{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 40000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.5917308926582336, "learning_rate": 4.5e-07, "loss": 1.2934, "step": 10 }, { "grad_norm": 0.5956218242645264, "learning_rate": 9.5e-07, "loss": 1.302, "step": 20 }, { "grad_norm": 0.5066128969192505, "learning_rate": 1.45e-06, "loss": 1.2878, "step": 30 }, { "grad_norm": 0.6079267859458923, "learning_rate": 1.95e-06, "loss": 1.2407, "step": 40 }, { "grad_norm": 0.35770943760871887, "learning_rate": 2.4500000000000003e-06, "loss": 1.2136, "step": 50 }, { "grad_norm": 0.40626829862594604, "learning_rate": 2.95e-06, "loss": 1.192, "step": 60 }, { "grad_norm": 0.4430682063102722, "learning_rate": 3.4500000000000004e-06, "loss": 1.1631, "step": 70 }, { "grad_norm": 0.4442620277404785, "learning_rate": 3.95e-06, "loss": 1.1396, "step": 80 }, { "grad_norm": 0.5642955303192139, "learning_rate": 4.45e-06, "loss": 1.1118, "step": 90 }, { "grad_norm": 0.47835057973861694, "learning_rate": 4.950000000000001e-06, "loss": 1.0963, "step": 100 }, { "grad_norm": 0.48458200693130493, "learning_rate": 5.45e-06, "loss": 1.0772, "step": 110 }, { "grad_norm": 0.5455941557884216, "learning_rate": 5.95e-06, "loss": 1.0762, "step": 120 }, { "grad_norm": 0.4653521180152893, "learning_rate": 6.45e-06, "loss": 1.074, "step": 130 }, { "grad_norm": 0.49748000502586365, "learning_rate": 6.950000000000001e-06, "loss": 1.0741, "step": 140 }, { "grad_norm": 0.5504190325737, "learning_rate": 7.45e-06, "loss": 1.0621, "step": 150 }, { "grad_norm": 0.552014946937561, "learning_rate": 7.95e-06, "loss": 1.0546, "step": 160 }, { "grad_norm": 0.7554380297660828, "learning_rate": 8.45e-06, "loss": 1.067, "step": 170 }, { "grad_norm": 1.1445671319961548, "learning_rate": 8.95e-06, "loss": 1.0303, "step": 180 }, { "grad_norm": 1.1424305438995361, "learning_rate": 9.450000000000001e-06, "loss": 1.0002, "step": 190 }, { "grad_norm": 0.9780583381652832, "learning_rate": 9.950000000000001e-06, "loss": 0.957, "step": 200 }, { "grad_norm": 1.2749563455581665, "learning_rate": 1.045e-05, "loss": 0.9144, "step": 210 }, { "grad_norm": 1.296170949935913, "learning_rate": 1.095e-05, "loss": 0.8653, "step": 220 }, { "grad_norm": 1.3219212293624878, "learning_rate": 1.145e-05, "loss": 0.7842, "step": 230 }, { "grad_norm": 1.4269319772720337, "learning_rate": 1.195e-05, "loss": 0.7355, "step": 240 }, { "grad_norm": 1.3941267728805542, "learning_rate": 1.2450000000000001e-05, "loss": 0.6891, "step": 250 }, { "grad_norm": 1.7447683811187744, "learning_rate": 1.2950000000000001e-05, "loss": 0.6479, "step": 260 }, { "grad_norm": 1.6563198566436768, "learning_rate": 1.3450000000000002e-05, "loss": 0.6014, "step": 270 }, { "grad_norm": 1.5465058088302612, "learning_rate": 1.3950000000000002e-05, "loss": 0.5615, "step": 280 }, { "grad_norm": 1.34519362449646, "learning_rate": 1.4449999999999999e-05, "loss": 0.5205, "step": 290 }, { "grad_norm": 1.6614229679107666, "learning_rate": 1.4950000000000001e-05, "loss": 0.4831, "step": 300 }, { "grad_norm": 1.8632866144180298, "learning_rate": 1.545e-05, "loss": 0.4476, "step": 310 }, { "grad_norm": 1.7526261806488037, "learning_rate": 1.595e-05, "loss": 0.3964, "step": 320 }, { "grad_norm": 1.5427043437957764, "learning_rate": 1.645e-05, "loss": 0.357, "step": 330 }, { "grad_norm": 1.7545642852783203, "learning_rate": 1.6950000000000002e-05, "loss": 0.3328, "step": 340 }, { "grad_norm": 1.8075200319290161, "learning_rate": 1.745e-05, "loss": 0.3203, "step": 350 }, { "grad_norm": 1.8086761236190796, "learning_rate": 1.795e-05, "loss": 0.3112, "step": 360 }, { "grad_norm": 1.4851826429367065, "learning_rate": 1.845e-05, "loss": 0.2961, "step": 370 }, { "grad_norm": 1.90358567237854, "learning_rate": 1.895e-05, "loss": 0.2777, "step": 380 }, { "grad_norm": 1.6338852643966675, "learning_rate": 1.9450000000000002e-05, "loss": 0.2705, "step": 390 }, { "grad_norm": 1.4127023220062256, "learning_rate": 1.995e-05, "loss": 0.255, "step": 400 }, { "grad_norm": 1.416651964187622, "learning_rate": 2.045e-05, "loss": 0.2467, "step": 410 }, { "grad_norm": 1.7859480381011963, "learning_rate": 2.095e-05, "loss": 0.2197, "step": 420 }, { "grad_norm": 2.083770751953125, "learning_rate": 2.145e-05, "loss": 0.2062, "step": 430 }, { "grad_norm": 1.9129972457885742, "learning_rate": 2.195e-05, "loss": 0.1921, "step": 440 }, { "grad_norm": 1.9008049964904785, "learning_rate": 2.245e-05, "loss": 0.1852, "step": 450 }, { "grad_norm": 1.6429637670516968, "learning_rate": 2.2950000000000002e-05, "loss": 0.1648, "step": 460 }, { "grad_norm": 1.797263264656067, "learning_rate": 2.345e-05, "loss": 0.1652, "step": 470 }, { "grad_norm": 1.5145591497421265, "learning_rate": 2.395e-05, "loss": 0.153, "step": 480 }, { "grad_norm": 1.593704104423523, "learning_rate": 2.445e-05, "loss": 0.1379, "step": 490 }, { "grad_norm": 1.8096568584442139, "learning_rate": 2.495e-05, "loss": 0.138, "step": 500 }, { "grad_norm": 1.6113165616989136, "learning_rate": 2.5450000000000002e-05, "loss": 0.1355, "step": 510 }, { "grad_norm": 1.3228784799575806, "learning_rate": 2.595e-05, "loss": 0.1283, "step": 520 }, { "grad_norm": 1.5226502418518066, "learning_rate": 2.6450000000000003e-05, "loss": 0.129, "step": 530 }, { "grad_norm": 1.7247620820999146, "learning_rate": 2.6950000000000005e-05, "loss": 0.1291, "step": 540 }, { "grad_norm": 1.562394142150879, "learning_rate": 2.7450000000000003e-05, "loss": 0.1238, "step": 550 }, { "grad_norm": 1.4812381267547607, "learning_rate": 2.7950000000000005e-05, "loss": 0.1253, "step": 560 }, { "grad_norm": 1.6975467205047607, "learning_rate": 2.845e-05, "loss": 0.1297, "step": 570 }, { "grad_norm": 1.819911003112793, "learning_rate": 2.895e-05, "loss": 0.132, "step": 580 }, { "grad_norm": 1.6957908868789673, "learning_rate": 2.945e-05, "loss": 0.1182, "step": 590 }, { "grad_norm": 1.2963670492172241, "learning_rate": 2.995e-05, "loss": 0.1217, "step": 600 }, { "grad_norm": 1.4479252099990845, "learning_rate": 3.045e-05, "loss": 0.1154, "step": 610 }, { "grad_norm": 1.4544728994369507, "learning_rate": 3.095e-05, "loss": 0.1187, "step": 620 }, { "grad_norm": 1.4913729429244995, "learning_rate": 3.145e-05, "loss": 0.1108, "step": 630 }, { "grad_norm": 1.678189992904663, "learning_rate": 3.1950000000000004e-05, "loss": 0.1083, "step": 640 }, { "grad_norm": 1.6236698627471924, "learning_rate": 3.245e-05, "loss": 0.1036, "step": 650 }, { "grad_norm": 1.2770849466323853, "learning_rate": 3.295e-05, "loss": 0.1167, "step": 660 }, { "grad_norm": 1.4118672609329224, "learning_rate": 3.345000000000001e-05, "loss": 0.1086, "step": 670 }, { "grad_norm": 1.441764235496521, "learning_rate": 3.3950000000000005e-05, "loss": 0.1041, "step": 680 }, { "grad_norm": 1.472336769104004, "learning_rate": 3.445e-05, "loss": 0.1038, "step": 690 }, { "grad_norm": 1.42439866065979, "learning_rate": 3.495e-05, "loss": 0.0993, "step": 700 }, { "grad_norm": 2.0421323776245117, "learning_rate": 3.545e-05, "loss": 0.1184, "step": 710 }, { "grad_norm": 1.3984744548797607, "learning_rate": 3.595e-05, "loss": 0.108, "step": 720 }, { "grad_norm": 1.4885045289993286, "learning_rate": 3.645e-05, "loss": 0.0931, "step": 730 }, { "grad_norm": 1.3463157415390015, "learning_rate": 3.6950000000000004e-05, "loss": 0.1053, "step": 740 }, { "grad_norm": 1.3699471950531006, "learning_rate": 3.745e-05, "loss": 0.1055, "step": 750 }, { "grad_norm": 1.5465822219848633, "learning_rate": 3.795e-05, "loss": 0.112, "step": 760 }, { "grad_norm": 1.5660563707351685, "learning_rate": 3.845e-05, "loss": 0.1009, "step": 770 }, { "grad_norm": 1.2822130918502808, "learning_rate": 3.8950000000000005e-05, "loss": 0.1017, "step": 780 }, { "grad_norm": 1.368312954902649, "learning_rate": 3.9450000000000003e-05, "loss": 0.101, "step": 790 }, { "grad_norm": 1.3434205055236816, "learning_rate": 3.995e-05, "loss": 0.1008, "step": 800 }, { "grad_norm": 1.1067501306533813, "learning_rate": 4.045000000000001e-05, "loss": 0.097, "step": 810 }, { "grad_norm": 1.519289493560791, "learning_rate": 4.095e-05, "loss": 0.101, "step": 820 }, { "grad_norm": 1.354209542274475, "learning_rate": 4.145e-05, "loss": 0.1003, "step": 830 }, { "grad_norm": 1.508072018623352, "learning_rate": 4.195e-05, "loss": 0.0992, "step": 840 }, { "grad_norm": 1.2704591751098633, "learning_rate": 4.245e-05, "loss": 0.1033, "step": 850 }, { "grad_norm": 1.4619317054748535, "learning_rate": 4.295e-05, "loss": 0.104, "step": 860 }, { "grad_norm": 1.2592161893844604, "learning_rate": 4.345e-05, "loss": 0.098, "step": 870 }, { "grad_norm": 1.5460516214370728, "learning_rate": 4.3950000000000004e-05, "loss": 0.1013, "step": 880 }, { "grad_norm": 1.3333797454833984, "learning_rate": 4.445e-05, "loss": 0.0999, "step": 890 }, { "grad_norm": 1.5117151737213135, "learning_rate": 4.495e-05, "loss": 0.0934, "step": 900 }, { "grad_norm": 1.5144648551940918, "learning_rate": 4.545000000000001e-05, "loss": 0.0951, "step": 910 }, { "grad_norm": 1.2717652320861816, "learning_rate": 4.5950000000000006e-05, "loss": 0.0888, "step": 920 }, { "grad_norm": 1.139288067817688, "learning_rate": 4.6450000000000004e-05, "loss": 0.0921, "step": 930 }, { "grad_norm": 1.526144027709961, "learning_rate": 4.695e-05, "loss": 0.087, "step": 940 }, { "grad_norm": 1.0298430919647217, "learning_rate": 4.745e-05, "loss": 0.096, "step": 950 }, { "grad_norm": 1.198772668838501, "learning_rate": 4.795e-05, "loss": 0.0891, "step": 960 }, { "grad_norm": 1.1340601444244385, "learning_rate": 4.845e-05, "loss": 0.0905, "step": 970 }, { "grad_norm": 1.3683668375015259, "learning_rate": 4.8950000000000004e-05, "loss": 0.0939, "step": 980 }, { "grad_norm": 1.563915491104126, "learning_rate": 4.945e-05, "loss": 0.095, "step": 990 }, { "grad_norm": 1.6076679229736328, "learning_rate": 4.995e-05, "loss": 0.0974, "step": 1000 }, { "grad_norm": 0.9983281493186951, "learning_rate": 5.045e-05, "loss": 0.09, "step": 1010 }, { "grad_norm": 1.2368406057357788, "learning_rate": 5.095e-05, "loss": 0.0923, "step": 1020 }, { "grad_norm": 1.1224356889724731, "learning_rate": 5.145e-05, "loss": 0.096, "step": 1030 }, { "grad_norm": 1.1620182991027832, "learning_rate": 5.1949999999999996e-05, "loss": 0.0858, "step": 1040 }, { "grad_norm": 0.8956158757209778, "learning_rate": 5.245e-05, "loss": 0.0842, "step": 1050 }, { "grad_norm": 0.9405304193496704, "learning_rate": 5.295e-05, "loss": 0.0854, "step": 1060 }, { "grad_norm": 1.4309135675430298, "learning_rate": 5.345e-05, "loss": 0.0967, "step": 1070 }, { "grad_norm": 1.2947970628738403, "learning_rate": 5.3950000000000004e-05, "loss": 0.0872, "step": 1080 }, { "grad_norm": 1.1429096460342407, "learning_rate": 5.445e-05, "loss": 0.0874, "step": 1090 }, { "grad_norm": 1.197623610496521, "learning_rate": 5.495e-05, "loss": 0.0962, "step": 1100 }, { "grad_norm": 1.0402482748031616, "learning_rate": 5.545e-05, "loss": 0.0807, "step": 1110 }, { "grad_norm": 0.915535032749176, "learning_rate": 5.5950000000000005e-05, "loss": 0.0878, "step": 1120 }, { "grad_norm": 1.352502703666687, "learning_rate": 5.645e-05, "loss": 0.0924, "step": 1130 }, { "grad_norm": 1.3966892957687378, "learning_rate": 5.695e-05, "loss": 0.0909, "step": 1140 }, { "grad_norm": 1.1338180303573608, "learning_rate": 5.745e-05, "loss": 0.089, "step": 1150 }, { "grad_norm": 0.7804926633834839, "learning_rate": 5.7950000000000006e-05, "loss": 0.0834, "step": 1160 }, { "grad_norm": 1.0618555545806885, "learning_rate": 5.8450000000000005e-05, "loss": 0.0838, "step": 1170 }, { "grad_norm": 1.1040382385253906, "learning_rate": 5.895e-05, "loss": 0.0861, "step": 1180 }, { "grad_norm": 1.023377537727356, "learning_rate": 5.945000000000001e-05, "loss": 0.0851, "step": 1190 }, { "grad_norm": 0.9269119501113892, "learning_rate": 5.995000000000001e-05, "loss": 0.0853, "step": 1200 }, { "grad_norm": 0.9269769191741943, "learning_rate": 6.0450000000000006e-05, "loss": 0.0786, "step": 1210 }, { "grad_norm": 1.2504351139068604, "learning_rate": 6.0950000000000004e-05, "loss": 0.0723, "step": 1220 }, { "grad_norm": 1.1691515445709229, "learning_rate": 6.145e-05, "loss": 0.0754, "step": 1230 }, { "grad_norm": 1.1351704597473145, "learning_rate": 6.195e-05, "loss": 0.0796, "step": 1240 }, { "grad_norm": 0.9207521080970764, "learning_rate": 6.245000000000001e-05, "loss": 0.0806, "step": 1250 }, { "grad_norm": 1.0673400163650513, "learning_rate": 6.295e-05, "loss": 0.0719, "step": 1260 }, { "grad_norm": 0.8698428869247437, "learning_rate": 6.345e-05, "loss": 0.0889, "step": 1270 }, { "grad_norm": 1.176358938217163, "learning_rate": 6.395e-05, "loss": 0.0816, "step": 1280 }, { "grad_norm": 1.0234766006469727, "learning_rate": 6.445e-05, "loss": 0.0824, "step": 1290 }, { "grad_norm": 0.9029611349105835, "learning_rate": 6.494999999999999e-05, "loss": 0.0801, "step": 1300 }, { "grad_norm": 1.2201935052871704, "learning_rate": 6.545e-05, "loss": 0.0751, "step": 1310 }, { "grad_norm": 1.1465617418289185, "learning_rate": 6.595e-05, "loss": 0.0943, "step": 1320 }, { "grad_norm": 1.0688234567642212, "learning_rate": 6.645e-05, "loss": 0.0813, "step": 1330 }, { "grad_norm": 1.0792423486709595, "learning_rate": 6.695e-05, "loss": 0.0817, "step": 1340 }, { "grad_norm": 0.8873389959335327, "learning_rate": 6.745e-05, "loss": 0.0853, "step": 1350 }, { "grad_norm": 0.9130085706710815, "learning_rate": 6.795e-05, "loss": 0.0832, "step": 1360 }, { "grad_norm": 0.9366294145584106, "learning_rate": 6.845e-05, "loss": 0.0758, "step": 1370 }, { "grad_norm": 0.972253680229187, "learning_rate": 6.895000000000001e-05, "loss": 0.0736, "step": 1380 }, { "grad_norm": 0.8338888883590698, "learning_rate": 6.945000000000001e-05, "loss": 0.0685, "step": 1390 }, { "grad_norm": 0.848437488079071, "learning_rate": 6.995e-05, "loss": 0.0698, "step": 1400 }, { "grad_norm": 1.148547887802124, "learning_rate": 7.045e-05, "loss": 0.0735, "step": 1410 }, { "grad_norm": 0.7256631255149841, "learning_rate": 7.095e-05, "loss": 0.0701, "step": 1420 }, { "grad_norm": 1.2470545768737793, "learning_rate": 7.145e-05, "loss": 0.0827, "step": 1430 }, { "grad_norm": 0.9509998559951782, "learning_rate": 7.195e-05, "loss": 0.0797, "step": 1440 }, { "grad_norm": 0.8372241258621216, "learning_rate": 7.245000000000001e-05, "loss": 0.0819, "step": 1450 }, { "grad_norm": 0.9517603516578674, "learning_rate": 7.295000000000001e-05, "loss": 0.084, "step": 1460 }, { "grad_norm": 0.8522399663925171, "learning_rate": 7.345000000000001e-05, "loss": 0.0783, "step": 1470 }, { "grad_norm": 1.0468336343765259, "learning_rate": 7.395000000000001e-05, "loss": 0.0687, "step": 1480 }, { "grad_norm": 0.8577043414115906, "learning_rate": 7.445000000000001e-05, "loss": 0.0708, "step": 1490 }, { "grad_norm": 0.9954223036766052, "learning_rate": 7.495e-05, "loss": 0.0747, "step": 1500 }, { "grad_norm": 0.960878849029541, "learning_rate": 7.545e-05, "loss": 0.081, "step": 1510 }, { "grad_norm": 0.8557776808738708, "learning_rate": 7.595e-05, "loss": 0.0814, "step": 1520 }, { "grad_norm": 0.8636218905448914, "learning_rate": 7.645e-05, "loss": 0.0765, "step": 1530 }, { "grad_norm": 0.7792890071868896, "learning_rate": 7.695e-05, "loss": 0.0827, "step": 1540 }, { "grad_norm": 0.8205713629722595, "learning_rate": 7.745e-05, "loss": 0.0744, "step": 1550 }, { "grad_norm": 0.8274639248847961, "learning_rate": 7.795e-05, "loss": 0.0731, "step": 1560 }, { "grad_norm": 0.8495892286300659, "learning_rate": 7.845e-05, "loss": 0.0738, "step": 1570 }, { "grad_norm": 0.7587372660636902, "learning_rate": 7.895000000000001e-05, "loss": 0.0658, "step": 1580 }, { "grad_norm": 0.7936840653419495, "learning_rate": 7.945e-05, "loss": 0.0659, "step": 1590 }, { "grad_norm": 0.8597129583358765, "learning_rate": 7.995e-05, "loss": 0.0761, "step": 1600 }, { "grad_norm": 0.6577185392379761, "learning_rate": 8.045e-05, "loss": 0.0691, "step": 1610 }, { "grad_norm": 0.9112058877944946, "learning_rate": 8.095e-05, "loss": 0.0712, "step": 1620 }, { "grad_norm": 0.652986466884613, "learning_rate": 8.145e-05, "loss": 0.0724, "step": 1630 }, { "grad_norm": 0.8828380107879639, "learning_rate": 8.195e-05, "loss": 0.0694, "step": 1640 }, { "grad_norm": 0.8754063248634338, "learning_rate": 8.245e-05, "loss": 0.0647, "step": 1650 }, { "grad_norm": 0.8403984308242798, "learning_rate": 8.295000000000001e-05, "loss": 0.0631, "step": 1660 }, { "grad_norm": 0.7847513556480408, "learning_rate": 8.345000000000001e-05, "loss": 0.0702, "step": 1670 }, { "grad_norm": 0.915621280670166, "learning_rate": 8.395000000000001e-05, "loss": 0.0728, "step": 1680 }, { "grad_norm": 0.7127467393875122, "learning_rate": 8.445e-05, "loss": 0.0685, "step": 1690 }, { "grad_norm": 0.684093713760376, "learning_rate": 8.495e-05, "loss": 0.0768, "step": 1700 }, { "grad_norm": 0.8383549451828003, "learning_rate": 8.545e-05, "loss": 0.0662, "step": 1710 }, { "grad_norm": 0.820927619934082, "learning_rate": 8.595e-05, "loss": 0.0677, "step": 1720 }, { "grad_norm": 0.7275295853614807, "learning_rate": 8.645000000000001e-05, "loss": 0.0689, "step": 1730 }, { "grad_norm": 0.8796226382255554, "learning_rate": 8.695000000000001e-05, "loss": 0.0686, "step": 1740 }, { "grad_norm": 0.7109848856925964, "learning_rate": 8.745000000000001e-05, "loss": 0.0654, "step": 1750 }, { "grad_norm": 0.6888343691825867, "learning_rate": 8.795e-05, "loss": 0.0651, "step": 1760 }, { "grad_norm": 0.6560230851173401, "learning_rate": 8.845e-05, "loss": 0.0796, "step": 1770 }, { "grad_norm": 0.6830241680145264, "learning_rate": 8.895e-05, "loss": 0.0699, "step": 1780 }, { "grad_norm": 0.5884499549865723, "learning_rate": 8.945e-05, "loss": 0.0655, "step": 1790 }, { "grad_norm": 0.7516936659812927, "learning_rate": 8.995e-05, "loss": 0.0717, "step": 1800 }, { "grad_norm": 0.7261505722999573, "learning_rate": 9.045e-05, "loss": 0.0654, "step": 1810 }, { "grad_norm": 0.7556334733963013, "learning_rate": 9.095e-05, "loss": 0.0696, "step": 1820 }, { "grad_norm": 0.7753201127052307, "learning_rate": 9.145e-05, "loss": 0.0648, "step": 1830 }, { "grad_norm": 0.9452122449874878, "learning_rate": 9.195e-05, "loss": 0.0671, "step": 1840 }, { "grad_norm": 0.7534818649291992, "learning_rate": 9.245e-05, "loss": 0.0653, "step": 1850 }, { "grad_norm": 0.495876669883728, "learning_rate": 9.295000000000001e-05, "loss": 0.0606, "step": 1860 }, { "grad_norm": 0.6785076856613159, "learning_rate": 9.345000000000001e-05, "loss": 0.0642, "step": 1870 }, { "grad_norm": 0.853222131729126, "learning_rate": 9.395000000000001e-05, "loss": 0.0658, "step": 1880 }, { "grad_norm": 0.5248989462852478, "learning_rate": 9.445e-05, "loss": 0.0603, "step": 1890 }, { "grad_norm": 0.6914570927619934, "learning_rate": 9.495e-05, "loss": 0.067, "step": 1900 }, { "grad_norm": 0.6811227202415466, "learning_rate": 9.545e-05, "loss": 0.0646, "step": 1910 }, { "grad_norm": 0.5717974901199341, "learning_rate": 9.595e-05, "loss": 0.058, "step": 1920 }, { "grad_norm": 0.6157114505767822, "learning_rate": 9.645000000000001e-05, "loss": 0.0601, "step": 1930 }, { "grad_norm": 0.5473312735557556, "learning_rate": 9.695000000000001e-05, "loss": 0.0715, "step": 1940 }, { "grad_norm": 0.6020218729972839, "learning_rate": 9.745000000000001e-05, "loss": 0.0599, "step": 1950 }, { "grad_norm": 0.5028707385063171, "learning_rate": 9.795000000000001e-05, "loss": 0.0599, "step": 1960 }, { "grad_norm": 0.5370204448699951, "learning_rate": 9.845000000000001e-05, "loss": 0.0643, "step": 1970 }, { "grad_norm": 0.6729667782783508, "learning_rate": 9.895e-05, "loss": 0.0684, "step": 1980 }, { "grad_norm": 0.6235915422439575, "learning_rate": 9.945e-05, "loss": 0.0612, "step": 1990 }, { "grad_norm": 0.6145341396331787, "learning_rate": 9.995e-05, "loss": 0.0603, "step": 2000 }, { "grad_norm": 0.4986555874347687, "learning_rate": 9.99999861593158e-05, "loss": 0.0596, "step": 2010 }, { "grad_norm": 0.6580787897109985, "learning_rate": 9.999993831498517e-05, "loss": 0.0611, "step": 2020 }, { "grad_norm": 0.6270045042037964, "learning_rate": 9.99998562961682e-05, "loss": 0.0558, "step": 2030 }, { "grad_norm": 0.665068507194519, "learning_rate": 9.99997401029209e-05, "loss": 0.0634, "step": 2040 }, { "grad_norm": 0.5271323323249817, "learning_rate": 9.999958973532271e-05, "loss": 0.0657, "step": 2050 }, { "grad_norm": 0.6869821548461914, "learning_rate": 9.999940519347642e-05, "loss": 0.0608, "step": 2060 }, { "grad_norm": 0.6624715924263, "learning_rate": 9.999918647750812e-05, "loss": 0.0637, "step": 2070 }, { "grad_norm": 0.5523958802223206, "learning_rate": 9.999893358756736e-05, "loss": 0.063, "step": 2080 }, { "grad_norm": 0.7380605340003967, "learning_rate": 9.999864652382692e-05, "loss": 0.0656, "step": 2090 }, { "grad_norm": 0.6707255244255066, "learning_rate": 9.999832528648307e-05, "loss": 0.0585, "step": 2100 }, { "grad_norm": 0.623630166053772, "learning_rate": 9.999796987575532e-05, "loss": 0.0602, "step": 2110 }, { "grad_norm": 0.5420946478843689, "learning_rate": 9.999758029188662e-05, "loss": 0.0566, "step": 2120 }, { "grad_norm": 0.6893917322158813, "learning_rate": 9.999715653514324e-05, "loss": 0.0622, "step": 2130 }, { "grad_norm": 0.6731513738632202, "learning_rate": 9.999669860581482e-05, "loss": 0.0611, "step": 2140 }, { "grad_norm": 0.5536237359046936, "learning_rate": 9.999620650421433e-05, "loss": 0.0566, "step": 2150 }, { "grad_norm": 0.6298080086708069, "learning_rate": 9.999568023067813e-05, "loss": 0.0496, "step": 2160 }, { "grad_norm": 0.6635470986366272, "learning_rate": 9.999511978556595e-05, "loss": 0.0565, "step": 2170 }, { "grad_norm": 0.5682566165924072, "learning_rate": 9.99945251692608e-05, "loss": 0.0592, "step": 2180 }, { "grad_norm": 0.6703065037727356, "learning_rate": 9.999389638216912e-05, "loss": 0.0545, "step": 2190 }, { "grad_norm": 0.8263062834739685, "learning_rate": 9.999323342472068e-05, "loss": 0.0529, "step": 2200 }, { "grad_norm": 0.4463250935077667, "learning_rate": 9.99925362973686e-05, "loss": 0.0522, "step": 2210 }, { "grad_norm": 0.6623143553733826, "learning_rate": 9.999180500058935e-05, "loss": 0.0565, "step": 2220 }, { "grad_norm": 0.5776143670082092, "learning_rate": 9.99910395348828e-05, "loss": 0.0571, "step": 2230 }, { "grad_norm": 0.6831008791923523, "learning_rate": 9.99902399007721e-05, "loss": 0.0566, "step": 2240 }, { "grad_norm": 0.5696154236793518, "learning_rate": 9.99894060988038e-05, "loss": 0.0636, "step": 2250 }, { "grad_norm": 0.69194096326828, "learning_rate": 9.998853812954783e-05, "loss": 0.0663, "step": 2260 }, { "grad_norm": 0.6933267712593079, "learning_rate": 9.998763599359739e-05, "loss": 0.064, "step": 2270 }, { "grad_norm": 1.035070538520813, "learning_rate": 9.998669969156912e-05, "loss": 0.0598, "step": 2280 }, { "grad_norm": 0.5412179827690125, "learning_rate": 9.998572922410294e-05, "loss": 0.0594, "step": 2290 }, { "grad_norm": 0.6390842199325562, "learning_rate": 9.998472459186219e-05, "loss": 0.0608, "step": 2300 }, { "grad_norm": 0.5526530742645264, "learning_rate": 9.99836857955335e-05, "loss": 0.0563, "step": 2310 }, { "grad_norm": 0.7223526835441589, "learning_rate": 9.998261283582688e-05, "loss": 0.0588, "step": 2320 }, { "grad_norm": 0.5605912208557129, "learning_rate": 9.998150571347571e-05, "loss": 0.063, "step": 2330 }, { "grad_norm": 0.49070218205451965, "learning_rate": 9.998036442923667e-05, "loss": 0.0591, "step": 2340 }, { "grad_norm": 0.808766782283783, "learning_rate": 9.997918898388983e-05, "loss": 0.0607, "step": 2350 }, { "grad_norm": 0.7511130571365356, "learning_rate": 9.997797937823859e-05, "loss": 0.0631, "step": 2360 }, { "grad_norm": 0.5840237736701965, "learning_rate": 9.997673561310971e-05, "loss": 0.064, "step": 2370 }, { "grad_norm": 0.514590322971344, "learning_rate": 9.997545768935331e-05, "loss": 0.0615, "step": 2380 }, { "grad_norm": 0.5617139339447021, "learning_rate": 9.997414560784282e-05, "loss": 0.0557, "step": 2390 }, { "grad_norm": 0.853462815284729, "learning_rate": 9.997279936947502e-05, "loss": 0.0627, "step": 2400 }, { "grad_norm": 0.47693106532096863, "learning_rate": 9.997141897517006e-05, "loss": 0.0572, "step": 2410 }, { "grad_norm": 0.6072409749031067, "learning_rate": 9.997000442587145e-05, "loss": 0.0559, "step": 2420 }, { "grad_norm": 0.5374108552932739, "learning_rate": 9.9968555722546e-05, "loss": 0.0566, "step": 2430 }, { "grad_norm": 0.7348878383636475, "learning_rate": 9.996707286618387e-05, "loss": 0.0529, "step": 2440 }, { "grad_norm": 0.5875288248062134, "learning_rate": 9.996555585779861e-05, "loss": 0.0537, "step": 2450 }, { "grad_norm": 0.5810160040855408, "learning_rate": 9.996400469842708e-05, "loss": 0.0601, "step": 2460 }, { "grad_norm": 0.5872305035591125, "learning_rate": 9.996241938912945e-05, "loss": 0.0509, "step": 2470 }, { "grad_norm": 0.643047571182251, "learning_rate": 9.99607999309893e-05, "loss": 0.0602, "step": 2480 }, { "grad_norm": 0.5010973811149597, "learning_rate": 9.99591463251135e-05, "loss": 0.0525, "step": 2490 }, { "grad_norm": 0.8063540458679199, "learning_rate": 9.995745857263226e-05, "loss": 0.0583, "step": 2500 }, { "grad_norm": 0.685634195804596, "learning_rate": 9.995573667469918e-05, "loss": 0.0606, "step": 2510 }, { "grad_norm": 0.6399229168891907, "learning_rate": 9.995398063249111e-05, "loss": 0.0594, "step": 2520 }, { "grad_norm": 0.6627663969993591, "learning_rate": 9.995219044720833e-05, "loss": 0.0551, "step": 2530 }, { "grad_norm": 0.479400634765625, "learning_rate": 9.99503661200744e-05, "loss": 0.0529, "step": 2540 }, { "grad_norm": 0.5183054208755493, "learning_rate": 9.994850765233624e-05, "loss": 0.0637, "step": 2550 }, { "grad_norm": 0.6447212100028992, "learning_rate": 9.994661504526406e-05, "loss": 0.0619, "step": 2560 }, { "grad_norm": 0.4498017132282257, "learning_rate": 9.994468830015148e-05, "loss": 0.0568, "step": 2570 }, { "grad_norm": 0.5737210512161255, "learning_rate": 9.99427274183154e-05, "loss": 0.0576, "step": 2580 }, { "grad_norm": 0.4102161228656769, "learning_rate": 9.994073240109606e-05, "loss": 0.0524, "step": 2590 }, { "grad_norm": 0.5932025909423828, "learning_rate": 9.993870324985703e-05, "loss": 0.0542, "step": 2600 }, { "grad_norm": 0.3742571175098419, "learning_rate": 9.993663996598523e-05, "loss": 0.0535, "step": 2610 }, { "grad_norm": 0.6032214760780334, "learning_rate": 9.993454255089089e-05, "loss": 0.0506, "step": 2620 }, { "grad_norm": 0.5854984521865845, "learning_rate": 9.993241100600757e-05, "loss": 0.0561, "step": 2630 }, { "grad_norm": 0.5534113049507141, "learning_rate": 9.993024533279215e-05, "loss": 0.0529, "step": 2640 }, { "grad_norm": 0.47667795419692993, "learning_rate": 9.992804553272487e-05, "loss": 0.0512, "step": 2650 }, { "grad_norm": 0.5408337712287903, "learning_rate": 9.992581160730926e-05, "loss": 0.0493, "step": 2660 }, { "grad_norm": 0.5402240753173828, "learning_rate": 9.992354355807218e-05, "loss": 0.0584, "step": 2670 }, { "grad_norm": 0.49478915333747864, "learning_rate": 9.992124138656385e-05, "loss": 0.0538, "step": 2680 }, { "grad_norm": 0.47173064947128296, "learning_rate": 9.991890509435774e-05, "loss": 0.0554, "step": 2690 }, { "grad_norm": 0.4049104154109955, "learning_rate": 9.991653468305071e-05, "loss": 0.053, "step": 2700 }, { "grad_norm": 0.5808486342430115, "learning_rate": 9.991413015426292e-05, "loss": 0.0555, "step": 2710 }, { "grad_norm": 0.5587334036827087, "learning_rate": 9.991169150963781e-05, "loss": 0.0535, "step": 2720 }, { "grad_norm": 0.5281392335891724, "learning_rate": 9.990921875084221e-05, "loss": 0.0594, "step": 2730 }, { "grad_norm": 0.4564759135246277, "learning_rate": 9.990671187956622e-05, "loss": 0.0611, "step": 2740 }, { "grad_norm": 0.5672823786735535, "learning_rate": 9.990417089752324e-05, "loss": 0.0546, "step": 2750 }, { "grad_norm": 0.5874357223510742, "learning_rate": 9.990159580645001e-05, "loss": 0.0536, "step": 2760 }, { "grad_norm": 0.6569625735282898, "learning_rate": 9.98989866081066e-05, "loss": 0.0521, "step": 2770 }, { "grad_norm": 0.5644649267196655, "learning_rate": 9.989634330427636e-05, "loss": 0.0549, "step": 2780 }, { "grad_norm": 0.472597599029541, "learning_rate": 9.989366589676596e-05, "loss": 0.052, "step": 2790 }, { "grad_norm": 0.5524618625640869, "learning_rate": 9.98909543874054e-05, "loss": 0.047, "step": 2800 }, { "grad_norm": 0.5984176993370056, "learning_rate": 9.988820877804795e-05, "loss": 0.0458, "step": 2810 }, { "grad_norm": 0.5490973591804504, "learning_rate": 9.988542907057022e-05, "loss": 0.0452, "step": 2820 }, { "grad_norm": 0.46087539196014404, "learning_rate": 9.988261526687212e-05, "loss": 0.0554, "step": 2830 }, { "grad_norm": 0.4732190668582916, "learning_rate": 9.987976736887685e-05, "loss": 0.0553, "step": 2840 }, { "grad_norm": 0.574191689491272, "learning_rate": 9.987688537853091e-05, "loss": 0.0464, "step": 2850 }, { "grad_norm": 0.45711010694503784, "learning_rate": 9.987396929780413e-05, "loss": 0.047, "step": 2860 }, { "grad_norm": 0.4652513265609741, "learning_rate": 9.987101912868962e-05, "loss": 0.0456, "step": 2870 }, { "grad_norm": 0.5198781490325928, "learning_rate": 9.98680348732038e-05, "loss": 0.0459, "step": 2880 }, { "grad_norm": 0.49708321690559387, "learning_rate": 9.986501653338636e-05, "loss": 0.0457, "step": 2890 }, { "grad_norm": 0.5528702735900879, "learning_rate": 9.986196411130031e-05, "loss": 0.0446, "step": 2900 }, { "grad_norm": 0.5838372707366943, "learning_rate": 9.985887760903197e-05, "loss": 0.0441, "step": 2910 }, { "grad_norm": 0.4455975294113159, "learning_rate": 9.985575702869093e-05, "loss": 0.0405, "step": 2920 }, { "grad_norm": 0.5345509648323059, "learning_rate": 9.985260237241008e-05, "loss": 0.0506, "step": 2930 }, { "grad_norm": 0.5924714207649231, "learning_rate": 9.984941364234557e-05, "loss": 0.0535, "step": 2940 }, { "grad_norm": 0.42364904284477234, "learning_rate": 9.98461908406769e-05, "loss": 0.0523, "step": 2950 }, { "grad_norm": 0.35990995168685913, "learning_rate": 9.98429339696068e-05, "loss": 0.0508, "step": 2960 }, { "grad_norm": 0.5237912535667419, "learning_rate": 9.983964303136133e-05, "loss": 0.0454, "step": 2970 }, { "grad_norm": 0.44863200187683105, "learning_rate": 9.983631802818981e-05, "loss": 0.0441, "step": 2980 }, { "grad_norm": 0.429414838552475, "learning_rate": 9.983295896236484e-05, "loss": 0.0417, "step": 2990 }, { "grad_norm": 0.5570700764656067, "learning_rate": 9.982956583618232e-05, "loss": 0.0468, "step": 3000 }, { "grad_norm": 0.45841488242149353, "learning_rate": 9.982613865196142e-05, "loss": 0.0495, "step": 3010 }, { "grad_norm": 0.6088969707489014, "learning_rate": 9.982267741204458e-05, "loss": 0.0459, "step": 3020 }, { "grad_norm": 0.5964875221252441, "learning_rate": 9.981918211879753e-05, "loss": 0.0441, "step": 3030 }, { "grad_norm": 0.5101991891860962, "learning_rate": 9.981565277460927e-05, "loss": 0.0443, "step": 3040 }, { "grad_norm": 0.4940824806690216, "learning_rate": 9.981208938189206e-05, "loss": 0.0453, "step": 3050 }, { "grad_norm": 0.49435660243034363, "learning_rate": 9.980849194308149e-05, "loss": 0.0457, "step": 3060 }, { "grad_norm": 0.4400814473628998, "learning_rate": 9.98048604606363e-05, "loss": 0.0472, "step": 3070 }, { "grad_norm": 0.343702495098114, "learning_rate": 9.980119493703864e-05, "loss": 0.0458, "step": 3080 }, { "grad_norm": 0.4056857228279114, "learning_rate": 9.979749537479383e-05, "loss": 0.0457, "step": 3090 }, { "grad_norm": 0.502017080783844, "learning_rate": 9.979376177643051e-05, "loss": 0.047, "step": 3100 }, { "grad_norm": 0.40009042620658875, "learning_rate": 9.978999414450052e-05, "loss": 0.0503, "step": 3110 }, { "grad_norm": 0.3976839482784271, "learning_rate": 9.978619248157904e-05, "loss": 0.0405, "step": 3120 }, { "grad_norm": 0.44004786014556885, "learning_rate": 9.978235679026445e-05, "loss": 0.0537, "step": 3130 }, { "grad_norm": 0.5610888600349426, "learning_rate": 9.97784870731784e-05, "loss": 0.0443, "step": 3140 }, { "grad_norm": 0.4681626856327057, "learning_rate": 9.977458333296584e-05, "loss": 0.0466, "step": 3150 }, { "grad_norm": 0.41892245411872864, "learning_rate": 9.977064557229492e-05, "loss": 0.0392, "step": 3160 }, { "grad_norm": 0.39787614345550537, "learning_rate": 9.976667379385705e-05, "loss": 0.0433, "step": 3170 }, { "grad_norm": 0.43207815289497375, "learning_rate": 9.97626680003669e-05, "loss": 0.044, "step": 3180 }, { "grad_norm": 0.49371442198753357, "learning_rate": 9.975862819456242e-05, "loss": 0.0448, "step": 3190 }, { "grad_norm": 0.39743760228157043, "learning_rate": 9.975455437920477e-05, "loss": 0.0411, "step": 3200 }, { "grad_norm": 0.34972071647644043, "learning_rate": 9.975044655707834e-05, "loss": 0.0423, "step": 3210 }, { "grad_norm": 0.5286634564399719, "learning_rate": 9.974630473099082e-05, "loss": 0.0453, "step": 3220 }, { "grad_norm": 0.3912527859210968, "learning_rate": 9.974212890377311e-05, "loss": 0.0495, "step": 3230 }, { "grad_norm": 0.4319349527359009, "learning_rate": 9.973791907827931e-05, "loss": 0.0445, "step": 3240 }, { "grad_norm": 0.3902151584625244, "learning_rate": 9.973367525738683e-05, "loss": 0.0442, "step": 3250 }, { "grad_norm": 0.49699145555496216, "learning_rate": 9.972939744399627e-05, "loss": 0.0454, "step": 3260 }, { "grad_norm": 0.553244948387146, "learning_rate": 9.97250856410315e-05, "loss": 0.0422, "step": 3270 }, { "grad_norm": 0.5028464794158936, "learning_rate": 9.972073985143955e-05, "loss": 0.0402, "step": 3280 }, { "grad_norm": 0.3909026086330414, "learning_rate": 9.971636007819074e-05, "loss": 0.0432, "step": 3290 }, { "grad_norm": 0.41603347659111023, "learning_rate": 9.971194632427863e-05, "loss": 0.0444, "step": 3300 }, { "grad_norm": 0.47250139713287354, "learning_rate": 9.970749859271997e-05, "loss": 0.0438, "step": 3310 }, { "grad_norm": 0.45235854387283325, "learning_rate": 9.970301688655473e-05, "loss": 0.0447, "step": 3320 }, { "grad_norm": 0.4322913587093353, "learning_rate": 9.969850120884612e-05, "loss": 0.0433, "step": 3330 }, { "grad_norm": 0.4597958028316498, "learning_rate": 9.969395156268055e-05, "loss": 0.0429, "step": 3340 }, { "grad_norm": 0.4990161657333374, "learning_rate": 9.968936795116768e-05, "loss": 0.0436, "step": 3350 }, { "grad_norm": 0.4223533272743225, "learning_rate": 9.968475037744036e-05, "loss": 0.0392, "step": 3360 }, { "grad_norm": 0.4667355716228485, "learning_rate": 9.968009884465465e-05, "loss": 0.041, "step": 3370 }, { "grad_norm": 0.4203506112098694, "learning_rate": 9.967541335598984e-05, "loss": 0.0403, "step": 3380 }, { "grad_norm": 0.48353323340415955, "learning_rate": 9.967069391464841e-05, "loss": 0.0421, "step": 3390 }, { "grad_norm": 0.34876126050949097, "learning_rate": 9.966594052385608e-05, "loss": 0.0425, "step": 3400 }, { "grad_norm": 0.4827471375465393, "learning_rate": 9.96611531868617e-05, "loss": 0.0434, "step": 3410 }, { "grad_norm": 0.4842453896999359, "learning_rate": 9.96563319069374e-05, "loss": 0.0397, "step": 3420 }, { "grad_norm": 0.3950054943561554, "learning_rate": 9.965147668737847e-05, "loss": 0.0388, "step": 3430 }, { "grad_norm": 0.5607830286026001, "learning_rate": 9.96465875315034e-05, "loss": 0.0457, "step": 3440 }, { "grad_norm": 1.1582181453704834, "learning_rate": 9.96416644426539e-05, "loss": 0.0503, "step": 3450 }, { "grad_norm": 0.45340532064437866, "learning_rate": 9.963670742419485e-05, "loss": 0.0616, "step": 3460 }, { "grad_norm": 0.6157346367835999, "learning_rate": 9.96317164795143e-05, "loss": 0.058, "step": 3470 }, { "grad_norm": 0.4600118100643158, "learning_rate": 9.962669161202356e-05, "loss": 0.0519, "step": 3480 }, { "grad_norm": 0.45524314045906067, "learning_rate": 9.962163282515705e-05, "loss": 0.0507, "step": 3490 }, { "grad_norm": 0.5055732131004333, "learning_rate": 9.961654012237241e-05, "loss": 0.046, "step": 3500 }, { "grad_norm": 0.7732868194580078, "learning_rate": 9.961141350715044e-05, "loss": 0.0533, "step": 3510 }, { "grad_norm": 0.471972793340683, "learning_rate": 9.960625298299516e-05, "loss": 0.0505, "step": 3520 }, { "grad_norm": 0.6170992255210876, "learning_rate": 9.960105855343372e-05, "loss": 0.0479, "step": 3530 }, { "grad_norm": 0.579897940158844, "learning_rate": 9.959583022201647e-05, "loss": 0.0486, "step": 3540 }, { "grad_norm": 0.5838022232055664, "learning_rate": 9.959056799231692e-05, "loss": 0.0488, "step": 3550 }, { "grad_norm": 0.4831591844558716, "learning_rate": 9.958527186793176e-05, "loss": 0.0463, "step": 3560 }, { "grad_norm": 0.44914644956588745, "learning_rate": 9.957994185248086e-05, "loss": 0.0475, "step": 3570 }, { "grad_norm": 0.5127126574516296, "learning_rate": 9.957457794960718e-05, "loss": 0.0478, "step": 3580 }, { "grad_norm": 0.42584675550460815, "learning_rate": 9.956918016297694e-05, "loss": 0.0452, "step": 3590 }, { "grad_norm": 0.4570040702819824, "learning_rate": 9.956374849627948e-05, "loss": 0.0459, "step": 3600 }, { "grad_norm": 0.4069882929325104, "learning_rate": 9.955828295322728e-05, "loss": 0.0428, "step": 3610 }, { "grad_norm": 0.4282301664352417, "learning_rate": 9.955278353755598e-05, "loss": 0.04, "step": 3620 }, { "grad_norm": 0.4540446996688843, "learning_rate": 9.954725025302439e-05, "loss": 0.0474, "step": 3630 }, { "grad_norm": 0.43784788250923157, "learning_rate": 9.954168310341445e-05, "loss": 0.0423, "step": 3640 }, { "grad_norm": 0.5955245494842529, "learning_rate": 9.953608209253126e-05, "loss": 0.0422, "step": 3650 }, { "grad_norm": 0.41016000509262085, "learning_rate": 9.953044722420307e-05, "loss": 0.0407, "step": 3660 }, { "grad_norm": 0.49255773425102234, "learning_rate": 9.952477850228124e-05, "loss": 0.044, "step": 3670 }, { "grad_norm": 0.4675139784812927, "learning_rate": 9.95190759306403e-05, "loss": 0.0411, "step": 3680 }, { "grad_norm": 0.4236554205417633, "learning_rate": 9.951333951317789e-05, "loss": 0.0471, "step": 3690 }, { "grad_norm": 0.4757644236087799, "learning_rate": 9.950756925381479e-05, "loss": 0.0416, "step": 3700 }, { "grad_norm": 0.44487130641937256, "learning_rate": 9.950176515649496e-05, "loss": 0.042, "step": 3710 }, { "grad_norm": 0.6058261394500732, "learning_rate": 9.949592722518542e-05, "loss": 0.0523, "step": 3720 }, { "grad_norm": 0.38880419731140137, "learning_rate": 9.949005546387631e-05, "loss": 0.0594, "step": 3730 }, { "grad_norm": 0.34032490849494934, "learning_rate": 9.948414987658098e-05, "loss": 0.0469, "step": 3740 }, { "grad_norm": 0.4579121768474579, "learning_rate": 9.94782104673358e-05, "loss": 0.0445, "step": 3750 }, { "grad_norm": 0.41695404052734375, "learning_rate": 9.947223724020034e-05, "loss": 0.0466, "step": 3760 }, { "grad_norm": 0.5001764297485352, "learning_rate": 9.94662301992572e-05, "loss": 0.041, "step": 3770 }, { "grad_norm": 0.4948504865169525, "learning_rate": 9.946018934861216e-05, "loss": 0.0494, "step": 3780 }, { "grad_norm": 0.3718605935573578, "learning_rate": 9.945411469239409e-05, "loss": 0.0434, "step": 3790 }, { "grad_norm": 0.45404714345932007, "learning_rate": 9.944800623475497e-05, "loss": 0.0452, "step": 3800 }, { "grad_norm": 0.4104216992855072, "learning_rate": 9.944186397986984e-05, "loss": 0.0469, "step": 3810 }, { "grad_norm": 0.5228665471076965, "learning_rate": 9.943568793193687e-05, "loss": 0.0494, "step": 3820 }, { "grad_norm": 0.4373098313808441, "learning_rate": 9.942947809517737e-05, "loss": 0.0437, "step": 3830 }, { "grad_norm": 0.3645937442779541, "learning_rate": 9.942323447383568e-05, "loss": 0.0431, "step": 3840 }, { "grad_norm": 0.4466964602470398, "learning_rate": 9.941695707217925e-05, "loss": 0.0474, "step": 3850 }, { "grad_norm": 0.5664576888084412, "learning_rate": 9.941064589449865e-05, "loss": 0.0503, "step": 3860 }, { "grad_norm": 0.28796425461769104, "learning_rate": 9.940430094510748e-05, "loss": 0.0415, "step": 3870 }, { "grad_norm": 0.42040345072746277, "learning_rate": 9.939792222834248e-05, "loss": 0.0412, "step": 3880 }, { "grad_norm": 0.5184802412986755, "learning_rate": 9.939150974856343e-05, "loss": 0.0454, "step": 3890 }, { "grad_norm": 0.47815224528312683, "learning_rate": 9.93850635101532e-05, "loss": 0.0483, "step": 3900 }, { "grad_norm": 0.4387677013874054, "learning_rate": 9.937858351751775e-05, "loss": 0.0472, "step": 3910 }, { "grad_norm": 0.4440391957759857, "learning_rate": 9.937206977508604e-05, "loss": 0.0522, "step": 3920 }, { "grad_norm": 0.4345382750034332, "learning_rate": 9.936552228731022e-05, "loss": 0.0404, "step": 3930 }, { "grad_norm": 0.6300784945487976, "learning_rate": 9.93589410586654e-05, "loss": 0.0439, "step": 3940 }, { "grad_norm": 0.39559289813041687, "learning_rate": 9.93523260936498e-05, "loss": 0.0435, "step": 3950 }, { "grad_norm": 0.34023699164390564, "learning_rate": 9.934567739678467e-05, "loss": 0.0391, "step": 3960 }, { "grad_norm": 0.42682841420173645, "learning_rate": 9.933899497261433e-05, "loss": 0.0391, "step": 3970 }, { "grad_norm": 0.550899088382721, "learning_rate": 9.933227882570617e-05, "loss": 0.0408, "step": 3980 }, { "grad_norm": 0.4052072763442993, "learning_rate": 9.932552896065062e-05, "loss": 0.0408, "step": 3990 }, { "grad_norm": 0.4651620388031006, "learning_rate": 9.931874538206114e-05, "loss": 0.0457, "step": 4000 }, { "grad_norm": 0.4131235182285309, "learning_rate": 9.931192809457423e-05, "loss": 0.0436, "step": 4010 }, { "grad_norm": 0.42805126309394836, "learning_rate": 9.930507710284944e-05, "loss": 0.0418, "step": 4020 }, { "grad_norm": 0.5135667324066162, "learning_rate": 9.92981924115694e-05, "loss": 0.0443, "step": 4030 }, { "grad_norm": 0.5325000882148743, "learning_rate": 9.929127402543968e-05, "loss": 0.0419, "step": 4040 }, { "grad_norm": 0.36313432455062866, "learning_rate": 9.928432194918895e-05, "loss": 0.0444, "step": 4050 }, { "grad_norm": 0.40591150522232056, "learning_rate": 9.927733618756889e-05, "loss": 0.0484, "step": 4060 }, { "grad_norm": 0.5068106651306152, "learning_rate": 9.92703167453542e-05, "loss": 0.0427, "step": 4070 }, { "grad_norm": 0.3860991597175598, "learning_rate": 9.926326362734263e-05, "loss": 0.0505, "step": 4080 }, { "grad_norm": 0.40083470940589905, "learning_rate": 9.925617683835489e-05, "loss": 0.0408, "step": 4090 }, { "grad_norm": 0.3314113914966583, "learning_rate": 9.924905638323472e-05, "loss": 0.0397, "step": 4100 }, { "grad_norm": 0.4032135307788849, "learning_rate": 9.92419022668489e-05, "loss": 0.0434, "step": 4110 }, { "grad_norm": 0.3014068305492401, "learning_rate": 9.923471449408723e-05, "loss": 0.0414, "step": 4120 }, { "grad_norm": 0.49859049916267395, "learning_rate": 9.922749306986244e-05, "loss": 0.0454, "step": 4130 }, { "grad_norm": 0.4207134544849396, "learning_rate": 9.922023799911033e-05, "loss": 0.0412, "step": 4140 }, { "grad_norm": 0.5193842649459839, "learning_rate": 9.921294928678968e-05, "loss": 0.0409, "step": 4150 }, { "grad_norm": 0.4103717505931854, "learning_rate": 9.920562693788222e-05, "loss": 0.0389, "step": 4160 }, { "grad_norm": 0.3476479947566986, "learning_rate": 9.919827095739273e-05, "loss": 0.0402, "step": 4170 }, { "grad_norm": 0.5540982484817505, "learning_rate": 9.919088135034898e-05, "loss": 0.039, "step": 4180 }, { "grad_norm": 0.41657528281211853, "learning_rate": 9.918345812180165e-05, "loss": 0.0443, "step": 4190 }, { "grad_norm": 0.40182608366012573, "learning_rate": 9.917600127682446e-05, "loss": 0.037, "step": 4200 }, { "grad_norm": 0.4191625416278839, "learning_rate": 9.91685108205141e-05, "loss": 0.0369, "step": 4210 }, { "grad_norm": 0.4609600007534027, "learning_rate": 9.916098675799024e-05, "loss": 0.0371, "step": 4220 }, { "grad_norm": 0.7378095984458923, "learning_rate": 9.915342909439548e-05, "loss": 0.043, "step": 4230 }, { "grad_norm": 0.43525341153144836, "learning_rate": 9.914583783489543e-05, "loss": 0.0471, "step": 4240 }, { "grad_norm": 0.5059776306152344, "learning_rate": 9.913821298467863e-05, "loss": 0.0448, "step": 4250 }, { "grad_norm": 0.542802631855011, "learning_rate": 9.913055454895661e-05, "loss": 0.0441, "step": 4260 }, { "grad_norm": 0.33282336592674255, "learning_rate": 9.912286253296381e-05, "loss": 0.0413, "step": 4270 }, { "grad_norm": 0.3834342658519745, "learning_rate": 9.911513694195768e-05, "loss": 0.0391, "step": 4280 }, { "grad_norm": 0.38425007462501526, "learning_rate": 9.910737778121859e-05, "loss": 0.0367, "step": 4290 }, { "grad_norm": 0.3097541928291321, "learning_rate": 9.909958505604984e-05, "loss": 0.0334, "step": 4300 }, { "grad_norm": 0.3587503135204315, "learning_rate": 9.909175877177767e-05, "loss": 0.0331, "step": 4310 }, { "grad_norm": 0.426146000623703, "learning_rate": 9.908389893375129e-05, "loss": 0.0376, "step": 4320 }, { "grad_norm": 0.420052170753479, "learning_rate": 9.907600554734283e-05, "loss": 0.0344, "step": 4330 }, { "grad_norm": 0.31017976999282837, "learning_rate": 9.906807861794734e-05, "loss": 0.0364, "step": 4340 }, { "grad_norm": 0.3344161808490753, "learning_rate": 9.906011815098279e-05, "loss": 0.042, "step": 4350 }, { "grad_norm": 0.39384156465530396, "learning_rate": 9.90521241518901e-05, "loss": 0.0359, "step": 4360 }, { "grad_norm": 0.3952242434024811, "learning_rate": 9.904409662613308e-05, "loss": 0.0392, "step": 4370 }, { "grad_norm": 0.5148122310638428, "learning_rate": 9.903603557919849e-05, "loss": 0.045, "step": 4380 }, { "grad_norm": 0.39166325330734253, "learning_rate": 9.902794101659594e-05, "loss": 0.0467, "step": 4390 }, { "grad_norm": 0.4391108453273773, "learning_rate": 9.901981294385803e-05, "loss": 0.0401, "step": 4400 }, { "grad_norm": 0.47732511162757874, "learning_rate": 9.901165136654018e-05, "loss": 0.0408, "step": 4410 }, { "grad_norm": 0.46537676453590393, "learning_rate": 9.900345629022079e-05, "loss": 0.041, "step": 4420 }, { "grad_norm": 0.41732439398765564, "learning_rate": 9.89952277205011e-05, "loss": 0.0374, "step": 4430 }, { "grad_norm": 0.371968537569046, "learning_rate": 9.898696566300527e-05, "loss": 0.0384, "step": 4440 }, { "grad_norm": 0.4395129382610321, "learning_rate": 9.897867012338032e-05, "loss": 0.0404, "step": 4450 }, { "grad_norm": 0.5281887054443359, "learning_rate": 9.897034110729617e-05, "loss": 0.0437, "step": 4460 }, { "grad_norm": 0.37140071392059326, "learning_rate": 9.896197862044564e-05, "loss": 0.0443, "step": 4470 }, { "grad_norm": 0.4122464060783386, "learning_rate": 9.895358266854442e-05, "loss": 0.041, "step": 4480 }, { "grad_norm": 0.5127971768379211, "learning_rate": 9.894515325733103e-05, "loss": 0.0468, "step": 4490 }, { "grad_norm": 0.3425811231136322, "learning_rate": 9.893669039256693e-05, "loss": 0.0417, "step": 4500 }, { "grad_norm": 0.41825205087661743, "learning_rate": 9.89281940800364e-05, "loss": 0.0396, "step": 4510 }, { "grad_norm": 0.4902319610118866, "learning_rate": 9.891966432554655e-05, "loss": 0.0394, "step": 4520 }, { "grad_norm": 0.42458033561706543, "learning_rate": 9.891110113492745e-05, "loss": 0.0434, "step": 4530 }, { "grad_norm": 0.37409138679504395, "learning_rate": 9.89025045140319e-05, "loss": 0.0406, "step": 4540 }, { "grad_norm": 0.38928353786468506, "learning_rate": 9.889387446873567e-05, "loss": 0.0366, "step": 4550 }, { "grad_norm": 0.44687047600746155, "learning_rate": 9.888521100493726e-05, "loss": 0.0392, "step": 4560 }, { "grad_norm": 0.4533497393131256, "learning_rate": 9.887651412855809e-05, "loss": 0.0358, "step": 4570 }, { "grad_norm": 0.4624086916446686, "learning_rate": 9.88677838455424e-05, "loss": 0.042, "step": 4580 }, { "grad_norm": 0.39846736192703247, "learning_rate": 9.885902016185725e-05, "loss": 0.0431, "step": 4590 }, { "grad_norm": 0.41557785868644714, "learning_rate": 9.885022308349252e-05, "loss": 0.0369, "step": 4600 }, { "grad_norm": 0.45077672600746155, "learning_rate": 9.884139261646095e-05, "loss": 0.0423, "step": 4610 }, { "grad_norm": 0.5392223596572876, "learning_rate": 9.883252876679807e-05, "loss": 0.042, "step": 4620 }, { "grad_norm": 0.331158310174942, "learning_rate": 9.882363154056225e-05, "loss": 0.0426, "step": 4630 }, { "grad_norm": 0.521746814250946, "learning_rate": 9.881470094383465e-05, "loss": 0.0411, "step": 4640 }, { "grad_norm": 0.445451021194458, "learning_rate": 9.880573698271924e-05, "loss": 0.0394, "step": 4650 }, { "grad_norm": 0.43372470140457153, "learning_rate": 9.879673966334282e-05, "loss": 0.0444, "step": 4660 }, { "grad_norm": 0.42027372121810913, "learning_rate": 9.878770899185496e-05, "loss": 0.0382, "step": 4670 }, { "grad_norm": 0.3330642580986023, "learning_rate": 9.877864497442804e-05, "loss": 0.0349, "step": 4680 }, { "grad_norm": 0.42061519622802734, "learning_rate": 9.876954761725723e-05, "loss": 0.0351, "step": 4690 }, { "grad_norm": 0.5111641883850098, "learning_rate": 9.876041692656052e-05, "loss": 0.0476, "step": 4700 }, { "grad_norm": 0.44750821590423584, "learning_rate": 9.87512529085786e-05, "loss": 0.0422, "step": 4710 }, { "grad_norm": 0.4870935380458832, "learning_rate": 9.874205556957503e-05, "loss": 0.0378, "step": 4720 }, { "grad_norm": 0.42123129963874817, "learning_rate": 9.873282491583608e-05, "loss": 0.038, "step": 4730 }, { "grad_norm": 0.42499566078186035, "learning_rate": 9.872356095367084e-05, "loss": 0.0359, "step": 4740 }, { "grad_norm": 0.28153273463249207, "learning_rate": 9.87142636894111e-05, "loss": 0.0365, "step": 4750 }, { "grad_norm": 0.45674583315849304, "learning_rate": 9.870493312941148e-05, "loss": 0.0388, "step": 4760 }, { "grad_norm": 0.40435102581977844, "learning_rate": 9.869556928004933e-05, "loss": 0.0449, "step": 4770 }, { "grad_norm": 0.37311580777168274, "learning_rate": 9.868617214772476e-05, "loss": 0.0415, "step": 4780 }, { "grad_norm": 0.3909149169921875, "learning_rate": 9.867674173886059e-05, "loss": 0.0449, "step": 4790 }, { "grad_norm": 0.40131092071533203, "learning_rate": 9.866727805990245e-05, "loss": 0.041, "step": 4800 }, { "grad_norm": 0.40467557311058044, "learning_rate": 9.865778111731865e-05, "loss": 0.0356, "step": 4810 }, { "grad_norm": 0.359013170003891, "learning_rate": 9.864825091760028e-05, "loss": 0.0361, "step": 4820 }, { "grad_norm": 0.3730357885360718, "learning_rate": 9.863868746726112e-05, "loss": 0.0333, "step": 4830 }, { "grad_norm": 0.29776039719581604, "learning_rate": 9.862909077283773e-05, "loss": 0.0327, "step": 4840 }, { "grad_norm": 0.42311009764671326, "learning_rate": 9.861946084088933e-05, "loss": 0.0384, "step": 4850 }, { "grad_norm": 0.3653348982334137, "learning_rate": 9.860979767799792e-05, "loss": 0.0353, "step": 4860 }, { "grad_norm": 0.4185222387313843, "learning_rate": 9.860010129076813e-05, "loss": 0.0395, "step": 4870 }, { "grad_norm": 0.3593049943447113, "learning_rate": 9.859037168582741e-05, "loss": 0.0361, "step": 4880 }, { "grad_norm": 0.41691964864730835, "learning_rate": 9.85806088698258e-05, "loss": 0.0358, "step": 4890 }, { "grad_norm": 0.3626408278942108, "learning_rate": 9.857081284943612e-05, "loss": 0.0362, "step": 4900 }, { "grad_norm": 0.4037771224975586, "learning_rate": 9.856098363135385e-05, "loss": 0.0337, "step": 4910 }, { "grad_norm": 0.40333476662635803, "learning_rate": 9.855112122229717e-05, "loss": 0.0369, "step": 4920 }, { "grad_norm": 0.38086313009262085, "learning_rate": 9.854122562900696e-05, "loss": 0.0396, "step": 4930 }, { "grad_norm": 0.4283423125743866, "learning_rate": 9.853129685824673e-05, "loss": 0.0324, "step": 4940 }, { "grad_norm": 0.484939843416214, "learning_rate": 9.852133491680273e-05, "loss": 0.0358, "step": 4950 }, { "grad_norm": 0.2919255793094635, "learning_rate": 9.851133981148385e-05, "loss": 0.0347, "step": 4960 }, { "grad_norm": 0.314150333404541, "learning_rate": 9.850131154912164e-05, "loss": 0.0309, "step": 4970 }, { "grad_norm": 0.406009316444397, "learning_rate": 9.849125013657031e-05, "loss": 0.0392, "step": 4980 }, { "grad_norm": 0.4154247045516968, "learning_rate": 9.848115558070676e-05, "loss": 0.0391, "step": 4990 }, { "grad_norm": 0.43685275316238403, "learning_rate": 9.847102788843055e-05, "loss": 0.0358, "step": 5000 }, { "grad_norm": 0.3762720823287964, "learning_rate": 9.84608670666638e-05, "loss": 0.0348, "step": 5010 }, { "grad_norm": 0.43157151341438293, "learning_rate": 9.845067312235138e-05, "loss": 0.0371, "step": 5020 }, { "grad_norm": 0.39032694697380066, "learning_rate": 9.844044606246074e-05, "loss": 0.0332, "step": 5030 }, { "grad_norm": 0.3274824917316437, "learning_rate": 9.843018589398199e-05, "loss": 0.0313, "step": 5040 }, { "grad_norm": 0.5311518311500549, "learning_rate": 9.841989262392785e-05, "loss": 0.0361, "step": 5050 }, { "grad_norm": 0.3489782512187958, "learning_rate": 9.840956625933367e-05, "loss": 0.0385, "step": 5060 }, { "grad_norm": 0.371101975440979, "learning_rate": 9.839920680725743e-05, "loss": 0.0361, "step": 5070 }, { "grad_norm": 0.47200626134872437, "learning_rate": 9.83888142747797e-05, "loss": 0.0373, "step": 5080 }, { "grad_norm": 0.4484921395778656, "learning_rate": 9.83783886690037e-05, "loss": 0.0358, "step": 5090 }, { "grad_norm": 0.4491329491138458, "learning_rate": 9.836792999705524e-05, "loss": 0.0348, "step": 5100 }, { "grad_norm": 0.46051913499832153, "learning_rate": 9.835743826608271e-05, "loss": 0.0359, "step": 5110 }, { "grad_norm": 0.3051442801952362, "learning_rate": 9.834691348325709e-05, "loss": 0.0342, "step": 5120 }, { "grad_norm": 0.33754962682724, "learning_rate": 9.833635565577199e-05, "loss": 0.0335, "step": 5130 }, { "grad_norm": 0.4594212472438812, "learning_rate": 9.832576479084359e-05, "loss": 0.0336, "step": 5140 }, { "grad_norm": 0.44353771209716797, "learning_rate": 9.831514089571064e-05, "loss": 0.0355, "step": 5150 }, { "grad_norm": 0.3325527012348175, "learning_rate": 9.830448397763447e-05, "loss": 0.0367, "step": 5160 }, { "grad_norm": 0.3852197825908661, "learning_rate": 9.8293794043899e-05, "loss": 0.0373, "step": 5170 }, { "grad_norm": 0.323600709438324, "learning_rate": 9.828307110181066e-05, "loss": 0.0328, "step": 5180 }, { "grad_norm": 0.3297925591468811, "learning_rate": 9.827231515869852e-05, "loss": 0.035, "step": 5190 }, { "grad_norm": 0.4018571972846985, "learning_rate": 9.826152622191413e-05, "loss": 0.0327, "step": 5200 }, { "grad_norm": 0.4512878954410553, "learning_rate": 