "learning_rate": 4.212163945442754e-06, "loss": 0.0055, "step": 70000 }, { "grad_norm": 0.2627243101596832, "learning_rate": 4.203864684470621e-06, "loss": 0.0064, "step": 70010 }, { "grad_norm": 0.11148163676261902, "learning_rate": 4.19557324880549e-06, "loss": 0.0088, "step": 70020 }, { "grad_norm": 0.16045676171779633, "learning_rate": 4.187289639864145e-06, "loss": 0.0101, "step": 70030 }, { "grad_norm": 0.18404076993465424, "learning_rate": 4.1790138590619974e-06, "loss": 0.0093, "step": 70040 }, { "grad_norm": 0.2607969045639038, "learning_rate": 4.170745907813195e-06, "loss": 0.01, "step": 70050 }, { "grad_norm": 0.18027378618717194, "learning_rate": 4.162485787530479e-06, "loss": 0.0117, "step": 70060 }, { "grad_norm": 0.1531076431274414, "learning_rate": 4.154233499625282e-06, "loss": 0.0078, "step": 70070 }, { "grad_norm": 0.12804798781871796, "learning_rate": 4.145989045507692e-06, "loss": 0.0082, "step": 70080 }, { "grad_norm": 0.07653429359197617, "learning_rate": 4.1377524265864666e-06, "loss": 0.0114, "step": 70090 }, { "grad_norm": 0.1177222728729248, "learning_rate": 4.1295236442690175e-06, "loss": 0.0093, "step": 70100 }, { "grad_norm": 0.39064615964889526, "learning_rate": 4.121302699961421e-06, "loss": 0.0082, "step": 70110 }, { "grad_norm": 0.2581164538860321, "learning_rate": 4.113089595068403e-06, "loss": 0.0129, "step": 70120 }, { "grad_norm": 0.20595216751098633, "learning_rate": 4.104884330993364e-06, "loss": 0.0107, "step": 70130 }, { "grad_norm": 0.1845446228981018, "learning_rate": 4.0966869091383585e-06, "loss": 0.0118, "step": 70140 }, { "grad_norm": 0.17517998814582825, "learning_rate": 4.088497330904101e-06, "loss": 0.0114, "step": 70150 }, { "grad_norm": 0.17625807225704193, "learning_rate": 4.080315597689976e-06, "loss": 0.011, "step": 70160 }, { "grad_norm": 0.11684300005435944, "learning_rate": 4.072141710893995e-06, "loss": 0.0122, "step": 70170 }, { "grad_norm": 0.0762934535741806, "learning_rate": 4.063975671912879e-06, "loss": 0.0143, "step": 70180 }, { "grad_norm": 0.19207604229450226, "learning_rate": 4.055817482141949e-06, "loss": 0.0082, "step": 70190 }, { "grad_norm": 0.1597544103860855, "learning_rate": 4.047667142975259e-06, "loss": 0.0081, "step": 70200 }, { "grad_norm": 0.27200570702552795, "learning_rate": 4.039524655805443e-06, "loss": 0.0139, "step": 70210 }, { "grad_norm": 0.14253215491771698, "learning_rate": 4.03139002202384e-06, "loss": 0.0072, "step": 70220 }, { "grad_norm": 0.18899160623550415, "learning_rate": 4.023263243020447e-06, "loss": 0.0104, "step": 70230 }, { "grad_norm": 0.08544453978538513, "learning_rate": 4.015144320183884e-06, "loss": 0.011, "step": 70240 }, { "grad_norm": 0.23676271736621857, "learning_rate": 4.007033254901482e-06, "loss": 0.0089, "step": 70250 }, { "grad_norm": 0.1528271585702896, "learning_rate": 3.9989300485591795e-06, "loss": 0.0106, "step": 70260 }, { "grad_norm": 0.17445623874664307, "learning_rate": 3.990834702541601e-06, "loss": 0.0104, "step": 70270 }, { "grad_norm": 0.1834058165550232, "learning_rate": 3.982747218232019e-06, "loss": 0.0085, "step": 70280 }, { "grad_norm": 0.8969126343727112, "learning_rate": 3.97466759701236e-06, "loss": 0.0123, "step": 70290 }, { "grad_norm": 0.20998810231685638, "learning_rate": 3.966595840263226e-06, "loss": 0.0093, "step": 70300 }, { "grad_norm": 0.2562495470046997, "learning_rate": 3.958531949363836e-06, "loss": 0.0082, "step": 70310 }, { "grad_norm": 0.18337978422641754, "learning_rate": 3.950475925692098e-06, "loss": 0.0106, "step": 70320 }, { "grad_norm": 0.17305997014045715, "learning_rate": 3.9424277706245685e-06, "loss": 0.0084, "step": 70330 }, { "grad_norm": 0.12587356567382812, "learning_rate": 3.934387485536451e-06, "loss": 0.0114, "step": 70340 }, { "grad_norm": 0.12296178936958313, "learning_rate": 3.926355071801619e-06, "loss": 0.0086, "step": 70350 }, { "grad_norm": 0.12107092142105103, "learning_rate": 3.9183305307925965e-06, "loss": 0.0081, "step": 70360 }, { "grad_norm": 0.1989259570837021, "learning_rate": 3.910313863880533e-06, "loss": 0.0067, "step": 70370 }, { "grad_norm": 0.1620519608259201, "learning_rate": 3.902305072435292e-06, "loss": 0.009, "step": 70380 }, { "grad_norm": 0.16587501764297485, "learning_rate": 3.894304157825329e-06, "loss": 0.0078, "step": 70390 }, { "grad_norm": 0.14362308382987976, "learning_rate": 3.886311121417791e-06, "loss": 0.0082, "step": 70400 }, { "grad_norm": 0.11515384912490845, "learning_rate": 3.878325964578472e-06, "loss": 0.0072, "step": 70410 }, { "grad_norm": 0.1476009041070938, "learning_rate": 3.870348688671815e-06, "loss": 0.006, "step": 70420 }, { "grad_norm": 0.25420331954956055, "learning_rate": 3.862379295060931e-06, "loss": 0.0086, "step": 70430 }, { "grad_norm": 0.09693294763565063, "learning_rate": 3.854417785107534e-06, "loss": 0.0168, "step": 70440 }, { "grad_norm": 0.16052867472171783, "learning_rate": 3.8464641601720755e-06, "loss": 0.0109, "step": 70450 }, { "grad_norm": 0.27121055126190186, "learning_rate": 3.838518421613579e-06, "loss": 0.0135, "step": 70460 }, { "grad_norm": 0.24143598973751068, "learning_rate": 3.830580570789766e-06, "loss": 0.0105, "step": 70470 }, { "grad_norm": 0.12848134338855743, "learning_rate": 3.822650609057005e-06, "loss": 0.0085, "step": 70480 }, { "grad_norm": 0.16746622323989868, "learning_rate": 3.814728537770285e-06, "loss": 0.0105, "step": 70490 }, { "grad_norm": 0.1798003911972046, "learning_rate": 3.806814358283306e-06, "loss": 0.0108, "step": 70500 }, { "grad_norm": 0.6143376231193542, "learning_rate": 3.7989080719483596e-06, "loss": 0.0074, "step": 70510 }, { "grad_norm": 0.08512531220912933, "learning_rate": 3.7910096801164143e-06, "loss": 0.0104, "step": 70520 }, { "grad_norm": 0.25537344813346863, "learning_rate": 3.7831191841371016e-06, "loss": 0.0093, "step": 70530 }, { "grad_norm": 0.17180100083351135, "learning_rate": 3.775236585358688e-06, "loss": 0.0112, "step": 70540 }, { "grad_norm": 0.11492366343736649, "learning_rate": 3.7673618851280843e-06, "loss": 0.0114, "step": 70550 }, { "grad_norm": 0.2556453049182892, "learning_rate": 3.759495084790887e-06, "loss": 0.0113, "step": 70560 }, { "grad_norm": 0.1715860366821289, "learning_rate": 3.751636185691282e-06, "loss": 0.0086, "step": 70570 }, { "grad_norm": 0.2522014379501343, "learning_rate": 3.7437851891721607e-06, "loss": 0.0099, "step": 70580 }, { "grad_norm": 0.2588691711425781, "learning_rate": 3.7359420965750404e-06, "loss": 0.0127, "step": 70590 }, { "grad_norm": 0.11081041395664215, "learning_rate": 3.7281069092400922e-06, "loss": 0.0094, "step": 70600 }, { "grad_norm": 0.14993664622306824, "learning_rate": 3.7202796285061348e-06, "loss": 0.0077, "step": 70610 }, { "grad_norm": 0.14669744670391083, "learning_rate": 3.712460255710637e-06, "loss": 0.0125, "step": 70620 }, { "grad_norm": 0.20663149654865265, "learning_rate": 3.704648792189719e-06, "loss": 0.0088, "step": 70630 }, { "grad_norm": 0.1964086890220642, "learning_rate": 3.696845239278124e-06, "loss": 0.0096, "step": 70640 }, { "grad_norm": 0.2988656163215637, "learning_rate": 3.689049598309302e-06, "loss": 0.0097, "step": 70650 }, { "grad_norm": 0.15628597140312195, "learning_rate": 3.681261870615288e-06, "loss": 0.0141, "step": 70660 }, { "grad_norm": 0.22041136026382446, "learning_rate": 3.6734820575268004e-06, "loss": 0.019, "step": 70670 }, { "grad_norm": 0.09288066625595093, "learning_rate": 3.665710160373209e-06, "loss": 0.0133, "step": 70680 }, { "grad_norm": 0.10356567800045013, "learning_rate": 3.65794618048248e-06, "loss": 0.0103, "step": 70690 }, { "grad_norm": 0.24917389452457428, "learning_rate": 3.6501901191813125e-06, "loss": 0.0071, "step": 70700 }, { "grad_norm": 0.11073322594165802, "learning_rate": 3.642441977794975e-06, "loss": 0.0064, "step": 70710 }, { "grad_norm": 0.28913530707359314, "learning_rate": 3.634701757647424e-06, "loss": 0.0075, "step": 70720 }, { "grad_norm": 0.1371561586856842, "learning_rate": 3.6269694600612468e-06, "loss": 0.0067, "step": 70730 }, { "grad_norm": 0.47384029626846313, "learning_rate": 3.619245086357681e-06, "loss": 0.0127, "step": 70740 }, { "grad_norm": 0.26521503925323486, "learning_rate": 3.611528637856615e-06, "loss": 0.0081, "step": 70750 }, { "grad_norm": 0.1782645434141159, "learning_rate": 3.6038201158765884e-06, "loss": 0.0088, "step": 70760 }, { "grad_norm": 0.45023608207702637, "learning_rate": 3.5961195217347534e-06, "loss": 0.0193, "step": 70770 }, { "grad_norm": 0.11623130738735199, "learning_rate": 3.588426856746946e-06, "loss": 0.0096, "step": 70780 }, { "grad_norm": 0.09392156451940536, "learning_rate": 3.5807421222276316e-06, "loss": 0.0159, "step": 70790 }, { "grad_norm": 0.33143293857574463, "learning_rate": 3.5730653194899156e-06, "loss": 0.0126, "step": 70800 }, { "grad_norm": 0.1512899249792099, "learning_rate": 3.565396449845554e-06, "loss": 0.0086, "step": 70810 }, { "grad_norm": 0.362153559923172, "learning_rate": 3.557735514604954e-06, "loss": 0.0105, "step": 70820 }, { "grad_norm": 0.16756118834018707, "learning_rate": 3.5500825150771633e-06, "loss": 0.0114, "step": 70830 }, { "grad_norm": 0.2025432139635086, "learning_rate": 3.5424374525698466e-06, "loss": 0.0096, "step": 70840 }, { "grad_norm": 0.18670564889907837, "learning_rate": 3.5348003283893704e-06, "loss": 0.0076, "step": 70850 }, { "grad_norm": 0.1100657507777214, "learning_rate": 3.527171143840685e-06, "loss": 0.0109, "step": 70860 }, { "grad_norm": 0.15346753597259521, "learning_rate": 3.519549900227409e-06, "loss": 0.0122, "step": 70870 }, { "grad_norm": 0.24941948056221008, "learning_rate": 3.511936598851828e-06, "loss": 0.0121, "step": 70880 }, { "grad_norm": 0.14455777406692505, "learning_rate": 3.504331241014813e-06, "loss": 0.0103, "step": 70890 }, { "grad_norm": 0.2977045178413391, "learning_rate": 3.4967338280159414e-06, "loss": 0.0091, "step": 70900 }, { "grad_norm": 0.08850227296352386, "learning_rate": 3.4891443611533846e-06, "loss": 0.0119, "step": 70910 }, { "grad_norm": 0.15457387268543243, "learning_rate": 3.481562841723984e-06, "loss": 0.0133, "step": 70920 }, { "grad_norm": 0.1775815635919571, "learning_rate": 3.473989271023215e-06, "loss": 0.0077, "step": 70930 }, { "grad_norm": 0.17885401844978333, "learning_rate": 3.4664236503451854e-06, "loss": 0.0086, "step": 70940 }, { "grad_norm": 0.14452576637268066, "learning_rate": 3.4588659809826674e-06, "loss": 0.0076, "step": 70950 }, { "grad_norm": 0.174913689494133, "learning_rate": 3.451316264227039e-06, "loss": 0.0109, "step": 70960 }, { "grad_norm": 0.19445696473121643, "learning_rate": 3.4437745013683574e-06, "loss": 0.009, "step": 70970 }, { "grad_norm": 0.17896614968776703, "learning_rate": 3.4362406936952916e-06, "loss": 0.0092, "step": 70980 }, { "grad_norm": 0.1863771378993988, "learning_rate": 3.4287148424951786e-06, "loss": 0.0094, "step": 70990 }, { "grad_norm": 0.22975821793079376, "learning_rate": 3.4211969490539617e-06, "loss": 0.0098, "step": 71000 }, { "grad_norm": 0.3499618470668793, "learning_rate": 3.4136870146562584e-06, "loss": 0.0121, "step": 71010 }, { "grad_norm": 0.13375495374202728, "learning_rate": 3.406185040585308e-06, "loss": 0.0095, "step": 71020 }, { "grad_norm": 0.17372755706310272, "learning_rate": 3.3986910281229966e-06, "loss": 0.0134, "step": 71030 }, { "grad_norm": 0.3328978717327118, "learning_rate": 3.391204978549828e-06, "loss": 0.0112, "step": 71040 }, { "grad_norm": 0.19070041179656982, "learning_rate": 3.3837268931449785e-06, "loss": 0.0122, "step": 71050 }, { "grad_norm": 0.19354307651519775, "learning_rate": 3.376256773186248e-06, "loss": 0.0115, "step": 71060 }, { "grad_norm": 0.22022183239459991, "learning_rate": 3.3687946199500664e-06, "loss": 0.0146, "step": 71070 }, { "grad_norm": 0.14324671030044556, "learning_rate": 3.3613404347115295e-06, "loss": 0.0087, "step": 71080 }, { "grad_norm": 0.12982206046581268, "learning_rate": 3.35389421874433e-06, "loss": 0.0078, "step": 71090 }, { "grad_norm": 0.11556926369667053, "learning_rate": 3.3464559733208446e-06, "loss": 0.0096, "step": 71100 }, { "grad_norm": 0.4260496199131012, "learning_rate": 3.3390256997120505e-06, "loss": 0.0114, "step": 71110 }, { "grad_norm": 0.08611436188220978, "learning_rate": 3.331603399187583e-06, "loss": 0.0107, "step": 71120 }, { "grad_norm": 0.08856958895921707, "learning_rate": 3.324189073015721e-06, "loss": 0.0093, "step": 71130 }, { "grad_norm": 0.13662298023700714, "learning_rate": 3.3167827224633465e-06, "loss": 0.011, "step": 71140 }, { "grad_norm": 0.2083790898323059, "learning_rate": 3.309384348796024e-06, "loss": 0.012, "step": 71150 }, { "grad_norm": 0.14641974866390228, "learning_rate": 3.301993953277921e-06, "loss": 0.0085, "step": 71160 }, { "grad_norm": 0.1509505659341812, "learning_rate": 3.29461153717186e-06, "loss": 0.0102, "step": 71170 }, { "grad_norm": 0.189808651804924, "learning_rate": 3.287237101739293e-06, "loss": 0.0122, "step": 71180 }, { "grad_norm": 0.22334691882133484, "learning_rate": 3.2798706482403075e-06, "loss": 0.0164, "step": 71190 }, { "grad_norm": 0.2786795496940613, "learning_rate": 3.2725121779336285e-06, "loss": 0.0131, "step": 71200 }, { "grad_norm": 0.30434563755989075, "learning_rate": 3.265161692076618e-06, "loss": 0.0132, "step": 71210 }, { "grad_norm": 0.06823267787694931, "learning_rate": 3.257819191925282e-06, "loss": 0.0062, "step": 71220 }, { "grad_norm": 0.1758534461259842, "learning_rate": 3.2504846787342392e-06, "loss": 0.0108, "step": 71230 }, { "grad_norm": 0.2406773716211319, "learning_rate": 3.243158153756759e-06, "loss": 0.0083, "step": 71240 }, { "grad_norm": 0.09821968525648117, "learning_rate": 3.2358396182447514e-06, "loss": 