9.825070429883167e-05, "loss": 0.0391, "step": 5210 }, { "grad_norm": 0.36004137992858887, "learning_rate": 9.82398493968478e-05, "loss": 0.0389, "step": 5220 }, { "grad_norm": 0.5662655830383301, "learning_rate": 9.822896152338173e-05, "loss": 0.034, "step": 5230 }, { "grad_norm": 0.36365842819213867, "learning_rate": 9.821804068587523e-05, "loss": 0.0335, "step": 5240 }, { "grad_norm": 0.3575446605682373, "learning_rate": 9.820708689179259e-05, "loss": 0.0363, "step": 5250 }, { "grad_norm": 0.3436412811279297, "learning_rate": 9.819610014862063e-05, "loss": 0.0318, "step": 5260 }, { "grad_norm": 0.3785577416419983, "learning_rate": 9.818508046386868e-05, "loss": 0.0318, "step": 5270 }, { "grad_norm": 0.4731779098510742, "learning_rate": 9.817402784506859e-05, "loss": 0.0317, "step": 5280 }, { "grad_norm": 0.3580905795097351, "learning_rate": 9.816294229977472e-05, "loss": 0.0337, "step": 5290 }, { "grad_norm": 0.5018748044967651, "learning_rate": 9.815182383556394e-05, "loss": 0.0371, "step": 5300 }, { "grad_norm": 0.3698716163635254, "learning_rate": 9.81406724600356e-05, "loss": 0.0376, "step": 5310 }, { "grad_norm": 0.4249502122402191, "learning_rate": 9.812948818081161e-05, "loss": 0.0403, "step": 5320 }, { "grad_norm": 0.32550695538520813, "learning_rate": 9.811827100553623e-05, "loss": 0.0323, "step": 5330 }, { "grad_norm": 0.3868997097015381, "learning_rate": 9.810702094187638e-05, "loss": 0.0346, "step": 5340 }, { "grad_norm": 0.4110497534275055, "learning_rate": 9.809573799752135e-05, "loss": 0.0329, "step": 5350 }, { "grad_norm": 0.39250779151916504, "learning_rate": 9.808442218018288e-05, "loss": 0.029, "step": 5360 }, { "grad_norm": 0.32975661754608154, "learning_rate": 9.807307349759527e-05, "loss": 0.0287, "step": 5370 }, { "grad_norm": 0.346955269575119, "learning_rate": 9.806169195751525e-05, "loss": 0.0308, "step": 5380 }, { "grad_norm": 0.3724677562713623, "learning_rate": 9.805027756772194e-05, "loss": 0.0295, "step": 5390 }, { "grad_norm": 0.31417420506477356, "learning_rate": 9.803883033601702e-05, "loss": 0.0291, "step": 5400 }, { "grad_norm": 0.28775063157081604, "learning_rate": 9.802735027022453e-05, "loss": 0.0305, "step": 5410 }, { "grad_norm": 0.35004326701164246, "learning_rate": 9.8015837378191e-05, "loss": 0.0282, "step": 5420 }, { "grad_norm": 0.3098219037055969, "learning_rate": 9.80042916677854e-05, "loss": 0.032, "step": 5430 }, { "grad_norm": 0.3610292375087738, "learning_rate": 9.799271314689908e-05, "loss": 0.0311, "step": 5440 }, { "grad_norm": 0.4279262125492096, "learning_rate": 9.798110182344588e-05, "loss": 0.0326, "step": 5450 }, { "grad_norm": 0.38609832525253296, "learning_rate": 9.796945770536204e-05, "loss": 0.0317, "step": 5460 }, { "grad_norm": 0.4591657817363739, "learning_rate": 9.795778080060615e-05, "loss": 0.0322, "step": 5470 }, { "grad_norm": 0.47080159187316895, "learning_rate": 9.79460711171593e-05, "loss": 0.0355, "step": 5480 }, { "grad_norm": 0.436164528131485, "learning_rate": 9.793432866302497e-05, "loss": 0.0323, "step": 5490 }, { "grad_norm": 0.507859468460083, "learning_rate": 9.792255344622897e-05, "loss": 0.0292, "step": 5500 }, { "grad_norm": 0.6121169328689575, "learning_rate": 9.791074547481957e-05, "loss": 0.0439, "step": 5510 }, { "grad_norm": 0.4760006368160248, "learning_rate": 9.789890475686742e-05, "loss": 0.0461, "step": 5520 }, { "grad_norm": 0.4754295349121094, "learning_rate": 9.788703130046552e-05, "loss": 0.0371, "step": 5530 }, { "grad_norm": 0.3550874888896942, "learning_rate": 9.78751251137293e-05, "loss": 0.0387, "step": 5540 }, { "grad_norm": 0.3438611626625061, "learning_rate": 9.786318620479646e-05, "loss": 0.0377, "step": 5550 }, { "grad_norm": 0.3641085922718048, "learning_rate": 9.78512145818272e-05, "loss": 0.034, "step": 5560 }, { "grad_norm": 0.3499623239040375, "learning_rate": 9.783921025300398e-05, "loss": 0.0401, "step": 5570 }, { "grad_norm": 0.49137043952941895, "learning_rate": 9.782717322653164e-05, "loss": 0.0343, "step": 5580 }, { "grad_norm": 0.4496133625507355, "learning_rate": 9.781510351063738e-05, "loss": 0.0401, "step": 5590 }, { "grad_norm": 0.3957708477973938, "learning_rate": 9.780300111357075e-05, "loss": 0.0369, "step": 5600 }, { "grad_norm": 0.4804340898990631, "learning_rate": 9.779086604360361e-05, "loss": 0.0342, "step": 5610 }, { "grad_norm": 0.543345034122467, "learning_rate": 9.777869830903019e-05, "loss": 0.0331, "step": 5620 }, { "grad_norm": 0.38005056977272034, "learning_rate": 9.776649791816698e-05, "loss": 0.0314, "step": 5630 }, { "grad_norm": 0.3417242765426636, "learning_rate": 9.775426487935285e-05, "loss": 0.0284, "step": 5640 }, { "grad_norm": 0.3320237994194031, "learning_rate": 9.774199920094898e-05, "loss": 0.0296, "step": 5650 }, { "grad_norm": 0.3443394899368286, "learning_rate": 9.772970089133884e-05, "loss": 0.0355, "step": 5660 }, { "grad_norm": 0.4236684739589691, "learning_rate": 9.77173699589282e-05, "loss": 0.031, "step": 5670 }, { "grad_norm": 0.42827358841896057, "learning_rate": 9.770500641214513e-05, "loss": 0.0311, "step": 5680 }, { "grad_norm": 0.39352044463157654, "learning_rate": 9.769261025944003e-05, "loss": 0.027, "step": 5690 }, { "grad_norm": 0.4287756085395813, "learning_rate": 9.768018150928552e-05, "loss": 0.0302, "step": 5700 }, { "grad_norm": 0.43585914373397827, "learning_rate": 9.766772017017654e-05, "loss": 0.0292, "step": 5710 }, { "grad_norm": 0.2929578721523285, "learning_rate": 9.765522625063032e-05, "loss": 0.0288, "step": 5720 }, { "grad_norm": 0.32489821314811707, "learning_rate": 9.76426997591863e-05, "loss": 0.0329, "step": 5730 }, { "grad_norm": 0.5409331321716309, "learning_rate": 9.763014070440625e-05, "loss": 0.0349, "step": 5740 }, { "grad_norm": 0.35163700580596924, "learning_rate": 9.761754909487415e-05, "loss": 0.0368, "step": 5750 }, { "grad_norm": 0.406637966632843, "learning_rate": 9.760492493919626e-05, "loss": 0.0298, "step": 5760 }, { "grad_norm": 0.2862915098667145, "learning_rate": 9.759226824600105e-05, "loss": 0.0319, "step": 5770 }, { "grad_norm": 0.33246761560440063, "learning_rate": 9.757957902393928e-05, "loss": 0.0295, "step": 5780 }, { "grad_norm": 0.42075204849243164, "learning_rate": 9.756685728168387e-05, "loss": 0.0334, "step": 5790 }, { "grad_norm": 0.5060274004936218, "learning_rate": 9.755410302793004e-05, "loss": 0.0318, "step": 5800 }, { "grad_norm": 0.357109397649765, "learning_rate": 9.754131627139522e-05, "loss": 0.0294, "step": 5810 }, { "grad_norm": 0.29132434725761414, "learning_rate": 9.752849702081901e-05, "loss": 0.0309, "step": 5820 }, { "grad_norm": 0.3393866717815399, "learning_rate": 9.751564528496324e-05, "loss": 0.0312, "step": 5830 }, { "grad_norm": 0.3994801938533783, "learning_rate": 9.750276107261197e-05, "loss": 0.0289, "step": 5840 }, { "grad_norm": 0.3093264698982239, "learning_rate": 9.748984439257142e-05, "loss": 0.0279, "step": 5850 }, { "grad_norm": 0.33789685368537903, "learning_rate": 9.747689525367005e-05, "loss": 0.0292, "step": 5860 }, { "grad_norm": 0.3011770248413086, "learning_rate": 9.746391366475845e-05, "loss": 0.0307, "step": 5870 }, { "grad_norm": 0.3812103569507599, "learning_rate": 9.745089963470942e-05, "loss": 0.0269, "step": 5880 }, { "grad_norm": 0.2675851285457611, "learning_rate": 9.743785317241791e-05, "loss": 0.0257, "step": 5890 }, { "grad_norm": 0.41212913393974304, "learning_rate": 9.742477428680108e-05, "loss": 0.0276, "step": 5900 }, { "grad_norm": 0.3197498321533203, "learning_rate": 9.741166298679821e-05, "loss": 0.0274, "step": 5910 }, { "grad_norm": 0.3131210505962372, "learning_rate": 9.739851928137076e-05, "loss": 0.0263, "step": 5920 }, { "grad_norm": 0.3226488530635834, "learning_rate": 9.738534317950232e-05, "loss": 0.025, "step": 5930 }, { "grad_norm": 0.3808702230453491, "learning_rate": 9.737213469019864e-05, "loss": 0.0306, "step": 5940 }, { "grad_norm": 0.3849978744983673, "learning_rate": 9.735889382248757e-05, "loss": 0.0278, "step": 5950 }, { "grad_norm": 0.3610084652900696, "learning_rate": 9.734562058541916e-05, "loss": 0.0262, "step": 5960 }, { "grad_norm": 0.3701421618461609, "learning_rate": 9.733231498806552e-05, "loss": 0.03, "step": 5970 }, { "grad_norm": 0.3068681061267853, "learning_rate": 9.731897703952088e-05, "loss": 0.0256, "step": 5980 }, { "grad_norm": 0.3766013979911804, "learning_rate": 9.730560674890165e-05, "loss": 0.0285, "step": 5990 }, { "grad_norm": 0.3017835319042206, "learning_rate": 9.729220412534628e-05, "loss": 0.0295, "step": 6000 }, { "grad_norm": 0.3335022032260895, "learning_rate": 9.72787691780153e-05, "loss": 0.0257, "step": 6010 }, { "grad_norm": 0.23918230831623077, "learning_rate": 9.726530191609142e-05, "loss": 0.0285, "step": 6020 }, { "grad_norm": 0.2561664581298828, "learning_rate": 9.725180234877937e-05, "loss": 0.0259, "step": 6030 }, { "grad_norm": 0.33037084341049194, "learning_rate": 9.723827048530595e-05, "loss": 0.0285, "step": 6040 }, { "grad_norm": 0.33766821026802063, "learning_rate": 9.722470633492011e-05, "loss": 0.0299, "step": 6050 }, { "grad_norm": 0.28815147280693054, "learning_rate": 9.721110990689278e-05, "loss": 0.0259, "step": 6060 }, { "grad_norm": 0.35508403182029724, "learning_rate": 9.719748121051699e-05, "loss": 0.0278, "step": 6070 }, { "grad_norm": 0.2768538296222687, "learning_rate": 9.718382025510785e-05, "loss": 0.0281, "step": 6080 }, { "grad_norm": 0.501425564289093, "learning_rate": 9.717012705000248e-05, "loss": 0.0244, "step": 6090 }, { "grad_norm": 0.39658665657043457, "learning_rate": 9.715640160456006e-05, "loss": 0.0305, "step": 6100 }, { "grad_norm": 0.35346224904060364, "learning_rate": 9.714264392816181e-05, "loss": 0.0259, "step": 6110 }, { "grad_norm": 0.3432239294052124, "learning_rate": 9.712885403021095e-05, "loss": 0.03, "step": 6120 }, { "grad_norm": 0.34005218744277954, "learning_rate": 9.711503192013276e-05, "loss": 0.0258, "step": 6130 }, { "grad_norm": 0.44375133514404297, "learning_rate": 9.710117760737454e-05, "loss": 0.0241, "step": 6140 }, { "grad_norm": 0.43594998121261597, "learning_rate": 9.708729110140554e-05, "loss": 0.0281, "step": 6150 }, { "grad_norm": 0.30307137966156006, "learning_rate": 9.707337241171709e-05, "loss": 0.0257, "step": 6160 }, { "grad_norm": 0.39222654700279236, "learning_rate": 9.705942154782245e-05, "loss": 0.028, "step": 6170 }, { "grad_norm": 0.383294939994812, "learning_rate": 9.704543851925696e-05, "loss": 0.026, "step": 6180 }, { "grad_norm": 0.3522627353668213, "learning_rate": 9.703142333557784e-05, "loss": 0.0238, "step": 6190 }, { "grad_norm": 0.391622394323349, "learning_rate": 9.701737600636436e-05, "loss": 0.0268, "step": 6200 }, { "grad_norm": 0.3366168141365051, "learning_rate": 9.700329654121771e-05, "loss": 0.0262, "step": 6210 }, { "grad_norm": 0.3479861319065094, "learning_rate": 9.698918494976109e-05, "loss": 0.0289, "step": 6220 }, { "grad_norm": 0.37148046493530273, "learning_rate": 9.697504124163965e-05, "loss": 0.0275, "step": 6230 }, { "grad_norm": 0.33736759424209595, "learning_rate": 9.696086542652045e-05, "loss": 0.0267, "step": 6240 }, { "grad_norm": 0.3022816479206085, "learning_rate": 9.694665751409256e-05, "loss": 0.0296, "step": 6250 }, { "grad_norm": 0.3437879979610443, "learning_rate": 9.693241751406694e-05, "loss": 0.0301, "step": 6260 }, { "grad_norm": 0.4188607633113861, "learning_rate": 9.69181454361765e-05, "loss": 0.0258, "step": 6270 }, { "grad_norm": 0.31633174419403076, "learning_rate": 9.690384129017605e-05, "loss": 0.0292, "step": 6280 }, { "grad_norm": 0.3604598343372345, "learning_rate": 9.688950508584237e-05, "loss": 0.0247, "step": 6290 }, { "grad_norm": 0.36579325795173645, "learning_rate": 9.68751368329741e-05, "loss": 0.0305, "step": 6300 }, { "grad_norm": 0.39172664284706116, "learning_rate": 9.68607365413918e-05, "loss": 0.0297, "step": 6310 }, { "grad_norm": 0.3745357096195221, "learning_rate": 9.684630422093797e-05, "loss": 0.0275, "step": 6320 }, { "grad_norm": 0.35378730297088623, "learning_rate": 9.683183988147693e-05, "loss": 0.0288, "step": 6330 }, { "grad_norm": 0.3415171504020691, "learning_rate": 9.681734353289491e-05, "loss": 0.0288, "step": 6340 }, { "grad_norm": 0.36937180161476135, "learning_rate": 9.680281518510006e-05, "loss": 0.0278, "step": 6350 }, { "grad_norm": 0.2759874761104584, "learning_rate": 9.678825484802236e-05, "loss": 0.0304, "step": 6360 }, { "grad_norm": 0.28627461194992065, "learning_rate": 9.677366253161365e-05, "loss": 0.0272, "step": 6370 }, { "grad_norm": 0.3976120352745056, "learning_rate": 9.675903824584765e-05, "loss": 0.029, "step": 6380 }, { "grad_norm": 0.3981839418411255, "learning_rate": 9.674438200071991e-05, "loss": 0.0281, "step": 6390 }, { "grad_norm": 0.2737092971801758, "learning_rate": 9.672969380624786e-05, "loss": 0.0241, "step": 6400 }, { "grad_norm": 0.31666669249534607, "learning_rate": 9.671497367247073e-05, "loss": 0.0293, "step": 6410 }, { "grad_norm": 0.4175805449485779, "learning_rate": 9.670022160944959e-05, "loss": 0.0279, "step": 6420 }, { "grad_norm": 0.30005937814712524, "learning_rate": 9.668543762726734e-05, "loss": 0.0243, "step": 6430 }, { "grad_norm": 0.37879127264022827, "learning_rate": 9.667062173602868e-05, "loss": 0.0259, "step": 6440 }, { "grad_norm": 0.3343585431575775, "learning_rate": 9.665577394586013e-05, "loss": 0.0248, "step": 6450 }, { "grad_norm": 0.20844262838363647, "learning_rate": 9.664089426691006e-05, "loss": 0.0237, "step": 6460 }, { "grad_norm": 0.31287726759910583, "learning_rate": 9.662598270934852e-05, "loss": 0.0225, "step": 6470 }, { "grad_norm": 0.28221395611763, "learning_rate": 9.661103928336748e-05, "loss": 0.0262, "step": 6480 }, { "grad_norm": 0.45214277505874634, "learning_rate": 9.659606399918058e-05, "loss": 0.0249, "step": 6490 }, { "grad_norm": 0.3857182264328003, "learning_rate": 9.658105686702334e-05, "loss": 0.0257, "step": 6500 }, { "grad_norm": 0.2931417226791382, "learning_rate": 9.656601789715294e-05, "loss": 0.0262, "step": 6510 }, { "grad_norm": 0.23492397367954254, "learning_rate": 9.655094709984842e-05, "loss": 0.0242, "step": 6520 }, { "grad_norm": 0.33071646094322205, "learning_rate": 9.653584448541048e-05, "loss": 0.0267, "step": 6530 }, { "grad_norm": 0.3229232728481293, "learning_rate": 9.652071006416166e-05, "loss": 0.0269, "step": 6540 }, { "grad_norm": 0.3106355369091034, "learning_rate": 9.650554384644617e-05, "loss": 0.0262, "step": 6550 }, { "grad_norm": 0.44788843393325806, "learning_rate": 9.649034584263e-05, "loss": 0.0272, "step": 6560 }, { "grad_norm": 0.34974396228790283, "learning_rate": 9.64751160631008e-05, "loss": 0.0297, "step": 6570 }, { "grad_norm": 0.366462767124176, "learning_rate": 9.645985451826803e-05, "loss": 0.0261, "step": 6580 }, { "grad_norm": 0.36912021040916443, "learning_rate": 9.644456121856275e-05, "loss": 0.0272, "step": 6590 }, { "grad_norm": 0.30439329147338867, "learning_rate": 9.642923617443784e-05, "loss": 0.0309, "step": 6600 }, { "grad_norm": 0.32857170701026917, "learning_rate": 9.641387939636782e-05, "loss": 0.0337, "step": 6610 }, { "grad_norm": 0.3059946298599243, "learning_rate": 9.639849089484888e-05, "loss": 0.0261, "step": 6620 }, { "grad_norm": 0.38943642377853394, "learning_rate": 9.638307068039893e-05, "loss": 0.0343, "step": 6630 }, { "grad_norm": 0.3318859040737152, "learning_rate": 9.636761876355753e-05, "loss": 0.0318, "step": 6640 }, { "grad_norm": 0.33079081773757935, "learning_rate": 9.635213515488593e-05, "loss": 0.0269, "step": 6650 }, { "grad_norm": 0.3413121998310089, "learning_rate": 9.633661986496702e-05, "loss": 0.0322, "step": 6660 }, { "grad_norm": 0.3010748028755188, "learning_rate": 9.632107290440535e-05, "loss": 0.0314, "step": 6670 }, { "grad_norm": 0.28510740399360657, "learning_rate": 9.630549428382715e-05, "loss": 0.0288, "step": 6680 }, { "grad_norm": 0.3600160777568817, "learning_rate": 9.628988401388024e-05, "loss": 0.0282, "step": 6690 }, { "grad_norm": 0.3702007830142975, "learning_rate": 9.627424210523407e-05, "loss": 0.0283, "step": 6700 }, { "grad_norm": 0.3439370095729828, "learning_rate": 9.62585685685798e-05, "loss": 0.0271, "step": 6710 }, { "grad_norm": 0.35799628496170044, "learning_rate": 9.624286341463008e-05, "loss": 0.0253, "step": 6720 }, { "grad_norm": 0.3263894021511078, "learning_rate": 9.622712665411927e-05, "loss": 0.0253, "step": 6730 }, { "grad_norm": 0.3921811878681183, "learning_rate": 9.621135829780328e-05, "loss": 0.0263, "step": 6740 }, { "grad_norm": 0.32349899411201477, "learning_rate": 9.619555835645964e-05, "loss": 0.0264, "step": 6750 }, { "grad_norm": 0.45370787382125854, "learning_rate": 9.617972684088747e-05, "loss": 0.0282, "step": 6760 }, { "grad_norm": 0.2929461896419525, "learning_rate": 9.616386376190745e-05, "loss": 0.0256, "step": 6770 }, { "grad_norm": 0.34853532910346985, "learning_rate": 9.614796913036184e-05, "loss": 0.0266, "step": 6780 }, { "grad_norm": 0.26920369267463684, "learning_rate": 9.61320429571145e-05, "loss": 0.0273, "step": 6790 }, { "grad_norm": 0.28838568925857544, "learning_rate": 9.611608525305078e-05, "loss": 0.0297, "step": 6800 }, { "grad_norm": 0.6444446444511414, "learning_rate": 9.610009602907764e-05, "loss": 0.0292, "step": 6810 }, { "grad_norm": 0.3721935749053955, "learning_rate": 9.608407529612357e-05, "loss": 0.0363, "step": 6820 }, { "grad_norm": 0.35031458735466003, "learning_rate": 9.606802306513857e-05, "loss": 0.0291, "step": 6830 }, { "grad_norm": 0.3541056513786316, "learning_rate": 9.605193934709421e-05, "loss": 0.0291, "step": 6840 }, { "grad_norm": 0.5677616000175476, "learning_rate": 9.603582415298354e-05, "loss": 0.0315, "step": 6850 }, { "grad_norm": 0.3849647343158722, "learning_rate": 9.601967749382117e-05, "loss": 0.0373, "step": 6860 }, { "grad_norm": 0.4373804032802582, "learning_rate": 9.600349938064316e-05, "loss": 0.0339, "step": 6870 }, { "grad_norm": 0.3532694876194, "learning_rate": 9.598728982450711e-05, "loss": 0.0359, "step": 6880 }, { "grad_norm": 0.3864944577217102, "learning_rate": 9.597104883649211e-05, "loss": 0.0258, "step": 6890 }, { "grad_norm": 0.40807661414146423, "learning_rate": 9.59547764276987e-05, "loss": 0.0255, "step": 6900 }, { "grad_norm": 0.47009220719337463, "learning_rate": 9.593847260924892e-05, "loss": 0.0239, "step": 6910 }, { "grad_norm": 0.3621138632297516, "learning_rate": 9.59221373922863e-05, "loss": 0.0292, "step": 6920 }, { "grad_norm": 0.4674352705478668, "learning_rate": 9.59057707879758e-05, "loss": 0.0281, "step": 6930 }, { "grad_norm": 0.24807977676391602, "learning_rate": 9.588937280750382e-05, "loss": 0.0259, "step": 6940 }, { "grad_norm": 0.28568190336227417, "learning_rate": 9.587294346207824e-05, "loss": 0.0258, "step": 6950 }, { "grad_norm": 0.3541107773780823, "learning_rate": 9.585648276292836e-05, "loss": 0.0295, "step": 6960 }, { "grad_norm": 0.2649005353450775, "learning_rate": 9.583999072130492e-05, "loss": 0.0256, "step": 6970 }, { "grad_norm": 0.34033310413360596, "learning_rate": 9.582346734848005e-05, "loss": 0.0296, "step": 6980 }, { "grad_norm": 0.3124617636203766, "learning_rate": 9.580691265574735e-05, "loss": 0.0267, "step": 6990 }, { "grad_norm": 0.29623937606811523, "learning_rate": 9.579032665442178e-05, "loss": 0.0278, "step": 7000 }, { "grad_norm": 0.2389025092124939, "learning_rate": 9.577370935583971e-05, "loss": 0.0309, "step": 7010 }, { "grad_norm": 0.3216720223426819, "learning_rate": 9.575706077135893e-05, "loss": 0.0318, "step": 7020 }, { "grad_norm": 0.4640809893608093, "learning_rate": 9.574038091235857e-05, "loss": 0.03, "step": 7030 }, { "grad_norm": 0.3932536840438843, "learning_rate": 9.572366979023916e-05, "loss": 0.037, "step": 7040 }, { "grad_norm": 0.3015384376049042, "learning_rate": 9.57069274164226e-05, "loss": 0.0292, "step": 7050 }, { "grad_norm": 0.32522523403167725, "learning_rate": 9.569015380235213e-05, "loss": 0.031, "step": 7060 }, { "grad_norm": 0.29311540722846985, "learning_rate": 9.567334895949238e-05, "loss": 0.0273, "step": 7070 }, { "grad_norm": 0.38889262080192566, "learning_rate": 9.565651289932928e-05, "loss": 0.0288, "step": 7080 }, { "grad_norm": 0.3859604597091675, "learning_rate": 9.563964563337015e-05, "loss": 0.0328, "step": 7090 }, { "grad_norm": 0.37999680638313293, "learning_rate": 9.562274717314357e-05, "loss": 0.0273, "step": 7100 }, { "grad_norm": 0.33482626080513, "learning_rate": 9.56058175301995e-05, "loss": 0.0291, "step": 7110 }, { "grad_norm": 0.3280470371246338, "learning_rate": 9.558885671610918e-05, "loss": 0.0272, "step": 7120 }, { "grad_norm": 0.33344748616218567, "learning_rate": 9.557186474246517e-05, "loss": 0.0236, "step": 7130 }, { "grad_norm": 0.2851226031780243, "learning_rate": 9.555484162088133e-05, "loss": 0.0265, "step": 7140 }, { "grad_norm": 0.347761332988739, "learning_rate": 9.553778736299279e-05, "loss": 0.0224, "step": 7150 }, { "grad_norm": 0.3169417381286621, "learning_rate": 9.552070198045599e-05, "loss": 0.0284, "step": 7160 }, { "grad_norm": 0.37750163674354553, "learning_rate": 9.550358548494863e-05, "loss": 0.0307, "step": 7170 }, { "grad_norm": 0.30811807513237, "learning_rate": 9.548643788816965e-05, "loss": 0.0319, "step": 7180 }, { "grad_norm": 0.40612977743148804, "learning_rate": 9.546925920183929e-05, "loss": 0.0292, "step": 7190 }, { "grad_norm": 0.36271241307258606, "learning_rate": 9.545204943769902e-05, "loss": 0.0284, "step": 7200 }, { "grad_norm": 0.3196602165699005, "learning_rate": 9.543480860751155e-05, "loss": 0.0258, "step": 7210 }, { "grad_norm": 0.3679661452770233, "learning_rate": 9.541753672306081e-05, "loss": 0.0265, "step": 7220 }, { "grad_norm": 0.2539997100830078, "learning_rate": 9.540023379615198e-05, "loss": 0.0265, "step": 7230 }, { "grad_norm": 0.34138229489326477, "learning_rate": 9.538289983861146e-05, "loss": 0.0238, "step": 7240 }, { "grad_norm": 0.39826157689094543, "learning_rate": 9.536553486228683e-05, "loss": 0.0232, "step": 7250 }, { "grad_norm": 0.3412843942642212, "learning_rate": 9.534813887904689e-05, "loss": 0.0226, "step": 7260 }, { "grad_norm": 0.4046872854232788, "learning_rate": 9.533071190078163e-05, "loss": 0.026, "step": 7270 }, { "grad_norm": 0.41153523325920105, "learning_rate": 9.531325393940225e-05, "loss": 0.0214, "step": 7280 }, { "grad_norm": 0.4240032732486725, "learning_rate": 9.529576500684105e-05, "loss": 0.0261, "step": 7290 }, { "grad_norm": 0.3030003011226654, "learning_rate": 9.52782451150516e-05, "loss": 0.0293, "step": 7300 }, { "grad_norm": 0.34802499413490295, "learning_rate": 9.526069427600857e-05, "loss": 0.0262, "step": 7310 }, { "grad_norm": 0.3565171957015991, "learning_rate": 9.524311250170776e-05, "loss": 0.0254, "step": 7320 }, { "grad_norm": 0.2782709300518036, "learning_rate": 9.522549980416619e-05, "loss": 0.029, "step": 7330 }, { "grad_norm": 0.29749128222465515, "learning_rate": 9.520785619542196e-05, "loss": 0.0243, "step": 7340 }, { "grad_norm": 0.4250601828098297, "learning_rate": 9.519018168753428e-05, "loss": 0.0278, "step": 7350 }, { "grad_norm": 0.35248470306396484, "learning_rate": 9.517247629258357e-05, "loss": 0.0217, "step": 7360 }, { "grad_norm": 0.3261374533176422, "learning_rate": 9.515474002267124e-05, "loss": 0.0244, "step": 7370 }, { "grad_norm": 0.3270490765571594, "learning_rate": 9.513697288991989e-05, "loss": 0.0243, "step": 7380 }, { "grad_norm": 0.3166464865207672, "learning_rate": 9.51191749064732e-05, "loss": 0.0248, "step": 7390 }, { "grad_norm": 0.3286330997943878, "learning_rate": 9.51013460844959e-05, "loss": 0.0225, "step": 7400 }, { "grad_norm": 0.34940582513809204, "learning_rate": 9.508348643617382e-05, "loss": 0.0279, "step": 7410 }, { "grad_norm": 0.27676379680633545, "learning_rate": 9.50655959737139e-05, "loss": 0.0233, "step": 7420 }, { "grad_norm": 0.2771269381046295, "learning_rate": 9.504767470934405e-05, "loss": 0.0312, "step": 7430 }, { "grad_norm": 0.290549635887146, "learning_rate": 9.502972265531332e-05, "loss": 0.0265, "step": 7440 }, { "grad_norm": 0.3350881338119507, "learning_rate": 9.501173982389175e-05, "loss": 0.0258, "step": 7450 }, { "grad_norm": 0.3134850561618805, "learning_rate": 9.499372622737047e-05, "loss": 0.025, "step": 7460 }, { "grad_norm": 0.32948216795921326, "learning_rate": 9.497568187806156e-05, "loss": 0.0284, "step": 7470 }, { "grad_norm": 0.2799876630306244, "learning_rate": 9.495760678829819e-05, "loss": 0.0242, "step": 7480 }, { "grad_norm": 0.31399837136268616, "learning_rate": 9.49395009704345e-05, "loss": 0.0246, "step": 7490 }, { "grad_norm": 0.3112129271030426, "learning_rate": 9.492136443684565e-05, "loss": 0.0245, "step": 7500 }, { "grad_norm": 0.3165290951728821, "learning_rate": 9.49031971999278e-05, "loss": 0.0208, "step": 7510 }, { "grad_norm": 0.323796808719635, "learning_rate": 9.488499927209806e-05, "loss": 0.0229, "step": 7520 }, { "grad_norm": 0.3182159960269928, "learning_rate": 9.486677066579456e-05, "loss": 0.0233, "step": 7530 }, { "grad_norm": 0.2507661283016205, "learning_rate": 9.484851139347639e-05, "loss": 0.0225, "step": 7540 }, { "grad_norm": 0.3296395242214203, "learning_rate": 9.483022146762358e-05, "loss": 0.0226, "step": 7550 }, { "grad_norm": 0.2933656573295593, "learning_rate": 9.48119009007371e-05, "loss": 0.0244, "step": 7560 }, { "grad_norm": 0.3776261508464813, "learning_rate": 9.479354970533892e-05, "loss": 0.0213, "step": 7570 }, { "grad_norm": 0.3889034390449524, "learning_rate": 9.477516789397189e-05, "loss": 0.0252, "step": 7580 }, { "grad_norm": 0.261442631483078, "learning_rate": 9.475675547919983e-05, "loss": 0.0202, "step": 7590 }, { "grad_norm": 0.2744413912296295, "learning_rate": 9.47383124736074e-05, "loss": 0.0225, "step": 7600 }, { "grad_norm": 0.2939872741699219, "learning_rate": 9.471983888980028e-05, "loss": 0.0228, "step": 7610 }, { "grad_norm": 0.38369423151016235, "learning_rate": 9.470133474040494e-05, "loss": 0.0269, "step": 7620 }, { "grad_norm": 0.246125727891922, "learning_rate": 9.468280003806882e-05, "loss": 0.0226, "step": 7630 }, { "grad_norm": 0.2650824785232544, "learning_rate": 9.46642347954602e-05, "loss": 0.0275, "step": 7640 }, { "grad_norm": 0.3302931487560272, "learning_rate": 9.464563902526826e-05, "loss": 0.0249, "step": 7650 }, { "grad_norm": 0.3329029381275177, "learning_rate": 9.462701274020303e-05, "loss": 0.0201, "step": 7660 }, { "grad_norm": 0.2561056911945343, "learning_rate": 9.46083559529954e-05, "loss": 0.0215, "step": 7670 }, { "grad_norm": 0.40313318371772766, "learning_rate": 9.458966867639712e-05, "loss": 0.0274, "step": 7680 }, { "grad_norm": 0.4033510684967041, "learning_rate": 9.457095092318074e-05, "loss": 0.0293, "step": 7690 }, { "grad_norm": 0.46851760149002075, "learning_rate": 9.455220270613969e-05, "loss": 0.0274, "step": 7700 }, { "grad_norm": 0.4006962478160858, "learning_rate": 9.45334240380882e-05, "loss": 0.0301, "step": 7710 }, { "grad_norm": 0.2959243059158325, "learning_rate": 9.451461493186129e-05, "loss": 0.0232, "step": 7720 }, { "grad_norm": 0.2678012549877167, "learning_rate": 9.449577540031482e-05, "loss": 0.028, "step": 7730 }, { "grad_norm": 0.3119176924228668, "learning_rate": 9.447690545632544e-05, "loss": 0.0239, "step": 7740 }, { "grad_norm": 0.2928768992424011, "learning_rate": 9.445800511279058e-05, "loss": 0.023, "step": 7750 }, { "grad_norm": 0.26890304684638977, "learning_rate": 9.443907438262844e-05, "loss": 0.0249, "step": 7760 }, { "grad_norm": 0.2803957462310791, "learning_rate": 9.442011327877798e-05, "loss": 0.0223, "step": 7770 }, { "grad_norm": 0.57781982421875, "learning_rate": 9.440112181419895e-05, "loss": 0.0292, "step": 7780 }, { "grad_norm": 0.3566320538520813, "learning_rate": 9.438210000187186e-05, "loss": 0.0281, "step": 7790 }, { "grad_norm": 0.21340058743953705, "learning_rate": 9.43630478547979e-05, "loss": 0.0256, "step": 7800 }, { "grad_norm": 0.3381194770336151, "learning_rate": 9.434396538599906e-05, "loss": 0.0296, "step": 7810 }, { "grad_norm": 0.23844532668590546, "learning_rate": 9.4324852608518e-05, "loss": 0.0245, "step": 7820 }, { "grad_norm": 0.24216328561306, "learning_rate": 9.430570953541816e-05, "loss": 0.023, "step": 7830 }, { "grad_norm": 0.3509272336959839, "learning_rate": 9.428653617978362e-05, "loss": 0.0239, "step": 7840 }, { "grad_norm": 0.36090847849845886, "learning_rate": 9.42673325547192e-05, "loss": 0.0204, "step": 7850 }, { "grad_norm": 0.22855381667613983, "learning_rate": 9.42480986733504e-05, "loss": 0.0214, "step": 7860 }, { "grad_norm": 0.2392604798078537, "learning_rate": 9.422883454882338e-05, "loss": 0.0225, "step": 7870 }, { "grad_norm": 0.26393553614616394, "learning_rate": 9.4209540194305e-05, "loss": 0.0266, "step": 7880 }, { "grad_norm": 0.24921749532222748, "learning_rate": 9.419021562298278e-05, "loss": 0.0212, "step": 7890 }, { "grad_norm": 0.2863638699054718, "learning_rate": 9.417086084806486e-05, "loss": 0.0217, "step": 7900 }, { "grad_norm": 0.27603864669799805, "learning_rate": 9.415147588278005e-05, "loss": 0.0245, "step": 7910 }, { "grad_norm": 0.30620500445365906, "learning_rate": 9.413206074037781e-05, "loss": 0.0205, "step": 7920 }, { "grad_norm": 0.2690567970275879, "learning_rate": 9.411261543412819e-05, "loss": 0.0235, "step": 7930 }, { "grad_norm": 0.2728281319141388, "learning_rate": 9.409313997732188e-05, "loss": 0.0241, "step": 7940 }, { "grad_norm": 0.3044789731502533, "learning_rate": 9.407363438327018e-05, "loss": 0.0196, "step": 7950 }, { "grad_norm": 0.36006563901901245, "learning_rate": 9.405409866530496e-05, "loss": 0.0225, "step": 7960 }, { "grad_norm": 0.347628116607666, "learning_rate": 9.40345328367787e-05, "loss": 0.0232, "step": 7970 }, { "grad_norm": 0.38302546739578247, "learning_rate": 9.401493691106446e-05, "loss": 0.025, "step": 7980 }, { "grad_norm": 0.33031973242759705, "learning_rate": 9.39953109015559e-05, "loss": 0.0209, "step": 7990 }, { "grad_norm": 0.32550278306007385, "learning_rate": 9.397565482166718e-05, "loss": 0.0216, "step": 8000 }, { "grad_norm": 0.3057543635368347, "learning_rate": 9.395596868483305e-05, "loss": 0.0239, "step": 8010 }, { "grad_norm": 0.3193571865558624, "learning_rate": 9.393625250450881e-05, "loss": 0.0267, "step": 8020 }, { "grad_norm": 0.38304442167282104, "learning_rate": 9.391650629417028e-05, "loss": 0.0215, "step": 8030 }, { "grad_norm": 0.37836164236068726, "learning_rate": 9.38967300673138e-05, "loss": 0.0246, "step": 8040 }, { "grad_norm": 0.27877843379974365, "learning_rate": 9.387692383745624e-05, "loss": 0.0223, "step": 8050 }, { "grad_norm": 0.25082525610923767, "learning_rate": 9.385708761813496e-05, "loss": 0.0231, "step": 8060 }, { "grad_norm": 0.28011199831962585, "learning_rate": 9.383722142290787e-05, "loss": 0.0245, "step": 8070 }, { "grad_norm": 0.31286248564720154, "learning_rate": 9.381732526535328e-05, "loss": 0.0242, "step": 8080 }, { "grad_norm": 0.2945137619972229, "learning_rate": 9.379739915907004e-05, "loss": 0.0232, "step": 8090 }, { "grad_norm": 0.2731167674064636, "learning_rate": 9.377744311767746e-05, "loss": 0.0262, "step": 8100 }, { "grad_norm": 0.3126373291015625, "learning_rate": 9.375745715481532e-05, "loss": 0.0292, "step": 8110 }, { "grad_norm": 0.31937429308891296, "learning_rate": 9.373744128414382e-05, "loss": 0.0242, "step": 8120 }, { "grad_norm": 0.30470988154411316, "learning_rate": 9.371739551934359e-05, "loss": 0.0249, "step": 8130 }, { "grad_norm": 0.34936606884002686, "learning_rate": 9.36973198741158e-05, "loss": 0.0274, "step": 8140 }, { "grad_norm": 0.23686762154102325, "learning_rate": 9.367721436218189e-05, "loss": 0.0283, "step": 8150 }, { "grad_norm": 0.36879727244377136, "learning_rate": 9.365707899728383e-05, "loss": 0.0264, "step": 8160 }, { "grad_norm": 0.2814793586730957, "learning_rate": 9.363691379318391e-05, "loss": 0.0234, "step": 8170 }, { "grad_norm": 0.3719989061355591, "learning_rate": 9.361671876366491e-05, "loss": 0.0236, "step": 8180 }, { "grad_norm": 0.2962248623371124, "learning_rate": 9.35964939225299e-05, "loss": 0.0238, "step": 8190 }, { "grad_norm": 0.27038657665252686, "learning_rate": 9.35762392836024e-05, "loss": 0.0195, "step": 8200 }, { "grad_norm": 0.2760765552520752, "learning_rate": 9.355595486072624e-05, "loss": 0.021, "step": 8210 }, { "grad_norm": 0.20536169409751892, "learning_rate": 9.353564066776563e-05, "loss": 0.0201, "step": 8220 }, { "grad_norm": 0.30800917744636536, "learning_rate": 9.351529671860516e-05, "loss": 0.0196, "step": 8230 }, { "grad_norm": 0.28578653931617737, "learning_rate": 9.349492302714969e-05, "loss": 0.0191, "step": 8240 }, { "grad_norm": 0.24491357803344727, "learning_rate": 9.347451960732447e-05, "loss": 0.02, "step": 8250 }, { "grad_norm": 0.3249106705188751, "learning_rate": 9.345408647307506e-05, "loss": 0.0209, "step": 8260 }, { "grad_norm": 0.26218241453170776, "learning_rate": 9.343362363836726e-05, "loss": 0.0225, "step": 8270 }, { "grad_norm": 0.20200441777706146, "learning_rate": 9.341313111718728e-05, "loss": 0.0217, "step": 8280 }, { "grad_norm": 0.23754185438156128, "learning_rate": 9.339260892354153e-05, "loss": 0.0215, "step": 8290 }, { "grad_norm": 0.21926580369472504, "learning_rate": 9.337205707145676e-05, "loss": 0.0195, "step": 8300 }, { "grad_norm": 0.30926090478897095, "learning_rate": 9.335147557497994e-05, "loss": 0.0236, "step": 8310 }, { "grad_norm": 0.28488072752952576, "learning_rate": 9.333086444817835e-05, "loss": 0.0212, "step": 8320 }, { "grad_norm": 0.27310556173324585, "learning_rate": 9.331022370513951e-05, "loss": 0.0213, "step": 8330 }, { "grad_norm": 0.34358054399490356, "learning_rate": 9.328955335997113e-05, "loss": 0.0238, "step": 8340 }, { "grad_norm": 0.2656618356704712, "learning_rate": 9.326885342680125e-05, "loss": 0.0219, "step": 8350 }, { "grad_norm": 0.30053600668907166, "learning_rate": 9.324812391977806e-05, "loss": 0.0212, "step": 8360 }, { "grad_norm": 0.30504316091537476, "learning_rate": 9.322736485306995e-05, "loss": 0.021, "step": 8370 }, { "grad_norm": 0.6061001420021057, "learning_rate": 9.32065762408656e-05, "loss": 0.0223, "step": 8380 }, { "grad_norm": 0.3934614658355713, "learning_rate": 9.318575809737377e-05, "loss": 0.0262, "step": 8390 }, { "grad_norm": 0.2864987850189209, "learning_rate": 9.316491043682351e-05, "loss": 0.0235, "step": 8400 }, { "grad_norm": 0.2924745976924896, "learning_rate": 9.3144033273464e-05, "loss": 0.0232, "step": 8410 }, { "grad_norm": 0.3007837235927582, "learning_rate": 9.312312662156457e-05, "loss": 0.0271, "step": 8420 }, { "grad_norm": 0.2871954143047333, "learning_rate": 9.310219049541471e-05, "loss": 0.024, "step": 8430 }, { "grad_norm": 0.3768567442893982, "learning_rate": 9.308122490932409e-05, "loss": 0.0298, "step": 8440 }, { "grad_norm": 0.3818131387233734, "learning_rate": 9.306022987762246e-05, "loss": 0.031, "step": 8450 }, { "grad_norm": 0.3188258111476898, "learning_rate": 9.303920541465974e-05, "loss": 0.0267, "step": 8460 }, { "grad_norm": 0.24169369041919708, "learning_rate": 9.301815153480598e-05, "loss": 0.0272, "step": 8470 }, { "grad_norm": 0.2634790241718292, "learning_rate": 9.299706825245126e-05, "loss": 0.023, "step": 8480 }, { "grad_norm": 0.2601926326751709, "learning_rate": 9.297595558200583e-05, "loss": 0.0235, "step": 8490 }, { "grad_norm": 0.3019915521144867, "learning_rate": 9.295481353790001e-05, "loss": 0.0241, "step": 8500 }, { "grad_norm": 0.23916202783584595, "learning_rate": 9.293364213458417e-05, "loss": 0.0246, "step": 8510 }, { "grad_norm": 0.2722492814064026, "learning_rate": 9.291244138652877e-05, "loss": 0.0239, "step": 8520 }, { "grad_norm": 0.2814244329929352, "learning_rate": 9.289121130822431e-05, "loss": 0.0241, "step": 8530 }, { "grad_norm": 0.3709714412689209, "learning_rate": 9.286995191418137e-05, "loss": 0.0235, "step": 8540 }, { "grad_norm": 0.2998107075691223, "learning_rate": 9.284866321893056e-05, "loss": 0.0234, "step": 8550 }, { "grad_norm": 0.27257293462753296, "learning_rate": 9.282734523702244e-05, "loss": 0.0233, "step": 8560 }, { "grad_norm": 0.26328161358833313, "learning_rate": 9.280599798302771e-05, "loss": 0.0173, "step": 8570 }, { "grad_norm": 0.3750743567943573, "learning_rate": 9.278462147153699e-05, "loss": 0.0223, "step": 8580 }, { "grad_norm": 0.3031505048274994, "learning_rate": 9.276321571716094e-05, "loss": 0.0201, "step": 8590 }, { "grad_norm": 0.27757030725479126, "learning_rate": 9.274178073453017e-05, "loss": 0.0189, "step": 8600 }, { "grad_norm": 0.3134070634841919, "learning_rate": 9.27203165382953e-05, "loss": 0.0209, "step": 8610 }, { "grad_norm": 0.25488704442977905, "learning_rate": 9.26988231431269e-05, "loss": 0.0215, "step": 8620 }, { "grad_norm": 0.3287521302700043, "learning_rate": 9.267730056371551e-05, "loss": 0.0185, "step": 8630 }, { "grad_norm": 0.3686034083366394, "learning_rate": 9.265574881477161e-05, "loss": 0.0212, "step": 8640 }, { "grad_norm": 0.22122502326965332, "learning_rate": 9.263416791102561e-05, "loss": 0.0248, "step": 8650 }, { "grad_norm": 0.268964558839798, "learning_rate": 9.261255786722786e-05, "loss": 0.0237, "step": 8660 }, { "grad_norm": 0.3201062083244324, "learning_rate": 9.259091869814864e-05, "loss": 0.0243, "step": 8670 }, { "grad_norm": 0.29489755630493164, "learning_rate": 9.256925041857807e-05, "loss": 0.0239, "step": 8680 }, { "grad_norm": 0.2608465850353241, "learning_rate": 9.254755304332626e-05, "loss": 0.0233, "step": 8690 }, { "grad_norm": 0.2776874899864197, "learning_rate": 9.252582658722313e-05, "loss": 0.0199, "step": 8700 }, { "grad_norm": 0.2557159662246704, "learning_rate": 9.250407106511853e-05, "loss": 0.0201, "step": 8710 }, { "grad_norm": 0.25940659642219543, "learning_rate": 9.248228649188215e-05, "loss": 0.0203, "step": 8720 }, { "grad_norm": 0.3397485315799713, "learning_rate": 9.246047288240354e-05, "loss": 0.0273, "step": 8730 }, { "grad_norm": 0.3333941698074341, "learning_rate": 9.243863025159208e-05, "loss": 0.0231, "step": 8740 }, { "grad_norm": 0.26720911264419556, "learning_rate": 9.241675861437704e-05, "loss": 0.0251, "step": 8750 }, { "grad_norm": 0.2728726267814636, "learning_rate": 9.239485798570742e-05, "loss": 0.0212, "step": 8760 }, { "grad_norm": 0.3979912996292114, "learning_rate": 9.237292838055215e-05, "loss": 0.0203, "step": 8770 }, { "grad_norm": 0.32767125964164734, "learning_rate": 9.235096981389988e-05, "loss": 0.0224, "step": 8780 }, { "grad_norm": 0.31145599484443665, "learning_rate": 9.232898230075909e-05, "loss": 0.0211, "step": 8790 }, { "grad_norm": 0.3739616870880127, "learning_rate": 9.230696585615801e-05, "loss": 0.0294, "step": 8800 }, { "grad_norm": 0.32902222871780396, "learning_rate": 9.228492049514473e-05, "loss": 0.0249, "step": 8810 }, { "grad_norm": 0.22337312996387482, "learning_rate": 9.2262846232787e-05, "loss": 0.0256, "step": 8820 }, { "grad_norm": 0.371844083070755, "learning_rate": 9.224074308417238e-05, "loss": 0.0286, "step": 8830 }, { "grad_norm": 0.34754884243011475, "learning_rate": 9.221861106440818e-05, "loss": 0.0265, "step": 8840 }, { "grad_norm": 0.3135865330696106, "learning_rate": 9.219645018862142e-05, "loss": 0.0262, "step": 8850 }, { "grad_norm": 0.33422181010246277, "learning_rate": 9.217426047195882e-05, "loss": 0.0226, "step": 8860 }, { "grad_norm": 0.23427367210388184, "learning_rate": 9.21520419295869e-05, "loss": 0.0228, "step": 8870 }, { "grad_norm": 0.31302115321159363, "learning_rate": 9.212979457669178e-05, "loss": 0.0202, "step": 8880 }, { "grad_norm": 0.30912432074546814, "learning_rate": 9.210751842847931e-05, "loss": 0.0233, "step": 8890 }, { "grad_norm": 0.2806941866874695, "learning_rate": 9.208521350017506e-05, "loss": 0.0178, "step": 8900 }, { "grad_norm": 0.3594443202018738, "learning_rate": 9.206287980702422e-05, "loss": 0.0221, "step": 8910 }, { "grad_norm": 0.2918371856212616, "learning_rate": 9.204051736429166e-05, "loss": 0.0259, "step": 8920 }, { "grad_norm": 0.33134064078330994, "learning_rate": 9.20181261872619e-05, "loss": 0.022, "step": 8930 }, { "grad_norm": 0.315949946641922, "learning_rate": 9.19957062912391e-05, "loss": 0.0215, "step": 8940 }, { "grad_norm": 0.36643534898757935, "learning_rate": 9.197325769154704e-05, "loss": 0.0229, "step": 8950 }, { "grad_norm": 0.29092341661453247, "learning_rate": 9.195078040352914e-05, "loss": 0.0242, "step": 8960 }, { "grad_norm": 0.2657340168952942, "learning_rate": 9.192827444254839e-05, "loss": 0.0239, "step": 8970 }, { "grad_norm": 0.272122323513031, "learning_rate": 9.190573982398744e-05, "loss": 0.0294, "step": 8980 }, { "grad_norm": 0.36400991678237915, "learning_rate": 9.188317656324845e-05, "loss": 0.0274, "step": 8990 }, { "grad_norm": 0.2967469096183777, "learning_rate": 9.186058467575322e-05, "loss": 0.029, "step": 9000 }, { "grad_norm": 0.2879854440689087, "learning_rate": 9.183796417694309e-05, "loss": 0.0261, "step": 9010 }, { "grad_norm": 0.40993720293045044, "learning_rate": 9.181531508227897e-05, "loss": 0.0257, "step": 9020 }, { "grad_norm": 0.33080819249153137, "learning_rate": 9.179263740724125e-05, "loss": 0.0297, "step": 9030 }, { "grad_norm": 0.30857032537460327, "learning_rate": 9.176993116732996e-05, "loss": 0.027, "step": 9040 }, { "grad_norm": 0.31333398818969727, "learning_rate": 9.17471963780646e-05, "loss": 0.0234, "step": 9050 }, { "grad_norm": 0.2722316384315491, "learning_rate": 9.172443305498414e-05, "loss": 0.0237, "step": 9060 }, { "grad_norm": 0.31908679008483887, "learning_rate": 9.170164121364714e-05, "loss": 0.0289, "step": 9070 }, { "grad_norm": 0.2415040135383606, "learning_rate": 9.167882086963158e-05, "loss": 0.0247, "step": 9080 }, { "grad_norm": 0.32852450013160706, "learning_rate": 9.165597203853499e-05, "loss": 0.023, "step": 9090 }, { "grad_norm": 0.34284457564353943, "learning_rate": 9.163309473597428e-05, "loss": 0.0227, "step": 9100 }, { "grad_norm": 0.5546903014183044, "learning_rate": 9.161018897758591e-05, "loss": 0.0275, "step": 9110 }, { "grad_norm": 0.2723963260650635, "learning_rate": 9.158725477902572e-05, "loss": 0.0261, "step": 9120 }, { "grad_norm": 0.24924980103969574, "learning_rate": 9.156429215596905e-05, "loss": 0.0237, "step": 9130 }, { "grad_norm": 0.27461233735084534, "learning_rate": 9.154130112411061e-05, "loss": 0.0248, "step": 9140 }, { "grad_norm": 0.20752064883708954, "learning_rate": 9.151828169916455e-05, "loss": 0.02, "step": 9150 }, { "grad_norm": 0.34056663513183594, "learning_rate": 9.149523389686445e-05, "loss": 0.0209, "step": 9160 }, { "grad_norm": 0.27026763558387756, "learning_rate": 9.147215773296327e-05, "loss": 0.0231, "step": 9170 }, { "grad_norm": 0.32843104004859924, "learning_rate": 9.144905322323334e-05, "loss": 0.0207, "step": 9180 }, { "grad_norm": 0.30195069313049316, "learning_rate": 9.142592038346635e-05, "loss": 0.0203, "step": 9190 }, { "grad_norm": 0.2826494872570038, "learning_rate": 9.140275922947343e-05, "loss": 0.0214, "step": 9200 }, { "grad_norm": 0.30520254373550415, "learning_rate": 9.137956977708496e-05, "loss": 0.0216, "step": 9210 }, { "grad_norm": 0.4038359820842743, "learning_rate": 9.135635204215075e-05, "loss": 0.0243, "step": 9220 }, { "grad_norm": 0.2647936940193176, "learning_rate": 9.133310604053986e-05, "loss": 0.0197, "step": 9230 }, { "grad_norm": 0.2838312089443207, "learning_rate": 9.130983178814077e-05, "loss": 0.022, "step": 9240 }, { "grad_norm": 0.30107539892196655, "learning_rate": 9.128652930086116e-05, "loss": 0.0196, "step": 9250 }, { "grad_norm": 0.27773764729499817, "learning_rate": 9.126319859462808e-05, "loss": 0.022, "step": 9260 }, { "grad_norm": 0.27599918842315674, "learning_rate": 9.123983968538787e-05, "loss": 0.0211, "step": 9270 }, { "grad_norm": 0.3079123795032501, "learning_rate": 9.121645258910609e-05, "loss": 0.0242, "step": 9280 }, { "grad_norm": 0.2594890892505646, "learning_rate": 9.119303732176762e-05, "loss": 0.0202, "step": 9290 }, { "grad_norm": 0.3679283559322357, "learning_rate": 9.116959389937656e-05, "loss": 0.0223, "step": 9300 }, { "grad_norm": 0.30648383498191833, "learning_rate": 9.114612233795628e-05, "loss": 0.0209, "step": 9310 }, { "grad_norm": 0.2742466628551483, "learning_rate": 9.112262265354937e-05, "loss": 0.0214, "step": 9320 }, { "grad_norm": 0.3279174566268921, "learning_rate": 9.109909486221763e-05, "loss": 0.0205, "step": 9330 }, { "grad_norm": 0.24027320742607117, "learning_rate": 9.107553898004208e-05, "loss": 0.0192, "step": 9340 }, { "grad_norm": 0.23402681946754456, "learning_rate": 9.105195502312298e-05, "loss": 0.0231, "step": 9350 }, { "grad_norm": 0.2306092381477356, "learning_rate": 9.102834300757972e-05, "loss": 0.0206, "step": 9360 }, { "grad_norm": 0.2574482858181, "learning_rate": 9.100470294955087e-05, "loss": 0.0184, "step": 9370 }, { "grad_norm": 0.25825271010398865, "learning_rate": 9.098103486519423e-05, "loss": 0.0206, "step": 9380 }, { "grad_norm": 0.3102331757545471, "learning_rate": 9.095733877068666e-05, "loss": 0.0174, "step": 9390 }, { "grad_norm": 0.3421568274497986, "learning_rate": 9.093361468222426e-05, "loss": 0.0204, "step": 9400 }, { "grad_norm": 0.2031109631061554, "learning_rate": 9.090986261602221e-05, "loss": 0.0169, "step": 9410 }, { "grad_norm": 0.29651665687561035, "learning_rate": 9.088608258831482e-05, "loss": 0.0204, "step": 9420 }, { "grad_norm": 0.3126530051231384, "learning_rate": 9.08622746153555e-05, "loss": 0.0216, "step": 9430 }, { "grad_norm": 0.3007679879665375, "learning_rate": 9.083843871341679e-05, "loss": 0.0191, "step": 9440 }, { "grad_norm": 0.34244203567504883, "learning_rate": 9.08145748987903e-05, "loss": 0.0221, "step": 9450 }, { "grad_norm": 0.2908690869808197, "learning_rate": 9.079068318778674e-05, "loss": 0.0175, "step": 9460 }, { "grad_norm": 0.26639142632484436, "learning_rate": 9.076676359673586e-05, "loss": 0.0175, "step": 9470 }, { "grad_norm": 0.26679447293281555, "learning_rate": 9.074281614198645e-05, "loss": 0.0204, "step": 9480 }, { "grad_norm": 0.3058103621006012, "learning_rate": 9.071884083990638e-05, "loss": 0.0215, "step": 9490 }, { "grad_norm": 0.3446187973022461, "learning_rate": 9.069483770688259e-05, "loss": 0.0225, "step": 9500 }, { "grad_norm": 0.31324899196624756, "learning_rate": 9.067080675932093e-05, "loss": 0.0245, "step": 9510 }, { "grad_norm": 0.28332024812698364, "learning_rate": 9.064674801364634e-05, "loss": 0.0211, "step": 9520 }, { "grad_norm": 0.32847753167152405, "learning_rate": 9.062266148630279e-05, "loss": 0.0187, "step": 9530 }, { "grad_norm": 0.28180113434791565, "learning_rate": 9.059854719375313e-05, "loss": 0.0221, "step": 9540 }, { "grad_norm": 0.2491273432970047, "learning_rate": 9.057440515247932e-05, "loss": 0.0259, "step": 9550 }, { "grad_norm": 0.31000250577926636, "learning_rate": 9.055023537898215e-05, "loss": 0.0227, "step": 9560 }, { "grad_norm": 0.28854069113731384, "learning_rate": 9.052603788978149e-05, "loss": 0.0222, "step": 9570 }, { "grad_norm": 0.261015921831131, "learning_rate": 9.050181270141606e-05, "loss": 0.0227, "step": 9580 }, { "grad_norm": 0.3794471323490143, "learning_rate": 9.047755983044355e-05, "loss": 0.0243, "step": 9590 }, { "grad_norm": 0.3174976110458374, "learning_rate": 9.045327929344058e-05, "loss": 0.0183, "step": 9600 }, { "grad_norm": 0.32054170966148376, "learning_rate": 9.042897110700268e-05, "loss": 0.022, "step": 9610 }, { "grad_norm": 0.21991439163684845, "learning_rate": 9.040463528774423e-05, "loss": 0.0172, "step": 9620 }, { "grad_norm": 0.30816084146499634, "learning_rate": 9.038027185229856e-05, "loss": 0.0201, "step": 9630 }, { "grad_norm": 0.2386881560087204, "learning_rate": 9.035588081731784e-05, "loss": 0.0243, "step": 9640 }, { "grad_norm": 0.35595861077308655, "learning_rate": 9.033146219947313e-05, "loss": 0.0255, "step": 9650 }, { "grad_norm": 0.23816582560539246, "learning_rate": 9.030701601545429e-05, "loss": 0.0246, "step": 9660 }, { "grad_norm": 0.28731852769851685, "learning_rate": 9.028254228197009e-05, "loss": 0.0198, "step": 9670 }, { "grad_norm": 0.42923617362976074, "learning_rate": 9.025804101574808e-05, "loss": 0.0186, "step": 9680 }, { "grad_norm": 0.33992618322372437, "learning_rate": 9.023351223353463e-05, "loss": 0.0231, "step": 9690 }, { "grad_norm": 0.33790022134780884, "learning_rate": 9.020895595209498e-05, "loss": 0.0224, "step": 9700 }, { "grad_norm": 0.22800271213054657, "learning_rate": 9.018437218821305e-05, "loss": 0.021, "step": 9710 }, { "grad_norm": 0.2129819691181183, "learning_rate": 9.015976095869166e-05, "loss": 0.0195, "step": 9720 }, { "grad_norm": 0.3659260869026184, "learning_rate": 9.013512228035235e-05, "loss": 0.0198, "step": 9730 }, { "grad_norm": 0.3038254976272583, "learning_rate": 9.01104561700354e-05, "loss": 0.0214, "step": 9740 }, { "grad_norm": 0.3098638653755188, "learning_rate": 9.008576264459987e-05, "loss": 0.022, "step": 9750 }, { "grad_norm": 0.28229770064353943, "learning_rate": 9.006104172092357e-05, "loss": 0.0212, "step": 9760 }, { "grad_norm": 0.3240756094455719, "learning_rate": 9.0036293415903e-05, "loss": 0.0209, "step": 9770 }, { "grad_norm": 0.29588645696640015, "learning_rate": 9.00115177464534e-05, "loss": 0.018, "step": 9780 }, { "grad_norm": 0.25745847821235657, "learning_rate": 8.998671472950872e-05, "loss": 0.018, "step": 9790 }, { "grad_norm": 0.2794647514820099, "learning_rate": 8.996188438202156e-05, "loss": 0.0204, "step": 9800 }, { "grad_norm": 0.2639206349849701, "learning_rate": 8.993702672096324e-05, "loss": 0.0267, "step": 9810 }, { "grad_norm": 0.42432236671447754, "learning_rate": 8.991214176332377e-05, "loss": 0.0216, "step": 9820 }, { "grad_norm": 0.479707807302475, "learning_rate": 8.988722952611171e-05, "loss": 0.0259, "step": 9830 }, { "grad_norm": 0.30785858631134033, "learning_rate": 8.986229002635441e-05, "loss": 0.026, "step": 9840 }, { "grad_norm": 0.26946720480918884, "learning_rate": 8.983732328109775e-05, "loss": 0.0243, "step": 9850 }, { "grad_norm": 0.23735150694847107, "learning_rate": 8.981232930740627e-05, "loss": 0.0213, "step": 9860 }, { "grad_norm": 0.17269600927829742, "learning_rate": 8.978730812236313e-05, "loss": 0.0164, "step": 9870 }, { "grad_norm": 0.3725918233394623, "learning_rate": 8.976225974307004e-05, "loss": 0.0184, "step": 9880 }, { "grad_norm": 0.2276921570301056, "learning_rate": 8.973718418664736e-05, "loss": 0.0204, "step": 9890 }, { "grad_norm": 0.31990763545036316, "learning_rate": 8.971208147023397e-05, "loss": 0.0191, "step": 9900 }, { "grad_norm": 0.24362117052078247, "learning_rate": 8.968695161098739e-05, "loss": 0.0178, "step": 9910 }, { "grad_norm": 0.22522203624248505, "learning_rate": 8.96617946260836e-05, "loss": 0.0183, "step": 9920 }, { "grad_norm": 0.3107154071331024, "learning_rate": 8.963661053271716e-05, "loss": 0.019, "step": 9930 }, { "grad_norm": 0.3076677918434143, "learning_rate": 8.961139934810117e-05, "loss": 0.0147, "step": 9940 }, { "grad_norm": 0.28497597575187683, "learning_rate": 8.958616108946725e-05, "loss": 0.0223, "step": 9950 }, { "grad_norm": 0.2254512906074524, "learning_rate": 8.95608957740655e-05, "loss": 0.0176, "step": 9960 }, { "grad_norm": 0.2308811992406845, "learning_rate": 8.953560341916452e-05, "loss": 0.0164, "step": 9970 }, { "grad_norm": 0.2263970971107483, "learning_rate": 8.95102840420514e-05, "loss": 0.0152, "step": 9980 }, { "grad_norm": 0.2766692340373993, "learning_rate": 8.94849376600317e-05, "loss": 0.0169, "step": 9990 }, { "grad_norm": 0.31766477227211, "learning_rate": 8.945956429042943e-05, "loss": 0.0178, "step": 10000 }, { "grad_norm": 0.29792314767837524, "learning_rate": 8.943416395058705e-05, "loss": 0.0171, "step": 10010 }, { "grad_norm": 0.2503995895385742, "learning_rate": 8.940873665786544e-05, "loss": 0.0194, "step": 10020 }, { "grad_norm": 0.31836411356925964, "learning_rate": 8.938328242964394e-05, "loss": 0.027, "step": 10030 }, { "grad_norm": 0.35540130734443665, "learning_rate": 8.935780128332026e-05, "loss": 0.0201, "step": 10040 }, { "grad_norm": 0.25901898741722107, "learning_rate": 8.933229323631052e-05, "loss": 0.0201, "step": 10050 }, { "grad_norm": 0.3097890317440033, "learning_rate": 8.930675830604925e-05, "loss": 0.0174, "step": 10060 }, { "grad_norm": 0.205769345164299, "learning_rate": 8.92811965099893e-05, "loss": 0.0182, "step": 10070 }, { "grad_norm": 0.267990380525589, "learning_rate": 8.925560786560194e-05, "loss": 0.0236, "step": 10080 }, { "grad_norm": 0.26998281478881836, "learning_rate": 8.922999239037677e-05, "loss": 0.0185, "step": 10090 }, { "grad_norm": 0.22766414284706116, "learning_rate": 8.920435010182171e-05, "loss": 0.0166, "step": 10100 }, { "grad_norm": 0.27738210558891296, "learning_rate": 8.917868101746302e-05, "loss": 0.0157, "step": 10110 }, { "grad_norm": 0.27550092339515686, "learning_rate": 8.91529851548453e-05, "loss": 0.0189, "step": 10120 }, { "grad_norm": 0.21589022874832153, "learning_rate": 8.912726253153142e-05, "loss": 0.0199, "step": 10130 }, { "grad_norm": 0.2924022972583771, "learning_rate": 8.910151316510255e-05, "loss": 0.0176, "step": 10140 }, { "grad_norm": 0.33295124769210815, "learning_rate": 8.907573707315813e-05, "loss": 0.0169, "step": 10150 }, { "grad_norm": 0.2818719148635864, "learning_rate": 8.904993427331588e-05, "loss": 0.0213, "step": 10160 }, { "grad_norm": 0.2982214391231537, "learning_rate": 8.902410478321176e-05, "loss": 0.0218, "step": 10170 }, { "grad_norm": 0.6453559398651123, "learning_rate": 8.899824862050002e-05, "loss": 0.0218, "step": 10180 }, { "grad_norm": 0.30061057209968567, "learning_rate": 8.897236580285308e-05, "loss": 0.0205, "step": 10190 }, { "grad_norm": 0.36885517835617065, "learning_rate": 8.894645634796159e-05, "loss": 0.0244, "step": 10200 }, { "grad_norm": 0.3002922534942627, "learning_rate": 8.892052027353444e-05, "loss": 0.0252, "step": 10210 }, { "grad_norm": 0.31840160489082336, "learning_rate": 8.889455759729866e-05, "loss": 0.0257, "step": 10220 }, { "grad_norm": 0.34565749764442444, "learning_rate": 8.886856833699955e-05, "loss": 0.0232, "step": 10230 }, { "grad_norm": 0.25653329491615295, "learning_rate": 8.884255251040046e-05, "loss": 0.0222, "step": 10240 }, { "grad_norm": 0.2973921597003937, "learning_rate": 8.8816510135283e-05, "loss": 0.0201, "step": 10250 }, { "grad_norm": 0.2952231466770172, "learning_rate": 8.879044122944688e-05, "loss": 0.0259, "step": 10260 }, { "grad_norm": 0.24726416170597076, "learning_rate": 8.876434581070996e-05, "loss": 0.0239, "step": 10270 }, { "grad_norm": 0.324343204498291, "learning_rate": 8.87382238969082e-05, "loss": 0.0231, "step": 10280 }, { "grad_norm": 0.3007510304450989, "learning_rate": 8.871207550589568e-05, "loss": 0.0204, "step": 10290 }, { "grad_norm": 0.2827986776828766, "learning_rate": 8.868590065554458e-05, "loss": 0.0212, "step": 10300 }, { "grad_norm": 0.2504040598869324, "learning_rate": 8.865969936374519e-05, "loss": 0.0205, "step": 10310 }, { "grad_norm": 0.2714235484600067, "learning_rate": 8.863347164840581e-05, "loss": 0.0201, "step": 10320 }, { "grad_norm": 0.2943003475666046, "learning_rate": 8.860721752745285e-05, "loss": 0.0212, "step": 10330 }, { "grad_norm": 0.21939675509929657, "learning_rate": 8.858093701883077e-05, "loss": 0.0185, "step": 10340 }, { "grad_norm": 0.26465746760368347, "learning_rate": 8.8554630140502e-05, "loss": 0.0239, "step": 10350 }, { "grad_norm": 0.2426869422197342, "learning_rate": 8.85282969104471e-05, "loss": 0.0216, "step": 10360 }, { "grad_norm": 0.281040757894516, "learning_rate": 8.850193734666456e-05, "loss": 0.02, "step": 10370 }, { "grad_norm": 0.7273247241973877, "learning_rate": 8.84755514671709e-05, "loss": 0.0187, "step": 10380 }, { "grad_norm": 0.23791298270225525, "learning_rate": 8.84491392900006e-05, "loss": 0.0219, "step": 10390 }, { "grad_norm": 0.3730471134185791, "learning_rate": 8.842270083320617e-05, "loss": 0.0277, "step": 10400 }, { "grad_norm": 0.35629770159721375, "learning_rate": 8.839623611485801e-05, "loss": 0.0253, "step": 10410 }, { "grad_norm": 0.25686943531036377, "learning_rate": 8.836974515304453e-05, "loss": 0.022, "step": 10420 }, { "grad_norm": 0.3103864789009094, "learning_rate": 8.834322796587204e-05, "loss": 0.0256, "step": 10430 }, { "grad_norm": 0.26124605536460876, "learning_rate": 8.831668457146478e-05, "loss": 0.0233, "step": 10440 }, { "grad_norm": 0.24605919420719147, "learning_rate": 8.829011498796493e-05, "loss": 0.0221, "step": 10450 }, { "grad_norm": 0.273269921541214, "learning_rate": 8.826351923353253e-05, "loss": 0.0268, "step": 10460 }, { "grad_norm": 0.2910027503967285, "learning_rate": 8.823689732634555e-05, "loss": 0.028, "step": 10470 }, { "grad_norm": 0.2835399806499481, "learning_rate": 8.82102492845998e-05, "loss": 0.0257, "step": 10480 }, { "grad_norm": 0.2964680790901184, "learning_rate": 8.818357512650896e-05, "loss": 0.025, "step": 10490 }, { "grad_norm": 0.34536075592041016, "learning_rate": 8.815687487030458e-05, "loss": 0.0228, "step": 10500 }, { "grad_norm": 0.44811156392097473, "learning_rate": 8.8130148534236e-05, "loss": 0.0285, "step": 10510 }, { "grad_norm": 0.32745155692100525, "learning_rate": 8.810339613657047e-05, "loss": 0.0323, "step": 10520 }, { "grad_norm": 0.28503286838531494, "learning_rate": 8.807661769559295e-05, "loss": 0.0232, "step": 10530 }, { "grad_norm": 0.3217129111289978, "learning_rate": 8.804981322960628e-05, "loss": 0.0239, "step": 10540 }, { "grad_norm": 0.27616575360298157, "learning_rate": 8.802298275693106e-05, "loss": 0.0255, "step": 10550 }, { "grad_norm": 0.2841913104057312, "learning_rate": 8.799612629590568e-05, "loss": 0.021, "step": 10560 }, { "grad_norm": 0.3041156530380249, "learning_rate": 8.796924386488624e-05, "loss": 0.023, "step": 10570 }, { "grad_norm": 0.23580271005630493, "learning_rate": 8.794233548224666e-05, "loss": 0.0202, "step": 10580 }, { "grad_norm": 0.2977970540523529, "learning_rate": 8.791540116637853e-05, "loss": 0.0249, "step": 10590 }, { "grad_norm": 0.43907707929611206, "learning_rate": 8.788844093569124e-05, "loss": 0.0254, "step": 10600 }, { "grad_norm": 0.23038145899772644, "learning_rate": 8.786145480861184e-05, "loss": 0.0249, "step": 10610 }, { "grad_norm": 0.284637451171875, "learning_rate": 8.783444280358507e-05, "loss": 0.0269, "step": 10620 }, { "grad_norm": 0.27748891711235046, "learning_rate": 8.780740493907342e-05, "loss": 0.0209, "step": 10630 }, { "grad_norm": 0.2858564853668213, "learning_rate": 8.778034123355698e-05, "loss": 0.0249, "step": 10640 }, { "grad_norm": 0.29742056131362915, "learning_rate": 8.775325170553357e-05, "loss": 0.022, "step": 10650 }, { "grad_norm": 0.25033801794052124, "learning_rate": 8.77261363735186e-05, "loss": 0.0214, "step": 10660 }, { "grad_norm": 0.23548564314842224, "learning_rate": 8.769899525604517e-05, "loss": 0.0186, "step": 10670 }, { "grad_norm": 0.39665958285331726, "learning_rate": 8.767182837166397e-05, "loss": 0.0185, "step": 10680 }, { "grad_norm": 0.24252761900424957, "learning_rate": 8.764463573894328e-05, "loss": 0.0189, "step": 10690 }, { "grad_norm": 0.24133914709091187, "learning_rate": 8.761741737646902e-05, "loss": 0.0225, "step": 10700 }, { "grad_norm": 0.3181835114955902, "learning_rate": 8.759017330284471e-05, "loss": 0.0212, "step": 10710 }, { "grad_norm": 0.24749727547168732, "learning_rate": 8.756290353669142e-05, "loss": 0.0219, "step": 10720 }, { "grad_norm": 0.33169689774513245, "learning_rate": 8.753560809664774e-05, "loss": 0.0202, "step": 10730 }, { "grad_norm": 0.2946588397026062, "learning_rate": 8.750828700136986e-05, "loss": 0.0206, "step": 10740 }, { "grad_norm": 0.2614910304546356, "learning_rate": 8.74809402695315e-05, "loss": 0.02, "step": 10750 }, { "grad_norm": 0.23079811036586761, "learning_rate": 8.745356791982391e-05, "loss": 0.0199, "step": 10760 }, { "grad_norm": 0.25305140018463135, "learning_rate": 8.742616997095578e-05, "loss": 0.0212, "step": 10770 }, { "grad_norm": 0.2551364004611969, "learning_rate": 8.739874644165341e-05, "loss": 0.0174, "step": 10780 }, { "grad_norm": 0.2598411440849304, "learning_rate": 8.737129735066048e-05, "loss": 0.0224, "step": 10790 }, { "grad_norm": 0.2879733741283417, "learning_rate": 8.734382271673821e-05, "loss": 0.0189, "step": 10800 }, { "grad_norm": 0.3198516070842743, "learning_rate": 8.731632255866525e-05, "loss": 0.0158, "step": 10810 }, { "grad_norm": 0.28900960087776184, "learning_rate": 8.728879689523767e-05, "loss": 0.021, "step": 10820 }, { "grad_norm": 0.3177364468574524, "learning_rate": 8.726124574526905e-05, "loss": 0.0197, "step": 10830 }, { "grad_norm": 0.31065669655799866, "learning_rate": 8.72336691275903e-05, "loss": 0.0166, "step": 10840 }, { "grad_norm": 0.24437315762043, "learning_rate": 8.720606706104979e-05, "loss": 0.0159, "step": 10850 }, { "grad_norm": 0.24875350296497345, "learning_rate": 8.71784395645133e-05, "loss": 0.021, "step": 10860 }, { "grad_norm": 0.25882142782211304, "learning_rate": 8.715078665686392e-05, "loss": 0.0187, "step": 10870 }, { "grad_norm": 0.282785028219223, "learning_rate": 8.712310835700218e-05, "loss": 0.0179, "step": 10880 }, { "grad_norm": 0.2441740483045578, "learning_rate": 8.709540468384591e-05, "loss": 0.0201, "step": 10890 }, { "grad_norm": 0.27429628372192383, "learning_rate": 8.706767565633033e-05, "loss": 0.0158, "step": 10900 }, { "grad_norm": 0.24232643842697144, "learning_rate": 8.7039921293408e-05, "loss": 0.0178, "step": 10910 }, { "grad_norm": 0.21512213349342346, "learning_rate": 8.70121416140487e-05, "loss": 0.0211, "step": 10920 }, { "grad_norm": 0.25655242800712585, "learning_rate": 8.698433663723962e-05, "loss": 0.0209, "step": 10930 }, { "grad_norm": 0.24901820719242096, "learning_rate": 8.695650638198518e-05, "loss": 0.02, "step": 10940 }, { "grad_norm": 0.21694032847881317, "learning_rate": 8.692865086730713e-05, "loss": 0.0155, "step": 10950 }, { "grad_norm": 0.283024400472641, "learning_rate": 8.69007701122444e-05, "loss": 0.0155, "step": 10960 }, { "grad_norm": 0.2541472315788269, "learning_rate": 8.687286413585328e-05, "loss": 0.0154, "step": 10970 }, { "grad_norm": 0.19784383475780487, "learning_rate": 8.684493295720719e-05, "loss": 0.0175, "step": 10980 }, { "grad_norm": 0.2858920097351074, "learning_rate": 8.681697659539685e-05, "loss": 0.0188, "step": 10990 }, { "grad_norm": 0.21711397171020508, "learning_rate": 8.678899506953019e-05, "loss": 0.0164, "step": 11000 }, { "grad_norm": 0.24877570569515228, "learning_rate": 8.676098839873227e-05, "loss": 0.018, "step": 11010 }, { "grad_norm": 0.22969305515289307, "learning_rate": 8.673295660214545e-05, "loss": 0.0175, "step": 11020 }, { "grad_norm": 0.2390137016773224, "learning_rate": 8.670489969892914e-05, "loss": 0.02, "step": 11030 }, { "grad_norm": 0.2634517550468445, "learning_rate": 8.667681770826e-05, "loss": 0.0167, "step": 11040 }, { "grad_norm": 0.32164087891578674, "learning_rate": 8.66487106493318e-05, "loss": 0.019, "step": 11050 }, { "grad_norm": 0.24975894391536713, "learning_rate": 8.662057854135544e-05, "loss": 0.0184, "step": 11060 }, { "grad_norm": 0.22839771211147308, "learning_rate": 8.659242140355897e-05, "loss": 0.018, "step": 11070 }, { "grad_norm": 0.23943349719047546, "learning_rate": 8.65642392551875e-05, "loss": 0.019, "step": 11080 }, { "grad_norm": 0.23711685836315155, "learning_rate": 8.65360321155033e-05, "loss": 0.017, "step": 11090 }, { "grad_norm": 0.23708224296569824, "learning_rate": 8.650780000378566e-05, "loss": 0.0185, "step": 11100 }, { "grad_norm": 0.2583813965320587, "learning_rate": 8.647954293933096e-05, "loss": 0.0168, "step": 11110 }, { "grad_norm": 0.19159670174121857, "learning_rate": 8.645126094145264e-05, "loss": 0.0172, "step": 11120 }, { "grad_norm": 0.17908143997192383, "learning_rate": 8.642295402948117e-05, "loss": 0.0169, "step": 11130 }, { "grad_norm": 0.2544824182987213, "learning_rate": 8.639462222276409e-05, "loss": 0.0183, "step": 11140 }, { "grad_norm": 0.283939927816391, "learning_rate": 8.636626554066589e-05, "loss": 0.0218, "step": 11150 }, { "grad_norm": 0.257051944732666, "learning_rate": 8.633788400256811e-05, "loss": 0.0177, "step": 11160 }, { "grad_norm": 0.3092561960220337, "learning_rate": 8.630947762786927e-05, "loss": 0.0182, "step": 11170 }, { "grad_norm": 0.2877769470214844, "learning_rate": 8.628104643598483e-05, "loss": 0.0205, "step": 11180 }, { "grad_norm": 0.27821916341781616, "learning_rate": 8.625259044634726e-05, "loss": 0.0191, "step": 11190 }, { "grad_norm": 0.1942397803068161, "learning_rate": 8.622410967840597e-05, "loss": 0.0181, "step": 11200 }, { "grad_norm": 0.3516511619091034, "learning_rate": 8.619560415162731e-05, "loss": 0.0218, "step": 11210 }, { "grad_norm": 0.33502620458602905, "learning_rate": 8.616707388549447e-05, "loss": 0.0176, "step": 11220 }, { "grad_norm": 0.2493496835231781, "learning_rate": 8.613851889950771e-05, "loss": 0.0173, "step": 11230 }, { "grad_norm": 0.23060613870620728, "learning_rate": 8.610993921318402e-05, "loss": 0.0156, "step": 11240 }, { "grad_norm": 0.2536062002182007, "learning_rate": 8.608133484605738e-05, "loss": 0.0176, "step": 11250 }, { "grad_norm": 0.37334680557250977, "learning_rate": 8.605270581767859e-05, "loss": 0.0214, "step": 11260 }, { "grad_norm": 0.35451823472976685, "learning_rate": 8.602405214761536e-05, "loss": 0.0249, "step": 11270 }, { "grad_norm": 0.24696768820285797, "learning_rate": 8.599537385545215e-05, "loss": 0.0182, "step": 11280 }, { "grad_norm": 0.21161037683486938, "learning_rate": 8.596667096079032e-05, "loss": 0.0171, "step": 11290 }, { "grad_norm": 0.2706752121448517, "learning_rate": 8.593794348324806e-05, "loss": 0.0179, "step": 11300 }, { "grad_norm": 0.2898147404193878, "learning_rate": 8.590919144246028e-05, "loss": 0.0241, "step": 11310 }, { "grad_norm": 0.33923056721687317, "learning_rate": 8.588041485807876e-05, "loss": 0.0182, "step": 11320 }, { "grad_norm": 0.21419532597064972, "learning_rate": 8.585161374977202e-05, "loss": 0.0204, "step": 11330 }, { "grad_norm": 0.2697407007217407, "learning_rate": 8.582278813722533e-05, "loss": 0.0221, "step": 11340 }, { "grad_norm": 0.2700018286705017, "learning_rate": 8.579393804014076e-05, "loss": 0.0204, "step": 11350 }, { "grad_norm": 0.2716735303401947, "learning_rate": 8.576506347823703e-05, "loss": 0.0168, "step": 11360 }, { "grad_norm": 0.29677027463912964, "learning_rate": 8.573616447124968e-05, "loss": 0.0189, "step": 11370 }, { "grad_norm": 0.27782851457595825, "learning_rate": 8.570724103893086e-05, "loss": 0.0199, "step": 11380 }, { "grad_norm": 0.2413795292377472, "learning_rate": 8.567829320104951e-05, "loss": 0.0176, "step": 11390 }, { "grad_norm": 0.2599356174468994, "learning_rate": 8.564932097739118e-05, "loss": 0.0162, "step": 11400 }, { "grad_norm": 0.21136751770973206, "learning_rate": 8.562032438775811e-05, "loss": 0.0178, "step": 11410 }, { "grad_norm": 0.24865679442882538, "learning_rate": 8.559130345196921e-05, "loss": 0.0202, "step": 11420 }, { "grad_norm": 0.16671067476272583, "learning_rate": 8.556225818986e-05, "loss": 0.0192, "step": 11430 }, { "grad_norm": 0.3253313899040222, "learning_rate": 8.553318862128265e-05, "loss": 0.0178, "step": 11440 }, { "grad_norm": 0.2460712343454361, "learning_rate": 8.550409476610593e-05, "loss": 0.0164, "step": 11450 }, { "grad_norm": 0.22152110934257507, "learning_rate": 8.547497664421522e-05, "loss": 0.0169, "step": 11460 }, { "grad_norm": 0.2510722875595093, "learning_rate": 8.544583427551249e-05, "loss": 0.0189, "step": 11470 }, { "grad_norm": 0.24909311532974243, "learning_rate": 8.541666767991628e-05, "loss": 0.0172, "step": 11480 }, { "grad_norm": 0.27610087394714355, "learning_rate": 8.538747687736166e-05, "loss": 0.0147, "step": 11490 }, { "grad_norm": 0.13699927926063538, "learning_rate": 8.535826188780026e-05, "loss": 0.0189, "step": 11500 }, { "grad_norm": 0.23326706886291504, "learning_rate": 8.532902273120029e-05, "loss": 0.0173, "step": 11510 }, { "grad_norm": 0.3244427442550659, "learning_rate": 8.52997594275464e-05, "loss": 0.021, "step": 11520 }, { "grad_norm": 0.3238634765148163, "learning_rate": 8.52704719968398e-05, "loss": 0.0191, "step": 11530 }, { "grad_norm": 0.31925472617149353, "learning_rate": 8.524116045909818e-05, "loss": 0.0208, "step": 11540 }, { "grad_norm": 0.24778801202774048, "learning_rate": 8.521182483435569e-05, "loss": 0.0172, "step": 11550 }, { "grad_norm": 0.2721612751483917, "learning_rate": 8.518246514266295e-05, "loss": 0.0229, "step": 11560 }, { "grad_norm": 0.34612545371055603, "learning_rate": 8.515308140408703e-05, "loss": 0.0218, "step": 11570 }, { "grad_norm": 0.29691097140312195, "learning_rate": 8.512367363871145e-05, "loss": 0.0175, "step": 11580 }, { "grad_norm": 0.24368806183338165, "learning_rate": 8.509424186663614e-05, "loss": 0.0176, "step": 11590 }, { "grad_norm": 0.2078675776720047, "learning_rate": 8.506478610797743e-05, "loss": 0.0168, "step": 11600 }, { "grad_norm": 0.23645898699760437, "learning_rate": 8.503530638286805e-05, "loss": 0.0169, "step": 11610 }, { "grad_norm": 0.29893749952316284, "learning_rate": 8.500580271145712e-05, "loss": 0.0218, "step": 11620 }, { "grad_norm": 0.20446628332138062, "learning_rate": 8.497627511391014e-05, "loss": 0.0158, "step": 11630 }, { "grad_norm": 0.3495118021965027, "learning_rate": 8.494672361040891e-05, "loss": 0.0173, "step": 11640 }, { "grad_norm": 0.24124416708946228, "learning_rate": 8.491714822115162e-05, "loss": 0.0172, "step": 11650 }, { "grad_norm": 0.24722379446029663, "learning_rate": 8.488754896635277e-05, "loss": 0.0167, "step": 11660 }, { "grad_norm": 0.24427548050880432, "learning_rate": 8.485792586624317e-05, "loss": 0.0149, "step": 11670 }, { "grad_norm": 0.2813063859939575, "learning_rate": 8.482827894106993e-05, "loss": 0.0151, "step": 11680 }, { "grad_norm": 0.3263258934020996, "learning_rate": 8.479860821109646e-05, "loss": 0.0189, "step": 11690 }, { "grad_norm": 0.23868468403816223, "learning_rate": 8.476891369660239e-05, "loss": 0.0182, "step": 11700 }, { "grad_norm": 0.29374393820762634, "learning_rate": 8.473919541788366e-05, "loss": 0.0163, "step": 11710 }, { "grad_norm": 0.22102773189544678, "learning_rate": 8.470945339525245e-05, "loss": 0.0202, "step": 11720 }, { "grad_norm": 0.18900831043720245, "learning_rate": 8.467968764903713e-05, "loss": 0.0145, "step": 11730 }, { "grad_norm": 0.23665930330753326, "learning_rate": 8.46498981995823e-05, "loss": 0.0169, "step": 11740 }, { "grad_norm": 0.28004100918769836, "learning_rate": 8.462008506724879e-05, "loss": 0.0177, "step": 11750 }, { "grad_norm": 0.2250245362520218, "learning_rate": 8.459024827241359e-05, "loss": 0.0181, "step": 11760 }, { "grad_norm": 0.2522324025630951, "learning_rate": 8.456038783546985e-05, "loss": 0.0167, "step": 11770 }, { "grad_norm": 0.25597524642944336, "learning_rate": 8.453050377682691e-05, "loss": 0.0209, "step": 11780 }, { "grad_norm": 0.27141866087913513, "learning_rate": 8.450059611691026e-05, "loss": 0.0171, "step": 11790 }, { "grad_norm": 0.2770340144634247, "learning_rate": 8.447066487616146e-05, "loss": 0.0206, "step": 11800 }, { "grad_norm": 0.287557989358902, "learning_rate": 8.444071007503826e-05, "loss": 0.0169, "step": 11810 }, { "grad_norm": 0.3411196768283844, "learning_rate": 8.441073173401449e-05, "loss": 0.0218, "step": 11820 }, { "grad_norm": 0.24131685495376587, "learning_rate": 8.438072987358006e-05, "loss": 0.0175, "step": 11830 }, { "grad_norm": 0.24745085835456848, "learning_rate": 8.435070451424094e-05, "loss": 0.0169, "step": 11840 }, { "grad_norm": 0.2678869962692261, "learning_rate": 8.432065567651919e-05, "loss": 0.0208, "step": 11850 }, { "grad_norm": 0.2378748655319214, "learning_rate": 8.429058338095291e-05, "loss": 0.0213, "step": 11860 }, { "grad_norm": 0.21940645575523376, "learning_rate": 8.426048764809624e-05, "loss": 0.0207, "step": 11870 }, { "grad_norm": 0.2921586334705353, "learning_rate": 8.423036849851932e-05, "loss": 0.0177, "step": 11880 }, { "grad_norm": 0.24591344594955444, "learning_rate": 8.42002259528083e-05, "loss": 0.019, "step": 11890 }, { "grad_norm": 0.28400811553001404, "learning_rate": 8.417006003156532e-05, "loss": 0.0255, "step": 11900 }, { "grad_norm": 0.24868111312389374, "learning_rate": 8.413987075540852e-05, "loss": 0.0169, "step": 11910 }, { "grad_norm": 0.322573721408844, "learning_rate": 8.4109658144972e-05, "loss": 0.0187, "step": 11920 }, { "grad_norm": 0.2865610718727112, "learning_rate": 8.407942222090573e-05, "loss": 0.0181, "step": 11930 }, { "grad_norm": 0.33976882696151733, "learning_rate": 8.404916300387576e-05, "loss": 0.0161, "step": 11940 }, { "grad_norm": 0.3077263832092285, "learning_rate": 8.401888051456391e-05, "loss": 0.0221, "step": 11950 }, { "grad_norm": 0.30882909893989563, "learning_rate": 8.398857477366803e-05, "loss": 0.0243, "step": 11960 }, { "grad_norm": 0.30852892994880676, "learning_rate": 8.395824580190178e-05, "loss": 0.019, "step": 11970 }, { "grad_norm": 0.31951117515563965, "learning_rate": 8.392789361999473e-05, "loss": 0.0232, "step": 11980 }, { "grad_norm": 0.26009342074394226, "learning_rate": 8.38975182486923e-05, "loss": 0.0185, "step": 11990 }, { "grad_norm": 0.27863144874572754, "learning_rate": 8.386711970875581e-05, "loss": 0.0184, "step": 12000 }, { "grad_norm": 0.28495749831199646, "learning_rate": 8.383669802096232e-05, "loss": 0.0196, "step": 12010 }, { "grad_norm": 0.3122473359107971, "learning_rate": 8.38062532061048e-05, "loss": 0.0186, "step": 12020 }, { "grad_norm": 0.2997320890426636, "learning_rate": 8.3775785284992e-05, "loss": 0.0191, "step": 12030 }, { "grad_norm": 0.2694462835788727, "learning_rate": 8.374529427844843e-05, "loss": 0.0203, "step": 12040 }, { "grad_norm": 0.2700246274471283, "learning_rate": 8.371478020731442e-05, "loss": 0.0164, "step": 12050 }, { "grad_norm": 0.3079600930213928, "learning_rate": 8.368424309244607e-05, "loss": 0.0176, "step": 12060 }, { "grad_norm": 0.274985671043396, "learning_rate": 8.365368295471517e-05, "loss": 0.0173, "step": 12070 }, { "grad_norm": 0.26069504022598267, "learning_rate": 8.362309981500931e-05, "loss": 0.0168, "step": 12080 }, { "grad_norm": 0.24339774250984192, "learning_rate": 8.359249369423177e-05, "loss": 0.0155, "step": 12090 }, { "grad_norm": 0.27321189641952515, "learning_rate": 8.356186461330155e-05, "loss": 0.0188, "step": 12100 }, { "grad_norm": 0.26037314534187317, "learning_rate": 8.353121259315334e-05, "loss": 0.0163, "step": 12110 }, { "grad_norm": 0.19435349106788635, "learning_rate": 8.350053765473751e-05, "loss": 0.0168, "step": 12120 }, { "grad_norm": 0.23397991061210632, "learning_rate": 8.346983981902005e-05, "loss": 0.0162, "step": 12130 }, { "grad_norm": 0.21918846666812897, "learning_rate": 8.343911910698271e-05, "loss": 0.0131, "step": 12140 }, { "grad_norm": 0.2809174358844757, "learning_rate": 8.340837553962278e-05, "loss": 0.0151, "step": 12150 }, { "grad_norm": 0.21387194097042084, "learning_rate": 8.337760913795316e-05, "loss": 0.0174, "step": 12160 }, { "grad_norm": 0.20651578903198242, "learning_rate": 8.334681992300244e-05, "loss": 0.0169, "step": 12170 }, { "grad_norm": 0.19342441856861115, "learning_rate": 8.331600791581475e-05, "loss": 0.015, "step": 12180 }, { "grad_norm": 0.19391119480133057, "learning_rate": 8.328517313744978e-05, "loss": 0.0163, "step": 12190 }, { "grad_norm": 0.2624305784702301, "learning_rate": 8.325431560898286e-05, "loss": 0.02, "step": 12200 }, { "grad_norm": 0.26653051376342773, "learning_rate": 8.322343535150478e-05, "loss": 0.0191, "step": 12210 }, { "grad_norm": 0.31483158469200134, "learning_rate": 8.319253238612191e-05, "loss": 0.0196, "step": 12220 }, { "grad_norm": 0.30761268734931946, "learning_rate": 8.316160673395614e-05, "loss": 0.0162, "step": 12230 }, { "grad_norm": 0.2702692449092865, "learning_rate": 8.313065841614487e-05, "loss": 0.0155, "step": 12240 }, { "grad_norm": 0.3006288707256317, "learning_rate": 8.309968745384096e-05, "loss": 0.0202, "step": 12250 }, { "grad_norm": 0.23596268892288208, "learning_rate": 8.306869386821282e-05, "loss": 0.0184, "step": 12260 }, { "grad_norm": 0.23787237703800201, "learning_rate": 8.30376776804442e-05, "loss": 0.0215, "step": 12270 }, { "grad_norm": 0.35502931475639343, "learning_rate": 8.300663891173443e-05, "loss": 0.0196, "step": 12280 }, { "grad_norm": 0.323270320892334, "learning_rate": 8.297557758329822e-05, "loss": 0.0262, "step": 12290 }, { "grad_norm": 0.36198338866233826, "learning_rate": 8.294449371636564e-05, "loss": 0.0221, "step": 12300 }, { "grad_norm": 0.2799721956253052, "learning_rate": 8.291338733218226e-05, "loss": 0.0213, "step": 12310 }, { "grad_norm": 0.38302865624427795, "learning_rate": 8.2882258452009e-05, "loss": 0.0197, "step": 12320 }, { "grad_norm": 0.2698972821235657, "learning_rate": 8.285110709712214e-05, "loss": 0.0191, "step": 12330 }, { "grad_norm": 0.2609735429286957, "learning_rate": 8.281993328881337e-05, "loss": 0.021, "step": 12340 }, { "grad_norm": 0.18290141224861145, "learning_rate": 8.278873704838964e-05, "loss": 0.0173, "step": 12350 }, { "grad_norm": 0.2912469804286957, "learning_rate": 8.275751839717334e-05, "loss": 0.0217, "step": 12360 }, { "grad_norm": 0.23594126105308533, "learning_rate": 8.272627735650208e-05, "loss": 0.0204, "step": 12370 }, { "grad_norm": 0.2104303538799286, "learning_rate": 8.269501394772884e-05, "loss": 0.0197, "step": 12380 }, { "grad_norm": 0.2931503355503082, "learning_rate": 8.266372819222189e-05, "loss": 0.0169, "step": 12390 }, { "grad_norm": 0.27308666706085205, "learning_rate": 8.26324201113647e-05, "loss": 0.02, "step": 12400 }, { "grad_norm": 0.21324539184570312, "learning_rate": 8.260108972655606e-05, "loss": 0.0173, "step": 12410 }, { "grad_norm": 0.23349253833293915, "learning_rate": 8.256973705921e-05, "loss": 0.0161, "step": 12420 }, { "grad_norm": 0.20363041758537292, "learning_rate": 8.25383621307558e-05, "loss": 0.0144, "step": 12430 }, { "grad_norm": 0.292523056268692, "learning_rate": 8.25069649626379e-05, "loss": 0.0158, "step": 12440 }, { "grad_norm": 0.2349836677312851, "learning_rate": 8.247554557631596e-05, "loss": 0.0231, "step": 12450 }, { "grad_norm": 0.2574256956577301, "learning_rate": 8.244410399326483e-05, "loss": 0.0164, "step": 12460 }, { "grad_norm": 0.24138526618480682, "learning_rate": 8.241264023497457e-05, "loss": 0.0203, "step": 12470 }, { "grad_norm": 0.18927815556526184, "learning_rate": 8.238115432295034e-05, "loss": 0.0138, "step": 12480 }, { "grad_norm": 0.20067629218101501, "learning_rate": 8.234964627871247e-05, "loss": 0.0145, "step": 12490 }, { "grad_norm": 0.23857522010803223, "learning_rate": 8.231811612379639e-05, "loss": 0.0145, "step": 12500 }, { "grad_norm": 0.20295369625091553, "learning_rate": 8.228656387975268e-05, "loss": 0.0141, "step": 12510 }, { "grad_norm": 0.18277665972709656, "learning_rate": 8.225498956814702e-05, "loss": 0.0159, "step": 12520 }, { "grad_norm": 0.21665266156196594, "learning_rate": 8.222339321056014e-05, "loss": 0.0178, "step": 12530 }, { "grad_norm": 0.24993127584457397, "learning_rate": 8.219177482858785e-05, "loss": 0.0162, "step": 12540 }, { "grad_norm": 0.28904807567596436, "learning_rate": 8.216013444384099e-05, "loss": 0.0163, "step": 12550 }, { "grad_norm": 0.20956997573375702, "learning_rate": 8.21284720779455e-05, "loss": 0.0156, "step": 12560 }, { "grad_norm": 0.19736088812351227, "learning_rate": 8.209678775254231e-05, "loss": 0.0162, "step": 12570 }, { "grad_norm": 0.21154403686523438, "learning_rate": 8.206508148928733e-05, "loss": 0.0184, "step": 12580 }, { "grad_norm": 0.28182724118232727, "learning_rate": 8.203335330985151e-05, "loss": 0.0243, "step": 12590 }, { "grad_norm": 0.26114732027053833, "learning_rate": 8.200160323592076e-05, "loss": 0.0268, "step": 12600 }, { "grad_norm": 0.23068714141845703, "learning_rate": 8.196983128919598e-05, "loss": 0.0218, "step": 12610 }, { "grad_norm": 0.21480272710323334, "learning_rate": 8.193803749139295e-05, "loss": 0.0223, "step": 12620 }, { "grad_norm": 0.4722104072570801, "learning_rate": 8.190622186424244e-05, "loss": 0.0195, "step": 12630 }, { "grad_norm": 0.33448970317840576, "learning_rate": 8.187438442949016e-05, "loss": 0.0205, "step": 12640 }, { "grad_norm": 0.22934865951538086, "learning_rate": 8.184252520889668e-05, "loss": 0.0173, "step": 12650 }, { "grad_norm": 0.28194451332092285, "learning_rate": 8.181064422423748e-05, "loss": 0.0161, "step": 12660 }, { "grad_norm": 0.23818764090538025, "learning_rate": 8.177874149730289e-05, "loss": 0.0178, "step": 12670 }, { "grad_norm": 0.2727178633213043, "learning_rate": 8.174681704989816e-05, "loss": 0.0182, "step": 12680 }, { "grad_norm": 0.2581961750984192, "learning_rate": 8.171487090384333e-05, "loss": 0.0185, "step": 12690 }, { "grad_norm": 0.34345969557762146, "learning_rate": 8.168290308097328e-05, "loss": 0.0168, "step": 12700 }, { "grad_norm": 0.3752903342247009, "learning_rate": 8.165091360313774e-05, "loss": 0.0198, "step": 12710 }, { "grad_norm": 0.21826490759849548, "learning_rate": 8.161890249220119e-05, "loss": 0.0194, "step": 12720 }, { "grad_norm": 0.267732709646225, "learning_rate": 8.158686977004295e-05, "loss": 0.0164, "step": 12730 }, { "grad_norm": 0.18924736976623535, "learning_rate": 8.155481545855706e-05, "loss": 0.0161, "step": 12740 }, { "grad_norm": 0.27329128980636597, "learning_rate": 8.152273957965233e-05, "loss": 0.0163, "step": 12750 }, { "grad_norm": 0.21599484980106354, "learning_rate": 8.149064215525237e-05, "loss": 0.0172, "step": 12760 }, { "grad_norm": 0.19974306225776672, "learning_rate": 8.14585232072954e-05, "loss": 0.018, "step": 12770 }, { "grad_norm": 0.20387540757656097, "learning_rate": 8.142638275773449e-05, "loss": 0.0201, "step": 12780 }, { "grad_norm": 0.2573419511318207, "learning_rate": 8.139422082853729e-05, "loss": 0.0165, "step": 12790 }, { "grad_norm": 0.3068048655986786, "learning_rate": 8.136203744168618e-05, "loss": 0.0188, "step": 12800 }, { "grad_norm": 0.24711620807647705, "learning_rate": 8.132983261917819e-05, "loss": 0.0144, "step": 12810 }, { "grad_norm": 0.2586219310760498, "learning_rate": 8.129760638302504e-05, "loss": 0.016, "step": 12820 }, { "grad_norm": 0.203657329082489, "learning_rate": 8.126535875525305e-05, "loss": 0.0171, "step": 12830 }, { "grad_norm": 0.2132851928472519, "learning_rate": 8.123308975790316e-05, "loss": 0.0164, "step": 12840 }, { "grad_norm": 0.27973493933677673, "learning_rate": 8.120079941303094e-05, "loss": 0.0173, "step": 12850 }, { "grad_norm": 0.3510672450065613, "learning_rate": 8.116848774270651e-05, "loss": 0.0182, "step": 12860 }, { "grad_norm": 0.24534450471401215, "learning_rate": 8.113615476901461e-05, "loss": 0.0146, "step": 12870 }, { "grad_norm": 0.24077753722667694, "learning_rate": 8.110380051405454e-05, "loss": 0.0146, "step": 12880 }, { "grad_norm": 0.2202301174402237, "learning_rate": 8.107142499994009e-05, "loss": 0.0167, "step": 12890 }, { "grad_norm": 0.2056407779455185, "learning_rate": 8.103902824879966e-05, "loss": 0.0173, "step": 12900 }, { "grad_norm": 0.30725008249282837, "learning_rate": 8.10066102827761e-05, "loss": 0.0158, "step": 12910 }, { "grad_norm": 0.2197410613298416, "learning_rate": 8.097417112402676e-05, "loss": 0.0171, "step": 12920 }, { "grad_norm": 0.26173076033592224, "learning_rate": 8.094171079472355e-05, "loss": 0.0158, "step": 12930 }, { "grad_norm": 0.27025073766708374, "learning_rate": 8.090922931705277e-05, "loss": 0.0164, "step": 12940 }, { "grad_norm": 0.20663760602474213, "learning_rate": 8.08767267132152e-05, "loss": 0.0143, "step": 12950 }, { "grad_norm": 0.2732468843460083, "learning_rate": 8.084420300542608e-05, "loss": 0.0165, "step": 12960 }, { "grad_norm": 0.2542497217655182, "learning_rate": 8.081165821591505e-05, "loss": 0.0138, "step": 12970 }, { "grad_norm": 0.23824100196361542, "learning_rate": 8.077909236692615e-05, "loss": 0.0166, "step": 12980 }, { "grad_norm": 0.2974347472190857, "learning_rate": 8.074650548071787e-05, "loss": 0.0255, "step": 12990 }, { "grad_norm": 0.30506035685539246, "learning_rate": 8.071389757956301e-05, "loss": 0.0182, "step": 13000 }, { "grad_norm": 0.35865291953086853, "learning_rate": 8.068126868574876e-05, "loss": 0.0186, "step": 13010 }, { "grad_norm": 0.2914121150970459, "learning_rate": 8.064861882157668e-05, "loss": 0.0179, "step": 13020 }, { "grad_norm": 0.23027923703193665, "learning_rate": 8.061594800936263e-05, "loss": 0.0163, "step": 13030 }, { "grad_norm": 0.25501933693885803, "learning_rate": 8.058325627143681e-05, "loss": 0.0208, "step": 13040 }, { "grad_norm": 0.22977802157402039, "learning_rate": 8.055054363014372e-05, "loss": 0.0197, "step": 13050 }, { "grad_norm": 0.29685279726982117, "learning_rate": 8.051781010784211e-05, "loss": 0.0164, "step": 13060 }, { "grad_norm": 0.2529321014881134, "learning_rate": 8.048505572690506e-05, "loss": 0.0155, "step": 13070 }, { "grad_norm": 0.20063245296478271, "learning_rate": 8.045228050971988e-05, "loss": 0.0149, "step": 13080 }, { "grad_norm": 0.1757207214832306, "learning_rate": 8.041948447868814e-05, "loss": 0.015, "step": 13090 }, { "grad_norm": 0.23408374190330505, "learning_rate": 8.038666765622558e-05, "loss": 0.0117, "step": 13100 }, { "grad_norm": 0.23120121657848358, "learning_rate": 8.03538300647622e-05, "loss": 0.0165, "step": 13110 }, { "grad_norm": 0.2360689789056778, "learning_rate": 8.03209717267422e-05, "loss": 0.0133, "step": 13120 }, { "grad_norm": 0.1880175918340683, "learning_rate": 8.028809266462395e-05, "loss": 0.0151, "step": 13130 }, { "grad_norm": 0.18290594220161438, "learning_rate": 8.025519290087994e-05, "loss": 0.0156, "step": 13140 }, { "grad_norm": 0.23764319717884064, "learning_rate": 8.022227245799688e-05, "loss": 0.017, "step": 13150 }, { "grad_norm": 0.27492183446884155, "learning_rate": 8.018933135847557e-05, "loss": 0.014, "step": 13160 }, { "grad_norm": 0.30232948064804077, "learning_rate": 8.015636962483096e-05, "loss": 0.0195, "step": 13170 }, { "grad_norm": 0.23120515048503876, "learning_rate": 8.012338727959205e-05, "loss": 0.0152, "step": 13180 }, { "grad_norm": 0.25524917244911194, "learning_rate": 8.009038434530198e-05, "loss": 0.0173, "step": 13190 }, { "grad_norm": 0.31191253662109375, "learning_rate": 8.005736084451796e-05, "loss": 0.0167, "step": 13200 }, { "grad_norm": 0.2304830700159073, "learning_rate": 8.002431679981122e-05, "loss": 0.0156, "step": 13210 }, { "grad_norm": 0.2319273203611374, "learning_rate": 7.999125223376706e-05, "loss": 0.0186, "step": 13220 }, { "grad_norm": 0.2762177288532257, "learning_rate": 7.99581671689848e-05, "loss": 0.0202, "step": 13230 }, { "grad_norm": 0.31683430075645447, "learning_rate": 7.992506162807775e-05, "loss": 0.0205, "step": 13240 }, { "grad_norm": 0.24324186146259308, "learning_rate": 7.989193563367328e-05, "loss": 0.0175, "step": 13250 }, { "grad_norm": 0.2905685007572174, "learning_rate": 7.985878920841266e-05, "loss": 0.0169, "step": 13260 }, { "grad_norm": 0.2420824021100998, "learning_rate": 7.982562237495117e-05, "loss": 0.0142, "step": 13270 }, { "grad_norm": 0.23417174816131592, "learning_rate": 7.979243515595802e-05, "loss": 0.0159, "step": 13280 }, { "grad_norm": 0.13682812452316284, "learning_rate": 7.975922757411636e-05, "loss": 0.0122, "step": 13290 }, { "grad_norm": 0.27547359466552734, "learning_rate": 7.972599965212329e-05, "loss": 0.0176, "step": 13300 }, { "grad_norm": 0.19693729281425476, "learning_rate": 7.969275141268973e-05, "loss": 0.0154, "step": 13310 }, { "grad_norm": 0.2507711946964264, "learning_rate": 7.96594828785406e-05, "loss": 0.0144, "step": 13320 }, { "grad_norm": 0.2981087863445282, "learning_rate": 7.962619407241456e-05, "loss": 0.0158, "step": 13330 }, { "grad_norm": 0.2891901731491089, "learning_rate": 7.959288501706424e-05, "loss": 0.015, "step": 13340 }, { "grad_norm": 0.21141546964645386, "learning_rate": 7.955955573525605e-05, "loss": 0.0148, "step": 13350 }, { "grad_norm": 0.2097148895263672, "learning_rate": 7.952620624977026e-05, "loss": 0.019, "step": 13360 }, { "grad_norm": 0.24071533977985382, "learning_rate": 7.949283658340089e-05, "loss": 0.018, "step": 13370 }, { "grad_norm": 0.24027597904205322, "learning_rate": 7.945944675895585e-05, "loss": 0.0162, "step": 13380 }, { "grad_norm": 0.270802766084671, "learning_rate": 7.942603679925671e-05, "loss": 0.017, "step": 13390 }, { "grad_norm": 0.2675418257713318, "learning_rate": 7.93926067271389e-05, "loss": 0.0172, "step": 13400 }, { "grad_norm": 0.20585080981254578, "learning_rate": 7.935915656545155e-05, "loss": 0.0159, "step": 13410 }, { "grad_norm": 0.2996295392513275, "learning_rate": 7.932568633705752e-05, "loss": 0.0152, "step": 13420 }, { "grad_norm": 0.29721954464912415, "learning_rate": 7.929219606483341e-05, "loss": 0.0189, "step": 13430 }, { "grad_norm": 0.24892069399356842, "learning_rate": 7.925868577166948e-05, "loss": 0.0207, "step": 13440 }, { "grad_norm": 0.29297736287117004, "learning_rate": 7.922515548046974e-05, "loss": 0.0178, "step": 13450 }, { "grad_norm": 0.2212260216474533, "learning_rate": 7.919160521415179e-05, "loss": 0.0178, "step": 13460 }, { "grad_norm": 0.20860420167446136, "learning_rate": 7.915803499564694e-05, "loss": 0.0199, "step": 13470 }, { "grad_norm": 0.2951616644859314, "learning_rate": 7.912444484790013e-05, "loss": 0.0174, "step": 13480 }, { "grad_norm": 0.27953341603279114, "learning_rate": 7.909083479386987e-05, "loss": 0.0171, "step": 13490 }, { "grad_norm": 0.20664116740226746, "learning_rate": 7.905720485652836e-05, "loss": 0.0152, "step": 13500 }, { "grad_norm": 0.2791363596916199, "learning_rate": 7.902355505886132e-05, "loss": 0.0177, "step": 13510 }, { "grad_norm": 0.20436958968639374, "learning_rate": 7.898988542386805e-05, "loss": 0.0152, "step": 13520 }, { "grad_norm": 0.22952382266521454, "learning_rate": 7.895619597456147e-05, "loss": 0.0147, "step": 13530 }, { "grad_norm": 0.28600093722343445, "learning_rate": 7.892248673396798e-05, "loss": 0.0135, "step": 13540 }, { "grad_norm": 0.22455671429634094, "learning_rate": 7.888875772512754e-05, "loss": 0.015, "step": 13550 }, { "grad_norm": 0.22174561023712158, "learning_rate": 7.885500897109359e-05, "loss": 0.0192, "step": 13560 }, { "grad_norm": 0.2929903268814087, "learning_rate": 7.882124049493309e-05, "loss": 0.0157, "step": 13570 }, { "grad_norm": 0.2567988932132721, "learning_rate": 7.878745231972649e-05, "loss": 0.0139, "step": 13580 }, { "grad_norm": 0.18682906031608582, "learning_rate": 7.875364446856766e-05, "loss": 0.0157, "step": 13590 }, { "grad_norm": 0.23805029690265656, "learning_rate": 7.871981696456398e-05, "loss": 0.0123, "step": 13600 }, { "grad_norm": 0.22198061645030975, "learning_rate": 7.868596983083623e-05, "loss": 0.016, "step": 13610 }, { "grad_norm": 0.17113880813121796, "learning_rate": 7.865210309051858e-05, "loss": 0.014, "step": 13620 }, { "grad_norm": 0.27274227142333984, "learning_rate": 7.861821676675863e-05, "loss": 0.015, "step": 13630 }, { "grad_norm": 0.20355355739593506, "learning_rate": 7.858431088271739e-05, "loss": 0.0157, "step": 13640 }, { "grad_norm": 0.28771764039993286, "learning_rate": 7.855038546156918e-05, "loss": 0.0165, "step": 13650 }, { "grad_norm": 0.2393818497657776, "learning_rate": 7.851644052650173e-05, "loss": 0.0134, "step": 13660 }, { "grad_norm": 0.3116772770881653, "learning_rate": 7.848247610071609e-05, "loss": 0.0166, "step": 13670 }, { "grad_norm": 0.3284268081188202, "learning_rate": 7.844849220742658e-05, "loss": 0.0159, "step": 13680 }, { "grad_norm": 0.20632249116897583, "learning_rate": 7.841448886986092e-05, "loss": 0.0158, "step": 13690 }, { "grad_norm": 0.2303599864244461, "learning_rate": 7.838046611126004e-05, "loss": 0.015, "step": 13700 }, { "grad_norm": 0.2749268710613251, "learning_rate": 7.834642395487819e-05, "loss": 0.0161, "step": 13710 }, { "grad_norm": 0.1767609566450119, "learning_rate": 7.831236242398285e-05, "loss": 0.0151, "step": 13720 }, { "grad_norm": 0.3195914924144745, "learning_rate": 7.827828154185477e-05, "loss": 0.0144, "step": 13730 }, { "grad_norm": 0.1976119428873062, "learning_rate": 7.82441813317879e-05, "loss": 0.0164, "step": 13740 }, { "grad_norm": 0.27307459712028503, "learning_rate": 7.821006181708944e-05, "loss": 0.0133, "step": 13750 }, { "grad_norm": 0.21520282328128815, "learning_rate": 7.81759230210797e-05, "loss": 0.0126, "step": 13760 }, { "grad_norm": 0.22775055468082428, "learning_rate": 7.814176496709227e-05, "loss": 0.0154, "step": 13770 }, { "grad_norm": 0.22262713313102722, "learning_rate": 7.810758767847385e-05, "loss": 0.0174, "step": 13780 }, { "grad_norm": 0.2969236969947815, "learning_rate": 7.807339117858427e-05, "loss": 0.0148, "step": 13790 }, { "grad_norm": 0.2467334121465683, "learning_rate": 7.803917549079655e-05, "loss": 0.0148, "step": 13800 }, { "grad_norm": 0.24575646221637726, "learning_rate": 7.800494063849679e-05, "loss": 0.0137, "step": 13810 }, { "grad_norm": 0.23038055002689362, "learning_rate": 7.797068664508416e-05, "loss": 0.0143, "step": 13820 }, { "grad_norm": 0.2643640339374542, "learning_rate": 7.793641353397096e-05, "loss": 0.0129, "step": 13830 }, { "grad_norm": 0.2274516224861145, "learning_rate": 7.790212132858253e-05, "loss": 0.0128, "step": 13840 }, { "grad_norm": 0.24406123161315918, "learning_rate": 7.786781005235728e-05, "loss": 0.0156, "step": 13850 }, { "grad_norm": 0.23363418877124786, "learning_rate": 7.783347972874662e-05, "loss": 0.0164, "step": 13860 }, { "grad_norm": 0.19697047770023346, "learning_rate": 7.779913038121504e-05, "loss": 0.0155, "step": 13870 }, { "grad_norm": 0.20449207723140717, "learning_rate": 7.776476203323997e-05, "loss": 0.0145, "step": 13880 }, { "grad_norm": 0.19439749419689178, "learning_rate": 7.773037470831185e-05, "loss": 0.0132, "step": 13890 }, { "grad_norm": 0.24354250729084015, "learning_rate": 7.76959684299341e-05, "loss": 0.0154, "step": 13900 }, { "grad_norm": 0.17208349704742432, "learning_rate": 7.76615432216231e-05, "loss": 0.0114, "step": 13910 }, { "grad_norm": 0.20648854970932007, "learning_rate": 7.762709910690811e-05, "loss": 0.0125, "step": 13920 }, { "grad_norm": 0.27076828479766846, "learning_rate": 7.759263610933141e-05, "loss": 0.0134, "step": 13930 }, { "grad_norm": 0.1984148621559143, "learning_rate": 7.755815425244811e-05, "loss": 0.0145, "step": 13940 }, { "grad_norm": 0.18209360539913177, "learning_rate": 7.752365355982624e-05, "loss": 0.0143, "step": 13950 }, { "grad_norm": 0.16620470583438873, "learning_rate": 7.748913405504668e-05, "loss": 0.0139, "step": 13960 }, { "grad_norm": 0.2937784492969513, "learning_rate": 7.745459576170322e-05, "loss": 0.0136, "step": 13970 }, { "grad_norm": 0.23546582460403442, "learning_rate": 7.742003870340242e-05, "loss": 0.0142, "step": 13980 }, { "grad_norm": 0.23338183760643005, "learning_rate": 7.738546290376373e-05, "loss": 0.0161, "step": 13990 }, { "grad_norm": 0.18541684746742249, "learning_rate": 7.735086838641937e-05, "loss": 0.0174, "step": 14000 }, { "grad_norm": 0.246803879737854, "learning_rate": 7.731625517501437e-05, "loss": 0.012, "step": 14010 }, { "grad_norm": 0.2555334270000458, "learning_rate": 7.728162329320655e-05, "loss": 0.0145, "step": 14020 }, { "grad_norm": 0.25577327609062195, "learning_rate": 7.724697276466645e-05, "loss": 0.0141, "step": 14030 }, { "grad_norm": 0.24839909374713898, "learning_rate": 7.721230361307738e-05, "loss": 0.0151, "step": 14040 }, { "grad_norm": 0.21603132784366608, "learning_rate": 7.71776158621354e-05, "loss": 0.0159, "step": 14050 }, { "grad_norm": 0.24831442534923553, "learning_rate": 7.714290953554925e-05, "loss": 0.0156, "step": 14060 }, { "grad_norm": 0.18926580250263214, "learning_rate": 7.710818465704037e-05, "loss": 0.0152, "step": 14070 }, { "grad_norm": 0.29884791374206543, "learning_rate": 7.707344125034288e-05, "loss": 0.0146, "step": 14080 }, { "grad_norm": 0.1956091970205307, "learning_rate": 7.703867933920359e-05, "loss": 0.0151, "step": 14090 }, { "grad_norm": 0.2352083921432495, "learning_rate": 7.700389894738194e-05, "loss": 0.0146, "step": 14100 }, { "grad_norm": 0.22398404777050018, "learning_rate": 7.696910009864999e-05, "loss": 0.0144, "step": 14110 }, { "grad_norm": 0.19092968106269836, "learning_rate": 7.693428281679241e-05, "loss": 0.0143, "step": 14120 }, { "grad_norm": 0.17841246724128723, "learning_rate": 7.689944712560652e-05, "loss": 0.0153, "step": 14130 }, { "grad_norm": 0.20235787332057953, "learning_rate": 7.686459304890214e-05, "loss": 0.0138, "step": 14140 }, { "grad_norm": 0.24455077946186066, "learning_rate": 7.682972061050175e-05, "loss": 0.0147, "step": 14150 }, { "grad_norm": 0.26294150948524475, "learning_rate": 7.679482983424032e-05, "loss": 0.0168, "step": 14160 }, { "grad_norm": 0.2933233082294464, "learning_rate": 7.675992074396534e-05, "loss": 0.014, "step": 14170 }, { "grad_norm": 0.21351614594459534, "learning_rate": 7.672499336353687e-05, "loss": 0.0181, "step": 14180 }, { "grad_norm": 0.267890065908432, "learning_rate": 7.669004771682744e-05, "loss": 0.0141, "step": 14190 }, { "grad_norm": 0.21355514228343964, "learning_rate": 7.665508382772206e-05, "loss": 0.0176, "step": 14200 }, { "grad_norm": 0.3134532570838928, "learning_rate": 7.662010172011824e-05, "loss": 0.0126, "step": 14210 }, { "grad_norm": 0.2669408917427063, "learning_rate": 7.658510141792588e-05, "loss": 0.0157, "step": 14220 }, { "grad_norm": 0.3157544434070587, "learning_rate": 7.65500829450674e-05, "loss": 0.0139, "step": 14230 }, { "grad_norm": 0.2446998804807663, "learning_rate": 7.651504632547759e-05, "loss": 0.0172, "step": 14240 }, { "grad_norm": 0.2110620141029358, "learning_rate": 7.647999158310364e-05, "loss": 0.0165, "step": 14250 }, { "grad_norm": 0.2302870899438858, "learning_rate": 7.644491874190512e-05, "loss": 0.0157, "step": 14260 }, { "grad_norm": 0.21135833859443665, "learning_rate": 7.6409827825854e-05, "loss": 0.0148, "step": 14270 }, { "grad_norm": 0.218619242310524, "learning_rate": 7.637471885893459e-05, "loss": 0.0159, "step": 14280 }, { "grad_norm": 0.26966723799705505, "learning_rate": 7.633959186514354e-05, "loss": 0.0149, "step": 14290 }, { "grad_norm": 0.25487515330314636, "learning_rate": 7.630444686848984e-05, "loss": 0.0177, "step": 14300 }, { "grad_norm": 0.2411964237689972, "learning_rate": 7.626928389299471e-05, "loss": 0.0135, "step": 14310 }, { "grad_norm": 0.22101984918117523, "learning_rate": 7.623410296269175e-05, "loss": 0.0151, "step": 14320 }, { "grad_norm": 0.1915147602558136, "learning_rate": 7.61989041016268e-05, "loss": 0.015, "step": 14330 }, { "grad_norm": 0.20425495505332947, "learning_rate": 7.616368733385793e-05, "loss": 0.0135, "step": 14340 }, { "grad_norm": 0.27115964889526367, "learning_rate": 7.612845268345547e-05, "loss": 0.0168, "step": 14350 }, { "grad_norm": 0.206074520945549, "learning_rate": 7.609320017450199e-05, "loss": 0.0146, "step": 14360 }, { "grad_norm": 0.2551981210708618, "learning_rate": 7.605792983109222e-05, "loss": 0.0122, "step": 14370 }, { "grad_norm": 0.1978178769350052, "learning_rate": 7.602264167733313e-05, "loss": 0.0168, "step": 14380 }, { "grad_norm": 0.2589010000228882, "learning_rate": 7.598733573734384e-05, "loss": 0.0166, "step": 14390 }, { "grad_norm": 0.3041294813156128, "learning_rate": 7.595201203525561e-05, "loss": 0.0149, "step": 14400 }, { "grad_norm": 0.23556829988956451, "learning_rate": 7.591667059521187e-05, "loss": 0.0169, "step": 14410 }, { "grad_norm": 0.2663807272911072, "learning_rate": 7.588131144136815e-05, "loss": 0.0154, "step": 14420 }, { "grad_norm": 0.18149212002754211, "learning_rate": 7.584593459789212e-05, "loss": 0.0132, "step": 14430 }, { "grad_norm": 0.16501854360103607, "learning_rate": 7.58105400889635e-05, "loss": 0.0127, "step": 14440 }, { "grad_norm": 0.28323957324028015, "learning_rate": 7.57751279387741e-05, "loss": 0.0135, "step": 14450 }, { "grad_norm": 0.262543261051178, "learning_rate": 7.573969817152782e-05, "loss": 0.0149, "step": 14460 }, { "grad_norm": 0.21249084174633026, "learning_rate": 7.570425081144052e-05, "loss": 0.0145, "step": 14470 }, { "grad_norm": 0.2547721564769745, "learning_rate": 7.56687858827402e-05, "loss": 0.0139, "step": 14480 }, { "grad_norm": 0.32797861099243164, "learning_rate": 7.563330340966675e-05, "loss": 0.0157, "step": 14490 }, { "grad_norm": 0.24078959226608276, "learning_rate": 7.559780341647212e-05, "loss": 0.0141, "step": 14500 }, { "grad_norm": 0.2453860640525818, "learning_rate": 7.556228592742026e-05, "loss": 0.0155, "step": 14510 }, { "grad_norm": 0.24073605239391327, "learning_rate": 7.552675096678696e-05, "loss": 0.013, "step": 14520 }, { "grad_norm": 0.2501186728477478, "learning_rate": 7.549119855886012e-05, "loss": 0.0151, "step": 14530 }, { "grad_norm": 0.1741497665643692, "learning_rate": 7.545562872793941e-05, "loss": 0.0163, "step": 14540 }, { "grad_norm": 0.22305426001548767, "learning_rate": 7.542004149833648e-05, "loss": 0.0135, "step": 14550 }, { "grad_norm": 0.17846503853797913, "learning_rate": 7.538443689437492e-05, "loss": 0.0125, "step": 14560 }, { "grad_norm": 0.20522762835025787, "learning_rate": 7.53488149403901e-05, "loss": 0.0138, "step": 14570 }, { "grad_norm": 0.27433356642723083, "learning_rate": 7.531317566072929e-05, "loss": 0.0137, "step": 14580 }, { "grad_norm": 0.18284747004508972, "learning_rate": 7.527751907975158e-05, "loss": 0.0131, "step": 14590 }, { "grad_norm": 0.2391151487827301, "learning_rate": 7.524184522182793e-05, "loss": 0.0128, "step": 14600 }, { "grad_norm": 0.17460982501506805, "learning_rate": 7.520615411134112e-05, "loss": 0.0136, "step": 14610 }, { "grad_norm": 0.2813291847705841, "learning_rate": 7.517044577268564e-05, "loss": 0.0151, "step": 14620 }, { "grad_norm": 0.30469661951065063, "learning_rate": 7.513472023026782e-05, "loss": 0.0188, "step": 14630 }, { "grad_norm": 0.18961846828460693, "learning_rate": 7.509897750850572e-05, "loss": 0.0162, "step": 14640 }, { "grad_norm": 0.20361432433128357, "learning_rate": 7.506321763182918e-05, "loss": 0.0138, "step": 14650 }, { "grad_norm": 0.22410954535007477, "learning_rate": 7.50274406246797e-05, "loss": 0.018, "step": 14660 }, { "grad_norm": 0.2834399342536926, "learning_rate": 7.499164651151056e-05, "loss": 0.0136, "step": 14670 }, { "grad_norm": 0.18275614082813263, "learning_rate": 7.495583531678669e-05, "loss": 0.0137, "step": 14680 }, { "grad_norm": 0.22027723491191864, "learning_rate": 7.492000706498469e-05, "loss": 0.0159, "step": 14690 }, { "grad_norm": 0.19754092395305634, "learning_rate": 7.488416178059284e-05, "loss": 0.0156, "step": 14700 }, { "grad_norm": 0.25370585918426514, "learning_rate": 7.484829948811107e-05, "loss": 0.0131, "step": 14710 }, { "grad_norm": 0.22822852432727814, "learning_rate": 7.48124202120509e-05, "loss": 0.0155, "step": 14720 }, { "grad_norm": 0.24947351217269897, "learning_rate": 7.477652397693549e-05, "loss": 0.0155, "step": 14730 }, { "grad_norm": 0.32576313614845276, "learning_rate": 7.474061080729955e-05, "loss": 0.0176, "step": 14740 }, { "grad_norm": 0.3357251286506653, "learning_rate": 7.470468072768941e-05, "loss": 0.015, "step": 14750 }, { "grad_norm": 0.2512498199939728, "learning_rate": 7.466873376266297e-05, "loss": 0.0138, "step": 14760 }, { "grad_norm": 0.23220369219779968, "learning_rate": 7.46327699367896e-05, "loss": 0.0159, "step": 14770 }, { "grad_norm": 0.28336212038993835, "learning_rate": 7.459678927465026e-05, "loss": 0.0145, "step": 14780 }, { "grad_norm": 0.19396787881851196, "learning_rate": 7.456079180083737e-05, "loss": 0.016, "step": 14790 }, { "grad_norm": 0.2004605531692505, "learning_rate": 7.452477753995489e-05, "loss": 0.015, "step": 14800 }, { "grad_norm": 0.2348499447107315, "learning_rate": 7.448874651661823e-05, "loss": 0.0136, "step": 14810 }, { "grad_norm": 0.24573257565498352, "learning_rate": 7.445269875545423e-05, "loss": 0.0167, "step": 14820 }, { "grad_norm": 0.24361838400363922, "learning_rate": 7.44166342811012e-05, "loss": 0.0173, "step": 14830 }, { "grad_norm": 0.2649342715740204, "learning_rate": 7.438055311820886e-05, "loss": 0.0169, "step": 14840 }, { "grad_norm": 0.23645029962062836, "learning_rate": 7.434445529143837e-05, "loss": 0.0148, "step": 14850 }, { "grad_norm": 0.22955673933029175, "learning_rate": 7.430834082546225e-05, "loss": 0.015, "step": 14860 }, { "grad_norm": 0.21441224217414856, "learning_rate": 7.427220974496438e-05, "loss": 0.0135, "step": 14870 }, { "grad_norm": 0.23026734590530396, "learning_rate": 7.423606207464005e-05, "loss": 0.0165, "step": 14880 }, { "grad_norm": 0.1894923895597458, "learning_rate": 7.419989783919578e-05, "loss": 0.0163, "step": 14890 }, { "grad_norm": 0.2910066246986389, "learning_rate": 7.416371706334956e-05, "loss": 0.0176, "step": 14900 }, { "grad_norm": 0.306314080953598, "learning_rate": 7.412751977183056e-05, "loss": 0.0132, "step": 14910 }, { "grad_norm": 0.230758398771286, "learning_rate": 7.409130598937932e-05, "loss": 0.0174, "step": 14920 }, { "grad_norm": 0.2927529811859131, "learning_rate": 7.40550757407476e-05, "loss": 0.0168, "step": 14930 }, { "grad_norm": 0.21456986665725708, "learning_rate": 7.401882905069843e-05, "loss": 0.0168, "step": 14940 }, { "grad_norm": 0.26683396100997925, "learning_rate": 7.39825659440061e-05, "loss": 0.0129, "step": 14950 }, { "grad_norm": 0.1907789260149002, "learning_rate": 7.394628644545609e-05, "loss": 0.0125, "step": 14960 }, { "grad_norm": 0.22627191245555878, "learning_rate": 7.390999057984507e-05, "loss": 0.0186, "step": 14970 }, { "grad_norm": 0.20038722455501556, "learning_rate": 7.387367837198097e-05, "loss": 0.0136, "step": 14980 }, { "grad_norm": 0.23780646920204163, "learning_rate": 7.383734984668281e-05, "loss": 0.0129, "step": 14990 }, { "grad_norm": 0.23646123707294464, "learning_rate": 7.38010050287808e-05, "loss": 0.0157, "step": 15000 }, { "grad_norm": 0.2874116003513336, "learning_rate": 7.376464394311628e-05, "loss": 0.0135, "step": 15010 }, { "grad_norm": 0.24947625398635864, "learning_rate": 7.372826661454172e-05, "loss": 0.0157, "step": 15020 }, { "grad_norm": 0.21288587152957916, "learning_rate": 7.369187306792068e-05, "loss": 0.0129, "step": 15030 }, { "grad_norm": 0.22099658846855164, "learning_rate": 7.365546332812779e-05, "loss": 0.0152, "step": 15040 }, { "grad_norm": 0.19745735824108124, "learning_rate": 7.361903742004876e-05, "loss": 0.0138, "step": 15050 }, { "grad_norm": 0.21884384751319885, "learning_rate": 7.358259536858039e-05, "loss": 0.0125, "step": 15060 }, { "grad_norm": 0.2131662368774414, "learning_rate": 7.354613719863044e-05, "loss": 0.0135, "step": 15070 }, { "grad_norm": 0.22273826599121094, "learning_rate": 7.350966293511776e-05, "loss": 0.014, "step": 15080 }, { "grad_norm": 0.24823518097400665, "learning_rate": 7.347317260297212e-05, "loss": 0.0136, "step": 15090 }, { "grad_norm": 0.26920509338378906, "learning_rate": 7.343666622713437e-05, "loss": 0.0137, "step": 15100 }, { "grad_norm": 0.25272396206855774, "learning_rate": 7.340014383255624e-05, "loss": 0.0148, "step": 15110 }, { "grad_norm": 0.23034341633319855, "learning_rate": 7.336360544420044e-05, "loss": 0.0141, "step": 15120 }, { "grad_norm": 0.18818239867687225, "learning_rate": 7.332705108704064e-05, "loss": 0.0142, "step": 15130 }, { "grad_norm": 0.23472405970096588, "learning_rate": 7.329048078606138e-05, "loss": 0.0152, "step": 15140 }, { "grad_norm": 0.2066391110420227, "learning_rate": 7.32538945662581e-05, "loss": 0.0158, "step": 15150 }, { "grad_norm": 0.21992227435112, "learning_rate": 7.321729245263718e-05, "loss": 0.0146, "step": 15160 }, { "grad_norm": 0.27123451232910156, "learning_rate": 7.318067447021578e-05, "loss": 0.0148, "step": 15170 }, { "grad_norm": 0.24990223348140717, "learning_rate": 7.314404064402198e-05, "loss": 0.0145, "step": 15180 }, { "grad_norm": 0.1982785016298294, "learning_rate": 7.310739099909461e-05, "loss": 0.016, "step": 15190 }, { "grad_norm": 0.2539767324924469, "learning_rate": 7.307072556048339e-05, "loss": 0.0137, "step": 15200 }, { "grad_norm": 0.24860171973705292, "learning_rate": 7.30340443532488e-05, "loss": 0.0149, "step": 15210 }, { "grad_norm": 0.21611399948596954, "learning_rate": 7.299734740246208e-05, "loss": 0.0127, "step": 15220 }, { "grad_norm": 0.14487992227077484, "learning_rate": 7.296063473320528e-05, "loss": 0.0116, "step": 15230 }, { "grad_norm": 