0.0086, "step": 71250 }, { "grad_norm": 0.11224880814552307, "learning_rate": 3.2285290734487496e-06, "loss": 0.0085, "step": 71260 }, { "grad_norm": 0.09553036093711853, "learning_rate": 3.221226520617926e-06, "loss": 0.0134, "step": 71270 }, { "grad_norm": 0.2952558696269989, "learning_rate": 3.213931961000094e-06, "loss": 0.0124, "step": 71280 }, { "grad_norm": 0.18407663702964783, "learning_rate": 3.2066453958416733e-06, "loss": 0.0076, "step": 71290 }, { "grad_norm": 0.16312624514102936, "learning_rate": 3.199366826387773e-06, "loss": 0.0093, "step": 71300 }, { "grad_norm": 0.544469952583313, "learning_rate": 3.192096253882071e-06, "loss": 0.0099, "step": 71310 }, { "grad_norm": 0.2486603856086731, "learning_rate": 3.1848336795669177e-06, "loss": 0.0102, "step": 71320 }, { "grad_norm": 0.12390357255935669, "learning_rate": 3.1775791046833035e-06, "loss": 0.0115, "step": 71330 }, { "grad_norm": 0.13671772181987762, "learning_rate": 3.170332530470804e-06, "loss": 0.0098, "step": 71340 }, { "grad_norm": 0.1692969799041748, "learning_rate": 3.1630939581677012e-06, "loss": 0.011, "step": 71350 }, { "grad_norm": 0.10634710639715195, "learning_rate": 3.155863389010838e-06, "loss": 0.0121, "step": 71360 }, { "grad_norm": 0.16564929485321045, "learning_rate": 3.1486408242357323e-06, "loss": 0.0146, "step": 71370 }, { "grad_norm": 0.3243042826652527, "learning_rate": 3.1414262650765248e-06, "loss": 0.0233, "step": 71380 }, { "grad_norm": 0.25592240691185, "learning_rate": 3.134219712765979e-06, "loss": 0.0114, "step": 71390 }, { "grad_norm": 0.14342659711837769, "learning_rate": 3.1270211685355044e-06, "loss": 0.0095, "step": 71400 }, { "grad_norm": 0.18146033585071564, "learning_rate": 3.1198306336151338e-06, "loss": 0.0094, "step": 71410 }, { "grad_norm": 0.15868748724460602, "learning_rate": 3.11264810923354e-06, "loss": 0.0091, "step": 71420 }, { "grad_norm": 0.06076240539550781, "learning_rate": 3.105473596618008e-06, "loss": 0.0059, "step": 71430 }, { "grad_norm": 0.22132690250873566, "learning_rate": 3.0983070969944683e-06, "loss": 0.0091, "step": 71440 }, { "grad_norm": 0.11008361726999283, "learning_rate": 3.09114861158748e-06, "loss": 0.0111, "step": 71450 }, { "grad_norm": 0.11517976969480515, "learning_rate": 3.0839981416202314e-06, "loss": 0.0116, "step": 71460 }, { "grad_norm": 0.11957673728466034, "learning_rate": 3.0768556883145507e-06, "loss": 0.0084, "step": 71470 }, { "grad_norm": 0.1276118904352188, "learning_rate": 3.0697212528908834e-06, "loss": 0.0072, "step": 71480 }, { "grad_norm": 0.191132590174675, "learning_rate": 3.0625948365683e-06, "loss": 0.0108, "step": 71490 }, { "grad_norm": 0.14616599678993225, "learning_rate": 3.055476440564531e-06, "loss": 0.0118, "step": 71500 }, { "grad_norm": 0.12218176573514938, "learning_rate": 3.0483660660958924e-06, "loss": 0.01, "step": 71510 }, { "grad_norm": 0.4079191982746124, "learning_rate": 3.0412637143773624e-06, "loss": 0.0132, "step": 71520 }, { "grad_norm": 0.21210706233978271, "learning_rate": 3.034169386622554e-06, "loss": 0.0079, "step": 71530 }, { "grad_norm": 0.18241560459136963, "learning_rate": 3.027083084043658e-06, "loss": 0.0091, "step": 71540 }, { "grad_norm": 0.23521000146865845, "learning_rate": 3.0200048078515676e-06, "loss": 0.0096, "step": 71550 }, { "grad_norm": 0.14320261776447296, "learning_rate": 3.012934559255742e-06, "loss": 0.0072, "step": 71560 }, { "grad_norm": 0.2042425274848938, "learning_rate": 3.0058723394642994e-06, "loss": 0.0091, "step": 71570 }, { "grad_norm": 0.1960412710905075, "learning_rate": 2.998818149683985e-06, "loss": 0.0163, "step": 71580 }, { "grad_norm": 0.20206445455551147, "learning_rate": 2.9917719911201627e-06, "loss": 0.0091, "step": 71590 }, { "grad_norm": 0.16836325824260712, "learning_rate": 2.9847338649768352e-06, "loss": 0.0101, "step": 71600 }, { "grad_norm": 0.36205053329467773, "learning_rate": 2.977703772456608e-06, "loss": 0.0114, "step": 71610 }, { "grad_norm": 0.22542722523212433, "learning_rate": 2.9706817147607535e-06, "loss": 0.0102, "step": 71620 }, { "grad_norm": 0.1066044420003891, "learning_rate": 2.9636676930891393e-06, "loss": 0.0073, "step": 71630 }, { "grad_norm": 0.2218819111585617, "learning_rate": 2.9566617086402625e-06, "loss": 0.0094, "step": 71640 }, { "grad_norm": 0.1663617491722107, "learning_rate": 2.9496637626112655e-06, "loss": 0.0077, "step": 71650 }, { "grad_norm": 0.1663067638874054, "learning_rate": 2.9426738561979027e-06, "loss": 0.0092, "step": 71660 }, { "grad_norm": 0.35453277826309204, "learning_rate": 2.9356919905945524e-06, "loss": 0.0222, "step": 71670 }, { "grad_norm": 0.14143340289592743, "learning_rate": 2.928718166994243e-06, "loss": 0.0074, "step": 71680 }, { "grad_norm": 0.2643415629863739, "learning_rate": 2.9217523865885833e-06, "loss": 0.0104, "step": 71690 }, { "grad_norm": 0.15313731133937836, "learning_rate": 2.9147946505678537e-06, "loss": 0.0095, "step": 71700 }, { "grad_norm": 0.14949460327625275, "learning_rate": 2.9078449601209313e-06, "loss": 0.0094, "step": 71710 }, { "grad_norm": 0.3093187212944031, "learning_rate": 2.900903316435333e-06, "loss": 0.0126, "step": 71720 }, { "grad_norm": 0.20394159853458405, "learning_rate": 2.8939697206971983e-06, "loss": 0.0098, "step": 71730 }, { "grad_norm": 0.22739867866039276, "learning_rate": 2.8870441740912746e-06, "loss": 0.0124, "step": 71740 }, { "grad_norm": 0.12558019161224365, "learning_rate": 2.8801266778009762e-06, "loss": 0.0102, "step": 71750 }, { "grad_norm": 0.11270487308502197, "learning_rate": 2.8732172330082798e-06, "loss": 0.0107, "step": 71760 }, { "grad_norm": 0.08274172246456146, "learning_rate": 2.8663158408938517e-06, "loss": 0.0101, "step": 71770 }, { "grad_norm": 0.1292734146118164, "learning_rate": 2.859422502636938e-06, "loss": 0.0068, "step": 71780 }, { "grad_norm": 0.19401472806930542, "learning_rate": 2.8525372194154076e-06, "loss": 0.0085, "step": 71790 }, { "grad_norm": 0.20799429714679718, "learning_rate": 2.8456599924057913e-06, "loss": 0.0103, "step": 71800 }, { "grad_norm": 0.1046215370297432, "learning_rate": 2.8387908227831995e-06, "loss": 0.0098, "step": 71810 }, { "grad_norm": 0.08001206815242767, "learning_rate": 2.831929711721404e-06, "loss": 0.0091, "step": 71820 }, { "grad_norm": 0.17777502536773682, "learning_rate": 2.825076660392767e-06, "loss": 0.0093, "step": 71830 }, { "grad_norm": 0.1578776240348816, "learning_rate": 2.8182316699682908e-06, "loss": 0.015, "step": 71840 }, { "grad_norm": 0.09903103858232498, "learning_rate": 2.811394741617601e-06, "loss": 0.0087, "step": 71850 }, { "grad_norm": 0.43275508284568787, "learning_rate": 2.8045658765089356e-06, "loss": 0.0155, "step": 71860 }, { "grad_norm": 0.19774620234966278, "learning_rate": 2.7977450758091605e-06, "loss": 0.0145, "step": 71870 }, { "grad_norm": 0.18536853790283203, "learning_rate": 2.790932340683783e-06, "loss": 0.0131, "step": 71880 }, { "grad_norm": 0.22346659004688263, "learning_rate": 2.7841276722968823e-06, "loss": 0.0125, "step": 71890 }, { "grad_norm": 0.24393537640571594, "learning_rate": 2.7773310718112067e-06, "loss": 0.0082, "step": 71900 }, { "grad_norm": 0.18192033469676971, "learning_rate": 2.7705425403881102e-06, "loss": 0.0113, "step": 71910 }, { "grad_norm": 0.3899857699871063, "learning_rate": 2.763762079187565e-06, "loss": 0.0102, "step": 71920 }, { "grad_norm": 0.10882122814655304, "learning_rate": 2.756989689368178e-06, "loss": 0.0082, "step": 71930 }, { "grad_norm": 0.15763379633426666, "learning_rate": 2.750225372087134e-06, "loss": 0.0107, "step": 71940 }, { "grad_norm": 0.255869597196579, "learning_rate": 2.7434691285003033e-06, "loss": 0.0069, "step": 71950 }, { "grad_norm": 0.20702867209911346, "learning_rate": 2.7367209597621178e-06, "loss": 0.0165, "step": 71960 }, { "grad_norm": 0.1826402097940445, "learning_rate": 2.729980867025683e-06, "loss": 0.0095, "step": 71970 }, { "grad_norm": 0.17845416069030762, "learning_rate": 2.7232488514426724e-06, "loss": 0.0104, "step": 71980 }, { "grad_norm": 0.2570517361164093, "learning_rate": 2.7165249141634097e-06, "loss": 0.0083, "step": 71990 }, { "grad_norm": 0.18792638182640076, "learning_rate": 2.709809056336837e-06, "loss": 0.0102, "step": 72000 }, { "grad_norm": 0.16350214183330536, "learning_rate": 2.703101279110498e-06, "loss": 0.008, "step": 72010 }, { "grad_norm": 0.1601325124502182, "learning_rate": 2.696401583630587e-06, "loss": 0.0096, "step": 72020 }, { "grad_norm": 0.1384793519973755, "learning_rate": 2.689709971041887e-06, "loss": 0.0064, "step": 72030 }, { "grad_norm": 0.18384571373462677, "learning_rate": 2.6830264424878126e-06, "loss": 0.0081, "step": 72040 }, { "grad_norm": 0.23483091592788696, "learning_rate": 2.676350999110394e-06, "loss": 0.0155, "step": 72050 }, { "grad_norm": 0.18489712476730347, "learning_rate": 2.6696836420502856e-06, "loss": 0.0154, "step": 72060 }, { "grad_norm": 0.12114666402339935, "learning_rate": 2.663024372446765e-06, "loss": 0.0075, "step": 72070 }, { "grad_norm": 0.18231379985809326, "learning_rate": 2.6563731914377055e-06, "loss": 0.0072, "step": 72080 }, { "grad_norm": 0.12492285668849945, "learning_rate": 2.6497301001596087e-06, "loss": 0.011, "step": 72090 }, { "grad_norm": 0.2740975022315979, "learning_rate": 2.643095099747611e-06, "loss": 0.0102, "step": 72100 }, { "grad_norm": 0.13132400810718536, "learning_rate": 2.636468191335445e-06, "loss": 0.0094, "step": 72110 }, { "grad_norm": 0.1752447932958603, "learning_rate": 2.629849376055471e-06, "loss": 0.013, "step": 72120 }, { "grad_norm": 0.17862731218338013, "learning_rate": 2.6232386550386678e-06, "loss": 0.0093, "step": 72130 }, { "grad_norm": 0.159030020236969, "learning_rate": 2.6166360294146097e-06, "loss": 0.01, "step": 72140 }, { "grad_norm": 0.30573615431785583, "learning_rate": 2.6100415003115275e-06, "loss": 0.0154, "step": 72150 }, { "grad_norm": 0.206992968916893, "learning_rate": 2.603455068856225e-06, "loss": 0.0113, "step": 72160 }, { "grad_norm": 0.08333294838666916, "learning_rate": 2.5968767361741584e-06, "loss": 0.0121, "step": 72170 }, { "grad_norm": 0.19496142864227295, "learning_rate": 2.590306503389378e-06, "loss": 0.0127, "step": 72180 }, { "grad_norm": 0.1172037348151207, "learning_rate": 2.583744371624558e-06, "loss": 0.0144, "step": 72190 }, { "grad_norm": 0.14375019073486328, "learning_rate": 2.57719034200099e-06, "loss": 0.0113, "step": 72200 }, { "grad_norm": 0.20829634368419647, "learning_rate": 2.570644415638568e-06, "loss": 0.0092, "step": 72210 }, { "grad_norm": 0.25088098645210266, "learning_rate": 2.5641065936558295e-06, "loss": 0.0119, "step": 72220 }, { "grad_norm": 0.15309184789657593, "learning_rate": 2.557576877169898e-06, "loss": 0.0088, "step": 72230 }, { "grad_norm": 0.1328693926334381, "learning_rate": 2.5510552672965205e-06, "loss": 0.0077, "step": 72240 }, { "grad_norm": 0.15413732826709747, "learning_rate": 2.544541765150077e-06, "loss": 0.0088, "step": 72250 }, { "grad_norm": 0.2542642652988434, "learning_rate": 2.5380363718435165e-06, "loss": 0.0114, "step": 72260 }, { "grad_norm": 0.1041647344827652, "learning_rate": 2.5315390884884714e-06, "loss": 0.008, "step": 72270 }, { "grad_norm": 0.16649672389030457, "learning_rate": 2.5250499161951214e-06, "loss": 0.0129, "step": 72280 }, { "grad_norm": 0.15508319437503815, "learning_rate": 2.518568856072301e-06, "loss": 0.0073, "step": 72290 }, { "grad_norm": 0.334071546792984, "learning_rate": 2.512095909227441e-06, "loss": 0.01, "step": 72300 }, { "grad_norm": 0.17328041791915894, "learning_rate": 2.50563107676659e-06, "loss": 0.0096, "step": 72310 }, { "grad_norm": 0.1260654628276825, "learning_rate": 2.4991743597944152e-06, "loss": 0.009, "step": 72320 }, { "grad_norm": 0.21542593836784363, "learning_rate": 2.492725759414205e-06, "loss": 0.0098, "step": 72330 }, { "grad_norm": 0.17937743663787842, "learning_rate": 2.4862852767278234e-06, "loss": 0.0101, "step": 72340 }, { "grad_norm": 0.07612771540880203, "learning_rate": 2.479852912835784e-06, "loss": 0.0085, "step": 72350 }, { "grad_norm": 0.20061719417572021, "learning_rate": 2.4734286688372075e-06, "loss": 0.0098, "step": 72360 }, { "grad_norm": 0.9380538463592529, "learning_rate": 2.467012545829811e-06, "loss": 0.0138, "step": 72370 }, { "grad_norm": 0.11284074187278748, "learning_rate": 2.460604544909945e-06, "loss": 0.0067, "step": 72380 }, { "grad_norm": 0.23049592971801758, "learning_rate": 2.454204667172555e-06, "loss": 0.01, "step": 72390 }, { "grad_norm": 0.13632610440254211, "learning_rate": 2.447812913711217e-06, "loss": 0.0086, "step": 72400 }, { "grad_norm": 0.23102577030658722, "learning_rate": 2.44142928561808e-06, "loss": 0.011, "step": 72410 }, { "grad_norm": 0.14640380442142487, "learning_rate": 2.435053783983965e-06, "loss": 0.0097, "step": 72420 }, { "grad_norm": 0.12178178876638412, "learning_rate": 2.4286864098982453e-06, "loss": 0.0101, "step": 72430 }, { "grad_norm": 0.10458970069885254, "learning_rate": 2.4223271644489397e-06, "loss": 0.0132, "step": 72440 }, { "grad_norm": 0.14753805100917816, "learning_rate": 2.415976048722679e-06, "loss": 0.008, "step": 72450 }, { "grad_norm": 0.18960055708885193, "learning_rate": 2.4096330638046673e-06, "loss": 0.0076, "step": 72460 }, { "grad_norm": 0.16599369049072266, "learning_rate": 2.4032982107787816e-06, "loss": 0.0079, "step": 72470 }, { "grad_norm": 0.30093103647232056, "learning_rate": 2.396971490727451e-06, "loss": 0.0121, "step": 72480 }, { "grad_norm": 0.12631820142269135, "learning_rate": 2.3906529047317493e-06, "loss": 0.0066, "step": 72490 }, { "grad_norm": 0.1908789575099945, "learning_rate": 2.3843424538713465e-06, "loss": 0.0102, "step": 