0.1767863780260086, "learning_rate": 7.292390637057113e-05, "loss": 0.0124, "step": 15240 }, { "grad_norm": 0.19220994412899017, "learning_rate": 7.288716233966314e-05, "loss": 0.0127, "step": 15250 }, { "grad_norm": 0.23652790486812592, "learning_rate": 7.285040266559551e-05, "loss": 0.0148, "step": 15260 }, { "grad_norm": 0.33571091294288635, "learning_rate": 7.281362737349312e-05, "loss": 0.0139, "step": 15270 }, { "grad_norm": 0.3335040807723999, "learning_rate": 7.277683648849153e-05, "loss": 0.015, "step": 15280 }, { "grad_norm": 0.311247318983078, "learning_rate": 7.2740030035737e-05, "loss": 0.0132, "step": 15290 }, { "grad_norm": 0.29124540090560913, "learning_rate": 7.270320804038634e-05, "loss": 0.0161, "step": 15300 }, { "grad_norm": 0.22069507837295532, "learning_rate": 7.266637052760708e-05, "loss": 0.0128, "step": 15310 }, { "grad_norm": 0.24980896711349487, "learning_rate": 7.262951752257728e-05, "loss": 0.0142, "step": 15320 }, { "grad_norm": 0.20158757269382477, "learning_rate": 7.259264905048564e-05, "loss": 0.013, "step": 15330 }, { "grad_norm": 0.24210186302661896, "learning_rate": 7.255576513653142e-05, "loss": 0.0149, "step": 15340 }, { "grad_norm": 0.2642565071582794, "learning_rate": 7.251886580592439e-05, "loss": 0.0144, "step": 15350 }, { "grad_norm": 0.2102830559015274, "learning_rate": 7.248195108388496e-05, "loss": 0.0137, "step": 15360 }, { "grad_norm": 0.17867116630077362, "learning_rate": 7.244502099564395e-05, "loss": 0.0158, "step": 15370 }, { "grad_norm": 0.20124879479408264, "learning_rate": 7.240807556644271e-05, "loss": 0.0157, "step": 15380 }, { "grad_norm": 0.2926758825778961, "learning_rate": 7.237111482153314e-05, "loss": 0.0131, "step": 15390 }, { "grad_norm": 0.2246326506137848, "learning_rate": 7.233413878617751e-05, "loss": 0.0158, "step": 15400 }, { "grad_norm": 0.20268331468105316, "learning_rate": 7.229714748564864e-05, "loss": 0.016, "step": 15410 }, { "grad_norm": 0.23721539974212646, "learning_rate": 7.22601409452297e-05, "loss": 0.0156, "step": 15420 }, { "grad_norm": 0.16326677799224854, "learning_rate": 7.222311919021433e-05, "loss": 0.0158, "step": 15430 }, { "grad_norm": 0.19136366248130798, "learning_rate": 7.218608224590655e-05, "loss": 0.0156, "step": 15440 }, { "grad_norm": 0.20553681254386902, "learning_rate": 7.214903013762074e-05, "loss": 0.0137, "step": 15450 }, { "grad_norm": 0.2354578673839569, "learning_rate": 7.21119628906817e-05, "loss": 0.0146, "step": 15460 }, { "grad_norm": 0.2808685898780823, "learning_rate": 7.207488053042454e-05, "loss": 0.018, "step": 15470 }, { "grad_norm": 0.24005649983882904, "learning_rate": 7.203778308219467e-05, "loss": 0.0165, "step": 15480 }, { "grad_norm": 0.27551016211509705, "learning_rate": 7.200067057134787e-05, "loss": 0.0216, "step": 15490 }, { "grad_norm": 0.26045775413513184, "learning_rate": 7.196354302325019e-05, "loss": 0.0191, "step": 15500 }, { "grad_norm": 0.19112707674503326, "learning_rate": 7.192640046327795e-05, "loss": 0.0173, "step": 15510 }, { "grad_norm": 0.2121107280254364, "learning_rate": 7.188924291681777e-05, "loss": 0.0155, "step": 15520 }, { "grad_norm": 0.22180278599262238, "learning_rate": 7.185207040926643e-05, "loss": 0.0151, "step": 15530 }, { "grad_norm": 0.18105237185955048, "learning_rate": 7.181488296603103e-05, "loss": 0.0141, "step": 15540 }, { "grad_norm": 0.23583900928497314, "learning_rate": 7.177768061252885e-05, "loss": 0.0149, "step": 15550 }, { "grad_norm": 0.23473188281059265, "learning_rate": 7.174046337418729e-05, "loss": 0.0149, "step": 15560 }, { "grad_norm": 0.2059093415737152, "learning_rate": 7.170323127644403e-05, "loss": 0.0142, "step": 15570 }, { "grad_norm": 0.19660869240760803, "learning_rate": 7.166598434474683e-05, "loss": 0.0125, "step": 15580 }, { "grad_norm": 0.2936795949935913, "learning_rate": 7.162872260455364e-05, "loss": 0.0159, "step": 15590 }, { "grad_norm": 0.22672335803508759, "learning_rate": 7.159144608133248e-05, "loss": 0.0162, "step": 15600 }, { "grad_norm": 0.3303685784339905, "learning_rate": 7.155415480056153e-05, "loss": 0.0167, "step": 15610 }, { "grad_norm": 0.2859678864479065, "learning_rate": 7.151684878772902e-05, "loss": 0.0182, "step": 15620 }, { "grad_norm": 0.1798611879348755, "learning_rate": 7.147952806833324e-05, "loss": 0.02, "step": 15630 }, { "grad_norm": 0.26298996806144714, "learning_rate": 7.14421926678826e-05, "loss": 0.0171, "step": 15640 }, { "grad_norm": 0.2201356142759323, "learning_rate": 7.140484261189543e-05, "loss": 0.0162, "step": 15650 }, { "grad_norm": 0.2824734151363373, "learning_rate": 7.136747792590017e-05, "loss": 0.0158, "step": 15660 }, { "grad_norm": 0.24494460225105286, "learning_rate": 7.133009863543524e-05, "loss": 0.015, "step": 15670 }, { "grad_norm": 0.25488799810409546, "learning_rate": 7.129270476604901e-05, "loss": 0.0166, "step": 15680 }, { "grad_norm": 0.2192799150943756, "learning_rate": 7.125529634329988e-05, "loss": 0.0163, "step": 15690 }, { "grad_norm": 0.21994487941265106, "learning_rate": 7.12178733927561e-05, "loss": 0.0138, "step": 15700 }, { "grad_norm": 0.22542308270931244, "learning_rate": 7.118043593999593e-05, "loss": 0.0138, "step": 15710 }, { "grad_norm": 0.31799057126045227, "learning_rate": 7.114298401060752e-05, "loss": 0.0156, "step": 15720 }, { "grad_norm": 0.20968852937221527, "learning_rate": 7.11055176301889e-05, "loss": 0.0191, "step": 15730 }, { "grad_norm": 0.2807031571865082, "learning_rate": 7.1068036824348e-05, "loss": 0.0153, "step": 15740 }, { "grad_norm": 0.20982608199119568, "learning_rate": 7.10305416187026e-05, "loss": 0.0152, "step": 15750 }, { "grad_norm": 0.24046629667282104, "learning_rate": 7.099303203888029e-05, "loss": 0.0159, "step": 15760 }, { "grad_norm": 0.16957391798496246, "learning_rate": 7.095550811051855e-05, "loss": 0.0167, "step": 15770 }, { "grad_norm": 0.29200389981269836, "learning_rate": 7.09179698592646e-05, "loss": 0.0155, "step": 15780 }, { "grad_norm": 0.28612369298934937, "learning_rate": 7.088041731077551e-05, "loss": 0.0144, "step": 15790 }, { "grad_norm": 0.27232587337493896, "learning_rate": 7.084285049071806e-05, "loss": 0.0158, "step": 15800 }, { "grad_norm": 0.21006141602993011, "learning_rate": 7.080526942476886e-05, "loss": 0.0125, "step": 15810 }, { "grad_norm": 0.1978549212217331, "learning_rate": 7.076767413861418e-05, "loss": 0.019, "step": 15820 }, { "grad_norm": 0.183803990483284, "learning_rate": 7.073006465795005e-05, "loss": 0.0121, "step": 15830 }, { "grad_norm": 0.16762909293174744, "learning_rate": 7.06924410084822e-05, "loss": 0.0157, "step": 15840 }, { "grad_norm": 0.21937693655490875, "learning_rate": 7.065480321592604e-05, "loss": 0.0124, "step": 15850 }, { "grad_norm": 0.2010355144739151, "learning_rate": 7.061715130600663e-05, "loss": 0.0146, "step": 15860 }, { "grad_norm": 0.26451602578163147, "learning_rate": 7.057948530445873e-05, "loss": 0.0126, "step": 15870 }, { "grad_norm": 0.1689828634262085, "learning_rate": 7.054180523702668e-05, "loss": 0.0134, "step": 15880 }, { "grad_norm": 0.20583677291870117, "learning_rate": 7.050411112946442e-05, "loss": 0.0132, "step": 15890 }, { "grad_norm": 0.24602735042572021, "learning_rate": 7.046640300753557e-05, "loss": 0.0199, "step": 15900 }, { "grad_norm": 0.279078871011734, "learning_rate": 7.042868089701325e-05, "loss": 0.0176, "step": 15910 }, { "grad_norm": 0.2350328415632248, "learning_rate": 7.039094482368016e-05, "loss": 0.0188, "step": 15920 }, { "grad_norm": 0.2658314108848572, "learning_rate": 7.035319481332858e-05, "loss": 0.0141, "step": 15930 }, { "grad_norm": 0.3158702254295349, "learning_rate": 7.031543089176023e-05, "loss": 0.0143, "step": 15940 }, { "grad_norm": 0.21520280838012695, "learning_rate": 7.027765308478644e-05, "loss": 0.0162, "step": 15950 }, { "grad_norm": 0.20133481919765472, "learning_rate": 7.023986141822798e-05, "loss": 0.0136, "step": 15960 }, { "grad_norm": 0.18913012742996216, "learning_rate": 7.02020559179151e-05, "loss": 0.0127, "step": 15970 }, { "grad_norm": 0.1981777846813202, "learning_rate": 7.016423660968748e-05, "loss": 0.0124, "step": 15980 }, { "grad_norm": 0.16558583080768585, "learning_rate": 7.012640351939428e-05, "loss": 0.0123, "step": 15990 }, { "grad_norm": 0.4108463227748871, "learning_rate": 7.008855667289404e-05, "loss": 0.0162, "step": 16000 }, { "grad_norm": 0.22934590280056, "learning_rate": 7.005069609605476e-05, "loss": 0.0171, "step": 16010 }, { "grad_norm": 0.20720091462135315, "learning_rate": 7.001282181475377e-05, "loss": 0.0149, "step": 16020 }, { "grad_norm": 0.2366197258234024, "learning_rate": 6.997493385487775e-05, "loss": 0.0132, "step": 16030 }, { "grad_norm": 0.26887765526771545, "learning_rate": 6.99370322423228e-05, "loss": 0.0125, "step": 16040 }, { "grad_norm": 0.2489815205335617, "learning_rate": 6.989911700299433e-05, "loss": 0.0134, "step": 16050 }, { "grad_norm": 0.22673283517360687, "learning_rate": 6.9861188162807e-05, "loss": 0.0146, "step": 16060 }, { "grad_norm": 0.22783176600933075, "learning_rate": 6.982324574768487e-05, "loss": 0.0113, "step": 16070 }, { "grad_norm": 0.20309080183506012, "learning_rate": 6.978528978356117e-05, "loss": 0.0129, "step": 16080 }, { "grad_norm": 0.24471652507781982, "learning_rate": 6.974732029637846e-05, "loss": 0.0116, "step": 16090 }, { "grad_norm": 0.3030321002006531, "learning_rate": 6.970933731208855e-05, "loss": 0.0121, "step": 16100 }, { "grad_norm": 0.24205006659030914, "learning_rate": 6.967134085665244e-05, "loss": 0.0123, "step": 16110 }, { "grad_norm": 0.17474733293056488, "learning_rate": 6.963333095604034e-05, "loss": 0.0126, "step": 16120 }, { "grad_norm": 0.24246342480182648, "learning_rate": 6.959530763623166e-05, "loss": 0.014, "step": 16130 }, { "grad_norm": 0.21474164724349976, "learning_rate": 6.955727092321497e-05, "loss": 0.0128, "step": 16140 }, { "grad_norm": 0.25480958819389343, "learning_rate": 6.951922084298803e-05, "loss": 0.0127, "step": 16150 }, { "grad_norm": 0.3232339918613434, "learning_rate": 6.948115742155769e-05, "loss": 0.0137, "step": 16160 }, { "grad_norm": 0.20951049029827118, "learning_rate": 6.944308068493996e-05, "loss": 0.0119, "step": 16170 }, { "grad_norm": 0.23668651282787323, "learning_rate": 6.940499065915992e-05, "loss": 0.0124, "step": 16180 }, { "grad_norm": 0.177730530500412, "learning_rate": 6.936688737025173e-05, "loss": 0.0121, "step": 16190 }, { "grad_norm": 0.20958413183689117, "learning_rate": 6.932877084425867e-05, "loss": 0.0111, "step": 16200 }, { "grad_norm": 0.30547773838043213, "learning_rate": 6.929064110723297e-05, "loss": 0.0118, "step": 16210 }, { "grad_norm": 0.23914441466331482, "learning_rate": 6.925249818523598e-05, "loss": 0.013, "step": 16220 }, { "grad_norm": 0.21129122376441956, "learning_rate": 6.921434210433801e-05, "loss": 0.0118, "step": 16230 }, { "grad_norm": 0.3118106424808502, "learning_rate": 6.917617289061841e-05, "loss": 0.0141, "step": 16240 }, { "grad_norm": 0.21031102538108826, "learning_rate": 6.913799057016547e-05, "loss": 0.0106, "step": 16250 }, { "grad_norm": 0.23303067684173584, "learning_rate": 6.909979516907641e-05, "loss": 0.0154, "step": 16260 }, { "grad_norm": 0.23319590091705322, "learning_rate": 6.906158671345746e-05, "loss": 0.0114, "step": 16270 }, { "grad_norm": 0.25321680307388306, "learning_rate": 6.902336522942374e-05, "loss": 0.0155, "step": 16280 }, { "grad_norm": 0.16708441078662872, "learning_rate": 6.898513074309924e-05, "loss": 0.0108, "step": 16290 }, { "grad_norm": 0.20277652144432068, "learning_rate": 6.894688328061693e-05, "loss": 0.0123, "step": 16300 }, { "grad_norm": 0.19893597066402435, "learning_rate": 6.890862286811853e-05, "loss": 0.0142, "step": 16310 }, { "grad_norm": 0.2159937620162964, "learning_rate": 6.88703495317547e-05, "loss": 0.0121, "step": 16320 }, { "grad_norm": 0.16900841891765594, "learning_rate": 6.883206329768492e-05, "loss": 0.0105, "step": 16330 }, { "grad_norm": 0.19518043100833893, "learning_rate": 6.879376419207743e-05, "loss": 0.0103, "step": 16340 }, { "grad_norm": 0.16644762456417084, "learning_rate": 6.875545224110935e-05, "loss": 0.013, "step": 16350 }, { "grad_norm": 0.2206379771232605, "learning_rate": 6.871712747096651e-05, "loss": 0.015, "step": 16360 }, { "grad_norm": 0.2008971869945526, "learning_rate": 6.867878990784353e-05, "loss": 0.0107, "step": 16370 }, { "grad_norm": 0.21678544580936432, "learning_rate": 6.864043957794377e-05, "loss": 0.0144, "step": 16380 }, { "grad_norm": 0.17731142044067383, "learning_rate": 6.860207650747934e-05, "loss": 0.0116, "step": 16390 }, { "grad_norm": 0.18433301150798798, "learning_rate": 6.856370072267104e-05, "loss": 0.015, "step": 16400 }, { "grad_norm": 0.25954699516296387, "learning_rate": 6.852531224974831e-05, "loss": 0.0151, "step": 16410 }, { "grad_norm": 0.21043485403060913, "learning_rate": 6.848691111494936e-05, "loss": 0.0145, "step": 16420 }, { "grad_norm": 0.24283112585544586, "learning_rate": 6.844849734452097e-05, "loss": 0.0118, "step": 16430 }, { "grad_norm": 0.26104748249053955, "learning_rate": 6.841007096471862e-05, "loss": 0.0131, "step": 16440 }, { "grad_norm": 0.2607068717479706, "learning_rate": 6.837163200180636e-05, "loss": 0.0148, "step": 16450 }, { "grad_norm": 0.20014610886573792, "learning_rate": 6.833318048205684e-05, "loss": 0.0152, "step": 16460 }, { "grad_norm": 0.2347070276737213, "learning_rate": 6.829471643175136e-05, "loss": 0.014, "step": 16470 }, { "grad_norm": 0.2113821655511856, "learning_rate": 6.825623987717969e-05, "loss": 0.0155, "step": 16480 }, { "grad_norm": 0.20232345163822174, "learning_rate": 6.821775084464022e-05, "loss": 0.0134, "step": 16490 }, { "grad_norm": 0.2627025544643402, "learning_rate": 6.817924936043982e-05, "loss": 0.014, "step": 16500 }, { "grad_norm": 0.2315029799938202, "learning_rate": 6.81407354508939e-05, "loss": 0.0101, "step": 16510 }, { "grad_norm": 0.2385248839855194, "learning_rate": 6.810220914232636e-05, "loss": 0.013, "step": 16520 }, { "grad_norm": 0.23693107068538666, "learning_rate": 6.806367046106959e-05, "loss": 0.0135, "step": 16530 }, { "grad_norm": 0.212192565202713, "learning_rate": 6.802511943346435e-05, "loss": 0.0116, "step": 16540 }, { "grad_norm": 0.23794448375701904, "learning_rate": 6.798655608585997e-05, "loss": 0.0098, "step": 16550 }, { "grad_norm": 0.24232998490333557, "learning_rate": 6.79479804446141e-05, "loss": 0.013, "step": 16560 }, { "grad_norm": 0.18332262337207794, "learning_rate": 6.790939253609284e-05, "loss": 0.0147, "step": 16570 }, { "grad_norm": 0.2148701250553131, "learning_rate": 6.787079238667065e-05, "loss": 0.0128, "step": 16580 }, { "grad_norm": 0.13398578763008118, "learning_rate": 6.783218002273039e-05, "loss": 0.0115, "step": 16590 }, { "grad_norm": 0.17778266966342926, "learning_rate": 6.779355547066322e-05, "loss": 0.0123, "step": 16600 }, { "grad_norm": 0.18499422073364258, "learning_rate": 6.775491875686865e-05, "loss": 0.0123, "step": 16610 }, { "grad_norm": 0.20900948345661163, "learning_rate": 6.771626990775457e-05, "loss": 0.0116, "step": 16620 }, { "grad_norm": 0.19623742997646332, "learning_rate": 6.767760894973704e-05, "loss": 0.0143, "step": 16630 }, { "grad_norm": 0.20104621350765228, "learning_rate": 6.763893590924048e-05, "loss": 0.0122, "step": 16640 }, { "grad_norm": 0.2138518989086151, "learning_rate": 6.760025081269756e-05, "loss": 0.0167, "step": 16650 }, { "grad_norm": 0.2444993406534195, "learning_rate": 6.756155368654915e-05, "loss": 0.011, "step": 16660 }, { "grad_norm": 0.3233400285243988, "learning_rate": 6.752284455724442e-05, "loss": 0.0141, "step": 16670 }, { "grad_norm": 0.24678559601306915, "learning_rate": 6.748412345124065e-05, "loss": 0.0123, "step": 16680 }, { "grad_norm": 0.19043444097042084, "learning_rate": 6.744539039500335e-05, "loss": 0.0105, "step": 16690 }, { "grad_norm": 0.1923423409461975, "learning_rate": 6.740664541500625e-05, "loss": 0.0107, "step": 16700 }, { "grad_norm": 0.1657031774520874, "learning_rate": 6.736788853773112e-05, "loss": 0.0123, "step": 16710 }, { "grad_norm": 0.16729678213596344, "learning_rate": 6.732911978966796e-05, "loss": 0.0111, "step": 16720 }, { "grad_norm": 0.2652471959590912, "learning_rate": 6.729033919731482e-05, "loss": 0.012, "step": 16730 }, { "grad_norm": 0.22834250330924988, "learning_rate": 6.725154678717787e-05, "loss": 0.0151, "step": 16740 }, { "grad_norm": 0.25786417722702026, "learning_rate": 6.721274258577138e-05, "loss": 0.0139, "step": 16750 }, { "grad_norm": 0.20459270477294922, "learning_rate": 6.717392661961763e-05, "loss": 0.0126, "step": 16760 }, { "grad_norm": 0.21208730340003967, "learning_rate": 6.713509891524697e-05, "loss": 0.015, "step": 16770 }, { "grad_norm": 0.22279317677021027, "learning_rate": 6.709625949919777e-05, "loss": 0.0144, "step": 16780 }, { "grad_norm": 0.18275386095046997, "learning_rate": 6.705740839801642e-05, "loss": 0.0117, "step": 16790 }, { "grad_norm": 0.1752253919839859, "learning_rate": 6.701854563825727e-05, "loss": 0.0121, "step": 16800 }, { "grad_norm": 0.15058079361915588, "learning_rate": 6.697967124648266e-05, "loss": 0.0087, "step": 16810 }, { "grad_norm": 0.21604712307453156, "learning_rate": 6.694078524926285e-05, "loss": 0.0113, "step": 16820 }, { "grad_norm": 0.23863816261291504, "learning_rate": 6.690188767317607e-05, "loss": 0.0126, "step": 16830 }, { "grad_norm": 0.22243742644786835, "learning_rate": 6.686297854480843e-05, "loss": 0.0117, "step": 16840 }, { "grad_norm": 0.2652052938938141, "learning_rate": 6.682405789075398e-05, "loss": 0.0143, "step": 16850 }, { "grad_norm": 0.2681061923503876, "learning_rate": 6.67851257376146e-05, "loss": 0.0114, "step": 16860 }, { "grad_norm": 0.2063470333814621, "learning_rate": 6.674618211200004e-05, "loss": 0.0112, "step": 16870 }, { "grad_norm": 0.19032126665115356, "learning_rate": 6.670722704052792e-05, "loss": 0.0116, "step": 16880 }, { "grad_norm": 0.13939310610294342, "learning_rate": 6.666826054982365e-05, "loss": 0.0123, "step": 16890 }, { "grad_norm": 0.13565681874752045, "learning_rate": 6.662928266652048e-05, "loss": 0.0122, "step": 16900 }, { "grad_norm": 0.20774218440055847, "learning_rate": 6.659029341725941e-05, "loss": 0.0109, "step": 16910 }, { "grad_norm": 0.2382713407278061, "learning_rate": 6.655129282868923e-05, "loss": 0.0157, "step": 16920 }, { "grad_norm": 0.15194351971149445, "learning_rate": 6.651228092746646e-05, "loss": 0.0122, "step": 16930 }, { "grad_norm": 0.19999302923679352, "learning_rate": 6.647325774025539e-05, "loss": 0.0113, "step": 16940 }, { "grad_norm": 0.2552894949913025, "learning_rate": 6.643422329372798e-05, "loss": 0.0135, "step": 16950 }, { "grad_norm": 0.20413820445537567, "learning_rate": 6.639517761456392e-05, "loss": 0.0146, "step": 16960 }, { "grad_norm": 0.1965920478105545, "learning_rate": 6.635612072945054e-05, "loss": 0.0139, "step": 16970 }, { "grad_norm": 0.23634441196918488, "learning_rate": 6.631705266508289e-05, "loss": 0.0127, "step": 16980 }, { "grad_norm": 0.2337864339351654, "learning_rate": 6.62779734481636e-05, "loss": 0.0123, "step": 16990 }, { "grad_norm": 0.25014710426330566, "learning_rate": 6.623888310540294e-05, "loss": 0.0144, "step": 17000 }, { "grad_norm": 0.3278335630893707, "learning_rate": 6.619978166351882e-05, "loss": 0.0174, "step": 17010 }, { "grad_norm": 0.2311357706785202, "learning_rate": 6.616066914923666e-05, "loss": 0.0149, "step": 17020 }, { "grad_norm": 0.18033547699451447, "learning_rate": 6.612154558928955e-05, "loss": 0.0129, "step": 17030 }, { "grad_norm": 0.1814996898174286, "learning_rate": 6.608241101041804e-05, "loss": 0.0147, "step": 17040 }, { "grad_norm": 0.29492247104644775, "learning_rate": 6.604326543937025e-05, "loss": 0.0154, "step": 17050 }, { "grad_norm": 0.20469063520431519, "learning_rate": 6.60041089029018e-05, "loss": 0.0135, "step": 17060 }, { "grad_norm": 0.15723203122615814, "learning_rate": 6.596494142777583e-05, "loss": 0.0128, "step": 17070 }, { "grad_norm": 0.18681611120700836, "learning_rate": 6.592576304076294e-05, "loss": 0.0114, "step": 17080 }, { "grad_norm": 0.18957270681858063, "learning_rate": 6.588657376864119e-05, "loss": 0.0138, "step": 17090 }, { "grad_norm": 0.16725938022136688, "learning_rate": 6.584737363819605e-05, "loss": 0.0113, "step": 17100 }, { "grad_norm": 0.19342631101608276, "learning_rate": 6.580816267622048e-05, "loss": 0.011, "step": 17110 }, { "grad_norm": 0.24649974703788757, "learning_rate": 6.576894090951478e-05, "loss": 0.0119, "step": 17120 }, { "grad_norm": 0.21125203371047974, "learning_rate": 6.572970836488665e-05, "loss": 0.0113, "step": 17130 }, { "grad_norm": 0.23351268470287323, "learning_rate": 6.569046506915119e-05, "loss": 0.0133, "step": 17140 }, { "grad_norm": 0.2091691941022873, "learning_rate": 6.56512110491308e-05, "loss": 0.0119, "step": 17150 }, { "grad_norm": 0.20418481528759003, "learning_rate": 6.561194633165523e-05, "loss": 0.0129, "step": 17160 }, { "grad_norm": 0.1752399206161499, "learning_rate": 6.557267094356155e-05, "loss": 0.0101, "step": 17170 }, { "grad_norm": 0.15727898478507996, "learning_rate": 6.553338491169414e-05, "loss": 0.0102, "step": 17180 }, { "grad_norm": 0.18705853819847107, "learning_rate": 6.54940882629046e-05, "loss": 0.0145, "step": 17190 }, { "grad_norm": 0.24441002309322357, "learning_rate": 6.545478102405184e-05, "loss": 0.0097, "step": 17200 }, { "grad_norm": 0.13403542339801788, "learning_rate": 6.541546322200199e-05, "loss": 0.0094, "step": 17210 }, { "grad_norm": 0.40081363916397095, "learning_rate": 6.537613488362837e-05, "loss": 0.0136, "step": 17220 }, { "grad_norm": 0.24452365934848785, "learning_rate": 6.533679603581155e-05, "loss": 0.0174, "step": 17230 }, { "grad_norm": 0.19342660903930664, "learning_rate": 6.529744670543926e-05, "loss": 0.0168, "step": 17240 }, { "grad_norm": 0.21609897911548615, "learning_rate": 6.52580869194064e-05, "loss": 0.0127, "step": 17250 }, { "grad_norm": 0.14574483036994934, "learning_rate": 6.521871670461499e-05, "loss": 0.0136, "step": 17260 }, { "grad_norm": 0.1804082989692688, "learning_rate": 6.517933608797422e-05, "loss": 0.0125, "step": 17270 }, { "grad_norm": 0.17018786072731018, "learning_rate": 6.513994509640038e-05, "loss": 0.0121, "step": 17280 }, { "grad_norm": 0.1527400016784668, "learning_rate": 6.510054375681682e-05, "loss": 0.0104, "step": 17290 }, { "grad_norm": 0.21618245542049408, "learning_rate": 6.506113209615398e-05, "loss": 0.0132, "step": 17300 }, { "grad_norm": 0.20424777269363403, "learning_rate": 6.502171014134938e-05, "loss": 0.0112, "step": 17310 }, { "grad_norm": 0.1677703708410263, "learning_rate": 6.498227791934755e-05, "loss": 0.0124, "step": 17320 }, { "grad_norm": 0.21026256680488586, "learning_rate": 6.494283545710003e-05, "loss": 0.0147, "step": 17330 }, { "grad_norm": 0.18611322343349457, "learning_rate": 6.490338278156538e-05, "loss": 0.0132, "step": 17340 }, { "grad_norm": 0.21041521430015564, "learning_rate": 6.486391991970913e-05, "loss": 0.0152, "step": 17350 }, { "grad_norm": 0.1801346391439438, "learning_rate": 6.482444689850377e-05, "loss": 0.0125, "step": 17360 }, { "grad_norm": 0.22849228978157043, "learning_rate": 6.478496374492875e-05, "loss": 0.0114, "step": 17370 }, { "grad_norm": 0.1473359912633896, "learning_rate": 6.474547048597042e-05, "loss": 0.0119, "step": 17380 }, { "grad_norm": 0.21384385228157043, "learning_rate": 6.470596714862205e-05, "loss": 0.0135, "step": 17390 }, { "grad_norm": 0.1798485964536667, "learning_rate": 6.46664537598838e-05, "loss": 0.0148, "step": 17400 }, { "grad_norm": 0.17963837087154388, "learning_rate": 6.462693034676271e-05, "loss": 0.0132, "step": 17410 }, { "grad_norm": 0.2607434093952179, "learning_rate": 6.458739693627265e-05, "loss": 0.013, "step": 17420 }, { "grad_norm": 0.1924399733543396, "learning_rate": 6.454785355543432e-05, "loss": 0.0122, "step": 17430 }, { "grad_norm": 0.19810214638710022, "learning_rate": 6.450830023127528e-05, "loss": 0.0127, "step": 17440 }, { "grad_norm": 0.29822129011154175, "learning_rate": 6.446873699082982e-05, "loss": 0.015, "step": 17450 }, { "grad_norm": 0.15320192277431488, "learning_rate": 6.44291638611391e-05, "loss": 0.0123, "step": 17460 }, { "grad_norm": 0.2598561942577362, "learning_rate": 6.43895808692509e-05, "loss": 0.0138, "step": 17470 }, { "grad_norm": 0.22188641130924225, "learning_rate": 6.434998804221986e-05, "loss": 0.0132, "step": 17480 }, { "grad_norm": 0.18199186027050018, "learning_rate": 6.431038540710732e-05, "loss": 0.0123, "step": 17490 }, { "grad_norm": 0.22404859960079193, "learning_rate": 6.427077299098129e-05, "loss": 0.0126, "step": 17500 }, { "grad_norm": 0.1739821881055832, "learning_rate": 6.423115082091651e-05, "loss": 0.0133, "step": 17510 }, { "grad_norm": 0.19021564722061157, "learning_rate": 6.419151892399429e-05, "loss": 0.0096, "step": 17520 }, { "grad_norm": 0.18552109599113464, "learning_rate": 6.415187732730273e-05, "loss": 0.0118, "step": 17530 }, { "grad_norm": 0.17731668055057526, "learning_rate": 6.411222605793645e-05, "loss": 0.0115, "step": 17540 }, { "grad_norm": 0.18594492971897125, "learning_rate": 6.407256514299674e-05, "loss": 0.0103, "step": 17550 }, { "grad_norm": 0.22089113295078278, "learning_rate": 6.403289460959147e-05, "loss": 0.0114, "step": 17560 }, { "grad_norm": 0.23900249600410461, "learning_rate": 6.399321448483501e-05, "loss": 0.012, "step": 17570 }, { "grad_norm": 0.1714361160993576, "learning_rate": 6.395352479584844e-05, "loss": 0.0119, "step": 17580 }, { "grad_norm": 0.174729585647583, "learning_rate": 6.391382556975923e-05, "loss": 0.0128, "step": 17590 }, { "grad_norm": 0.20615029335021973, "learning_rate": 6.387411683370144e-05, "loss": 0.01, "step": 17600 }, { "grad_norm": 0.21080389618873596, "learning_rate": 6.383439861481562e-05, "loss": 0.0121, "step": 17610 }, { "grad_norm": 0.16536085307598114, "learning_rate": 6.379467094024879e-05, "loss": 0.0105, "step": 17620 }, { "grad_norm": 0.21954679489135742, "learning_rate": 6.375493383715445e-05, "loss": 0.0125, "step": 17630 }, { "grad_norm": 0.22958506643772125, "learning_rate": 6.371518733269254e-05, "loss": 0.0119, "step": 17640 }, { "grad_norm": 0.2401142716407776, "learning_rate": 6.367543145402942e-05, "loss": 0.0092, "step": 17650 }, { "grad_norm": 0.18704234063625336, "learning_rate": 6.363566622833785e-05, "loss": 0.0096, "step": 17660 }, { "grad_norm": 0.18224477767944336, "learning_rate": 6.359589168279698e-05, "loss": 0.0118, "step": 17670 }, { "grad_norm": 0.20670746266841888, "learning_rate": 6.355610784459235e-05, "loss": 0.0142, "step": 17680 }, { "grad_norm": 0.25006353855133057, "learning_rate": 6.351631474091585e-05, "loss": 0.0117, "step": 17690 }, { "grad_norm": 0.22785402834415436, "learning_rate": 6.347651239896566e-05, "loss": 0.0114, "step": 17700 }, { "grad_norm": 0.2479984313249588, "learning_rate": 6.343670084594633e-05, "loss": 0.0105, "step": 17710 }, { "grad_norm": 0.1603367179632187, "learning_rate": 6.339688010906866e-05, "loss": 0.0134, "step": 17720 }, { "grad_norm": 0.2037036120891571, "learning_rate": 6.335705021554975e-05, "loss": 0.0108, "step": 17730 }, { "grad_norm": 0.1740315556526184, "learning_rate": 6.3317211192613e-05, "loss": 0.0139, "step": 17740 }, { "grad_norm": 0.2056688368320465, "learning_rate": 6.327736306748795e-05, "loss": 0.0099, "step": 17750 }, { "grad_norm": 0.2593599259853363, "learning_rate": 6.323750586741047e-05, "loss": 0.0169, "step": 17760 }, { "grad_norm": 0.24928446114063263, "learning_rate": 6.319763961962252e-05, "loss": 0.0132, "step": 17770 }, { "grad_norm": 0.1432783156633377, "learning_rate": 6.315776435137233e-05, "loss": 0.0123, "step": 17780 }, { "grad_norm": 0.14631131291389465, "learning_rate": 6.311788008991432e-05, "loss": 0.0128, "step": 17790 }, { "grad_norm": 0.23280447721481323, "learning_rate": 6.307798686250891e-05, "loss": 0.0114, "step": 17800 }, { "grad_norm": 0.18469072878360748, "learning_rate": 6.303808469642284e-05, "loss": 0.0138, "step": 17810 }, { "grad_norm": 0.27009567618370056, "learning_rate": 6.29981736189288e-05, "loss": 0.013, "step": 17820 }, { "grad_norm": 0.17146198451519012, "learning_rate": 6.295825365730567e-05, "loss": 0.0119, "step": 17830 }, { "grad_norm": 0.22593528032302856, "learning_rate": 6.291832483883835e-05, "loss": 0.0132, "step": 17840 }, { "grad_norm": 0.16426438093185425, "learning_rate": 6.28783871908178e-05, "loss": 0.0121, "step": 17850 }, { "grad_norm": 0.20307084918022156, "learning_rate": 6.283844074054107e-05, "loss": 0.0128, "step": 17860 }, { "grad_norm": 0.20329298079013824, "learning_rate": 6.279848551531112e-05, "loss": 0.0156, "step": 17870 }, { "grad_norm": 0.1890866607427597, "learning_rate": 6.275852154243702e-05, "loss": 0.0107, "step": 17880 }, { "grad_norm": 0.20791000127792358, "learning_rate": 6.271854884923377e-05, "loss": 0.0143, "step": 17890 }, { "grad_norm": 0.20941922068595886, "learning_rate": 6.267856746302228e-05, "loss": 0.0117, "step": 17900 }, { "grad_norm": 0.20501922070980072, "learning_rate": 6.263857741112948e-05, "loss": 0.014, "step": 17910 }, { "grad_norm": 0.2959318459033966, "learning_rate": 6.259857872088821e-05, "loss": 0.0156, "step": 17920 }, { "grad_norm": 0.21754077076911926, "learning_rate": 6.255857141963719e-05, "loss": 0.0112, "step": 17930 }, { "grad_norm": 0.20882244408130646, "learning_rate": 6.251855553472101e-05, "loss": 0.0128, "step": 17940 }, { "grad_norm": 0.2249075472354889, "learning_rate": 6.247853109349016e-05, "loss": 0.0097, "step": 17950 }, { "grad_norm": 0.2873305678367615, "learning_rate": 6.243849812330098e-05, "loss": 0.0144, "step": 17960 }, { "grad_norm": 0.28106528520584106, "learning_rate": 6.239845665151563e-05, "loss": 0.0122, "step": 17970 }, { "grad_norm": 0.20733124017715454, "learning_rate": 6.235840670550204e-05, "loss": 0.0115, "step": 17980 }, { "grad_norm": 0.2312072217464447, "learning_rate": 6.231834831263403e-05, "loss": 0.0128, "step": 17990 }, { "grad_norm": 0.164296954870224, "learning_rate": 6.22782815002911e-05, "loss": 0.0115, "step": 18000 }, { "grad_norm": 0.22490985691547394, "learning_rate": 6.223820629585852e-05, "loss": 0.0123, "step": 18010 }, { "grad_norm": 0.1829034835100174, "learning_rate": 6.219812272672737e-05, "loss": 0.0107, "step": 18020 }, { "grad_norm": 0.16603952646255493, "learning_rate": 6.215803082029434e-05, "loss": 0.0106, "step": 18030 }, { "grad_norm": 0.17466230690479279, "learning_rate": 6.211793060396188e-05, "loss": 0.0105, "step": 18040 }, { "grad_norm": 0.2181352823972702, "learning_rate": 6.207782210513811e-05, "loss": 0.0127, "step": 18050 }, { "grad_norm": 0.26199761033058167, "learning_rate": 6.203770535123683e-05, "loss": 0.0139, "step": 18060 }, { "grad_norm": 0.20279742777347565, "learning_rate": 6.199758036967747e-05, "loss": 0.0146, "step": 18070 }, { "grad_norm": 0.16310648620128632, "learning_rate": 6.195744718788503e-05, "loss": 0.0111, "step": 18080 }, { "grad_norm": 0.2205197960138321, "learning_rate": 6.191730583329021e-05, "loss": 0.0128, "step": 18090 }, { "grad_norm": 0.21941719949245453, "learning_rate": 6.187715633332921e-05, "loss": 0.013, "step": 18100 }, { "grad_norm": 0.2093987613916397, "learning_rate": 6.183699871544386e-05, "loss": 0.0114, "step": 18110 }, { "grad_norm": 0.18696363270282745, "learning_rate": 6.179683300708152e-05, "loss": 0.0178, "step": 18120 }, { "grad_norm": 0.17873598635196686, "learning_rate": 6.175665923569503e-05, "loss": 0.0137, "step": 18130 }, { "grad_norm": 0.20563708245754242, "learning_rate": 6.171647742874281e-05, "loss": 0.0103, "step": 18140 }, { "grad_norm": 0.16173863410949707, "learning_rate": 6.167628761368875e-05, "loss": 0.0096, "step": 18150 }, { "grad_norm": 0.18937376141548157, "learning_rate": 6.163608981800222e-05, "loss": 0.0111, "step": 18160 }, { "grad_norm": 0.22307202219963074, "learning_rate": 6.159588406915803e-05, "loss": 0.0111, "step": 18170 }, { "grad_norm": 0.25771066546440125, "learning_rate": 6.155567039463639e-05, "loss": 0.0144, "step": 18180 }, { "grad_norm": 0.21677352488040924, "learning_rate": 6.151544882192302e-05, "loss": 0.0143, "step": 18190 }, { "grad_norm": 0.19849896430969238, "learning_rate": 6.147521937850895e-05, "loss": 0.0139, "step": 18200 }, { "grad_norm": 0.22558321058750153, "learning_rate": 6.143498209189066e-05, "loss": 0.0121, "step": 18210 }, { "grad_norm": 0.24213765561580658, "learning_rate": 6.139473698956993e-05, "loss": 0.0124, "step": 18220 }, { "grad_norm": 0.21648502349853516, "learning_rate": 6.13544840990539e-05, "loss": 0.0135, "step": 18230 }, { "grad_norm": 0.1977127343416214, "learning_rate": 6.131422344785507e-05, "loss": 0.0133, "step": 18240 }, { "grad_norm": 0.21918386220932007, "learning_rate": 6.127395506349119e-05, "loss": 0.0129, "step": 18250 }, { "grad_norm": 0.21811197698116302, "learning_rate": 6.123367897348533e-05, "loss": 0.0125, "step": 18260 }, { "grad_norm": 0.183407723903656, "learning_rate": 6.119339520536584e-05, "loss": 0.0138, "step": 18270 }, { "grad_norm": 0.18533970415592194, "learning_rate": 6.115310378666625e-05, "loss": 0.0126, "step": 18280 }, { "grad_norm": 0.19355608522891998, "learning_rate": 6.11128047449254e-05, "loss": 0.0111, "step": 18290 }, { "grad_norm": 0.1632988601922989, "learning_rate": 6.107249810768729e-05, "loss": 0.0105, "step": 18300 }, { "grad_norm": 0.2526598870754242, "learning_rate": 6.1032183902501125e-05, "loss": 0.0148, "step": 18310 }, { "grad_norm": 0.24014195799827576, "learning_rate": 6.099186215692131e-05, "loss": 0.0105, "step": 18320 }, { "grad_norm": 0.20755355060100555, "learning_rate": 6.095153289850734e-05, "loss": 0.0147, "step": 18330 }, { "grad_norm": 0.1959662288427353, "learning_rate": 6.0911196154823904e-05, "loss": 0.0127, "step": 18340 }, { "grad_norm": 0.14950594305992126, "learning_rate": 6.087085195344079e-05, "loss": 0.0115, "step": 18350 }, { "grad_norm": 0.15336540341377258, "learning_rate": 6.083050032193286e-05, "loss": 0.0106, "step": 18360 }, { "grad_norm": 0.19440245628356934, "learning_rate": 6.0790141287880097e-05, "loss": 0.0093, "step": 18370 }, { "grad_norm": 0.17678683996200562, "learning_rate": 6.0749774878867496e-05, "loss": 0.0115, "step": 18380 }, { "grad_norm": 0.19451995193958282, "learning_rate": 6.0709401122485146e-05, "loss": 0.0101, "step": 18390 }, { "grad_norm": 0.2447141706943512, "learning_rate": 6.066902004632811e-05, "loss": 0.0105, "step": 18400 }, { "grad_norm": 0.22536629438400269, "learning_rate": 6.062863167799646e-05, "loss": 0.0106, "step": 18410 }, { "grad_norm": 0.17821651697158813, "learning_rate": 6.058823604509529e-05, "loss": 0.0108, "step": 18420 }, { "grad_norm": 0.1798604279756546, "learning_rate": 6.054783317523462e-05, "loss": 0.0122, "step": 18430 }, { "grad_norm": 0.20219627022743225, "learning_rate": 6.050742309602944e-05, "loss": 0.014, "step": 18440 }, { "grad_norm": 0.23235780000686646, "learning_rate": 6.046700583509965e-05, "loss": 0.0132, "step": 18450 }, { "grad_norm": 0.22304530441761017, "learning_rate": 6.042658142007007e-05, "loss": 0.0096, "step": 18460 }, { "grad_norm": 0.15612220764160156, "learning_rate": 6.038614987857041e-05, "loss": 0.0097, "step": 18470 }, { "grad_norm": 0.19289901852607727, "learning_rate": 6.0345711238235224e-05, "loss": 0.0105, "step": 18480 }, { "grad_norm": 0.16860169172286987, "learning_rate": 6.030526552670399e-05, "loss": 0.0137, "step": 18490 }, { "grad_norm": 0.17526276409626007, "learning_rate": 6.0264812771620925e-05, "loss": 0.0116, "step": 18500 }, { "grad_norm": 0.21275852620601654, "learning_rate": 6.022435300063512e-05, "loss": 0.0111, "step": 18510 }, { "grad_norm": 0.21825841069221497, "learning_rate": 6.0183886241400466e-05, "loss": 0.0118, "step": 18520 }, { "grad_norm": 0.20979316532611847, "learning_rate": 6.0143412521575584e-05, "loss": 0.012, "step": 18530 }, { "grad_norm": 0.2955944538116455, "learning_rate": 6.010293186882389e-05, "loss": 0.0128, "step": 18540 }, { "grad_norm": 0.2641124725341797, "learning_rate": 6.0062444310813525e-05, "loss": 0.0134, "step": 18550 }, { "grad_norm": 0.21018671989440918, "learning_rate": 6.0021949875217355e-05, "loss": 0.0128, "step": 18560 }, { "grad_norm": 0.16372352838516235, "learning_rate": 5.998144858971295e-05, "loss": 0.0127, "step": 18570 }, { "grad_norm": 0.2330281287431717, "learning_rate": 5.994094048198257e-05, "loss": 0.0099, "step": 18580 }, { "grad_norm": 0.20876072347164154, "learning_rate": 5.990042557971307e-05, "loss": 0.0133, "step": 18590 }, { "grad_norm": 0.21477490663528442, "learning_rate": 5.985990391059607e-05, "loss": 0.0112, "step": 18600 }, { "grad_norm": 0.21195469796657562, "learning_rate": 5.981937550232771e-05, "loss": 0.0132, "step": 18610 }, { "grad_norm": 0.19788043200969696, "learning_rate": 5.9778840382608794e-05, "loss": 0.014, "step": 18620 }, { "grad_norm": 0.28831368684768677, "learning_rate": 5.9738298579144695e-05, "loss": 0.0135, "step": 18630 }, { "grad_norm": 0.17832762002944946, "learning_rate": 5.9697750119645314e-05, "loss": 0.0116, "step": 18640 }, { "grad_norm": 0.20771755278110504, "learning_rate": 5.96571950318252e-05, "loss": 0.0101, "step": 18650 }, { "grad_norm": 0.25162434577941895, "learning_rate": 5.9616633343403316e-05, "loss": 0.0115, "step": 18660 }, { "grad_norm": 0.1886700987815857, "learning_rate": 5.957606508210324e-05, "loss": 0.0109, "step": 18670 }, { "grad_norm": 0.2130989283323288, "learning_rate": 5.953549027565297e-05, "loss": 0.0133, "step": 18680 }, { "grad_norm": 0.19219321012496948, "learning_rate": 5.949490895178501e-05, "loss": 0.0136, "step": 18690 }, { "grad_norm": 0.29250162839889526, "learning_rate": 5.945432113823632e-05, "loss": 0.0125, "step": 18700 }, { "grad_norm": 0.20967286825180054, "learning_rate": 5.9413726862748276e-05, "loss": 0.0118, "step": 18710 }, { "grad_norm": 0.21635356545448303, "learning_rate": 5.9373126153066694e-05, "loss": 0.0106, "step": 18720 }, { "grad_norm": 0.18672382831573486, "learning_rate": 5.933251903694177e-05, "loss": 0.0144, "step": 18730 }, { "grad_norm": 0.20725110173225403, "learning_rate": 5.929190554212807e-05, "loss": 0.0124, "step": 18740 }, { "grad_norm": 0.2744068503379822, "learning_rate": 5.9251285696384565e-05, "loss": 0.0122, "step": 18750 }, { "grad_norm": 0.1929134875535965, "learning_rate": 5.921065952747451e-05, "loss": 0.0089, "step": 18760 }, { "grad_norm": 0.192900612950325, "learning_rate": 5.917002706316552e-05, "loss": 0.01, "step": 18770 }, { "grad_norm": 0.17947670817375183, "learning_rate": 5.912938833122952e-05, "loss": 0.0108, "step": 18780 }, { "grad_norm": 0.1482563465833664, "learning_rate": 5.908874335944265e-05, "loss": 0.0094, "step": 18790 }, { "grad_norm": 0.2162598967552185, "learning_rate": 5.904809217558542e-05, "loss": 0.0115, "step": 18800 }, { "grad_norm": 0.1281696856021881, "learning_rate": 5.90074348074425e-05, "loss": 0.011, "step": 18810 }, { "grad_norm": 0.22245833277702332, "learning_rate": 5.8966771282802814e-05, "loss": 0.0143, "step": 18820 }, { "grad_norm": 0.20265080034732819, "learning_rate": 5.892610162945952e-05, "loss": 0.0119, "step": 18830 }, { "grad_norm": 0.2612321376800537, "learning_rate": 5.8885425875209924e-05, "loss": 0.0111, "step": 18840 }, { "grad_norm": 0.2233608365058899, "learning_rate": 5.884474404785553e-05, "loss": 0.011, "step": 18850 }, { "grad_norm": 0.23643648624420166, "learning_rate": 5.8804056175201983e-05, "loss": 0.0093, "step": 18860 }, { "grad_norm": 0.177827388048172, "learning_rate": 5.876336228505904e-05, "loss": 0.009, "step": 18870 }, { "grad_norm": 0.19280323386192322, "learning_rate": 5.872266240524062e-05, "loss": 0.0108, "step": 18880 }, { "grad_norm": 0.1571049690246582, "learning_rate": 5.86819565635647e-05, "loss": 0.0126, "step": 18890 }, { "grad_norm": 0.2178749442100525, "learning_rate": 5.8641244787853334e-05, "loss": 0.0097, "step": 18900 }, { "grad_norm": 0.18610738217830658, "learning_rate": 5.860052710593265e-05, "loss": 0.0104, "step": 18910 }, { "grad_norm": 0.14849212765693665, "learning_rate": 5.855980354563276e-05, "loss": 0.0111, "step": 18920 }, { "grad_norm": 0.2186020314693451, "learning_rate": 5.8519074134787874e-05, "loss": 0.0112, "step": 18930 }, { "grad_norm": 0.22710976004600525, "learning_rate": 5.847833890123614e-05, "loss": 0.0103, "step": 18940 }, { "grad_norm": 0.19561025500297546, "learning_rate": 5.8437597872819737e-05, "loss": 0.0114, "step": 18950 }, { "grad_norm": 0.27023300528526306, "learning_rate": 5.839685107738473e-05, "loss": 0.0121, "step": 18960 }, { "grad_norm": 0.40150538086891174, "learning_rate": 5.835609854278118e-05, "loss": 0.0116, "step": 18970 }, { "grad_norm": 0.5797971487045288, "learning_rate": 5.831534029686308e-05, "loss": 0.0127, "step": 18980 }, { "grad_norm": 0.24760903418064117, "learning_rate": 5.82745763674883e-05, "loss": 0.0134, "step": 18990 }, { "grad_norm": 0.22107934951782227, "learning_rate": 5.823380678251861e-05, "loss": 0.0136, "step": 19000 }, { "grad_norm": 0.15573303401470184, "learning_rate": 5.81930315698196e-05, "loss": 0.0115, "step": 19010 }, { "grad_norm": 0.16979561746120453, "learning_rate": 5.815225075726076e-05, "loss": 0.0123, "step": 19020 }, { "grad_norm": 0.22460977733135223, "learning_rate": 5.811146437271543e-05, "loss": 0.0135, "step": 19030 }, { "grad_norm": 0.1702624261379242, "learning_rate": 5.807067244406066e-05, "loss": 0.0094, "step": 19040 }, { "grad_norm": 0.1984783262014389, "learning_rate": 5.8029874999177405e-05, "loss": 0.011, "step": 19050 }, { "grad_norm": 0.148539736866951, "learning_rate": 5.798907206595029e-05, "loss": 0.0094, "step": 19060 }, { "grad_norm": 0.18777590990066528, "learning_rate": 5.794826367226773e-05, "loss": 0.0119, "step": 19070 }, { "grad_norm": 0.21213966608047485, "learning_rate": 5.790744984602193e-05, "loss": 0.014, "step": 19080 }, { "grad_norm": 0.15721094608306885, "learning_rate": 5.786663061510872e-05, "loss": 0.0102, "step": 19090 }, { "grad_norm": 0.1985965371131897, "learning_rate": 5.782580600742765e-05, "loss": 0.0111, "step": 19100 }, { "grad_norm": 0.15305368602275848, "learning_rate": 5.7784976050881965e-05, "loss": 0.0105, "step": 19110 }, { "grad_norm": 0.25520259141921997, "learning_rate": 5.774414077337855e-05, "loss": 0.0116, "step": 19120 }, { "grad_norm": 0.2743077874183655, "learning_rate": 5.770330020282796e-05, "loss": 0.0097, "step": 19130 }, { "grad_norm": 0.20914939045906067, "learning_rate": 5.7662454367144317e-05, "loss": 0.0094, "step": 19140 }, { "grad_norm": 0.2019609957933426, "learning_rate": 5.762160329424536e-05, "loss": 0.0103, "step": 19150 }, { "grad_norm": 0.1939324289560318, "learning_rate": 5.7580747012052416e-05, "loss": 0.0087, "step": 19160 }, { "grad_norm": 0.2114035040140152, "learning_rate": 5.753988554849037e-05, "loss": 0.0098, "step": 19170 }, { "grad_norm": 0.20488856732845306, "learning_rate": 5.749901893148766e-05, "loss": 0.012, "step": 19180 }, { "grad_norm": 0.16647684574127197, "learning_rate": 5.745814718897621e-05, "loss": 0.0104, "step": 19190 }, { "grad_norm": 0.22297441959381104, "learning_rate": 5.74172703488915e-05, "loss": 0.0102, "step": 19200 }, { "grad_norm": 0.1537398397922516, "learning_rate": 5.737638843917242e-05, "loss": 0.0089, "step": 19210 }, { "grad_norm": 0.15320590138435364, "learning_rate": 5.73355014877614e-05, "loss": 0.008, "step": 19220 }, { "grad_norm": 0.21149185299873352, "learning_rate": 5.7294609522604316e-05, "loss": 0.0097, "step": 19230 }, { "grad_norm": 0.24615095555782318, "learning_rate": 5.7253712571650376e-05, "loss": 0.0098, "step": 19240 }, { "grad_norm": 0.2737863063812256, "learning_rate": 5.721281066285229e-05, "loss": 0.0124, "step": 19250 }, { "grad_norm": 0.17411945760250092, "learning_rate": 5.717190382416615e-05, "loss": 0.0098, "step": 19260 }, { "grad_norm": 0.18765807151794434, "learning_rate": 5.713099208355135e-05, "loss": 0.0103, "step": 19270 }, { "grad_norm": 0.1474277675151825, "learning_rate": 5.709007546897074e-05, "loss": 0.0118, "step": 19280 }, { "grad_norm": 0.15674810111522675, "learning_rate": 5.704915400839037e-05, "loss": 0.0107, "step": 19290 }, { "grad_norm": 0.14522436261177063, "learning_rate": 5.700822772977971e-05, "loss": 0.0072, "step": 19300 }, { "grad_norm": 0.12245448678731918, "learning_rate": 5.696729666111148e-05, "loss": 0.0084, "step": 19310 }, { "grad_norm": 0.18135914206504822, "learning_rate": 5.692636083036168e-05, "loss": 0.0096, "step": 19320 }, { "grad_norm": 0.18457834422588348, "learning_rate": 5.688542026550958e-05, "loss": 0.0128, "step": 19330 }, { "grad_norm": 0.15001961588859558, "learning_rate": 5.684447499453763e-05, "loss": 0.011, "step": 19340 }, { "grad_norm": 0.13949881494045258, "learning_rate": 5.680352504543156e-05, "loss": 0.0098, "step": 19350 }, { "grad_norm": 0.2763650715351105, "learning_rate": 5.67625704461803e-05, "loss": 0.0103, "step": 19360 }, { "grad_norm": 0.18916867673397064, "learning_rate": 5.672161122477589e-05, "loss": 0.009, "step": 19370 }, { "grad_norm": 0.19581767916679382, "learning_rate": 5.668064740921359e-05, "loss": 0.0106, "step": 19380 }, { "grad_norm": 0.14993277192115784, "learning_rate": 5.663967902749179e-05, "loss": 0.0113, "step": 19390 }, { "grad_norm": 0.25379326939582825, "learning_rate": 5.6598706107611965e-05, "loss": 0.0103, "step": 19400 }, { "grad_norm": 0.20097176730632782, "learning_rate": 5.655772867757876e-05, "loss": 0.0114, "step": 19410 }, { "grad_norm": 0.15659788250923157, "learning_rate": 5.651674676539982e-05, "loss": 0.0082, "step": 19420 }, { "grad_norm": 0.15142428874969482, "learning_rate": 5.647576039908593e-05, "loss": 0.0086, "step": 19430 }, { "grad_norm": 0.17912398278713226, "learning_rate": 5.6434769606650864e-05, "loss": 0.0093, "step": 19440 }, { "grad_norm": 0.19297172129154205, "learning_rate": 5.639377441611143e-05, "loss": 0.0103, "step": 19450 }, { "grad_norm": 0.2166556566953659, "learning_rate": 5.635277485548751e-05, "loss": 0.0084, "step": 19460 }, { "grad_norm": 0.2243652641773224, "learning_rate": 5.631177095280186e-05, "loss": 0.0113, "step": 19470 }, { "grad_norm": 0.19560283422470093, "learning_rate": 5.627076273608027e-05, "loss": 0.0126, "step": 19480 }, { "grad_norm": 0.18126271665096283, "learning_rate": 5.622975023335148e-05, "loss": 0.0092, "step": 19490 }, { "grad_norm": 0.18809236586093903, "learning_rate": 5.618873347264716e-05, "loss": 0.0086, "step": 19500 }, { "grad_norm": 0.22609661519527435, "learning_rate": 5.614771248200188e-05, "loss": 0.0127, "step": 19510 }, { "grad_norm": 0.1677379310131073, "learning_rate": 5.6106687289453066e-05, "loss": 0.0095, "step": 19520 }, { "grad_norm": 0.16894777119159698, "learning_rate": 5.606565792304108e-05, "loss": 0.0103, "step": 19530 }, { "grad_norm": 0.2657540738582611, "learning_rate": 5.602462441080909e-05, "loss": 0.0105, "step": 19540 }, { "grad_norm": 0.2576170563697815, "learning_rate": 5.5983586780803135e-05, "loss": 0.013, "step": 19550 }, { "grad_norm": 0.16829128563404083, "learning_rate": 5.594254506107205e-05, "loss": 0.0099, "step": 19560 }, { "grad_norm": 0.17185792326927185, "learning_rate": 5.590149927966743e-05, "loss": 0.0117, "step": 19570 }, { "grad_norm": 0.17261038720607758, "learning_rate": 5.58604494646437e-05, "loss": 0.0086, "step": 19580 }, { "grad_norm": 0.2985573709011078, "learning_rate": 5.5819395644058025e-05, "loss": 0.0105, "step": 19590 }, { "grad_norm": 0.22178180515766144, "learning_rate": 5.577833784597031e-05, "loss": 0.012, "step": 19600 }, { "grad_norm": 0.20066215097904205, "learning_rate": 5.573727609844316e-05, "loss": 0.009, "step": 19610 }, { "grad_norm": 0.2802743911743164, "learning_rate": 5.5696210429541884e-05, "loss": 0.0105, "step": 19620 }, { "grad_norm": 0.25332412123680115, "learning_rate": 5.565514086733451e-05, "loss": 0.0089, "step": 19630 }, { "grad_norm": 0.2327655851840973, "learning_rate": 5.5614067439891657e-05, "loss": 0.0118, "step": 19640 }, { "grad_norm": 0.15153080224990845, "learning_rate": 5.557299017528666e-05, "loss": 0.01, "step": 19650 }, { "grad_norm": 0.2006833702325821, "learning_rate": 5.5531909101595436e-05, "loss": 0.0096, "step": 19660 }, { "grad_norm": 0.18255050480365753, "learning_rate": 5.549082424689649e-05, "loss": 0.0102, "step": 19670 }, { "grad_norm": 0.1680525243282318, "learning_rate": 5.544973563927095e-05, "loss": 0.0115, "step": 19680 }, { "grad_norm": 0.18319258093833923, "learning_rate": 5.540864330680249e-05, "loss": 0.0092, "step": 19690 }, { "grad_norm": 0.20732779800891876, "learning_rate": 5.536754727757733e-05, "loss": 0.0083, "step": 19700 }, { "grad_norm": 0.21219028532505035, "learning_rate": 5.532644757968422e-05, "loss": 0.0117, "step": 19710 }, { "grad_norm": 0.1748468577861786, "learning_rate": 5.528534424121441e-05, "loss": 0.0104, "step": 19720 }, { "grad_norm": 0.2432827353477478, "learning_rate": 5.524423729026165e-05, "loss": 0.0132, "step": 19730 }, { "grad_norm": 0.1711128205060959, "learning_rate": 5.5203126754922164e-05, "loss": 0.0093, "step": 19740 }, { "grad_norm": 0.14353500306606293, "learning_rate": 5.5162012663294585e-05, "loss": 0.01, "step": 19750 }, { "grad_norm": 0.1517045944929123, "learning_rate": 5.512089504348003e-05, "loss": 0.0085, "step": 19760 }, { "grad_norm": 0.274541974067688, "learning_rate": 5.5079773923582e-05, "loss": 0.0114, "step": 19770 }, { "grad_norm": 0.16219612956047058, "learning_rate": 5.50386493317064e-05, "loss": 0.0102, "step": 19780 }, { "grad_norm": 0.19512368738651276, "learning_rate": 5.49975212959615e-05, "loss": 0.01, "step": 19790 }, { "grad_norm": 0.12554149329662323, "learning_rate": 5.4956389844457904e-05, "loss": 0.0081, "step": 19800 }, { "grad_norm": 0.1922522634267807, "learning_rate": 5.491525500530859e-05, "loss": 0.0104, "step": 19810 }, { "grad_norm": 0.2295917570590973, "learning_rate": 5.487411680662882e-05, "loss": 0.0107, "step": 19820 }, { "grad_norm": 0.1964811235666275, "learning_rate": 5.483297527653618e-05, "loss": 0.0114, "step": 19830 }, { "grad_norm": 0.20608924329280853, "learning_rate": 5.4791830443150516e-05, "loss": 0.0111, "step": 19840 }, { "grad_norm": 0.15238547325134277, "learning_rate": 5.475068233459392e-05, "loss": 0.0113, "step": 19850 }, { "grad_norm": 0.1833222210407257, "learning_rate": 5.470953097899075e-05, "loss": 0.0116, "step": 19860 }, { "grad_norm": 0.1991361379623413, "learning_rate": 5.466837640446756e-05, "loss": 0.0132, "step": 19870 }, { "grad_norm": 0.1968451887369156, "learning_rate": 5.462721863915312e-05, "loss": 0.0126, "step": 19880 }, { "grad_norm": 0.21767789125442505, "learning_rate": 5.4586057711178374e-05, "loss": 0.0109, "step": 19890 }, { "grad_norm": 0.22600705921649933, "learning_rate": 5.454489364867642e-05, "loss": 0.0115, "step": 19900 }, { "grad_norm": 0.18054422736167908, "learning_rate": 5.4503726479782523e-05, "loss": 0.0131, "step": 19910 }, { "grad_norm": 0.20037922263145447, "learning_rate": 5.446255623263403e-05, "loss": 0.01, "step": 19920 }, { "grad_norm": 0.2311311811208725, "learning_rate": 5.4421382935370445e-05, "loss": 0.0096, "step": 19930 }, { "grad_norm": 0.16880697011947632, "learning_rate": 5.438020661613331e-05, "loss": 0.0112, "step": 19940 }, { "grad_norm": 0.18167656660079956, "learning_rate": 5.433902730306625e-05, "loss": 0.0098, "step": 19950 }, { "grad_norm": 0.17317089438438416, "learning_rate": 5.429784502431495e-05, "loss": 0.0122, "step": 19960 }, { "grad_norm": 0.2291475534439087, "learning_rate": 5.42566598080271e-05, "loss": 0.0136, "step": 19970 }, { "grad_norm": 0.20447802543640137, "learning_rate": 5.421547168235241e-05, "loss": 0.0102, "step": 19980 }, { "grad_norm": 0.16594387590885162, "learning_rate": 5.417428067544258e-05, "loss": 0.0091, "step": 19990 }, { "grad_norm": 0.174717515707016, "learning_rate": 5.413308681545126e-05, "loss": 0.011, "step": 20000 }, { "grad_norm": 0.2184642106294632, "learning_rate": 5.409189013053408e-05, "loss": 0.0129, "step": 20010 }, { "grad_norm": 0.16367460787296295, "learning_rate": 5.4050690648848576e-05, "loss": 0.0103, "step": 20020 }, { "grad_norm": 0.1548432856798172, "learning_rate": 5.400948839855421e-05, "loss": 0.0104, "step": 20030 }, { "grad_norm": 0.24376748502254486, "learning_rate": 5.396828340781234e-05, "loss": 0.0094, "step": 20040 }, { "grad_norm": 0.1996598094701767, "learning_rate": 5.392707570478617e-05, "loss": 0.0088, "step": 20050 }, { "grad_norm": 0.2040952444076538, "learning_rate": 5.388586531764078e-05, "loss": 0.0104, "step": 20060 }, { "grad_norm": 0.30119433999061584, "learning_rate": 5.384465227454311e-05, "loss": 0.012, "step": 20070 }, { "grad_norm": 0.2188057005405426, "learning_rate": 5.380343660366184e-05, "loss": 0.0099, "step": 20080 }, { "grad_norm": 0.18198952078819275, "learning_rate": 5.376221833316752e-05, "loss": 0.0116, "step": 20090 }, { "grad_norm": 0.2761377692222595, "learning_rate": 5.3720997491232436e-05, "loss": 0.0153, "step": 20100 }, { "grad_norm": 0.27416297793388367, "learning_rate": 5.367977410603068e-05, "loss": 0.0125, "step": 20110 }, { "grad_norm": 0.1926116943359375, "learning_rate": 5.3638548205738004e-05, "loss": 0.0138, "step": 20120 }, { "grad_norm": 0.22945143282413483, "learning_rate": 5.359731981853194e-05, "loss": 0.0114, "step": 20130 }, { "grad_norm": 0.21453115344047546, "learning_rate": 5.35560889725917e-05, "loss": 0.0111, "step": 20140 }, { "grad_norm": 0.19036425650119781, "learning_rate": 5.3514855696098176e-05, "loss": 0.0126, "step": 20150 }, { "grad_norm": 0.1644294410943985, "learning_rate": 5.347362001723394e-05, "loss": 0.0097, "step": 20160 }, { "grad_norm": 0.1879550814628601, "learning_rate": 5.3432381964183176e-05, "loss": 0.0096, "step": 20170 }, { "grad_norm": 0.17368996143341064, "learning_rate": 5.3391141565131685e-05, "loss": 0.0095, "step": 20180 }, { "grad_norm": 0.22475875914096832, "learning_rate": 5.3349898848266935e-05, "loss": 0.0102, "step": 20190 }, { "grad_norm": 0.1811273992061615, "learning_rate": 5.330865384177789e-05, "loss": 0.009, "step": 20200 }, { "grad_norm": 0.1627761721611023, "learning_rate": 5.326740657385515e-05, "loss": 0.0098, "step": 20210 }, { "grad_norm": 0.19658954441547394, "learning_rate": 5.322615707269083e-05, "loss": 0.0108, "step": 20220 }, { "grad_norm": 0.1852594017982483, "learning_rate": 5.318490536647856e-05, "loss": 0.0075, "step": 20230 }, { "grad_norm": 0.24519219994544983, "learning_rate": 5.3143651483413524e-05, "loss": 0.0083, "step": 20240 }, { "grad_norm": 0.1753709465265274, "learning_rate": 5.310239545169232e-05, "loss": 0.0108, "step": 20250 }, { "grad_norm": 0.17013408243656158, "learning_rate": 5.30611372995131e-05, "loss": 0.0116, "step": 20260 }, { "grad_norm": 0.21563304960727692, "learning_rate": 5.30198770550754e-05, "loss": 0.0095, "step": 20270 }, { "grad_norm": 0.21341657638549805, "learning_rate": 5.297861474658019e-05, "loss": 0.0096, "step": 20280 }, { "grad_norm": 0.17294447124004364, "learning_rate": 5.29373504022299e-05, "loss": 0.01, "step": 20290 }, { "grad_norm": 0.18722330033779144, "learning_rate": 5.28960840502283e-05, "loss": 0.0083, "step": 20300 }, { "grad_norm": 0.18419650197029114, "learning_rate": 5.285481571878056e-05, "loss": 0.0122, "step": 20310 }, { "grad_norm": 0.20801174640655518, "learning_rate": 5.281354543609321e-05, "loss": 0.0092, "step": 20320 }, { "grad_norm": 0.22419290244579315, "learning_rate": 5.277227323037406e-05, "loss": 0.0103, "step": 20330 }, { "grad_norm": 0.2179199606180191, "learning_rate": 5.273099912983233e-05, "loss": 0.0102, "step": 20340 }, { "grad_norm": 0.2049124389886856, "learning_rate": 5.268972316267843e-05, "loss": 0.0124, "step": 20350 }, { "grad_norm": 0.153096005320549, "learning_rate": 5.26484453571241e-05, "loss": 0.0103, "step": 20360 }, { "grad_norm": 0.16775348782539368, "learning_rate": 5.260716574138235e-05, "loss": 0.0106, "step": 20370 }, { "grad_norm": 0.19456073641777039, "learning_rate": 5.256588434366739e-05, "loss": 0.0116, "step": 20380 }, { "grad_norm": 0.17815053462982178, "learning_rate": 5.25246011921947e-05, "loss": 0.0127, "step": 20390 }, { "grad_norm": 0.2592252790927887, "learning_rate": 5.248331631518089e-05, "loss": 0.0105, "step": 20400 }, { "grad_norm": 0.19680273532867432, "learning_rate": 5.244202974084379e-05, "loss": 0.0096, "step": 20410 }, { "grad_norm": 0.20499958097934723, "learning_rate": 5.240074149740239e-05, "loss": 0.0074, "step": 20420 }, { "grad_norm": 0.1446528136730194, "learning_rate": 5.2359451613076814e-05, "loss": 0.0069, "step": 20430 }, { "grad_norm": 0.1581238955259323, "learning_rate": 5.231816011608832e-05, "loss": 0.01, "step": 20440 }, { "grad_norm": 0.16084209084510803, "learning_rate": 5.227686703465924e-05, "loss": 0.0081, "step": 20450 }, { "grad_norm": 0.16806325316429138, "learning_rate": 5.2235572397013e-05, "loss": 0.0094, "step": 20460 }, { "grad_norm": 0.17718519270420074, "learning_rate": 5.2194276231374114e-05, "loss": 0.0093, "step": 20470 }, { "grad_norm": 0.2307499796152115, "learning_rate": 5.21529785659681e-05, "loss": 0.0108, "step": 20480 }, { "grad_norm": 0.183832049369812, "learning_rate": 5.2111679429021565e-05, "loss": 0.012, "step": 20490 }, { "grad_norm": 0.1927802711725235, "learning_rate": 5.207037884876205e-05, "loss": 0.009, "step": 20500 }, { "grad_norm": 0.16744348406791687, "learning_rate": 5.202907685341809e-05, "loss": 0.0085, "step": 20510 }, { "grad_norm": 0.18598133325576782, "learning_rate": 5.198777347121926e-05, "loss": 0.0132, "step": 20520 }, { "grad_norm": 0.19934764504432678, "learning_rate": 5.194646873039598e-05, "loss": 0.0117, "step": 20530 }, { "grad_norm": 0.17800134420394897, "learning_rate": 5.1905162659179696e-05, "loss": 0.0076, "step": 20540 }, { "grad_norm": 0.18138623237609863, "learning_rate": 5.18638552858027e-05, "loss": 0.01, "step": 20550 }, { "grad_norm": 0.1792236864566803, "learning_rate": 5.182254663849818e-05, "loss": 0.012, "step": 20560 }, { "grad_norm": 0.2799525558948517, "learning_rate": 5.178123674550023e-05, "loss": 0.0103, "step": 20570 }, { "grad_norm": 0.17229534685611725, "learning_rate": 5.173992563504375e-05, "loss": 0.0115, "step": 20580 }, { "grad_norm": 0.204354926943779, "learning_rate": 5.169861333536451e-05, "loss": 0.01, "step": 20590 }, { "grad_norm": 0.1940380036830902, "learning_rate": 5.165729987469907e-05, "loss": 0.0104, "step": 20600 }, { "grad_norm": 0.20120514929294586, "learning_rate": 5.161598528128478e-05, "loss": 0.0104, "step": 20610 }, { "grad_norm": 0.13733407855033875, "learning_rate": 5.157466958335981e-05, "loss": 0.0079, "step": 20620 }, { "grad_norm": 0.21351028978824615, "learning_rate": 5.1533352809163025e-05, "loss": 0.0106, "step": 20630 }, { "grad_norm": 0.15568393468856812, "learning_rate": 5.1492034986934046e-05, "loss": 0.0093, "step": 20640 }, { "grad_norm": 0.20253430306911469, "learning_rate": 5.1450716144913225e-05, "loss": 0.0121, "step": 20650 }, { "grad_norm": 0.16283921897411346, "learning_rate": 5.1409396311341595e-05, "loss": 0.0128, "step": 20660 }, { "grad_norm": 0.14864325523376465, "learning_rate": 5.136807551446089e-05, "loss": 0.014, "step": 20670 }, { "grad_norm": 0.20105548202991486, "learning_rate": 5.132675378251346e-05, "loss": 0.0115, "step": 20680 }, { "grad_norm": 0.19192489981651306, "learning_rate": 5.1285431143742325e-05, "loss": 0.0111, "step": 20690 }, { "grad_norm": 0.3614121675491333, "learning_rate": 5.1244107626391136e-05, "loss": 0.0117, "step": 20700 }, { "grad_norm": 0.19915184378623962, "learning_rate": 5.12027832587041e-05, "loss": 0.0107, "step": 20710 }, { "grad_norm": 0.2683389186859131, "learning_rate": 5.116145806892607e-05, "loss": 0.0149, "step": 20720 }, { "grad_norm": 0.25863879919052124, "learning_rate": 5.1120132085302384e-05, "loss": 0.0168, "step": 20730 }, { "grad_norm": 0.31834399700164795, "learning_rate": 5.107880533607898e-05, "loss": 0.0153, "step": 20740 }, { "grad_norm": 0.3010008633136749, "learning_rate": 5.103747784950231e-05, "loss": 0.0129, "step": 20750 }, { "grad_norm": 0.2574952244758606, "learning_rate": 5.09961496538193e-05, "loss": 0.0133, "step": 20760 }, { "grad_norm": 0.2137957364320755, "learning_rate": 5.095482077727742e-05, "loss": 0.0148, "step": 20770 }, { "grad_norm": 0.21177031099796295, "learning_rate": 5.091349124812452e-05, "loss": 0.0105, "step": 20780 }, { "grad_norm": 0.16476310789585114, "learning_rate": 5.087216109460897e-05, "loss": 0.0107, "step": 20790 }, { "grad_norm": 0.2119765281677246, "learning_rate": 5.083083034497954e-05, "loss": 0.0105, "step": 20800 }, { "grad_norm": 0.2134970873594284, "learning_rate": 5.07894990274854e-05, "loss": 0.0092, "step": 20810 }, { "grad_norm": 0.15553301572799683, "learning_rate": 5.074816717037614e-05, "loss": 0.0091, "step": 20820 }, { "grad_norm": 0.19187168776988983, "learning_rate": 5.070683480190165e-05, "loss": 0.0097, "step": 20830 }, { "grad_norm": 0.2491847574710846, "learning_rate": 5.066550195031223e-05, "loss": 0.0083, "step": 20840 }, { "grad_norm": 0.24104061722755432, "learning_rate": 5.062416864385852e-05, "loss": 0.0094, "step": 20850 }, { "grad_norm": 0.4499252438545227, "learning_rate": 5.058283491079142e-05, "loss": 0.0107, "step": 20860 }, { "grad_norm": 0.19872884452342987, "learning_rate": 5.054150077936216e-05, "loss": 0.0099, "step": 20870 }, { "grad_norm": 0.2293088436126709, "learning_rate": 5.0500166277822214e-05, "loss": 0.0115, "step": 20880 }, { "grad_norm": 0.21488459408283234, "learning_rate": 5.0458831434423334e-05, "loss": 0.0088, "step": 20890 }, { "grad_norm": 0.28084322810173035, "learning_rate": 5.0417496277417506e-05, "loss": 0.0081, "step": 20900 }, { "grad_norm": 0.19063489139080048, "learning_rate": 5.037616083505691e-05, "loss": 0.0107, "step": 20910 }, { "grad_norm": 0.17174765467643738, "learning_rate": 5.0334825135593935e-05, "loss": 0.0094, "step": 20920 }, { "grad_norm": 0.14216721057891846, "learning_rate": 5.029348920728111e-05, "loss": 0.0086, "step": 20930 }, { "grad_norm": 0.202788308262825, "learning_rate": 5.0252153078371186e-05, "loss": 0.0119, "step": 20940 }, { "grad_norm": 0.12287764251232147, "learning_rate": 5.021081677711704e-05, "loss": 0.0094, "step": 20950 }, { "grad_norm": 0.17829181253910065, "learning_rate": 5.016948033177159e-05, "loss": 0.0098, "step": 20960 }, { "grad_norm": 0.1816859394311905, "learning_rate": 5.012814377058793e-05, "loss": 0.0108, "step": 20970 }, { "grad_norm": 0.17818035185337067, "learning_rate": 5.008680712181921e-05, "loss": 0.0132, "step": 20980 }, { "grad_norm": 0.14904199540615082, "learning_rate": 5.0045470413718645e-05, "loss": 0.01, "step": 20990 }, { "grad_norm": 0.15184564888477325, "learning_rate": 5.00041336745395e-05, "loss": 0.0131, "step": 21000 }, { "grad_norm": 0.1672070473432541, "learning_rate": 4.996279693253499e-05, "loss": 0.0122, "step": 21010 }, { "grad_norm": 0.18233776092529297, "learning_rate": 4.992146021595847e-05, "loss": 0.0111, "step": 21020 }, { "grad_norm": 0.2157333493232727, "learning_rate": 4.988012355306313e-05, "loss": 0.0106, "step": 21030 }, { "grad_norm": 0.22944404184818268, "learning_rate": 4.98387869721022e-05, "loss": 0.0128, "step": 21040 }, { "grad_norm": 0.19442352652549744, "learning_rate": 4.9797450501328866e-05, "loss": 0.0143, "step": 21050 }, { "grad_norm": 0.2079644650220871, "learning_rate": 4.97561141689962e-05, "loss": 0.0093, "step": 21060 }, { "grad_norm": 0.21584229171276093, "learning_rate": 4.971477800335721e-05, "loss": 0.0135, "step": 21070 }, { "grad_norm": 0.19744297862052917, "learning_rate": 4.967344203266475e-05, "loss": 0.011, "step": 21080 }, { "grad_norm": 0.16303898394107819, "learning_rate": 4.9632106285171584e-05, "loss": 0.0096, "step": 21090 }, { "grad_norm": 0.170233815908432, "learning_rate": 4.959077078913031e-05, "loss": 0.0097, "step": 21100 }, { "grad_norm": 0.17353826761245728, "learning_rate": 4.954943557279333e-05, "loss": 0.0116, "step": 21110 }, { "grad_norm": 0.2620907425880432, "learning_rate": 4.9508100664412916e-05, "loss": 0.01, "step": 21120 }, { "grad_norm": 0.24191227555274963, "learning_rate": 4.946676609224105e-05, "loss": 0.0124, "step": 21130 }, { "grad_norm": 0.2437361478805542, "learning_rate": 4.942543188452952e-05, "loss": 0.0116, "step": 21140 }, { "grad_norm": 0.24679356813430786, "learning_rate": 4.938409806952988e-05, "loss": 0.0124, "step": 21150 }, { "grad_norm": 0.30946511030197144, "learning_rate": 4.93427646754934e-05, "loss": 0.0104, "step": 21160 }, { "grad_norm": 0.118576779961586, "learning_rate": 4.930143173067108e-05, "loss": 0.0099, "step": 21170 }, { "grad_norm": 0.1988079696893692, "learning_rate": 4.9260099263313565e-05, "loss": 0.0105, "step": 21180 }, { "grad_norm": 0.2299494445323944, "learning_rate": 4.921876730167123e-05, "loss": 0.0113, "step": 21190 }, { "grad_norm": 0.1611454337835312, "learning_rate": 4.917743587399409e-05, "loss": 0.0097, "step": 21200 }, { "grad_norm": 0.13829204440116882, "learning_rate": 4.913610500853178e-05, "loss": 0.0078, "step": 21210 }, { "grad_norm": 0.20395272970199585, "learning_rate": 4.909477473353354e-05, "loss": 0.0092, "step": 21220 }, { "grad_norm": 0.17966701090335846, "learning_rate": 4.9053445077248236e-05, "loss": 0.0097, "step": 21230 }, { "grad_norm": 0.14279773831367493, "learning_rate": 4.901211606792429e-05, "loss": 0.009, "step": 21240 }, { "grad_norm": 0.15580877661705017, "learning_rate": 4.89707877338097e-05, "loss": 0.0078, "step": 21250 }, { "grad_norm": 0.1879449039697647, "learning_rate": 4.892946010315199e-05, "loss": 0.0082, "step": 21260 }, { "grad_norm": 0.12411419302225113, "learning_rate": 4.8888133204198204e-05, "loss": 0.0099, "step": 21270 }, { "grad_norm": 0.1245381087064743, "learning_rate": 4.8846807065194886e-05, "loss": 0.0087, "step": 21280 }, { "grad_norm": 0.1645684838294983, "learning_rate": 4.880548171438806e-05, "loss": 0.0094, "step": 21290 }, { "grad_norm": 0.16743957996368408, "learning_rate": 4.8764157180023245e-05, "loss": 0.0082, "step": 21300 }, { "grad_norm": 0.24112296104431152, "learning_rate": 4.872283349034533e-05, "loss": 0.0111, "step": 21310 }, { "grad_norm": 0.15392784774303436, "learning_rate": 4.8681510673598674e-05, "loss": 0.0098, "step": 21320 }, { "grad_norm": 0.2084713578224182, "learning_rate": 4.8640188758027046e-05, "loss": 0.0071, "step": 21330 }, { "grad_norm": 0.19407854974269867, "learning_rate": 4.859886777187357e-05, "loss": 0.0102, "step": 21340 }, { "grad_norm": 0.11782129108905792, "learning_rate": 4.855754774338077e-05, "loss": 0.0092, "step": 21350 }, { "grad_norm": 0.19940832257270813, "learning_rate": 4.851622870079048e-05, "loss": 0.0084, "step": 21360 }, { "grad_norm": 0.1907869577407837, "learning_rate": 4.847491067234389e-05, "loss": 0.0083, "step": 21370 }, { "grad_norm": 0.17044267058372498, "learning_rate": 4.843359368628146e-05, "loss": 0.0078, "step": 21380 }, { "grad_norm": 0.20956037938594818, "learning_rate": 4.8392277770842975e-05, "loss": 0.01, "step": 21390 }, { "grad_norm": 0.19160787761211395, "learning_rate": 4.83509629542675e-05, "loss": 0.0086, "step": 21400 }, { "grad_norm": 0.13880866765975952, "learning_rate": 4.830964926479329e-05, "loss": 0.0096, "step": 21410 }, { "grad_norm": 0.156000018119812, "learning_rate": 4.826833673065785e-05, "loss": 0.0089, "step": 21420 }, { "grad_norm": 0.20276834070682526, "learning_rate": 4.822702538009794e-05, "loss": 0.0087, "step": 21430 }, { "grad_norm": 0.15172651410102844, "learning_rate": 4.818571524134945e-05, "loss": 0.0098, "step": 21440 }, { "grad_norm": 0.20554037392139435, "learning_rate": 4.8144406342647496e-05, "loss": 0.0128, "step": 21450 }, { "grad_norm": 0.17808769643306732, "learning_rate": 4.81030987122263e-05, "loss": 0.0083, "step": 21460 }, { "grad_norm": 0.1315649300813675, "learning_rate": 4.806179237831926e-05, "loss": 0.0073, "step": 21470 }, { "grad_norm": 0.15167854726314545, "learning_rate": 4.802048736915884e-05, "loss": 0.0077, "step": 21480 }, { "grad_norm": 0.22588610649108887, "learning_rate": 4.797918371297666e-05, "loss": 0.0092, "step": 21490 }, { "grad_norm": 0.14207495748996735, "learning_rate": 4.793788143800334e-05, "loss": 0.0104, "step": 21500 }, { "grad_norm": 0.17262455821037292, "learning_rate": 4.789658057246862e-05, "loss": 0.007, "step": 21510 }, { "grad_norm": 0.18432040512561798, "learning_rate": 4.7855281144601227e-05, "loss": 0.0089, "step": 21520 }, { "grad_norm": 0.1762229949235916, "learning_rate": 4.781398318262897e-05, "loss": 0.0098, "step": 21530 }, { "grad_norm": 0.16297142207622528, "learning_rate": 4.777268671477858e-05, "loss": 0.0111, "step": 21540 }, { "grad_norm": 0.205332949757576, "learning_rate": 4.773139176927582e-05, "loss": 0.0097, "step": 21550 }, { "grad_norm": 0.17664140462875366, "learning_rate": 4.769009837434539e-05, "loss": 0.009, "step": 21560 }, { "grad_norm": 0.14415697753429413, "learning_rate": 4.764880655821095e-05, "loss": 0.008, "step": 21570 }, { "grad_norm": 0.19213618338108063, "learning_rate": 4.760751634909508e-05, "loss": 0.011, "step": 21580 }, { "grad_norm": 0.22426292300224304, "learning_rate": 4.756622777521919e-05, "loss": 0.0101, "step": 21590 }, { "grad_norm": 0.21092604100704193, "learning_rate": 4.752494086480368e-05, "loss": 0.0114, "step": 21600 }, { "grad_norm": 0.26706770062446594, "learning_rate": 4.7483655646067744e-05, "loss": 0.0091, "step": 21610 }, { "grad_norm": 0.17943859100341797, "learning_rate": 4.744237214722944e-05, "loss": 0.0107, "step": 21620 }, { "grad_norm": 0.1381855458021164, "learning_rate": 4.740109039650567e-05, "loss": 0.0094, "step": 21630 }, { "grad_norm": 0.11480539292097092, "learning_rate": 4.73598104221121e-05, "loss": 0.0091, "step": 21640 }, { "grad_norm": 0.1499001830816269, "learning_rate": 4.731853225226322e-05, "loss": 0.0077, "step": 21650 }, { "grad_norm": 0.25512316823005676, "learning_rate": 4.727725591517225e-05, "loss": 0.0089, "step": 21660 }, { "grad_norm": 0.15669405460357666, "learning_rate": 4.723598143905119e-05, "loss": 0.0088, "step": 21670 }, { "grad_norm": 0.14266172051429749, "learning_rate": 4.719470885211077e-05, "loss": 0.0097, "step": 21680 }, { "grad_norm": 0.14008761942386627, "learning_rate": 4.7153438182560387e-05, "loss": 0.0104, "step": 21690 }, { "grad_norm": 0.1613425463438034, "learning_rate": 4.711216945860815e-05, "loss": 0.0095, "step": 21700 }, { "grad_norm": 0.21001818776130676, "learning_rate": 4.707090270846088e-05, "loss": 0.0101, "step": 21710 }, { "grad_norm": 0.18894925713539124, "learning_rate": 4.702963796032397e-05, "loss": 0.0138, "step": 21720 }, { "grad_norm": 0.25832414627075195, "learning_rate": 4.6988375242401514e-05, "loss": 0.013, "step": 21730 }, { "grad_norm": 0.19231443107128143, "learning_rate": 4.694711458289618e-05, "loss": 0.0096, "step": 21740 }, { "grad_norm": 0.16606499254703522, "learning_rate": 4.690585601000925e-05, "loss": 0.0088, "step": 21750 }, { "grad_norm": 0.18898995220661163, "learning_rate": 4.686459955194055e-05, "loss": 0.0092, "step": 21760 }, { "grad_norm": 0.15776371955871582, "learning_rate": 4.6823345236888504e-05, "loss": 0.0113, "step": 21770 }, { "grad_norm": 0.2387278825044632, "learning_rate": 4.678209309305002e-05, "loss": 0.0095, "step": 21780 }, { "grad_norm": 0.168409526348114, "learning_rate": 4.674084314862057e-05, "loss": 0.0099, "step": 21790 }, { "grad_norm": 0.22160394489765167, "learning_rate": 4.669959543179409e-05, "loss": 0.0091, "step": 21800 }, { "grad_norm": 0.18966180086135864, "learning_rate": 4.665834997076303e-05, "loss": 0.0099, "step": 21810 }, { "grad_norm": 0.20710746943950653, "learning_rate": 4.661710679371823e-05, "loss": 0.0078, "step": 21820 }, { "grad_norm": 0.1563362181186676, "learning_rate": 4.657586592884905e-05, "loss": 0.0092, "step": 21830 }, { "grad_norm": 0.18051165342330933, "learning_rate": 4.653462740434322e-05, "loss": 0.0083, "step": 21840 }, { "grad_norm": 0.15942874550819397, "learning_rate": 4.649339124838689e-05, "loss": 0.0103, "step": 21850 }, { "grad_norm": 0.12603691220283508, "learning_rate": 4.6452157489164574e-05, "loss": 0.0088, "step": 21860 }, { "grad_norm": 0.17615394294261932, "learning_rate": 4.6410926154859155e-05, "loss": 0.0104, "step": 21870 }, { "grad_norm": 0.18882869184017181, "learning_rate": 4.636969727365186e-05, "loss": 0.0075, "step": 21880 }, { "grad_norm": 0.19918477535247803, "learning_rate": 4.632847087372226e-05, "loss": 0.009, "step": 21890 }, { "grad_norm": 0.18057405948638916, "learning_rate": 4.628724698324818e-05, "loss": 0.0097, "step": 21900 }, { "grad_norm": 0.22062474489212036, "learning_rate": 4.6246025630405795e-05, "loss": 0.0086, "step": 21910 }, { "grad_norm": 0.20532628893852234, "learning_rate": 4.6204806843369474e-05, "loss": 0.0091, "step": 21920 }, { "grad_norm": 0.13884253799915314, "learning_rate": 4.616359065031191e-05, "loss": 0.0091, "step": 21930 }, { "grad_norm": 0.11827415972948074, "learning_rate": 4.6122377079403946e-05, "loss": 0.0107, "step": 21940 }, { "grad_norm": 0.16734202206134796, "learning_rate": 4.6081166158814695e-05, "loss": 0.0075, "step": 21950 }, { "grad_norm": 0.1525958776473999, "learning_rate": 4.603995791671144e-05, "loss": 0.0098, "step": 21960 }, { "grad_norm": 0.16174748539924622, "learning_rate": 4.599875238125957e-05, "loss": 0.0108, "step": 21970 }, { "grad_norm": 0.23992913961410522, "learning_rate": 4.595754958062273e-05, "loss": 0.0106, "step": 21980 }, { "grad_norm": 0.1582466959953308, "learning_rate": 4.591634954296265e-05, "loss": 0.0089, "step": 21990 }, { "grad_norm": 0.17007730901241302, "learning_rate": 4.587515229643913e-05, "loss": 0.0071, "step": 22000 }, { "grad_norm": 0.15822510421276093, "learning_rate": 4.583395786921013e-05, "loss": 0.0083, "step": 22010 }, { "grad_norm": 0.1222064346075058, "learning_rate": 4.579276628943164e-05, "loss": 0.0082, "step": 22020 }, { "grad_norm": 0.22639931738376617, "learning_rate": 4.575157758525772e-05, "loss": 0.0087, "step": 22030 }, { "grad_norm": 0.2005172222852707, "learning_rate": 4.571039178484046e-05, "loss": 0.0096, "step": 22040 }, { "grad_norm": 0.14983366429805756, "learning_rate": 4.566920891632998e-05, "loss": 0.0079, "step": 22050 }, { "grad_norm": 0.15227444469928741, "learning_rate": 4.562802900787436e-05, "loss": 0.007, "step": 22060 }, { "grad_norm": 0.18994258344173431, "learning_rate": 4.558685208761968e-05, "loss": 0.009, "step": 22070 }, { "grad_norm": 0.195846825838089, "learning_rate": 4.554567818370998e-05, "loss": 0.0073, "step": 22080 }, { "grad_norm": 0.23060965538024902, "learning_rate": 4.550450732428726e-05, "loss": 0.0082, "step": 22090 }, { "grad_norm": 0.1699407696723938, "learning_rate": 4.546333953749137e-05, "loss": 0.011, "step": 22100 }, { "grad_norm": 0.14155060052871704, "learning_rate": 4.5422174851460154e-05, "loss": 0.0082, "step": 22110 }, { "grad_norm": 0.1486051082611084, "learning_rate": 4.538101329432924e-05, "loss": 0.0084, "step": 22120 }, { "grad_norm": 0.19812995195388794, "learning_rate": 4.5339854894232195e-05, "loss": 0.0123, "step": 22130 }, { "grad_norm": 0.11135821044445038, "learning_rate": 4.52986996793004e-05, "loss": 0.0087, "step": 22140 }, { "grad_norm": 0.13855168223381042, "learning_rate": 4.5257547677663024e-05, "loss": 0.0081, "step": 22150 }, { "grad_norm": 0.1780128926038742, "learning_rate": 4.52163989174471e-05, "loss": 0.0071, "step": 22160 }, { "grad_norm": 0.14325395226478577, "learning_rate": 4.51752534267774e-05, "loss": 0.0081, "step": 22170 }, { "grad_norm": 0.2593461275100708, "learning_rate": 4.513411123377649e-05, "loss": 0.009, "step": 22180 }, { "grad_norm": 0.19101360440254211, "learning_rate": 4.5092972366564675e-05, "loss": 0.0087, "step": 22190 }, { "grad_norm": 0.23085899651050568, "learning_rate": 4.505183685325997e-05, "loss": 0.008, "step": 22200 }, { "grad_norm": 0.1457575410604477, "learning_rate": 4.5010704721978125e-05, "loss": 0.0088, "step": 22210 }, { "grad_norm": 0.19140022993087769, "learning_rate": 4.496957600083255e-05, "loss": 0.0107, "step": 22220 }, { "grad_norm": 0.1938515305519104, "learning_rate": 4.4928450717934343e-05, "loss": 0.009, "step": 22230 }, { "grad_norm": 0.2096714824438095, "learning_rate": 4.488732890139227e-05, "loss": 0.0099, "step": 22240 }, { "grad_norm": 0.20769086480140686, "learning_rate": 4.4846210579312665e-05, "loss": 0.0129, "step": 22250 }, { "grad_norm": 0.16219154000282288, "learning_rate": 4.480509577979953e-05, "loss": 0.0105, "step": 22260 }, { "grad_norm": 0.14923645555973053, "learning_rate": 4.476398453095445e-05, "loss": 0.0094, "step": 22270 }, { "grad_norm": 0.16352269053459167, "learning_rate": 4.472287686087656e-05, "loss": 0.0088, "step": 22280 }, { "grad_norm": 0.14912372827529907, "learning_rate": 4.468177279766259e-05, "loss": 0.008, "step": 22290 }, { "grad_norm": 0.15419337153434753, "learning_rate": 4.4640672369406746e-05, "loss": 0.0069, "step": 22300 }, { "grad_norm": 0.2052694857120514, "learning_rate": 4.459957560420082e-05, "loss": 0.0087, "step": 22310 }, { "grad_norm": 0.15307186543941498, "learning_rate": 4.455848253013403e-05, "loss": 0.0081, "step": 22320 }, { "grad_norm": 0.220290407538414, "learning_rate": 4.4517393175293146e-05, "loss": 0.011, "step": 22330 }, { "grad_norm": 0.19359493255615234, "learning_rate": 4.447630756776232e-05, "loss": 0.0098, "step": 22340 }, { "grad_norm": 0.18384630978107452, "learning_rate": 4.443522573562318e-05, "loss": 0.0079, "step": 22350 }, { "grad_norm": 0.15978170931339264, "learning_rate": 4.4394147706954776e-05, "loss": 0.0085, "step": 22360 }, { "grad_norm": 0.13294357061386108, "learning_rate": 4.435307350983355e-05, "loss": 0.008, "step": 22370 }, { "grad_norm": 0.12724179029464722, "learning_rate": 4.4312003172333326e-05, "loss": 0.0073, "step": 22380 }, { "grad_norm": 0.24389833211898804, "learning_rate": 4.427093672252531e-05, "loss": 0.0115, "step": 22390 }, { "grad_norm": 0.2377956211566925, "learning_rate": 4.422987418847802e-05, "loss": 0.0105, "step": 22400 }, { "grad_norm": 0.18506285548210144, "learning_rate": 4.4188815598257325e-05, "loss": 0.0087, "step": 22410 }, { "grad_norm": 0.14076107740402222, "learning_rate": 4.414776097992638e-05, "loss": 0.0105, "step": 22420 }, { "grad_norm": 0.2149863839149475, "learning_rate": 4.4106710361545595e-05, "loss": 0.0103, "step": 22430 }, { "grad_norm": 0.18309377133846283, "learning_rate": 4.406566377117272e-05, "loss": 0.0107, "step": 22440 }, { "grad_norm": 0.1591002196073532, "learning_rate": 4.40246212368627e-05, "loss": 0.0084, "step": 22450 }, { "grad_norm": 0.20935548841953278, "learning_rate": 4.3983582786667715e-05, "loss": 0.0081, "step": 22460 }, { "grad_norm": 0.172771155834198, "learning_rate": 4.394254844863716e-05, "loss": 0.0079, "step": 22470 }, { "grad_norm": 0.24580219388008118, "learning_rate": 4.390151825081762e-05, "loss": 0.0086, "step": 22480 }, { "grad_norm": 0.21395090222358704, "learning_rate": 4.386049222125286e-05, "loss": 0.0089, "step": 22490 }, { "grad_norm": 0.16055022180080414, "learning_rate": 4.3819470387983774e-05, "loss": 0.0094, "step": 22500 }, { "grad_norm": 0.16589276492595673, "learning_rate": 4.377845277904841e-05, "loss": 0.0091, "step": 22510 }, { "grad_norm": 0.12403848022222519, "learning_rate": 4.37374394224819e-05, "loss": 0.0096, "step": 22520 }, { "grad_norm": 0.16811592876911163, "learning_rate": 4.369643034631648e-05, "loss": 0.0097, "step": 22530 }, { "grad_norm": 0.20632357895374298, "learning_rate": 4.365542557858149e-05, "loss": 0.0097, "step": 22540 }, { "grad_norm": 0.1813460737466812, "learning_rate": 4.361442514730329e-05, "loss": 0.0083, "step": 22550 }, { "grad_norm": 0.1533062905073166, "learning_rate": 4.357342908050528e-05, "loss": 0.009, "step": 22560 }, { "grad_norm": 0.1997690498828888, "learning_rate": 4.3532437406207895e-05, "loss": 0.0082, "step": 22570 }, { "grad_norm": 0.15319295227527618, "learning_rate": 4.349145015242856e-05, "loss": 0.0068, "step": 22580 }, { "grad_norm": 0.17305558919906616, "learning_rate": 4.345046734718168e-05, "loss": 0.0077, "step": 22590 }, { "grad_norm": 0.19833584129810333, "learning_rate": 4.34094890184786e-05, "loss": 0.0084, "step": 22600 }, { "grad_norm": 0.14125917851924896, "learning_rate": 4.336851519432765e-05, "loss": 0.0071, "step": 22610 }, { "grad_norm": 0.14561761915683746, "learning_rate": 4.332754590273403e-05, "loss": 0.0061, "step": 22620 }, { "grad_norm": 0.13985282182693481, "learning_rate": 4.3286581171699855e-05, "loss": 0.0072, "step": 22630 }, { "grad_norm": 0.23073925077915192, "learning_rate": 4.324562102922416e-05, "loss": 0.0084, "step": 22640 }, { "grad_norm": 0.23034100234508514, "learning_rate": 4.320466550330278e-05, "loss": 0.0106, "step": 22650 }, { "grad_norm": 0.12506753206253052, "learning_rate": 4.3163714621928466e-05, "loss": 0.0082, "step": 22660 }, { "grad_norm": 0.1995018720626831, "learning_rate": 4.312276841309074e-05, "loss": 0.0119, "step": 22670 }, { "grad_norm": 0.19759902358055115, "learning_rate": 4.3081826904775945e-05, "loss": 0.0088, "step": 22680 }, { "grad_norm": 0.19250483810901642, "learning_rate": 4.3040890124967246e-05, "loss": 0.009, "step": 22690 }, { "grad_norm": 0.162439227104187, "learning_rate": 4.2999958101644537e-05, "loss": 0.0093, "step": 22700 }, { "grad_norm": 0.13206429779529572, "learning_rate": 4.2959030862784435e-05, "loss": 0.0099, "step": 22710 }, { "grad_norm": 0.19317986071109772, "learning_rate": 4.291810843636036e-05, "loss": 0.0074, "step": 22720 }, { "grad_norm": 0.1734963059425354, "learning_rate": 4.2877190850342375e-05, "loss": 0.0072, "step": 22730 }, { "grad_norm": 0.2057945430278778, "learning_rate": 4.2836278132697294e-05, "loss": 0.0092, "step": 22740 }, { "grad_norm": 0.16880574822425842, "learning_rate": 4.279537031138855e-05, "loss": 0.0091, "step": 22750 }, { "grad_norm": 0.1388416737318039, "learning_rate": 4.275446741437625e-05, "loss": 0.0088, "step": 22760 }, { "grad_norm": 0.156434565782547, "learning_rate": 4.2713569469617176e-05, "loss": 0.0093, "step": 22770 }, { "grad_norm": 0.16585613787174225, "learning_rate": 4.267267650506465e-05, "loss": 0.0089, "step": 22780 }, { "grad_norm": 0.16879647970199585, "learning_rate": 4.263178854866866e-05, "loss": 0.0093, "step": 22790 }, { "grad_norm": 0.17882925271987915, "learning_rate": 4.259090562837571e-05, "loss": 0.0087, "step": 22800 }, { "grad_norm": 0.21578805148601532, "learning_rate": 4.255002777212888e-05, "loss": 0.0109, "step": 22810 }, { "grad_norm": 0.1591101735830307, "learning_rate": 4.250915500786783e-05, "loss": 0.0065, "step": 22820 }, { "grad_norm": 0.15424738824367523, "learning_rate": 4.24682873635287e-05, "loss": 0.0092, "step": 22830 }, { "grad_norm": 0.139204204082489, "learning_rate": 4.242742486704414e-05, "loss": 0.0071, "step": 22840 }, { "grad_norm": 0.18622401356697083, "learning_rate": 4.238656754634327e-05, "loss": 0.0089, "step": 22850 }, { "grad_norm": 0.15677717328071594, "learning_rate": 4.234571542935168e-05, "loss": 0.0087, "step": 22860 }, { "grad_norm": 0.1744554489850998, "learning_rate": 4.230486854399144e-05, "loss": 0.0086, "step": 22870 }, { "grad_norm": 0.1544899046421051, "learning_rate": 4.226402691818098e-05, "loss": 0.0093, "step": 22880 }, { "grad_norm": 0.16111211478710175, "learning_rate": 4.2223190579835196e-05, "loss": 0.0085, "step": 22890 }, { "grad_norm": 0.18733537197113037, "learning_rate": 4.218235955686531e-05, "loss": 0.01, "step": 22900 }, { "grad_norm": 0.2023991346359253, "learning_rate": 4.214153387717894e-05, "loss": 0.0078, "step": 22910 }, { "grad_norm": 0.16989505290985107, "learning_rate": 4.210071356868007e-05, "loss": 0.0086, "step": 22920 }, { "grad_norm": 0.23689253628253937, "learning_rate": 4.205989865926898e-05, "loss": 0.0081, "step": 22930 }, { "grad_norm": 0.1558697670698166, "learning_rate": 4.2019089176842294e-05, "loss": 0.0083, "step": 22940 }, { "grad_norm": 0.1733696162700653, "learning_rate": 4.1978285149292894e-05, "loss": 0.0094, "step": 22950 }, { "grad_norm": 0.1979406774044037, "learning_rate": 4.193748660450996e-05, "loss": 0.0094, "step": 22960 }, { "grad_norm": 0.1922760307788849, "learning_rate": 4.189669357037891e-05, "loss": 0.0088, "step": 22970 }, { "grad_norm": 0.1800045669078827, "learning_rate": 4.1855906074781405e-05, "loss": 0.0081, "step": 22980 }, { "grad_norm": 0.1371593177318573, "learning_rate": 4.1815124145595285e-05, "loss": 0.0067, "step": 22990 }, { "grad_norm": 0.16254325211048126, "learning_rate": 4.1774347810694644e-05, "loss": 0.0093, "step": 23000 }, { "grad_norm": 0.1606978327035904, "learning_rate": 4.17335770979497e-05, "loss": 0.0076, "step": 23010 }, { "grad_norm": 0.1766461730003357, "learning_rate": 4.169281203522687e-05, "loss": 0.0079, "step": 23020 }, { "grad_norm": 0.1476108878850937, "learning_rate": 4.1652052650388674e-05, "loss": 0.0066, "step": 23030 }, { "grad_norm": 0.12639275193214417, "learning_rate": 4.1611298971293786e-05, "loss": 0.0088, "step": 23040 }, { "grad_norm": 0.17192651331424713, "learning_rate": 4.1570551025796935e-05, "loss": 0.0077, "step": 23050 }, { "grad_norm": 0.19655263423919678, "learning_rate": 4.152980884174897e-05, "loss": 0.0092, "step": 23060 }, { "grad_norm": 0.1548653244972229, "learning_rate": 4.148907244699682e-05, "loss": 0.0081, "step": 23070 }, { "grad_norm": 0.24891552329063416, "learning_rate": 4.1448341869383395e-05, "loss": 0.0097, "step": 23080 }, { "grad_norm": 0.15609048306941986, "learning_rate": 4.140761713674765e-05, "loss": 0.0081, "step": 23090 }, { "grad_norm": 0.1757558137178421, "learning_rate": 4.1366898276924574e-05, "loss": 0.0086, "step": 23100 }, { "grad_norm": 0.16185300052165985, "learning_rate": 4.132618531774512e-05, "loss": 0.0084, "step": 23110 }, { "grad_norm": 0.1560438573360443, "learning_rate": 4.128547828703622e-05, "loss": 0.0076, "step": 23120 }, { "grad_norm": 0.1356988251209259, "learning_rate": 4.1244777212620725e-05, "loss": 0.0077, "step": 23130 }, { "grad_norm": 0.157565176486969, "learning_rate": 4.120408212231746e-05, "loss": 0.0074, "step": 23140 }, { "grad_norm": 0.17512445151805878, "learning_rate": 4.116339304394111e-05, "loss": 0.009, "step": 23150 }, { "grad_norm": 0.1776580810546875, "learning_rate": 4.112271000530229e-05, "loss": 0.008, "step": 23160 }, { "grad_norm": 0.1996711641550064, "learning_rate": 4.10820330342075e-05, "loss": 0.0078, "step": 23170 }, { "grad_norm": 0.16334062814712524, "learning_rate": 4.1041362158459027e-05, "loss": 0.0076, "step": 23180 }, { "grad_norm": 0.19703897833824158, "learning_rate": 4.1000697405855024e-05, "loss": 0.0106, "step": 23190 }, { "grad_norm": 0.21380487084388733, "learning_rate": 4.096003880418951e-05, "loss": 0.0123, "step": 23200 }, { "grad_norm": 0.17952370643615723, "learning_rate": 4.0919386381252215e-05, "loss": 0.0078, "step": 23210 }, { "grad_norm": 0.17793060839176178, "learning_rate": 4.087874016482872e-05, "loss": 0.0069, "step": 23220 }, { "grad_norm": 0.1665341705083847, "learning_rate": 4.0838100182700295e-05, "loss": 0.0091, "step": 23230 }, { "grad_norm": 0.1578567624092102, "learning_rate": 4.079746646264402e-05, "loss": 0.0086, "step": 23240 }, { "grad_norm": 0.18761056661605835, "learning_rate": 4.075683903243262e-05, "loss": 0.0065, "step": 23250 }, { "grad_norm": 0.15713830292224884, "learning_rate": 4.071621791983462e-05, "loss": 0.0077, "step": 23260 }, { "grad_norm": 0.21674519777297974, "learning_rate": 4.06756031526141e-05, "loss": 0.0079, "step": 23270 }, { "grad_norm": 0.19605214893817902, "learning_rate": 4.063499475853092e-05, "loss": 0.0089, "step": 23280 }, { "grad_norm": 0.1762504130601883, "learning_rate": 4.0594392765340506e-05, "loss": 0.0089, "step": 23290 }, { "grad_norm": 0.1730760633945465, "learning_rate": 4.0553797200793954e-05, "loss": 0.0082, "step": 23300 }, { "grad_norm": 0.17276573181152344, "learning_rate": 4.0513208092637926e-05, "loss": 0.0083, "step": 23310 }, { "grad_norm": 0.20151448249816895, "learning_rate": 4.0472625468614735e-05, "loss": 0.0092, "step": 23320 }, { "grad_norm": 0.19556930661201477, "learning_rate": 4.043204935646218e-05, "loss": 0.0073, "step": 23330 }, { "grad_norm": 0.1770869344472885, "learning_rate": 4.0391479783913675e-05, "loss": 0.0063, "step": 23340 }, { "grad_norm": 0.12383395433425903, "learning_rate": 4.0350916778698155e-05, "loss": 0.0076, "step": 23350 }, { "grad_norm": 0.18294380605220795, "learning_rate": 4.031036036854001e-05, "loss": 0.0094, "step": 23360 }, { "grad_norm": 0.16200917959213257, "learning_rate": 4.026981058115918e-05, "loss": 0.0094, "step": 23370 }, { "grad_norm": 0.17175962030887604, "learning_rate": 4.022926744427108e-05, "loss": 0.0111, "step": 23380 }, { "grad_norm": 0.12882211804389954, "learning_rate": 4.018873098558654e-05, "loss": 0.0079, "step": 23390 }, { "grad_norm": 0.16675937175750732, "learning_rate": 4.014820123281186e-05, "loss": 0.0079, "step": 23400 }, { "grad_norm": 0.19125838577747345, "learning_rate": 4.0107678213648735e-05, "loss": 0.0078, "step": 23410 }, { "grad_norm": 0.14635977149009705, "learning_rate": 4.006716195579428e-05, "loss": 0.0099, "step": 23420 }, { "grad_norm": 0.21946807205677032, "learning_rate": 4.002665248694096e-05, "loss": 0.0098, "step": 23430 }, { "grad_norm": 0.16706973314285278, "learning_rate": 3.998614983477664e-05, "loss": 0.0077, "step": 23440 }, { "grad_norm": 0.22416195273399353, "learning_rate": 3.994565402698448e-05, "loss": 0.0085, "step": 23450 }, { "grad_norm": 0.19663158059120178, "learning_rate": 3.9905165091242975e-05, "loss": 0.0093, "step": 23460 }, { "grad_norm": 0.1186285987496376, "learning_rate": 3.9864683055225936e-05, "loss": 0.0089, "step": 23470 }, { "grad_norm": 0.15343177318572998, "learning_rate": 3.982420794660247e-05, "loss": 0.007, "step": 23480 }, { "grad_norm": 0.15247435867786407, "learning_rate": 3.978373979303691e-05, "loss": 0.0058, "step": 23490 }, { "grad_norm": 0.15074530243873596, "learning_rate": 3.974327862218888e-05, "loss": 0.0082, "step": 23500 }, { "grad_norm": 0.13763247430324554, "learning_rate": 3.970282446171318e-05, "loss": 0.0093, "step": 23510 }, { "grad_norm": 0.23996825516223907, "learning_rate": 3.966237733925988e-05, "loss": 0.0096, "step": 23520 }, { "grad_norm": 0.18791598081588745, "learning_rate": 3.962193728247418e-05, "loss": 0.0101, "step": 23530 }, { "grad_norm": 0.19516319036483765, "learning_rate": 3.958150431899651e-05, "loss": 0.0098, "step": 23540 }, { "grad_norm": 0.14818871021270752, "learning_rate": 3.954107847646238e-05, "loss": 0.008, "step": 23550 }, { "grad_norm": 0.16722472012043, "learning_rate": 3.950065978250249e-05, "loss": 0.0063, "step": 23560 }, { "grad_norm": 0.1480305790901184, "learning_rate": 3.9460248264742624e-05, "loss": 0.0077, "step": 23570 }, { "grad_norm": 0.22256779670715332, "learning_rate": 3.941984395080371e-05, "loss": 0.0113, "step": 23580 }, { "grad_norm": 0.13491585850715637, "learning_rate": 3.937944686830167e-05, "loss": 0.0085, "step": 23590 }, { "grad_norm": 0.1825946569442749, "learning_rate": 3.933905704484756e-05, "loss": 0.008, "step": 23600 }, { "grad_norm": 0.13052822649478912, "learning_rate": 3.929867450804743e-05, "loss": 0.0068, "step": 23610 }, { "grad_norm": 0.12660571932792664, "learning_rate": 3.925829928550237e-05, "loss": 0.0062, "step": 23620 }, { "grad_norm": 0.17227061092853546, "learning_rate": 3.921793140480847e-05, "loss": 0.0073, "step": 23630 }, { "grad_norm": 0.1841241419315338, "learning_rate": 3.917757089355677e-05, "loss": 0.0069, "step": 23640 }, { "grad_norm": 0.141392320394516, "learning_rate": 3.9137217779333326e-05, "loss": 0.0064, "step": 23650 }, { "grad_norm": 0.13457843661308289, "learning_rate": 3.9096872089719083e-05, "loss": 0.0068, "step": 23660 }, { "grad_norm": 0.19155238568782806, "learning_rate": 3.905653385228996e-05, "loss": 0.0077, "step": 23670 }, { "grad_norm": 0.13471585512161255, "learning_rate": 3.901620309461677e-05, "loss": 0.0065, "step": 23680 }, { "grad_norm": 0.17748698592185974, "learning_rate": 3.897587984426518e-05, "loss": 0.0078, "step": 23690 }, { "grad_norm": 0.12703795731067657, "learning_rate": 3.893556412879577e-05, "loss": 0.0081, "step": 23700 }, { "grad_norm": 0.1497892588376999, "learning_rate": 3.889525597576395e-05, "loss": 0.0072, "step": 23710 }, { "grad_norm": 0.21725055575370789, "learning_rate": 3.8854955412719965e-05, "loss": 0.0098, "step": 23720 }, { "grad_norm": 0.18111689388751984, "learning_rate": 3.881466246720887e-05, "loss": 0.0079, "step": 23730 }, { "grad_norm": 0.1608337163925171, "learning_rate": 3.8774377166770484e-05, "loss": 0.0087, "step": 23740 }, { "grad_norm": 0.16650697588920593, "learning_rate": 3.8734099538939474e-05, "loss": 0.0079, "step": 23750 }, { "grad_norm": 0.2065918743610382, "learning_rate": 3.869382961124518e-05, "loss": 0.0092, "step": 23760 }, { "grad_norm": 0.1716650277376175, "learning_rate": 3.8653567411211736e-05, "loss": 0.0072, "step": 23770 }, { "grad_norm": 0.16889120638370514, "learning_rate": 3.8613312966357987e-05, "loss": 0.0081, "step": 23780 }, { "grad_norm": 0.21493414044380188, "learning_rate": 3.857306630419745e-05, "loss": 0.0085, "step": 23790 }, { "grad_norm": 0.2624231278896332, "learning_rate": 3.853282745223834e-05, "loss": 0.0091, "step": 23800 }, { "grad_norm": 0.1542797088623047, "learning_rate": 3.8492596437983546e-05, "loss": 0.0075, "step": 23810 }, { "grad_norm": 0.22859138250350952, "learning_rate": 3.8452373288930586e-05, "loss": 0.009, "step": 23820 }, { "grad_norm": 0.1228453740477562, "learning_rate": 3.841215803257159e-05, "loss": 0.006, "step": 23830 }, { "grad_norm": 0.1139363944530487, "learning_rate": 3.83719506963933e-05, "loss": 0.0085, "step": 23840 }, { "grad_norm": 0.14256709814071655, "learning_rate": 3.8331751307877087e-05, "loss": 0.0074, "step": 23850 }, { "grad_norm": 0.14341118931770325, "learning_rate": 3.82915598944988e-05, "loss": 0.0065, "step": 23860 }, { "grad_norm": 0.20234425365924835, "learning_rate": 3.825137648372893e-05, "loss": 0.0076, "step": 23870 }, { "grad_norm": 0.1278318464756012, "learning_rate": 3.8211201103032465e-05, "loss": 0.0076, "step": 23880 }, { "grad_norm": 0.19347906112670898, "learning_rate": 3.817103377986887e-05, "loss": 0.009, "step": 23890 }, { "grad_norm": 0.16960358619689941, "learning_rate": 3.813087454169215e-05, "loss": 0.0084, "step": 23900 }, { "grad_norm": 0.19831758737564087, "learning_rate": 3.809072341595078e-05, "loss": 0.0078, "step": 23910 }, { "grad_norm": 0.13680081069469452, "learning_rate": 3.8050580430087636e-05, "loss": 0.0068, "step": 23920 }, { "grad_norm": 0.19194252789020538, "learning_rate": 3.8010445611540096e-05, "loss": 0.0064, "step": 23930 }, { "grad_norm": 0.19938336312770844, "learning_rate": 3.797031898773992e-05, "loss": 0.0073, "step": 23940 }, { "grad_norm": 0.2048262655735016, "learning_rate": 3.793020058611329e-05, "loss": 0.0078, "step": 23950 }, { "grad_norm": 0.1952211707830429, "learning_rate": 3.789009043408074e-05, "loss": 0.0093, "step": 23960 }, { "grad_norm": 0.12429679930210114, "learning_rate": 3.7849988559057194e-05, "loss": 0.0072, "step": 23970 }, { "grad_norm": 0.13791006803512573, "learning_rate": 3.78098949884519e-05, "loss": 0.0058, "step": 23980 }, { "grad_norm": 0.17031438648700714, "learning_rate": 3.776980974966843e-05, "loss": 0.007, "step": 23990 }, { "grad_norm": 0.21916024386882782, "learning_rate": 3.772973287010468e-05, "loss": 0.008, "step": 24000 }, { "grad_norm": 0.19029687345027924, "learning_rate": 3.768966437715283e-05, "loss": 0.0074, "step": 24010 }, { "grad_norm": 0.15484748780727386, "learning_rate": 3.7649604298199274e-05, "loss": 0.0088, "step": 24020 }, { "grad_norm": 0.1331060379743576, "learning_rate": 3.760955266062473e-05, "loss": 0.0069, "step": 24030 }, { "grad_norm": 0.140456885099411, "learning_rate": 3.75695094918041e-05, "loss": 0.0083, "step": 24040 }, { "grad_norm": 0.17000830173492432, "learning_rate": 3.752947481910652e-05, "loss": 0.008, "step": 24050 }, { "grad_norm": 0.13221104443073273, "learning_rate": 3.7489448669895324e-05, "loss": 0.0092, "step": 24060 }, { "grad_norm": 0.14909908175468445, "learning_rate": 3.744943107152798e-05, "loss": 0.0071, "step": 24070 }, { "grad_norm": 0.132433220744133, "learning_rate": 3.7409422051356165e-05, "loss": 0.0059, "step": 24080 }, { "grad_norm": 0.1660618931055069, "learning_rate": 3.736942163672564e-05, "loss": 0.0087, "step": 24090 }, { "grad_norm": 0.16307806968688965, "learning_rate": 3.732942985497636e-05, "loss": 0.0077, "step": 24100 }, { "grad_norm": 0.19078993797302246, "learning_rate": 3.728944673344228e-05, "loss": 0.0074, "step": 24110 }, { "grad_norm": 0.25010907649993896, "learning_rate": 3.72494722994515e-05, "loss": 0.0083, "step": 24120 }, { "grad_norm": 0.16527526080608368, "learning_rate": 3.720950658032617e-05, "loss": 0.0074, "step": 24130 }, { "grad_norm": 0.15365129709243774, "learning_rate": 3.716954960338249e-05, "loss": 0.0074, "step": 24140 }, { "grad_norm": 0.13794627785682678, "learning_rate": 3.712960139593066e-05, "loss": 0.0077, "step": 24150 }, { "grad_norm": 0.10652756690979004, "learning_rate": 3.708966198527493e-05, "loss": 0.0073, "step": 24160 }, { "grad_norm": 0.16847233474254608, "learning_rate": 3.704973139871349e-05, "loss": 0.0062, "step": 24170 }, { "grad_norm": 0.18298286199569702, "learning_rate": 3.700980966353853e-05, "loss": 0.0093, "step": 24180 }, { "grad_norm": 0.1630856990814209, "learning_rate": 3.696989680703619e-05, "loss": 0.0097, "step": 24190 }, { "grad_norm": 0.12839588522911072, "learning_rate": 3.69299928564865e-05, "loss": 0.0075, "step": 24200 }, { "grad_norm": 0.1388130486011505, "learning_rate": 3.689009783916345e-05, "loss": 0.0073, "step": 24210 }, { "grad_norm": 0.1544593721628189, "learning_rate": 3.6850211782334895e-05, "loss": 0.0108, "step": 24220 }, { "grad_norm": 0.13851146399974823, "learning_rate": 3.681033471326261e-05, "loss": 0.0094, "step": 24230 }, { "grad_norm": 0.1720241904258728, "learning_rate": 3.677046665920216e-05, "loss": 0.0082, "step": 24240 }, { "grad_norm": 0.29232287406921387, "learning_rate": 3.6730607647403005e-05, "loss": 0.008, "step": 24250 }, { "grad_norm": 0.19856879115104675, "learning_rate": 3.6690757705108416e-05, "loss": 0.0078, "step": 24260 }, { "grad_norm": 0.16406770050525665, "learning_rate": 3.665091685955542e-05, "loss": 0.0079, "step": 24270 }, { "grad_norm": 0.1879664957523346, "learning_rate": 3.6611085137974896e-05, "loss": 0.0079, "step": 24280 }, { "grad_norm": 0.19557124376296997, "learning_rate": 3.657126256759143e-05, "loss": 0.0098, "step": 24290 }, { "grad_norm": 0.12134312838315964, "learning_rate": 3.653144917562335e-05, "loss": 0.0084, "step": 24300 }, { "grad_norm": 0.12476914376020432, "learning_rate": 3.649164498928277e-05, "loss": 0.0084, "step": 24310 }, { "grad_norm": 0.22408874332904816, "learning_rate": 3.645185003577546e-05, "loss": 0.0077, "step": 24320 }, { "grad_norm": 0.13916191458702087, "learning_rate": 3.6412064342300906e-05, "loss": 0.0073, "step": 24330 }, { "grad_norm": 0.2742511034011841, "learning_rate": 3.637228793605224e-05, "loss": 0.008, "step": 24340 }, { "grad_norm": 0.2037440687417984, "learning_rate": 3.6332520844216264e-05, "loss": 0.0076, "step": 24350 }, { "grad_norm": 0.1748759150505066, "learning_rate": 3.6292763093973425e-05, "loss": 0.0089, "step": 24360 }, { "grad_norm": 0.20875723659992218, "learning_rate": 3.6253014712497754e-05, "loss": 0.0089, "step": 24370 }, { "grad_norm": 0.1699463278055191, "learning_rate": 3.621327572695692e-05, "loss": 0.0074, "step": 24380 }, { "grad_norm": 0.15634293854236603, "learning_rate": 3.617354616451211e-05, "loss": 0.0083, "step": 24390 }, { "grad_norm": 0.11984633654356003, "learning_rate": 3.6133826052318116e-05, "loss": 0.0073, "step": 24400 }, { "grad_norm": 0.1345951408147812, "learning_rate": 3.609411541752327e-05, "loss": 0.0081, "step": 24410 }, { "grad_norm": 0.15233813226222992, "learning_rate": 3.6054414287269405e-05, "loss": 0.0072, "step": 24420 }, { "grad_norm": 0.18675920367240906, "learning_rate": 3.601472268869188e-05, "loss": 0.0084, "step": 24430 }, { "grad_norm": 0.16417083144187927, "learning_rate": 3.597504064891952e-05, "loss": 0.0092, "step": 24440 }, { "grad_norm": 0.12476322799921036, "learning_rate": 3.5935368195074636e-05, "loss": 0.0074, "step": 24450 }, { "grad_norm": 0.11895349621772766, "learning_rate": 3.589570535427297e-05, "loss": 0.0065, "step": 24460 }, { "grad_norm": 0.1603594720363617, "learning_rate": 3.585605215362371e-05, "loss": 0.0069, "step": 24470 }, { "grad_norm": 0.19862821698188782, "learning_rate": 3.581640862022941e-05, "loss": 0.0074, "step": 24480 }, { "grad_norm": 0.21112573146820068, "learning_rate": 3.57767747811861e-05, "loss": 0.0088, "step": 24490 }, { "grad_norm": 0.21644169092178345, "learning_rate": 3.573715066358308e-05, "loss": 0.0102, "step": 24500 }, { "grad_norm": 0.16444426774978638, "learning_rate": 3.569753629450311e-05, "loss": 0.0076, "step": 24510 }, { "grad_norm": 0.1549115926027298, "learning_rate": 3.565793170102221e-05, "loss": 0.0085, "step": 24520 }, { "grad_norm": 0.1044875755906105, "learning_rate": 3.561833691020976e-05, "loss": 0.0074, "step": 24530 }, { "grad_norm": 0.19541732966899872, "learning_rate": 3.5578751949128415e-05, "loss": 0.0112, "step": 24540 }, { "grad_norm": 0.11727695912122726, "learning_rate": 3.5539176844834125e-05, "loss": 0.0079, "step": 24550 }, { "grad_norm": 0.13816307485103607, "learning_rate": 3.5499611624376125e-05, "loss": 0.0078, "step": 24560 }, { "grad_norm": 0.11582665145397186, "learning_rate": 3.546005631479684e-05, "loss": 0.0091, "step": 24570 }, { "grad_norm": 0.15484051406383514, "learning_rate": 3.542051094313196e-05, "loss": 0.0065, "step": 24580 }, { "grad_norm": 0.10491538792848587, "learning_rate": 3.5380975536410364e-05, "loss": 0.006, "step": 24590 }, { "grad_norm": 0.15678025782108307, "learning_rate": 3.534145012165415e-05, "loss": 0.0091, "step": 24600 }, { "grad_norm": 0.16082708537578583, "learning_rate": 3.5301934725878546e-05, "loss": 0.0075, "step": 24610 }, { "grad_norm": 0.16432532668113708, "learning_rate": 3.526242937609197e-05, "loss": 0.0087, "step": 24620 }, { "grad_norm": 0.20645460486412048, "learning_rate": 3.522293409929595e-05, "loss": 0.0066, "step": 24630 }, { "grad_norm": 0.20506775379180908, "learning_rate": 3.518344892248513e-05, "loss": 0.0062, "step": 24640 }, { "grad_norm": 0.16093303263187408, "learning_rate": 3.514397387264725e-05, "loss": 0.0073, "step": 24650 }, { "grad_norm": 0.13702842593193054, "learning_rate": 3.5104508976763176e-05, "loss": 0.0076, "step": 24660 }, { "grad_norm": 0.14750301837921143, "learning_rate": 3.506505426180674e-05, "loss": 0.0072, "step": 24670 }, { "grad_norm": 0.16847500205039978, "learning_rate": 3.502560975474488e-05, "loss": 0.0082, "step": 24680 }, { "grad_norm": 0.16072167456150055, "learning_rate": 3.4986175482537566e-05, "loss": 0.0093, "step": 24690 }, { "grad_norm": 0.1256771981716156, "learning_rate": 3.4946751472137725e-05, "loss": 0.0074, "step": 24700 }, { "grad_norm": 0.1766708642244339, "learning_rate": 3.490733775049132e-05, "loss": 0.0069, "step": 24710 }, { "grad_norm": 0.23068317770957947, "learning_rate": 3.4867934344537236e-05, "loss": 0.0096, "step": 24720 }, { "grad_norm": 0.20195023715496063, "learning_rate": 3.482854128120735e-05, "loss": 0.007, "step": 24730 }, { "grad_norm": 0.22380219399929047, "learning_rate": 3.478915858742643e-05, "loss": 0.0077, "step": 24740 }, { "grad_norm": 0.13403373956680298, "learning_rate": 3.4749786290112205e-05, "loss": 0.0076, "step": 24750 }, { "grad_norm": 0.21480624377727509, "learning_rate": 3.471042441617524e-05, "loss": 0.0104, "step": 24760 }, { "grad_norm": 0.15881574153900146, "learning_rate": 3.467107299251902e-05, "loss": 0.0064, "step": 24770 }, { "grad_norm": 0.12091132998466492, "learning_rate": 3.463173204603984e-05, "loss": 0.0072, "step": 24780 }, { "grad_norm": 0.1288212686777115, "learning_rate": 3.4592401603626924e-05, "loss": 0.0082, "step": 24790 }, { "grad_norm": 0.1754387468099594, "learning_rate": 3.45530816921622e-05, "loss": 0.0071, "step": 24800 }, { "grad_norm": 0.20053182542324066, "learning_rate": 3.451377233852051e-05, "loss": 0.0086, "step": 24810 }, { "grad_norm": 0.21422439813613892, "learning_rate": 3.4474473569569385e-05, "loss": 0.0062, "step": 24820 }, { "grad_norm": 0.14685381948947906, "learning_rate": 3.443518541216918e-05, "loss": 0.0068, "step": 24830 }, { "grad_norm": 0.12069250643253326, "learning_rate": 3.439590789317299e-05, "loss": 0.0062, "step": 24840 }, { "grad_norm": 0.21216554939746857, "learning_rate": 3.4356641039426607e-05, "loss": 0.0078, "step": 24850 }, { "grad_norm": 0.1428918093442917, "learning_rate": 3.431738487776857e-05, "loss": 0.0056, "step": 24860 }, { "grad_norm": 0.14330369234085083, "learning_rate": 3.4278139435030084e-05, "loss": 0.0063, "step": 24870 }, { "grad_norm": 0.1215585470199585, "learning_rate": 3.423890473803504e-05, "loss": 0.0066, "step": 24880 }, { "grad_norm": 0.17011886835098267, "learning_rate": 3.41996808136e-05, "loss": 0.0051, "step": 24890 }, { "grad_norm": 0.15071089565753937, "learning_rate": 3.416046768853413e-05, "loss": 0.0071, "step": 24900 }, { "grad_norm": 0.16558139026165009, "learning_rate": 3.412126538963925e-05, "loss": 0.0069, "step": 24910 }, { "grad_norm": 0.19848401844501495, "learning_rate": 3.4082073943709727e-05, "loss": 0.0089, "step": 24920 }, { "grad_norm": 0.21778330206871033, "learning_rate": 3.404289337753258e-05, "loss": 0.0074, "step": 24930 }, { "grad_norm": 0.26717162132263184, "learning_rate": 3.400372371788736e-05, "loss": 0.0072, "step": 24940 }, { "grad_norm": 0.1551148146390915, "learning_rate": 3.3964564991546124e-05, "loss": 0.008, "step": 24950 }, { "grad_norm": 0.15967687964439392, "learning_rate": 3.392541722527351e-05, "loss": 0.0078, "step": 24960 }, { "grad_norm": 0.17469939589500427, "learning_rate": 3.3886280445826644e-05, "loss": 0.0061, "step": 24970 }, { "grad_norm": 0.17400102317333221, "learning_rate": 3.3847154679955154e-05, "loss": 0.009, "step": 24980 }, { "grad_norm": 0.15491466224193573, "learning_rate": 3.380803995440113e-05, "loss": 0.0079, "step": 24990 }, { "grad_norm": 0.15050525963306427, "learning_rate": 3.3768936295899115e-05, "loss": 0.0076, "step": 25000 }, { "grad_norm": 0.14526648819446564, "learning_rate": 3.3729843731176094e-05, "loss": 0.0067, "step": 25010 }, { "grad_norm": 0.13332775235176086, "learning_rate": 3.369076228695146e-05, "loss": 0.0062, "step": 25020 }, { "grad_norm": 0.14968839287757874, "learning_rate": 3.365169198993703e-05, "loss": 0.0079, "step": 25030 }, { "grad_norm": 0.12281221896409988, "learning_rate": 3.361263286683697e-05, "loss": 0.0059, "step": 25040 }, { "grad_norm": 0.14164075255393982, "learning_rate": 3.35735849443478e-05, "loss": 0.0072, "step": 25050 }, { "grad_norm": 0.12881749868392944, "learning_rate": 3.3534548249158435e-05, "loss": 0.0069, "step": 25060 }, { "grad_norm": 0.19527044892311096, "learning_rate": 3.3495522807950086e-05, "loss": 0.0091, "step": 25070 }, { "grad_norm": 0.1134667694568634, "learning_rate": 3.345650864739627e-05, "loss": 0.0073, "step": 25080 }, { "grad_norm": 0.1818232238292694, "learning_rate": 3.3417505794162794e-05, "loss": 0.0063, "step": 25090 }, { "grad_norm": 0.13013581931591034, "learning_rate": 3.3378514274907745e-05, "loss": 0.0072, "step": 25100 }, { "grad_norm": 0.15935204923152924, "learning_rate": 3.333953411628147e-05, "loss": 0.007, "step": 25110 }, { "grad_norm": 0.15025505423545837, "learning_rate": 3.330056534492653e-05, "loss": 0.008, "step": 25120 }, { "grad_norm": 0.1916799396276474, "learning_rate": 3.32616079874777e-05, "loss": 0.0067, "step": 25130 }, { "grad_norm": 0.14521045982837677, "learning_rate": 3.322266207056197e-05, "loss": 0.0082, "step": 25140 }, { "grad_norm": 0.1575697660446167, "learning_rate": 3.318372762079852e-05, "loss": 0.0073, "step": 25150 }, { "grad_norm": 0.1392722874879837, "learning_rate": 3.3144804664798666e-05, "loss": 0.0062, "step": 25160 }, { "grad_norm": 0.18009623885154724, "learning_rate": 3.3105893229165894e-05, "loss": 0.0069, "step": 25170 }, { "grad_norm": 