72500 }, { "grad_norm": 0.4067094027996063, "learning_rate": 2.3780401392245298e-06, "loss": 0.0103, "step": 72510 }, { "grad_norm": 0.11362456530332565, "learning_rate": 2.3717459618681883e-06, "loss": 0.0086, "step": 72520 }, { "grad_norm": 0.12615680694580078, "learning_rate": 2.3654599228778274e-06, "loss": 0.0093, "step": 72530 }, { "grad_norm": 0.21405543386936188, "learning_rate": 2.3591820233275607e-06, "loss": 0.0107, "step": 72540 }, { "grad_norm": 0.15115606784820557, "learning_rate": 2.3529122642901024e-06, "loss": 0.0123, "step": 72550 }, { "grad_norm": 0.1393333524465561, "learning_rate": 2.3466506468367845e-06, "loss": 0.0181, "step": 72560 }, { "grad_norm": 0.6361685395240784, "learning_rate": 2.3403971720375506e-06, "loss": 0.0089, "step": 72570 }, { "grad_norm": 0.18699024617671967, "learning_rate": 2.334151840960952e-06, "loss": 0.0067, "step": 72580 }, { "grad_norm": 0.08532271534204483, "learning_rate": 2.3279146546741347e-06, "loss": 0.008, "step": 72590 }, { "grad_norm": 0.1644131988286972, "learning_rate": 2.32168561424288e-06, "loss": 0.0092, "step": 72600 }, { "grad_norm": 0.18601326644420624, "learning_rate": 2.3154647207315307e-06, "loss": 0.011, "step": 72610 }, { "grad_norm": 0.08166458457708359, "learning_rate": 2.309251975203103e-06, "loss": 0.0108, "step": 72620 }, { "grad_norm": 0.1651613712310791, "learning_rate": 2.303047378719159e-06, "loss": 0.0098, "step": 72630 }, { "grad_norm": 0.21313686668872833, "learning_rate": 2.2968509323399e-06, "loss": 0.0128, "step": 72640 }, { "grad_norm": 0.1786326766014099, "learning_rate": 2.290662637124147e-06, "loss": 0.0067, "step": 72650 }, { "grad_norm": 0.14425458014011383, "learning_rate": 2.2844824941292807e-06, "loss": 0.0072, "step": 72660 }, { "grad_norm": 0.2784111499786377, "learning_rate": 2.2783105044113406e-06, "loss": 0.0109, "step": 72670 }, { "grad_norm": 0.2433384507894516, "learning_rate": 2.272146669024944e-06, "loss": 0.0079, "step": 72680 }, { "grad_norm": 0.21164017915725708, "learning_rate": 2.26599098902332e-06, "loss": 0.0083, "step": 72690 }, { "grad_norm": 0.19604885578155518, "learning_rate": 2.2598434654583113e-06, "loss": 0.0094, "step": 72700 }, { "grad_norm": 0.2616209089756012, "learning_rate": 2.253704099380355e-06, "loss": 0.0162, "step": 72710 }, { "grad_norm": 0.25643280148506165, "learning_rate": 2.247572891838512e-06, "loss": 0.0097, "step": 72720 }, { "grad_norm": 0.21924220025539398, "learning_rate": 2.241449843880422e-06, "loss": 0.0116, "step": 72730 }, { "grad_norm": 0.07733826339244843, "learning_rate": 2.235334956552354e-06, "loss": 0.0085, "step": 72740 }, { "grad_norm": 0.13895361125469208, "learning_rate": 2.2292282308991775e-06, "loss": 0.0165, "step": 72750 }, { "grad_norm": 0.28454771637916565, "learning_rate": 2.223129667964363e-06, "loss": 0.0102, "step": 72760 }, { "grad_norm": 0.18681196868419647, "learning_rate": 2.2170392687899834e-06, "loss": 0.0081, "step": 72770 }, { "grad_norm": 0.08122827112674713, "learning_rate": 2.210957034416733e-06, "loss": 0.0073, "step": 72780 }, { "grad_norm": 0.24329061806201935, "learning_rate": 2.2048829658838867e-06, "loss": 0.0111, "step": 72790 }, { "grad_norm": 0.28210732340812683, "learning_rate": 2.1988170642293525e-06, "loss": 0.0064, "step": 72800 }, { "grad_norm": 0.13456614315509796, "learning_rate": 2.1927593304896075e-06, "loss": 0.0083, "step": 72810 }, { "grad_norm": 0.11115731298923492, "learning_rate": 2.1867097656997626e-06, "loss": 0.0096, "step": 72820 }, { "grad_norm": 0.19626998901367188, "learning_rate": 2.180668370893524e-06, "loss": 0.0097, "step": 72830 }, { "grad_norm": 0.19811974465847015, "learning_rate": 2.174635147103199e-06, "loss": 0.0099, "step": 72840 }, { "grad_norm": 0.18408025801181793, "learning_rate": 2.1686100953597075e-06, "loss": 0.0112, "step": 72850 }, { "grad_norm": 0.2550579905509949, "learning_rate": 2.1625932166925433e-06, "loss": 0.0073, "step": 72860 }, { "grad_norm": 0.2477063089609146, "learning_rate": 2.1565845121298556e-06, "loss": 0.0096, "step": 72870 }, { "grad_norm": 0.21890996396541595, "learning_rate": 2.150583982698351e-06, "loss": 0.0146, "step": 72880 }, { "grad_norm": 0.08092671632766724, "learning_rate": 2.144591629423359e-06, "loss": 0.0077, "step": 72890 }, { "grad_norm": 0.11732333153486252, "learning_rate": 2.138607453328817e-06, "loss": 0.0113, "step": 72900 }, { "grad_norm": 0.1374499797821045, "learning_rate": 2.132631455437234e-06, "loss": 0.0096, "step": 72910 }, { "grad_norm": 0.22027361392974854, "learning_rate": 2.1266636367697714e-06, "loss": 0.0149, "step": 72920 }, { "grad_norm": 0.20423200726509094, "learning_rate": 2.120703998346152e-06, "loss": 0.0094, "step": 72930 }, { "grad_norm": 0.20751191675662994, "learning_rate": 2.1147525411847114e-06, "loss": 0.0102, "step": 72940 }, { "grad_norm": 0.1420670747756958, "learning_rate": 2.108809266302403e-06, "loss": 0.0089, "step": 72950 }, { "grad_norm": 0.2770601212978363, "learning_rate": 2.102874174714764e-06, "loss": 0.0104, "step": 72960 }, { "grad_norm": 0.1719207912683487, "learning_rate": 2.0969472674359393e-06, "loss": 0.0084, "step": 72970 }, { "grad_norm": 0.0730421170592308, "learning_rate": 2.091028545478668e-06, "loss": 0.0104, "step": 72980 }, { "grad_norm": 0.1449117511510849, "learning_rate": 2.08511800985432e-06, "loss": 0.0097, "step": 72990 }, { "grad_norm": 0.3524344563484192, "learning_rate": 2.07921566157282e-06, "loss": 0.0166, "step": 73000 }, { "grad_norm": 0.1528049260377884, "learning_rate": 2.073321501642728e-06, "loss": 0.0113, "step": 73010 }, { "grad_norm": 0.09857439249753952, "learning_rate": 2.0674355310711937e-06, "loss": 0.0091, "step": 73020 }, { "grad_norm": 0.17076165974140167, "learning_rate": 2.0615577508639682e-06, "loss": 0.0065, "step": 73030 }, { "grad_norm": 0.06928077340126038, "learning_rate": 2.055688162025399e-06, "loss": 0.0085, "step": 73040 }, { "grad_norm": 0.092485211789608, "learning_rate": 2.0498267655584546e-06, "loss": 0.0135, "step": 73050 }, { "grad_norm": 0.17685462534427643, "learning_rate": 2.0439735624646626e-06, "loss": 0.0103, "step": 73060 }, { "grad_norm": 0.06384380161762238, "learning_rate": 2.0381285537442008e-06, "loss": 0.0071, "step": 73070 }, { "grad_norm": 0.16109561920166016, "learning_rate": 2.032291740395803e-06, "loss": 0.0109, "step": 73080 }, { "grad_norm": 0.374675452709198, "learning_rate": 2.0264631234168276e-06, "loss": 0.0143, "step": 73090 }, { "grad_norm": 0.2383914291858673, "learning_rate": 2.0206427038032326e-06, "loss": 0.0117, "step": 73100 }, { "grad_norm": 0.25948673486709595, "learning_rate": 2.0148304825495457e-06, "loss": 0.0087, "step": 73110 }, { "grad_norm": 0.21440398693084717, "learning_rate": 2.0090264606489496e-06, "loss": 0.0097, "step": 73120 }, { "grad_norm": 0.19260229170322418, "learning_rate": 2.003230639093162e-06, "loss": 0.0085, "step": 73130 }, { "grad_norm": 0.31505417823791504, "learning_rate": 1.9974430188725525e-06, "loss": 0.0121, "step": 73140 }, { "grad_norm": 0.17646940052509308, "learning_rate": 1.9916636009760513e-06, "loss": 0.0113, "step": 73150 }, { "grad_norm": 0.10813908278942108, "learning_rate": 1.985892386391208e-06, "loss": 0.0115, "step": 73160 }, { "grad_norm": 0.4562138617038727, "learning_rate": 1.980129376104173e-06, "loss": 0.0071, "step": 73170 }, { "grad_norm": 0.13753536343574524, "learning_rate": 1.9743745710996796e-06, "loss": 0.0084, "step": 73180 }, { "grad_norm": 0.32032740116119385, "learning_rate": 1.9686279723610757e-06, "loss": 0.012, "step": 73190 }, { "grad_norm": 0.1414540559053421, "learning_rate": 1.962889580870281e-06, "loss": 0.0077, "step": 73200 }, { "grad_norm": 0.16921740770339966, "learning_rate": 1.9571593976078385e-06, "loss": 0.0103, "step": 73210 }, { "grad_norm": 0.16028261184692383, "learning_rate": 1.9514374235528876e-06, "loss": 0.0107, "step": 73220 }, { "grad_norm": 0.16717267036437988, "learning_rate": 1.9457236596831408e-06, "loss": 0.0056, "step": 73230 }, { "grad_norm": 0.13435573875904083, "learning_rate": 1.940018106974939e-06, "loss": 0.0104, "step": 73240 }, { "grad_norm": 0.22147993743419647, "learning_rate": 1.9343207664032025e-06, "loss": 0.0078, "step": 73250 }, { "grad_norm": 0.09880167245864868, "learning_rate": 1.928631638941436e-06, "loss": 0.0095, "step": 73260 }, { "grad_norm": 0.1737501174211502, "learning_rate": 1.922950725561784e-06, "loss": 0.0093, "step": 73270 }, { "grad_norm": 0.26105499267578125, "learning_rate": 1.9172780272349312e-06, "loss": 0.0072, "step": 73280 }, { "grad_norm": 0.2577492594718933, "learning_rate": 1.911613544930202e-06, "loss": 0.0097, "step": 73290 }, { "grad_norm": 0.16301390528678894, "learning_rate": 1.9059572796155e-06, "loss": 0.011, "step": 73300 }, { "grad_norm": 0.1333831548690796, "learning_rate": 1.9003092322573192e-06, "loss": 0.008, "step": 73310 }, { "grad_norm": 0.18901067972183228, "learning_rate": 1.8946694038207647e-06, "loss": 0.0115, "step": 73320 }, { "grad_norm": 0.23028424382209778, "learning_rate": 1.8890377952695215e-06, "loss": 0.0062, "step": 73330 }, { "grad_norm": 0.127760648727417, "learning_rate": 1.8834144075658865e-06, "loss": 0.0078, "step": 73340 }, { "grad_norm": 0.050599947571754456, "learning_rate": 1.8777992416707357e-06, "loss": 0.0087, "step": 73350 }, { "grad_norm": 0.10680841654539108, "learning_rate": 1.8721922985435458e-06, "loss": 0.0071, "step": 73360 }, { "grad_norm": 0.23561714589595795, "learning_rate": 1.8665935791424006e-06, "loss": 0.0145, "step": 73370 }, { "grad_norm": 0.3100549578666687, "learning_rate": 1.8610030844239512e-06, "loss": 0.0091, "step": 73380 }, { "grad_norm": 0.15350070595741272, "learning_rate": 1.8554208153434838e-06, "loss": 0.0115, "step": 73390 }, { "grad_norm": 0.05802018940448761, "learning_rate": 1.8498467728548296e-06, "loss": 0.0103, "step": 73400 }, { "grad_norm": 0.11766830831766129, "learning_rate": 1.8442809579104547e-06, "loss": 0.0075, "step": 73410 }, { "grad_norm": 0.20965196192264557, "learning_rate": 1.8387233714614039e-06, "loss": 0.0084, "step": 73420 }, { "grad_norm": 0.24687890708446503, "learning_rate": 1.8331740144573174e-06, "loss": 0.0165, "step": 73430 }, { "grad_norm": 0.13364198803901672, "learning_rate": 1.8276328878464255e-06, "loss": 0.0094, "step": 73440 }, { "grad_norm": 0.33056747913360596, "learning_rate": 1.82209999257556e-06, "loss": 0.011, "step": 73450 }, { "grad_norm": 0.15644720196723938, "learning_rate": 1.8165753295901311e-06, "loss": 0.008, "step": 73460 }, { "grad_norm": 0.1091703474521637, "learning_rate": 1.8110588998341616e-06, "loss": 0.0094, "step": 73470 }, { "grad_norm": 0.16835156083106995, "learning_rate": 1.805550704250253e-06, "loss": 0.0141, "step": 73480 }, { "grad_norm": 0.24071617424488068, "learning_rate": 1.8000507437796077e-06, "loss": 0.0111, "step": 73490 }, { "grad_norm": 0.11074800789356232, "learning_rate": 1.7945590193620242e-06, "loss": 0.0099, "step": 73500 }, { "grad_norm": 0.14429253339767456, "learning_rate": 1.7890755319358742e-06, "loss": 0.0094, "step": 73510 }, { "grad_norm": 0.1796741932630539, "learning_rate": 1.7836002824381525e-06, "loss": 0.0115, "step": 73520 }, { "grad_norm": 0.10003405064344406, "learning_rate": 1.7781332718044165e-06, "loss": 0.0078, "step": 73530 }, { "grad_norm": 0.23473180830478668, "learning_rate": 1.772674500968835e-06, "loss": 0.011, "step": 73540 }, { "grad_norm": 0.22006964683532715, "learning_rate": 1.7672239708641624e-06, "loss": 0.0098, "step": 73550 }, { "grad_norm": 0.1348816454410553, "learning_rate": 1.7617816824217371e-06, "loss": 0.0081, "step": 73560 }, { "grad_norm": 0.191865473985672, "learning_rate": 1.7563476365715148e-06, "loss": 0.0146, "step": 73570 }, { "grad_norm": 0.16282212734222412, "learning_rate": 1.750921834242003e-06, "loss": 0.0115, "step": 73580 }, { "grad_norm": 0.3042212724685669, "learning_rate": 1.7455042763603436e-06, "loss": 0.0124, "step": 73590 }, { "grad_norm": 0.28670012950897217, "learning_rate": 1.7400949638522345e-06, "loss": 0.0075, "step": 73600 }, { "grad_norm": 0.09198998659849167, "learning_rate": 1.734693897641987e-06, "loss": 0.0186, "step": 73610 }, { "grad_norm": 0.36254382133483887, "learning_rate": 1.7293010786524955e-06, "loss": 0.0089, "step": 73620 }, { "grad_norm": 0.1768667995929718, "learning_rate": 1.72391650780524e-06, "loss": 0.0118, "step": 73630 }, { "grad_norm": 0.24182246625423431, "learning_rate": 1.718540186020301e-06, "loss": 0.012, "step": 73640 }, { "grad_norm": 0.11033201962709427, "learning_rate": 1.7131721142163437e-06, "loss": 0.0115, "step": 73650 }, { "grad_norm": 0.292177677154541, "learning_rate": 1.7078122933106233e-06, "loss": 0.0111, "step": 73660 }, { "grad_norm": 0.15468652546405792, "learning_rate": 1.7024607242189905e-06, "loss": 0.01, "step": 73670 }, { "grad_norm": 0.12049923092126846, "learning_rate": 1.6971174078558749e-06, "loss": 0.0064, "step": 73680 }, { "grad_norm": 0.1116986945271492, "learning_rate": 1.6917823451343073e-06, "loss": 0.0124, "step": 73690 }, { "grad_norm": 0.1858634650707245, "learning_rate": 1.6864555369659142e-06, "loss": 0.0084, "step": 73700 }, { "grad_norm": 0.12042461335659027, "learning_rate": 1.681136984260878e-06, "loss": 0.0119, "step": 73710 }, { "grad_norm": 0.17754799127578735, "learning_rate": 1.6758266879280172e-06, "loss": 0.0103, "step": 73720 }, { "grad_norm": 0.1476297229528427, "learning_rate": 1.6705246488746996e-06, "loss": 0.0106, "step": 73730 }, { "grad_norm": 0.22689710557460785, "learning_rate": 1.6652308680069062e-06, "loss": 0.0091, "step": 73740 }, { "grad_norm": 0.1873220056295395, "learning_rate": 1.6599453462292081e-06, "loss": 0.0096, "step": 73750 }, { "grad_norm": 0.5247069001197815, "learning_rate": 1.6546680844447326e-06, "loss": 0.0118, "step": 73760 }, { "grad_norm": 0.14291894435882568, "learning_rate": 1.6493990835552475e-06, "loss": 0.0105, "step": 73770 }, { "grad_norm": 0.15185098350048065, "learning_rate": 1.6441383444610547e-06, "loss": 0.0086, "step": 73780 }, { "grad_norm": 0.2070305347442627, "learning_rate": 1.6388858680610908e-06, "loss": 0.0117, "step": 73790 }, { "grad_norm": 0.17339341342449188, "learning_rate": 1.6336416552528543e-06, "loss": 0.0102, "step": 73800 }, { "grad_norm": 0.15278850495815277, "learning_rate": 1.6284057069324288e-06, "loss": 0.0125, "step": 73810 }, { "grad_norm": 0.12526117265224457, "learning_rate": 1.623178023994504e-06, "loss": 0.0101, "step": 73820 }, { "grad_norm": 0.15602616965770721, "learning_rate": 1.6179586073323483e-06, "loss": 0.0108, "step": 73830 }, { "grad_norm": 0.25298792123794556, "learning_rate": 1.612747457837821e-06, "loss": 0.0099, "step": 73840 }, { "grad_norm": 0.169661745429039, "learning_rate": 1.607544576401354e-06, "loss": 0.0111, "step": 73850 }, { "grad_norm": 0.14371290802955627, "learning_rate": 1.6023499639119754e-06, "loss": 0.011, "step": 73860 }, { "grad_norm": 0.2831047773361206, "learning_rate": 1.5971636212573138e-06, "loss": 0.0097, "step": 73870 }, { "grad_norm": 0.20181706547737122, "learning_rate": 1.5919855493235714e-06, "loss": 0.0101, "step": 73880 }, { "grad_norm": 0.19805008172988892, "learning_rate": 1.586815748995535e-06, "loss": 0.0073, "step": 73890 }, { "grad_norm": 0.08488349616527557, "learning_rate": 1.5816542211565866e-06, "loss": 0.0077, "step": 73900 }, { "grad_norm": 0.14130181074142456, "learning_rate": 1.5765009666886765e-06, "loss": 0.0083, "step": 73910 }, { "grad_norm": 0.2589937448501587, "learning_rate": 1.5713559864723781e-06, "loss": 0.0084, "step": 73920 }, { "grad_norm": 0.26916706562042236, "learning_rate": 1.5662192813868104e-06, "loss": 0.0087, "step": 73930 }, { "grad_norm": 0.125687375664711, "learning_rate": 1.561090852309699e-06, "loss": 0.0158, "step": 73940 }, { "grad_norm": 0.22344663739204407, "learning_rate": 1.5559707001173651e-06, "loss": 0.0135, "step": 73950 }, { "grad_norm": 0.21760617196559906, "learning_rate": 1.5508588256846757e-06, "loss": 0.0108, "step": 73960 }, { "grad_norm": 0.11696401238441467, "learning_rate": 1.5457552298851319e-06, "loss": 0.0071, "step": 73970 }, { "grad_norm": 0.25273674726486206, "learning_rate": 1.5406599135907918e-06, "loss": 0.0099, "step": 73980 }, { "grad_norm": 0.12732768058776855, "learning_rate": 1.5355728776723088e-06, "loss": 0.008, "step": 73990 }, { "grad_norm": 0.11833947151899338, "learning_rate": 1.5304941229989155e-06, "loss": 0.0113, "step": 74000 }, { "grad_norm": 0.17652778327465057, "learning_rate": 1.5254236504384345e-06, "loss": 0.0091, "step": 74010 }, { "grad_norm": 0.14339816570281982, "learning_rate": 1.5203614608572726e-06, "loss": 0.0096, "step": 74020 }, { "grad_norm": 0.1572837382555008, "learning_rate": 1.5153075551204044e-06, "loss": 0.0077, "step": 74030 }, { "grad_norm": 0.19504545629024506, "learning_rate": 1.5102619340914225e-06, "loss": 0.0159, "step": 74040 }, { "grad_norm": 0.28267207741737366, "learning_rate": 1.505224598632482e-06, "loss": 0.0076, "step": 74050 }, { "grad_norm": 0.12488728016614914, "learning_rate": 1.5001955496043164e-06, "loss": 0.0125, "step": 74060 }, { "grad_norm": 0.31826603412628174, "learning_rate": 1.4951747878662604e-06, "loss": 0.0067, "step": 74070 }, { "grad_norm": 0.09632355719804764, "learning_rate": 1.4901623142762221e-06, "loss": 0.0056, "step": 74080 }, { "grad_norm": 0.12009105831384659, "learning_rate": 1.4851581296907001e-06, "loss": 0.0117, "step": 74090 }, { "grad_norm": 0.16637751460075378, "learning_rate": 1.4801622349647714e-06, "loss": 0.0074, "step": 74100 }, { "grad_norm": 0.23733389377593994, "learning_rate": 1.4751746309520976e-06, "loss": 0.009, "step": 74110 }, { "grad_norm": 0.17033660411834717, "learning_rate": 1.4701953185049132e-06, "loss": 0.0108, "step": 74120 }, { "grad_norm": 0.09626029431819916, "learning_rate": 1.4652242984740661e-06, "loss": 0.0115, "step": 74130 }, { "grad_norm": 0.17255344986915588, "learning_rate": 1.4602615717089484e-06, "loss": 0.014, "step": 74140 }, { "grad_norm": 0.17429432272911072, "learning_rate": 1.4553071390575657e-06, "loss": 0.0095, "step": 74150 }, { "grad_norm": 0.10353133082389832, "learning_rate": 1.450361001366496e-06, "loss": 0.0119, "step": 74160 }, { "grad_norm": 0.11510712653398514, "learning_rate": 1.4454231594809021e-06, "loss": 0.01, "step": 74170 }, { "grad_norm": 0.22631147503852844, "learning_rate": 1.4404936142445036e-06, "loss": 0.0123, "step": 74180 }, { "grad_norm": 0.3190782070159912, "learning_rate": 1.4355723664996546e-06, "loss": 0.0099, "step": 74190 }, { "grad_norm": 0.7451736330986023, "learning_rate": 1.4306594170872433e-06, "loss": 0.0132, "step": 74200 }, { "grad_norm": 0.17436504364013672, "learning_rate": 1.4257547668467598e-06, "loss": 0.0137, "step": 74210 }, { "grad_norm": 0.11965823918581009, "learning_rate": 1.4208584166162886e-06, "loss": 0.0075, "step": 74220 }, { "grad_norm": 0.19146643579006195, "learning_rate": 1.4159703672324553e-06, "loss": 0.0092, "step": 74230 }, { "grad_norm": 0.14276839792728424, "learning_rate": 1.4110906195305252e-06, "loss": 0.0098, "step": 74240 }, { "grad_norm": 0.19742633402347565, "learning_rate": 1.406219174344292e-06, "loss": 0.0079, "step": 74250 }, { "grad_norm": 0.12103129923343658, "learning_rate": 1.401356032506157e-06, "loss": 0.0068, "step": 74260 }, { "grad_norm": 0.15458695590496063, "learning_rate": 1.3965011948471051e-06, "loss": 0.0117, "step": 74270 }, { "grad_norm": 0.26206669211387634, "learning_rate": 1.3916546621966842e-06, "loss": 0.0087, "step": 74280 }, { "grad_norm": 0.19351999461650848, "learning_rate": 1.3868164353830482e-06, "loss": 0.0069, "step": 74290 }, { "grad_norm": 0.18247109651565552, "learning_rate": 1.3819865152329082e-06, "loss": 0.007, "step": 74300 }, { "grad_norm": 0.148654505610466, "learning_rate": 1.377164902571565e-06, "loss": 0.011, "step": 74310 }, { "grad_norm": 0.7934630513191223, "learning_rate": 1.372351598222904e-06, "loss": 0.0099, "step": 74320 }, { "grad_norm": 0.1663845330476761, "learning_rate": 1.3675466030093843e-06, "loss": 0.0103, "step": 74330 }, { "grad_norm": 0.26938578486442566, "learning_rate": 1.3627499177520486e-06, "loss": 0.0098, "step": 74340 }, { "grad_norm": 0.34039801359176636, "learning_rate": 1.3579615432705196e-06, "loss": 0.011, "step": 74350 }, { "grad_norm": 0.14949220418930054, "learning_rate": 1.3531814803829978e-06, "loss": 0.009, "step": 74360 }, { "grad_norm": 0.09204111993312836, "learning_rate": 1.3484097299062803e-06, "loss": 0.0102, "step": 74370 }, { "grad_norm": 0.22100573778152466, "learning_rate": 1.343646292655698e-06, "loss": 0.0104, "step": 74380 }, { "grad_norm": 0.17008046805858612, "learning_rate": 1.3388911694452277e-06, "loss": 0.0088, "step": 74390 }, { "grad_norm": 0.3571472764015198, "learning_rate": 1.3341443610873638e-06, "loss": 0.0141, "step": 74400 }, { "grad_norm": 0.27499285340309143, "learning_rate": 1.329405868393213e-06, "loss": 0.0109, "step": 74410 }, { "grad_norm": 0.2546228766441345, "learning_rate": 1.3246756921724613e-06, "loss": 0.0129, "step": 74420 }, { "grad_norm": 0.1804027259349823, "learning_rate": 1.3199538332333506e-06, "loss": 0.0092, "step": 74430 }, { "grad_norm": 0.05792775750160217, "learning_rate": 1.3152402923827411e-06, "loss": 0.0119, "step": 74440 }, { "grad_norm": 0.1309347152709961, "learning_rate": 1.3105350704260277e-06, "loss": 0.0067, "step": 74450 }, { "grad_norm": 0.07607053965330124, "learning_rate": 1.305838168167206e-06, "loss": 0.0076, "step": 74460 }, { "grad_norm": 0.10986582934856415, "learning_rate": 1.301149586408862e-06, "loss": 0.0066, "step": 74470 }, { "grad_norm": 0.12770508229732513, "learning_rate": 1.2964693259521322e-06, "loss": 0.0104, "step": 74480 }, { "grad_norm": 0.354339063167572, "learning_rate": 1.2917973875967548e-06, "loss": 0.0069, "step": 74490 }, { "grad_norm": 0.374737411737442, "learning_rate": 1.2871337721410249e-06, "loss": 0.0135, "step": 74500 }, { "grad_norm": 0.15704257786273956, "learning_rate": 1.2824784803818379e-06, "loss": 0.0093, "step": 74510 }, { "grad_norm": 0.2508698105812073, "learning_rate": 1.2778315131146467e-06, "loss": 0.0081, "step": 74520 }, { "grad_norm": 0.21622925996780396, "learning_rate": 1.2731928711334994e-06, "loss": 0.0096, "step": 74530 }, { "grad_norm": 0.1198374480009079, "learning_rate": 1.2685625552310122e-06, "loss": 0.007, "step": 74540 }, { "grad_norm": 0.21461045742034912, "learning_rate": 1.263940566198374e-06, "loss": 0.0091, "step": 74550 }, { "grad_norm": 0.06955330073833466, "learning_rate": 1.259326904825353e-06, "loss": 0.0123, "step": 74560 }, { "grad_norm": 0.3086768686771393, "learning_rate": 1.2547215719003137e-06, "loss": 0.0113, "step": 74570 }, { "grad_norm": 0.2576957643032074, "learning_rate": 1.2501245682101703e-06, "loss": 0.0112, "step": 74580 }, { "grad_norm": 0.19344711303710938, "learning_rate": 1.2455358945404171e-06, "loss": 0.012, "step": 74590 }, { "grad_norm": 0.12854401767253876, "learning_rate": 1.2409555516751493e-06, "loss": 0.0097, "step": 74600 }, { "grad_norm": 0.18023450672626495, "learning_rate": 1.2363835403970125e-06, "loss": 0.0089, "step": 74610 }, { "grad_norm": 0.15723060071468353, "learning_rate": 1.2318198614872489e-06, "loss": 0.0107, "step": 74620 }, { "grad_norm": 0.15629155933856964, "learning_rate": 1.2272645157256457e-06, "loss": 0.0072, "step": 74630 }, { "grad_norm": 0.16821402311325073, "learning_rate": 1.222717503890608e-06, "loss": 0.0115, "step": 74640 }, { "grad_norm": 0.2854056656360626, "learning_rate": 1.218178826759081e-06, "loss": 0.0157, "step": 74650 }, { "grad_norm": 0.2658041715621948, "learning_rate": 1.213648485106611e-06, "loss": 0.0133, "step": 74660 }, { "grad_norm": 0.29039672017097473, "learning_rate": 1.2091264797073066e-06, "loss": 0.0146, "step": 74670 }, { "grad_norm": 0.2165188193321228, "learning_rate": 1.2046128113338494e-06, "loss": 0.0179, "step": 74680 }, { "grad_norm": 0.19248028099536896, "learning_rate": 1.200107480757512e-06, "loss": 0.0124, "step": 74690 }, { "grad_norm": 0.08259785920381546, "learning_rate": 1.1956104887481168e-06, "loss": 0.0092, "step": 74700 }, { "grad_norm": 0.09505675733089447, "learning_rate": 1.191121836074094e-06, "loss": 0.0072, "step": 74710 }, { "grad_norm": 0.2305295467376709, "learning_rate": 1.1866415235024186e-06, "loss": 0.01, "step": 74720 }, { "grad_norm": 0.21430498361587524, "learning_rate": 1.182169551798662e-06, "loss": 0.0128, "step": 74730 }, { "grad_norm": 0.18739579617977142, "learning_rate": 1.177705921726957e-06, "loss": 0.008, "step": 74740 }, { "grad_norm": 0.12854152917861938, "learning_rate": 1.173250634050016e-06, "loss": 0.012, "step": 74750 }, { "grad_norm": 0.1472039818763733, "learning_rate": 1.1688036895291354e-06, "loss": 0.0085, "step": 74760 }, { "grad_norm": 0.2588614523410797, "learning_rate": 1.1643650889241574e-06, "loss": 0.0089, "step": 74770 }, { "grad_norm": 0.18756122887134552, "learning_rate": 1.1599348329935311e-06, "loss": 0.0093, "step": 74780 }, { "grad_norm": 0.12155602127313614, "learning_rate": 1.1555129224942673e-06, "loss": 0.016, "step": 74790 }, { "grad_norm": 0.15418469905853271, "learning_rate": 1.1510993581819396e-06, "loss": 0.0121, "step": 74800 }, { "grad_norm": 0.15347018837928772, "learning_rate": 1.1466941408107112e-06, "loss": 0.0107, "step": 74810 }, { "grad_norm": 0.13794010877609253, "learning_rate": 1.1422972711333247e-06, "loss": 0.008, "step": 74820 }, { "grad_norm": 0.181688129901886, "learning_rate": 1.1379087499010565e-06, "loss": 0.011, "step": 74830 }, { "grad_norm": 0.1609921008348465, "learning_rate": 1.1335285778638127e-06, "loss": 0.0105, "step": 74840 }, { "grad_norm": 0.28567585349082947, "learning_rate": 1.1291567557700333e-06, "loss": 0.0098, "step": 74850 }, { "grad_norm": 0.20481517910957336, "learning_rate": 1.124793284366743e-06, "loss": 0.0096, "step": 74860 }, { "grad_norm": 0.13755781948566437, "learning_rate": 1.1204381643995455e-06, "loss": 0.0091, "step": 74870 }, { "grad_norm": 0.38006576895713806, "learning_rate": 1.116091396612595e-06, "loss": 0.0115, "step": 74880 }, { "grad_norm": 0.1700538843870163, "learning_rate": 1.1117529817486538e-06, "loss": 0.0073, "step": 74890 }, { "grad_norm": 0.16624420881271362, "learning_rate": 1.107422920549034e-06, "loss": 0.0111, "step": 74900 }, { "grad_norm": 0.2225167453289032, "learning_rate": 1.1031012137536157e-06, "loss": 0.0089, "step": 74910 }, { "grad_norm": 0.21570873260498047, "learning_rate": 1.0987878621008695e-06, "loss": 0.0074, "step": 74920 }, { "grad_norm": 0.1861090362071991, "learning_rate": 1.0944828663278284e-06, "loss": 0.0098, "step": 74930 }, { "grad_norm": 0.06994806975126266, "learning_rate": 1.0901862271700925e-06, "loss": 0.0114, "step": 74940 }, { "grad_norm": 0.14591076970100403, "learning_rate": 1.0858979453618467e-06, "loss": 0.0097, "step": 74950 }, { "grad_norm": 0.17138628661632538, "learning_rate": 1.0816180216358441e-06, "loss": 0.0135, "step": 74960 }, { "grad_norm": 0.25254496932029724, "learning_rate": 1.0773464567233937e-06, "loss": 0.0095, "step": 74970 }, { "grad_norm": 0.2494296431541443, "learning_rate": 1.0730832513543953e-06, "loss": 0.0116, "step": 74980 }, { "grad_norm": 0.10099197924137115, "learning_rate": 1.0688284062573218e-06, "loss": 0.0079, "step": 74990 }, { "grad_norm": 0.2655671536922455, "learning_rate": 1.0645819221591969e-06, "loss": 0.0094, "step": 