0.1294155716896057, "learning_rate": 3.30669933404958e-05, "loss": 0.0076, "step": 25180 }, { "grad_norm": 0.19902783632278442, "learning_rate": 3.302810502537609e-05, "loss": 0.0076, "step": 25190 }, { "grad_norm": 0.18002432584762573, "learning_rate": 3.298922831038655e-05, "loss": 0.0059, "step": 25200 }, { "grad_norm": 0.1425257921218872, "learning_rate": 3.2950363222099073e-05, "loss": 0.0065, "step": 25210 }, { "grad_norm": 0.18748794496059418, "learning_rate": 3.291150978707758e-05, "loss": 0.0097, "step": 25220 }, { "grad_norm": 0.17420026659965515, "learning_rate": 3.287266803187798e-05, "loss": 0.0086, "step": 25230 }, { "grad_norm": 0.17774954438209534, "learning_rate": 3.283383798304829e-05, "loss": 0.0075, "step": 25240 }, { "grad_norm": 0.1384962946176529, "learning_rate": 3.279501966712847e-05, "loss": 0.0066, "step": 25250 }, { "grad_norm": 0.11868537962436676, "learning_rate": 3.275621311065047e-05, "loss": 0.0063, "step": 25260 }, { "grad_norm": 0.14864353835582733, "learning_rate": 3.271741834013822e-05, "loss": 0.008, "step": 25270 }, { "grad_norm": 0.14181162416934967, "learning_rate": 3.267863538210756e-05, "loss": 0.0065, "step": 25280 }, { "grad_norm": 0.1473168134689331, "learning_rate": 3.2639864263066296e-05, "loss": 0.0084, "step": 25290 }, { "grad_norm": 0.1327088177204132, "learning_rate": 3.26011050095141e-05, "loss": 0.008, "step": 25300 }, { "grad_norm": 0.18808075785636902, "learning_rate": 3.256235764794259e-05, "loss": 0.0097, "step": 25310 }, { "grad_norm": 0.18096043169498444, "learning_rate": 3.2523622204835194e-05, "loss": 0.0085, "step": 25320 }, { "grad_norm": 0.11234124004840851, "learning_rate": 3.2484898706667214e-05, "loss": 0.0069, "step": 25330 }, { "grad_norm": 0.11206568032503128, "learning_rate": 3.2446187179905806e-05, "loss": 0.0065, "step": 25340 }, { "grad_norm": 0.16554486751556396, "learning_rate": 3.240748765100995e-05, "loss": 0.0059, "step": 25350 }, { "grad_norm": 0.12253739684820175, "learning_rate": 3.236880014643039e-05, "loss": 0.0056, "step": 25360 }, { "grad_norm": 0.24005600810050964, "learning_rate": 3.233012469260969e-05, "loss": 0.0083, "step": 25370 }, { "grad_norm": 0.17757432162761688, "learning_rate": 3.229146131598213e-05, "loss": 0.0085, "step": 25380 }, { "grad_norm": 0.20709000527858734, "learning_rate": 3.2252810042973794e-05, "loss": 0.0086, "step": 25390 }, { "grad_norm": 0.15632416307926178, "learning_rate": 3.2214170900002456e-05, "loss": 0.0086, "step": 25400 }, { "grad_norm": 0.11972818523645401, "learning_rate": 3.217554391347758e-05, "loss": 0.0107, "step": 25410 }, { "grad_norm": 0.11626319587230682, "learning_rate": 3.213692910980037e-05, "loss": 0.0091, "step": 25420 }, { "grad_norm": 0.11454514414072037, "learning_rate": 3.2098326515363666e-05, "loss": 0.0065, "step": 25430 }, { "grad_norm": 0.17402653396129608, "learning_rate": 3.205973615655199e-05, "loss": 0.0077, "step": 25440 }, { "grad_norm": 0.14940489828586578, "learning_rate": 3.202115805974149e-05, "loss": 0.0054, "step": 25450 }, { "grad_norm": 0.1881413459777832, "learning_rate": 3.1982592251299916e-05, "loss": 0.0084, "step": 25460 }, { "grad_norm": 0.18329265713691711, "learning_rate": 3.1944038757586656e-05, "loss": 0.0107, "step": 25470 }, { "grad_norm": 0.16334889829158783, "learning_rate": 3.190549760495263e-05, "loss": 0.0095, "step": 25480 }, { "grad_norm": 0.15437784790992737, "learning_rate": 3.186696881974039e-05, "loss": 0.0085, "step": 25490 }, { "grad_norm": 0.2226571887731552, "learning_rate": 3.1828452428283986e-05, "loss": 0.007, "step": 25500 }, { "grad_norm": 0.1443440467119217, "learning_rate": 3.178994845690898e-05, "loss": 0.0076, "step": 25510 }, { "grad_norm": 0.16522443294525146, "learning_rate": 3.17514569319325e-05, "loss": 0.0086, "step": 25520 }, { "grad_norm": 0.17743290960788727, "learning_rate": 3.171297787966312e-05, "loss": 0.0088, "step": 25530 }, { "grad_norm": 0.11704829335212708, "learning_rate": 3.167451132640093e-05, "loss": 0.0075, "step": 25540 }, { "grad_norm": 0.13997876644134521, "learning_rate": 3.163605729843746e-05, "loss": 0.0054, "step": 25550 }, { "grad_norm": 0.1600646823644638, "learning_rate": 3.159761582205565e-05, "loss": 0.0059, "step": 25560 }, { "grad_norm": 0.15422628819942474, "learning_rate": 3.155918692352992e-05, "loss": 0.0069, "step": 25570 }, { "grad_norm": 0.13281117379665375, "learning_rate": 3.152077062912602e-05, "loss": 0.0074, "step": 25580 }, { "grad_norm": 0.21918800473213196, "learning_rate": 3.148236696510117e-05, "loss": 0.0089, "step": 25590 }, { "grad_norm": 0.13347363471984863, "learning_rate": 3.144397595770388e-05, "loss": 0.0079, "step": 25600 }, { "grad_norm": 0.15900632739067078, "learning_rate": 3.1405597633174036e-05, "loss": 0.0064, "step": 25610 }, { "grad_norm": 0.17425723373889923, "learning_rate": 3.136723201774289e-05, "loss": 0.0072, "step": 25620 }, { "grad_norm": 0.20270495116710663, "learning_rate": 3.132887913763295e-05, "loss": 0.0078, "step": 25630 }, { "grad_norm": 0.1973976343870163, "learning_rate": 3.129053901905806e-05, "loss": 0.0064, "step": 25640 }, { "grad_norm": 0.1257094144821167, "learning_rate": 3.125221168822335e-05, "loss": 0.0066, "step": 25650 }, { "grad_norm": 0.1874718815088272, "learning_rate": 3.1213897171325154e-05, "loss": 0.0068, "step": 25660 }, { "grad_norm": 0.1754520684480667, "learning_rate": 3.1175595494551116e-05, "loss": 0.0093, "step": 25670 }, { "grad_norm": 0.16771647334098816, "learning_rate": 3.1137306684080045e-05, "loss": 0.0079, "step": 25680 }, { "grad_norm": 0.12413648515939713, "learning_rate": 3.1099030766081985e-05, "loss": 0.0051, "step": 25690 }, { "grad_norm": 0.1073007583618164, "learning_rate": 3.106076776671818e-05, "loss": 0.0077, "step": 25700 }, { "grad_norm": 0.2134809046983719, "learning_rate": 3.102251771214101e-05, "loss": 0.0082, "step": 25710 }, { "grad_norm": 0.15798419713974, "learning_rate": 3.098428062849404e-05, "loss": 0.0058, "step": 25720 }, { "grad_norm": 0.1634490191936493, "learning_rate": 3.094605654191195e-05, "loss": 0.0075, "step": 25730 }, { "grad_norm": 0.1832493394613266, "learning_rate": 3.090784547852055e-05, "loss": 0.0087, "step": 25740 }, { "grad_norm": 0.20955489575862885, "learning_rate": 3.0869647464436746e-05, "loss": 0.008, "step": 25750 }, { "grad_norm": 0.19975583255290985, "learning_rate": 3.0831462525768496e-05, "loss": 0.008, "step": 25760 }, { "grad_norm": 0.13235588371753693, "learning_rate": 3.079329068861488e-05, "loss": 0.006, "step": 25770 }, { "grad_norm": 0.1647636890411377, "learning_rate": 3.075513197906597e-05, "loss": 0.0069, "step": 25780 }, { "grad_norm": 0.18677178025245667, "learning_rate": 3.071698642320286e-05, "loss": 0.0067, "step": 25790 }, { "grad_norm": 0.12464045733213425, "learning_rate": 3.067885404709772e-05, "loss": 0.0053, "step": 25800 }, { "grad_norm": 0.11205881834030151, "learning_rate": 3.0640734876813636e-05, "loss": 0.0064, "step": 25810 }, { "grad_norm": 0.16776634752750397, "learning_rate": 3.060262893840473e-05, "loss": 0.0081, "step": 25820 }, { "grad_norm": 0.15686576068401337, "learning_rate": 3.056453625791603e-05, "loss": 0.0088, "step": 25830 }, { "grad_norm": 0.20781691372394562, "learning_rate": 3.052645686138353e-05, "loss": 0.0073, "step": 25840 }, { "grad_norm": 0.13573528826236725, "learning_rate": 3.0488390774834153e-05, "loss": 0.0073, "step": 25850 }, { "grad_norm": 0.22599287331104279, "learning_rate": 3.0450338024285684e-05, "loss": 0.0103, "step": 25860 }, { "grad_norm": 0.15338249504566193, "learning_rate": 3.0412298635746855e-05, "loss": 0.0058, "step": 25870 }, { "grad_norm": 0.1502707302570343, "learning_rate": 3.03742726352172e-05, "loss": 0.0101, "step": 25880 }, { "grad_norm": 0.19466528296470642, "learning_rate": 3.0336260048687125e-05, "loss": 0.0072, "step": 25890 }, { "grad_norm": 0.20132456719875336, "learning_rate": 3.0298260902137897e-05, "loss": 0.0067, "step": 25900 }, { "grad_norm": 0.21745577454566956, "learning_rate": 3.0260275221541566e-05, "loss": 0.0093, "step": 25910 }, { "grad_norm": 0.1555107980966568, "learning_rate": 3.0222303032860987e-05, "loss": 0.0086, "step": 25920 }, { "grad_norm": 0.2144501805305481, "learning_rate": 3.018434436204979e-05, "loss": 0.0059, "step": 25930 }, { "grad_norm": 0.139094278216362, "learning_rate": 3.014639923505237e-05, "loss": 0.0056, "step": 25940 }, { "grad_norm": 0.1482415646314621, "learning_rate": 3.0108467677803863e-05, "loss": 0.0075, "step": 25950 }, { "grad_norm": 0.1371762454509735, "learning_rate": 3.0070549716230156e-05, "loss": 0.006, "step": 25960 }, { "grad_norm": 0.16126373410224915, "learning_rate": 3.003264537624777e-05, "loss": 0.0062, "step": 25970 }, { "grad_norm": 0.2098364233970642, "learning_rate": 2.9994754683764e-05, "loss": 0.0062, "step": 25980 }, { "grad_norm": 0.15703190863132477, "learning_rate": 2.9956877664676754e-05, "loss": 0.0061, "step": 25990 }, { "grad_norm": 0.12547078728675842, "learning_rate": 2.9919014344874636e-05, "loss": 0.0056, "step": 26000 }, { "grad_norm": 0.15533339977264404, "learning_rate": 2.9881164750236857e-05, "loss": 0.01, "step": 26010 }, { "grad_norm": 0.15699702501296997, "learning_rate": 2.984332890663326e-05, "loss": 0.0061, "step": 26020 }, { "grad_norm": 0.13862892985343933, "learning_rate": 2.9805506839924292e-05, "loss": 0.0088, "step": 26030 }, { "grad_norm": 0.15038882195949554, "learning_rate": 2.9767698575960968e-05, "loss": 0.0064, "step": 26040 }, { "grad_norm": 0.13580457866191864, "learning_rate": 2.9729904140584913e-05, "loss": 0.0059, "step": 26050 }, { "grad_norm": 0.1600235551595688, "learning_rate": 2.9692123559628234e-05, "loss": 0.0063, "step": 26060 }, { "grad_norm": 0.13856671750545502, "learning_rate": 2.9654356858913596e-05, "loss": 0.0071, "step": 26070 }, { "grad_norm": 0.12321677803993225, "learning_rate": 2.9616604064254206e-05, "loss": 0.0064, "step": 26080 }, { "grad_norm": 0.12734036147594452, "learning_rate": 2.9578865201453732e-05, "loss": 0.0067, "step": 26090 }, { "grad_norm": 0.17272697389125824, "learning_rate": 2.9541140296306335e-05, "loss": 0.007, "step": 26100 }, { "grad_norm": 0.17229802906513214, "learning_rate": 2.9503429374596627e-05, "loss": 0.0057, "step": 26110 }, { "grad_norm": 0.1408192664384842, "learning_rate": 2.946573246209967e-05, "loss": 0.0053, "step": 26120 }, { "grad_norm": 0.13830511271953583, "learning_rate": 2.942804958458094e-05, "loss": 0.0064, "step": 26130 }, { "grad_norm": 0.1494608074426651, "learning_rate": 2.9390380767796343e-05, "loss": 0.0069, "step": 26140 }, { "grad_norm": 0.15492428839206696, "learning_rate": 2.9352726037492174e-05, "loss": 0.0079, "step": 26150 }, { "grad_norm": 0.10966068506240845, "learning_rate": 2.9315085419405052e-05, "loss": 0.0057, "step": 26160 }, { "grad_norm": 0.1628919094800949, "learning_rate": 2.927745893926199e-05, "loss": 0.0083, "step": 26170 }, { "grad_norm": 0.12596742808818817, "learning_rate": 2.9239846622780358e-05, "loss": 0.0058, "step": 26180 }, { "grad_norm": 0.15992262959480286, "learning_rate": 2.9202248495667788e-05, "loss": 0.0059, "step": 26190 }, { "grad_norm": 0.20856964588165283, "learning_rate": 2.916466458362227e-05, "loss": 0.0079, "step": 26200 }, { "grad_norm": 0.13814876973628998, "learning_rate": 2.9127094912332033e-05, "loss": 0.0086, "step": 26210 }, { "grad_norm": 0.16276516020298004, "learning_rate": 2.9089539507475606e-05, "loss": 0.0067, "step": 26220 }, { "grad_norm": 0.15777833759784698, "learning_rate": 2.9051998394721748e-05, "loss": 0.0088, "step": 26230 }, { "grad_norm": 0.12440195679664612, "learning_rate": 2.901447159972948e-05, "loss": 0.0057, "step": 26240 }, { "grad_norm": 0.1150512844324112, "learning_rate": 2.8976959148148e-05, "loss": 0.0051, "step": 26250 }, { "grad_norm": 0.12758955359458923, "learning_rate": 2.8939461065616674e-05, "loss": 0.007, "step": 26260 }, { "grad_norm": 0.11921824514865875, "learning_rate": 2.8901977377765127e-05, "loss": 0.0063, "step": 26270 }, { "grad_norm": 0.1336388736963272, "learning_rate": 2.8864508110213094e-05, "loss": 0.0046, "step": 26280 }, { "grad_norm": 0.11376829445362091, "learning_rate": 2.8827053288570503e-05, "loss": 0.0051, "step": 26290 }, { "grad_norm": 0.12082868069410324, "learning_rate": 2.8789612938437315e-05, "loss": 0.0067, "step": 26300 }, { "grad_norm": 0.1973475068807602, "learning_rate": 2.8752187085403683e-05, "loss": 0.0052, "step": 26310 }, { "grad_norm": 0.12103558331727982, "learning_rate": 2.8714775755049818e-05, "loss": 0.0065, "step": 26320 }, { "grad_norm": 0.15749366581439972, "learning_rate": 2.867737897294604e-05, "loss": 0.0076, "step": 26330 }, { "grad_norm": 0.1398387998342514, "learning_rate": 2.8639996764652653e-05, "loss": 0.0078, "step": 26340 }, { "grad_norm": 0.12288416922092438, "learning_rate": 2.8602629155720084e-05, "loss": 0.0095, "step": 26350 }, { "grad_norm": 0.20454829931259155, "learning_rate": 2.8565276171688703e-05, "loss": 0.0061, "step": 26360 }, { "grad_norm": 0.11975521594285965, "learning_rate": 2.8527937838088943e-05, "loss": 0.0055, "step": 26370 }, { "grad_norm": 0.16598327457904816, "learning_rate": 2.84906141804412e-05, "loss": 0.0073, "step": 26380 }, { "grad_norm": 0.14148980379104614, "learning_rate": 2.8453305224255867e-05, "loss": 0.0077, "step": 26390 }, { "grad_norm": 0.11345098912715912, "learning_rate": 2.8416010995033216e-05, "loss": 0.0064, "step": 26400 }, { "grad_norm": 0.14312008023262024, "learning_rate": 2.8378731518263524e-05, "loss": 0.0066, "step": 26410 }, { "grad_norm": 0.14447733759880066, "learning_rate": 2.834146681942696e-05, "loss": 0.006, "step": 26420 }, { "grad_norm": 0.1631643921136856, "learning_rate": 2.8304216923993622e-05, "loss": 0.007, "step": 26430 }, { "grad_norm": 0.13611342012882233, "learning_rate": 2.8266981857423413e-05, "loss": 0.0068, "step": 26440 }, { "grad_norm": 0.12432707101106644, "learning_rate": 2.8229761645166197e-05, "loss": 0.0061, "step": 26450 }, { "grad_norm": 0.12074702233076096, "learning_rate": 2.81925563126616e-05, "loss": 0.0068, "step": 26460 }, { "grad_norm": 0.12638495862483978, "learning_rate": 2.8155365885339124e-05, "loss": 0.0061, "step": 26470 }, { "grad_norm": 0.13546304404735565, "learning_rate": 2.8118190388618093e-05, "loss": 0.0065, "step": 26480 }, { "grad_norm": 0.14377251267433167, "learning_rate": 2.8081029847907607e-05, "loss": 0.0076, "step": 26490 }, { "grad_norm": 0.13038845360279083, "learning_rate": 2.8043884288606525e-05, "loss": 0.0063, "step": 26500 }, { "grad_norm": 0.15376786887645721, "learning_rate": 2.8006753736103496e-05, "loss": 0.0071, "step": 26510 }, { "grad_norm": 0.30768173933029175, "learning_rate": 2.7969638215776918e-05, "loss": 0.0067, "step": 26520 }, { "grad_norm": 0.16350440680980682, "learning_rate": 2.793253775299487e-05, "loss": 0.0066, "step": 26530 }, { "grad_norm": 0.16124969720840454, "learning_rate": 2.7895452373115184e-05, "loss": 0.0086, "step": 26540 }, { "grad_norm": 0.15984328091144562, "learning_rate": 2.785838210148539e-05, "loss": 0.0067, "step": 26550 }, { "grad_norm": 0.12811757624149323, "learning_rate": 2.782132696344263e-05, "loss": 0.0068, "step": 26560 }, { "grad_norm": 0.14065444469451904, "learning_rate": 2.7784286984313745e-05, "loss": 0.0088, "step": 26570 }, { "grad_norm": 0.16054227948188782, "learning_rate": 2.7747262189415236e-05, "loss": 0.0089, "step": 26580 }, { "grad_norm": 0.18773473799228668, "learning_rate": 2.7710252604053205e-05, "loss": 0.0065, "step": 26590 }, { "grad_norm": 0.10415136814117432, "learning_rate": 2.767325825352332e-05, "loss": 0.006, "step": 26600 }, { "grad_norm": 0.12890638411045074, "learning_rate": 2.7636279163110913e-05, "loss": 0.0081, "step": 26610 }, { "grad_norm": 0.1120980829000473, "learning_rate": 2.7599315358090795e-05, "loss": 0.0082, "step": 26620 }, { "grad_norm": 0.16573169827461243, "learning_rate": 2.7562366863727407e-05, "loss": 0.0062, "step": 26630 }, { "grad_norm": 0.1286522001028061, "learning_rate": 2.7525433705274695e-05, "loss": 0.0061, "step": 26640 }, { "grad_norm": 0.08535593748092651, "learning_rate": 2.748851590797614e-05, "loss": 0.0055, "step": 26650 }, { "grad_norm": 0.15177224576473236, "learning_rate": 2.7451613497064675e-05, "loss": 0.0064, "step": 26660 }, { "grad_norm": 0.1353566199541092, "learning_rate": 2.7414726497762765e-05, "loss": 0.0057, "step": 26670 }, { "grad_norm": 0.09841427206993103, "learning_rate": 2.737785493528232e-05, "loss": 0.0054, "step": 26680 }, { "grad_norm": 0.12875285744667053, "learning_rate": 2.7340998834824745e-05, "loss": 0.0056, "step": 26690 }, { "grad_norm": 0.1848643720149994, "learning_rate": 2.7304158221580777e-05, "loss": 0.0062, "step": 26700 }, { "grad_norm": 0.17202681303024292, "learning_rate": 2.7267333120730675e-05, "loss": 0.0076, "step": 26710 }, { "grad_norm": 0.17423544824123383, "learning_rate": 2.7230523557444017e-05, "loss": 0.0066, "step": 26720 }, { "grad_norm": 0.13097509741783142, "learning_rate": 2.7193729556879798e-05, "loss": 0.0056, "step": 26730 }, { "grad_norm": 0.14203894138336182, "learning_rate": 2.715695114418637e-05, "loss": 0.0049, "step": 26740 }, { "grad_norm": 0.15796194970607758, "learning_rate": 2.7120188344501475e-05, "loss": 0.0065, "step": 26750 }, { "grad_norm": 0.20876413583755493, "learning_rate": 2.7083441182952067e-05, "loss": 0.0068, "step": 26760 }, { "grad_norm": 0.13370278477668762, "learning_rate": 2.7046709684654527e-05, "loss": 0.0055, "step": 26770 }, { "grad_norm": 0.16409419476985931, "learning_rate": 2.700999387471448e-05, "loss": 0.0082, "step": 26780 }, { "grad_norm": 0.0967918336391449, "learning_rate": 2.6973293778226854e-05, "loss": 0.0043, "step": 26790 }, { "grad_norm": 0.17133347690105438, "learning_rate": 2.6936609420275804e-05, "loss": 0.007, "step": 26800 }, { "grad_norm": 0.15138830244541168, "learning_rate": 2.689994082593472e-05, "loss": 0.0061, "step": 26810 }, { "grad_norm": 0.20689769089221954, "learning_rate": 2.6863288020266264e-05, "loss": 0.0064, "step": 26820 }, { "grad_norm": 0.13649210333824158, "learning_rate": 2.682665102832228e-05, "loss": 0.0071, "step": 26830 }, { "grad_norm": 0.1762411892414093, "learning_rate": 2.67900298751438e-05, "loss": 0.0063, "step": 26840 }, { "grad_norm": 0.10006038844585419, "learning_rate": 2.6753424585761067e-05, "loss": 0.0056, "step": 26850 }, { "grad_norm": 0.1648489236831665, "learning_rate": 2.671683518519341e-05, "loss": 0.005, "step": 26860 }, { "grad_norm": 0.16594198346138, "learning_rate": 2.668026169844936e-05, "loss": 0.0071, "step": 26870 }, { "grad_norm": 0.15711469948291779, "learning_rate": 2.6643704150526538e-05, "loss": 0.0067, "step": 26880 }, { "grad_norm": 0.15970200300216675, "learning_rate": 2.6607162566411716e-05, "loss": 0.008, "step": 26890 }, { "grad_norm": 0.1583797186613083, "learning_rate": 2.6570636971080697e-05, "loss": 0.0067, "step": 26900 }, { "grad_norm": 0.15338332951068878, "learning_rate": 2.6534127389498364e-05, "loss": 0.0059, "step": 26910 }, { "grad_norm": 0.13812333345413208, "learning_rate": 2.6497633846618696e-05, "loss": 0.0065, "step": 26920 }, { "grad_norm": 0.13477516174316406, "learning_rate": 2.6461156367384677e-05, "loss": 0.0065, "step": 26930 }, { "grad_norm": 0.13607573509216309, "learning_rate": 2.6424694976728316e-05, "loss": 0.0068, "step": 26940 }, { "grad_norm": 0.2080758512020111, "learning_rate": 2.6388249699570667e-05, "loss": 0.008, "step": 26950 }, { "grad_norm": 0.13043753802776337, "learning_rate": 2.6351820560821672e-05, "loss": 0.0064, "step": 26960 }, { "grad_norm": 0.1854000985622406, "learning_rate": 2.631540758538034e-05, "loss": 0.0064, "step": 26970 }, { "grad_norm": 0.18018141388893127, "learning_rate": 2.6279010798134597e-05, "loss": 0.0059, "step": 26980 }, { "grad_norm": 0.1506587415933609, "learning_rate": 2.6242630223961305e-05, "loss": 0.0072, "step": 26990 }, { "grad_norm": 0.1603609174489975, "learning_rate": 2.6206265887726244e-05, "loss": 0.0055, "step": 27000 }, { "grad_norm": 0.14449851214885712, "learning_rate": 2.6169917814284066e-05, "loss": 0.0073, "step": 27010 }, { "grad_norm": 0.11731120944023132, "learning_rate": 2.6133586028478364e-05, "loss": 0.0059, "step": 27020 }, { "grad_norm": 0.12777478992938995, "learning_rate": 2.609727055514155e-05, "loss": 0.006, "step": 27030 }, { "grad_norm": 0.15402749180793762, "learning_rate": 2.606097141909494e-05, "loss": 0.0056, "step": 27040 }, { "grad_norm": 0.12432769685983658, "learning_rate": 2.6024688645148644e-05, "loss": 0.0055, "step": 27050 }, { "grad_norm": 0.21014338731765747, "learning_rate": 2.5988422258101564e-05, "loss": 0.0071, "step": 27060 }, { "grad_norm": 0.18832804262638092, "learning_rate": 2.5952172282741453e-05, "loss": 0.0073, "step": 27070 }, { "grad_norm": 0.13483698666095734, "learning_rate": 2.5915938743844853e-05, "loss": 0.0071, "step": 27080 }, { "grad_norm": 0.12548397481441498, "learning_rate": 2.5879721666177003e-05, "loss": 0.0051, "step": 27090 }, { "grad_norm": 0.16530101001262665, "learning_rate": 2.5843521074491972e-05, "loss": 0.0055, "step": 27100 }, { "grad_norm": 0.15931524336338043, "learning_rate": 2.5807336993532487e-05, "loss": 0.0053, "step": 27110 }, { "grad_norm": 0.08651816844940186, "learning_rate": 2.577116944803004e-05, "loss": 0.0065, "step": 27120 }, { "grad_norm": 0.18387050926685333, "learning_rate": 2.5735018462704818e-05, "loss": 0.0059, "step": 27130 }, { "grad_norm": 0.09722812473773956, "learning_rate": 2.5698884062265665e-05, "loss": 0.005, "step": 27140 }, { "grad_norm": 0.14762887358665466, "learning_rate": 2.5662766271410134e-05, "loss": 0.0072, "step": 27150 }, { "grad_norm": 0.17557184398174286, "learning_rate": 2.5626665114824343e-05, "loss": 0.0074, "step": 27160 }, { "grad_norm": 0.12855081260204315, "learning_rate": 2.5590580617183148e-05, "loss": 0.007, "step": 27170 }, { "grad_norm": 0.1767789125442505, "learning_rate": 2.5554512803149912e-05, "loss": 0.0061, "step": 27180 }, { "grad_norm": 0.17713700234889984, "learning_rate": 2.5518461697376662e-05, "loss": 0.0078, "step": 27190 }, { "grad_norm": 0.1831461489200592, "learning_rate": 2.548242732450402e-05, "loss": 0.0059, "step": 27200 }, { "grad_norm": 0.13670490682125092, "learning_rate": 2.5446409709161095e-05, "loss": 0.006, "step": 27210 }, { "grad_norm": 0.20351850986480713, "learning_rate": 2.541040887596561e-05, "loss": 0.0057, "step": 27220 }, { "grad_norm": 0.16466350853443146, "learning_rate": 2.537442484952378e-05, "loss": 0.0055, "step": 27230 }, { "grad_norm": 0.10987237840890884, "learning_rate": 2.533845765443037e-05, "loss": 0.0055, "step": 27240 }, { "grad_norm": 0.14886100590229034, "learning_rate": 2.530250731526863e-05, "loss": 0.0077, "step": 27250 }, { "grad_norm": 0.16677001118659973, "learning_rate": 2.5266573856610253e-05, "loss": 0.0074, "step": 27260 }, { "grad_norm": 0.166002094745636, "learning_rate": 2.5230657303015403e-05, "loss": 0.0072, "step": 27270 }, { "grad_norm": 0.12833565473556519, "learning_rate": 2.5194757679032728e-05, "loss": 0.0065, "step": 27280 }, { "grad_norm": 0.1585167646408081, "learning_rate": 2.5158875009199278e-05, "loss": 0.0076, "step": 27290 }, { "grad_norm": 0.17915602028369904, "learning_rate": 2.5123009318040537e-05, "loss": 0.0088, "step": 27300 }, { "grad_norm": 0.1303253024816513, "learning_rate": 2.508716063007034e-05, "loss": 0.0073, "step": 27310 }, { "grad_norm": 0.19314751029014587, "learning_rate": 2.5051328969790934e-05, "loss": 0.0063, "step": 27320 }, { "grad_norm": 0.17855209112167358, "learning_rate": 2.501551436169292e-05, "loss": 0.0057, "step": 27330 }, { "grad_norm": 0.08469472825527191, "learning_rate": 2.4979716830255255e-05, "loss": 0.0052, "step": 27340 }, { "grad_norm": 0.21582506597042084, "learning_rate": 2.4943936399945233e-05, "loss": 0.0054, "step": 27350 }, { "grad_norm": 0.15786750614643097, "learning_rate": 2.4908173095218412e-05, "loss": 0.0049, "step": 27360 }, { "grad_norm": 0.18043744564056396, "learning_rate": 2.4872426940518663e-05, "loss": 0.0066, "step": 27370 }, { "grad_norm": 0.14266155660152435, "learning_rate": 2.4836697960278156e-05, "loss": 0.0062, "step": 27380 }, { "grad_norm": 0.10236154496669769, "learning_rate": 2.480098617891732e-05, "loss": 0.0054, "step": 27390 }, { "grad_norm": 0.13992059230804443, "learning_rate": 2.4765291620844837e-05, "loss": 0.0059, "step": 27400 }, { "grad_norm": 0.15763965249061584, "learning_rate": 2.472961431045756e-05, "loss": 0.006, "step": 27410 }, { "grad_norm": 0.13185451924800873, "learning_rate": 2.4693954272140622e-05, "loss": 0.0059, "step": 27420 }, { "grad_norm": 0.16249503195285797, "learning_rate": 2.4658311530267315e-05, "loss": 0.0071, "step": 27430 }, { "grad_norm": 0.19610220193862915, "learning_rate": 2.4622686109199124e-05, "loss": 0.0059, "step": 27440 }, { "grad_norm": 0.20248942077159882, "learning_rate": 2.4587078033285695e-05, "loss": 0.0075, "step": 27450 }, { "grad_norm": 0.1634969264268875, "learning_rate": 2.45514873268648e-05, "loss": 0.0052, "step": 27460 }, { "grad_norm": 0.14124803245067596, "learning_rate": 2.4515914014262336e-05, "loss": 0.0068, "step": 27470 }, { "grad_norm": 0.11191587150096893, "learning_rate": 2.4480358119792345e-05, "loss": 0.0068, "step": 27480 }, { "grad_norm": 0.08961435407400131, "learning_rate": 2.4444819667756942e-05, "loss": 0.0049, "step": 27490 }, { "grad_norm": 0.14258575439453125, "learning_rate": 2.4409298682446346e-05, "loss": 0.0062, "step": 27500 }, { "grad_norm": 0.11569180339574814, "learning_rate": 2.437379518813877e-05, "loss": 0.0058, "step": 27510 }, { "grad_norm": 0.1535245180130005, "learning_rate": 2.4338309209100547e-05, "loss": 0.0054, "step": 27520 }, { "grad_norm": 0.14340552687644958, "learning_rate": 2.4302840769586004e-05, "loss": 0.0045, "step": 27530 }, { "grad_norm": 0.18681864440441132, "learning_rate": 2.42673898938375e-05, "loss": 0.006, "step": 27540 }, { "grad_norm": 0.12625056505203247, "learning_rate": 2.4231956606085343e-05, "loss": 0.0063, "step": 27550 }, { "grad_norm": 0.12203049659729004, "learning_rate": 2.419654093054789e-05, "loss": 0.0052, "step": 27560 }, { "grad_norm": 0.16208800673484802, "learning_rate": 2.4161142891431375e-05, "loss": 0.0068, "step": 27570 }, { "grad_norm": 0.16500279307365417, "learning_rate": 2.412576251293005e-05, "loss": 0.0059, "step": 27580 }, { "grad_norm": 0.15474164485931396, "learning_rate": 2.4090399819226068e-05, "loss": 0.0075, "step": 27590 }, { "grad_norm": 0.1576789766550064, "learning_rate": 2.4055054834489514e-05, "loss": 0.0063, "step": 27600 }, { "grad_norm": 0.14471697807312012, "learning_rate": 2.401972758287832e-05, "loss": 0.0066, "step": 27610 }, { "grad_norm": 0.16565774381160736, "learning_rate": 2.398441808853834e-05, "loss": 0.0092, "step": 27620 }, { "grad_norm": 0.19621792435646057, "learning_rate": 2.3949126375603288e-05, "loss": 0.0073, "step": 27630 }, { "grad_norm": 0.13223408162593842, "learning_rate": 2.3913852468194724e-05, "loss": 0.0058, "step": 27640 }, { "grad_norm": 0.14265747368335724, "learning_rate": 2.387859639042201e-05, "loss": 0.0058, "step": 27650 }, { "grad_norm": 0.16990526020526886, "learning_rate": 2.3843358166382368e-05, "loss": 0.0053, "step": 27660 }, { "grad_norm": 0.11270267516374588, "learning_rate": 2.3808137820160757e-05, "loss": 0.005, "step": 27670 }, { "grad_norm": 0.138695627450943, "learning_rate": 2.3772935375829975e-05, "loss": 0.0043, "step": 27680 }, { "grad_norm": 0.17257677018642426, "learning_rate": 2.3737750857450553e-05, "loss": 0.0067, "step": 27690 }, { "grad_norm": 0.13703851401805878, "learning_rate": 2.3702584289070805e-05, "loss": 0.0064, "step": 27700 }, { "grad_norm": 0.12907195091247559, "learning_rate": 2.36674356947267e-05, "loss": 0.0056, "step": 27710 }, { "grad_norm": 0.15303047001361847, "learning_rate": 2.3632305098442004e-05, "loss": 0.005, "step": 27720 }, { "grad_norm": 0.13752534985542297, "learning_rate": 2.3597192524228156e-05, "loss": 0.0052, "step": 27730 }, { "grad_norm": 0.0982188731431961, "learning_rate": 2.356209799608424e-05, "loss": 0.0063, "step": 27740 }, { "grad_norm": 0.13047181069850922, "learning_rate": 2.352702153799704e-05, "loss": 0.007, "step": 27750 }, { "grad_norm": 0.15817797183990479, "learning_rate": 2.3491963173941018e-05, "loss": 0.006, "step": 27760 }, { "grad_norm": 0.1966986060142517, "learning_rate": 2.3456922927878196e-05, "loss": 0.0074, "step": 27770 }, { "grad_norm": 0.20075765252113342, "learning_rate": 2.3421900823758257e-05, "loss": 0.0049, "step": 27780 }, { "grad_norm": 0.18719714879989624, "learning_rate": 2.3386896885518496e-05, "loss": 0.0064, "step": 27790 }, { "grad_norm": 0.1570589393377304, "learning_rate": 2.335191113708378e-05, "loss": 0.006, "step": 27800 }, { "grad_norm": 0.14171545207500458, "learning_rate": 2.331694360236651e-05, "loss": 0.0055, "step": 27810 }, { "grad_norm": 0.18050238490104675, "learning_rate": 2.3281994305266702e-05, "loss": 0.0067, "step": 27820 }, { "grad_norm": 0.17654353380203247, "learning_rate": 2.3247063269671826e-05, "loss": 0.0053, "step": 27830 }, { "grad_norm": 0.17491476237773895, "learning_rate": 2.321215051945695e-05, "loss": 0.0051, "step": 27840 }, { "grad_norm": 0.14039631187915802, "learning_rate": 2.3177256078484588e-05, "loss": 0.0056, "step": 27850 }, { "grad_norm": 0.10800901800394058, "learning_rate": 2.3142379970604798e-05, "loss": 0.0057, "step": 27860 }, { "grad_norm": 0.17325127124786377, "learning_rate": 2.3107522219655025e-05, "loss": 0.0068, "step": 27870 }, { "grad_norm": 0.16513580083847046, "learning_rate": 2.3072682849460236e-05, "loss": 0.0061, "step": 27880 }, { "grad_norm": 0.13895341753959656, "learning_rate": 2.303786188383281e-05, "loss": 0.0076, "step": 27890 }, { "grad_norm": 0.1421516239643097, "learning_rate": 2.300305934657257e-05, "loss": 0.0051, "step": 27900 }, { "grad_norm": 0.14200274646282196, "learning_rate": 2.2968275261466677e-05, "loss": 0.0079, "step": 27910 }, { "grad_norm": 0.2242487072944641, "learning_rate": 2.293350965228977e-05, "loss": 0.0051, "step": 27920 }, { "grad_norm": 0.19617153704166412, "learning_rate": 2.2898762542803776e-05, "loss": 0.0073, "step": 27930 }, { "grad_norm": 0.14432376623153687, "learning_rate": 2.286403395675803e-05, "loss": 0.0055, "step": 27940 }, { "grad_norm": 0.1400161236524582, "learning_rate": 2.28293239178892e-05, "loss": 0.0073, "step": 27950 }, { "grad_norm": 0.16043967008590698, "learning_rate": 2.2794632449921287e-05, "loss": 0.0055, "step": 27960 }, { "grad_norm": 0.13402949273586273, "learning_rate": 2.275995957656555e-05, "loss": 0.0061, "step": 27970 }, { "grad_norm": 0.1808519810438156, "learning_rate": 2.272530532152058e-05, "loss": 0.0057, "step": 27980 }, { "grad_norm": 0.1761670857667923, "learning_rate": 2.2690669708472233e-05, "loss": 0.0068, "step": 27990 }, { "grad_norm": 0.16620555520057678, "learning_rate": 2.2656052761093655e-05, "loss": 0.0056, "step": 28000 }, { "grad_norm": 0.12877553701400757, "learning_rate": 2.262145450304517e-05, "loss": 0.0061, "step": 28010 }, { "grad_norm": 0.15194645524024963, "learning_rate": 2.2586874957974352e-05, "loss": 0.0057, "step": 28020 }, { "grad_norm": 0.17739041149616241, "learning_rate": 2.2552314149516012e-05, "loss": 0.0059, "step": 28030 }, { "grad_norm": 0.1283172219991684, "learning_rate": 2.2517772101292133e-05, "loss": 0.0059, "step": 28040 }, { "grad_norm": 0.184153750538826, "learning_rate": 2.248324883691188e-05, "loss": 0.0074, "step": 28050 }, { "grad_norm": 0.15943564474582672, "learning_rate": 2.24487443799716e-05, "loss": 0.0053, "step": 28060 }, { "grad_norm": 0.15143705904483795, "learning_rate": 2.241425875405472e-05, "loss": 0.005, "step": 28070 }, { "grad_norm": 0.1503656655550003, "learning_rate": 2.2379791982731868e-05, "loss": 0.0054, "step": 28080 }, { "grad_norm": 0.1807641088962555, "learning_rate": 2.2345344089560756e-05, "loss": 0.0046, "step": 28090 }, { "grad_norm": 0.16688752174377441, "learning_rate": 2.2310915098086206e-05, "loss": 0.0082, "step": 28100 }, { "grad_norm": 0.1903723031282425, "learning_rate": 2.227650503184009e-05, "loss": 0.0049, "step": 28110 }, { "grad_norm": 0.13306398689746857, "learning_rate": 2.2242113914341357e-05, "loss": 0.0051, "step": 28120 }, { "grad_norm": 0.1277138739824295, "learning_rate": 2.220774176909602e-05, "loss": 0.0053, "step": 28130 }, { "grad_norm": 0.11837726831436157, "learning_rate": 2.2173388619597114e-05, "loss": 0.0058, "step": 28140 }, { "grad_norm": 0.14504855871200562, "learning_rate": 2.21390544893247e-05, "loss": 0.0042, "step": 28150 }, { "grad_norm": 0.11903876811265945, "learning_rate": 2.210473940174585e-05, "loss": 0.0061, "step": 28160 }, { "grad_norm": 0.16069668531417847, "learning_rate": 2.207044338031456e-05, "loss": 0.0069, "step": 28170 }, { "grad_norm": 0.1219392791390419, "learning_rate": 2.203616644847186e-05, "loss": 0.0048, "step": 28180 }, { "grad_norm": 0.11759835481643677, "learning_rate": 2.200190862964571e-05, "loss": 0.0064, "step": 28190 }, { "grad_norm": 0.1491711139678955, "learning_rate": 2.1967669947251024e-05, "loss": 0.0055, "step": 28200 }, { "grad_norm": 0.11299634724855423, "learning_rate": 2.1933450424689583e-05, "loss": 0.0057, "step": 28210 }, { "grad_norm": 0.18776695430278778, "learning_rate": 2.1899250085350142e-05, "loss": 0.0057, "step": 28220 }, { "grad_norm": 0.11327066272497177, "learning_rate": 2.1865068952608277e-05, "loss": 0.0048, "step": 28230 }, { "grad_norm": 0.11453918367624283, "learning_rate": 2.1830907049826487e-05, "loss": 0.0059, "step": 28240 }, { "grad_norm": 0.1596139371395111, "learning_rate": 2.179676440035411e-05, "loss": 0.0053, "step": 28250 }, { "grad_norm": 0.10666511952877045, "learning_rate": 2.1762641027527337e-05, "loss": 0.0059, "step": 28260 }, { "grad_norm": 0.12204969674348831, "learning_rate": 2.1728536954669143e-05, "loss": 0.0048, "step": 28270 }, { "grad_norm": 0.11297915130853653, "learning_rate": 2.169445220508936e-05, "loss": 0.0052, "step": 28280 }, { "grad_norm": 0.11534420400857925, "learning_rate": 2.166038680208461e-05, "loss": 0.0049, "step": 28290 }, { "grad_norm": 0.13172730803489685, "learning_rate": 2.162634076893823e-05, "loss": 0.0055, "step": 28300 }, { "grad_norm": 0.09835556894540787, "learning_rate": 2.1592314128920388e-05, "loss": 0.0065, "step": 28310 }, { "grad_norm": 0.14210839569568634, "learning_rate": 2.155830690528799e-05, "loss": 0.0075, "step": 28320 }, { "grad_norm": 0.10537828505039215, "learning_rate": 2.1524319121284613e-05, "loss": 0.0049, "step": 28330 }, { "grad_norm": 0.14112655818462372, "learning_rate": 2.1490350800140607e-05, "loss": 0.005, "step": 28340 }, { "grad_norm": 0.17489385604858398, "learning_rate": 2.1456401965073002e-05, "loss": 0.0051, "step": 28350 }, { "grad_norm": 0.12805697321891785, "learning_rate": 2.1422472639285524e-05, "loss": 0.0046, "step": 28360 }, { "grad_norm": 0.1080198660492897, "learning_rate": 2.13885628459685e-05, "loss": 0.0051, "step": 28370 }, { "grad_norm": 0.13170520961284637, "learning_rate": 2.135467260829901e-05, "loss": 0.0054, "step": 28380 }, { "grad_norm": 0.1560436189174652, "learning_rate": 2.1320801949440654e-05, "loss": 0.0054, "step": 28390 }, { "grad_norm": 0.14911405742168427, "learning_rate": 2.1286950892543744e-05, "loss": 0.0049, "step": 28400 }, { "grad_norm": 0.1624315232038498, "learning_rate": 2.125311946074515e-05, "loss": 0.0084, "step": 28410 }, { "grad_norm": 0.13480088114738464, "learning_rate": 2.1219307677168355e-05, "loss": 0.0052, "step": 28420 }, { "grad_norm": 0.14159086346626282, "learning_rate": 2.118551556492336e-05, "loss": 0.0065, "step": 28430 }, { "grad_norm": 0.09919893741607666, "learning_rate": 2.1151743147106774e-05, "loss": 0.0057, "step": 28440 }, { "grad_norm": 0.14948225021362305, "learning_rate": 2.111799044680172e-05, "loss": 0.0056, "step": 28450 }, { "grad_norm": 0.23907390236854553, "learning_rate": 2.1084257487077873e-05, "loss": 0.0053, "step": 28460 }, { "grad_norm": 0.17495496571063995, "learning_rate": 2.1050544290991357e-05, "loss": 0.0072, "step": 28470 }, { "grad_norm": 0.2636236250400543, "learning_rate": 2.101685088158486e-05, "loss": 0.0061, "step": 28480 }, { "grad_norm": 0.11058512330055237, "learning_rate": 2.0983177281887472e-05, "loss": 0.005, "step": 28490 }, { "grad_norm": 0.09874161332845688, "learning_rate": 2.0949523514914798e-05, "loss": 0.0055, "step": 28500 }, { "grad_norm": 0.15058277547359467, "learning_rate": 2.0915889603668876e-05, "loss": 0.004, "step": 28510 }, { "grad_norm": 0.10945755243301392, "learning_rate": 2.0882275571138175e-05, "loss": 0.0048, "step": 28520 }, { "grad_norm": 0.10346242785453796, "learning_rate": 2.0848681440297545e-05, "loss": 0.0062, "step": 28530 }, { "grad_norm": 0.16538405418395996, "learning_rate": 2.081510723410827e-05, "loss": 0.0054, "step": 28540 }, { "grad_norm": 0.16407771408557892, "learning_rate": 2.0781552975518003e-05, "loss": 0.0059, "step": 28550 }, { "grad_norm": 0.17005862295627594, "learning_rate": 2.074801868746078e-05, "loss": 0.0061, "step": 28560 }, { "grad_norm": 0.1347462683916092, "learning_rate": 2.0714504392856955e-05, "loss": 0.006, "step": 28570 }, { "grad_norm": 0.12230703979730606, "learning_rate": 2.0681010114613215e-05, "loss": 0.0052, "step": 28580 }, { "grad_norm": 0.17996209859848022, "learning_rate": 2.0647535875622597e-05, "loss": 0.0057, "step": 28590 }, { "grad_norm": 0.11497489362955093, "learning_rate": 2.0614081698764432e-05, "loss": 0.0049, "step": 28600 }, { "grad_norm": 0.1329309493303299, "learning_rate": 2.0580647606904334e-05, "loss": 0.005, "step": 28610 }, { "grad_norm": 0.1690351963043213, "learning_rate": 2.0547233622894208e-05, "loss": 0.0059, "step": 28620 }, { "grad_norm": 0.1357627958059311, "learning_rate": 2.0513839769572157e-05, "loss": 0.0077, "step": 28630 }, { "grad_norm": 0.13955479860305786, "learning_rate": 2.0480466069762584e-05, "loss": 0.005, "step": 28640 }, { "grad_norm": 0.1831568479537964, "learning_rate": 2.0447112546276104e-05, "loss": 0.0061, "step": 28650 }, { "grad_norm": 0.11694049835205078, "learning_rate": 2.0413779221909547e-05, "loss": 0.0045, "step": 28660 }, { "grad_norm": 0.10144365578889847, "learning_rate": 2.0380466119445912e-05, "loss": 0.0048, "step": 28670 }, { "grad_norm": 0.12744548916816711, "learning_rate": 2.0347173261654373e-05, "loss": 0.0058, "step": 28680 }, { "grad_norm": 0.13079945743083954, "learning_rate": 2.03139006712903e-05, "loss": 0.0053, "step": 28690 }, { "grad_norm": 0.12179137766361237, "learning_rate": 2.028064837109519e-05, "loss": 0.0061, "step": 28700 }, { "grad_norm": 0.11074129492044449, "learning_rate": 2.0247416383796685e-05, "loss": 0.0059, "step": 28710 }, { "grad_norm": 0.15897172689437866, "learning_rate": 2.0214204732108548e-05, "loss": 0.0044, "step": 28720 }, { "grad_norm": 0.1414836198091507, "learning_rate": 2.0181013438730596e-05, "loss": 0.0054, "step": 28730 }, { "grad_norm": 0.10092137008905411, "learning_rate": 2.0147842526348783e-05, "loss": 0.0047, "step": 28740 }, { "grad_norm": 0.12561799585819244, "learning_rate": 2.011469201763511e-05, "loss": 0.006, "step": 28750 }, { "grad_norm": 0.13839110732078552, "learning_rate": 2.0081561935247665e-05, "loss": 0.0052, "step": 28760 }, { "grad_norm": 0.14304082095623016, "learning_rate": 2.0048452301830523e-05, "loss": 0.0062, "step": 28770 }, { "grad_norm": 0.12406601756811142, "learning_rate": 2.0015363140013788e-05, "loss": 0.006, "step": 28780 }, { "grad_norm": 0.1114921048283577, "learning_rate": 1.9982294472413606e-05, "loss": 0.0048, "step": 28790 }, { "grad_norm": 0.1468164175748825, "learning_rate": 1.9949246321632103e-05, "loss": 0.005, "step": 28800 }, { "grad_norm": 0.1645595282316208, "learning_rate": 1.9916218710257377e-05, "loss": 0.0053, "step": 28810 }, { "grad_norm": 0.16082939505577087, "learning_rate": 1.988321166086351e-05, "loss": 0.0078, "step": 28820 }, { "grad_norm": 0.15519067645072937, "learning_rate": 1.9850225196010468e-05, "loss": 0.0045, "step": 28830 }, { "grad_norm": 0.13307839632034302, "learning_rate": 1.981725933824421e-05, "loss": 0.0057, "step": 28840 }, { "grad_norm": 0.11755575984716415, "learning_rate": 1.978431411009661e-05, "loss": 0.0056, "step": 28850 }, { "grad_norm": 0.15059417486190796, "learning_rate": 1.9751389534085375e-05, "loss": 0.0049, "step": 28860 }, { "grad_norm": 0.12291593104600906, "learning_rate": 1.9718485632714184e-05, "loss": 0.0056, "step": 28870 }, { "grad_norm": 0.10015998780727386, "learning_rate": 1.968560242847251e-05, "loss": 0.0054, "step": 28880 }, { "grad_norm": 0.10282839089632034, "learning_rate": 1.965273994383573e-05, "loss": 0.0048, "step": 28890 }, { "grad_norm": 0.11617647856473923, "learning_rate": 1.961989820126504e-05, "loss": 0.0057, "step": 28900 }, { "grad_norm": 0.12462837994098663, "learning_rate": 1.958707722320746e-05, "loss": 0.0048, "step": 28910 }, { "grad_norm": 0.1195271834731102, "learning_rate": 1.955427703209584e-05, "loss": 0.004, "step": 28920 }, { "grad_norm": 0.12338364869356155, "learning_rate": 1.9521497650348764e-05, "loss": 0.0065, "step": 28930 }, { "grad_norm": 0.12356576323509216, "learning_rate": 1.948873910037067e-05, "loss": 0.0043, "step": 28940 }, { "grad_norm": 0.11978600174188614, "learning_rate": 1.9456001404551678e-05, "loss": 0.0054, "step": 28950 }, { "grad_norm": 0.16821111738681793, "learning_rate": 1.942328458526771e-05, "loss": 0.0069, "step": 28960 }, { "grad_norm": 0.13464099168777466, "learning_rate": 1.9390588664880427e-05, "loss": 0.0048, "step": 28970 }, { "grad_norm": 0.19904997944831848, "learning_rate": 1.9357913665737145e-05, "loss": 0.0054, "step": 28980 }, { "grad_norm": 0.1410253643989563, "learning_rate": 1.932525961017093e-05, "loss": 0.0061, "step": 28990 }, { "grad_norm": 0.1816716492176056, "learning_rate": 1.9292626520500533e-05, "loss": 0.0057, "step": 29000 }, { "grad_norm": 0.11037237197160721, "learning_rate": 1.9260014419030354e-05, "loss": 0.0051, "step": 29010 }, { "grad_norm": 0.10695615410804749, "learning_rate": 1.9227423328050475e-05, "loss": 0.0052, "step": 29020 }, { "grad_norm": 0.07433179765939713, "learning_rate": 1.9194853269836582e-05, "loss": 0.0042, "step": 29030 }, { "grad_norm": 0.14459358155727386, "learning_rate": 1.916230426664999e-05, "loss": 0.008, "step": 29040 }, { "grad_norm": 0.11783802509307861, "learning_rate": 1.912977634073765e-05, "loss": 0.0033, "step": 29050 }, { "grad_norm": 0.11109954118728638, "learning_rate": 1.9097269514332083e-05, "loss": 0.0049, "step": 29060 }, { "grad_norm": 0.15951575338840485, "learning_rate": 1.9064783809651433e-05, "loss": 0.0053, "step": 29070 }, { "grad_norm": 0.13953597843647003, "learning_rate": 1.9032319248899333e-05, "loss": 0.0054, "step": 29080 }, { "grad_norm": 0.14122046530246735, "learning_rate": 1.8999875854265015e-05, "loss": 0.0056, "step": 29090 }, { "grad_norm": 0.15615850687026978, "learning_rate": 1.8967453647923232e-05, "loss": 0.0057, "step": 29100 }, { "grad_norm": 0.15763907134532928, "learning_rate": 1.893505265203427e-05, "loss": 0.0053, "step": 29110 }, { "grad_norm": 0.17356504499912262, "learning_rate": 1.8902672888743907e-05, "loss": 0.0052, "step": 29120 }, { "grad_norm": 0.13444961607456207, "learning_rate": 1.8870314380183396e-05, "loss": 0.0057, "step": 29130 }, { "grad_norm": 0.12430556118488312, "learning_rate": 1.8837977148469448e-05, "loss": 0.0069, "step": 29140 }, { "grad_norm": 0.17788958549499512, "learning_rate": 1.880566121570429e-05, "loss": 0.0063, "step": 29150 }, { "grad_norm": 0.15710829198360443, "learning_rate": 1.877336660397554e-05, "loss": 0.006, "step": 29160 }, { "grad_norm": 0.14574965834617615, "learning_rate": 1.874109333535628e-05, "loss": 0.0053, "step": 29170 }, { "grad_norm": 0.12246666848659515, "learning_rate": 1.870884143190496e-05, "loss": 0.0058, "step": 29180 }, { "grad_norm": 0.1602354496717453, "learning_rate": 1.867661091566546e-05, "loss": 0.0053, "step": 29190 }, { "grad_norm": 0.08242920786142349, "learning_rate": 1.864440180866704e-05, "loss": 0.0039, "step": 29200 }, { "grad_norm": 0.13110625743865967, "learning_rate": 1.8612214132924317e-05, "loss": 0.0048, "step": 29210 }, { "grad_norm": 0.17105704545974731, "learning_rate": 1.858004791043728e-05, "loss": 0.0063, "step": 29220 }, { "grad_norm": 0.20148561894893646, "learning_rate": 1.854790316319123e-05, "loss": 0.0055, "step": 29230 }, { "grad_norm": 0.1610250473022461, "learning_rate": 1.8515779913156766e-05, "loss": 0.0069, "step": 29240 }, { "grad_norm": 0.10963888466358185, "learning_rate": 1.848367818228986e-05, "loss": 0.0055, "step": 29250 }, { "grad_norm": 0.09146904945373535, "learning_rate": 1.8451597992531733e-05, "loss": 0.0044, "step": 29260 }, { "grad_norm": 0.12458931654691696, "learning_rate": 1.8419539365808914e-05, "loss": 0.005, "step": 29270 }, { "grad_norm": 0.08046894520521164, "learning_rate": 1.838750232403313e-05, "loss": 0.0048, "step": 29280 }, { "grad_norm": 0.11947188526391983, "learning_rate": 1.835548688910142e-05, "loss": 0.0057, "step": 29290 }, { "grad_norm": 0.15482695400714874, "learning_rate": 1.8323493082896037e-05, "loss": 0.0065, "step": 29300 }, { "grad_norm": 0.12435461580753326, "learning_rate": 1.8291520927284454e-05, "loss": 0.0049, "step": 29310 }, { "grad_norm": 0.10809113830327988, "learning_rate": 1.8259570444119305e-05, "loss": 0.0055, "step": 29320 }, { "grad_norm": 0.11149376630783081, "learning_rate": 1.8227641655238488e-05, "loss": 0.0057, "step": 29330 }, { "grad_norm": 0.10246103256940842, "learning_rate": 1.819573458246498e-05, "loss": 0.0043, "step": 29340 }, { "grad_norm": 0.16928696632385254, "learning_rate": 1.816384924760699e-05, "loss": 0.0059, "step": 29350 }, { "grad_norm": 0.13819344341754913, "learning_rate": 1.813198567245784e-05, "loss": 0.0059, "step": 29360 }, { "grad_norm": 0.19171810150146484, "learning_rate": 1.8100143878796006e-05, "loss": 0.0062, "step": 29370 }, { "grad_norm": 0.16220633685588837, "learning_rate": 1.8068323888385015e-05, "loss": 0.0051, "step": 29380 }, { "grad_norm": 0.11613140255212784, "learning_rate": 1.803652572297355e-05, "loss": 0.0049, "step": 29390 }, { "grad_norm": 0.09655337035655975, "learning_rate": 1.8004749404295353e-05, "loss": 0.0045, "step": 29400 }, { "grad_norm": 0.12029808759689331, "learning_rate": 1.797299495406926e-05, "loss": 0.004, "step": 29410 }, { "grad_norm": 0.15412282943725586, "learning_rate": 1.7941262393999103e-05, "loss": 0.0039, "step": 29420 }, { "grad_norm": 0.1728457808494568, "learning_rate": 1.7909551745773816e-05, "loss": 0.006, "step": 29430 }, { "grad_norm": 0.1464657038450241, "learning_rate": 1.7877863031067304e-05, "loss": 0.0053, "step": 29440 }, { "grad_norm": 0.1184283047914505, "learning_rate": 1.7846196271538516e-05, "loss": 0.0055, "step": 29450 }, { "grad_norm": 0.11077054589986801, "learning_rate": 1.7814551488831384e-05, "loss": 0.0062, "step": 29460 }, { "grad_norm": 0.14072611927986145, "learning_rate": 1.7782928704574835e-05, "loss": 0.004, "step": 29470 }, { "grad_norm": 0.135696142911911, "learning_rate": 1.775132794038271e-05, "loss": 0.0048, "step": 29480 }, { "grad_norm": 0.12929904460906982, "learning_rate": 1.7719749217853855e-05, "loss": 0.0041, "step": 29490 }, { "grad_norm": 0.15497413277626038, "learning_rate": 1.7688192558572038e-05, "loss": 0.0053, "step": 29500 }, { "grad_norm": 0.10146047919988632, "learning_rate": 1.7656657984105906e-05, "loss": 0.0064, "step": 29510 }, { "grad_norm": 0.18756887316703796, "learning_rate": 1.7625145516009068e-05, "loss": 0.006, "step": 29520 }, { "grad_norm": 0.11574511229991913, "learning_rate": 1.7593655175820005e-05, "loss": 0.0045, "step": 29530 }, { "grad_norm": 0.16710199415683746, "learning_rate": 1.7562186985062046e-05, "loss": 0.008, "step": 29540 }, { "grad_norm": 0.12185034900903702, "learning_rate": 1.7530740965243403e-05, "loss": 0.0053, "step": 29550 }, { "grad_norm": 0.16730250418186188, "learning_rate": 1.7499317137857153e-05, "loss": 0.0059, "step": 29560 }, { "grad_norm": 0.17357569932937622, "learning_rate": 1.7467915524381184e-05, "loss": 0.0047, "step": 29570 }, { "grad_norm": 0.1420164257287979, "learning_rate": 1.7436536146278182e-05, "loss": 0.0044, "step": 29580 }, { "grad_norm": 0.10614503920078278, "learning_rate": 1.7405179024995688e-05, "loss": 0.0049, "step": 29590 }, { "grad_norm": 0.16357704997062683, "learning_rate": 1.737384418196596e-05, "loss": 0.0066, "step": 29600 }, { "grad_norm": 0.12631584703922272, "learning_rate": 1.734253163860609e-05, "loss": 0.0097, "step": 29610 }, { "grad_norm": 0.1242184191942215, "learning_rate": 1.7311241416317896e-05, "loss": 0.0058, "step": 29620 }, { "grad_norm": 0.10281024128198624, "learning_rate": 1.7279973536487982e-05, "loss": 0.0045, "step": 29630 }, { "grad_norm": 0.14979879558086395, "learning_rate": 1.724872802048761e-05, "loss": 0.0063, "step": 29640 }, { "grad_norm": 0.10079079866409302, "learning_rate": 1.7217504889672803e-05, "loss": 0.005, "step": 29650 }, { "grad_norm": 0.18133948743343353, "learning_rate": 1.7186304165384287e-05, "loss": 0.005, "step": 29660 }, { "grad_norm": 0.12236025184392929, "learning_rate": 1.7155125868947475e-05, "loss": 0.0043, "step": 29670 }, { "grad_norm": 0.15151840448379517, "learning_rate": 1.7123970021672404e-05, "loss": 0.0042, "step": 29680 }, { "grad_norm": 0.1866491138935089, "learning_rate": 1.709283664485384e-05, "loss": 0.0052, "step": 29690 }, { "grad_norm": 0.14042451977729797, "learning_rate": 1.7061725759771113e-05, "loss": 0.0056, "step": 29700 }, { "grad_norm": 0.16516481339931488, "learning_rate": 1.7030637387688248e-05, "loss": 0.0056, "step": 29710 }, { "grad_norm": 0.15015563368797302, "learning_rate": 1.6999571549853836e-05, "loss": 0.0056, "step": 29720 }, { "grad_norm": 0.1007499024271965, "learning_rate": 1.696852826750112e-05, "loss": 0.0077, "step": 29730 }, { "grad_norm": 0.12916579842567444, "learning_rate": 1.6937507561847844e-05, "loss": 0.0039, "step": 29740 }, { "grad_norm": 0.12863707542419434, "learning_rate": 1.6906509454096385e-05, "loss": 0.0063, "step": 29750 }, { "grad_norm": 0.15283054113388062, "learning_rate": 1.687553396543367e-05, "loss": 0.0056, "step": 29760 }, { "grad_norm": 0.11159290373325348, "learning_rate": 1.6844581117031154e-05, "loss": 0.0065, "step": 29770 }, { "grad_norm": 0.13455389440059662, "learning_rate": 1.681365093004481e-05, "loss": 0.0051, "step": 29780 }, { "grad_norm": 0.1278572678565979, "learning_rate": 1.678274342561511e-05, "loss": 0.0047, "step": 29790 }, { "grad_norm": 0.17933954298496246, "learning_rate": 1.675185862486706e-05, "loss": 0.0074, "step": 29800 }, { "grad_norm": 0.15726567804813385, "learning_rate": 1.6720996548910127e-05, "loss": 0.0038, "step": 29810 }, { "grad_norm": 0.10723353922367096, "learning_rate": 1.6690157218838247e-05, "loss": 0.0047, "step": 29820 }, { "grad_norm": 0.10104188323020935, "learning_rate": 1.665934065572984e-05, "loss": 0.005, "step": 29830 }, { "grad_norm": 0.1553337275981903, "learning_rate": 1.6628546880647688e-05, "loss": 0.0048, "step": 29840 }, { "grad_norm": 0.2264203280210495, "learning_rate": 1.6597775914639076e-05, "loss": 0.0061, "step": 29850 }, { "grad_norm": 0.14072322845458984, "learning_rate": 1.6567027778735654e-05, "loss": 0.0067, "step": 29860 }, { "grad_norm": 0.1742134690284729, "learning_rate": 1.653630249395351e-05, "loss": 0.0057, "step": 29870 }, { "grad_norm": 0.1298227459192276, "learning_rate": 1.6505600081293072e-05, "loss": 0.0074, "step": 29880 }, { "grad_norm": 0.15953263640403748, "learning_rate": 1.647492056173912e-05, "loss": 0.0061, "step": 29890 }, { "grad_norm": 0.15339979529380798, "learning_rate": 1.644426395626085e-05, "loss": 0.0051, "step": 29900 }, { "grad_norm": 0.18514534831047058, "learning_rate": 1.641363028581175e-05, "loss": 0.0069, "step": 29910 }, { "grad_norm": 0.14013026654720306, "learning_rate": 1.638301957132965e-05, "loss": 0.0049, "step": 29920 }, { "grad_norm": 0.14526209235191345, "learning_rate": 1.6352431833736703e-05, "loss": 0.0076, "step": 29930 }, { "grad_norm": 0.1323464810848236, "learning_rate": 1.6321867093939298e-05, "loss": 0.0061, "step": 29940 }, { "grad_norm": 0.14310048520565033, "learning_rate": 1.629132537282817e-05, "loss": 0.0057, "step": 29950 }, { "grad_norm": 0.09937398135662079, "learning_rate": 1.62608066912783e-05, "loss": 0.0039, "step": 29960 }, { "grad_norm": 0.10457789897918701, "learning_rate": 1.623031107014893e-05, "loss": 0.0043, "step": 29970 }, { "grad_norm": 0.11270957440137863, "learning_rate": 1.619983853028351e-05, "loss": 0.0037, "step": 29980 }, { "grad_norm": 0.13177627325057983, "learning_rate": 1.6169389092509724e-05, "loss": 0.0041, "step": 29990 }, { "grad_norm": 0.1314675360918045, "learning_rate": 1.6138962777639494e-05, "loss": 0.0048, "step": 30000 }, { "grad_norm": 0.11183817684650421, "learning_rate": 1.610855960646891e-05, "loss": 0.0042, "step": 30010 }, { "grad_norm": 0.11857301741838455, "learning_rate": 1.607817959977826e-05, "loss": 0.0077, "step": 30020 }, { "grad_norm": 0.11255652457475662, "learning_rate": 1.6047822778332005e-05, "loss": 0.005, "step": 30030 }, { "grad_norm": 0.08063672482967377, "learning_rate": 1.6017489162878713e-05, "loss": 0.0039, "step": 30040 }, { "grad_norm": 0.11761800944805145, "learning_rate": 1.5987178774151147e-05, "loss": 0.0064, "step": 30050 }, { "grad_norm": 0.11791124194860458, "learning_rate": 1.5956891632866184e-05, "loss": 0.0068, "step": 30060 }, { "grad_norm": 0.10862471908330917, "learning_rate": 1.592662775972476e-05, "loss": 0.0043, "step": 30070 }, { "grad_norm": 0.11193399876356125, "learning_rate": 1.5896387175411986e-05, "loss": 0.0041, "step": 30080 }, { "grad_norm": 0.12002132087945938, "learning_rate": 1.586616990059699e-05, "loss": 0.0059, "step": 30090 }, { "grad_norm": 0.11392777413129807, "learning_rate": 1.5835975955933002e-05, "loss": 0.0051, "step": 30100 }, { "grad_norm": 0.11525142192840576, "learning_rate": 1.5805805362057295e-05, "loss": 0.0061, "step": 30110 }, { "grad_norm": 0.07070829719305038, "learning_rate": 1.57756581395912e-05, "loss": 0.0043, "step": 30120 }, { "grad_norm": 0.11280879378318787, "learning_rate": 1.574553430914006e-05, "loss": 0.0044, "step": 30130 }, { "grad_norm": 0.14918199181556702, "learning_rate": 1.5715433891293206e-05, "loss": 0.0076, "step": 30140 }, { "grad_norm": 0.1419016271829605, "learning_rate": 1.5685356906624016e-05, "loss": 0.0056, "step": 30150 }, { "grad_norm": 0.10316576063632965, "learning_rate": 1.56553033756898e-05, "loss": 0.0047, "step": 30160 }, { "grad_norm": 0.09054248034954071, "learning_rate": 1.5625273319031873e-05, "loss": 0.0062, "step": 30170 }, { "grad_norm": 0.12586559355258942, "learning_rate": 1.5595266757175518e-05, "loss": 0.0044, "step": 30180 }, { "grad_norm": 0.10956878960132599, "learning_rate": 1.55652837106299e-05, "loss": 0.0054, "step": 30190 }, { "grad_norm": 0.13742561638355255, "learning_rate": 1.5535324199888164e-05, "loss": 0.0064, "step": 30200 }, { "grad_norm": 0.1437482237815857, "learning_rate": 1.5505388245427355e-05, "loss": 0.0059, "step": 30210 }, { "grad_norm": 0.17963171005249023, "learning_rate": 1.5475475867708418e-05, "loss": 0.0074, "step": 30220 }, { "grad_norm": 0.14074403047561646, "learning_rate": 1.5445587087176195e-05, "loss": 0.0045, "step": 30230 }, { "grad_norm": 0.11704261600971222, "learning_rate": 1.5415721924259346e-05, "loss": 0.0051, "step": 30240 }, { "grad_norm": 0.1573241949081421, "learning_rate": 1.538588039937047e-05, "loss": 0.0084, "step": 30250 }, { "grad_norm": 0.13099047541618347, "learning_rate": 1.535606253290594e-05, "loss": 0.0043, "step": 30260 }, { "grad_norm": 0.12369658797979355, "learning_rate": 1.5326268345245993e-05, "loss": 0.0053, "step": 30270 }, { "grad_norm": 0.14774206280708313, "learning_rate": 1.5296497856754693e-05, "loss": 0.0071, "step": 30280 }, { "grad_norm": 0.12513311207294464, "learning_rate": 1.5266751087779858e-05, "loss": 0.0059, "step": 30290 }, { "grad_norm": 0.13826324045658112, "learning_rate": 1.5237028058653142e-05, "loss": 0.0044, "step": 30300 }, { "grad_norm": 0.0805259719491005, "learning_rate": 1.5207328789689946e-05, "loss": 0.0043, "step": 30310 }, { "grad_norm": 0.10304173827171326, "learning_rate": 1.5177653301189454e-05, "loss": 0.0057, "step": 30320 }, { "grad_norm": 0.09116100519895554, "learning_rate": 1.5148001613434587e-05, "loss": 0.005, "step": 30330 }, { "grad_norm": 0.13684764504432678, "learning_rate": 1.5118373746691983e-05, "loss": 0.006, "step": 30340 }, { "grad_norm": 0.16563184559345245, "learning_rate": 1.5088769721211992e-05, "loss": 0.0052, "step": 30350 }, { "grad_norm": 0.15074372291564941, "learning_rate": 1.505918955722871e-05, "loss": 0.0054, "step": 30360 }, { "grad_norm": 0.10626298189163208, "learning_rate": 1.5029633274959886e-05, "loss": 0.0057, "step": 30370 }, { "grad_norm": 0.11087101697921753, "learning_rate": 1.5000100894606989e-05, "loss": 0.0044, "step": 30380 }, { "grad_norm": 0.11397495865821838, "learning_rate": 1.4970592436355079e-05, "loss": 0.0046, "step": 30390 }, { "grad_norm": 0.09559259563684464, "learning_rate": 1.4941107920372938e-05, "loss": 0.0035, "step": 30400 }, { "grad_norm": 0.13090749084949493, "learning_rate": 1.4911647366812947e-05, "loss": 0.0047, "step": 30410 }, { "grad_norm": 0.11845515668392181, "learning_rate": 1.4882210795811114e-05, "loss": 0.0045, "step": 30420 }, { "grad_norm": 0.13502995669841766, "learning_rate": 1.485279822748708e-05, "loss": 0.0032, "step": 30430 }, { "grad_norm": 0.07157184928655624, "learning_rate": 1.4823409681944034e-05, "loss": 0.0036, "step": 30440 }, { "grad_norm": 0.1966133415699005, "learning_rate": 1.4794045179268763e-05, "loss": 0.0041, "step": 30450 }, { "grad_norm": 0.1463066190481186, "learning_rate": 1.4764704739531643e-05, "loss": 0.0041, "step": 30460 }, { "grad_norm": 0.14321564137935638, "learning_rate": 1.473538838278658e-05, "loss": 0.0045, "step": 30470 }, { "grad_norm": 0.16603443026542664, "learning_rate": 1.4706096129071045e-05, "loss": 0.0038, "step": 30480 }, { "grad_norm": 0.16231957077980042, "learning_rate": 1.4676827998405996e-05, "loss": 0.0049, "step": 30490 }, { "grad_norm": 0.1015484482049942, "learning_rate": 1.4647584010795928e-05, "loss": 0.0053, "step": 30500 }, { "grad_norm": 0.12623901665210724, "learning_rate": 1.4618364186228834e-05, "loss": 0.0044, "step": 30510 }, { "grad_norm": 0.12383095920085907, "learning_rate": 1.4589168544676207e-05, "loss": 0.0048, "step": 30520 }, { "grad_norm": 0.1125410944223404, "learning_rate": 1.455999710609296e-05, "loss": 0.006, "step": 30530 }, { "grad_norm": 0.13382519781589508, "learning_rate": 1.453084989041753e-05, "loss": 0.0044, "step": 30540 }, { "grad_norm": 0.11402243375778198, "learning_rate": 1.4501726917571728e-05, "loss": 0.0043, "step": 30550 }, { "grad_norm": 0.12273932993412018, "learning_rate": 1.4472628207460848e-05, "loss": 0.0059, "step": 30560 }, { "grad_norm": 0.11834068596363068, "learning_rate": 1.4443553779973584e-05, "loss": 0.0069, "step": 30570 }, { "grad_norm": 0.16365693509578705, "learning_rate": 1.4414503654982037e-05, "loss": 0.0055, "step": 30580 }, { "grad_norm": 0.08754004538059235, "learning_rate": 1.4385477852341706e-05, "loss": 0.0038, "step": 30590 }, { "grad_norm": 0.10683095455169678, "learning_rate": 1.4356476391891421e-05, "loss": 0.0045, "step": 30600 }, { "grad_norm": 0.12732020020484924, "learning_rate": 1.4327499293453423e-05, "loss": 0.0059, "step": 30610 }, { "grad_norm": 0.09014645963907242, "learning_rate": 1.4298546576833304e-05, "loss": 0.0034, "step": 30620 }, { "grad_norm": 0.12498389929533005, "learning_rate": 1.426961826181994e-05, "loss": 0.0056, "step": 30630 }, { "grad_norm": 0.13887637853622437, "learning_rate": 1.4240714368185592e-05, "loss": 0.0054, "step": 30640 }, { "grad_norm": 0.10583792626857758, "learning_rate": 1.4211834915685767e-05, "loss": 0.0042, "step": 30650 }, { "grad_norm": 0.10149334371089935, "learning_rate": 1.4182979924059313e-05, "loss": 0.0059, "step": 30660 }, { "grad_norm": 0.112715944647789, "learning_rate": 1.415414941302835e-05, "loss": 0.0042, "step": 30670 }, { "grad_norm": 0.11668100208044052, "learning_rate": 1.4125343402298257e-05, "loss": 0.0044, "step": 30680 }, { "grad_norm": 0.09042107313871384, "learning_rate": 1.4096561911557688e-05, "loss": 0.0054, "step": 30690 }, { "grad_norm": 0.11510219424962997, "learning_rate": 1.4067804960478481e-05, "loss": 0.0041, "step": 30700 }, { "grad_norm": 0.13704459369182587, "learning_rate": 1.4039072568715784e-05, "loss": 0.005, "step": 30710 }, { "grad_norm": 0.10257727652788162, "learning_rate": 1.401036475590788e-05, "loss": 0.0048, "step": 30720 }, { "grad_norm": 0.09420502930879593, "learning_rate": 1.3981681541676301e-05, "loss": 0.0052, "step": 30730 }, { "grad_norm": 0.12313235551118851, "learning_rate": 1.3953022945625772e-05, "loss": 0.0051, "step": 30740 }, { "grad_norm": 0.1003507748246193, "learning_rate": 1.3924388987344138e-05, "loss": 0.0034, "step": 30750 }, { "grad_norm": 0.1534586101770401, "learning_rate": 1.3895779686402465e-05, "loss": 0.0052, "step": 30760 }, { "grad_norm": 0.16025866568088531, "learning_rate": 1.3867195062354931e-05, "loss": 0.0051, "step": 30770 }, { "grad_norm": 0.11806933581829071, "learning_rate": 1.3838635134738858e-05, "loss": 0.0063, "step": 30780 }, { "grad_norm": 0.12296465039253235, "learning_rate": 1.3810099923074704e-05, "loss": 0.0034, "step": 30790 }, { "grad_norm": 0.13520033657550812, "learning_rate": 1.3781589446866006e-05, "loss": 0.0058, "step": 30800 }, { "grad_norm": 0.18109634518623352, "learning_rate": 1.3753103725599387e-05, "loss": 0.0062, "step": 30810 }, { "grad_norm": 0.1297190934419632, "learning_rate": 1.3724642778744579e-05, "loss": 0.0055, "step": 30820 }, { "grad_norm": 0.12471601366996765, "learning_rate": 1.3696206625754376e-05, "loss": 0.0053, "step": 30830 }, { "grad_norm": 0.12372085452079773, "learning_rate": 1.3667795286064639e-05, "loss": 0.0036, "step": 30840 }, { "grad_norm": 0.1367180198431015, "learning_rate": 1.3639408779094215e-05, "loss": 0.0043, "step": 30850 }, { "grad_norm": 0.15073713660240173, "learning_rate": 1.361104712424503e-05, "loss": 0.0052, "step": 30860 }, { "grad_norm": 0.2182995080947876, "learning_rate": 1.358271034090201e-05, "loss": 0.0069, "step": 30870 }, { "grad_norm": 0.13854360580444336, "learning_rate": 1.3554398448433076e-05, "loss": 0.0053, "step": 30880 }, { "grad_norm": 0.09876500815153122, "learning_rate": 1.3526111466189156e-05, "loss": 0.004, "step": 30890 }, { "grad_norm": 0.15225623548030853, "learning_rate": 1.3497849413504121e-05, "loss": 0.0048, "step": 30900 }, { "grad_norm": 0.12219996005296707, "learning_rate": 1.3469612309694802e-05, "loss": 0.0043, "step": 30910 }, { "grad_norm": 0.10831524431705475, "learning_rate": 1.3441400174061008e-05, "loss": 0.0048, "step": 30920 }, { "grad_norm": 0.08576048165559769, "learning_rate": 1.3413213025885469e-05, "loss": 0.0049, "step": 30930 }, { "grad_norm": 0.06949778646230698, "learning_rate": 1.3385050884433853e-05, "loss": 0.0036, "step": 30940 }, { "grad_norm": 0.08330493420362473, "learning_rate": 1.3356913768954676e-05, "loss": 0.0041, "step": 30950 }, { "grad_norm": 0.09385597705841064, "learning_rate": 1.3328801698679415e-05, "loss": 0.005, "step": 30960 }, { "grad_norm": 0.12756438553333282, "learning_rate": 1.3300714692822413e-05, "loss": 0.0043, "step": 30970 }, { "grad_norm": 0.1037381961941719, "learning_rate": 1.3272652770580857e-05, "loss": 0.0046, "step": 30980 }, { "grad_norm": 0.08728419989347458, "learning_rate": 1.3244615951134831e-05, "loss": 0.006, "step": 30990 }, { "grad_norm": 0.09438109397888184, "learning_rate": 1.3216604253647219e-05, "loss": 0.0044, "step": 31000 }, { "grad_norm": 0.09545506536960602, "learning_rate": 1.3188617697263738e-05, "loss": 0.0047, "step": 31010 }, { "grad_norm": 0.13259559869766235, "learning_rate": 1.3160656301112956e-05, "loss": 0.0051, "step": 31020 }, { "grad_norm": 0.1485103815793991, "learning_rate": 1.3132720084306211e-05, "loss": 0.0038, "step": 31030 }, { "grad_norm": 0.1383851319551468, "learning_rate": 1.3104809065937668e-05, "loss": 0.0062, "step": 31040 }, { "grad_norm": 0.1044149324297905, "learning_rate": 1.3076923265084212e-05, "loss": 0.0046, "step": 31050 }, { "grad_norm": 0.12428229302167892, "learning_rate": 1.3049062700805543e-05, "loss": 0.0041, "step": 31060 }, { "grad_norm": 0.10599098354578018, "learning_rate": 1.3021227392144092e-05, "loss": 0.0049, "step": 31070 }, { "grad_norm": 0.10304974019527435, "learning_rate": 1.2993417358125037e-05, "loss": 0.0054, "step": 31080 }, { "grad_norm": 0.11709524691104889, "learning_rate": 1.2965632617756257e-05, "loss": 0.0039, "step": 31090 }, { "grad_norm": 0.10663695633411407, "learning_rate": 1.2937873190028387e-05, "loss": 0.0048, "step": 31100 }, { "grad_norm": 0.11734573543071747, "learning_rate": 1.2910139093914697e-05, "loss": 0.0046, "step": 31110 }, { "grad_norm": 0.10849367082118988, "learning_rate": 1.2882430348371205e-05, "loss": 0.0052, "step": 31120 }, { "grad_norm": 0.1174854263663292, "learning_rate": 1.2854746972336567e-05, "loss": 0.0042, "step": 31130 }, { "grad_norm": 0.16444897651672363, "learning_rate": 1.2827088984732132e-05, "loss": 0.0047, "step": 31140 }, { "grad_norm": 0.14788788557052612, "learning_rate": 1.279945640446184e-05, "loss": 0.0041, "step": 31150 }, { "grad_norm": 0.12257318943738937, "learning_rate": 1.2771849250412315e-05, "loss": 0.0051, "step": 31160 }, { "grad_norm": 0.16703782975673676, "learning_rate": 1.2744267541452792e-05, "loss": 0.0045, "step": 31170 }, { "grad_norm": 0.1485775262117386, "learning_rate": 1.2716711296435112e-05, "loss": 0.0052, "step": 31180 }, { "grad_norm": 0.13100500404834747, "learning_rate": 1.268918053419369e-05, "loss": 0.0039, "step": 31190 }, { "grad_norm": 0.1236020103096962, "learning_rate": 1.2661675273545564e-05, "loss": 0.0038, "step": 31200 }, { "grad_norm": 0.10096847265958786, "learning_rate": 1.26341955332903e-05, "loss": 0.0046, "step": 31210 }, { "grad_norm": 0.18263116478919983, "learning_rate": 1.2606741332210048e-05, "loss": 0.0063, "step": 31220 }, { "grad_norm": 0.13134874403476715, "learning_rate": 1.2579312689069495e-05, "loss": 0.005, "step": 31230 }, { "grad_norm": 0.10586321353912354, "learning_rate": 1.2551909622615881e-05, "loss": 0.0061, "step": 31240 }, { "grad_norm": 0.08362504839897156, "learning_rate": 1.252453215157891e-05, "loss": 0.0047, "step": 31250 }, { "grad_norm": 0.11912242323160172, "learning_rate": 1.2497180294670841e-05, "loss": 0.0042, "step": 31260 }, { "grad_norm": 0.1305776834487915, "learning_rate": 1.2469854070586423e-05, "loss": 0.0051, "step": 31270 }, { "grad_norm": 0.1245480328798294, "learning_rate": 1.2442553498002845e-05, "loss": 0.0074, "step": 31280 }, { "grad_norm": 0.11092516779899597, "learning_rate": 1.241527859557981e-05, "loss": 0.006, "step": 31290 }, { "grad_norm": 0.16289684176445007, "learning_rate": 1.238802938195946e-05, "loss": 0.0066, "step": 31300 }, { "grad_norm": 0.08441206067800522, "learning_rate": 1.2360805875766357e-05, "loss": 0.0037, "step": 31310 }, { "grad_norm": 0.09832562506198883, "learning_rate": 1.2333608095607524e-05, "loss": 0.0044, "step": 31320 }, { "grad_norm": 0.18936683237552643, "learning_rate": 1.230643606007238e-05, "loss": 0.0048, "step": 31330 }, { "grad_norm": 0.12320369482040405, "learning_rate": 1.2279289787732778e-05, "loss": 0.0048, "step": 31340 }, { "grad_norm": 0.12652523815631866, "learning_rate": 1.2252169297142906e-05, "loss": 0.0054, "step": 31350 }, { "grad_norm": 0.17107941210269928, "learning_rate": 1.2225074606839392e-05, "loss": 0.005, "step": 31360 }, { "grad_norm": 0.09919305145740509, "learning_rate": 1.2198005735341173e-05, "loss": 0.0058, "step": 31370 }, { "grad_norm": 0.12532758712768555, "learning_rate": 1.2170962701149584e-05, "loss": 0.0049, "step": 31380 }, { "grad_norm": 0.1356000453233719, "learning_rate": 1.2143945522748285e-05, "loss": 0.0062, "step": 31390 }, { "grad_norm": 0.06513912975788116, "learning_rate": 1.2116954218603266e-05, "loss": 0.0046, "step": 31400 }, { "grad_norm": 0.08039166778326035, "learning_rate": 1.2089988807162817e-05, "loss": 0.0054, "step": 31410 }, { "grad_norm": 0.14249174296855927, "learning_rate": 1.2063049306857543e-05, "loss": 0.0061, "step": 31420 }, { "grad_norm": 0.12141287326812744, "learning_rate": 1.2036135736100346e-05, "loss": 0.0045, "step": 31430 }, { "grad_norm": 0.15044543147087097, "learning_rate": 1.200924811328641e-05, "loss": 0.0047, "step": 31440 }, { "grad_norm": 0.13487792015075684, "learning_rate": 1.1982386456793143e-05, "loss": 0.004, "step": 31450 }, { "grad_norm": 0.1309709995985031, "learning_rate": 1.1955550784980257e-05, "loss": 0.0049, "step": 31460 }, { "grad_norm": 0.09660349041223526, "learning_rate": 1.1928741116189663e-05, "loss": 0.0043, "step": 31470 }, { "grad_norm": 0.1267641931772232, "learning_rate": 1.1901957468745529e-05, "loss": 0.0042, "step": 31480 }, { "grad_norm": 0.12571197748184204, "learning_rate": 1.1875199860954217e-05, "loss": 0.005, "step": 31490 }, { "grad_norm": 0.1566440761089325, "learning_rate": 1.1848468311104316e-05, "loss": 0.0047, "step": 31500 }, { "grad_norm": 0.13948380947113037, "learning_rate": 1.1821762837466566e-05, "loss": 0.0042, "step": 31510 }, { "grad_norm": 0.13378961384296417, "learning_rate": 1.179508345829392e-05, "loss": 0.0038, "step": 31520 }, { "grad_norm": 0.1566176861524582, "learning_rate": 1.1768430191821478e-05, "loss": 0.0039, "step": 31530 }, { "grad_norm": 0.14356748759746552, "learning_rate": 1.174180305626651e-05, "loss": 0.0059, "step": 31540 }, { "grad_norm": 0.1215025931596756, "learning_rate": 1.1715202069828401e-05, "loss": 0.0073, "step": 31550 }, { "grad_norm": 0.13655956089496613, "learning_rate": 1.1688627250688661e-05, "loss": 0.0039, "step": 31560 }, { "grad_norm": 0.10359955579042435, "learning_rate": 1.1662078617010935e-05, "loss": 0.0036, "step": 31570 }, { "grad_norm": 0.0782545730471611, "learning_rate": 1.1635556186940972e-05, "loss": 0.0033, "step": 31580 }, { "grad_norm": 0.09289096295833588, "learning_rate": 1.1609059978606602e-05, "loss": 0.0038, "step": 31590 }, { "grad_norm": 0.13128440082073212, "learning_rate": 1.1582590010117745e-05, "loss": 0.0045, "step": 31600 }, { "grad_norm": 0.10486738383769989, "learning_rate": 1.1556146299566345e-05, "loss": 0.0042, "step": 31610 }, { "grad_norm": 0.16015148162841797, "learning_rate": 1.1529728865026452e-05, "loss": 0.0063, "step": 31620 }, { "grad_norm": 0.14412961900234222, "learning_rate": 1.150333772455412e-05, "loss": 0.0031, "step": 31630 }, { "grad_norm": 0.16061650216579437, "learning_rate": 1.1476972896187466e-05, "loss": 0.0057, "step": 31640 }, { "grad_norm": 0.12240578234195709, "learning_rate": 1.1450634397946592e-05, "loss": 0.0051, "step": 31650 }, { "grad_norm": 0.15798915922641754, "learning_rate": 1.142432224783359e-05, "loss": 0.0057, "step": 31660 }, { "grad_norm": 0.15237422287464142, "learning_rate": 1.1398036463832584e-05, "loss": 0.0058, "step": 31670 }, { "grad_norm": 0.1288473755121231, "learning_rate": 1.1371777063909661e-05, "loss": 0.005, "step": 31680 }, { "grad_norm": 0.09752330183982849, "learning_rate": 1.1345544066012874e-05, "loss": 0.0037, "step": 31690 }, { "grad_norm": 0.1308353841304779, "learning_rate": 1.131933748807224e-05, "loss": 0.0049, "step": 31700 }, { "grad_norm": 0.12956999242305756, "learning_rate": 1.1293157347999683e-05, "loss": 0.0042, "step": 31710 }, { "grad_norm": 0.11102873086929321, "learning_rate": 1.1267003663689097e-05, "loss": 0.0037, "step": 31720 }, { "grad_norm": 0.14426928758621216, "learning_rate": 1.1240876453016275e-05, "loss": 0.004, "step": 31730 }, { "grad_norm": 0.11255908012390137, "learning_rate": 1.1214775733838933e-05, "loss": 0.0043, "step": 31740 }, { "grad_norm": 0.11260043829679489, "learning_rate": 1.118870152399666e-05, "loss": 0.0053, "step": 31750 }, { "grad_norm": 0.12168056517839432, "learning_rate": 1.1162653841310905e-05, "loss": 0.0042, "step": 31760 }, { "grad_norm": 0.14218196272850037, "learning_rate": 1.1136632703585043e-05, "loss": 0.0047, "step": 31770 }, { "grad_norm": 0.08924791216850281, "learning_rate": 1.111063812860426e-05, "loss": 0.004, "step": 31780 }, { "grad_norm": 0.19842007756233215, "learning_rate": 1.1084670134135616e-05, "loss": 0.0045, "step": 31790 }, { "grad_norm": 0.16725029051303864, "learning_rate": 1.1058728737927998e-05, "loss": 0.0047, "step": 31800 }, { "grad_norm": 0.07571911066770554, "learning_rate": 1.1032813957712086e-05, "loss": 0.0043, "step": 31810 }, { "grad_norm": 0.11845355480909348, "learning_rate": 1.1006925811200385e-05, "loss": 0.0039, "step": 31820 }, { "grad_norm": 0.13075466454029083, "learning_rate": 1.0981064316087231e-05, "loss": 0.005, "step": 31830 }, { "grad_norm": 0.12100022286176682, "learning_rate": 1.0955229490048675e-05, "loss": 0.004, "step": 31840 }, { "grad_norm": 0.10384991765022278, "learning_rate": 1.0929421350742603e-05, "loss": 0.0045, "step": 31850 }, { "grad_norm": 0.12403598427772522, "learning_rate": 1.090363991580861e-05, "loss": 0.0042, "step": 31860 }, { "grad_norm": 0.12810532748699188, "learning_rate": 1.0877885202868083e-05, "loss": 0.0042, "step": 31870 }, { "grad_norm": 0.08036315441131592, "learning_rate": 1.0852157229524118e-05, "loss": 0.0039, "step": 31880 }, { "grad_norm": 0.08439286053180695, "learning_rate": 1.0826456013361542e-05, "loss": 0.0034, "step": 31890 }, { "grad_norm": 0.1084400862455368, "learning_rate": 1.080078157194691e-05, "loss": 0.004, "step": 31900 }, { "grad_norm": 0.228777214884758, "learning_rate": 1.077513392282844e-05, "loss": 0.0048, "step": 31910 }, { "grad_norm": 0.1442943811416626, "learning_rate": 1.0749513083536073e-05, "loss": 0.004, "step": 31920 }, { "grad_norm": 0.15206842124462128, "learning_rate": 1.072391907158139e-05, "loss": 0.0041, "step": 31930 }, { "grad_norm": 0.12638676166534424, "learning_rate": 1.0698351904457676e-05, "loss": 0.0061, "step": 31940 }, { "grad_norm": 0.1128656417131424, "learning_rate": 1.0672811599639849e-05, "loss": 0.004, "step": 31950 }, { "grad_norm": 0.12207602709531784, "learning_rate": 1.0647298174584453e-05, "loss": 0.0066, "step": 31960 }, { "grad_norm": 0.08543581515550613, "learning_rate": 1.0621811646729673e-05, "loss": 0.004, "step": 31970 }, { "grad_norm": 0.11338957399129868, "learning_rate": 1.0596352033495328e-05, "loss": 0.0055, "step": 31980 }, { "grad_norm": 0.1415969729423523, "learning_rate": 1.057091935228281e-05, "loss": 0.0039, "step": 31990 }, { "grad_norm": 0.11948467791080475, "learning_rate": 1.0545513620475134e-05, "loss": 0.0043, "step": 32000 }, { "grad_norm": 0.10635076463222504, "learning_rate": 1.0520134855436863e-05, "loss": 0.0061, "step": 32010 }, { "grad_norm": 0.202535942196846, "learning_rate": 1.049478307451413e-05, "loss": 0.0043, "step": 32020 }, { "grad_norm": 0.13175509870052338, "learning_rate": 1.0469458295034656e-05, "loss": 0.0036, "step": 32030 }, { "grad_norm": 0.06166895851492882, "learning_rate": 1.044416053430769e-05, "loss": 0.0033, "step": 32040 }, { "grad_norm": 0.09727636724710464, "learning_rate": 1.0418889809624021e-05, "loss": 0.0032, "step": 32050 }, { "grad_norm": 0.08014319837093353, "learning_rate": 1.0393646138255925e-05, "loss": 0.0041, "step": 32060 }, { "grad_norm": 0.11058583110570908, "learning_rate": 1.036842953745723e-05, "loss": 0.0029, "step": 32070 }, { "grad_norm": 0.13100285828113556, "learning_rate": 1.034324002446324e-05, "loss": 0.0049, "step": 32080 }, { "grad_norm": 0.08117925375699997, "learning_rate": 1.0318077616490751e-05, "loss": 0.0045, "step": 32090 }, { "grad_norm": 0.1405777484178543, "learning_rate": 1.0292942330738054e-05, "loss": 0.0036, "step": 32100 }, { "grad_norm": 0.0671190619468689, "learning_rate": 1.026783418438485e-05, "loss": 0.0039, "step": 32110 }, { "grad_norm": 0.13130898773670197, "learning_rate": 1.0242753194592319e-05, "loss": 0.0045, "step": 32120 }, { "grad_norm": 0.1048402264714241, "learning_rate": 1.0217699378503092e-05, "loss": 0.0036, "step": 32130 }, { "grad_norm": 0.05334865301847458, "learning_rate": 1.0192672753241218e-05, "loss": 0.0047, "step": 32140 }, { "grad_norm": 0.08306780457496643, "learning_rate": 1.0167673335912164e-05, "loss": 0.0059, "step": 32150 }, { "grad_norm": 0.08770028501749039, "learning_rate": 1.0142701143602779e-05, "loss": 0.0038, "step": 32160 }, { "grad_norm": 0.09474324434995651, "learning_rate": 1.0117756193381328e-05, "loss": 0.0035, "step": 32170 }, { "grad_norm": 0.11565227061510086, "learning_rate": 1.0092838502297453e-05, "loss": 0.004, "step": 32180 }, { "grad_norm": 0.08201964944601059, "learning_rate": 1.0067948087382162e-05, "loss": 0.0028, "step": 32190 }, { "grad_norm": 0.0825108140707016, "learning_rate": 1.0043084965647826e-05, "loss": 0.0037, "step": 32200 }, { "grad_norm": 0.09168665111064911, "learning_rate": 1.001824915408815e-05, "loss": 0.0071, "step": 32210 }, { "grad_norm": 0.17830125987529755, "learning_rate": 9.99344066967815e-06, "loss": 0.0042, "step": 32220 }, { "grad_norm": 0.11532022058963776, "learning_rate": 9.968659529374219e-06, "loss": 0.0045, "step": 32230 }, { "grad_norm": 0.1562311202287674, "learning_rate": 9.943905750114025e-06, "loss": 0.0043, "step": 32240 }, { "grad_norm": 0.10641294717788696, "learning_rate": 9.919179348816559e-06, "loss": 0.0036, "step": 32250 }, { "grad_norm": 0.0956353098154068, "learning_rate": 9.894480342382056e-06, "loss": 0.0056, "step": 32260 }, { "grad_norm": 0.1235186755657196, "learning_rate": 9.869808747692061e-06, "loss": 0.0052, "step": 32270 }, { "grad_norm": 0.19914510846138, "learning_rate": 9.84516458160939e-06, "loss": 0.005, "step": 32280 }, { "grad_norm": 0.15397070348262787, "learning_rate": 9.820547860978097e-06, "loss": 0.0049, "step": 32290 }, { "grad_norm": 0.14484070241451263, "learning_rate": 9.795958602623461e-06, "loss": 0.0043, "step": 32300 }, { "grad_norm": 0.0903182327747345, "learning_rate": 9.77139682335203e-06, "loss": 0.0039, "step": 32310 }, { "grad_norm": 0.09152639657258987, "learning_rate": 9.74686253995153e-06, "loss": 0.0028, "step": 32320 }, { "grad_norm": 0.10423610359430313, "learning_rate": 9.722355769190911e-06, "loss": 0.0043, "step": 32330 }, { "grad_norm": 0.12893380224704742, "learning_rate": 9.697876527820333e-06, "loss": 0.0039, "step": 32340 }, { "grad_norm": 0.1299532949924469, "learning_rate": 9.673424832571132e-06, "loss": 0.0042, "step": 32350 }, { "grad_norm": 0.0921315997838974, "learning_rate": 9.649000700155791e-06, "loss": 0.0044, "step": 32360 }, { "grad_norm": 0.10451903194189072, "learning_rate": 9.624604147267984e-06, "loss": 0.0044, "step": 32370 }, { "grad_norm": 0.08792388439178467, "learning_rate": 9.600235190582524e-06, "loss": 0.0033, "step": 32380 }, { "grad_norm": 0.14097215235233307, "learning_rate": 9.57589384675538e-06, "loss": 0.0061, "step": 32390 }, { "grad_norm": 0.1123630478978157, "learning_rate": 9.551580132423598e-06, "loss": 0.0045, "step": 32400 }, { "grad_norm": 0.15332607924938202, "learning_rate": 9.527294064205405e-06, "loss": 0.0069, "step": 32410 }, { "grad_norm": 0.12739498913288116, "learning_rate": 9.503035658700072e-06, "loss": 0.005, "step": 32420 }, { "grad_norm": 0.1756032407283783, "learning_rate": 9.478804932488e-06, "loss": 0.0039, "step": 32430 }, { "grad_norm": 0.09953779727220535, "learning_rate": 9.454601902130667e-06, "loss": 0.004, "step": 32440 }, { "grad_norm": 0.08513659983873367, "learning_rate": 9.430426584170626e-06, "loss": 0.0042, "step": 32450 }, { "grad_norm": 0.11048275232315063, "learning_rate": 9.40627899513145e-06, "loss": 0.0053, "step": 32460 }, { "grad_norm": 0.1020234078168869, "learning_rate": 9.38215915151781e-06, "loss": 0.0055, "step": 32470 }, { "grad_norm": 0.12611638009548187, "learning_rate": 9.358067069815396e-06, "loss": 0.0054, "step": 32480 }, { "grad_norm": 0.11057810485363007, "learning_rate": 9.334002766490901e-06, "loss": 0.0038, "step": 32490 }, { "grad_norm": 0.12109537422657013, "learning_rate": 9.309966257992054e-06, "loss": 0.0036, "step": 32500 }, { "grad_norm": 0.11366192996501923, "learning_rate": 9.2859575607476e-06, "loss": 0.004, "step": 32510 }, { "grad_norm": 0.16445808112621307, "learning_rate": 9.26197669116723e-06, "loss": 0.0045, "step": 32520 }, { "grad_norm": 0.1489102691411972, "learning_rate": 9.238023665641655e-06, "loss": 0.0039, "step": 32530 }, { "grad_norm": 0.20510098338127136, "learning_rate": 9.21409850054254e-06, "loss": 0.0077, "step": 32540 }, { "grad_norm": 0.194519504904747, "learning_rate": 9.190201212222526e-06, "loss": 0.0042, "step": 32550 }, { "grad_norm": 0.11080455780029297, "learning_rate": 9.166331817015155e-06, "loss": 0.0037, "step": 32560 }, { "grad_norm": 0.16668742895126343, "learning_rate": 9.142490331234965e-06, "loss": 0.0044, "step": 32570 }, { "grad_norm": 0.14454907178878784, "learning_rate": 9.11867677117736e-06, "loss": 0.0057, "step": 32580 }, { "grad_norm": 0.11062964797019958, "learning_rate": 9.09489115311869e-06, "loss": 0.0062, "step": 32590 }, { "grad_norm": 0.11933940649032593, "learning_rate": 9.071133493316202e-06, "loss": 0.0064, "step": 32600 }, { "grad_norm": 0.11964009702205658, "learning_rate": 9.047403808008053e-06, "loss": 0.0031, "step": 32610 }, { "grad_norm": 0.11967720091342926, "learning_rate": 9.023702113413224e-06, "loss": 0.0027, "step": 32620 }, { "grad_norm": 0.09507623314857483, "learning_rate": 9.00002842573162e-06, "loss": 0.005, "step": 32630 }, { "grad_norm": 0.12390821427106857, "learning_rate": 8.976382761143981e-06, "loss": 0.0036, "step": 32640 }, { "grad_norm": 0.12378298491239548, "learning_rate": 8.952765135811909e-06, "loss": 0.0029, "step": 32650 }, { "grad_norm": 0.12415184825658798, "learning_rate": 8.929175565877795e-06, "loss": 0.0045, "step": 32660 }, { "grad_norm": 0.12528224289417267, "learning_rate": 8.905614067464924e-06, "loss": 0.005, "step": 32670 }, { "grad_norm": 0.09328286349773407, "learning_rate": 8.882080656677333e-06, "loss": 0.0057, "step": 32680 }, { "grad_norm": 0.11221228539943695, "learning_rate": 8.85857534959989e-06, "loss": 0.0041, "step": 32690 }, { "grad_norm": 0.17357468605041504, "learning_rate": 8.835098162298256e-06, "loss": 0.0042, "step": 32700 }, { "grad_norm": 0.08748792111873627, "learning_rate": 8.811649110818875e-06, "loss": 0.0038, "step": 32710 }, { "grad_norm": 0.07851627469062805, "learning_rate": 8.78822821118893e-06, "loss": 0.0038, "step": 32720 }, { "grad_norm": 0.10433702170848846, "learning_rate": 8.764835479416405e-06, "loss": 0.0049, "step": 32730 }, { "grad_norm": 0.10293769836425781, "learning_rate": 8.741470931490003e-06, "loss": 0.0029, "step": 32740 }, { "grad_norm": 0.11862944811582565, "learning_rate": 8.718134583379189e-06, "loss": 0.0043, "step": 32750 }, { "grad_norm": 0.11793361604213715, "learning_rate": 8.69482645103412e-06, "loss": 0.0035, "step": 32760 }, { "grad_norm": 0.09985620528459549, "learning_rate": 8.671546550385679e-06, "loss": 0.005, "step": 32770 }, { "grad_norm": 0.1421767920255661, "learning_rate": 8.648294897345466e-06, "loss": 0.0041, "step": 32780 }, { "grad_norm": 0.1389259696006775, "learning_rate": 8.62507150780576e-06, "loss": 0.0036, "step": 32790 }, { "grad_norm": 0.0771096721291542, "learning_rate": 8.601876397639542e-06, "loss": 0.0044, "step": 32800 }, { "grad_norm": 0.14386455714702606, "learning_rate": 8.578709582700456e-06, "loss": 0.0034, "step": 32810 }, { "grad_norm": 0.09778507798910141, "learning_rate": 8.555571078822783e-06, "loss": 0.0033, "step": 32820 }, { "grad_norm": 0.16540184617042542, "learning_rate": 8.532460901821477e-06, "loss": 0.0062, "step": 32830 }, { "grad_norm": 0.1419306993484497, "learning_rate": 8.509379067492128e-06, "loss": 0.0032, "step": 32840 }, { "grad_norm": 0.1335560828447342, "learning_rate": 8.486325591610967e-06, "loss": 0.0037, "step": 32850 }, { "grad_norm": 0.12630431354045868, "learning_rate": 8.463300489934811e-06, "loss": 0.0037, "step": 32860 }, { "grad_norm": 0.10387948900461197, "learning_rate": 8.440303778201093e-06, "loss": 0.0039, "step": 32870 }, { "grad_norm": 0.09631961584091187, "learning_rate": 8.417335472127858e-06, "loss": 0.0054, "step": 32880 }, { "grad_norm": 0.07043223828077316, "learning_rate": 8.394395587413722e-06, "loss": 0.0031, "step": 32890 }, { "grad_norm": 0.0806792825460434, "learning_rate": 8.371484139737889e-06, "loss": 0.0036, "step": 32900 }, { "grad_norm": 0.05554519221186638, "learning_rate": 8.348601144760127e-06, "loss": 0.0031, "step": 32910 }, { "grad_norm": 0.09401749819517136, "learning_rate": 8.325746618120716e-06, "loss": 0.004, "step": 32920 }, { "grad_norm": 0.15230271220207214, "learning_rate": 8.302920575440531e-06, "loss": 0.0038, "step": 32930 }, { "grad_norm": 0.10142494738101959, "learning_rate": 8.280123032320953e-06, "loss": 0.0057, "step": 32940 }, { "grad_norm": 0.1529642939567566, "learning_rate": 8.257354004343892e-06, "loss": 0.0045, "step": 32950 }, { "grad_norm": 0.11053799092769623, "learning_rate": 8.234613507071754e-06, "loss": 0.0044, "step": 32960 }, { "grad_norm": 0.16794028878211975, "learning_rate": 8.211901556047475e-06, "loss": 0.0035, "step": 32970 }, { "grad_norm": 0.140781432390213, "learning_rate": 8.189218166794427e-06, "loss": 0.0038, "step": 32980 }, { "grad_norm": 0.08927538990974426, "learning_rate": 8.166563354816509e-06, "loss": 0.0047, "step": 32990 }, { "grad_norm": 0.1280791312456131, "learning_rate": 8.143937135598074e-06, "loss": 0.0043, "step": 33000 }, { "grad_norm": 0.1219049021601677, "learning_rate": 8.121339524603943e-06, "loss": 0.0036, "step": 33010 }, { "grad_norm": 0.12072762101888657, "learning_rate": 8.098770537279338e-06, "loss": 0.0042, "step": 33020 }, { "grad_norm": 0.08342470228672028, "learning_rate": 8.076230189049966e-06, "loss": 0.0053, "step": 33030 }, { "grad_norm": 0.08836820721626282, "learning_rate": 8.05371849532196e-06, "loss": 0.0048, "step": 33040 }, { "grad_norm": 0.08776549249887466, "learning_rate": 8.031235471481813e-06, "loss": 0.0045, "step": 33050 }, { "grad_norm": 0.0773559957742691, "learning_rate": 8.008781132896486e-06, "loss": 0.0037, "step": 33060 }, { "grad_norm": 0.11484235525131226, "learning_rate": 7.986355494913312e-06, "loss": 0.0037, "step": 33070 }, { "grad_norm": 0.07688996195793152, "learning_rate": 7.963958572859975e-06, "loss": 0.0054, "step": 33080 }, { "grad_norm": 0.13353203237056732, "learning_rate": 7.941590382044583e-06, "loss": 0.0058, "step": 33090 }, { "grad_norm": 0.13235455751419067, "learning_rate": 7.919250937755574e-06, "loss": 0.0042, "step": 33100 }, { "grad_norm": 0.10949045419692993, "learning_rate": 7.896940255261764e-06, "loss": 0.0043, "step": 33110 }, { "grad_norm": 0.10692273080348969, "learning_rate": 7.874658349812269e-06, "loss": 0.0038, "step": 33120 }, { "grad_norm": 0.09302773326635361, "learning_rate": 7.85240523663659e-06, "loss": 0.004, "step": 33130 }, { "grad_norm": 0.18548576533794403, "learning_rate": 7.830180930944487e-06, "loss": 0.0061, "step": 33140 }, { "grad_norm": 0.13294236361980438, "learning_rate": 7.80798544792608e-06, "loss": 0.0045, "step": 33150 }, { "grad_norm": 0.11763844639062881, "learning_rate": 7.785818802751776e-06, "loss": 0.0044, "step": 33160 }, { "grad_norm": 0.09282340854406357, "learning_rate": 7.763681010572277e-06, "loss": 0.0054, "step": 33170 }, { "grad_norm": 0.11887186765670776, "learning_rate": 7.741572086518528e-06, "loss": 0.0046, "step": 33180 }, { "grad_norm": 0.1397680938243866, "learning_rate": 7.719492045701782e-06, "loss": 0.0034, "step": 33190 }, { "grad_norm": 0.1261807382106781, "learning_rate": 7.697440903213549e-06, "loss": 0.0037, "step": 33200 }, { "grad_norm": 0.11746427416801453, "learning_rate": 7.675418674125573e-06, "loss": 0.0058, "step": 33210 }, { "grad_norm": 0.132172092795372, "learning_rate": 7.65342537348983e-06, "loss": 0.004, "step": 33220 }, { "grad_norm": 0.13457506895065308, "learning_rate": 7.631461016338549e-06, "loss": 0.0042, "step": 33230 }, { "grad_norm": 0.10311032831668854, "learning_rate": 7.609525617684138e-06, "loss": 0.0036, "step": 33240 }, { "grad_norm": 0.08776789158582687, "learning_rate": 7.587619192519252e-06, "loss": 0.004, "step": 33250 }, { "grad_norm": 0.10577615350484848, "learning_rate": 7.565741755816725e-06, "loss": 0.0061, "step": 33260 }, { "grad_norm": 0.10550834983587265, "learning_rate": 7.543893322529588e-06, "loss": 0.0045, "step": 33270 }, { "grad_norm": 0.0956292375922203, "learning_rate": 7.522073907591015e-06, "loss": 0.0044, "step": 33280 }, { "grad_norm": 0.0810326486825943, "learning_rate": 7.500283525914381e-06, "loss": 0.0043, "step": 33290 }, { "grad_norm": 0.10386461764574051, "learning_rate": 7.478522192393211e-06, "loss": 0.0037, "step": 33300 }, { "grad_norm": 0.10033393651247025, "learning_rate": 7.456789921901181e-06, "loss": 0.0043, "step": 33310 }, { "grad_norm": 0.10634670406579971, "learning_rate": 7.435086729292079e-06, "loss": 0.0029, "step": 33320 }, { "grad_norm": 0.1120009794831276, "learning_rate": 7.413412629399824e-06, "loss": 0.0067, "step": 33330 }, { "grad_norm": 0.16823318600654602, "learning_rate": 7.391767637038466e-06, "loss": 0.004, "step": 33340 }, { "grad_norm": 0.0697619840502739, "learning_rate": 7.370151767002159e-06, "loss": 0.0034, "step": 33350 }, { "grad_norm": 0.13310332596302032, "learning_rate": 7.348565034065147e-06, "loss": 0.0039, "step": 33360 }, { "grad_norm": 0.09563051909208298, "learning_rate": 7.327007452981765e-06, "loss": 0.0039, "step": 33370 }, { "grad_norm": 0.1462254524230957, "learning_rate": 7.305479038486396e-06, "loss": 0.005, "step": 33380 }, { "grad_norm": 0.06069145351648331, "learning_rate": 7.283979805293517e-06, "loss": 0.0044, "step": 33390 }, { "grad_norm": 0.10020074248313904, "learning_rate": 7.262509768097653e-06, "loss": 0.004, "step": 33400 }, { "grad_norm": 0.09090231359004974, "learning_rate": 7.241068941573381e-06, "loss": 0.0042, "step": 33410 }, { "grad_norm": 0.10855713486671448, "learning_rate": 7.219657340375296e-06, "loss": 0.0033, "step": 33420 }, { "grad_norm": 0.10569004714488983, "learning_rate": 7.198274979138014e-06, "loss": 0.0043, "step": 33430 }, { "grad_norm": 0.12802855670452118, "learning_rate": 7.176921872476189e-06, "loss": 0.0085, "step": 33440 }, { "grad_norm": 0.11563669890165329, "learning_rate": 7.1555980349844674e-06, "loss": 0.0036, "step": 33450 }, { "grad_norm": 0.06847187876701355, "learning_rate": 7.134303481237492e-06, "loss": 0.0031, "step": 33460 }, { "grad_norm": 0.12498139590024948, "learning_rate": 7.113038225789892e-06, "loss": 0.0049, "step": 33470 }, { "grad_norm": 0.0980992540717125, "learning_rate": 7.091802283176258e-06, "loss": 0.0056, "step": 33480 }, { "grad_norm": 0.08297126740217209, "learning_rate": 7.070595667911162e-06, "loss": 0.0038, "step": 33490 }, { "grad_norm": 0.07942487299442291, "learning_rate": 7.049418394489127e-06, "loss": 0.0029, "step": 33500 }, { "grad_norm": 0.14236962795257568, "learning_rate": 7.028270477384636e-06, "loss": 0.0059, "step": 33510 }, { "grad_norm": 0.08501803129911423, "learning_rate": 7.0071519310520695e-06, "loss": 0.0028, "step": 33520 }, { "grad_norm": 0.13374818861484528, "learning_rate": 6.98606276992575e-06, "loss": 0.0056, "step": 33530 }, { "grad_norm": 0.10020502656698227, "learning_rate": 6.965003008419929e-06, "loss": 0.0031, "step": 33540 }, { "grad_norm": 0.09406572580337524, "learning_rate": 6.943972660928755e-06, "loss": 0.0044, "step": 33550 }, { "grad_norm": 0.06586291640996933, "learning_rate": 6.922971741826273e-06, "loss": 0.0035, "step": 33560 }, { "grad_norm": 0.08429066091775894, "learning_rate": 6.902000265466424e-06, "loss": 0.0036, "step": 33570 }, { "grad_norm": 0.10648038238286972, "learning_rate": 6.881058246182998e-06, "loss": 0.0043, "step": 33580 }, { "grad_norm": 0.10560210794210434, "learning_rate": 6.860145698289666e-06, "loss": 0.0038, "step": 33590 }, { "grad_norm": 0.08922210335731506, "learning_rate": 6.839262636079985e-06, "loss": 0.0035, "step": 33600 }, { "grad_norm": 0.08854181319475174, "learning_rate": 6.818409073827292e-06, "loss": 0.005, "step": 33610 }, { "grad_norm": 0.10487346351146698, "learning_rate": 6.79758502578483e-06, "loss": 0.0034, "step": 33620 }, { "grad_norm": 0.1053224727511406, "learning_rate": 6.776790506185621e-06, "loss": 0.0042, "step": 33630 }, { "grad_norm": 0.10107452422380447, "learning_rate": 6.756025529242533e-06, "loss": 0.0036, "step": 33640 }, { "grad_norm": 0.2047710418701172, "learning_rate": 6.735290109148223e-06, "loss": 0.0052, "step": 33650 }, { "grad_norm": 0.10034523904323578, "learning_rate": 6.714584260075163e-06, "loss": 0.0039, "step": 33660 }, { "grad_norm": 0.14727236330509186, "learning_rate": 6.693907996175619e-06, "loss": 0.0042, "step": 33670 }, { "grad_norm": 0.09890954941511154, "learning_rate": 6.673261331581593e-06, "loss": 0.0032, "step": 33680 }, { "grad_norm": 0.09348536282777786, "learning_rate": 6.652644280404912e-06, "loss": 0.0058, "step": 33690 }, { "grad_norm": 0.08689448237419128, "learning_rate": 6.6320568567371175e-06, "loss": 0.0031, "step": 33700 }, { "grad_norm": 0.11115644127130508, "learning_rate": 6.611499074649518e-06, "loss": 0.0042, "step": 33710 }, { "grad_norm": 0.09864206612110138, "learning_rate": 6.590970948193187e-06, "loss": 0.0042, "step": 33720 }, { "grad_norm": 0.10704230517148972, "learning_rate": 6.5704724913988825e-06, "loss": 0.0035, "step": 33730 }, { "grad_norm": 0.08964058756828308, "learning_rate": 6.550003718277109e-06, "loss": 0.0033, "step": 33740 }, { "grad_norm": 0.10951913893222809, "learning_rate": 6.529564642818087e-06, "loss": 0.0041, "step": 33750 }, { "grad_norm": 0.0898650735616684, "learning_rate": 6.50915527899173e-06, "loss": 0.0037, "step": 33760 }, { "grad_norm": 0.10676794499158859, "learning_rate": 6.4887756407476575e-06, "loss": 0.0037, "step": 33770 }, { "grad_norm": 0.09391137212514877, "learning_rate": 6.468425742015149e-06, "loss": 0.0035, "step": 33780 }, { "grad_norm": 0.0922347903251648, "learning_rate": 6.448105596703152e-06, "loss": 0.004, "step": 33790 }, { "grad_norm": 0.09709522128105164, "learning_rate": 6.427815218700317e-06, "loss": 0.0043, "step": 33800 }, { "grad_norm": 0.11257806420326233, "learning_rate": 6.407554621874911e-06, "loss": 0.0042, "step": 33810 }, { "grad_norm": 0.08344800770282745, "learning_rate": 6.3873238200748864e-06, "loss": 0.0029, "step": 33820 }, { "grad_norm": 0.0985189750790596, "learning_rate": 6.367122827127775e-06, "loss": 0.0042, "step": 33830 }, { "grad_norm": 0.12847799062728882, "learning_rate": 6.346951656840777e-06, "loss": 0.003, "step": 33840 }, { "grad_norm": 0.14722305536270142, "learning_rate": 6.326810323000698e-06, "loss": 0.004, "step": 33850 }, { "grad_norm": 0.061545293778181076, "learning_rate": 6.306698839373948e-06, "loss": 0.0027, "step": 33860 }, { "grad_norm": 0.07445794343948364, "learning_rate": 6.286617219706547e-06, "loss": 0.0034, "step": 33870 }, { "grad_norm": 0.12765532732009888, "learning_rate": 6.26656547772409e-06, "loss": 0.0043, "step": 33880 }, { "grad_norm": 0.14234523475170135, "learning_rate": 6.246543627131734e-06, "loss": 0.0044, "step": 33890 }, { "grad_norm": 0.09162488579750061, "learning_rate": 6.22655168161424e-06, "loss": 0.0052, "step": 33900 }, { "grad_norm": 0.07871918380260468, "learning_rate": 6.20658965483591e-06, "loss": 0.0025, "step": 33910 }, { "grad_norm": 0.07403083890676498, "learning_rate": 6.186657560440612e-06, "loss": 0.0037, "step": 33920 }, { "grad_norm": 0.1132790744304657, "learning_rate": 6.166755412051734e-06, "loss": 0.0057, "step": 33930 }, { "grad_norm": 0.0933888852596283, "learning_rate": 6.1468832232721985e-06, "loss": 0.0037, "step": 33940 }, { "grad_norm": 0.10854902118444443, "learning_rate": 6.127041007684476e-06, "loss": 0.003, "step": 33950 }, { "grad_norm": 0.10687298327684402, "learning_rate": 6.107228778850527e-06, "loss": 0.0047, "step": 33960 }, { "grad_norm": 0.11989617347717285, "learning_rate": 6.087446550311837e-06, "loss": 0.0048, "step": 33970 }, { "grad_norm": 0.15156690776348114, "learning_rate": 6.067694335589353e-06, "loss": 0.0054, "step": 33980 }, { "grad_norm": 0.12149102240800858, "learning_rate": 6.047972148183528e-06, "loss": 0.004, "step": 33990 }, { "grad_norm": 0.16202481091022491, "learning_rate": 6.0282800015742936e-06, "loss": 0.0048, "step": 34000 }, { "grad_norm": 0.06925927102565765, "learning_rate": 6.008617909221054e-06, "loss": 0.0029, "step": 34010 }, { "grad_norm": 0.140586256980896, "learning_rate": 5.988985884562664e-06, "loss": 0.0039, "step": 34020 }, { "grad_norm": 0.07001437246799469, "learning_rate": 5.969383941017414e-06, "loss": 0.0052, "step": 34030 }, { "grad_norm": 0.06469070911407471, "learning_rate": 5.949812091983059e-06, "loss": 0.0038, "step": 34040 }, { "grad_norm": 0.06737949699163437, "learning_rate": 5.9302703508367695e-06, "loss": 0.0031, "step": 34050 }, { "grad_norm": 0.09980335831642151, "learning_rate": 5.910758730935157e-06, "loss": 0.0055, "step": 34060 }, { "grad_norm": 0.12215128540992737, "learning_rate": 5.8912772456142016e-06, "loss": 0.0039, "step": 34070 }, { "grad_norm": 0.0718945860862732, "learning_rate": 5.871825908189338e-06, "loss": 0.0045, "step": 34080 }, { "grad_norm": 0.12528353929519653, "learning_rate": 5.852404731955352e-06, "loss": 0.0045, "step": 34090 }, { "grad_norm": 0.06887051463127136, "learning_rate": 5.833013730186443e-06, "loss": 0.0027, "step": 34100 }, { "grad_norm": 0.11900301277637482, "learning_rate": 5.813652916136181e-06, "loss": 0.0055, "step": 34110 }, { "grad_norm": 0.10093013942241669, "learning_rate": 5.794322303037508e-06, "loss": 0.005, "step": 34120 }, { "grad_norm": 0.1598559468984604, "learning_rate": 5.775021904102701e-06, "loss": 0.0042, "step": 34130 }, { "grad_norm": 0.0965147316455841, "learning_rate": 5.755751732523401e-06, "loss": 0.0025, "step": 34140 }, { "grad_norm": 0.1158103421330452, "learning_rate": 5.736511801470595e-06, "loss": 0.0041, "step": 34150 }, { "grad_norm": 0.08495432883501053, "learning_rate": 5.717302124094603e-06, "loss": 0.0034, "step": 34160 }, { "grad_norm": 0.07705666869878769, "learning_rate": 5.6981227135250396e-06, "loss": 0.0042, "step": 34170 }, { "grad_norm": 0.08536031097173691, "learning_rate": 5.678973582870867e-06, "loss": 0.0041, "step": 34180 }, { "grad_norm": 0.15279099345207214, "learning_rate": 5.659854745220317e-06, "loss": 0.0034, "step": 34190 }, { "grad_norm": 0.1159173771739006, "learning_rate": 5.640766213640947e-06, "loss": 0.003, "step": 34200 }, { "grad_norm": 0.13019950687885284, "learning_rate": 5.6217080011795934e-06, "loss": 0.0029, "step": 34210 }, { "grad_norm": 0.10801558196544647, "learning_rate": 5.602680120862364e-06, "loss": 0.0033, "step": 34220 }, { "grad_norm": 0.14303268492221832, "learning_rate": 5.583682585694616e-06, "loss": 0.0037, "step": 34230 }, { "grad_norm": 0.12891319394111633, "learning_rate": 5.564715408661009e-06, "loss": 0.0048, "step": 34240 }, { "grad_norm": 0.06247710436582565, "learning_rate": 5.5457786027254265e-06, "loss": 0.0042, "step": 34250 }, { "grad_norm": 0.10719195753335953, "learning_rate": 5.526872180830977e-06, "loss": 0.0044, "step": 34260 }, { "grad_norm": 0.11667023599147797, "learning_rate": 5.507996155900041e-06, "loss": 0.004, "step": 34270 }, { "grad_norm": 0.10163608193397522, "learning_rate": 5.489150540834203e-06, "loss": 0.0046, "step": 34280 }, { "grad_norm": 0.15981197357177734, "learning_rate": 5.47033534851425e-06, "loss": 0.0038, "step": 34290 }, { "grad_norm": 0.08227571099996567, "learning_rate": 5.451550591800198e-06, "loss": 0.0028, "step": 34300 }, { "grad_norm": 0.1383172571659088, "learning_rate": 5.43279628353125e-06, "loss": 0.003, "step": 34310 }, { "grad_norm": 0.16411307454109192, "learning_rate": 5.4140724365258075e-06, "loss": 0.005, "step": 34320 }, { "grad_norm": 0.08319058269262314, "learning_rate": 5.395379063581424e-06, "loss": 0.0039, "step": 34330 }, { "grad_norm": 0.09223373234272003, "learning_rate": 5.376716177474866e-06, "loss": 0.003, "step": 34340 }, { "grad_norm": 0.1029282957315445, "learning_rate": 5.358083790962015e-06, "loss": 0.0041, "step": 34350 }, { "grad_norm": 0.15982909500598907, "learning_rate": 5.33948191677795e-06, "loss": 0.0042, "step": 34360 }, { "grad_norm": 0.11747629195451736, "learning_rate": 5.320910567636866e-06, "loss": 0.0033, "step": 34370 }, { "grad_norm": 0.11903700232505798, "learning_rate": 5.302369756232123e-06, "loss": 0.0038, "step": 34380 }, { "grad_norm": 0.2048349231481552, "learning_rate": 5.283859495236165e-06, "loss": 0.0065, "step": 34390 }, { "grad_norm": 0.1091570183634758, "learning_rate": 5.265379797300602e-06, "loss": 0.0031, "step": 34400 }, { "grad_norm": 0.10874579101800919, "learning_rate": 5.246930675056117e-06, "loss": 0.0031, "step": 34410 }, { "grad_norm": 0.0796835869550705, "learning_rate": 5.228512141112535e-06, "loss": 0.0032, "step": 34420 }, { "grad_norm": 0.11896440386772156, "learning_rate": 5.21012420805872e-06, "loss": 0.0038, "step": 34430 }, { "grad_norm": 0.08550026267766953, "learning_rate": 5.191766888462679e-06, "loss": 0.0028, "step": 34440 }, { "grad_norm": 0.09624554961919785, "learning_rate": 5.173440194871437e-06, "loss": 0.0042, "step": 34450 }, { "grad_norm": 0.10505911707878113, "learning_rate": 5.155144139811141e-06, "loss": 0.0037, "step": 34460 }, { "grad_norm": 0.06934694200754166, "learning_rate": 5.13687873578696e-06, "loss": 0.003, "step": 34470 }, { "grad_norm": 0.11232375353574753, "learning_rate": 5.11864399528314e-06, "loss": 0.004, "step": 34480 }, { "grad_norm": 0.09632185846567154, "learning_rate": 5.1004399307629346e-06, "loss": 0.0034, "step": 34490 }, { "grad_norm": 0.07744576036930084, "learning_rate": 5.082266554668652e-06, "loss": 0.0035, "step": 34500 }, { "grad_norm": 0.11207045614719391, "learning_rate": 5.06412387942164e-06, "loss": 0.0039, "step": 34510 }, { "grad_norm": 0.11576416343450546, "learning_rate": 5.046011917422239e-06, "loss": 0.0042, "step": 34520 }, { "grad_norm": 0.10296923667192459, "learning_rate": 5.0279306810498026e-06, "loss": 0.0045, "step": 34530 }, { "grad_norm": 0.06413029134273529, "learning_rate": 5.009880182662679e-06, "loss": 0.0038, "step": 34540 }, { "grad_norm": 0.06693041324615479, "learning_rate": 4.991860434598217e-06, "loss": 0.0033, "step": 34550 }, { "grad_norm": 0.10059382766485214, "learning_rate": 4.973871449172746e-06, "loss": 0.0037, "step": 34560 }, { "grad_norm": 0.08150885254144669, "learning_rate": 4.9559132386815695e-06, "loss": 0.0035, "step": 34570 }, { "grad_norm": 0.0867142379283905, "learning_rate": 4.93798581539896e-06, "loss": 0.0047, "step": 34580 }, { "grad_norm": 0.11549031734466553, "learning_rate": 4.920089191578131e-06, "loss": 0.0034, "step": 34590 }, { "grad_norm": 0.09983906149864197, "learning_rate": 4.90222337945126e-06, "loss": 0.0028, "step": 34600 }, { "grad_norm": 0.060929328203201294, "learning_rate": 4.8843883912294605e-06, "loss": 0.0041, "step": 34610 }, { "grad_norm": 0.08800256997346878, "learning_rate": 4.866584239102795e-06, "loss": 0.0044, "step": 34620 }, { "grad_norm": 0.09540706872940063, "learning_rate": 4.8488109352402164e-06, "loss": 0.0052, "step": 34630 }, { "grad_norm": 0.19549062848091125, "learning_rate": 4.831068491789603e-06, "loss": 0.0048, "step": 34640 }, { "grad_norm": 0.14069566130638123, "learning_rate": 4.81335692087776e-06, "loss": 0.0062, "step": 34650 }, { "grad_norm": 0.10337714850902557, "learning_rate": 4.795676234610374e-06, "loss": 0.0028, "step": 34660 }, { "grad_norm": 0.11924330145120621, "learning_rate": 4.778026445072037e-06, "loss": 0.0034, "step": 34670 }, { "grad_norm": 0.10668422281742096, "learning_rate": 4.760407564326208e-06, "loss": 0.0043, "step": 34680 }, { "grad_norm": 0.11382598429918289, "learning_rate": 4.742819604415222e-06, "loss": 0.0034, "step": 34690 }, { "grad_norm": 0.09917318820953369, "learning_rate": 4.725262577360284e-06, "loss": 0.0046, "step": 34700 }, { "grad_norm": 0.08016540855169296, "learning_rate": 4.707736495161458e-06, "loss": 0.0038, "step": 34710 }, { "grad_norm": 0.14519785344600677, "learning_rate": 4.690241369797671e-06, "loss": 0.0038, "step": 34720 }, { "grad_norm": 0.10969899594783783, "learning_rate": 4.672777213226659e-06, "loss": 0.0031, "step": 34730 }, { "grad_norm": 0.09164915233850479, "learning_rate": 4.655344037385001e-06, "loss": 0.0049, "step": 34740 }, { "grad_norm": 0.08026550710201263, "learning_rate": 4.6379418541881215e-06, "loss": 0.0042, "step": 34750 }, { "grad_norm": 0.11056502163410187, "learning_rate": 4.620570675530245e-06, "loss": 0.0032, "step": 34760 }, { "grad_norm": 0.1283838152885437, "learning_rate": 4.603230513284407e-06, "loss": 0.0035, "step": 34770 }, { "grad_norm": 0.07549533993005753, "learning_rate": 4.58592137930246e-06, "loss": 0.003, "step": 34780 }, { "grad_norm": 0.1341305524110794, "learning_rate": 4.568643285415009e-06, "loss": 0.003, "step": 34790 }, { "grad_norm": 0.10239654779434204, "learning_rate": 4.551396243431483e-06, "loss": 0.0033, "step": 34800 }, { "grad_norm": 0.13159938156604767, "learning_rate": 4.534180265140075e-06, "loss": 0.004, "step": 34810 }, { "grad_norm": 0.12606680393218994, "learning_rate": 4.5169953623077356e-06, "loss": 0.0031, "step": 34820 }, { "grad_norm": 0.05581028014421463, "learning_rate": 4.499841546680189e-06, "loss": 0.0041, "step": 34830 }, { "grad_norm": 0.10099323838949203, "learning_rate": 4.4827188299819025e-06, "loss": 0.0031, "step": 34840 }, { "grad_norm": 0.07736280560493469, "learning_rate": 4.465627223916086e-06, "loss": 0.0032, "step": 34850 }, { "grad_norm": 0.08794655650854111, "learning_rate": 4.448566740164706e-06, "loss": 0.0027, "step": 34860 }, { "grad_norm": 0.08546878397464752, "learning_rate": 4.431537390388429e-06, "loss": 0.0052, "step": 34870 }, { "grad_norm": 0.05301795154809952, "learning_rate": 4.414539186226674e-06, "loss": 0.0031, "step": 34880 }, { "grad_norm": 0.07016777992248535, "learning_rate": 4.397572139297529e-06, "loss": 0.0039, "step": 34890 }, { "grad_norm": 0.102296844124794, "learning_rate": 4.380636261197829e-06, "loss": 0.003, "step": 34900 }, { "grad_norm": 0.1176643967628479, "learning_rate": 4.36373156350307e-06, "loss": 0.0028, "step": 34910 }, { "grad_norm": 0.12016759067773819, "learning_rate": 4.3468580577674665e-06, "loss": 0.0033, "step": 34920 }, { "grad_norm": 0.09493686258792877, "learning_rate": 4.330015755523908e-06, "loss": 0.0033, "step": 34930 }, { "grad_norm": 0.1432352066040039, "learning_rate": 4.313204668283926e-06, "loss": 0.0049, "step": 34940 }, { "grad_norm": 0.15033677220344543, "learning_rate": 4.296424807537752e-06, "loss": 0.0034, "step": 34950 }, { "grad_norm": 0.08998481184244156, "learning_rate": 4.279676184754267e-06, "loss": 0.0038, "step": 34960 }, { "grad_norm": 0.060178954154253006, "learning_rate": 4.262958811380985e-06, "loss": 0.0032, "step": 34970 }, { "grad_norm": 0.08550170063972473, "learning_rate": 4.246272698844095e-06, "loss": 0.0048, "step": 34980 }, { "grad_norm": 0.08733254671096802, "learning_rate": 4.229617858548374e-06, "loss": 0.0036, "step": 34990 }, { "grad_norm": 0.10395945608615875, "learning_rate": 4.212994301877271e-06, "loss": 0.0034, "step": 35000 }, { "grad_norm": 0.10918872058391571, "learning_rate": 4.196402040192804e-06, "loss": 0.0066, "step": 35010 }, { "grad_norm": 0.09511607140302658, "learning_rate": 4.179841084835639e-06, "loss": 0.0037, "step": 35020 }, { "grad_norm": 0.12096681445837021, "learning_rate": 4.163311447125046e-06, "loss": 0.0042, "step": 35030 }, { "grad_norm": 0.09122002869844437, "learning_rate": 4.146813138358851e-06, "loss": 0.0043, "step": 35040 }, { "grad_norm": 0.1022442877292633, "learning_rate": 4.130346169813509e-06, "loss": 0.0032, "step": 35050 }, { "grad_norm": 0.07762724906206131, "learning_rate": 4.113910552744038e-06, "loss": 0.0029, "step": 35060 }, { "grad_norm": 0.12477196753025055, "learning_rate": 4.097506298384024e-06, "loss": 0.0042, "step": 35070 }, { "grad_norm": 0.08460105210542679, "learning_rate": 4.081133417945626e-06, "loss": 0.0063, "step": 35080 }, { "grad_norm": 0.10797902196645737, "learning_rate": 4.0647919226195484e-06, "loss": 0.0043, "step": 35090 }, { "grad_norm": 0.1431213766336441, "learning_rate": 4.04848182357504e-06, "loss": 0.0038, "step": 35100 }, { "grad_norm": 0.11654757708311081, "learning_rate": 4.032203131959911e-06, "loss": 0.006, "step": 35110 }, { "grad_norm": 0.13820131123065948, "learning_rate": 4.015955858900483e-06, "loss": 0.0041, "step": 35120 }, { "grad_norm": 0.08831765502691269, "learning_rate": 3.999740015501641e-06, "loss": 0.0033, "step": 35130 }, { "grad_norm": 0.08680501580238342, "learning_rate": 3.9835556128467236e-06, "loss": 0.0029, "step": 35140 }, { "grad_norm": 0.10221081972122192, "learning_rate": 3.967402661997644e-06, "loss": 0.0033, "step": 35150 }, { "grad_norm": 0.12589576840400696, "learning_rate": 3.951281173994781e-06, "loss": 0.0032, "step": 35160 }, { "grad_norm": 0.07078519463539124, "learning_rate": 3.935191159857016e-06, "loss": 0.0029, "step": 35170 }, { "grad_norm": 0.08701300621032715, "learning_rate": 3.919132630581746e-06, "loss": 0.0053, "step": 35180 }, { "grad_norm": 0.07888630777597427, "learning_rate": 3.903105597144802e-06, "loss": 0.0056, "step": 35190 }, { "grad_norm": 0.10134683549404144, "learning_rate": 3.887110070500505e-06, "loss": 0.0063, "step": 35200 }, { "grad_norm": 0.05550084263086319, "learning_rate": 3.871146061581654e-06, "loss": 0.0031, "step": 35210 }, { "grad_norm": 0.1593247950077057, "learning_rate": 3.855213581299505e-06, "loss": 0.0033, "step": 35220 }, { "grad_norm": 0.07804623246192932, "learning_rate": 3.839312640543763e-06, "loss": 0.0034, "step": 35230 }, { "grad_norm": 0.10687362402677536, "learning_rate": 3.823443250182551e-06, "loss": 0.0038, "step": 35240 }, { "grad_norm": 0.05918705463409424, "learning_rate": 3.80760542106246e-06, "loss": 0.0037, "step": 35250 }, { "grad_norm": 0.07757000625133514, "learning_rate": 3.7917991640084972e-06, "loss": 0.0039, "step": 35260 }, { "grad_norm": 0.11512293666601181, "learning_rate": 3.7760244898240925e-06, "loss": 0.0034, "step": 35270 }, { "grad_norm": 0.053974512964487076, "learning_rate": 3.760281409291078e-06, "loss": 0.0024, "step": 35280 }, { "grad_norm": 0.08659304678440094, "learning_rate": 3.7445699331697113e-06, "loss": 0.004, "step": 35290 }, { "grad_norm": 0.08528207987546921, "learning_rate": 3.7288900721986262e-06, "loss": 0.0043, "step": 35300 }, { "grad_norm": 0.1194339469075203, "learning_rate": 3.7132418370948663e-06, "loss": 0.0035, "step": 35310 }, { "grad_norm": 0.1283564269542694, "learning_rate": 3.6976252385538535e-06, "loss": 0.004, "step": 35320 }, { "grad_norm": 0.17751257121562958, "learning_rate": 3.6820402872493832e-06, "loss": 0.0045, "step": 35330 }, { "grad_norm": 0.0437438003718853, "learning_rate": 3.6664869938336343e-06, "loss": 0.0043, "step": 35340 }, { "grad_norm": 0.10017368197441101, "learning_rate": 3.650965368937115e-06, "loss": 0.0043, "step": 35350 }, { "grad_norm": 0.0760963037610054, "learning_rate": 3.6354754231687214e-06, "loss": 0.0032, "step": 35360 }, { "grad_norm": 0.07934436947107315, "learning_rate": 3.620017167115697e-06, "loss": 0.006, "step": 35370 }, { "grad_norm": 0.08390868455171585, "learning_rate": 3.604590611343589e-06, "loss": 0.0029, "step": 35380 }, { "grad_norm": 0.08080090582370758, "learning_rate": 3.589195766396325e-06, "loss": 0.0041, "step": 35390 }, { "grad_norm": 0.13456401228904724, "learning_rate": 3.573832642796121e-06, "loss": 0.0039, "step": 35400 }, { "grad_norm": 0.09478971362113953, "learning_rate": 3.5585012510435336e-06, "loss": 0.003, "step": 35410 }, { "grad_norm": 0.12513136863708496, "learning_rate": 3.5432016016174184e-06, "loss": 0.005, "step": 35420 }, { "grad_norm": 0.07305371761322021, "learning_rate": 3.5279337049749506e-06, "loss": 0.0041, "step": 35430 }, { "grad_norm": 0.10470398515462875, "learning_rate": 3.5126975715515985e-06, "loss": 0.0058, "step": 35440 }, { "grad_norm": 0.11192943900823593, "learning_rate": 3.4974932117610994e-06, "loss": 0.0039, "step": 35450 }, { "grad_norm": 0.07151461392641068, "learning_rate": 3.4823206359955065e-06, "loss": 0.0045, "step": 35460 }, { "grad_norm": 0.0902549996972084, "learning_rate": 3.4671798546251133e-06, "loss": 0.0044, "step": 35470 }, { "grad_norm": 0.106941357254982, "learning_rate": 3.4520708779985088e-06, "loss": 0.0043, "step": 35480 }, { "grad_norm": 0.09966153651475906, "learning_rate": 3.4369937164425493e-06, "loss": 0.0022, "step": 35490 }, { "grad_norm": 0.08682692050933838, "learning_rate": 3.421948380262313e-06, "loss": 0.0047, "step": 35500 }, { "grad_norm": 0.08856180310249329, "learning_rate": 3.406934879741158e-06, "loss": 0.0056, "step": 35510 }, { "grad_norm": 0.07798032462596893, "learning_rate": 3.391953225140665e-06, "loss": 0.0042, "step": 35520 }, { "grad_norm": 0.10433397442102432, "learning_rate": 3.3770034267006533e-06, "loss": 0.0041, "step": 35530 }, { "grad_norm": 0.12587742507457733, "learning_rate": 3.362085494639178e-06, "loss": 0.0038, "step": 35540 }, { "grad_norm": 0.10642985254526138, "learning_rate": 3.3471994391524986e-06, "loss": 0.0031, "step": 35550 }, { "grad_norm": 0.052647680044174194, "learning_rate": 3.3323452704150825e-06, "loss": 0.0039, "step": 35560 }, { "grad_norm": 0.09645097702741623, "learning_rate": 3.3175229985796253e-06, "loss": 0.0037, "step": 35570 }, { "grad_norm": 0.1030443087220192, "learning_rate": 3.3027326337770115e-06, "loss": 0.0031, "step": 35580 }, { "grad_norm": 0.11345706880092621, "learning_rate": 3.2879741861163216e-06, "loss": 0.0035, "step": 35590 }, { "grad_norm": 0.09662766009569168, "learning_rate": 3.2732476656847976e-06, "loss": 0.0038, "step": 35600 }, { "grad_norm": 0.0909786969423294, "learning_rate": 3.2585530825478883e-06, "loss": 0.0035, "step": 35610 }, { "grad_norm": 0.10541228950023651, "learning_rate": 3.243890446749204e-06, "loss": 0.0048, "step": 35620 }, { "grad_norm": 0.10751161724328995, "learning_rate": 3.2292597683105107e-06, "loss": 0.0051, "step": 35630 }, { "grad_norm": 0.19057020545005798, "learning_rate": 3.214661057231755e-06, "loss": 0.0052, "step": 35640 }, { "grad_norm": 0.13011249899864197, "learning_rate": 3.2000943234910097e-06, "loss": 0.0035, "step": 35650 }, { "grad_norm": 0.06817211955785751, "learning_rate": 3.1855595770444835e-06, "loss": 0.0035, "step": 35660 }, { "grad_norm": 0.08564140647649765, "learning_rate": 3.1710568278265574e-06, "loss": 0.0034, "step": 35670 }, { "grad_norm": 0.07203496247529984, "learning_rate": 3.156586085749719e-06, "loss": 0.0039, "step": 35680 }, { "grad_norm": 0.14957278966903687, "learning_rate": 3.1421473607045893e-06, "loss": 0.0053, "step": 35690 }, { "grad_norm": 0.09967726469039917, "learning_rate": 3.127740662559886e-06, "loss": 0.0046, "step": 35700 }, { "grad_norm": 0.14874985814094543, "learning_rate": 3.1133660011624655e-06, "loss": 0.0032, "step": 35710 }, { "grad_norm": 0.10500947386026382, "learning_rate": 3.099023386337263e-06, "loss": 0.0059, "step": 35720 }, { "grad_norm": 0.10476453602313995, "learning_rate": 3.0847128278873317e-06, "loss": 0.0042, "step": 35730 }, { "grad_norm": 0.0812622532248497, "learning_rate": 3.0704343355938027e-06, "loss": 0.0032, "step": 35740 }, { "grad_norm": 0.09100857377052307, "learning_rate": 3.056187919215886e-06, "loss": 0.0034, "step": 35750 }, { "grad_norm": 0.08917110413312912, "learning_rate": 3.041973588490865e-06, "loss": 0.0033, "step": 35760 }, { "grad_norm": 0.1205216571688652, "learning_rate": 3.0277913531341128e-06, "loss": 0.0044, "step": 35770 }, { "grad_norm": 0.07507552951574326, "learning_rate": 3.0136412228390476e-06, "loss": 0.0023, "step": 35780 }, { "grad_norm": 0.11115548014640808, "learning_rate": 2.999523207277155e-06, "loss": 0.0029, "step": 35790 }, { "grad_norm": 0.15694497525691986, "learning_rate": 2.9854373160979552e-06, "loss": 0.0045, "step": 35800 }, { "grad_norm": 0.137578085064888, "learning_rate": 2.9713835589290362e-06, "loss": 0.0046, "step": 35810 }, { "grad_norm": 0.08479297906160355, "learning_rate": 2.9573619453759913e-06, "loss": 0.0029, "step": 35820 }, { "grad_norm": 0.07929375022649765, "learning_rate": 2.943372485022489e-06, "loss": 0.0033, "step": 35830 }, { "grad_norm": 0.10288994014263153, "learning_rate": 2.9294151874301623e-06, "loss": 0.0035, "step": 35840 }, { "grad_norm": 0.06006420776247978, "learning_rate": 2.9154900621387093e-06, "loss": 0.0035, "step": 35850 }, { "grad_norm": 0.08204763382673264, "learning_rate": 2.9015971186658164e-06, "loss": 0.0025, "step": 35860 }, { "grad_norm": 0.09177490323781967, "learning_rate": 2.887736366507182e-06, "loss": 0.0052, "step": 35870 }, { "grad_norm": 0.09458590298891068, "learning_rate": 2.873907815136501e-06, "loss": 0.0027, "step": 35880 }, { "grad_norm": 0.12126480787992477, "learning_rate": 2.8601114740054556e-06, "loss": 0.0031, "step": 35890 }, { "grad_norm": 0.0848330408334732, "learning_rate": 2.8463473525437155e-06, "loss": 0.0027, "step": 35900 }, { "grad_norm": 0.11264874041080475, "learning_rate": 2.8326154601589216e-06, "loss": 0.005, "step": 35910 }, { "grad_norm": 0.07782037556171417, "learning_rate": 2.818915806236705e-06, "loss": 0.0025, "step": 35920 }, { "grad_norm": 0.06906133145093918, "learning_rate": 2.8052484001406478e-06, "loss": 0.0032, "step": 35930 }, { "grad_norm": 0.055715303868055344, "learning_rate": 2.791613251212277e-06, "loss": 0.0028, "step": 35940 }, { "grad_norm": 0.15297621488571167, "learning_rate": 2.7780103687711166e-06, "loss": 0.0034, "step": 35950 }, { "grad_norm": 0.053766511380672455, "learning_rate": 2.764439762114579e-06, "loss": 0.0028, "step": 35960 }, { "grad_norm": 0.14104063808918, "learning_rate": 2.750901440518061e-06, "loss": 0.0037, "step": 35970 }, { "grad_norm": 0.12919391691684723, "learning_rate": 2.7373954132348776e-06, "loss": 0.0039, "step": 35980 }, { "grad_norm": 0.14518487453460693, "learning_rate": 2.7239216894962784e-06, "loss": 0.004, "step": 35990 }, { "grad_norm": 0.08829747140407562, "learning_rate": 2.710480278511407e-06, "loss": 0.0045, "step": 36000 }, { "grad_norm": 0.07192979753017426, "learning_rate": 2.6970711894673483e-06, "loss": 0.0039, "step": 36010 }, { "grad_norm": 0.143467977643013, "learning_rate": 2.6836944315291047e-06, "loss": 0.0032, "step": 36020 }, { "grad_norm": 0.11751117557287216, "learning_rate": 2.67035001383954e-06, "loss": 0.0056, "step": 36030 }, { "grad_norm": 0.05593769997358322, "learning_rate": 2.6570379455194527e-06, "loss": 0.0041, "step": 36040 }, { "grad_norm": 0.08163795620203018, "learning_rate": 2.6437582356675206e-06, "loss": 0.0031, "step": 36050 }, { "grad_norm": 0.06511399894952774, "learning_rate": 2.6305108933602775e-06, "loss": 0.0044, "step": 36060 }, { "grad_norm": 0.07613523304462433, "learning_rate": 2.6172959276521803e-06, "loss": 0.0035, "step": 36070 }, { "grad_norm": 0.11958958953619003, "learning_rate": 2.6041133475755153e-06, "loss": 0.0039, "step": 36080 }, { "grad_norm": 0.09891034662723541, "learning_rate": 2.590963162140475e-06, "loss": 0.0037, "step": 36090 }, { "grad_norm": 0.11062053591012955, "learning_rate": 2.5778453803350587e-06, "loss": 0.0048, "step": 36100 }, { "grad_norm": 0.09703183174133301, "learning_rate": 2.5647600111251723e-06, "loss": 0.0039, "step": 36110 }, { "grad_norm": 0.08503583818674088, "learning_rate": 2.551707063454528e-06, "loss": 0.0038, "step": 36120 }, { "grad_norm": 0.14105182886123657, "learning_rate": 2.5386865462446897e-06, "loss": 0.0033, "step": 36130 }, { "grad_norm": 0.07012965530157089, "learning_rate": 2.52569846839506e-06, "loss": 0.0028, "step": 36140 }, { "grad_norm": 0.07849352806806564, "learning_rate": 2.512742838782883e-06, "loss": 0.0035, "step": 36150 }, { "grad_norm": 0.05180754512548447, "learning_rate": 2.4998196662631924e-06, "loss": 0.0025, "step": 36160 }, { "grad_norm": 0.07016915827989578, "learning_rate": 2.4869289596688604e-06, "loss": 0.0044, "step": 36170 }, { "grad_norm": 0.0849151685833931, "learning_rate": 2.4740707278105567e-06, "loss": 0.0034, "step": 36180 }, { "grad_norm": 0.085825614631176, "learning_rate": 2.461244979476779e-06, "loss": 0.0033, "step": 36190 }, { "grad_norm": 0.05650397390127182, "learning_rate": 2.4484517234337866e-06, "loss": 0.003, "step": 36200 }, { "grad_norm": 0.08756648749113083, "learning_rate": 2.4356909684256634e-06, "loss": 0.0028, "step": 36210 }, { "grad_norm": 0.0621427483856678, "learning_rate": 2.4229627231742545e-06, "loss": 0.0026, "step": 36220 }, { "grad_norm": 0.11166735738515854, "learning_rate": 2.4102669963791956e-06, "loss": 0.0046, "step": 36230 }, { "grad_norm": 0.12749245762825012, "learning_rate": 2.3976037967179065e-06, "loss": 0.0038, "step": 36240 }, { "grad_norm": 0.06266201287508011, "learning_rate": 2.3849731328455582e-06, "loss": 0.0036, "step": 36250 }, { "grad_norm": 0.0767425075173378, "learning_rate": 2.3723750133950894e-06, "loss": 0.0033, "step": 36260 }, { "grad_norm": 0.08363364636898041, "learning_rate": 2.3598094469771957e-06, "loss": 0.0033, "step": 36270 }, { "grad_norm": 0.10748903453350067, "learning_rate": 2.347276442180324e-06, "loss": 0.0041, "step": 36280 }, { "grad_norm": 0.06768933683633804, "learning_rate": 2.3347760075706715e-06, "loss": 0.0032, "step": 36290 }, { "grad_norm": 0.113580122590065, "learning_rate": 2.3223081516921595e-06, "loss": 0.0035, "step": 36300 }, { "grad_norm": 0.12115823477506638, "learning_rate": 2.3098728830664494e-06, "loss": 0.0032, "step": 36310 }, { "grad_norm": 0.0692443698644638, "learning_rate": 2.2974702101929314e-06, "loss": 0.0026, "step": 36320 }, { "grad_norm": 0.09919312596321106, "learning_rate": 2.2851001415487185e-06, "loss": 0.0043, "step": 36330 }, { "grad_norm": 0.11230067908763885, "learning_rate": 2.272762685588631e-06, "loss": 0.0044, "step": 36340 }, { "grad_norm": 0.0836358591914177, "learning_rate": 2.260457850745218e-06, "loss": 0.0033, "step": 36350 }, { "grad_norm": 0.05905741825699806, "learning_rate": 2.2481856454286963e-06, "loss": 0.0024, "step": 36360 }, { "grad_norm": 0.0580589659512043, "learning_rate": 2.235946078027018e-06, "loss": 0.0035, "step": 36370 }, { "grad_norm": 0.13386918604373932, "learning_rate": 2.223739156905802e-06, "loss": 0.0044, "step": 36380 }, { "grad_norm": 0.09166477620601654, "learning_rate": 2.2115648904083796e-06, "loss": 0.0049, "step": 36390 }, { "grad_norm": 0.07788615673780441, "learning_rate": 2.199423286855734e-06, "loss": 0.003, "step": 36400 }, { "grad_norm": 0.1066918894648552, "learning_rate": 2.1873143545465323e-06, "loss": 0.0031, "step": 36410 }, { "grad_norm": 0.10124900192022324, "learning_rate": 2.1752381017571267e-06, "loss": 0.0031, "step": 36420 }, { "grad_norm": 0.0976029708981514, "learning_rate": 2.1631945367415142e-06, "loss": 0.0038, "step": 36430 }, { "grad_norm": 0.12947742640972137, "learning_rate": 2.15118366773136e-06, "loss": 0.0034, "step": 36440 }, { "grad_norm": 0.14337728917598724, "learning_rate": 2.1392055029359926e-06, "loss": 0.0035, "step": 36450 }, { "grad_norm": 0.06898064911365509, "learning_rate": 2.127260050542357e-06, "loss": 0.0026, "step": 36460 }, { "grad_norm": 0.09669665992259979, "learning_rate": 2.1153473187150608e-06, "loss": 0.0031, "step": 36470 }, { "grad_norm": 0.045183371752500534, "learning_rate": 2.1034673155963467e-06, "loss": 0.0033, "step": 36480 }, { "grad_norm": 0.09917625784873962, "learning_rate": 2.0916200493060976e-06, "loss": 0.003, "step": 36490 }, { "grad_norm": 0.1350312978029251, "learning_rate": 2.0798055279417916e-06, "loss": 0.0031, "step": 36500 }, { "grad_norm": 0.05931716039776802, "learning_rate": 2.0680237595785356e-06, "loss": 0.0024, "step": 36510 }, { "grad_norm": 0.049417078495025635, "learning_rate": 2.0562747522690774e-06, "loss": 0.0024, "step": 36520 }, { "grad_norm": 0.04213760420680046, "learning_rate": 2.0445585140437373e-06, "loss": 0.0027, "step": 36530 }, { "grad_norm": 0.09213299304246902, "learning_rate": 2.0328750529104602e-06, "loss": 0.0042, "step": 36540 }, { "grad_norm": 0.09906043112277985, "learning_rate": 2.021224376854791e-06, "loss": 0.003, "step": 36550 }, { "grad_norm": 0.06177371367812157, "learning_rate": 2.0096064938398386e-06, "loss": 0.0025, "step": 36560 }, { "grad_norm": 0.07626412063837051, "learning_rate": 1.998021411806328e-06, "loss": 0.0026, "step": 36570 }, { "grad_norm": 0.09790221601724625, "learning_rate": 1.9864691386725586e-06, "loss": 0.0056, "step": 36580 }, { "grad_norm": 0.0893969014286995, "learning_rate": 1.9749496823343814e-06, "loss": 0.0048, "step": 36590 }, { "grad_norm": 0.09566424787044525, "learning_rate": 1.963463050665254e-06, "loss": 0.0043, "step": 36600 }, { "grad_norm": 0.08928710967302322, "learning_rate": 1.9520092515161682e-06, "loss": 0.0041, "step": 36610 }, { "grad_norm": 0.06732720136642456, "learning_rate": 1.940588292715695e-06, "loss": 0.0039, "step": 36620 }, { "grad_norm": 0.11824356764554977, "learning_rate": 1.929200182069951e-06, "loss": 0.0031, "step": 36630 }, { "grad_norm": 0.09389873594045639, "learning_rate": 1.9178449273625997e-06, "loss": 0.0029, "step": 36640 }, { "grad_norm": 0.0792841985821724, "learning_rate": 1.90652253635486e-06, "loss": 0.0041, "step": 36650 }, { "grad_norm": 0.08923054486513138, "learning_rate": 1.8952330167854693e-06, "loss": 0.0023, "step": 36660 }, { "grad_norm": 0.0696134939789772, "learning_rate": 1.8839763763707107e-06, "loss": 0.0033, "step": 36670 }, { "grad_norm": 0.05171077325940132, "learning_rate": 1.8727526228043912e-06, "loss": 0.003, "step": 36680 }, { "grad_norm": 0.0923975333571434, "learning_rate": 1.8615617637578408e-06, "loss": 0.0041, "step": 36690 }, { "grad_norm": 0.08490816503763199, "learning_rate": 1.850403806879908e-06, "loss": 0.0036, "step": 36700 }, { "grad_norm": 0.05342545732855797, "learning_rate": 1.8392787597969475e-06, "loss": 0.0031, "step": 36710 }, { "grad_norm": 0.10284788906574249, "learning_rate": 1.828186630112827e-06, "loss": 0.0037, "step": 36720 }, { "grad_norm": 0.06929697096347809, "learning_rate": 1.8171274254089098e-06, "loss": 0.0026, "step": 36730 }, { "grad_norm": 0.09085581451654434, "learning_rate": 1.806101153244061e-06, "loss": 0.003, "step": 36740 }, { "grad_norm": 0.047913808375597, "learning_rate": 1.7951078211546401e-06, "loss": 0.0047, "step": 36750 }, { "grad_norm": 0.07762959599494934, "learning_rate": 1.7841474366544764e-06, "loss": 0.0046, "step": 36760 }, { "grad_norm": 0.09535930305719376, "learning_rate": 1.7732200072348825e-06, "loss": 0.0022, "step": 36770 }, { "grad_norm": 0.059752918779850006, "learning_rate": 1.7623255403646677e-06, "loss": 0.0029, "step": 36780 }, { "grad_norm": 0.06928420811891556, "learning_rate": 1.7514640434900865e-06, "loss": 0.0034, "step": 36790 }, { "grad_norm": 0.11707992851734161, "learning_rate": 1.7406355240348838e-06, "loss": 0.0042, "step": 36800 }, { "grad_norm": 0.04899121820926666, "learning_rate": 1.7298399894002392e-06, "loss": 0.0028, "step": 36810 }, { "grad_norm": 0.09780579805374146, "learning_rate": 1.7190774469647952e-06, "loss": 0.0036, "step": 36820 }, { "grad_norm": 0.12383049726486206, "learning_rate": 1.7083479040846617e-06, "loss": 0.0043, "step": 36830 }, { "grad_norm": 0.12413901835680008, "learning_rate": 1.697651368093367e-06, "loss": 0.0029, "step": 36840 }, { "grad_norm": 0.13420070707798004, "learning_rate": 1.6869878463019017e-06, "loss": 0.0039, "step": 36850 }, { "grad_norm": 0.07177441567182541, "learning_rate": 1.6763573459986858e-06, "loss": 0.0038, "step": 36860 }, { "grad_norm": 0.07205943763256073, "learning_rate": 1.6657598744495516e-06, "loss": 0.0031, "step": 36870 }, { "grad_norm": 0.07045775651931763, "learning_rate": 1.6551954388977774e-06, "loss": 0.0041, "step": 36880 }, { "grad_norm": 0.0764811709523201, "learning_rate": 1.644664046564054e-06, "loss": 0.0025, "step": 36890 }, { "grad_norm": 0.07808586210012436, "learning_rate": 1.634165704646501e-06, "loss": 0.0031, "step": 36900 }, { "grad_norm": 0.14612284302711487, "learning_rate": 1.6237004203206175e-06, "loss": 0.0057, "step": 36910 }, { "grad_norm": 0.06231095269322395, "learning_rate": 1.6132682007393373e-06, "loss": 0.0033, "step": 36920 }, { "grad_norm": 0.05396682024002075, "learning_rate": 1.6028690530329847e-06, "loss": 0.0027, "step": 36930 }, { "grad_norm": 0.12952829897403717, "learning_rate": 1.5925029843092742e-06, "loss": 0.0045, "step": 36940 }, { "grad_norm": 0.14067718386650085, "learning_rate": 1.5821700016533214e-06, "loss": 0.0051, "step": 36950 }, { "grad_norm": 0.08428674936294556, "learning_rate": 1.5718701121276268e-06, "loss": 0.0066, "step": 36960 }, { "grad_norm": 0.12048659473657608, "learning_rate": 1.5616033227720483e-06, "loss": 0.004, "step": 36970 }, { "grad_norm": 0.04738893359899521, "learning_rate": 1.5513696406038558e-06, "loss": 0.0032, "step": 36980 }, { "grad_norm": 0.14830611646175385, "learning_rate": 1.5411690726176653e-06, "loss": 0.0029, "step": 36990 }, { "grad_norm": 0.06116188317537308, "learning_rate": 1.5310016257854832e-06, "loss": 0.0044, "step": 37000 }, { "grad_norm": 0.07593926042318344, "learning_rate": 1.5208673070566504e-06, "loss": 0.0035, "step": 37010 }, { "grad_norm": 0.06972458958625793, "learning_rate": 1.5107661233578818e-06, "loss": 0.0029, "step": 37020 }, { "grad_norm": 0.1016182079911232, "learning_rate": 1.5006980815932437e-06, "loss": 0.0031, "step": 37030 }, { "grad_norm": 0.07411924749612808, "learning_rate": 1.490663188644148e-06, "loss": 0.0053, "step": 37040 }, { "grad_norm": 0.08324994891881943, "learning_rate": 1.4806614513693418e-06, "loss": 0.0038, "step": 37050 }, { "grad_norm": 0.07209443300962448, "learning_rate": 1.4706928766049289e-06, "loss": 0.0048, "step": 37060 }, { "grad_norm": 0.07061842083930969, "learning_rate": 1.4607574711643257e-06, "loss": 0.003, "step": 37070 }, { "grad_norm": 0.08352752774953842, "learning_rate": 1.4508552418382949e-06, "loss": 0.0029, "step": 37080 }, { "grad_norm": 0.08434557914733887, "learning_rate": 1.4409861953949112e-06, "loss": 0.0043, "step": 37090 }, { "grad_norm": 0.08274346590042114, "learning_rate": 1.43115033857959e-06, "loss": 0.0034, "step": 37100 }, { "grad_norm": 0.09229736030101776, "learning_rate": 1.4213476781150258e-06, "loss": 0.0033, "step": 37110 }, { "grad_norm": 0.13185888528823853, "learning_rate": 1.411578220701254e-06, "loss": 0.0025, "step": 37120 }, { "grad_norm": 0.05708669498562813, "learning_rate": 1.4018419730156108e-06, "loss": 0.0029, "step": 37130 }, { "grad_norm": 0.06120669096708298, "learning_rate": 1.3921389417127283e-06, "loss": 0.0026, "step": 37140 }, { "grad_norm": 0.09401502460241318, "learning_rate": 1.3824691334245298e-06, "loss": 0.0037, "step": 37150 }, { "grad_norm": 0.06648958474397659, "learning_rate": 1.3728325547602505e-06, "loss": 0.003, "step": 37160 }, { "grad_norm": 0.05630118399858475, "learning_rate": 1.3632292123063828e-06, "loss": 0.0042, "step": 37170 }, { "grad_norm": 0.09126687049865723, "learning_rate": 1.3536591126267317e-06, "loss": 0.0042, "step": 37180 }, { "grad_norm": 0.11890853941440582, "learning_rate": 1.3441222622623761e-06, "loss": 0.0057, "step": 37190 }, { "grad_norm": 0.056072358042001724, "learning_rate": 1.3346186677316575e-06, "loss": 0.0041, "step": 37200 }, { "grad_norm": 0.07635647803544998, "learning_rate": 1.3251483355301852e-06, "loss": 0.0052, "step": 37210 }, { "grad_norm": 0.06347796320915222, "learning_rate": 1.315711272130854e-06, "loss": 0.0037, "step": 37220 }, { "grad_norm": 0.06625036150217056, "learning_rate": 1.306307483983804e-06, "loss": 0.0047, "step": 37230 }, { "grad_norm": 0.09879548102617264, "learning_rate": 1.2969369775164275e-06, "loss": 0.0029, "step": 37240 }, { "grad_norm": 0.09617100656032562, "learning_rate": 1.28759975913339e-06, "loss": 0.0038, "step": 37250 }, { "grad_norm": 0.0590745285153389, "learning_rate": 1.2782958352165863e-06, "loss": 0.0033, "step": 37260 }, { "grad_norm": 0.08772985637187958, "learning_rate": 1.2690252121251633e-06, "loss": 0.0029, "step": 37270 }, { "grad_norm": 0.08147455751895905, "learning_rate": 1.2597878961954968e-06, "loss": 0.0045, "step": 37280 }, { "grad_norm": 0.076816126704216, "learning_rate": 1.250583893741214e-06, "loss": 0.0025, "step": 37290 }, { "grad_norm": 0.09545647352933884, "learning_rate": 1.2414132110531607e-06, "loss": 0.004, "step": 37300 }, { "grad_norm": 0.07028541713953018, "learning_rate": 1.2322758543994062e-06, "loss": 0.0027, "step": 37310 }, { "grad_norm": 0.07544303685426712, "learning_rate": 1.2231718300252603e-06, "loss": 0.0059, "step": 37320 }, { "grad_norm": 0.15630003809928894, "learning_rate": 1.214101144153229e-06, "loss": 0.004, "step": 37330 }, { "grad_norm": 0.07923002541065216, "learning_rate": 1.205063802983042e-06, "loss": 0.004, "step": 37340 }, { "grad_norm": 0.07397213578224182, "learning_rate": 1.1960598126916301e-06, "loss": 0.004, "step": 37350 }, { "grad_norm": 0.09886118024587631, "learning_rate": 1.187089179433154e-06, "loss": 0.0042, "step": 37360 }, { "grad_norm": 0.062754787504673, "learning_rate": 1.178151909338937e-06, "loss": 0.0032, "step": 37370 }, { "grad_norm": 0.1447913646697998, "learning_rate": 1.1692480085175317e-06, "loss": 0.0034, "step": 37380 }, { "grad_norm": 0.07154176384210587, "learning_rate": 1.1603774830546587e-06, "loss": 0.0042, "step": 37390 }, { "grad_norm": 0.07056315988302231, "learning_rate": 1.1515403390132573e-06, "loss": 0.0036, "step": 37400 }, { "grad_norm": 0.05143751949071884, "learning_rate": 1.1427365824334124e-06, "loss": 0.0051, "step": 37410 }, { "grad_norm": 0.07302361726760864, "learning_rate": 1.133966219332422e-06, "loss": 0.003, "step": 37420 }, { "grad_norm": 0.10522732883691788, "learning_rate": 1.1252292557047405e-06, "loss": 0.0026, "step": 37430 }, { "grad_norm": 0.07006357610225677, "learning_rate": 1.1165256975219974e-06, "loss": 0.0058, "step": 37440 }, { "grad_norm": 0.08389067649841309, "learning_rate": 1.1078555507330002e-06, "loss": 0.0034, "step": 37450 }, { "grad_norm": 0.09075920283794403, "learning_rate": 1.0992188212637089e-06, "loss": 0.0044, "step": 37460 }, { "grad_norm": 0.08424749225378036, "learning_rate": 1.0906155150172458e-06, "loss": 0.0039, "step": 37470 }, { "grad_norm": 0.08676276355981827, "learning_rate": 1.0820456378738852e-06, "loss": 0.0036, "step": 37480 }, { "grad_norm": 0.06748569756746292, "learning_rate": 1.0735091956910692e-06, "loss": 0.004, "step": 37490 }, { "grad_norm": 0.08301907777786255, "learning_rate": 1.0650061943033695e-06, "loss": 0.0042, "step": 37500 }, { "grad_norm": 0.04612937197089195, "learning_rate": 1.0565366395225041e-06, "loss": 0.0033, "step": 37510 }, { "grad_norm": 0.07952476292848587, "learning_rate": 1.0481005371373366e-06, "loss": 0.0053, "step": 37520 }, { "grad_norm": 0.10494980961084366, "learning_rate": 1.0396978929138602e-06, "loss": 0.0038, "step": 37530 }, { "grad_norm": 0.06368803232908249, "learning_rate": 1.0313287125952032e-06, "loss": 0.0037, "step": 37540 }, { "grad_norm": 0.07405368983745575, "learning_rate": 1.0229930019016288e-06, "loss": 0.0032, "step": 37550 }, { "grad_norm": 0.05406048148870468, "learning_rate": 1.0146907665305183e-06, "loss": 0.0034, "step": 37560 }, { "grad_norm": 0.08474171906709671, "learning_rate": 1.0064220121563551e-06, "loss": 0.0044, "step": 37570 }, { "grad_norm": 0.052151869982481, "learning_rate": 9.98186744430768e-07, "loss": 0.003, "step": 37580 }, { "grad_norm": 0.10639168322086334, "learning_rate": 9.899849689824826e-07, "loss": 0.0031, "step": 37590 }, { "grad_norm": 0.04607314243912697, "learning_rate": 9.818166914173365e-07, "loss": 0.0021, "step": 37600 }, { "grad_norm": 0.10035179555416107, "learning_rate": 9.736819173182642e-07, "loss": 0.0031, "step": 37610 }, { "grad_norm": 0.12667371332645416, "learning_rate": 9.655806522453126e-07, "loss": 0.0026, "step": 37620 }, { "grad_norm": 0.07982123643159866, "learning_rate": 9.575129017356187e-07, "loss": 0.0029, "step": 37630 }, { "grad_norm": 0.06955044716596603, "learning_rate": 9.494786713034109e-07, "loss": 0.0026, "step": 37640 }, { "grad_norm": 0.12150324881076813, "learning_rate": 9.414779664400075e-07, "loss": 0.004, "step": 37650 }, { "grad_norm": 0.05733887106180191, "learning_rate": 9.335107926138342e-07, "loss": 0.0043, "step": 37660 }, { "grad_norm": 0.11610966920852661, "learning_rate": 9.255771552703519e-07, "loss": 0.0049, "step": 37670 }, { "grad_norm": 0.08617644757032394, "learning_rate": 9.176770598321449e-07, "loss": 0.0054, "step": 37680 }, { "grad_norm": 0.08912172168493271, "learning_rate": 9.098105116988498e-07, "loss": 0.0042, "step": 37690 }, { "grad_norm": 0.06277260184288025, "learning_rate": 9.019775162471767e-07, "loss": 0.0027, "step": 37700 }, { "grad_norm": 0.05050168186426163, "learning_rate": 8.94178078830904e-07, "loss": 0.0033, "step": 37710 }, { "grad_norm": 0.05717487633228302, "learning_rate": 8.864122047808731e-07, "loss": 0.0028, "step": 37720 }, { "grad_norm": 0.055931463837623596, "learning_rate": 8.786798994049882e-07, "loss": 0.0028, "step": 37730 }, { "grad_norm": 0.07537644356489182, "learning_rate": 8.709811679882052e-07, "loss": 0.0033, "step": 37740 }, { "grad_norm": 0.07589969784021378, "learning_rate": 8.633160157925313e-07, "loss": 0.0033, "step": 37750 }, { "grad_norm": 0.075672447681427, "learning_rate": 8.556844480570369e-07, "loss": 0.0032, "step": 37760 }, { "grad_norm": 0.08743893355131149, "learning_rate": 8.480864699978164e-07, "loss": 0.0027, "step": 37770 }, { "grad_norm": 0.06871683150529861, "learning_rate": 8.405220868080154e-07, "loss": 0.0048, "step": 37780 }, { "grad_norm": 0.08618547022342682, "learning_rate": 8.329913036578263e-07, "loss": 0.0023, "step": 37790 }, { "grad_norm": 0.11018714308738708, "learning_rate": 8.254941256944593e-07, "loss": 0.0031, "step": 37800 }, { "grad_norm": 0.05905044823884964, "learning_rate": 8.18030558042171e-07, "loss": 0.0036, "step": 37810 }, { "grad_norm": 0.16185835003852844, "learning_rate": 8.106006058022419e-07, "loss": 0.0035, "step": 37820 }, { "grad_norm": 0.104793019592762, "learning_rate": 8.03204274052971e-07, "loss": 0.0025, "step": 37830 }, { "grad_norm": 0.10641723871231079, "learning_rate": 7.958415678496756e-07, "loss": 0.0038, "step": 37840 }, { "grad_norm": 0.07891490310430527, "learning_rate": 7.885124922247022e-07, "loss": 0.004, "step": 37850 }, { "grad_norm": 0.059117503464221954, "learning_rate": 7.812170521874051e-07, "loss": 0.0029, "step": 37860 }, { "grad_norm": 0.0652887299656868, "learning_rate": 7.739552527241456e-07, "loss": 0.0024, "step": 37870 }, { "grad_norm": 0.09077727794647217, "learning_rate": 7.667270987983033e-07, "loss": 0.0031, "step": 37880 }, { "grad_norm": 0.06163763254880905, "learning_rate": 7.595325953502375e-07, "loss": 0.0029, "step": 37890 }, { "grad_norm": 0.05790649354457855, "learning_rate": 7.523717472973368e-07, "loss": 0.0025, "step": 37900 }, { "grad_norm": 0.06148087978363037, "learning_rate": 7.452445595339696e-07, "loss": 0.0029, "step": 37910 }, { "grad_norm": 0.09258551150560379, "learning_rate": 7.381510369315059e-07, "loss": 0.0028, "step": 37920 }, { "grad_norm": 0.09183194488286972, "learning_rate": 7.310911843382895e-07, "loss": 0.0032, "step": 37930 }, { "grad_norm": 0.06122231483459473, "learning_rate": 7.240650065796716e-07, "loss": 0.0025, "step": 37940 }, { "grad_norm": 0.06945431977510452, "learning_rate": 7.170725084579777e-07, "loss": 0.003, "step": 37950 }, { "grad_norm": 0.04544636979699135, "learning_rate": 7.101136947525233e-07, "loss": 0.0033, "step": 37960 }, { "grad_norm": 0.06141199544072151, "learning_rate": 7.031885702195707e-07, "loss": 0.0034, "step": 37970 }, { "grad_norm": 0.07863331586122513, "learning_rate": 6.962971395924e-07, "loss": 0.0033, "step": 37980 }, { "grad_norm": 0.03558903932571411, "learning_rate": 6.894394075812217e-07, "loss": 0.0024, "step": 37990 }, { "grad_norm": 0.08458532392978668, "learning_rate": 6.826153788732359e-07, "loss": 0.0048, "step": 38000 }, { "grad_norm": 0.06962225586175919, "learning_rate": 6.758250581326009e-07, "loss": 0.0033, "step": 38010 }, { "grad_norm": 0.09874125570058823, "learning_rate": 6.690684500004485e-07, "loss": 0.0027, "step": 38020 }, { "grad_norm": 0.06569092720746994, "learning_rate": 6.623455590948458e-07, "loss": 0.0025, "step": 38030 }, { "grad_norm": 0.04067116230726242, "learning_rate": 6.556563900108226e-07, "loss": 0.003, "step": 38040 }, { "grad_norm": 0.11610128730535507, "learning_rate": 6.490009473203662e-07, "loss": 0.0039, "step": 38050 }, { "grad_norm": 0.09680502116680145, "learning_rate": 6.423792355724156e-07, "loss": 0.0032, "step": 38060 }, { "grad_norm": 0.04880959540605545, "learning_rate": 6.35791259292845e-07, "loss": 0.0024, "step": 38070 }, { "grad_norm": 0.11094807833433151, "learning_rate": 6.292370229844691e-07, "loss": 0.0039, "step": 38080 }, { "grad_norm": 0.07232040911912918, "learning_rate": 6.227165311270488e-07, "loss": 0.0046, "step": 38090 }, { "grad_norm": 0.12905238568782806, "learning_rate": 6.16229788177275e-07, "loss": 0.005, "step": 38100 }, { "grad_norm": 0.06562509387731552, "learning_rate": 6.097767985687841e-07, "loss": 0.003, "step": 38110 }, { "grad_norm": 0.062100883573293686, "learning_rate": 6.033575667121317e-07, "loss": 0.0035, "step": 38120 }, { "grad_norm": 0.11078928411006927, "learning_rate": 5.969720969947967e-07, "loss": 0.0049, "step": 38130 }, { "grad_norm": 0.06813259422779083, "learning_rate": 5.906203937811883e-07, "loss": 0.0024, "step": 38140 }, { "grad_norm": 0.11495836079120636, "learning_rate": 5.843024614126447e-07, "loss": 0.0027, "step": 38150 }, { "grad_norm": 0.0670345276594162, "learning_rate": 5.780183042074061e-07, "loss": 0.0034, "step": 38160 }, { "grad_norm": 0.1139722689986229, "learning_rate": 5.717679264606312e-07, "loss": 0.0045, "step": 38170 }, { "grad_norm": 0.03557564690709114, "learning_rate": 5.655513324444028e-07, "loss": 0.0024, "step": 38180 }, { "grad_norm": 0.12123450636863708, "learning_rate": 5.593685264076942e-07, "loss": 0.0035, "step": 38190 }, { "grad_norm": 0.0694769024848938, "learning_rate": 5.532195125764029e-07, "loss": 0.0029, "step": 38200 }, { "grad_norm": 0.058396074920892715, "learning_rate": 5.471042951533167e-07, "loss": 0.0035, "step": 38210 }, { "grad_norm": 0.09863200783729553, "learning_rate": 5.410228783181371e-07, "loss": 0.0033, "step": 38220 }, { "grad_norm": 0.052933406084775925, "learning_rate": 5.349752662274444e-07, "loss": 0.004, "step": 38230 }, { "grad_norm": 0.0694415271282196, "learning_rate": 5.289614630147321e-07, "loss": 0.0026, "step": 38240 }, { "grad_norm": 0.07751217484474182, "learning_rate": 5.229814727903737e-07, "loss": 0.0036, "step": 38250 }, { "grad_norm": 0.0738883912563324, "learning_rate": 5.170352996416439e-07, "loss": 0.0032, "step": 38260 }, { "grad_norm": 0.08964871615171432, "learning_rate": 5.111229476326918e-07, "loss": 0.0028, "step": 38270 }, { "grad_norm": 0.10054054111242294, "learning_rate": 5.052444208045459e-07, "loss": 0.0031, "step": 38280 }, { "grad_norm": 0.08532460778951645, "learning_rate": 4.993997231751312e-07, "loss": 0.0034, "step": 38290 }, { "grad_norm": 0.07911572605371475, "learning_rate": 4.935888587392468e-07, "loss": 0.003, "step": 38300 }, { "grad_norm": 0.09945521503686905, "learning_rate": 4.878118314685598e-07, "loss": 0.0044, "step": 38310 }, { "grad_norm": 0.2051452398300171, "learning_rate": 4.820686453116119e-07, "loss": 0.0032, "step": 38320 }, { "grad_norm": 0.07892810553312302, "learning_rate": 4.763593041938186e-07, "loss": 0.0039, "step": 38330 }, { "grad_norm": 0.040404852479696274, "learning_rate": 4.706838120174584e-07, "loss": 0.0059, "step": 38340 }, { "grad_norm": 0.0686190277338028, "learning_rate": 4.650421726616838e-07, "loss": 0.0031, "step": 38350 }, { "grad_norm": 0.10974720865488052, "learning_rate": 4.594343899824993e-07, "loss": 0.0031, "step": 38360 }, { "grad_norm": 0.1328694075345993, "learning_rate": 4.538604678127667e-07, "loss": 0.004, "step": 38370 }, { "grad_norm": 0.14834532141685486, "learning_rate": 4.483204099622107e-07, "loss": 0.0035, "step": 38380 }, { "grad_norm": 0.10017460584640503, "learning_rate": 4.4281422021740793e-07, "loss": 0.0028, "step": 38390 }, { "grad_norm": 0.04500976577401161, "learning_rate": 4.373419023417924e-07, "loss": 0.0025, "step": 38400 }, { "grad_norm": 0.10372109711170197, "learning_rate": 4.319034600756389e-07, "loss": 0.0051, "step": 38410 }, { "grad_norm": 0.061886854469776154, "learning_rate": 4.264988971360739e-07, "loss": 0.0041, "step": 38420 }, { "grad_norm": 0.07963667809963226, "learning_rate": 4.211282172170594e-07, "loss": 0.0035, "step": 38430 }, { "grad_norm": 0.08038107305765152, "learning_rate": 4.1579142398940894e-07, "loss": 0.0029, "step": 38440 }, { "grad_norm": 0.08127538114786148, "learning_rate": 4.1048852110076585e-07, "loss": 0.0029, "step": 38450 }, { "grad_norm": 0.09031713008880615, "learning_rate": 4.052195121756197e-07, "loss": 0.0043, "step": 38460 }, { "grad_norm": 0.07867104560136795, "learning_rate": 3.9998440081528974e-07, "loss": 0.0037, "step": 38470 }, { "grad_norm": 0.07797030359506607, "learning_rate": 3.9478319059791936e-07, "loss": 0.0032, "step": 38480 }, { "grad_norm": 0.06176122650504112, "learning_rate": 3.8961588507848143e-07, "loss": 0.0036, "step": 38490 }, { "grad_norm": 0.03810844197869301, "learning_rate": 3.844824877887954e-07, "loss": 0.0051, "step": 38500 }, { "grad_norm": 0.050001922994852066, "learning_rate": 3.7938300223748224e-07, "loss": 0.004, "step": 38510 }, { "grad_norm": 0.05353359878063202, "learning_rate": 3.743174319099929e-07, "loss": 0.0024, "step": 38520 }, { "grad_norm": 0.08626502007246017, "learning_rate": 3.6928578026859673e-07, "loss": 0.0029, "step": 38530 }, { "grad_norm": 0.09743721038103104, "learning_rate": 3.642880507523816e-07, "loss": 0.0046, "step": 38540 }, { "grad_norm": 0.05433259904384613, "learning_rate": 3.5932424677724285e-07, "loss": 0.0051, "step": 38550 }, { "grad_norm": 0.06863778829574585, "learning_rate": 3.5439437173589996e-07, "loss": 0.0028, "step": 38560 }, { "grad_norm": 0.04673503711819649, "learning_rate": 3.4949842899787423e-07, "loss": 0.0032, "step": 38570 }, { "grad_norm": 0.08241458237171173, "learning_rate": 3.446364219094944e-07, "loss": 0.0032, "step": 38580 }, { "grad_norm": 0.10094642639160156, "learning_rate": 3.3980835379389675e-07, "loss": 0.004, "step": 38590 }, { "grad_norm": 0.04557700827717781, "learning_rate": 3.350142279510249e-07, "loss": 0.0036, "step": 38600 }, { "grad_norm": 0.06008422374725342, "learning_rate": 3.3025404765761324e-07, "loss": 0.0035, "step": 38610 }, { "grad_norm": 0.07611653208732605, "learning_rate": 3.255278161672037e-07, "loss": 0.0049, "step": 38620 }, { "grad_norm": 0.05150912329554558, "learning_rate": 3.20835536710129e-07, "loss": 0.003, "step": 38630 }, { "grad_norm": 0.09409184753894806, "learning_rate": 3.1617721249351804e-07, "loss": 0.003, "step": 38640 }, { "grad_norm": 0.08071240782737732, "learning_rate": 3.115528467012963e-07, "loss": 0.0036, "step": 38650 }, { "grad_norm": 0.10707080364227295, "learning_rate": 3.0696244249416305e-07, "loss": 0.0035, "step": 38660 }, { "grad_norm": 0.10662992298603058, "learning_rate": 3.024060030096254e-07, "loss": 0.0026, "step": 38670 }, { "grad_norm": 0.059756457805633545, "learning_rate": 2.9788353136196435e-07, "loss": 0.0039, "step": 38680 }, { "grad_norm": 0.08757945150136948, "learning_rate": 2.9339503064224615e-07, "loss": 0.0034, "step": 38690 }, { "grad_norm": 0.05311063304543495, "learning_rate": 2.889405039183224e-07, "loss": 0.0028, "step": 38700 }, { "grad_norm": 0.07136080414056778, "learning_rate": 2.8451995423481315e-07, "loss": 0.0037, "step": 38710 }, { "grad_norm": 0.0699232891201973, "learning_rate": 2.801333846131293e-07, "loss": 0.0042, "step": 38720 }, { "grad_norm": 0.13638681173324585, "learning_rate": 2.757807980514504e-07, "loss": 0.0047, "step": 38730 }, { "grad_norm": 0.10786887258291245, "learning_rate": 2.714621975247189e-07, "loss": 0.0037, "step": 38740 }, { "grad_norm": 0.11806883662939072, "learning_rate": 2.671775859846626e-07, "loss": 0.004, "step": 38750 }, { "grad_norm": 0.10497433692216873, "learning_rate": 2.629269663597722e-07, "loss": 0.0026, "step": 38760 }, { "grad_norm": 0.14455144107341766, "learning_rate": 2.587103415553127e-07, "loss": 0.0043, "step": 38770 }, { "grad_norm": 0.09129461646080017, "learning_rate": 2.545277144533009e-07, "loss": 0.0031, "step": 38780 }, { "grad_norm": 0.1058952659368515, "learning_rate": 2.5037908791252227e-07, "loss": 0.003, "step": 38790 }, { "grad_norm": 0.07399415224790573, "learning_rate": 2.462644647685308e-07, "loss": 0.0033, "step": 38800 }, { "grad_norm": 0.0969686508178711, "learning_rate": 2.421838478336269e-07, "loss": 0.0043, "step": 38810 }, { "grad_norm": 0.0697701945900917, "learning_rate": 2.3813723989687398e-07, "loss": 0.0039, "step": 38820 }, { "grad_norm": 0.0925728902220726, "learning_rate": 2.3412464372409293e-07, "loss": 0.0033, "step": 38830 }, { "grad_norm": 0.07318971306085587, "learning_rate": 2.301460620578455e-07, "loss": 0.0037, "step": 38840 }, { "grad_norm": 0.07618126273155212, "learning_rate": 2.262014976174731e-07, "loss": 0.0034, "step": 38850 }, { "grad_norm": 0.06806603074073792, "learning_rate": 2.2229095309903025e-07, "loss": 0.0057, "step": 38860 }, { "grad_norm": 0.1509149968624115, "learning_rate": 2.184144311753511e-07, "loss": 0.0033, "step": 38870 }, { "grad_norm": 0.10756536573171616, "learning_rate": 2.14571934495994e-07, "loss": 0.0035, "step": 38880 }, { "grad_norm": 0.09961532056331635, "learning_rate": 2.1076346568726923e-07, "loss": 0.0034, "step": 38890 }, { "grad_norm": 0.08524517714977264, "learning_rate": 2.0698902735223346e-07, "loss": 0.0041, "step": 38900 }, { "grad_norm": 0.03991768881678581, "learning_rate": 2.0324862207067863e-07, "loss": 0.0034, "step": 38910 }, { "grad_norm": 0.086479052901268, "learning_rate": 1.9954225239913195e-07, "loss": 0.0031, "step": 38920 }, { "grad_norm": 0.04087769612669945, "learning_rate": 1.9586992087087252e-07, "loss": 0.0022, "step": 38930 }, { "grad_norm": 0.030903486534953117, "learning_rate": 1.9223162999588705e-07, "loss": 0.003, "step": 38940 }, { "grad_norm": 0.10157950222492218, "learning_rate": 1.886273822609308e-07, "loss": 0.005, "step": 38950 }, { "grad_norm": 0.09723910689353943, "learning_rate": 1.8505718012946094e-07, "loss": 0.003, "step": 38960 }, { "grad_norm": 0.10781238973140717, "learning_rate": 1.8152102604168664e-07, "loss": 0.0033, "step": 38970 }, { "grad_norm": 0.07347603142261505, "learning_rate": 1.7801892241452457e-07, "loss": 0.0031, "step": 38980 }, { "grad_norm": 0.03443975746631622, "learning_rate": 1.745508716416433e-07, "loss": 0.0044, "step": 38990 }, { "grad_norm": 0.08395753055810928, "learning_rate": 1.7111687609341342e-07, "loss": 0.003, "step": 39000 }, { "grad_norm": 0.08216559886932373, "learning_rate": 1.6771693811693523e-07, "loss": 0.0046, "step": 39010 }, { "grad_norm": 0.07050071656703949, "learning_rate": 1.643510600360443e-07, "loss": 0.0029, "step": 39020 }, { "grad_norm": 0.04319359362125397, "learning_rate": 1.610192441512781e-07, "loss": 0.0037, "step": 39030 }, { "grad_norm": 0.14616888761520386, "learning_rate": 1.5772149273990956e-07, "loss": 0.0046, "step": 39040 }, { "grad_norm": 0.10152345150709152, "learning_rate": 1.5445780805591337e-07, "loss": 0.0025, "step": 39050 }, { "grad_norm": 0.06809142976999283, "learning_rate": 1.5122819232998852e-07, "loss": 0.0043, "step": 39060 }, { "grad_norm": 0.10577453672885895, "learning_rate": 1.4803264776955262e-07, "loss": 0.0044, "step": 39070 }, { "grad_norm": 0.07055288553237915, "learning_rate": 1.448711765587196e-07, "loss": 0.0022, "step": 39080 }, { "grad_norm": 0.06490276753902435, "learning_rate": 1.4174378085833885e-07, "loss": 0.0035, "step": 39090 }, { "grad_norm": 0.051440611481666565, "learning_rate": 1.3865046280594486e-07, "loss": 0.0036, "step": 39100 }, { "grad_norm": 0.06634476035833359, "learning_rate": 1.3559122451579643e-07, "loss": 0.0024, "step": 39110 }, { "grad_norm": 0.09101849794387817, "learning_rate": 1.3256606807884874e-07, "loss": 0.0045, "step": 39120 }, { "grad_norm": 0.05355556681752205, "learning_rate": 1.2957499556277563e-07, "loss": 0.0046, "step": 39130 }, { "grad_norm": 0.09740503877401352, "learning_rate": 1.266180090119473e-07, "loss": 0.0038, "step": 39140 }, { "grad_norm": 0.053655918687582016, "learning_rate": 1.2369511044742488e-07, "loss": 0.0026, "step": 39150 }, { "grad_norm": 0.08844219893217087, "learning_rate": 1.2080630186698805e-07, "loss": 0.003, "step": 39160 }, { "grad_norm": 0.02672494947910309, "learning_rate": 1.1795158524511851e-07, "loss": 0.0025, "step": 39170 }, { "grad_norm": 0.0488639660179615, "learning_rate": 1.1513096253297773e-07, "loss": 0.0025, "step": 39180 }, { "grad_norm": 0.06916936486959457, "learning_rate": 1.1234443565843466e-07, "loss": 0.004, "step": 39190 }, { "grad_norm": 0.08156905323266983, "learning_rate": 1.0959200652606028e-07, "loss": 0.0028, "step": 39200 }, { "grad_norm": 0.1077633798122406, "learning_rate": 1.0687367701711082e-07, "loss": 0.0035, "step": 39210 }, { "grad_norm": 0.07118070125579834, "learning_rate": 1.0418944898953898e-07, "loss": 0.0033, "step": 39220 }, { "grad_norm": 0.05485805124044418, "learning_rate": 1.0153932427798829e-07, "loss": 0.0036, "step": 39230 }, { "grad_norm": 0.1757539063692093, "learning_rate": 9.892330469379874e-08, "loss": 0.0059, "step": 39240 }, { "grad_norm": 0.049438755959272385, "learning_rate": 9.634139202498449e-08, "loss": 0.0022, "step": 39250 }, { "grad_norm": 0.0617816261947155, "learning_rate": 9.379358803627281e-08, "loss": 0.0023, "step": 39260 }, { "grad_norm": 0.11222995072603226, "learning_rate": 9.12798944690485e-08, "loss": 0.0025, "step": 39270 }, { "grad_norm": 0.14186906814575195, "learning_rate": 8.880031304141501e-08, "loss": 0.0055, "step": 39280 }, { "grad_norm": 0.06928688287734985, "learning_rate": 8.635484544812777e-08, "loss": 0.0046, "step": 39290 }, { "grad_norm": 0.08903535455465317, "learning_rate": 8.394349336064422e-08, "loss": 0.0031, "step": 39300 }, { "grad_norm": 0.04605250060558319, "learning_rate": 8.156625842710153e-08, "loss": 0.0038, "step": 39310 }, { "grad_norm": 0.16185806691646576, "learning_rate": 7.92231422723222e-08, "loss": 0.0038, "step": 39320 }, { "grad_norm": 0.054133184254169464, "learning_rate": 7.691414649780293e-08, "loss": 0.0044, "step": 39330 }, { "grad_norm": 0.04841142147779465, "learning_rate": 7.463927268171467e-08, "loss": 0.0023, "step": 39340 }, { "grad_norm": 0.11202026158571243, "learning_rate": 7.239852237891365e-08, "loss": 0.0043, "step": 39350 }, { "grad_norm": 0.03682930767536163, "learning_rate": 7.019189712093588e-08, "loss": 0.0043, "step": 39360 }, { "grad_norm": 0.08698239922523499, "learning_rate": 6.801939841598603e-08, "loss": 0.0027, "step": 39370 }, { "grad_norm": 0.08431082218885422, "learning_rate": 6.588102774894856e-08, "loss": 0.0031, "step": 39380 }, { "grad_norm": 0.04179135337471962, "learning_rate": 6.377678658137653e-08, "loss": 0.0026, "step": 39390 }, { "grad_norm": 0.07070119678974152, "learning_rate": 6.17066763514973e-08, "loss": 0.0029, "step": 39400 }, { "grad_norm": 0.0768052339553833, "learning_rate": 5.967069847421792e-08, "loss": 0.003, "step": 39410 }, { "grad_norm": 0.08356370031833649, "learning_rate": 5.7668854341108583e-08, "loss": 0.0035, "step": 39420 }, { "grad_norm": 0.111392080783844, "learning_rate": 5.570114532040815e-08, "loss": 0.0042, "step": 39430 }, { "grad_norm": 0.05680064857006073, "learning_rate": 5.3767572757029706e-08, "loss": 0.0046, "step": 39440 }, { "grad_norm": 0.03686903417110443, "learning_rate": 5.186813797254386e-08, "loss": 0.0038, "step": 39450 }, { "grad_norm": 0.11734307557344437, "learning_rate": 5.000284226521213e-08, "loss": 0.003, "step": 39460 }, { "grad_norm": 0.0724804475903511, "learning_rate": 4.8171686909931346e-08, "loss": 0.0028, "step": 39470 }, { "grad_norm": 0.06303409487009048, "learning_rate": 4.637467315828925e-08, "loss": 0.0027, "step": 39480 }, { "grad_norm": 0.24899189174175262, "learning_rate": 4.461180223852002e-08, "loss": 0.0037, "step": 39490 }, { "grad_norm": 0.12167216092348099, "learning_rate": 4.288307535553204e-08, "loss": 0.0029, "step": 39500 }, { "grad_norm": 0.12809698283672333, "learning_rate": 4.1188493690896836e-08, "loss": 0.0062, "step": 39510 }, { "grad_norm": 0.12647272646427155, "learning_rate": 3.952805840284346e-08, "loss": 0.0044, "step": 39520 }, { "grad_norm": 0.03911247476935387, "learning_rate": 3.790177062625855e-08, "loss": 0.0027, "step": 39530 }, { "grad_norm": 0.0796186625957489, "learning_rate": 3.6309631472702946e-08, "loss": 0.0032, "step": 39540 }, { "grad_norm": 0.06810999661684036, "learning_rate": 3.475164203038395e-08, "loss": 0.0026, "step": 39550 }, { "grad_norm": 0.07070747762918472, "learning_rate": 3.3227803364171974e-08, "loss": 0.0026, "step": 39560 }, { "grad_norm": 0.07571015506982803, "learning_rate": 3.173811651560055e-08, "loss": 0.0038, "step": 39570 }, { "grad_norm": 0.0941091775894165, "learning_rate": 3.0282582502855204e-08, "loss": 0.0047, "step": 39580 }, { "grad_norm": 0.09818371385335922, "learning_rate": 2.8861202320773496e-08, "loss": 0.0039, "step": 39590 }, { "grad_norm": 0.09372760355472565, "learning_rate": 2.7473976940867175e-08, "loss": 0.0028, "step": 39600 }, { "grad_norm": 0.0536353662610054, "learning_rate": 2.6120907311277808e-08, "loss": 0.004, "step": 39610 }, { "grad_norm": 0.10158183425664902, "learning_rate": 2.480199435682673e-08, "loss": 0.0024, "step": 39620 }, { "grad_norm": 0.08852094411849976, "learning_rate": 2.3517238978976174e-08, "loss": 0.0032, "step": 39630 }, { "grad_norm": 0.10440581291913986, "learning_rate": 2.2266642055834842e-08, "loss": 0.0031, "step": 39640 }, { "grad_norm": 0.06409130990505219, "learning_rate": 2.105020444219119e-08, "loss": 0.0024, "step": 39650 }, { "grad_norm": 0.05289595201611519, "learning_rate": 1.9867926969452388e-08, "loss": 0.0032, "step": 39660 }, { "grad_norm": 0.07127179950475693, "learning_rate": 1.8719810445705366e-08, "loss": 0.0037, "step": 39670 }, { "grad_norm": 0.07064493000507355, "learning_rate": 1.760585565567241e-08, "loss": 0.0034, "step": 39680 }, { "grad_norm": 0.06511950492858887, "learning_rate": 1.6526063360727816e-08, "loss": 0.0029, "step": 39690 }, { "grad_norm": 0.05327765643596649, "learning_rate": 1.5480434298908997e-08, "loss": 0.0035, "step": 39700 }, { "grad_norm": 0.05886894837021828, "learning_rate": 1.4468969184883164e-08, "loss": 0.0041, "step": 39710 }, { "grad_norm": 0.09314825385808945, "learning_rate": 1.3491668709980643e-08, "loss": 0.0032, "step": 39720 }, { "grad_norm": 0.13107387721538544, "learning_rate": 1.254853354217822e-08, "loss": 0.0036, "step": 39730 }, { "grad_norm": 0.1660250872373581, "learning_rate": 1.1639564326099139e-08, "loss": 0.0061, "step": 39740 }, { "grad_norm": 0.045603349804878235, "learning_rate": 1.0764761683018653e-08, "loss": 0.0022, "step": 39750 }, { "grad_norm": 0.1257924735546112, "learning_rate": 9.924126210852925e-09, "loss": 0.0035, "step": 39760 }, { "grad_norm": 0.06126097962260246, "learning_rate": 9.117658484170122e-09, "loss": 0.003, "step": 39770 }, { "grad_norm": 0.060451049357652664, "learning_rate": 8.345359054179324e-09, "loss": 0.0036, "step": 39780 }, { "grad_norm": 0.06815160810947418, "learning_rate": 7.60722844874162e-09, "loss": 0.0049, "step": 39790 }, { "grad_norm": 0.06619760394096375, "learning_rate": 6.903267172364558e-09, "loss": 0.0034, "step": 39800 }, { "grad_norm": 0.05004052072763443, "learning_rate": 6.23347570619659e-09, "loss": 0.0035, "step": 39810 }, { "grad_norm": 0.08709901571273804, "learning_rate": 5.59785450803818e-09, "loss": 0.0048, "step": 39820 }, { "grad_norm": 0.07261746376752853, "learning_rate": 4.996404012325151e-09, "loss": 0.0029, "step": 39830 }, { "grad_norm": 0.09909741580486298, "learning_rate": 4.429124630145331e-09, "loss": 0.0034, "step": 39840 }, { "grad_norm": 0.1185336485505104, "learning_rate": 3.89601674922746e-09, "loss": 0.0044, "step": 39850 }, { "grad_norm": 0.06628697365522385, "learning_rate": 3.3970807339467336e-09, "loss": 0.0031, "step": 39860 }, { "grad_norm": 0.056499045342206955, "learning_rate": 2.9323169253248073e-09, "loss": 0.0027, "step": 39870 }, { "grad_norm": 0.09278372675180435, "learning_rate": 2.501725641018693e-09, "loss": 0.0028, "step": 39880 }, { "grad_norm": 0.11156842112541199, "learning_rate": 2.105307175331861e-09, "loss": 0.0045, "step": 39890 }, { "grad_norm": 0.06887213885784149, "learning_rate": 1.7430617992197918e-09, "loss": 0.0047, "step": 39900 }, { "grad_norm": 0.08170315623283386, "learning_rate": 1.4149897602677708e-09, "loss": 0.0032, "step": 39910 }, { "grad_norm": 0.06162683293223381, "learning_rate": 1.1210912827130937e-09, "loss": 0.0039, "step": 39920 }, { "grad_norm": 0.09886502474546432, "learning_rate": 8.613665674284121e-10, "loss": 0.0051, "step": 39930 }, { "grad_norm": 0.0343267023563385, "learning_rate": 6.358157919383878e-10, "loss": 0.0027, "step": 39940 }, { "grad_norm": 0.04672392085194588, "learning_rate": 4.4443911040303877e-10, "loss": 0.0023, "step": 39950 }, { "grad_norm": 0.10608924180269241, "learning_rate": 2.8723665362329066e-10, "loss": 0.0037, "step": 39960 }, { "grad_norm": 0.10477261245250702, "learning_rate": 1.6420852905207895e-10, "loss": 0.0025, "step": 39970 }, { "grad_norm": 0.042411383241415024, "learning_rate": 7.535482076659328e-11, "loss": 0.0033, "step": 39980 }, { "grad_norm": 0.06973801553249359, "learning_rate": 2.0675589512686443e-11, "loss": 0.0027, "step": 39990 }, { "grad_norm": 0.08920072019100189, "learning_rate": 1.7087265491433358e-13, "loss": 0.0027, "step": 40000 } ], "logging_steps": 10, "max_steps": 40000, "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": 48, "trial_name": null, "trial_params": null }