75000 }, { "grad_norm": 0.19119048118591309, "learning_rate": 1.0603437997856347e-06, "loss": 0.0079, "step": 75010 }, { "grad_norm": 0.08710745722055435, "learning_rate": 1.0561140398608228e-06, "loss": 0.0098, "step": 75020 }, { "grad_norm": 0.08482621610164642, "learning_rate": 1.0518926431074937e-06, "loss": 0.0076, "step": 75030 }, { "grad_norm": 0.12877978384494781, "learning_rate": 1.0476796102469877e-06, "loss": 0.0098, "step": 75040 }, { "grad_norm": 0.17786210775375366, "learning_rate": 1.0434749419991786e-06, "loss": 0.0112, "step": 75050 }, { "grad_norm": 0.17829100787639618, "learning_rate": 1.0392786390825415e-06, "loss": 0.0102, "step": 75060 }, { "grad_norm": 0.2697197496891022, "learning_rate": 1.0350907022141087e-06, "loss": 0.0101, "step": 75070 }, { "grad_norm": 0.2705243229866028, "learning_rate": 1.0309111321094744e-06, "loss": 0.0097, "step": 75080 }, { "grad_norm": 0.1368514597415924, "learning_rate": 1.026739929482834e-06, "loss": 0.0083, "step": 75090 }, { "grad_norm": 0.29903286695480347, "learning_rate": 1.022577095046906e-06, "loss": 0.013, "step": 75100 }, { "grad_norm": 0.14833664894104004, "learning_rate": 1.0184226295130217e-06, "loss": 0.0084, "step": 75110 }, { "grad_norm": 0.1794784963130951, "learning_rate": 1.0142765335910575e-06, "loss": 0.011, "step": 75120 }, { "grad_norm": 0.18159151077270508, "learning_rate": 1.0101388079894746e-06, "loss": 0.0089, "step": 75130 }, { "grad_norm": 0.1750345677137375, "learning_rate": 1.0060094534152908e-06, "loss": 0.0084, "step": 75140 }, { "grad_norm": 0.15610331296920776, "learning_rate": 1.0018884705741028e-06, "loss": 0.0087, "step": 75150 }, { "grad_norm": 0.11194483935832977, "learning_rate": 9.977758601700805e-07, "loss": 0.0127, "step": 75160 }, { "grad_norm": 0.11728513985872269, "learning_rate": 9.93671622905945e-07, "loss": 0.0152, "step": 75170 }, { "grad_norm": 0.18856072425842285, "learning_rate": 9.895757594830024e-07, "loss": 0.0094, "step": 75180 }, { "grad_norm": 0.10595479607582092, "learning_rate": 9.854882706011204e-07, "loss": 0.0078, "step": 75190 }, { "grad_norm": 0.2000010907649994, "learning_rate": 9.814091569587513e-07, "loss": 0.0092, "step": 75200 }, { "grad_norm": 0.31886497139930725, "learning_rate": 9.77338419252888e-07, "loss": 0.0113, "step": 75210 }, { "grad_norm": 0.21635693311691284, "learning_rate": 9.732760581791289e-07, "loss": 0.0076, "step": 75220 }, { "grad_norm": 0.17624592781066895, "learning_rate": 9.692220744315972e-07, "loss": 0.0107, "step": 75230 }, { "grad_norm": 0.08794091641902924, "learning_rate": 9.651764687030162e-07, "loss": 0.0072, "step": 75240 }, { "grad_norm": 0.16372814774513245, "learning_rate": 9.611392416846776e-07, "loss": 0.0115, "step": 75250 }, { "grad_norm": 0.11108998209238052, "learning_rate": 9.571103940664238e-07, "loss": 0.0101, "step": 75260 }, { "grad_norm": 0.1212928518652916, "learning_rate": 9.530899265366822e-07, "loss": 0.0115, "step": 75270 }, { "grad_norm": 0.19232213497161865, "learning_rate": 9.490778397824307e-07, "loss": 0.0079, "step": 75280 }, { "grad_norm": 0.2951856255531311, "learning_rate": 9.450741344892322e-07, "loss": 0.009, "step": 75290 }, { "grad_norm": 0.30735278129577637, "learning_rate": 9.410788113412116e-07, "loss": 0.01, "step": 75300 }, { "grad_norm": 0.1499732881784439, "learning_rate": 9.370918710210563e-07, "loss": 0.0079, "step": 75310 }, { "grad_norm": 0.16771988570690155, "learning_rate": 9.331133142100268e-07, "loss": 0.0101, "step": 75320 }, { "grad_norm": 0.20250821113586426, "learning_rate": 9.291431415879459e-07, "loss": 0.0087, "step": 75330 }, { "grad_norm": 0.12926724553108215, "learning_rate": 9.251813538332155e-07, "loss": 0.0125, "step": 75340 }, { "grad_norm": 0.20418085157871246, "learning_rate": 9.212279516227884e-07, "loss": 0.0088, "step": 75350 }, { "grad_norm": 0.16821898519992828, "learning_rate": 9.172829356322021e-07, "loss": 0.0085, "step": 75360 }, { "grad_norm": 0.08267582952976227, "learning_rate": 9.133463065355452e-07, "loss": 0.0067, "step": 75370 }, { "grad_norm": 0.25968489050865173, "learning_rate": 9.094180650054796e-07, "loss": 0.0106, "step": 75380 }, { "grad_norm": 0.27992916107177734, "learning_rate": 9.054982117132404e-07, "loss": 0.0082, "step": 75390 }, { "grad_norm": 0.22749820351600647, "learning_rate": 9.015867473286143e-07, "loss": 0.0108, "step": 75400 }, { "grad_norm": 0.2457589954137802, "learning_rate": 8.976836725199778e-07, "loss": 0.0078, "step": 75410 }, { "grad_norm": 0.15478460490703583, "learning_rate": 8.937889879542416e-07, "loss": 0.0096, "step": 75420 }, { "grad_norm": 0.23184117674827576, "learning_rate": 8.899026942969068e-07, "loss": 0.0074, "step": 75430 }, { "grad_norm": 0.1282159388065338, "learning_rate": 8.860247922120424e-07, "loss": 0.0106, "step": 75440 }, { "grad_norm": 0.273298442363739, "learning_rate": 8.821552823622737e-07, "loss": 0.01, "step": 75450 }, { "grad_norm": 0.15008953213691711, "learning_rate": 8.782941654087884e-07, "loss": 0.0095, "step": 75460 }, { "grad_norm": 0.2626603841781616, "learning_rate": 8.744414420113478e-07, "loss": 0.0096, "step": 75470 }, { "grad_norm": 0.21362832188606262, "learning_rate": 8.705971128282753e-07, "loss": 0.0087, "step": 75480 }, { "grad_norm": 0.2084031105041504, "learning_rate": 8.66761178516473e-07, "loss": 0.0088, "step": 75490 }, { "grad_norm": 0.11408697068691254, "learning_rate": 8.629336397313781e-07, "loss": 0.0076, "step": 75500 }, { "grad_norm": 0.11397583037614822, "learning_rate": 8.591144971270282e-07, "loss": 0.0103, "step": 75510 }, { "grad_norm": 0.1468387395143509, "learning_rate": 8.553037513560069e-07, "loss": 0.0138, "step": 75520 }, { "grad_norm": 0.1275852918624878, "learning_rate": 8.515014030694546e-07, "loss": 0.0066, "step": 75530 }, { "grad_norm": 0.11302409321069717, "learning_rate": 8.47707452917107e-07, "loss": 0.0079, "step": 75540 }, { "grad_norm": 0.16468459367752075, "learning_rate": 8.439219015472343e-07, "loss": 0.0097, "step": 75550 }, { "grad_norm": 0.18777011334896088, "learning_rate": 8.401447496066861e-07, "loss": 0.007, "step": 75560 }, { "grad_norm": 0.2245745211839676, "learning_rate": 8.363759977408792e-07, "loss": 0.0097, "step": 75570 }, { "grad_norm": 0.21519756317138672, "learning_rate": 8.326156465937817e-07, "loss": 0.0118, "step": 75580 }, { "grad_norm": 0.35896360874176025, "learning_rate": 8.28863696807941e-07, "loss": 0.0162, "step": 75590 }, { "grad_norm": 0.07431303709745407, "learning_rate": 8.251201490244609e-07, "loss": 0.0088, "step": 75600 }, { "grad_norm": 0.14180342853069305, "learning_rate": 8.213850038830129e-07, "loss": 0.0104, "step": 75610 }, { "grad_norm": 0.10078161209821701, "learning_rate": 8.176582620218309e-07, "loss": 0.0104, "step": 75620 }, { "grad_norm": 0.24531793594360352, "learning_rate": 8.139399240777057e-07, "loss": 0.0082, "step": 75630 }, { "grad_norm": 0.25433290004730225, "learning_rate": 8.102299906860122e-07, "loss": 0.0108, "step": 75640 }, { "grad_norm": 0.28849130868911743, "learning_rate": 8.065284624806657e-07, "loss": 0.0096, "step": 75650 }, { "grad_norm": 0.2554985582828522, "learning_rate": 8.028353400941601e-07, "loss": 0.0107, "step": 75660 }, { "grad_norm": 0.059761933982372284, "learning_rate": 7.991506241575575e-07, "loss": 0.0117, "step": 75670 }, { "grad_norm": 0.26638734340667725, "learning_rate": 7.954743153004541e-07, "loss": 0.0162, "step": 75680 }, { "grad_norm": 0.18276691436767578, "learning_rate": 7.918064141510528e-07, "loss": 0.0149, "step": 75690 }, { "grad_norm": 0.08261129260063171, "learning_rate": 7.881469213360859e-07, "loss": 0.0137, "step": 75700 }, { "grad_norm": 0.13105957210063934, "learning_rate": 7.844958374808642e-07, "loss": 0.0077, "step": 75710 }, { "grad_norm": 0.1950872540473938, "learning_rate": 7.808531632092608e-07, "loss": 0.0085, "step": 75720 }, { "grad_norm": 0.21357759833335876, "learning_rate": 7.772188991436946e-07, "loss": 0.0083, "step": 75730 }, { "grad_norm": 0.17147096991539001, "learning_rate": 7.735930459051799e-07, "loss": 0.0092, "step": 75740 }, { "grad_norm": 0.23900829255580902, "learning_rate": 7.699756041132655e-07, "loss": 0.0101, "step": 75750 }, { "grad_norm": 0.2305433601140976, "learning_rate": 7.663665743860793e-07, "loss": 0.008, "step": 75760 }, { "grad_norm": 0.18945971131324768, "learning_rate": 7.627659573403001e-07, "loss": 0.009, "step": 75770 }, { "grad_norm": 0.2184353917837143, "learning_rate": 7.591737535911802e-07, "loss": 0.0098, "step": 75780 }, { "grad_norm": 0.07199209183454514, "learning_rate": 7.555899637525288e-07, "loss": 0.0091, "step": 75790 }, { "grad_norm": 0.25179311633110046, "learning_rate": 7.520145884367058e-07, "loss": 0.0068, "step": 75800 }, { "grad_norm": 0.16192923486232758, "learning_rate": 7.484476282546615e-07, "loss": 0.0102, "step": 75810 }, { "grad_norm": 0.2652813792228699, "learning_rate": 7.448890838158806e-07, "loss": 0.0127, "step": 75820 }, { "grad_norm": 0.13107770681381226, "learning_rate": 7.413389557284267e-07, "loss": 0.0126, "step": 75830 }, { "grad_norm": 0.0829724371433258, "learning_rate": 7.377972445989201e-07, "loss": 0.0091, "step": 75840 }, { "grad_norm": 0.16009940207004547, "learning_rate": 7.342639510325377e-07, "loss": 0.0106, "step": 75850 }, { "grad_norm": 0.2505733072757721, "learning_rate": 7.307390756330246e-07, "loss": 0.0061, "step": 75860 }, { "grad_norm": 0.2168862521648407, "learning_rate": 7.272226190026876e-07, "loss": 0.0084, "step": 75870 }, { "grad_norm": 0.3180362582206726, "learning_rate": 7.237145817423907e-07, "loss": 0.0061, "step": 75880 }, { "grad_norm": 0.16861295700073242, "learning_rate": 7.202149644515654e-07, "loss": 0.0092, "step": 75890 }, { "grad_norm": 0.6911920309066772, "learning_rate": 7.167237677281946e-07, "loss": 0.0123, "step": 75900 }, { "grad_norm": 0.18580546975135803, "learning_rate": 7.132409921688288e-07, "loss": 0.0086, "step": 75910 }, { "grad_norm": 0.07240644842386246, "learning_rate": 7.09766638368592e-07, "loss": 0.0068, "step": 75920 }, { "grad_norm": 0.21012918651103973, "learning_rate": 7.063007069211313e-07, "loss": 0.0098, "step": 75930 }, { "grad_norm": 0.20231090486049652, "learning_rate": 7.028431984187067e-07, "loss": 0.0092, "step": 75940 }, { "grad_norm": 0.17105428874492645, "learning_rate": 6.993941134520898e-07, "loss": 0.0112, "step": 75950 }, { "grad_norm": 0.24241669476032257, "learning_rate": 6.959534526106537e-07, "loss": 0.0083, "step": 75960 }, { "grad_norm": 0.11203745752573013, "learning_rate": 6.925212164822948e-07, "loss": 0.006, "step": 75970 }, { "grad_norm": 0.6068598628044128, "learning_rate": 6.890974056535049e-07, "loss": 0.0125, "step": 75980 }, { "grad_norm": 0.28752583265304565, "learning_rate": 6.856820207093106e-07, "loss": 0.0058, "step": 75990 }, { "grad_norm": 0.17956310510635376, "learning_rate": 6.822750622333063e-07, "loss": 0.0095, "step": 76000 }, { "grad_norm": 0.20154821872711182, "learning_rate": 6.788765308076539e-07, "loss": 0.0093, "step": 76010 }, { "grad_norm": 0.28358355164527893, "learning_rate": 6.754864270130668e-07, "loss": 0.0096, "step": 76020 }, { "grad_norm": 0.18275849521160126, "learning_rate": 6.721047514288203e-07, "loss": 0.0097, "step": 76030 }, { "grad_norm": 0.26118066906929016, "learning_rate": 6.687315046327469e-07, "loss": 0.0092, "step": 76040 }, { "grad_norm": 0.38507524132728577, "learning_rate": 6.65366687201252e-07, "loss": 0.0101, "step": 76050 }, { "grad_norm": 0.31377139687538147, "learning_rate": 6.620102997092814e-07, "loss": 0.0127, "step": 76060 }, { "grad_norm": 0.13060416281223297, "learning_rate": 6.586623427303596e-07, "loss": 0.0069, "step": 76070 }, { "grad_norm": 0.0987744852900505, "learning_rate": 6.553228168365455e-07, "loss": 0.0126, "step": 76080 }, { "grad_norm": 0.28535035252571106, "learning_rate": 6.519917225984884e-07, "loss": 0.0109, "step": 76090 }, { "grad_norm": 0.17438966035842896, "learning_rate": 6.486690605853718e-07, "loss": 0.0086, "step": 76100 }, { "grad_norm": 0.28936320543289185, "learning_rate": 6.453548313649527e-07, "loss": 0.0117, "step": 76110 }, { "grad_norm": 0.22133736312389374, "learning_rate": 6.420490355035446e-07, "loss": 0.0095, "step": 76120 }, { "grad_norm": 0.308993399143219, "learning_rate": 6.387516735660071e-07, "loss": 0.0094, "step": 76130 }, { "grad_norm": 0.18603278696537018, "learning_rate": 6.35462746115778e-07, "loss": 0.0095, "step": 76140 }, { "grad_norm": 0.16586348414421082, "learning_rate": 6.321822537148358e-07, "loss": 0.0094, "step": 76150 }, { "grad_norm": 0.15800170600414276, "learning_rate": 6.289101969237432e-07, "loss": 0.0085, "step": 76160 }, { "grad_norm": 0.25157472491264343, "learning_rate": 6.256465763015918e-07, "loss": 0.0074, "step": 76170 }, { "grad_norm": 0.1006886288523674, "learning_rate": 6.22391392406052e-07, "loss": 0.0102, "step": 76180 }, { "grad_norm": 0.10966452956199646, "learning_rate": 6.191446457933403e-07, "loss": 0.0062, "step": 76190 }, { "grad_norm": 0.07498754560947418, "learning_rate": 6.159063370182406e-07, "loss": 0.0065, "step": 76200 }, { "grad_norm": 0.20256155729293823, "learning_rate": 6.126764666340879e-07, "loss": 0.0073, "step": 76210 }, { "grad_norm": 0.10226549953222275, "learning_rate": 6.094550351927852e-07, "loss": 0.0094, "step": 76220 }, { "grad_norm": 0.21705593168735504, "learning_rate": 6.062420432447757e-07, "loss": 0.0083, "step": 76230 }, { "grad_norm": 0.14994972944259644, "learning_rate": 6.030374913390813e-07, "loss": 0.0072, "step": 76240 }, { "grad_norm": 0.22209954261779785, "learning_rate": 5.998413800232694e-07, "loss": 0.0084, "step": 76250 }, { "grad_norm": 0.14327247440814972, "learning_rate": 5.966537098434755e-07, "loss": 0.0089, "step": 76260 }, { "grad_norm": 0.2356313318014145, "learning_rate": 5.934744813443694e-07, "loss": 0.0078, "step": 76270 }, { "grad_norm": 0.3327234983444214, "learning_rate": 5.903036950692054e-07, "loss": 0.0099, "step": 76280 }, { "grad_norm": 0.21133926510810852, "learning_rate": 5.871413515597835e-07, "loss": 0.0111, "step": 76290 }, { "grad_norm": 0.3159639239311218, "learning_rate": 5.839874513564547e-07, "loss": 0.0082, "step": 76300 }, { "grad_norm": 0.2136743813753128, "learning_rate": 5.808419949981436e-07, "loss": 0.0107, "step": 76310 }, { "grad_norm": 0.24201169610023499, "learning_rate": 5.777049830223257e-07, "loss": 0.0138, "step": 76320 }, { "grad_norm": 0.3011925220489502, "learning_rate": 5.745764159650114e-07, "loss": 0.0114, "step": 76330 }, { "grad_norm": 0.15088608860969543, "learning_rate": 5.71456294360806e-07, "loss": 0.0076, "step": 76340 }, { "grad_norm": 0.42471978068351746, "learning_rate": 5.683446187428443e-07, "loss": 0.0083, "step": 76350 }, { "grad_norm": 0.21115346252918243, "learning_rate": 5.652413896428288e-07, "loss": 0.0097, "step": 76360 }, { "grad_norm": 0.26222676038742065, "learning_rate": 5.621466075910131e-07, "loss": 0.0109, "step": 76370 }, { "grad_norm": 0.08267492800951004, "learning_rate": 5.590602731162187e-07, "loss": 0.007, "step": 76380 }, { "grad_norm": 0.1812623143196106, "learning_rate": 5.559823867458125e-07, "loss": 0.0093, "step": 76390 }, { "grad_norm": 0.1713634580373764, "learning_rate": 5.529129490057128e-07, "loss": 0.0129, "step": 76400 }, { "grad_norm": 0.12710075080394745, "learning_rate": 5.498519604204167e-07, "loss": 0.0105, "step": 76410 }, { "grad_norm": 0.15100982785224915, "learning_rate": 5.467994215129501e-07, "loss": 0.0115, "step": 76420 }, { "grad_norm": 0.11751366406679153, "learning_rate": 5.437553328049183e-07, "loss": 0.0114, "step": 76430 }, { "grad_norm": 0.156174898147583, "learning_rate": 5.407196948164661e-07, "loss": 0.0092, "step": 76440 }, { "grad_norm": 0.18515431880950928, "learning_rate": 5.376925080663009e-07, "loss": 0.0123, "step": 76450 }, { "grad_norm": 0.4359505772590637, "learning_rate": 5.346737730716977e-07, "loss": 0.0095, "step": 76460 }, { "grad_norm": 0.11096150428056717, "learning_rate": 5.316634903484607e-07, "loss": 0.0097, "step": 76470 }, { "grad_norm": 0.09772218763828278, "learning_rate": 5.286616604109728e-07, "loss": 0.0076, "step": 76480 }, { "grad_norm": 0.19240239262580872, "learning_rate": 5.256682837721627e-07, "loss": 0.0171, "step": 76490 }, { "grad_norm": 0.1741793304681778, "learning_rate": 5.226833609435156e-07, "loss": 0.011, "step": 76500 }, { "grad_norm": 0.17908240854740143, "learning_rate": 5.19706892435079e-07, "loss": 0.0109, "step": 76510 }, { "grad_norm": 0.19102133810520172, "learning_rate": 5.167388787554406e-07, "loss": 0.0095, "step": 76520 }, { "grad_norm": 0.23181577026844025, "learning_rate": 5.137793204117614e-07, "loss": 0.0097, "step": 76530 }, { "grad_norm": 0.28791162371635437, "learning_rate": 5.108282179097479e-07, "loss": 0.0079, "step": 76540 }, { "grad_norm": 0.3068058490753174, "learning_rate": 5.078855717536523e-07, "loss": 0.0141, "step": 76550 }, { "grad_norm": 0.22781948745250702, "learning_rate": 5.049513824463059e-07, "loss": 0.0084, "step": 76560 }, { "grad_norm": 0.11819276213645935, "learning_rate": 5.020256504890742e-07, "loss": 0.0086, "step": 76570 }, { "grad_norm": 0.11986158043146133, "learning_rate": 4.991083763818849e-07, "loss": 0.0108, "step": 76580 }, { "grad_norm": 0.23383446037769318, "learning_rate": 4.961995606232228e-07, "loss": 0.0104, "step": 76590 }, { "grad_norm": 0.14534680545330048, "learning_rate": 4.932992037101236e-07, "loss": 0.0089, "step": 76600 }, { "grad_norm": 0.192199245095253, "learning_rate": 4.904073061381798e-07, "loss": 0.0113, "step": 76610 }, { "grad_norm": 0.2702814042568207, "learning_rate": 4.875238684015349e-07, "loss": 0.0141, "step": 76620 }, { "grad_norm": 0.17610056698322296, "learning_rate": 4.846488909928948e-07, "loss": 0.0076, "step": 76630 }, { "grad_norm": 0.08210263401269913, "learning_rate": 4.81782374403511e-07, "loss": 0.0126, "step": 76640 }, { "grad_norm": 0.18172267079353333, "learning_rate": 4.789243191231918e-07, "loss": 0.0072, "step": 76650 }, { "grad_norm": 0.19137844443321228, "learning_rate": 4.76074725640302e-07, "loss": 0.0073, "step": 76660 }, { "grad_norm": 0.2551300525665283, "learning_rate": 4.732335944417632e-07, "loss": 0.0168, "step": 76670 }, { "grad_norm": 0.19524365663528442, "learning_rate": 4.704009260130371e-07, "loss": 0.0091, "step": 76680 }, { "grad_norm": 0.08837839215993881, "learning_rate": 4.675767208381587e-07, "loss": 0.0096, "step": 76690 }, { "grad_norm": 0.35375478863716125, "learning_rate": 4.647609793997032e-07, "loss": 0.0107, "step": 76700 }, { "grad_norm": 0.0603601299226284, "learning_rate": 4.619537021788023e-07, "loss": 0.0091, "step": 76710 }, { "grad_norm": 0.1314718872308731, "learning_rate": 4.5915488965515005e-07, "loss": 0.0115, "step": 76720 }, { "grad_norm": 0.19174528121948242, "learning_rate": 4.563645423069807e-07, "loss": 0.0096, "step": 76730 }, { "grad_norm": 0.27377647161483765, "learning_rate": 4.535826606110849e-07, "loss": 0.0089, "step": 76740 }, { "grad_norm": 0.21373572945594788, "learning_rate": 4.508092450428214e-07, "loss": 0.0103, "step": 76750 }, { "grad_norm": 0.23645050823688507, "learning_rate": 4.480442960760778e-07, "loss": 0.0098, "step": 76760 }, { "grad_norm": 0.0926937386393547, "learning_rate": 4.4528781418332053e-07, "loss": 0.0094, "step": 76770 }, { "grad_norm": 0.3863723874092102, "learning_rate": 4.4253979983555073e-07, "loss": 0.0137, "step": 76780 }, { "grad_norm": 0.17160379886627197, "learning_rate": 4.3980025350233154e-07, "loss": 0.0076, "step": 76790 }, { "grad_norm": 0.12573425471782684, "learning_rate": 4.370691756517664e-07, "loss": 0.0084, "step": 76800 }, { "grad_norm": 0.15641780197620392, "learning_rate": 4.3434656675053753e-07, "loss": 0.0068, "step": 76810 }, { "grad_norm": 0.23336756229400635, "learning_rate": 4.3163242726385613e-07, "loss": 0.0115, "step": 76820 }, { "grad_norm": 0.1201588362455368, "learning_rate": 4.289267576554956e-07, "loss": 0.0079, "step": 76830 }, { "grad_norm": 0.16254405677318573, "learning_rate": 4.262295583877807e-07, "loss": 0.0078, "step": 76840 }, { "grad_norm": 0.21801118552684784, "learning_rate": 4.235408299215815e-07, "loss": 0.0095, "step": 76850 }, { "grad_norm": 0.15043950080871582, "learning_rate": 4.2086057271634173e-07, "loss": 0.0084, "step": 76860 }, { "grad_norm": 0.12253142148256302, "learning_rate": 4.181887872300394e-07, "loss": 0.0092, "step": 76870 }, { "grad_norm": 0.2013503909111023, "learning_rate": 4.155254739192038e-07, "loss": 0.0092, "step": 76880 }, { "grad_norm": 0.1488076150417328, "learning_rate": 4.128706332389265e-07, "loss": 0.0164, "step": 76890 }, { "grad_norm": 0.22666318714618683, "learning_rate": 4.1022426564284453e-07, "loss": 0.0075, "step": 76900 }, { "grad_norm": 0.17163977026939392, "learning_rate": 4.075863715831574e-07, "loss": 0.0085, "step": 76910 }, { "grad_norm": 0.2394288182258606, "learning_rate": 4.0495695151059887e-07, "loss": 0.0106, "step": 76920 }, { "grad_norm": 0.21446987986564636, "learning_rate": 4.023360058744763e-07, "loss": 0.0069, "step": 76930 }, { "grad_norm": 0.09770802408456802, "learning_rate": 3.997235351226314e-07, "loss": 0.0067, "step": 76940 }, { "grad_norm": 0.21701711416244507, "learning_rate": 3.971195397014571e-07, "loss": 0.0098, "step": 76950 }, { "grad_norm": 0.23507744073867798, "learning_rate": 3.9452402005591417e-07, "loss": 0.0084, "step": 76960 }, { "grad_norm": 0.1572970598936081, "learning_rate": 3.9193697662950334e-07, "loss": 0.0095, "step": 76970 }, { "grad_norm": 0.34291619062423706, "learning_rate": 3.893584098642822e-07, "loss": 0.0109, "step": 76980 }, { "grad_norm": 0.1420021951198578, "learning_rate": 3.867883202008538e-07, "loss": 0.0109, "step": 76990 }, { "grad_norm": 0.25066718459129333, "learning_rate": 3.842267080783779e-07, "loss": 0.01, "step": 77000 }, { "grad_norm": 0.197540283203125, "learning_rate": 3.816735739345656e-07, "loss": 0.01, "step": 77010 }, { "grad_norm": 0.16972112655639648, "learning_rate": 3.791289182056679e-07, "loss": 0.009, "step": 77020 }, { "grad_norm": 0.7620073556900024, "learning_rate": 3.7659274132650913e-07, "loss": 0.0115, "step": 77030 }, { "grad_norm": 0.18537385761737823, "learning_rate": 3.7406504373044826e-07, "loss": 0.0103, "step": 77040 }, { "grad_norm": 0.2221171110868454, "learning_rate": 3.7154582584939533e-07, "loss": 0.0108, "step": 77050 }, { "grad_norm": 0.14004039764404297, "learning_rate": 3.690350881138227e-07, "loss": 0.0088, "step": 77060 }, { "grad_norm": 0.16633810102939606, "learning_rate": 3.665328309527427e-07, "loss": 0.0081, "step": 77070 }, { "grad_norm": 0.11889977008104324, "learning_rate": 3.640390547937245e-07, "loss": 0.0062, "step": 77080 }, { "grad_norm": 0.13977858424186707, "learning_rate": 3.615537600628882e-07, "loss": 0.0114, "step": 77090 }, { "grad_norm": 0.20659948885440826, "learning_rate": 3.5907694718489427e-07, "loss": 0.0125, "step": 77100 }, { "grad_norm": 0.17465448379516602, "learning_rate": 3.566086165829707e-07, "loss": 0.0141, "step": 77110 }, { "grad_norm": 0.3149881362915039, "learning_rate": 3.5414876867888024e-07, "loss": 0.0104, "step": 77120 }, { "grad_norm": 0.09578234702348709, "learning_rate": 3.5169740389295326e-07, "loss": 0.0109, "step": 77130 }, { "grad_norm": 0.40138885378837585, "learning_rate": 3.4925452264405466e-07, "loss": 0.0105, "step": 77140 }, { "grad_norm": 0.21076147258281708, "learning_rate": 3.468201253496062e-07, "loss": 0.0075, "step": 77150 }, { "grad_norm": 0.18841704726219177, "learning_rate": 3.443942124255861e-07, "loss": 0.0088, "step": 77160 }, { "grad_norm": 0.18626706302165985, "learning_rate": 3.4197678428650183e-07, "loss": 0.0112, "step": 77170 }, { "grad_norm": 0.07513191550970078, "learning_rate": 3.3956784134544504e-07, "loss": 0.0075, "step": 77180 }, { "grad_norm": 0.12139341235160828, "learning_rate": 3.3716738401402547e-07, "loss": 0.0112, "step": 77190 }, { "grad_norm": 0.23840859532356262, "learning_rate": 3.3477541270241495e-07, "loss": 0.0083, "step": 77200 }, { "grad_norm": 0.10430620610713959, "learning_rate": 3.3239192781934215e-07, "loss": 0.0098, "step": 77210 }, { "grad_norm": 0.30554354190826416, "learning_rate": 3.3001692977207563e-07, "loss": 0.0098, "step": 77220 }, { "grad_norm": 0.12404945492744446, "learning_rate": 3.276504189664409e-07, "loss": 0.0089, "step": 77230 }, { "grad_norm": 0.11238039284944534, "learning_rate": 3.252923958068088e-07, "loss": 0.01, "step": 77240 }, { "grad_norm": 0.1286768913269043, "learning_rate": 3.2294286069609046e-07, "loss": 0.0118, "step": 77250 }, { "grad_norm": 0.06488855928182602, "learning_rate": 3.2060181403577583e-07, "loss": 0.0082, "step": 77260 }, { "grad_norm": 0.13541348278522491, "learning_rate": 3.1826925622587823e-07, "loss": 0.0067, "step": 77270 }, { "grad_norm": 0.1152571588754654, "learning_rate": 3.159451876649622e-07, "loss": 0.0106, "step": 77280 }, { "grad_norm": 0.2307400107383728, "learning_rate": 3.1362960875015444e-07, "loss": 0.0096, "step": 77290 }, { "grad_norm": 0.3153517246246338, "learning_rate": 3.1132251987711637e-07, "loss": 0.0121, "step": 77300 }, { "grad_norm": 0.15381242334842682, "learning_rate": 3.0902392144007695e-07, "loss": 0.0099, "step": 77310 }, { "grad_norm": 0.1443888396024704, "learning_rate": 3.0673381383179436e-07, "loss": 0.0109, "step": 77320 }, { "grad_norm": 0.2107418328523636, "learning_rate": 3.0445219744358875e-07, "loss": 0.0086, "step": 77330 }, { "grad_norm": 0.235101580619812, "learning_rate": 3.021790726653262e-07, "loss": 0.0091, "step": 77340 }, { "grad_norm": 0.3089804947376251, "learning_rate": 2.9991443988542366e-07, "loss": 0.0095, "step": 77350 }, { "grad_norm": 0.13730239868164062, "learning_rate": 2.976582994908439e-07, "loss": 0.0084, "step": 77360 }, { "grad_norm": 0.19700023531913757, "learning_rate": 2.954106518671007e-07, "loss": 0.0088, "step": 77370 }, { "grad_norm": 0.12965016067028046, "learning_rate": 2.9317149739825356e-07, "loss": 0.0077, "step": 77380 }, { "grad_norm": 0.2716918885707855, "learning_rate": 2.909408364669075e-07, "loss": 0.0102, "step": 77390 }, { "grad_norm": 0.3142831325531006, "learning_rate": 2.887186694542299e-07, "loss": 0.0113, "step": 77400 }, { "grad_norm": 0.14496947824954987, "learning_rate": 2.865049967399225e-07, "loss": 0.0088, "step": 77410 }, { "grad_norm": 0.10943775624036789, "learning_rate": 2.842998187022439e-07, "loss": 0.0077, "step": 77420 }, { "grad_norm": 0.25057387351989746, "learning_rate": 2.8210313571800374e-07, "loss": 0.0099, "step": 77430 }, { "grad_norm": 0.20640292763710022, "learning_rate": 2.7991494816255184e-07, "loss": 0.0085, "step": 77440 }, { "grad_norm": 0.25445738434791565, "learning_rate": 2.777352564097779e-07, "loss": 0.0098, "step": 77450 }, { "grad_norm": 0.15672865509986877, "learning_rate": 2.755640608321508e-07, "loss": 0.0087, "step": 77460 }, { "grad_norm": 0.41879576444625854, "learning_rate": 2.73401361800657e-07, "loss": 0.0105, "step": 77470 }, { "grad_norm": 0.1339551955461502, "learning_rate": 2.7124715968484537e-07, "loss": 0.0092, "step": 77480 }, { "grad_norm": 0.21139176189899445, "learning_rate": 2.691014548528104e-07, "loss": 0.0107, "step": 77490 }, { "grad_norm": 0.1421167254447937, "learning_rate": 2.669642476711864e-07, "loss": 0.0086, "step": 77500 }, { "grad_norm": 0.1019664853811264, "learning_rate": 2.648355385051815e-07, "loss": 0.0069, "step": 77510 }, { "grad_norm": 0.13859817385673523, "learning_rate": 2.627153277185157e-07, "loss": 0.0075, "step": 77520 }, { "grad_norm": 0.3535127341747284, "learning_rate": 2.606036156734881e-07, "loss": 0.0068, "step": 77530 }, { "grad_norm": 0.12550008296966553, "learning_rate": 2.5850040273092127e-07, "loss": 0.0087, "step": 77540 }, { "grad_norm": 0.14892153441905975, "learning_rate": 2.5640568925020536e-07, "loss": 0.0126, "step": 77550 }, { "grad_norm": 0.14168253540992737, "learning_rate": 2.5431947558927083e-07, "loss": 0.0117, "step": 77560 }, { "grad_norm": 0.167704239487648, "learning_rate": 2.522417621045825e-07, "loss": 0.0087, "step": 77570 }, { "grad_norm": 0.238917738199234, "learning_rate": 2.501725491511786e-07, "loss": 0.0133, "step": 77580 }, { "grad_norm": 0.11400322616100311, "learning_rate": 2.481118370826263e-07, "loss": 0.0108, "step": 77590 }, { "grad_norm": 0.21942046284675598, "learning_rate": 2.4605962625104393e-07, "loss": 0.0086, "step": 77600 }, { "grad_norm": 0.28708112239837646, "learning_rate": 2.440159170070955e-07, "loss": 0.0148, "step": 77610 }, { "grad_norm": 0.14546538889408112, "learning_rate": 2.41980709699996e-07, "loss": 0.0104, "step": 77620 }, { "grad_norm": 0.2730499804019928, "learning_rate": 2.3995400467751174e-07, "loss": 0.0099, "step": 77630 }, { "grad_norm": 0.2717773914337158, "learning_rate": 2.3793580228594902e-07, "loss": 0.0198, "step": 77640 }, { "grad_norm": 0.14949160814285278, "learning_rate": 2.3592610287015982e-07, "loss": 0.0076, "step": 77650 }, { "grad_norm": 0.17452341318130493, "learning_rate": 2.3392490677354718e-07, "loss": 0.0102, "step": 77660 }, { "grad_norm": 0.3230712115764618, "learning_rate": 2.3193221433806533e-07, "loss": 0.0063, "step": 77670 }, { "grad_norm": 0.404073029756546, "learning_rate": 2.299480259042086e-07, "loss": 0.0096, "step": 77680 }, { "grad_norm": 0.09224392473697662, "learning_rate": 2.2797234181102244e-07, "loss": 0.0077, "step": 77690 }, { "grad_norm": 0.16900493204593658, "learning_rate": 2.2600516239609238e-07, "loss": 0.0093, "step": 77700 }, { "grad_norm": 0.14243070781230927, "learning_rate": 2.240464879955606e-07, "loss": 0.0125, "step": 77710 }, { "grad_norm": 0.17986495792865753, "learning_rate": 2.2209631894410387e-07, "loss": 0.0105, "step": 77720 }, { "grad_norm": 0.16668209433555603, "learning_rate": 2.2015465557496117e-07, "loss": 0.0073, "step": 77730 }, { "grad_norm": 0.16838379204273224, "learning_rate": 2.1822149821990602e-07, "loss": 0.0092, "step": 77740 }, { "grad_norm": 0.40444889664649963, "learning_rate": 2.1629684720926303e-07, "loss": 0.01, "step": 77750 }, { "grad_norm": 0.11269642412662506, "learning_rate": 2.1438070287189693e-07, "loss": 0.0107, "step": 77760 }, { "grad_norm": 0.15202274918556213, "learning_rate": 2.1247306553523472e-07, "loss": 0.0098, "step": 77770 }, { "grad_norm": 0.37254706025123596, "learning_rate": 2.105739355252323e-07, "loss": 0.0109, "step": 77780 }, { "grad_norm": 0.12295780330896378, "learning_rate": 2.0868331316639678e-07, "loss": 0.0106, "step": 77790 }, { "grad_norm": 0.14258600771427155, "learning_rate": 2.0680119878179193e-07, "loss": 0.0087, "step": 77800 }, { "grad_norm": 0.23991475999355316, "learning_rate": 2.0492759269301054e-07, "loss": 0.0109, "step": 77810 }, { "grad_norm": 0.3136146664619446, "learning_rate": 2.0306249522021315e-07, "loss": 0.0092, "step": 77820 }, { "grad_norm": 0.1540869176387787, "learning_rate": 2.0120590668207816e-07, "loss": 0.0066, "step": 77830 }, { "grad_norm": 0.16953912377357483, "learning_rate": 1.993578273958574e-07, "loss": 0.0054, "step": 77840 }, { "grad_norm": 0.0890602394938469, "learning_rate": 1.975182576773371e-07, "loss": 0.0083, "step": 77850 }, { "grad_norm": 0.38012611865997314, "learning_rate": 1.9568719784083812e-07, "loss": 0.0101, "step": 77860 }, { "grad_norm": 0.13875488936901093, "learning_rate": 1.9386464819924898e-07, "loss": 0.0137, "step": 77870 }, { "grad_norm": 0.1346869170665741, "learning_rate": 1.9205060906399285e-07, "loss": 0.0102, "step": 77880 }, { "grad_norm": 0.14180640876293182, "learning_rate": 1.9024508074503845e-07, "loss": 0.0086, "step": 77890 }, { "grad_norm": 0.13448694348335266, "learning_rate": 1.8844806355089452e-07, "loss": 0.0163, "step": 77900 }, { "grad_norm": 0.1634257435798645, "learning_rate": 1.8665955778863208e-07, "loss": 0.0143, "step": 77910 }, { "grad_norm": 0.3042353689670563, "learning_rate": 1.848795637638512e-07, "loss": 0.0105, "step": 77920 }, { "grad_norm": 0.22305747866630554, "learning_rate": 1.831080817807085e-07, "loss": 0.0102, "step": 77930 }, { "grad_norm": 0.23020020127296448, "learning_rate": 1.813451121418952e-07, "loss": 0.0174, "step": 77940 }, { "grad_norm": 0.1741584837436676, "learning_rate": 1.795906551486648e-07, "loss": 0.0069, "step": 77950 }, { "grad_norm": 0.22467972338199615, "learning_rate": 1.7784471110079414e-07, "loss": 0.0102, "step": 77960 }, { "grad_norm": 0.25202760100364685, "learning_rate": 1.7610728029662793e-07, "loss": 0.0097, "step": 77970 }, { "grad_norm": 0.10568339377641678, "learning_rate": 1.7437836303303979e-07, "loss": 0.0082, "step": 77980 }, { "grad_norm": 0.10377916693687439, "learning_rate": 1.726579596054545e-07, "loss": 0.0078, "step": 77990 }, { "grad_norm": 0.23885412514209747, "learning_rate": 1.7094607030784803e-07, "loss": 0.0143, "step": 78000 }, { "grad_norm": 0.19146431982517242, "learning_rate": 1.6924269543272532e-07, "loss": 0.0091, "step": 78010 }, { "grad_norm": 0.16544702649116516, "learning_rate": 1.6754783527115348e-07, "loss": 0.0081, "step": 78020 }, { "grad_norm": 0.13655985891819, "learning_rate": 1.6586149011273422e-07, "loss": 0.0091, "step": 78030 }, { "grad_norm": 0.22969453036785126, "learning_rate": 1.6418366024562038e-07, "loss": 0.0149, "step": 78040 }, { "grad_norm": 0.14903338253498077, "learning_rate": 1.6251434595651037e-07, "loss": 0.01, "step": 78050 }, { "grad_norm": 0.18529759347438812, "learning_rate": 1.608535475306372e-07, "loss": 0.0084, "step": 78060 }, { "grad_norm": 0.294102281332016, "learning_rate": 1.5920126525179048e-07, "loss": 0.0089, "step": 78070 }, { "grad_norm": 0.20618890225887299, "learning_rate": 1.5755749940229435e-07, "loss": 0.0083, "step": 78080 }, { "grad_norm": 0.200266033411026, "learning_rate": 1.559222502630353e-07, "loss": 0.008, "step": 78090 }, { "grad_norm": 0.27655312418937683, "learning_rate": 1.5429551811341757e-07, "loss": 0.0104, "step": 78100 }, { "grad_norm": 0.17555318772792816, "learning_rate": 1.5267730323141882e-07, "loss": 0.0108, "step": 78110 }, { "grad_norm": 0.14612619578838348, "learning_rate": 1.5106760589353454e-07, "loss": 0.0063, "step": 78120 }, { "grad_norm": 0.14658662676811218, "learning_rate": 1.4946642637483355e-07, "loss": 0.009, "step": 78130 }, { "grad_norm": 0.17273983359336853, "learning_rate": 1.4787376494889703e-07, "loss": 0.0064, "step": 78140 }, { "grad_norm": 0.24358414113521576, "learning_rate": 1.4628962188787955e-07, "loss": 0.0082, "step": 78150 }, { "grad_norm": 0.19568254053592682, "learning_rate": 1.4471399746247006e-07, "loss": 0.0103, "step": 78160 }, { "grad_norm": 0.11384899169206619, "learning_rate": 1.4314689194188103e-07, "loss": 0.0099, "step": 78170 }, { "grad_norm": 0.3090202212333679, "learning_rate": 1.4158830559390933e-07, "loss": 0.0122, "step": 78180 }, { "grad_norm": 0.1563843935728073, "learning_rate": 1.4003823868486422e-07, "loss": 0.0198, "step": 78190 }, { "grad_norm": 0.1449790894985199, "learning_rate": 1.3849669147960598e-07, "loss": 0.0114, "step": 78200 }, { "grad_norm": 0.0879223495721817, "learning_rate": 1.3696366424155726e-07, "loss": 0.008, "step": 78210 }, { "grad_norm": 0.38029491901397705, "learning_rate": 1.35439157232653e-07, "loss": 0.0135, "step": 78220 }, { "grad_norm": 0.1611327975988388, "learning_rate": 1.3392317071340144e-07, "loss": 0.0085, "step": 78230 }, { "grad_norm": 0.20424652099609375, "learning_rate": 1.3241570494283984e-07, "loss": 0.0099, "step": 78240 }, { "grad_norm": 0.28529369831085205, "learning_rate": 1.3091676017855103e-07, "loss": 0.0131, "step": 78250 }, { "grad_norm": 0.1604568511247635, "learning_rate": 1.2942633667666904e-07, "loss": 0.0096, "step": 78260 }, { "grad_norm": 0.1991545557975769, "learning_rate": 1.2794443469185679e-07, "loss": 0.0117, "step": 78270 }, { "grad_norm": 0.15556684136390686, "learning_rate": 1.2647105447734508e-07, "loss": 0.0071, "step": 78280 }, { "grad_norm": 0.151603564620018, "learning_rate": 1.2500619628488254e-07, "loss": 0.0102, "step": 78290 }, { "grad_norm": 0.6584950685501099, "learning_rate": 1.2354986036477446e-07, "loss": 0.011, "step": 78300 }, { "grad_norm": 0.2770606279373169, "learning_rate": 1.221020469658718e-07, "loss": 0.0094, "step": 78310 }, { "grad_norm": 0.28825515508651733, "learning_rate": 1.2066275633556556e-07, "loss": 0.013, "step": 78320 }, { "grad_norm": 0.1601514220237732, "learning_rate": 1.192319887197979e-07, "loss": 0.0081, "step": 78330 }, { "grad_norm": 0.39880990982055664, "learning_rate": 1.1780974436303438e-07, "loss": 0.009, "step": 78340 }, { "grad_norm": 0.17104440927505493, "learning_rate": 1.1639602350830836e-07, "loss": 0.0102, "step": 78350 }, { "grad_norm": 0.16837053000926971, "learning_rate": 1.1499082639718217e-07, "loss": 0.0098, "step": 78360 }, { "grad_norm": 0.11006239056587219, "learning_rate": 1.1359415326976374e-07, "loss": 0.0072, "step": 78370 }, { "grad_norm": 0.10298112034797668, "learning_rate": 1.1220600436470663e-07, "loss": 0.0121, "step": 78380 }, { "grad_norm": 0.15371310710906982, "learning_rate": 1.1082637991920996e-07, "loss": 0.0099, "step": 78390 }, { "grad_norm": 0.18280307948589325, "learning_rate": 1.0945528016901851e-07, "loss": 0.01, "step": 78400 }, { "grad_norm": 0.32957836985588074, "learning_rate": 1.0809270534840599e-07, "loss": 0.009, "step": 78410 }, { "grad_norm": 0.09367434680461884, "learning_rate": 1.067386556902028e-07, "loss": 0.0104, "step": 78420 }, { "grad_norm": 0.2244032472372055, "learning_rate": 1.05393131425785e-07, "loss": 0.0074, "step": 78430 }, { "grad_norm": 0.257917582988739, "learning_rate": 1.0405613278505199e-07, "loss": 0.011, "step": 78440 }, { "grad_norm": 0.28454816341400146, "learning_rate": 1.027276599964766e-07, "loss": 0.0105, "step": 78450 }, { "grad_norm": 0.2926829159259796, "learning_rate": 1.0140771328704391e-07, "loss": 0.0098, "step": 78460 }, { "grad_norm": 0.09278242290019989, "learning_rate": 1.0009629288231237e-07, "loss": 0.0075, "step": 78470 }, { "grad_norm": 0.1347128450870514, "learning_rate": 9.879339900635276e-08, "loss": 0.012, "step": 78480 }, { "grad_norm": 0.2912716865539551, "learning_rate": 9.74990318817981e-08, "loss": 0.0095, "step": 78490 }, { "grad_norm": 0.16693753004074097, "learning_rate": 9.621319172982701e-08, "loss": 0.0085, "step": 78500 }, { "grad_norm": 0.0701625719666481, "learning_rate": 9.493587877015264e-08, "loss": 0.0076, "step": 78510 }, { "grad_norm": 0.23131179809570312, "learning_rate": 9.366709322102263e-08, "loss": 0.0077, "step": 78520 }, { "grad_norm": 0.30639469623565674, "learning_rate": 9.240683529924688e-08, "loss": 0.0078, "step": 78530 }, { "grad_norm": 0.25254783034324646, "learning_rate": 9.115510522016979e-08, "loss": 0.0083, "step": 78540 }, { "grad_norm": 0.1800266057252884, "learning_rate": 8.991190319767583e-08, "loss": 0.0089, "step": 78550 }, { "grad_norm": 0.15760977566242218, "learning_rate": 8.867722944419509e-08, "loss": 0.0082, "step": 78560 }, { "grad_norm": 0.11783663183450699, "learning_rate": 8.745108417069214e-08, "loss": 0.006, "step": 78570 }, { "grad_norm": 0.14554068446159363, "learning_rate": 8.623346758669381e-08, "loss": 0.0059, "step": 78580 }, { "grad_norm": 0.16812944412231445, "learning_rate": 8.502437990025037e-08, "loss": 0.0091, "step": 78590 }, { "grad_norm": 0.12586066126823425, "learning_rate": 8.382382131795764e-08, "loss": 0.0088, "step": 78600 }, { "grad_norm": 0.1257195621728897, "learning_rate": 8.263179204496818e-08, "loss": 0.0102, "step": 78610 }, { "grad_norm": 0.2438051551580429, "learning_rate": 8.144829228496354e-08, "loss": 0.0127, "step": 78620 }, { "grad_norm": 0.2155696600675583, "learning_rate": 8.027332224016526e-08, "loss": 0.0091, "step": 78630 }, { "grad_norm": 0.1901504099369049, "learning_rate": 7.910688211135164e-08, "loss": 0.0106, "step": 78640 }, { "grad_norm": 0.14654377102851868, "learning_rate": 7.794897209783546e-08, "loss": 0.0114, "step": 78650 }, { "grad_norm": 0.10008706152439117, "learning_rate": 7.679959239746403e-08, "loss": 0.0103, "step": 78660 }, { "grad_norm": 0.09494451433420181, "learning_rate": 7.565874320664135e-08, "loss": 0.0105, "step": 78670 }, { "grad_norm": 0.2057863473892212, "learning_rate": 7.45264247203059e-08, "loss": 0.0085, "step": 78680 }, { "grad_norm": 0.10725869238376617, "learning_rate": 7.340263713194184e-08, "loss": 0.0054, "step": 78690 }, { "grad_norm": 0.19257205724716187, "learning_rate": 7.228738063356777e-08, "loss": 0.0117, "step": 78700 }, { "grad_norm": 0.5127773880958557, "learning_rate": 7.118065541575903e-08, "loss": 0.0129, "step": 78710 }, { "grad_norm": 0.1487509161233902, "learning_rate": 7.008246166761435e-08, "loss": 0.0075, "step": 78720 }, { "grad_norm": 0.1450805515050888, "learning_rate": 6.899279957679472e-08, "loss": 0.0102, "step": 78730 }, { "grad_norm": 0.1365513801574707, "learning_rate": 6.791166932949011e-08, "loss": 0.0069, "step": 78740 }, { "grad_norm": 0.2800103724002838, "learning_rate": 6.683907111043608e-08, "loss": 0.0083, "step": 78750 }, { "grad_norm": 0.08419058471918106, "learning_rate": 6.577500510290824e-08, "loss": 0.0087, "step": 78760 }, { "grad_norm": 0.17830297350883484, "learning_rate": 6.47194714887278e-08, "loss": 0.0183, "step": 78770 }, { "grad_norm": 0.10442820191383362, "learning_rate": 6.367247044825608e-08, "loss": 0.0088, "step": 78780 }, { "grad_norm": 0.18202830851078033, "learning_rate": 6.263400216039994e-08, "loss": 0.0096, "step": 78790 }, { "grad_norm": 0.28057464957237244, "learning_rate": 6.160406680260078e-08, "loss": 0.0181, "step": 78800 }, { "grad_norm": 0.05186135321855545, "learning_rate": 6.058266455084561e-08, "loss": 0.009, "step": 78810 }, { "grad_norm": 0.13239459693431854, "learning_rate": 5.956979557967257e-08, "loss": 0.0089, "step": 78820 }, { "grad_norm": 0.15836374461650848, "learning_rate": 5.856546006214325e-08, "loss": 0.0096, "step": 78830 }, { "grad_norm": 0.14753998816013336, "learning_rate": 5.7569658169881466e-08, "loss": 0.0066, "step": 78840 }, { "grad_norm": 0.1426624357700348, "learning_rate": 5.658239007303445e-08, "loss": 0.0125, "step": 78850 }, { "grad_norm": 0.20587749779224396, "learning_rate": 5.560365594030059e-08, "loss": 0.0084, "step": 78860 }, { "grad_norm": 0.1287745237350464, "learning_rate": 5.463345593891833e-08, "loss": 0.0071, "step": 78870 }, { "grad_norm": 0.2519264817237854, "learning_rate": 5.367179023467173e-08, "loss": 0.0105, "step": 78880 }, { "grad_norm": 0.16472579538822174, "learning_rate": 5.271865899187378e-08, "loss": 0.0078, "step": 78890 }, { "grad_norm": 0.25208011269569397, "learning_rate": 5.177406237340532e-08, "loss": 0.0238, "step": 78900 }, { "grad_norm": 0.16394075751304626, "learning_rate": 5.083800054065946e-08, "loss": 0.007, "step": 78910 }, { "grad_norm": 0.14067856967449188, "learning_rate": 4.991047365358603e-08, "loss": 0.0089, "step": 78920 }, { "grad_norm": 0.4667564332485199, "learning_rate": 4.899148187067493e-08, "loss": 0.0157, "step": 78930 }, { "grad_norm": 0.09814093261957169, "learning_rate": 4.808102534895609e-08, "loss": 0.0073, "step": 78940 }, { "grad_norm": 0.13633094727993011, "learning_rate": 4.717910424400507e-08, "loss": 0.0085, "step": 78950 }, { "grad_norm": 0.19923843443393707, "learning_rate": 4.628571870993193e-08, "loss": 0.0098, "step": 78960 }, { "grad_norm": 0.20703797042369843, "learning_rate": 4.5400868899392324e-08, "loss": 0.0117, "step": 78970 }, { "grad_norm": 0.27748537063598633, "learning_rate": 4.4524554963576436e-08, "loss": 0.0134, "step": 78980 }, { "grad_norm": 0.21810927987098694, "learning_rate": 4.36567770522367e-08, "loss": 0.0084, "step": 78990 }, { "grad_norm": 0.09206213057041168, "learning_rate": 4.279753531364339e-08, "loss": 0.0073, "step": 79000 }, { "grad_norm": 0.2272140234708786, "learning_rate": 4.1946829894617955e-08, "loss": 0.0112, "step": 79010 }, { "grad_norm": 0.10980173945426941, "learning_rate": 4.11046609405219e-08, "loss": 0.0138, "step": 79020 }, { "grad_norm": 0.2142428457736969, "learning_rate": 4.027102859526233e-08, "loss": 0.0067, "step": 79030 }, { "grad_norm": 0.29263055324554443, "learning_rate": 3.9445933001280856e-08, "loss": 0.0107, "step": 79040 }, { "grad_norm": 0.16094747185707092, "learning_rate": 3.8629374299564704e-08, "loss": 0.0119, "step": 79050 }, { "grad_norm": 0.15971282124519348, "learning_rate": 3.78213526296467e-08, "loss": 0.0108, "step": 79060 }, { "grad_norm": 0.11346074193716049, "learning_rate": 3.702186812958308e-08, "loss": 0.0103, "step": 79070 }, { "grad_norm": 0.22934618592262268, "learning_rate": 3.623092093599789e-08, "loss": 0.0121, "step": 79080 }, { "grad_norm": 0.21652629971504211, "learning_rate": 3.544851118403303e-08, "loss": 0.011, "step": 79090 }, { "grad_norm": 0.13263852894306183, "learning_rate": 3.46746390073871e-08, "loss": 0.014, "step": 79100 }, { "grad_norm": 0.25020262598991394, "learning_rate": 3.390930453829322e-08, "loss": 0.0155, "step": 79110 }, { "grad_norm": 0.16334418952465057, "learning_rate": 3.3152507907519005e-08, "loss": 0.0071, "step": 79120 }, { "grad_norm": 0.13063983619213104, "learning_rate": 3.240424924438323e-08, "loss": 0.0079, "step": 79130 }, { "grad_norm": 0.1423242688179016, "learning_rate": 3.1664528676750295e-08, "loss": 0.0103, "step": 79140 }, { "grad_norm": 0.1570015549659729, "learning_rate": 3.093334633100797e-08, "loss": 0.0177, "step": 79150 }, { "grad_norm": 0.219749316573143, "learning_rate": 3.021070233210077e-08, "loss": 0.0104, "step": 79160 }, { "grad_norm": 0.19603140652179718, "learning_rate": 2.9496596803507692e-08, "loss": 0.0075, "step": 79170 }, { "grad_norm": 0.16455243527889252, "learning_rate": 2.879102986725335e-08, "loss": 0.0089, "step": 79180 }, { "grad_norm": 0.13128884136676788, "learning_rate": 2.809400164389131e-08, "loss": 0.0088, "step": 79190 }, { "grad_norm": 0.1558995246887207, "learning_rate": 2.740551225253185e-08, "loss": 0.0102, "step": 79200 }, { "grad_norm": 0.1739976406097412, "learning_rate": 2.6725561810819754e-08, "loss": 0.0108, "step": 79210 }, { "grad_norm": 0.21300072968006134, "learning_rate": 2.605415043493431e-08, "loss": 0.0083, "step": 79220 }, { "grad_norm": 0.16588528454303741, "learning_rate": 2.5391278239605965e-08, "loss": 0.008, "step": 79230 }, { "grad_norm": 0.21829907596111298, "learning_rate": 2.4736945338094118e-08, "loss": 0.0086, "step": 79240 }, { "grad_norm": 0.1579282581806183, "learning_rate": 2.4091151842214887e-08, "loss": 0.0081, "step": 79250 }, { "grad_norm": 0.17694567143917084, "learning_rate": 2.3453897862318884e-08, "loss": 0.0118, "step": 79260 }, { "grad_norm": 0.3300510048866272, "learning_rate": 2.2825183507285686e-08, "loss": 0.0084, "step": 79270 }, { "grad_norm": 0.4981583058834076, "learning_rate": 2.220500888455157e-08, "loss": 0.0137, "step": 79280 }, { "grad_norm": 0.23821434378623962, "learning_rate": 2.1593374100081777e-08, "loss": 0.013, "step": 79290 }, { "grad_norm": 0.2096848338842392, "learning_rate": 2.09902792583927e-08, "loss": 0.0139, "step": 79300 }, { "grad_norm": 0.1297718733549118, "learning_rate": 2.0395724462540788e-08, "loss": 0.0085, "step": 79310 }, { "grad_norm": 0.21177923679351807, "learning_rate": 1.9809709814111453e-08, "loss": 0.0111, "step": 79320 }, { "grad_norm": 0.12739942967891693, "learning_rate": 1.923223541324126e-08, "loss": 0.0115, "step": 79330 }, { "grad_norm": 0.14658868312835693, "learning_rate": 1.8663301358606833e-08, "loss": 0.0136, "step": 79340 }, { "grad_norm": 0.1291198432445526, "learning_rate": 1.8102907747419295e-08, "loss": 0.0104, "step": 79350 }, { "grad_norm": 0.15811465680599213, "learning_rate": 1.7551054675435385e-08, "loss": 0.0104, "step": 79360 }, { "grad_norm": 0.10529165714979172, "learning_rate": 1.7007742236957447e-08, "loss": 0.0098, "step": 79370 }, { "grad_norm": 0.21440444886684418, "learning_rate": 1.647297052481678e-08, "loss": 0.0093, "step": 79380 }, { "grad_norm": 0.1859702616930008, "learning_rate": 1.5946739630390285e-08, "loss": 0.0137, "step": 79390 }, { "grad_norm": 0.1440856009721756, "learning_rate": 1.5429049643606032e-08, "loss": 0.0139, "step": 79400 }, { "grad_norm": 0.11843102425336838, "learning_rate": 1.4919900652909935e-08, "loss": 0.0097, "step": 79410 }, { "grad_norm": 0.14646147191524506, "learning_rate": 1.4419292745310175e-08, "loss": 0.0113, "step": 79420 }, { "grad_norm": 0.11540425568819046, "learning_rate": 1.3927226006343885e-08, "loss": 0.0103, "step": 79430 }, { "grad_norm": 0.1573750227689743, "learning_rate": 1.3443700520093805e-08, "loss": 0.0106, "step": 79440 }, { "grad_norm": 0.13177646696567535, "learning_rate": 1.296871636917718e-08, "loss": 0.0085, "step": 79450 }, { "grad_norm": 0.16434776782989502, "learning_rate": 1.2502273634762419e-08, "loss": 0.0084, "step": 79460 }, { "grad_norm": 0.32705339789390564, "learning_rate": 1.2044372396546876e-08, "loss": 0.0119, "step": 79470 }, { "grad_norm": 0.19668833911418915, "learning_rate": 1.1595012732773524e-08, "loss": 0.0092, "step": 79480 }, { "grad_norm": 0.14923326671123505, "learning_rate": 1.1154194720225386e-08, "loss": 0.0116, "step": 79490 }, { "grad_norm": 0.19615015387535095, "learning_rate": 1.0721918434231093e-08, "loss": 0.0096, "step": 79500 }, { "grad_norm": 0.13907364010810852, "learning_rate": 1.0298183948648232e-08, "loss": 0.0053, "step": 79510 }, { "grad_norm": 0.07273425906896591, "learning_rate": 9.88299133588555e-09, "loss": 0.0081, "step": 79520 }, { "grad_norm": 0.19263365864753723, "learning_rate": 9.476340666891847e-09, "loss": 0.0166, "step": 79530 }, { "grad_norm": 0.09952722489833832, "learning_rate": 9.078232011139332e-09, "loss": 0.0069, "step": 79540 }, { "grad_norm": 0.19179090857505798, "learning_rate": 8.68866543666802e-09, "loss": 0.0083, "step": 79550 }, { "grad_norm": 0.0710211917757988, "learning_rate": 8.307641010035782e-09, "loss": 0.0104, "step": 79560 }, { "grad_norm": 0.47984033823013306, "learning_rate": 7.93515879635165e-09, "loss": 0.0109, "step": 79570 }, { "grad_norm": 0.15848201513290405, "learning_rate": 7.571218859264706e-09, "loss": 0.008, "step": 79580 }, { "grad_norm": 0.2287750095129013, "learning_rate": 7.215821260958544e-09, "loss": 0.01, "step": 79590 }, { "grad_norm": 0.16934925317764282, "learning_rate": 6.868966062162363e-09, "loss": 0.0088, "step": 79600 }, { "grad_norm": 0.10019632428884506, "learning_rate": 6.530653322145419e-09, "loss": 0.0065, "step": 79610 }, { "grad_norm": 0.16408562660217285, "learning_rate": 6.200883098717025e-09, "loss": 0.0069, "step": 79620 }, { "grad_norm": 0.30422112345695496, "learning_rate": 5.879655448226551e-09, "loss": 0.0114, "step": 79630 }, { "grad_norm": 0.2017584890127182, "learning_rate": 5.566970425557872e-09, "loss": 0.0075, "step": 79640 }, { "grad_norm": 0.2535766661167145, "learning_rate": 5.26282808414047e-09, "loss": 0.0081, "step": 79650 }, { "grad_norm": 0.20477551221847534, "learning_rate": 4.967228475949437e-09, "loss": 0.0106, "step": 79660 }, { "grad_norm": 0.19289113581180573, "learning_rate": 4.680171651494369e-09, "loss": 0.0091, "step": 79670 }, { "grad_norm": 0.18253271281719208, "learning_rate": 4.40165765981937e-09, "loss": 0.0064, "step": 79680 }, { "grad_norm": 0.18312323093414307, "learning_rate": 4.131686548519698e-09, "loss": 0.0074, "step": 79690 }, { "grad_norm": 0.15302899479866028, "learning_rate": 3.8702583637251214e-09, "loss": 0.0105, "step": 79700 }, { "grad_norm": 0.07369186729192734, "learning_rate": 3.617373150105463e-09, "loss": 0.0076, "step": 79710 }, { "grad_norm": 0.36193203926086426, "learning_rate": 3.3730309508706036e-09, "loss": 0.0057, "step": 79720 }, { "grad_norm": 0.150720477104187, "learning_rate": 3.137231807781582e-09, "loss": 0.0073, "step": 79730 }, { "grad_norm": 0.17718198895454407, "learning_rate": 2.9099757611172894e-09, "loss": 0.0109, "step": 79740 }, { "grad_norm": 0.15957589447498322, "learning_rate": 2.6912628497133275e-09, "loss": 0.0099, "step": 79750 }, { "grad_norm": 0.2941063940525055, "learning_rate": 2.481093110945354e-09, "loss": 0.0111, "step": 79760 }, { "grad_norm": 0.12497548013925552, "learning_rate": 2.2794665807235327e-09, "loss": 0.0097, "step": 79770 }, { "grad_norm": 0.15629276633262634, "learning_rate": 2.0863832934980843e-09, "loss": 0.0079, "step": 79780 }, { "grad_norm": 0.1345752775669098, "learning_rate": 1.901843282264837e-09, "loss": 0.0095, "step": 79790 }, { "grad_norm": 1.2661168575286865, "learning_rate": 1.7258465785541245e-09, "loss": 0.0137, "step": 79800 }, { "grad_norm": 0.12577678263187408, "learning_rate": 1.5583932124418887e-09, "loss": 0.0093, "step": 79810 }, { "grad_norm": 0.09926698356866837, "learning_rate": 1.3994832125441282e-09, "loss": 0.0096, "step": 79820 }, { "grad_norm": 0.17839670181274414, "learning_rate": 1.2491166060057958e-09, "loss": 0.0114, "step": 79830 }, { "grad_norm": 0.18741963803768158, "learning_rate": 1.1072934185230032e-09, "loss": 0.0114, "step": 79840 }, { "grad_norm": 0.4380391836166382, "learning_rate": 9.740136743319195e-10, "loss": 0.0116, "step": 79850 }, { "grad_norm": 0.2054820954799652, "learning_rate": 8.492773962087696e-10, "loss": 0.0092, "step": 79860 }, { "grad_norm": 0.37495845556259155, "learning_rate": 7.330846054587337e-10, "loss": 0.0115, "step": 79870 }, { "grad_norm": 0.18735256791114807, "learning_rate": 6.254353219492526e-10, "loss": 0.0097, "step": 79880 }, { "grad_norm": 0.2598482370376587, "learning_rate": 5.263295640600685e-10, "loss": 0.0097, "step": 79890 }, { "grad_norm": 0.2038344442844391, "learning_rate": 4.357673487387359e-10, "loss": 0.0116, "step": 79900 }, { "grad_norm": 0.24011610448360443, "learning_rate": 3.537486914506616e-10, "loss": 0.0103, "step": 79910 }, { "grad_norm": 0.3294242322444916, "learning_rate": 2.8027360621241117e-10, "loss": 0.0107, "step": 79920 }, { "grad_norm": 0.18765851855278015, "learning_rate": 2.153421055806071e-10, "loss": 0.0096, "step": 79930 }, { "grad_norm": 0.13247494399547577, "learning_rate": 1.5895420064637733e-10, "loss": 0.0101, "step": 79940 }, { "grad_norm": 0.13410957157611847, "learning_rate": 1.1110990105200891e-10, "loss": 0.0091, "step": 79950 }, { "grad_norm": 0.26782897114753723, "learning_rate": 7.180921496874326e-11, "loss": 0.0122, "step": 79960 }, { "grad_norm": 0.1944667398929596, "learning_rate": 4.10521491134297e-11, "loss": 0.0083, "step": 79970 }, { "grad_norm": 0.26473742723464966, "learning_rate": 1.883870874297422e-11, "loss": 0.0093, "step": 79980 }, { "grad_norm": 0.2509427070617676, "learning_rate": 5.1688976432373584e-12, "loss": 0.0102, "step": 79990 }, { "grad_norm": 0.15460950136184692, "learning_rate": 4.271816234080461e-14, "loss": 0.0114, "step": 80000 } ], "logging_steps": 10, "max_steps": 80000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 20000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }