{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 6.761553805064404, "eval_steps": 500, "global_step": 200000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0003380776902532202, "grad_norm": 5.155960559844971, "learning_rate": 9e-08, "loss": 0.6157, "step": 10 }, { "epoch": 0.0006761553805064404, "grad_norm": 4.684703826904297, "learning_rate": 1.9e-07, "loss": 0.6358, "step": 20 }, { "epoch": 0.0010142330707596605, "grad_norm": 5.175259113311768, "learning_rate": 2.9e-07, "loss": 0.6697, "step": 30 }, { "epoch": 0.0013523107610128808, "grad_norm": 3.368594169616699, "learning_rate": 3.9e-07, "loss": 0.6062, "step": 40 }, { "epoch": 0.001690388451266101, "grad_norm": 4.975178241729736, "learning_rate": 4.9e-07, "loss": 0.6261, "step": 50 }, { "epoch": 0.002028466141519321, "grad_norm": 3.4694535732269287, "learning_rate": 5.9e-07, "loss": 0.6221, "step": 60 }, { "epoch": 0.0023665438317725415, "grad_norm": 3.3969645500183105, "learning_rate": 6.900000000000001e-07, "loss": 0.5891, "step": 70 }, { "epoch": 0.0027046215220257615, "grad_norm": 2.753671884536743, "learning_rate": 7.900000000000002e-07, "loss": 0.5372, "step": 80 }, { "epoch": 0.0030426992122789816, "grad_norm": 2.267770290374756, "learning_rate": 8.900000000000001e-07, "loss": 0.4839, "step": 90 }, { "epoch": 0.003380776902532202, "grad_norm": 1.747541069984436, "learning_rate": 9.9e-07, "loss": 0.4289, "step": 100 }, { "epoch": 0.003718854592785422, "grad_norm": 1.7670830488204956, "learning_rate": 1.0900000000000002e-06, "loss": 0.4067, "step": 110 }, { "epoch": 0.004056932283038642, "grad_norm": 1.2932764291763306, "learning_rate": 1.19e-06, "loss": 0.3606, "step": 120 }, { "epoch": 0.004395009973291862, "grad_norm": 1.119842529296875, "learning_rate": 1.2900000000000001e-06, "loss": 0.3038, "step": 130 }, { "epoch": 0.004733087663545083, "grad_norm": 0.9262852072715759, "learning_rate": 1.39e-06, "loss": 0.2849, "step": 140 }, { "epoch": 0.005071165353798303, "grad_norm": 1.247458815574646, "learning_rate": 1.4900000000000001e-06, "loss": 0.2738, "step": 150 }, { "epoch": 0.005409243044051523, "grad_norm": 0.7871404886245728, "learning_rate": 1.5900000000000002e-06, "loss": 0.2577, "step": 160 }, { "epoch": 0.005747320734304743, "grad_norm": 0.9975180625915527, "learning_rate": 1.69e-06, "loss": 0.2331, "step": 170 }, { "epoch": 0.006085398424557963, "grad_norm": 0.8277950882911682, "learning_rate": 1.79e-06, "loss": 0.2052, "step": 180 }, { "epoch": 0.006423476114811184, "grad_norm": 0.943140983581543, "learning_rate": 1.8900000000000001e-06, "loss": 0.2153, "step": 190 }, { "epoch": 0.006761553805064404, "grad_norm": 1.0434321165084839, "learning_rate": 1.99e-06, "loss": 0.1879, "step": 200 }, { "epoch": 0.007099631495317624, "grad_norm": 1.490017056465149, "learning_rate": 2.09e-06, "loss": 0.1997, "step": 210 }, { "epoch": 0.007437709185570844, "grad_norm": 0.798238217830658, "learning_rate": 2.19e-06, "loss": 0.1759, "step": 220 }, { "epoch": 0.007775786875824064, "grad_norm": 1.128237009048462, "learning_rate": 2.29e-06, "loss": 0.1761, "step": 230 }, { "epoch": 0.008113864566077284, "grad_norm": 0.7243601679801941, "learning_rate": 2.3900000000000004e-06, "loss": 0.1645, "step": 240 }, { "epoch": 0.008451942256330504, "grad_norm": 0.8919597864151001, "learning_rate": 2.49e-06, "loss": 0.1581, "step": 250 }, { "epoch": 0.008790019946583724, "grad_norm": 0.8304742574691772, "learning_rate": 2.59e-06, "loss": 0.1597, "step": 260 }, { "epoch": 0.009128097636836946, "grad_norm": 1.0768707990646362, "learning_rate": 2.69e-06, "loss": 0.1535, "step": 270 }, { "epoch": 0.009466175327090166, "grad_norm": 1.3286269903182983, "learning_rate": 2.7900000000000004e-06, "loss": 0.1557, "step": 280 }, { "epoch": 0.009804253017343386, "grad_norm": 1.332438349723816, "learning_rate": 2.89e-06, "loss": 0.1369, "step": 290 }, { "epoch": 0.010142330707596606, "grad_norm": 0.9580008387565613, "learning_rate": 2.99e-06, "loss": 0.154, "step": 300 }, { "epoch": 0.010480408397849826, "grad_norm": 1.122244954109192, "learning_rate": 3.09e-06, "loss": 0.1436, "step": 310 }, { "epoch": 0.010818486088103046, "grad_norm": 0.8896529674530029, "learning_rate": 3.19e-06, "loss": 0.1449, "step": 320 }, { "epoch": 0.011156563778356266, "grad_norm": 1.542036771774292, "learning_rate": 3.29e-06, "loss": 0.1383, "step": 330 }, { "epoch": 0.011494641468609486, "grad_norm": 1.497359037399292, "learning_rate": 3.39e-06, "loss": 0.1481, "step": 340 }, { "epoch": 0.011832719158862706, "grad_norm": 0.983357846736908, "learning_rate": 3.49e-06, "loss": 0.1329, "step": 350 }, { "epoch": 0.012170796849115926, "grad_norm": 1.1431751251220703, "learning_rate": 3.5900000000000004e-06, "loss": 0.1406, "step": 360 }, { "epoch": 0.012508874539369146, "grad_norm": 0.8942676186561584, "learning_rate": 3.6900000000000002e-06, "loss": 0.1366, "step": 370 }, { "epoch": 0.012846952229622368, "grad_norm": 0.9701046347618103, "learning_rate": 3.7900000000000006e-06, "loss": 0.1361, "step": 380 }, { "epoch": 0.013185029919875588, "grad_norm": 0.8894744515419006, "learning_rate": 3.89e-06, "loss": 0.1347, "step": 390 }, { "epoch": 0.013523107610128808, "grad_norm": 0.8924310207366943, "learning_rate": 3.99e-06, "loss": 0.1365, "step": 400 }, { "epoch": 0.013861185300382028, "grad_norm": 0.9468827843666077, "learning_rate": 4.09e-06, "loss": 0.1336, "step": 410 }, { "epoch": 0.014199262990635248, "grad_norm": 0.7663879990577698, "learning_rate": 4.1900000000000005e-06, "loss": 0.1319, "step": 420 }, { "epoch": 0.014537340680888468, "grad_norm": 0.7574809193611145, "learning_rate": 4.2900000000000004e-06, "loss": 0.1181, "step": 430 }, { "epoch": 0.014875418371141688, "grad_norm": 0.5894230008125305, "learning_rate": 4.39e-06, "loss": 0.1156, "step": 440 }, { "epoch": 0.015213496061394908, "grad_norm": 1.1868999004364014, "learning_rate": 4.49e-06, "loss": 0.1183, "step": 450 }, { "epoch": 0.015551573751648128, "grad_norm": 0.7034958004951477, "learning_rate": 4.590000000000001e-06, "loss": 0.119, "step": 460 }, { "epoch": 0.01588965144190135, "grad_norm": 0.6963277459144592, "learning_rate": 4.69e-06, "loss": 0.1103, "step": 470 }, { "epoch": 0.01622772913215457, "grad_norm": 0.8717629313468933, "learning_rate": 4.79e-06, "loss": 0.1199, "step": 480 }, { "epoch": 0.01656580682240779, "grad_norm": 0.9731774926185608, "learning_rate": 4.89e-06, "loss": 0.1145, "step": 490 }, { "epoch": 0.01690388451266101, "grad_norm": 1.1781518459320068, "learning_rate": 4.9900000000000005e-06, "loss": 0.1193, "step": 500 }, { "epoch": 0.01724196220291423, "grad_norm": 1.2089394330978394, "learning_rate": 5.09e-06, "loss": 0.1174, "step": 510 }, { "epoch": 0.01758003989316745, "grad_norm": 1.4143964052200317, "learning_rate": 5.19e-06, "loss": 0.1197, "step": 520 }, { "epoch": 0.01791811758342067, "grad_norm": 0.982684314250946, "learning_rate": 5.29e-06, "loss": 0.1226, "step": 530 }, { "epoch": 0.018256195273673892, "grad_norm": 1.5373514890670776, "learning_rate": 5.390000000000001e-06, "loss": 0.1182, "step": 540 }, { "epoch": 0.01859427296392711, "grad_norm": 1.5738344192504883, "learning_rate": 5.49e-06, "loss": 0.1142, "step": 550 }, { "epoch": 0.018932350654180332, "grad_norm": 1.1845108270645142, "learning_rate": 5.59e-06, "loss": 0.1169, "step": 560 }, { "epoch": 0.01927042834443355, "grad_norm": 0.8683966994285583, "learning_rate": 5.690000000000001e-06, "loss": 0.1245, "step": 570 }, { "epoch": 0.019608506034686772, "grad_norm": 0.7065423130989075, "learning_rate": 5.7900000000000005e-06, "loss": 0.1137, "step": 580 }, { "epoch": 0.01994658372493999, "grad_norm": 0.845619797706604, "learning_rate": 5.89e-06, "loss": 0.1045, "step": 590 }, { "epoch": 0.020284661415193212, "grad_norm": 0.6912538409233093, "learning_rate": 5.99e-06, "loss": 0.1148, "step": 600 }, { "epoch": 0.02062273910544643, "grad_norm": 0.8908045291900635, "learning_rate": 6.090000000000001e-06, "loss": 0.1084, "step": 610 }, { "epoch": 0.020960816795699652, "grad_norm": 1.415392518043518, "learning_rate": 6.19e-06, "loss": 0.1203, "step": 620 }, { "epoch": 0.02129889448595287, "grad_norm": 0.8393387198448181, "learning_rate": 6.29e-06, "loss": 0.117, "step": 630 }, { "epoch": 0.021636972176206092, "grad_norm": 0.7606134414672852, "learning_rate": 6.39e-06, "loss": 0.1044, "step": 640 }, { "epoch": 0.021975049866459314, "grad_norm": 2.0393807888031006, "learning_rate": 6.4900000000000005e-06, "loss": 0.1107, "step": 650 }, { "epoch": 0.022313127556712532, "grad_norm": 0.8991869688034058, "learning_rate": 6.5900000000000004e-06, "loss": 0.1021, "step": 660 }, { "epoch": 0.022651205246965754, "grad_norm": 0.691292941570282, "learning_rate": 6.69e-06, "loss": 0.0992, "step": 670 }, { "epoch": 0.022989282937218972, "grad_norm": 0.9650527238845825, "learning_rate": 6.79e-06, "loss": 0.0982, "step": 680 }, { "epoch": 0.023327360627472194, "grad_norm": 0.7697871327400208, "learning_rate": 6.890000000000001e-06, "loss": 0.092, "step": 690 }, { "epoch": 0.023665438317725412, "grad_norm": 0.8185405135154724, "learning_rate": 6.990000000000001e-06, "loss": 0.0884, "step": 700 }, { "epoch": 0.024003516007978634, "grad_norm": 0.8192614912986755, "learning_rate": 7.090000000000001e-06, "loss": 0.0989, "step": 710 }, { "epoch": 0.024341593698231852, "grad_norm": 1.1433123350143433, "learning_rate": 7.190000000000001e-06, "loss": 0.0991, "step": 720 }, { "epoch": 0.024679671388485074, "grad_norm": 0.796172022819519, "learning_rate": 7.290000000000001e-06, "loss": 0.1045, "step": 730 }, { "epoch": 0.025017749078738293, "grad_norm": 0.5429674983024597, "learning_rate": 7.3899999999999995e-06, "loss": 0.1001, "step": 740 }, { "epoch": 0.025355826768991514, "grad_norm": 1.0549572706222534, "learning_rate": 7.4899999999999994e-06, "loss": 0.1011, "step": 750 }, { "epoch": 0.025693904459244736, "grad_norm": 0.8394684791564941, "learning_rate": 7.59e-06, "loss": 0.0956, "step": 760 }, { "epoch": 0.026031982149497954, "grad_norm": 0.8143083453178406, "learning_rate": 7.69e-06, "loss": 0.1032, "step": 770 }, { "epoch": 0.026370059839751176, "grad_norm": 0.865418553352356, "learning_rate": 7.79e-06, "loss": 0.1013, "step": 780 }, { "epoch": 0.026708137530004394, "grad_norm": 0.7387815117835999, "learning_rate": 7.89e-06, "loss": 0.0991, "step": 790 }, { "epoch": 0.027046215220257616, "grad_norm": 0.6031355261802673, "learning_rate": 7.99e-06, "loss": 0.0926, "step": 800 }, { "epoch": 0.027384292910510834, "grad_norm": 0.695909857749939, "learning_rate": 8.09e-06, "loss": 0.0881, "step": 810 }, { "epoch": 0.027722370600764056, "grad_norm": 0.6291481256484985, "learning_rate": 8.190000000000001e-06, "loss": 0.0893, "step": 820 }, { "epoch": 0.028060448291017275, "grad_norm": 0.6022233963012695, "learning_rate": 8.29e-06, "loss": 0.0957, "step": 830 }, { "epoch": 0.028398525981270496, "grad_norm": 0.8651807904243469, "learning_rate": 8.390000000000001e-06, "loss": 0.102, "step": 840 }, { "epoch": 0.028736603671523715, "grad_norm": 0.8549692630767822, "learning_rate": 8.49e-06, "loss": 0.1002, "step": 850 }, { "epoch": 0.029074681361776936, "grad_norm": 0.9032987356185913, "learning_rate": 8.59e-06, "loss": 0.1008, "step": 860 }, { "epoch": 0.029412759052030158, "grad_norm": 0.8818564414978027, "learning_rate": 8.690000000000002e-06, "loss": 0.0956, "step": 870 }, { "epoch": 0.029750836742283376, "grad_norm": 0.7996283173561096, "learning_rate": 8.79e-06, "loss": 0.0934, "step": 880 }, { "epoch": 0.030088914432536598, "grad_norm": 0.6142864227294922, "learning_rate": 8.890000000000001e-06, "loss": 0.0914, "step": 890 }, { "epoch": 0.030426992122789816, "grad_norm": 1.2803574800491333, "learning_rate": 8.99e-06, "loss": 0.0924, "step": 900 }, { "epoch": 0.030765069813043038, "grad_norm": 0.6340155005455017, "learning_rate": 9.09e-06, "loss": 0.0886, "step": 910 }, { "epoch": 0.031103147503296256, "grad_norm": 0.8250855803489685, "learning_rate": 9.19e-06, "loss": 0.0873, "step": 920 }, { "epoch": 0.03144122519354948, "grad_norm": 0.5297457575798035, "learning_rate": 9.29e-06, "loss": 0.0904, "step": 930 }, { "epoch": 0.0317793028838027, "grad_norm": 0.956263542175293, "learning_rate": 9.39e-06, "loss": 0.0918, "step": 940 }, { "epoch": 0.032117380574055915, "grad_norm": 0.6928868889808655, "learning_rate": 9.49e-06, "loss": 0.0962, "step": 950 }, { "epoch": 0.03245545826430914, "grad_norm": 1.0160305500030518, "learning_rate": 9.59e-06, "loss": 0.0841, "step": 960 }, { "epoch": 0.03279353595456236, "grad_norm": 0.7149878740310669, "learning_rate": 9.69e-06, "loss": 0.0965, "step": 970 }, { "epoch": 0.03313161364481558, "grad_norm": 0.7188750505447388, "learning_rate": 9.790000000000001e-06, "loss": 0.093, "step": 980 }, { "epoch": 0.0334696913350688, "grad_norm": 0.8482939004898071, "learning_rate": 9.89e-06, "loss": 0.1041, "step": 990 }, { "epoch": 0.03380776902532202, "grad_norm": 0.7138314247131348, "learning_rate": 9.990000000000001e-06, "loss": 0.0918, "step": 1000 }, { "epoch": 0.03414584671557524, "grad_norm": 1.0867387056350708, "learning_rate": 1.0090000000000002e-05, "loss": 0.0897, "step": 1010 }, { "epoch": 0.03448392440582846, "grad_norm": 0.6685708165168762, "learning_rate": 1.019e-05, "loss": 0.0944, "step": 1020 }, { "epoch": 0.03482200209608168, "grad_norm": 0.6109023690223694, "learning_rate": 1.0290000000000001e-05, "loss": 0.0922, "step": 1030 }, { "epoch": 0.0351600797863349, "grad_norm": 1.111782193183899, "learning_rate": 1.039e-05, "loss": 0.087, "step": 1040 }, { "epoch": 0.03549815747658812, "grad_norm": 0.5832949280738831, "learning_rate": 1.049e-05, "loss": 0.0941, "step": 1050 }, { "epoch": 0.03583623516684134, "grad_norm": 0.5675982236862183, "learning_rate": 1.059e-05, "loss": 0.0889, "step": 1060 }, { "epoch": 0.03617431285709456, "grad_norm": 0.9915959239006042, "learning_rate": 1.069e-05, "loss": 0.0871, "step": 1070 }, { "epoch": 0.036512390547347784, "grad_norm": 0.5743878483772278, "learning_rate": 1.079e-05, "loss": 0.092, "step": 1080 }, { "epoch": 0.036850468237601, "grad_norm": 0.5849697589874268, "learning_rate": 1.089e-05, "loss": 0.097, "step": 1090 }, { "epoch": 0.03718854592785422, "grad_norm": 0.6304098963737488, "learning_rate": 1.099e-05, "loss": 0.0923, "step": 1100 }, { "epoch": 0.03752662361810744, "grad_norm": 0.6905559301376343, "learning_rate": 1.109e-05, "loss": 0.0868, "step": 1110 }, { "epoch": 0.037864701308360664, "grad_norm": 0.7441672086715698, "learning_rate": 1.1190000000000001e-05, "loss": 0.085, "step": 1120 }, { "epoch": 0.03820277899861388, "grad_norm": 0.6305090188980103, "learning_rate": 1.129e-05, "loss": 0.0806, "step": 1130 }, { "epoch": 0.0385408566888671, "grad_norm": 0.5369828343391418, "learning_rate": 1.1390000000000001e-05, "loss": 0.0854, "step": 1140 }, { "epoch": 0.03887893437912032, "grad_norm": 0.7512139081954956, "learning_rate": 1.149e-05, "loss": 0.0887, "step": 1150 }, { "epoch": 0.039217012069373544, "grad_norm": 0.671459436416626, "learning_rate": 1.159e-05, "loss": 0.092, "step": 1160 }, { "epoch": 0.03955508975962676, "grad_norm": 0.616829514503479, "learning_rate": 1.1690000000000002e-05, "loss": 0.0881, "step": 1170 }, { "epoch": 0.03989316744987998, "grad_norm": 0.7513883113861084, "learning_rate": 1.179e-05, "loss": 0.0861, "step": 1180 }, { "epoch": 0.0402312451401332, "grad_norm": 0.7204957604408264, "learning_rate": 1.1890000000000001e-05, "loss": 0.0843, "step": 1190 }, { "epoch": 0.040569322830386424, "grad_norm": 0.6346932053565979, "learning_rate": 1.199e-05, "loss": 0.0864, "step": 1200 }, { "epoch": 0.040907400520639646, "grad_norm": 0.6846521496772766, "learning_rate": 1.209e-05, "loss": 0.0819, "step": 1210 }, { "epoch": 0.04124547821089286, "grad_norm": 0.8567700982093811, "learning_rate": 1.219e-05, "loss": 0.0842, "step": 1220 }, { "epoch": 0.04158355590114608, "grad_norm": 0.6991008520126343, "learning_rate": 1.2290000000000001e-05, "loss": 0.088, "step": 1230 }, { "epoch": 0.041921633591399304, "grad_norm": 0.591948390007019, "learning_rate": 1.239e-05, "loss": 0.0876, "step": 1240 }, { "epoch": 0.042259711281652526, "grad_norm": 1.216821551322937, "learning_rate": 1.249e-05, "loss": 0.0854, "step": 1250 }, { "epoch": 0.04259778897190574, "grad_norm": 0.7426754236221313, "learning_rate": 1.2590000000000001e-05, "loss": 0.0899, "step": 1260 }, { "epoch": 0.04293586666215896, "grad_norm": 0.6344157457351685, "learning_rate": 1.2690000000000002e-05, "loss": 0.0864, "step": 1270 }, { "epoch": 0.043273944352412184, "grad_norm": 0.7530702948570251, "learning_rate": 1.2790000000000001e-05, "loss": 0.098, "step": 1280 }, { "epoch": 0.043612022042665406, "grad_norm": 0.6734321713447571, "learning_rate": 1.2889999999999999e-05, "loss": 0.0832, "step": 1290 }, { "epoch": 0.04395009973291863, "grad_norm": 0.886172890663147, "learning_rate": 1.299e-05, "loss": 0.084, "step": 1300 }, { "epoch": 0.04428817742317184, "grad_norm": 0.8004536032676697, "learning_rate": 1.309e-05, "loss": 0.0878, "step": 1310 }, { "epoch": 0.044626255113425065, "grad_norm": 0.8207268714904785, "learning_rate": 1.3189999999999999e-05, "loss": 0.083, "step": 1320 }, { "epoch": 0.044964332803678286, "grad_norm": 0.6262787580490112, "learning_rate": 1.329e-05, "loss": 0.0884, "step": 1330 }, { "epoch": 0.04530241049393151, "grad_norm": 1.1396807432174683, "learning_rate": 1.339e-05, "loss": 0.0891, "step": 1340 }, { "epoch": 0.04564048818418472, "grad_norm": 0.8013758063316345, "learning_rate": 1.349e-05, "loss": 0.0852, "step": 1350 }, { "epoch": 0.045978565874437945, "grad_norm": 0.7264882922172546, "learning_rate": 1.359e-05, "loss": 0.0878, "step": 1360 }, { "epoch": 0.046316643564691166, "grad_norm": 0.6602123379707336, "learning_rate": 1.369e-05, "loss": 0.08, "step": 1370 }, { "epoch": 0.04665472125494439, "grad_norm": 0.6034471988677979, "learning_rate": 1.379e-05, "loss": 0.0815, "step": 1380 }, { "epoch": 0.0469927989451976, "grad_norm": 0.5327042937278748, "learning_rate": 1.389e-05, "loss": 0.0826, "step": 1390 }, { "epoch": 0.047330876635450825, "grad_norm": 0.6664891242980957, "learning_rate": 1.399e-05, "loss": 0.0826, "step": 1400 }, { "epoch": 0.04766895432570405, "grad_norm": 0.6967563629150391, "learning_rate": 1.409e-05, "loss": 0.0867, "step": 1410 }, { "epoch": 0.04800703201595727, "grad_norm": 0.5746398568153381, "learning_rate": 1.4190000000000001e-05, "loss": 0.0811, "step": 1420 }, { "epoch": 0.04834510970621049, "grad_norm": 0.5105915069580078, "learning_rate": 1.429e-05, "loss": 0.0855, "step": 1430 }, { "epoch": 0.048683187396463705, "grad_norm": 0.7857663631439209, "learning_rate": 1.4390000000000001e-05, "loss": 0.0795, "step": 1440 }, { "epoch": 0.04902126508671693, "grad_norm": 0.8718492984771729, "learning_rate": 1.449e-05, "loss": 0.0858, "step": 1450 }, { "epoch": 0.04935934277697015, "grad_norm": 0.7079184651374817, "learning_rate": 1.4590000000000001e-05, "loss": 0.0937, "step": 1460 }, { "epoch": 0.04969742046722337, "grad_norm": 0.7561407685279846, "learning_rate": 1.4690000000000002e-05, "loss": 0.0896, "step": 1470 }, { "epoch": 0.050035498157476585, "grad_norm": 0.642855703830719, "learning_rate": 1.479e-05, "loss": 0.0811, "step": 1480 }, { "epoch": 0.05037357584772981, "grad_norm": 0.5102798938751221, "learning_rate": 1.4890000000000001e-05, "loss": 0.0732, "step": 1490 }, { "epoch": 0.05071165353798303, "grad_norm": 0.5152024030685425, "learning_rate": 1.499e-05, "loss": 0.0744, "step": 1500 }, { "epoch": 0.05104973122823625, "grad_norm": 0.4940118193626404, "learning_rate": 1.5090000000000001e-05, "loss": 0.0859, "step": 1510 }, { "epoch": 0.05138780891848947, "grad_norm": 0.7005085945129395, "learning_rate": 1.5190000000000002e-05, "loss": 0.0801, "step": 1520 }, { "epoch": 0.05172588660874269, "grad_norm": 0.6642612218856812, "learning_rate": 1.529e-05, "loss": 0.0778, "step": 1530 }, { "epoch": 0.05206396429899591, "grad_norm": 0.7608426213264465, "learning_rate": 1.539e-05, "loss": 0.0852, "step": 1540 }, { "epoch": 0.05240204198924913, "grad_norm": 0.6108283996582031, "learning_rate": 1.5490000000000002e-05, "loss": 0.0774, "step": 1550 }, { "epoch": 0.05274011967950235, "grad_norm": 0.5721437931060791, "learning_rate": 1.559e-05, "loss": 0.0776, "step": 1560 }, { "epoch": 0.05307819736975557, "grad_norm": 0.6113861799240112, "learning_rate": 1.569e-05, "loss": 0.0786, "step": 1570 }, { "epoch": 0.05341627506000879, "grad_norm": 0.5323138236999512, "learning_rate": 1.5790000000000003e-05, "loss": 0.0769, "step": 1580 }, { "epoch": 0.05375435275026201, "grad_norm": 0.6523522734642029, "learning_rate": 1.5890000000000002e-05, "loss": 0.0825, "step": 1590 }, { "epoch": 0.05409243044051523, "grad_norm": 0.5040363073348999, "learning_rate": 1.599e-05, "loss": 0.0733, "step": 1600 }, { "epoch": 0.05443050813076845, "grad_norm": 0.8144084215164185, "learning_rate": 1.609e-05, "loss": 0.0838, "step": 1610 }, { "epoch": 0.05476858582102167, "grad_norm": 0.7424857020378113, "learning_rate": 1.619e-05, "loss": 0.0895, "step": 1620 }, { "epoch": 0.05510666351127489, "grad_norm": 0.5825036764144897, "learning_rate": 1.6289999999999998e-05, "loss": 0.0785, "step": 1630 }, { "epoch": 0.05544474120152811, "grad_norm": 0.5619456171989441, "learning_rate": 1.639e-05, "loss": 0.0779, "step": 1640 }, { "epoch": 0.055782818891781334, "grad_norm": 0.5471194982528687, "learning_rate": 1.649e-05, "loss": 0.0744, "step": 1650 }, { "epoch": 0.05612089658203455, "grad_norm": 0.6421973705291748, "learning_rate": 1.659e-05, "loss": 0.0773, "step": 1660 }, { "epoch": 0.05645897427228777, "grad_norm": 0.6703518033027649, "learning_rate": 1.669e-05, "loss": 0.0749, "step": 1670 }, { "epoch": 0.05679705196254099, "grad_norm": 0.5437494516372681, "learning_rate": 1.679e-05, "loss": 0.0799, "step": 1680 }, { "epoch": 0.057135129652794214, "grad_norm": 0.5284901857376099, "learning_rate": 1.689e-05, "loss": 0.0857, "step": 1690 }, { "epoch": 0.05747320734304743, "grad_norm": 0.5970097184181213, "learning_rate": 1.699e-05, "loss": 0.0869, "step": 1700 }, { "epoch": 0.05781128503330065, "grad_norm": 0.6305414438247681, "learning_rate": 1.709e-05, "loss": 0.0814, "step": 1710 }, { "epoch": 0.05814936272355387, "grad_norm": 0.4963807463645935, "learning_rate": 1.719e-05, "loss": 0.085, "step": 1720 }, { "epoch": 0.058487440413807094, "grad_norm": 0.6050302982330322, "learning_rate": 1.7290000000000002e-05, "loss": 0.0842, "step": 1730 }, { "epoch": 0.058825518104060316, "grad_norm": 0.3957057595252991, "learning_rate": 1.739e-05, "loss": 0.0819, "step": 1740 }, { "epoch": 0.05916359579431353, "grad_norm": 1.059696912765503, "learning_rate": 1.749e-05, "loss": 0.0835, "step": 1750 }, { "epoch": 0.05950167348456675, "grad_norm": 1.0268698930740356, "learning_rate": 1.759e-05, "loss": 0.081, "step": 1760 }, { "epoch": 0.059839751174819975, "grad_norm": 0.4970506429672241, "learning_rate": 1.7690000000000002e-05, "loss": 0.0799, "step": 1770 }, { "epoch": 0.060177828865073196, "grad_norm": 0.5707601308822632, "learning_rate": 1.779e-05, "loss": 0.0775, "step": 1780 }, { "epoch": 0.06051590655532641, "grad_norm": 0.7345420122146606, "learning_rate": 1.789e-05, "loss": 0.0754, "step": 1790 }, { "epoch": 0.06085398424557963, "grad_norm": 0.6972876191139221, "learning_rate": 1.7990000000000002e-05, "loss": 0.0831, "step": 1800 }, { "epoch": 0.061192061935832855, "grad_norm": 0.6727140545845032, "learning_rate": 1.809e-05, "loss": 0.0815, "step": 1810 }, { "epoch": 0.061530139626086076, "grad_norm": 0.7114630937576294, "learning_rate": 1.819e-05, "loss": 0.0848, "step": 1820 }, { "epoch": 0.0618682173163393, "grad_norm": 0.5708699226379395, "learning_rate": 1.8290000000000003e-05, "loss": 0.0884, "step": 1830 }, { "epoch": 0.06220629500659251, "grad_norm": 0.5063364505767822, "learning_rate": 1.8390000000000002e-05, "loss": 0.0814, "step": 1840 }, { "epoch": 0.06254437269684573, "grad_norm": 0.6246423125267029, "learning_rate": 1.849e-05, "loss": 0.0795, "step": 1850 }, { "epoch": 0.06288245038709896, "grad_norm": 0.6105027198791504, "learning_rate": 1.8590000000000003e-05, "loss": 0.0882, "step": 1860 }, { "epoch": 0.06322052807735218, "grad_norm": 0.7434371113777161, "learning_rate": 1.8690000000000002e-05, "loss": 0.0886, "step": 1870 }, { "epoch": 0.0635586057676054, "grad_norm": 0.4846676290035248, "learning_rate": 1.879e-05, "loss": 0.0876, "step": 1880 }, { "epoch": 0.06389668345785862, "grad_norm": 0.9508723616600037, "learning_rate": 1.8890000000000004e-05, "loss": 0.086, "step": 1890 }, { "epoch": 0.06423476114811183, "grad_norm": 0.5683633685112, "learning_rate": 1.8990000000000003e-05, "loss": 0.0768, "step": 1900 }, { "epoch": 0.06457283883836505, "grad_norm": 0.5916979312896729, "learning_rate": 1.909e-05, "loss": 0.086, "step": 1910 }, { "epoch": 0.06491091652861827, "grad_norm": 0.9850573539733887, "learning_rate": 1.919e-05, "loss": 0.0757, "step": 1920 }, { "epoch": 0.0652489942188715, "grad_norm": 0.8154470920562744, "learning_rate": 1.929e-05, "loss": 0.0776, "step": 1930 }, { "epoch": 0.06558707190912472, "grad_norm": 0.6619365215301514, "learning_rate": 1.939e-05, "loss": 0.0755, "step": 1940 }, { "epoch": 0.06592514959937794, "grad_norm": 0.6861457228660583, "learning_rate": 1.949e-05, "loss": 0.0745, "step": 1950 }, { "epoch": 0.06626322728963116, "grad_norm": 0.3985060453414917, "learning_rate": 1.959e-05, "loss": 0.0766, "step": 1960 }, { "epoch": 0.06660130497988438, "grad_norm": 0.4608534872531891, "learning_rate": 1.969e-05, "loss": 0.0747, "step": 1970 }, { "epoch": 0.0669393826701376, "grad_norm": 0.7063499689102173, "learning_rate": 1.979e-05, "loss": 0.0798, "step": 1980 }, { "epoch": 0.06727746036039081, "grad_norm": 0.80928635597229, "learning_rate": 1.989e-05, "loss": 0.0773, "step": 1990 }, { "epoch": 0.06761553805064403, "grad_norm": 0.41274794936180115, "learning_rate": 1.999e-05, "loss": 0.0833, "step": 2000 }, { "epoch": 0.06795361574089726, "grad_norm": 0.5321969985961914, "learning_rate": 2.009e-05, "loss": 0.0762, "step": 2010 }, { "epoch": 0.06829169343115048, "grad_norm": 0.8127756118774414, "learning_rate": 2.019e-05, "loss": 0.0764, "step": 2020 }, { "epoch": 0.0686297711214037, "grad_norm": 0.4742896258831024, "learning_rate": 2.029e-05, "loss": 0.0772, "step": 2030 }, { "epoch": 0.06896784881165692, "grad_norm": 0.4958079159259796, "learning_rate": 2.039e-05, "loss": 0.0762, "step": 2040 }, { "epoch": 0.06930592650191014, "grad_norm": 0.46309006214141846, "learning_rate": 2.0490000000000002e-05, "loss": 0.0797, "step": 2050 }, { "epoch": 0.06964400419216336, "grad_norm": 0.5309962630271912, "learning_rate": 2.059e-05, "loss": 0.0783, "step": 2060 }, { "epoch": 0.06998208188241659, "grad_norm": 0.5701480507850647, "learning_rate": 2.069e-05, "loss": 0.0801, "step": 2070 }, { "epoch": 0.0703201595726698, "grad_norm": 0.5480809807777405, "learning_rate": 2.0790000000000003e-05, "loss": 0.0766, "step": 2080 }, { "epoch": 0.07065823726292302, "grad_norm": 0.6877966523170471, "learning_rate": 2.089e-05, "loss": 0.077, "step": 2090 }, { "epoch": 0.07099631495317624, "grad_norm": 0.8258718848228455, "learning_rate": 2.099e-05, "loss": 0.0773, "step": 2100 }, { "epoch": 0.07133439264342946, "grad_norm": 0.5237071514129639, "learning_rate": 2.1090000000000003e-05, "loss": 0.0764, "step": 2110 }, { "epoch": 0.07167247033368268, "grad_norm": 0.5490078330039978, "learning_rate": 2.1190000000000002e-05, "loss": 0.0713, "step": 2120 }, { "epoch": 0.0720105480239359, "grad_norm": 0.5079713463783264, "learning_rate": 2.129e-05, "loss": 0.0903, "step": 2130 }, { "epoch": 0.07234862571418912, "grad_norm": 0.3650508224964142, "learning_rate": 2.139e-05, "loss": 0.079, "step": 2140 }, { "epoch": 0.07268670340444235, "grad_norm": 0.5984599590301514, "learning_rate": 2.1490000000000003e-05, "loss": 0.0769, "step": 2150 }, { "epoch": 0.07302478109469557, "grad_norm": 0.4619382619857788, "learning_rate": 2.159e-05, "loss": 0.0744, "step": 2160 }, { "epoch": 0.07336285878494878, "grad_norm": 0.8106852769851685, "learning_rate": 2.169e-05, "loss": 0.0765, "step": 2170 }, { "epoch": 0.073700936475202, "grad_norm": 0.4582311511039734, "learning_rate": 2.1790000000000003e-05, "loss": 0.0748, "step": 2180 }, { "epoch": 0.07403901416545522, "grad_norm": 0.46059083938598633, "learning_rate": 2.1890000000000002e-05, "loss": 0.0732, "step": 2190 }, { "epoch": 0.07437709185570844, "grad_norm": 0.5939735174179077, "learning_rate": 2.199e-05, "loss": 0.0848, "step": 2200 }, { "epoch": 0.07471516954596166, "grad_norm": 0.5236706733703613, "learning_rate": 2.2090000000000004e-05, "loss": 0.08, "step": 2210 }, { "epoch": 0.07505324723621488, "grad_norm": 0.5894333720207214, "learning_rate": 2.219e-05, "loss": 0.0788, "step": 2220 }, { "epoch": 0.0753913249264681, "grad_norm": 0.4520858824253082, "learning_rate": 2.229e-05, "loss": 0.0776, "step": 2230 }, { "epoch": 0.07572940261672133, "grad_norm": 0.5394782423973083, "learning_rate": 2.239e-05, "loss": 0.0683, "step": 2240 }, { "epoch": 0.07606748030697454, "grad_norm": 0.7953131198883057, "learning_rate": 2.249e-05, "loss": 0.0801, "step": 2250 }, { "epoch": 0.07640555799722776, "grad_norm": 0.42779427766799927, "learning_rate": 2.259e-05, "loss": 0.081, "step": 2260 }, { "epoch": 0.07674363568748098, "grad_norm": 0.28214922547340393, "learning_rate": 2.269e-05, "loss": 0.0812, "step": 2270 }, { "epoch": 0.0770817133777342, "grad_norm": 0.7415865659713745, "learning_rate": 2.279e-05, "loss": 0.0798, "step": 2280 }, { "epoch": 0.07741979106798742, "grad_norm": 0.6466509699821472, "learning_rate": 2.289e-05, "loss": 0.0816, "step": 2290 }, { "epoch": 0.07775786875824064, "grad_norm": 0.7670685052871704, "learning_rate": 2.2990000000000002e-05, "loss": 0.0805, "step": 2300 }, { "epoch": 0.07809594644849387, "grad_norm": 0.4998510181903839, "learning_rate": 2.309e-05, "loss": 0.0841, "step": 2310 }, { "epoch": 0.07843402413874709, "grad_norm": 0.6086809039115906, "learning_rate": 2.319e-05, "loss": 0.0849, "step": 2320 }, { "epoch": 0.07877210182900031, "grad_norm": 0.479108989238739, "learning_rate": 2.3290000000000002e-05, "loss": 0.0775, "step": 2330 }, { "epoch": 0.07911017951925352, "grad_norm": 0.46661531925201416, "learning_rate": 2.339e-05, "loss": 0.0685, "step": 2340 }, { "epoch": 0.07944825720950674, "grad_norm": 0.5987676382064819, "learning_rate": 2.349e-05, "loss": 0.0782, "step": 2350 }, { "epoch": 0.07978633489975996, "grad_norm": 0.4220929741859436, "learning_rate": 2.359e-05, "loss": 0.0718, "step": 2360 }, { "epoch": 0.08012441259001318, "grad_norm": 0.5461485981941223, "learning_rate": 2.3690000000000002e-05, "loss": 0.0707, "step": 2370 }, { "epoch": 0.0804624902802664, "grad_norm": 0.43515992164611816, "learning_rate": 2.379e-05, "loss": 0.0774, "step": 2380 }, { "epoch": 0.08080056797051963, "grad_norm": 0.46257463097572327, "learning_rate": 2.389e-05, "loss": 0.0722, "step": 2390 }, { "epoch": 0.08113864566077285, "grad_norm": 0.5627842545509338, "learning_rate": 2.3990000000000002e-05, "loss": 0.0796, "step": 2400 }, { "epoch": 0.08147672335102607, "grad_norm": 0.4600749611854553, "learning_rate": 2.409e-05, "loss": 0.0738, "step": 2410 }, { "epoch": 0.08181480104127929, "grad_norm": 0.5646429061889648, "learning_rate": 2.419e-05, "loss": 0.0772, "step": 2420 }, { "epoch": 0.0821528787315325, "grad_norm": 0.5839623212814331, "learning_rate": 2.4290000000000003e-05, "loss": 0.0796, "step": 2430 }, { "epoch": 0.08249095642178572, "grad_norm": 0.574521005153656, "learning_rate": 2.4390000000000002e-05, "loss": 0.0791, "step": 2440 }, { "epoch": 0.08282903411203894, "grad_norm": 0.4836287498474121, "learning_rate": 2.449e-05, "loss": 0.076, "step": 2450 }, { "epoch": 0.08316711180229217, "grad_norm": 0.45668691396713257, "learning_rate": 2.4590000000000003e-05, "loss": 0.0761, "step": 2460 }, { "epoch": 0.08350518949254539, "grad_norm": 0.5593380928039551, "learning_rate": 2.4690000000000002e-05, "loss": 0.0724, "step": 2470 }, { "epoch": 0.08384326718279861, "grad_norm": 0.5085211992263794, "learning_rate": 2.479e-05, "loss": 0.0748, "step": 2480 }, { "epoch": 0.08418134487305183, "grad_norm": 0.39153769612312317, "learning_rate": 2.489e-05, "loss": 0.0828, "step": 2490 }, { "epoch": 0.08451942256330505, "grad_norm": 0.3843998610973358, "learning_rate": 2.4990000000000003e-05, "loss": 0.0756, "step": 2500 }, { "epoch": 0.08485750025355827, "grad_norm": 0.4356549084186554, "learning_rate": 2.5090000000000002e-05, "loss": 0.0734, "step": 2510 }, { "epoch": 0.08519557794381148, "grad_norm": 0.4590463638305664, "learning_rate": 2.519e-05, "loss": 0.0707, "step": 2520 }, { "epoch": 0.0855336556340647, "grad_norm": 0.6739285588264465, "learning_rate": 2.5290000000000004e-05, "loss": 0.0774, "step": 2530 }, { "epoch": 0.08587173332431793, "grad_norm": 0.45171815156936646, "learning_rate": 2.5390000000000003e-05, "loss": 0.0745, "step": 2540 }, { "epoch": 0.08620981101457115, "grad_norm": 0.46833911538124084, "learning_rate": 2.549e-05, "loss": 0.0787, "step": 2550 }, { "epoch": 0.08654788870482437, "grad_norm": 0.7191058397293091, "learning_rate": 2.5590000000000004e-05, "loss": 0.0763, "step": 2560 }, { "epoch": 0.08688596639507759, "grad_norm": 0.541989266872406, "learning_rate": 2.5690000000000003e-05, "loss": 0.0737, "step": 2570 }, { "epoch": 0.08722404408533081, "grad_norm": 0.34988313913345337, "learning_rate": 2.5790000000000002e-05, "loss": 0.0677, "step": 2580 }, { "epoch": 0.08756212177558403, "grad_norm": 0.4897046983242035, "learning_rate": 2.5890000000000005e-05, "loss": 0.0757, "step": 2590 }, { "epoch": 0.08790019946583726, "grad_norm": 0.6574180722236633, "learning_rate": 2.5990000000000004e-05, "loss": 0.0817, "step": 2600 }, { "epoch": 0.08823827715609046, "grad_norm": 0.4771818220615387, "learning_rate": 2.6090000000000003e-05, "loss": 0.0761, "step": 2610 }, { "epoch": 0.08857635484634369, "grad_norm": 0.7282753586769104, "learning_rate": 2.6190000000000005e-05, "loss": 0.0797, "step": 2620 }, { "epoch": 0.08891443253659691, "grad_norm": 0.7909854054450989, "learning_rate": 2.6290000000000004e-05, "loss": 0.078, "step": 2630 }, { "epoch": 0.08925251022685013, "grad_norm": 0.4389767646789551, "learning_rate": 2.6390000000000003e-05, "loss": 0.0708, "step": 2640 }, { "epoch": 0.08959058791710335, "grad_norm": 0.44053369760513306, "learning_rate": 2.6490000000000002e-05, "loss": 0.0746, "step": 2650 }, { "epoch": 0.08992866560735657, "grad_norm": 0.4017919898033142, "learning_rate": 2.6590000000000005e-05, "loss": 0.0687, "step": 2660 }, { "epoch": 0.0902667432976098, "grad_norm": 0.46746042370796204, "learning_rate": 2.6690000000000004e-05, "loss": 0.0733, "step": 2670 }, { "epoch": 0.09060482098786302, "grad_norm": 0.48043587803840637, "learning_rate": 2.6790000000000003e-05, "loss": 0.0707, "step": 2680 }, { "epoch": 0.09094289867811624, "grad_norm": 0.5575421452522278, "learning_rate": 2.689e-05, "loss": 0.0832, "step": 2690 }, { "epoch": 0.09128097636836945, "grad_norm": 0.3914690315723419, "learning_rate": 2.6989999999999997e-05, "loss": 0.0654, "step": 2700 }, { "epoch": 0.09161905405862267, "grad_norm": 0.4167788028717041, "learning_rate": 2.709e-05, "loss": 0.0691, "step": 2710 }, { "epoch": 0.09195713174887589, "grad_norm": 0.3304353952407837, "learning_rate": 2.719e-05, "loss": 0.0674, "step": 2720 }, { "epoch": 0.09229520943912911, "grad_norm": 0.6301798224449158, "learning_rate": 2.7289999999999998e-05, "loss": 0.0777, "step": 2730 }, { "epoch": 0.09263328712938233, "grad_norm": 0.31624341011047363, "learning_rate": 2.739e-05, "loss": 0.0771, "step": 2740 }, { "epoch": 0.09297136481963555, "grad_norm": 0.5439915657043457, "learning_rate": 2.749e-05, "loss": 0.0689, "step": 2750 }, { "epoch": 0.09330944250988878, "grad_norm": 0.9240056276321411, "learning_rate": 2.759e-05, "loss": 0.0744, "step": 2760 }, { "epoch": 0.093647520200142, "grad_norm": 0.5250310897827148, "learning_rate": 2.769e-05, "loss": 0.0698, "step": 2770 }, { "epoch": 0.0939855978903952, "grad_norm": 0.45452943444252014, "learning_rate": 2.779e-05, "loss": 0.0726, "step": 2780 }, { "epoch": 0.09432367558064843, "grad_norm": 0.48889586329460144, "learning_rate": 2.789e-05, "loss": 0.0741, "step": 2790 }, { "epoch": 0.09466175327090165, "grad_norm": 0.4627422094345093, "learning_rate": 2.7989999999999998e-05, "loss": 0.0767, "step": 2800 }, { "epoch": 0.09499983096115487, "grad_norm": 0.40158548951148987, "learning_rate": 2.809e-05, "loss": 0.0782, "step": 2810 }, { "epoch": 0.0953379086514081, "grad_norm": 0.48446547985076904, "learning_rate": 2.819e-05, "loss": 0.0799, "step": 2820 }, { "epoch": 0.09567598634166131, "grad_norm": 0.4773123264312744, "learning_rate": 2.829e-05, "loss": 0.073, "step": 2830 }, { "epoch": 0.09601406403191454, "grad_norm": 0.45729145407676697, "learning_rate": 2.839e-05, "loss": 0.078, "step": 2840 }, { "epoch": 0.09635214172216776, "grad_norm": 0.406588077545166, "learning_rate": 2.849e-05, "loss": 0.0723, "step": 2850 }, { "epoch": 0.09669021941242098, "grad_norm": 0.5658850073814392, "learning_rate": 2.859e-05, "loss": 0.0759, "step": 2860 }, { "epoch": 0.09702829710267419, "grad_norm": 0.5513496994972229, "learning_rate": 2.869e-05, "loss": 0.0717, "step": 2870 }, { "epoch": 0.09736637479292741, "grad_norm": 0.5243033766746521, "learning_rate": 2.879e-05, "loss": 0.0725, "step": 2880 }, { "epoch": 0.09770445248318063, "grad_norm": 0.49698880314826965, "learning_rate": 2.889e-05, "loss": 0.0773, "step": 2890 }, { "epoch": 0.09804253017343385, "grad_norm": 0.32042214274406433, "learning_rate": 2.8990000000000002e-05, "loss": 0.0648, "step": 2900 }, { "epoch": 0.09838060786368708, "grad_norm": 0.7985565662384033, "learning_rate": 2.909e-05, "loss": 0.0733, "step": 2910 }, { "epoch": 0.0987186855539403, "grad_norm": 0.5278567671775818, "learning_rate": 2.919e-05, "loss": 0.0733, "step": 2920 }, { "epoch": 0.09905676324419352, "grad_norm": 0.360675573348999, "learning_rate": 2.929e-05, "loss": 0.0794, "step": 2930 }, { "epoch": 0.09939484093444674, "grad_norm": 0.7039985656738281, "learning_rate": 2.939e-05, "loss": 0.0851, "step": 2940 }, { "epoch": 0.09973291862469996, "grad_norm": 0.6530967354774475, "learning_rate": 2.949e-05, "loss": 0.0702, "step": 2950 }, { "epoch": 0.10007099631495317, "grad_norm": 0.5308017730712891, "learning_rate": 2.959e-05, "loss": 0.0795, "step": 2960 }, { "epoch": 0.10040907400520639, "grad_norm": 0.5589978694915771, "learning_rate": 2.9690000000000002e-05, "loss": 0.0807, "step": 2970 }, { "epoch": 0.10074715169545961, "grad_norm": 0.45812708139419556, "learning_rate": 2.979e-05, "loss": 0.0731, "step": 2980 }, { "epoch": 0.10108522938571284, "grad_norm": 0.43373408913612366, "learning_rate": 2.989e-05, "loss": 0.0732, "step": 2990 }, { "epoch": 0.10142330707596606, "grad_norm": 0.39102935791015625, "learning_rate": 2.9990000000000003e-05, "loss": 0.0772, "step": 3000 }, { "epoch": 0.10176138476621928, "grad_norm": 0.39844515919685364, "learning_rate": 3.009e-05, "loss": 0.0794, "step": 3010 }, { "epoch": 0.1020994624564725, "grad_norm": 1.2043662071228027, "learning_rate": 3.019e-05, "loss": 0.0796, "step": 3020 }, { "epoch": 0.10243754014672572, "grad_norm": 0.472261518239975, "learning_rate": 3.0290000000000003e-05, "loss": 0.0736, "step": 3030 }, { "epoch": 0.10277561783697894, "grad_norm": 0.3474313020706177, "learning_rate": 3.0390000000000002e-05, "loss": 0.0689, "step": 3040 }, { "epoch": 0.10311369552723215, "grad_norm": 0.537964403629303, "learning_rate": 3.049e-05, "loss": 0.0693, "step": 3050 }, { "epoch": 0.10345177321748537, "grad_norm": 0.39553937315940857, "learning_rate": 3.0590000000000004e-05, "loss": 0.077, "step": 3060 }, { "epoch": 0.1037898509077386, "grad_norm": 0.3369290232658386, "learning_rate": 3.069e-05, "loss": 0.0743, "step": 3070 }, { "epoch": 0.10412792859799182, "grad_norm": 0.3111400902271271, "learning_rate": 3.079e-05, "loss": 0.0753, "step": 3080 }, { "epoch": 0.10446600628824504, "grad_norm": 0.3808109164237976, "learning_rate": 3.0890000000000004e-05, "loss": 0.0689, "step": 3090 }, { "epoch": 0.10480408397849826, "grad_norm": 0.4429011344909668, "learning_rate": 3.099e-05, "loss": 0.075, "step": 3100 }, { "epoch": 0.10514216166875148, "grad_norm": 0.33667972683906555, "learning_rate": 3.109e-05, "loss": 0.0682, "step": 3110 }, { "epoch": 0.1054802393590047, "grad_norm": 0.3878828287124634, "learning_rate": 3.1190000000000005e-05, "loss": 0.0672, "step": 3120 }, { "epoch": 0.10581831704925793, "grad_norm": 0.5110762715339661, "learning_rate": 3.129e-05, "loss": 0.0795, "step": 3130 }, { "epoch": 0.10615639473951113, "grad_norm": 0.3132859170436859, "learning_rate": 3.139e-05, "loss": 0.0729, "step": 3140 }, { "epoch": 0.10649447242976436, "grad_norm": 0.49295881390571594, "learning_rate": 3.1490000000000005e-05, "loss": 0.0747, "step": 3150 }, { "epoch": 0.10683255012001758, "grad_norm": 0.4575751721858978, "learning_rate": 3.159e-05, "loss": 0.0771, "step": 3160 }, { "epoch": 0.1071706278102708, "grad_norm": 0.43519967794418335, "learning_rate": 3.169e-05, "loss": 0.073, "step": 3170 }, { "epoch": 0.10750870550052402, "grad_norm": 0.33941128849983215, "learning_rate": 3.1790000000000006e-05, "loss": 0.0738, "step": 3180 }, { "epoch": 0.10784678319077724, "grad_norm": 0.4584379494190216, "learning_rate": 3.189e-05, "loss": 0.0758, "step": 3190 }, { "epoch": 0.10818486088103046, "grad_norm": 0.37240320444107056, "learning_rate": 3.1990000000000004e-05, "loss": 0.0766, "step": 3200 }, { "epoch": 0.10852293857128369, "grad_norm": 0.570612370967865, "learning_rate": 3.2090000000000006e-05, "loss": 0.072, "step": 3210 }, { "epoch": 0.1088610162615369, "grad_norm": 0.6042145490646362, "learning_rate": 3.219e-05, "loss": 0.0707, "step": 3220 }, { "epoch": 0.10919909395179012, "grad_norm": 0.4139137268066406, "learning_rate": 3.2290000000000004e-05, "loss": 0.0663, "step": 3230 }, { "epoch": 0.10953717164204334, "grad_norm": 0.4669465124607086, "learning_rate": 3.239000000000001e-05, "loss": 0.0742, "step": 3240 }, { "epoch": 0.10987524933229656, "grad_norm": 0.5595911145210266, "learning_rate": 3.249e-05, "loss": 0.0778, "step": 3250 }, { "epoch": 0.11021332702254978, "grad_norm": 0.5202668309211731, "learning_rate": 3.2590000000000005e-05, "loss": 0.0774, "step": 3260 }, { "epoch": 0.110551404712803, "grad_norm": 0.7767825126647949, "learning_rate": 3.269000000000001e-05, "loss": 0.0672, "step": 3270 }, { "epoch": 0.11088948240305622, "grad_norm": 0.4295240044593811, "learning_rate": 3.279e-05, "loss": 0.0668, "step": 3280 }, { "epoch": 0.11122756009330945, "grad_norm": 0.39326706528663635, "learning_rate": 3.2890000000000005e-05, "loss": 0.0779, "step": 3290 }, { "epoch": 0.11156563778356267, "grad_norm": 0.4967365562915802, "learning_rate": 3.299e-05, "loss": 0.0706, "step": 3300 }, { "epoch": 0.11190371547381588, "grad_norm": 0.34590741991996765, "learning_rate": 3.309e-05, "loss": 0.0638, "step": 3310 }, { "epoch": 0.1122417931640691, "grad_norm": 0.5582442283630371, "learning_rate": 3.319e-05, "loss": 0.0734, "step": 3320 }, { "epoch": 0.11257987085432232, "grad_norm": 0.7582153081893921, "learning_rate": 3.329e-05, "loss": 0.0708, "step": 3330 }, { "epoch": 0.11291794854457554, "grad_norm": 0.3517226278781891, "learning_rate": 3.339e-05, "loss": 0.0765, "step": 3340 }, { "epoch": 0.11325602623482876, "grad_norm": 0.33347344398498535, "learning_rate": 3.349e-05, "loss": 0.0706, "step": 3350 }, { "epoch": 0.11359410392508199, "grad_norm": 0.5725781917572021, "learning_rate": 3.359e-05, "loss": 0.0724, "step": 3360 }, { "epoch": 0.1139321816153352, "grad_norm": 0.4206569194793701, "learning_rate": 3.369e-05, "loss": 0.0745, "step": 3370 }, { "epoch": 0.11427025930558843, "grad_norm": 0.36960193514823914, "learning_rate": 3.379e-05, "loss": 0.0715, "step": 3380 }, { "epoch": 0.11460833699584165, "grad_norm": 0.5431758761405945, "learning_rate": 3.389e-05, "loss": 0.0667, "step": 3390 }, { "epoch": 0.11494641468609486, "grad_norm": 0.4312625229358673, "learning_rate": 3.399e-05, "loss": 0.0692, "step": 3400 }, { "epoch": 0.11528449237634808, "grad_norm": 0.41917115449905396, "learning_rate": 3.409e-05, "loss": 0.0715, "step": 3410 }, { "epoch": 0.1156225700666013, "grad_norm": 0.4955395758152008, "learning_rate": 3.419e-05, "loss": 0.0753, "step": 3420 }, { "epoch": 0.11596064775685452, "grad_norm": 0.39011624455451965, "learning_rate": 3.429e-05, "loss": 0.0694, "step": 3430 }, { "epoch": 0.11629872544710775, "grad_norm": 0.5306334495544434, "learning_rate": 3.439e-05, "loss": 0.0765, "step": 3440 }, { "epoch": 0.11663680313736097, "grad_norm": 0.36897507309913635, "learning_rate": 3.449e-05, "loss": 0.0664, "step": 3450 }, { "epoch": 0.11697488082761419, "grad_norm": 0.3673994243144989, "learning_rate": 3.459e-05, "loss": 0.0755, "step": 3460 }, { "epoch": 0.11731295851786741, "grad_norm": 0.5020207166671753, "learning_rate": 3.469e-05, "loss": 0.0672, "step": 3470 }, { "epoch": 0.11765103620812063, "grad_norm": 0.3197127878665924, "learning_rate": 3.479e-05, "loss": 0.0707, "step": 3480 }, { "epoch": 0.11798911389837384, "grad_norm": 0.42562684416770935, "learning_rate": 3.489e-05, "loss": 0.0784, "step": 3490 }, { "epoch": 0.11832719158862706, "grad_norm": 0.40347880125045776, "learning_rate": 3.499e-05, "loss": 0.0682, "step": 3500 }, { "epoch": 0.11866526927888028, "grad_norm": 0.4892159402370453, "learning_rate": 3.509e-05, "loss": 0.0689, "step": 3510 }, { "epoch": 0.1190033469691335, "grad_norm": 0.3372395634651184, "learning_rate": 3.519e-05, "loss": 0.0728, "step": 3520 }, { "epoch": 0.11934142465938673, "grad_norm": 0.4849388301372528, "learning_rate": 3.529e-05, "loss": 0.0706, "step": 3530 }, { "epoch": 0.11967950234963995, "grad_norm": 0.44825631380081177, "learning_rate": 3.539e-05, "loss": 0.0678, "step": 3540 }, { "epoch": 0.12001758003989317, "grad_norm": 0.34838712215423584, "learning_rate": 3.549e-05, "loss": 0.0654, "step": 3550 }, { "epoch": 0.12035565773014639, "grad_norm": 0.3989817500114441, "learning_rate": 3.559e-05, "loss": 0.0801, "step": 3560 }, { "epoch": 0.12069373542039961, "grad_norm": 0.3853890001773834, "learning_rate": 3.569e-05, "loss": 0.0782, "step": 3570 }, { "epoch": 0.12103181311065282, "grad_norm": 0.3836081326007843, "learning_rate": 3.579e-05, "loss": 0.0739, "step": 3580 }, { "epoch": 0.12136989080090604, "grad_norm": 0.5489023327827454, "learning_rate": 3.5890000000000004e-05, "loss": 0.0644, "step": 3590 }, { "epoch": 0.12170796849115927, "grad_norm": 0.3695901334285736, "learning_rate": 3.599e-05, "loss": 0.0682, "step": 3600 }, { "epoch": 0.12204604618141249, "grad_norm": 0.6558501720428467, "learning_rate": 3.609e-05, "loss": 0.0744, "step": 3610 }, { "epoch": 0.12238412387166571, "grad_norm": 0.4021861255168915, "learning_rate": 3.6190000000000004e-05, "loss": 0.0658, "step": 3620 }, { "epoch": 0.12272220156191893, "grad_norm": 0.54509037733078, "learning_rate": 3.629e-05, "loss": 0.071, "step": 3630 }, { "epoch": 0.12306027925217215, "grad_norm": 0.3900390565395355, "learning_rate": 3.639e-05, "loss": 0.0712, "step": 3640 }, { "epoch": 0.12339835694242537, "grad_norm": 0.2725812792778015, "learning_rate": 3.6490000000000005e-05, "loss": 0.0668, "step": 3650 }, { "epoch": 0.1237364346326786, "grad_norm": 0.476818710565567, "learning_rate": 3.659e-05, "loss": 0.0712, "step": 3660 }, { "epoch": 0.1240745123229318, "grad_norm": 0.6275448799133301, "learning_rate": 3.669e-05, "loss": 0.0759, "step": 3670 }, { "epoch": 0.12441259001318503, "grad_norm": 0.5113996863365173, "learning_rate": 3.6790000000000005e-05, "loss": 0.0687, "step": 3680 }, { "epoch": 0.12475066770343825, "grad_norm": 0.4301358461380005, "learning_rate": 3.689e-05, "loss": 0.0695, "step": 3690 }, { "epoch": 0.12508874539369147, "grad_norm": 0.29858797788619995, "learning_rate": 3.699e-05, "loss": 0.0724, "step": 3700 }, { "epoch": 0.12542682308394468, "grad_norm": 0.32833701372146606, "learning_rate": 3.7090000000000006e-05, "loss": 0.0778, "step": 3710 }, { "epoch": 0.1257649007741979, "grad_norm": 0.4491752088069916, "learning_rate": 3.719e-05, "loss": 0.0737, "step": 3720 }, { "epoch": 0.12610297846445112, "grad_norm": 0.4124026298522949, "learning_rate": 3.7290000000000004e-05, "loss": 0.0716, "step": 3730 }, { "epoch": 0.12644105615470436, "grad_norm": 0.4438646137714386, "learning_rate": 3.739e-05, "loss": 0.0761, "step": 3740 }, { "epoch": 0.12677913384495756, "grad_norm": 0.4042142331600189, "learning_rate": 3.749e-05, "loss": 0.0751, "step": 3750 }, { "epoch": 0.1271172115352108, "grad_norm": 0.4262681007385254, "learning_rate": 3.7590000000000004e-05, "loss": 0.0739, "step": 3760 }, { "epoch": 0.127455289225464, "grad_norm": 0.3941057026386261, "learning_rate": 3.769e-05, "loss": 0.0685, "step": 3770 }, { "epoch": 0.12779336691571724, "grad_norm": 0.46704167127609253, "learning_rate": 3.779e-05, "loss": 0.065, "step": 3780 }, { "epoch": 0.12813144460597045, "grad_norm": 0.3324509263038635, "learning_rate": 3.7890000000000005e-05, "loss": 0.0741, "step": 3790 }, { "epoch": 0.12846952229622366, "grad_norm": 0.6049647927284241, "learning_rate": 3.799e-05, "loss": 0.0746, "step": 3800 }, { "epoch": 0.1288075999864769, "grad_norm": 0.3537854254245758, "learning_rate": 3.809e-05, "loss": 0.07, "step": 3810 }, { "epoch": 0.1291456776767301, "grad_norm": 0.43633410334587097, "learning_rate": 3.8190000000000005e-05, "loss": 0.0718, "step": 3820 }, { "epoch": 0.12948375536698334, "grad_norm": 0.3530278205871582, "learning_rate": 3.829e-05, "loss": 0.0715, "step": 3830 }, { "epoch": 0.12982183305723655, "grad_norm": 0.4186595380306244, "learning_rate": 3.8390000000000003e-05, "loss": 0.0715, "step": 3840 }, { "epoch": 0.13015991074748978, "grad_norm": 0.379126638174057, "learning_rate": 3.8490000000000006e-05, "loss": 0.0664, "step": 3850 }, { "epoch": 0.130497988437743, "grad_norm": 0.42173850536346436, "learning_rate": 3.859e-05, "loss": 0.072, "step": 3860 }, { "epoch": 0.13083606612799623, "grad_norm": 0.5996578335762024, "learning_rate": 3.8690000000000004e-05, "loss": 0.0723, "step": 3870 }, { "epoch": 0.13117414381824943, "grad_norm": 0.44873303174972534, "learning_rate": 3.8790000000000006e-05, "loss": 0.0697, "step": 3880 }, { "epoch": 0.13151222150850264, "grad_norm": 0.42008307576179504, "learning_rate": 3.889e-05, "loss": 0.0629, "step": 3890 }, { "epoch": 0.13185029919875588, "grad_norm": 0.4466307461261749, "learning_rate": 3.8990000000000004e-05, "loss": 0.0716, "step": 3900 }, { "epoch": 0.13218837688900908, "grad_norm": 0.4051268994808197, "learning_rate": 3.909000000000001e-05, "loss": 0.0689, "step": 3910 }, { "epoch": 0.13252645457926232, "grad_norm": 0.4822331964969635, "learning_rate": 3.919e-05, "loss": 0.075, "step": 3920 }, { "epoch": 0.13286453226951553, "grad_norm": 0.4384876787662506, "learning_rate": 3.9290000000000005e-05, "loss": 0.0703, "step": 3930 }, { "epoch": 0.13320260995976876, "grad_norm": 0.443498820066452, "learning_rate": 3.939e-05, "loss": 0.0735, "step": 3940 }, { "epoch": 0.13354068765002197, "grad_norm": 0.5678014159202576, "learning_rate": 3.9489999999999996e-05, "loss": 0.0755, "step": 3950 }, { "epoch": 0.1338787653402752, "grad_norm": 0.44405341148376465, "learning_rate": 3.959e-05, "loss": 0.0687, "step": 3960 }, { "epoch": 0.13421684303052842, "grad_norm": 0.31187495589256287, "learning_rate": 3.969e-05, "loss": 0.0756, "step": 3970 }, { "epoch": 0.13455492072078162, "grad_norm": 0.3395049273967743, "learning_rate": 3.979e-05, "loss": 0.0778, "step": 3980 }, { "epoch": 0.13489299841103486, "grad_norm": 0.4395677447319031, "learning_rate": 3.989e-05, "loss": 0.0789, "step": 3990 }, { "epoch": 0.13523107610128807, "grad_norm": 0.3899039924144745, "learning_rate": 3.999e-05, "loss": 0.0719, "step": 4000 }, { "epoch": 0.1355691537915413, "grad_norm": 0.3107697367668152, "learning_rate": 4.009e-05, "loss": 0.0711, "step": 4010 }, { "epoch": 0.1359072314817945, "grad_norm": 0.5347669720649719, "learning_rate": 4.019e-05, "loss": 0.0716, "step": 4020 }, { "epoch": 0.13624530917204775, "grad_norm": 0.46638479828834534, "learning_rate": 4.029e-05, "loss": 0.0667, "step": 4030 }, { "epoch": 0.13658338686230095, "grad_norm": 0.33241400122642517, "learning_rate": 4.039e-05, "loss": 0.0758, "step": 4040 }, { "epoch": 0.1369214645525542, "grad_norm": 0.3670966923236847, "learning_rate": 4.049e-05, "loss": 0.0739, "step": 4050 }, { "epoch": 0.1372595422428074, "grad_norm": 0.3839397430419922, "learning_rate": 4.059e-05, "loss": 0.0672, "step": 4060 }, { "epoch": 0.1375976199330606, "grad_norm": 0.2882821261882782, "learning_rate": 4.069e-05, "loss": 0.0672, "step": 4070 }, { "epoch": 0.13793569762331384, "grad_norm": 0.5169243812561035, "learning_rate": 4.079e-05, "loss": 0.0654, "step": 4080 }, { "epoch": 0.13827377531356705, "grad_norm": 0.39581170678138733, "learning_rate": 4.089e-05, "loss": 0.0691, "step": 4090 }, { "epoch": 0.13861185300382028, "grad_norm": 0.588420569896698, "learning_rate": 4.099e-05, "loss": 0.0673, "step": 4100 }, { "epoch": 0.1389499306940735, "grad_norm": 0.44730398058891296, "learning_rate": 4.109e-05, "loss": 0.0703, "step": 4110 }, { "epoch": 0.13928800838432673, "grad_norm": 0.34184226393699646, "learning_rate": 4.1190000000000004e-05, "loss": 0.0663, "step": 4120 }, { "epoch": 0.13962608607457994, "grad_norm": 0.4705362617969513, "learning_rate": 4.129e-05, "loss": 0.0755, "step": 4130 }, { "epoch": 0.13996416376483317, "grad_norm": 0.3700624406337738, "learning_rate": 4.139e-05, "loss": 0.0719, "step": 4140 }, { "epoch": 0.14030224145508638, "grad_norm": 0.37471091747283936, "learning_rate": 4.1490000000000004e-05, "loss": 0.0688, "step": 4150 }, { "epoch": 0.1406403191453396, "grad_norm": 0.2797147333621979, "learning_rate": 4.159e-05, "loss": 0.0682, "step": 4160 }, { "epoch": 0.14097839683559282, "grad_norm": 0.4785003364086151, "learning_rate": 4.169e-05, "loss": 0.0703, "step": 4170 }, { "epoch": 0.14131647452584603, "grad_norm": 0.5545189380645752, "learning_rate": 4.179e-05, "loss": 0.0689, "step": 4180 }, { "epoch": 0.14165455221609927, "grad_norm": 0.31849977374076843, "learning_rate": 4.189e-05, "loss": 0.0741, "step": 4190 }, { "epoch": 0.14199262990635247, "grad_norm": 0.40208008885383606, "learning_rate": 4.199e-05, "loss": 0.0775, "step": 4200 }, { "epoch": 0.1423307075966057, "grad_norm": 0.49309471249580383, "learning_rate": 4.209e-05, "loss": 0.0706, "step": 4210 }, { "epoch": 0.14266878528685892, "grad_norm": 0.38846006989479065, "learning_rate": 4.219e-05, "loss": 0.0668, "step": 4220 }, { "epoch": 0.14300686297711215, "grad_norm": 0.40614694356918335, "learning_rate": 4.229e-05, "loss": 0.0587, "step": 4230 }, { "epoch": 0.14334494066736536, "grad_norm": 0.4071804881095886, "learning_rate": 4.239e-05, "loss": 0.0711, "step": 4240 }, { "epoch": 0.14368301835761857, "grad_norm": 0.362138569355011, "learning_rate": 4.249e-05, "loss": 0.0636, "step": 4250 }, { "epoch": 0.1440210960478718, "grad_norm": 0.340567946434021, "learning_rate": 4.2590000000000004e-05, "loss": 0.0656, "step": 4260 }, { "epoch": 0.144359173738125, "grad_norm": 0.3680441677570343, "learning_rate": 4.269e-05, "loss": 0.069, "step": 4270 }, { "epoch": 0.14469725142837825, "grad_norm": 0.5892869234085083, "learning_rate": 4.279e-05, "loss": 0.0713, "step": 4280 }, { "epoch": 0.14503532911863146, "grad_norm": 0.30516406893730164, "learning_rate": 4.2890000000000004e-05, "loss": 0.07, "step": 4290 }, { "epoch": 0.1453734068088847, "grad_norm": 0.405927449464798, "learning_rate": 4.299e-05, "loss": 0.0715, "step": 4300 }, { "epoch": 0.1457114844991379, "grad_norm": 0.2891083061695099, "learning_rate": 4.309e-05, "loss": 0.0649, "step": 4310 }, { "epoch": 0.14604956218939114, "grad_norm": 0.3005872666835785, "learning_rate": 4.3190000000000005e-05, "loss": 0.0661, "step": 4320 }, { "epoch": 0.14638763987964434, "grad_norm": 0.36418771743774414, "learning_rate": 4.329e-05, "loss": 0.0724, "step": 4330 }, { "epoch": 0.14672571756989755, "grad_norm": 0.36661073565483093, "learning_rate": 4.339e-05, "loss": 0.0681, "step": 4340 }, { "epoch": 0.1470637952601508, "grad_norm": 0.43230709433555603, "learning_rate": 4.3490000000000005e-05, "loss": 0.0679, "step": 4350 }, { "epoch": 0.147401872950404, "grad_norm": 0.3887544870376587, "learning_rate": 4.359e-05, "loss": 0.0716, "step": 4360 }, { "epoch": 0.14773995064065723, "grad_norm": 0.5609337687492371, "learning_rate": 4.3690000000000004e-05, "loss": 0.0713, "step": 4370 }, { "epoch": 0.14807802833091044, "grad_norm": 0.2930644452571869, "learning_rate": 4.3790000000000006e-05, "loss": 0.071, "step": 4380 }, { "epoch": 0.14841610602116367, "grad_norm": 0.34067997336387634, "learning_rate": 4.389e-05, "loss": 0.0672, "step": 4390 }, { "epoch": 0.14875418371141688, "grad_norm": 0.35093823075294495, "learning_rate": 4.3990000000000004e-05, "loss": 0.0695, "step": 4400 }, { "epoch": 0.14909226140167012, "grad_norm": 0.4354107081890106, "learning_rate": 4.4090000000000006e-05, "loss": 0.0706, "step": 4410 }, { "epoch": 0.14943033909192333, "grad_norm": 0.44938018918037415, "learning_rate": 4.419e-05, "loss": 0.0687, "step": 4420 }, { "epoch": 0.14976841678217653, "grad_norm": 0.45166105031967163, "learning_rate": 4.4290000000000005e-05, "loss": 0.0721, "step": 4430 }, { "epoch": 0.15010649447242977, "grad_norm": 0.48713141679763794, "learning_rate": 4.439000000000001e-05, "loss": 0.0729, "step": 4440 }, { "epoch": 0.15044457216268298, "grad_norm": 0.5012915134429932, "learning_rate": 4.449e-05, "loss": 0.0726, "step": 4450 }, { "epoch": 0.1507826498529362, "grad_norm": 0.5053837895393372, "learning_rate": 4.4590000000000005e-05, "loss": 0.0646, "step": 4460 }, { "epoch": 0.15112072754318942, "grad_norm": 0.5475727319717407, "learning_rate": 4.469e-05, "loss": 0.0671, "step": 4470 }, { "epoch": 0.15145880523344266, "grad_norm": 0.3156111240386963, "learning_rate": 4.479e-05, "loss": 0.0712, "step": 4480 }, { "epoch": 0.15179688292369586, "grad_norm": 0.335126131772995, "learning_rate": 4.4890000000000006e-05, "loss": 0.0688, "step": 4490 }, { "epoch": 0.15213496061394907, "grad_norm": 0.581258237361908, "learning_rate": 4.499e-05, "loss": 0.0706, "step": 4500 }, { "epoch": 0.1524730383042023, "grad_norm": 0.5209929943084717, "learning_rate": 4.5090000000000004e-05, "loss": 0.0704, "step": 4510 }, { "epoch": 0.15281111599445552, "grad_norm": 0.6086817383766174, "learning_rate": 4.5190000000000006e-05, "loss": 0.074, "step": 4520 }, { "epoch": 0.15314919368470875, "grad_norm": 0.3746090233325958, "learning_rate": 4.529e-05, "loss": 0.0689, "step": 4530 }, { "epoch": 0.15348727137496196, "grad_norm": 0.48051929473876953, "learning_rate": 4.5390000000000004e-05, "loss": 0.0728, "step": 4540 }, { "epoch": 0.1538253490652152, "grad_norm": 0.2750568985939026, "learning_rate": 4.549000000000001e-05, "loss": 0.0659, "step": 4550 }, { "epoch": 0.1541634267554684, "grad_norm": 0.5118159651756287, "learning_rate": 4.559e-05, "loss": 0.077, "step": 4560 }, { "epoch": 0.15450150444572164, "grad_norm": 0.486673504114151, "learning_rate": 4.569e-05, "loss": 0.065, "step": 4570 }, { "epoch": 0.15483958213597485, "grad_norm": 0.3565453588962555, "learning_rate": 4.579e-05, "loss": 0.0762, "step": 4580 }, { "epoch": 0.15517765982622805, "grad_norm": 0.34188997745513916, "learning_rate": 4.589e-05, "loss": 0.0663, "step": 4590 }, { "epoch": 0.1555157375164813, "grad_norm": 0.3584916591644287, "learning_rate": 4.599e-05, "loss": 0.0749, "step": 4600 }, { "epoch": 0.1558538152067345, "grad_norm": 0.48341235518455505, "learning_rate": 4.609e-05, "loss": 0.0804, "step": 4610 }, { "epoch": 0.15619189289698773, "grad_norm": 0.4194978177547455, "learning_rate": 4.619e-05, "loss": 0.0728, "step": 4620 }, { "epoch": 0.15652997058724094, "grad_norm": 0.44240036606788635, "learning_rate": 4.629e-05, "loss": 0.0687, "step": 4630 }, { "epoch": 0.15686804827749418, "grad_norm": 0.32886260747909546, "learning_rate": 4.639e-05, "loss": 0.0705, "step": 4640 }, { "epoch": 0.15720612596774738, "grad_norm": 0.3580717146396637, "learning_rate": 4.649e-05, "loss": 0.0746, "step": 4650 }, { "epoch": 0.15754420365800062, "grad_norm": 0.305351197719574, "learning_rate": 4.659e-05, "loss": 0.064, "step": 4660 }, { "epoch": 0.15788228134825383, "grad_norm": 0.4547359049320221, "learning_rate": 4.669e-05, "loss": 0.0602, "step": 4670 }, { "epoch": 0.15822035903850704, "grad_norm": 0.31770262122154236, "learning_rate": 4.679e-05, "loss": 0.0651, "step": 4680 }, { "epoch": 0.15855843672876027, "grad_norm": 0.43239766359329224, "learning_rate": 4.689e-05, "loss": 0.0725, "step": 4690 }, { "epoch": 0.15889651441901348, "grad_norm": 0.4101863205432892, "learning_rate": 4.699e-05, "loss": 0.0653, "step": 4700 }, { "epoch": 0.15923459210926671, "grad_norm": 0.4501327872276306, "learning_rate": 4.709e-05, "loss": 0.0803, "step": 4710 }, { "epoch": 0.15957266979951992, "grad_norm": 0.33576205372810364, "learning_rate": 4.719e-05, "loss": 0.0674, "step": 4720 }, { "epoch": 0.15991074748977316, "grad_norm": 0.299565851688385, "learning_rate": 4.729e-05, "loss": 0.0686, "step": 4730 }, { "epoch": 0.16024882518002637, "grad_norm": 0.2957133650779724, "learning_rate": 4.739e-05, "loss": 0.0621, "step": 4740 }, { "epoch": 0.1605869028702796, "grad_norm": 0.4034571647644043, "learning_rate": 4.749e-05, "loss": 0.0648, "step": 4750 }, { "epoch": 0.1609249805605328, "grad_norm": 0.506633996963501, "learning_rate": 4.7590000000000003e-05, "loss": 0.0722, "step": 4760 }, { "epoch": 0.16126305825078602, "grad_norm": 0.3023461103439331, "learning_rate": 4.769e-05, "loss": 0.0687, "step": 4770 }, { "epoch": 0.16160113594103925, "grad_norm": 0.43200868368148804, "learning_rate": 4.779e-05, "loss": 0.0693, "step": 4780 }, { "epoch": 0.16193921363129246, "grad_norm": 0.327826589345932, "learning_rate": 4.7890000000000004e-05, "loss": 0.0675, "step": 4790 }, { "epoch": 0.1622772913215457, "grad_norm": 0.5735267400741577, "learning_rate": 4.799e-05, "loss": 0.0703, "step": 4800 }, { "epoch": 0.1626153690117989, "grad_norm": 0.613783597946167, "learning_rate": 4.809e-05, "loss": 0.0701, "step": 4810 }, { "epoch": 0.16295344670205214, "grad_norm": 0.3431846499443054, "learning_rate": 4.8190000000000004e-05, "loss": 0.0697, "step": 4820 }, { "epoch": 0.16329152439230535, "grad_norm": 0.3682699501514435, "learning_rate": 4.829e-05, "loss": 0.0735, "step": 4830 }, { "epoch": 0.16362960208255858, "grad_norm": 0.4743006229400635, "learning_rate": 4.839e-05, "loss": 0.072, "step": 4840 }, { "epoch": 0.1639676797728118, "grad_norm": 0.3947041630744934, "learning_rate": 4.8490000000000005e-05, "loss": 0.0709, "step": 4850 }, { "epoch": 0.164305757463065, "grad_norm": 0.38029196858406067, "learning_rate": 4.859e-05, "loss": 0.0695, "step": 4860 }, { "epoch": 0.16464383515331824, "grad_norm": 0.36895257234573364, "learning_rate": 4.869e-05, "loss": 0.0713, "step": 4870 }, { "epoch": 0.16498191284357144, "grad_norm": 0.3597721457481384, "learning_rate": 4.8790000000000006e-05, "loss": 0.0703, "step": 4880 }, { "epoch": 0.16531999053382468, "grad_norm": 0.40473586320877075, "learning_rate": 4.889e-05, "loss": 0.0776, "step": 4890 }, { "epoch": 0.1656580682240779, "grad_norm": 0.24488134682178497, "learning_rate": 4.8990000000000004e-05, "loss": 0.0717, "step": 4900 }, { "epoch": 0.16599614591433112, "grad_norm": 0.3759375512599945, "learning_rate": 4.9090000000000006e-05, "loss": 0.0722, "step": 4910 }, { "epoch": 0.16633422360458433, "grad_norm": 0.3181911110877991, "learning_rate": 4.919e-05, "loss": 0.0684, "step": 4920 }, { "epoch": 0.16667230129483757, "grad_norm": 0.5038986206054688, "learning_rate": 4.9290000000000004e-05, "loss": 0.0704, "step": 4930 }, { "epoch": 0.16701037898509077, "grad_norm": 0.5045334100723267, "learning_rate": 4.939e-05, "loss": 0.0671, "step": 4940 }, { "epoch": 0.16734845667534398, "grad_norm": 0.3437768518924713, "learning_rate": 4.949e-05, "loss": 0.0758, "step": 4950 }, { "epoch": 0.16768653436559722, "grad_norm": 0.5231336355209351, "learning_rate": 4.9590000000000005e-05, "loss": 0.0739, "step": 4960 }, { "epoch": 0.16802461205585043, "grad_norm": 0.4673420786857605, "learning_rate": 4.969e-05, "loss": 0.0654, "step": 4970 }, { "epoch": 0.16836268974610366, "grad_norm": 0.34895485639572144, "learning_rate": 4.979e-05, "loss": 0.071, "step": 4980 }, { "epoch": 0.16870076743635687, "grad_norm": 0.746017575263977, "learning_rate": 4.9890000000000005e-05, "loss": 0.0711, "step": 4990 }, { "epoch": 0.1690388451266101, "grad_norm": 0.37887200713157654, "learning_rate": 4.999e-05, "loss": 0.0717, "step": 5000 }, { "epoch": 0.1693769228168633, "grad_norm": 0.39519885182380676, "learning_rate": 5.009e-05, "loss": 0.0665, "step": 5010 }, { "epoch": 0.16971500050711655, "grad_norm": 0.3428352177143097, "learning_rate": 5.0190000000000006e-05, "loss": 0.0707, "step": 5020 }, { "epoch": 0.17005307819736976, "grad_norm": 0.5686338543891907, "learning_rate": 5.029e-05, "loss": 0.0748, "step": 5030 }, { "epoch": 0.17039115588762296, "grad_norm": 0.5148627758026123, "learning_rate": 5.0390000000000004e-05, "loss": 0.071, "step": 5040 }, { "epoch": 0.1707292335778762, "grad_norm": 0.45334845781326294, "learning_rate": 5.0490000000000006e-05, "loss": 0.0653, "step": 5050 }, { "epoch": 0.1710673112681294, "grad_norm": 0.42467594146728516, "learning_rate": 5.059e-05, "loss": 0.0696, "step": 5060 }, { "epoch": 0.17140538895838264, "grad_norm": 0.29623278975486755, "learning_rate": 5.0690000000000004e-05, "loss": 0.0703, "step": 5070 }, { "epoch": 0.17174346664863585, "grad_norm": 0.44480374455451965, "learning_rate": 5.079000000000001e-05, "loss": 0.0732, "step": 5080 }, { "epoch": 0.1720815443388891, "grad_norm": 0.38201019167900085, "learning_rate": 5.089e-05, "loss": 0.0687, "step": 5090 }, { "epoch": 0.1724196220291423, "grad_norm": 0.33607757091522217, "learning_rate": 5.0990000000000005e-05, "loss": 0.066, "step": 5100 }, { "epoch": 0.17275769971939553, "grad_norm": 0.5431190133094788, "learning_rate": 5.109000000000001e-05, "loss": 0.0721, "step": 5110 }, { "epoch": 0.17309577740964874, "grad_norm": 0.6978402137756348, "learning_rate": 5.119e-05, "loss": 0.0753, "step": 5120 }, { "epoch": 0.17343385509990195, "grad_norm": 0.3304062485694885, "learning_rate": 5.1290000000000005e-05, "loss": 0.064, "step": 5130 }, { "epoch": 0.17377193279015518, "grad_norm": 0.4395250976085663, "learning_rate": 5.139000000000001e-05, "loss": 0.067, "step": 5140 }, { "epoch": 0.1741100104804084, "grad_norm": 0.33032602071762085, "learning_rate": 5.149e-05, "loss": 0.0682, "step": 5150 }, { "epoch": 0.17444808817066162, "grad_norm": 0.5579239726066589, "learning_rate": 5.1590000000000006e-05, "loss": 0.0707, "step": 5160 }, { "epoch": 0.17478616586091483, "grad_norm": 0.3347111642360687, "learning_rate": 5.169000000000001e-05, "loss": 0.0625, "step": 5170 }, { "epoch": 0.17512424355116807, "grad_norm": 0.42045924067497253, "learning_rate": 5.1790000000000004e-05, "loss": 0.0687, "step": 5180 }, { "epoch": 0.17546232124142128, "grad_norm": 0.35852542519569397, "learning_rate": 5.1890000000000006e-05, "loss": 0.0684, "step": 5190 }, { "epoch": 0.1758003989316745, "grad_norm": 0.35498595237731934, "learning_rate": 5.199000000000001e-05, "loss": 0.0676, "step": 5200 }, { "epoch": 0.17613847662192772, "grad_norm": 0.3570035994052887, "learning_rate": 5.2090000000000004e-05, "loss": 0.0726, "step": 5210 }, { "epoch": 0.17647655431218093, "grad_norm": 0.44041547179222107, "learning_rate": 5.219000000000001e-05, "loss": 0.0699, "step": 5220 }, { "epoch": 0.17681463200243416, "grad_norm": 0.4464341998100281, "learning_rate": 5.229e-05, "loss": 0.0669, "step": 5230 }, { "epoch": 0.17715270969268737, "grad_norm": 0.4389634430408478, "learning_rate": 5.2390000000000005e-05, "loss": 0.0738, "step": 5240 }, { "epoch": 0.1774907873829406, "grad_norm": 0.3145093023777008, "learning_rate": 5.249000000000001e-05, "loss": 0.0743, "step": 5250 }, { "epoch": 0.17782886507319381, "grad_norm": 0.5347403883934021, "learning_rate": 5.259e-05, "loss": 0.0723, "step": 5260 }, { "epoch": 0.17816694276344705, "grad_norm": 0.5367119312286377, "learning_rate": 5.2690000000000005e-05, "loss": 0.0652, "step": 5270 }, { "epoch": 0.17850502045370026, "grad_norm": 0.32778382301330566, "learning_rate": 5.279000000000001e-05, "loss": 0.0681, "step": 5280 }, { "epoch": 0.1788430981439535, "grad_norm": 0.3705621361732483, "learning_rate": 5.2890000000000004e-05, "loss": 0.0738, "step": 5290 }, { "epoch": 0.1791811758342067, "grad_norm": 0.3621100187301636, "learning_rate": 5.2990000000000006e-05, "loss": 0.0672, "step": 5300 }, { "epoch": 0.1795192535244599, "grad_norm": 0.28766191005706787, "learning_rate": 5.309000000000001e-05, "loss": 0.068, "step": 5310 }, { "epoch": 0.17985733121471315, "grad_norm": 0.524904727935791, "learning_rate": 5.3190000000000004e-05, "loss": 0.0683, "step": 5320 }, { "epoch": 0.18019540890496635, "grad_norm": 0.3655865788459778, "learning_rate": 5.3290000000000006e-05, "loss": 0.0734, "step": 5330 }, { "epoch": 0.1805334865952196, "grad_norm": 0.32284459471702576, "learning_rate": 5.339000000000001e-05, "loss": 0.0731, "step": 5340 }, { "epoch": 0.1808715642854728, "grad_norm": 0.2655646800994873, "learning_rate": 5.3490000000000005e-05, "loss": 0.0621, "step": 5350 }, { "epoch": 0.18120964197572603, "grad_norm": 0.316707044839859, "learning_rate": 5.359000000000001e-05, "loss": 0.0661, "step": 5360 }, { "epoch": 0.18154771966597924, "grad_norm": 0.3111066520214081, "learning_rate": 5.369000000000001e-05, "loss": 0.0679, "step": 5370 }, { "epoch": 0.18188579735623248, "grad_norm": 0.3469178080558777, "learning_rate": 5.3790000000000005e-05, "loss": 0.0678, "step": 5380 }, { "epoch": 0.18222387504648568, "grad_norm": 0.4741226136684418, "learning_rate": 5.389000000000001e-05, "loss": 0.0594, "step": 5390 }, { "epoch": 0.1825619527367389, "grad_norm": 0.41749879717826843, "learning_rate": 5.399000000000001e-05, "loss": 0.0711, "step": 5400 }, { "epoch": 0.18290003042699213, "grad_norm": 0.41824284195899963, "learning_rate": 5.4090000000000006e-05, "loss": 0.0736, "step": 5410 }, { "epoch": 0.18323810811724534, "grad_norm": 0.3256002962589264, "learning_rate": 5.419000000000001e-05, "loss": 0.0583, "step": 5420 }, { "epoch": 0.18357618580749857, "grad_norm": 0.33930888772010803, "learning_rate": 5.429000000000001e-05, "loss": 0.0698, "step": 5430 }, { "epoch": 0.18391426349775178, "grad_norm": 0.3863186836242676, "learning_rate": 5.4390000000000006e-05, "loss": 0.0626, "step": 5440 }, { "epoch": 0.18425234118800501, "grad_norm": 0.39126572012901306, "learning_rate": 5.449000000000001e-05, "loss": 0.0721, "step": 5450 }, { "epoch": 0.18459041887825822, "grad_norm": 0.3973144590854645, "learning_rate": 5.459000000000001e-05, "loss": 0.0716, "step": 5460 }, { "epoch": 0.18492849656851143, "grad_norm": 0.27249306440353394, "learning_rate": 5.469000000000001e-05, "loss": 0.0673, "step": 5470 }, { "epoch": 0.18526657425876467, "grad_norm": 0.30070143938064575, "learning_rate": 5.479000000000001e-05, "loss": 0.0642, "step": 5480 }, { "epoch": 0.18560465194901787, "grad_norm": 0.31851285696029663, "learning_rate": 5.4890000000000005e-05, "loss": 0.0714, "step": 5490 }, { "epoch": 0.1859427296392711, "grad_norm": 0.31655433773994446, "learning_rate": 5.499000000000001e-05, "loss": 0.0696, "step": 5500 }, { "epoch": 0.18628080732952432, "grad_norm": 0.47075992822647095, "learning_rate": 5.5089999999999996e-05, "loss": 0.0603, "step": 5510 }, { "epoch": 0.18661888501977755, "grad_norm": 0.43628984689712524, "learning_rate": 5.519e-05, "loss": 0.0637, "step": 5520 }, { "epoch": 0.18695696271003076, "grad_norm": 0.3189806342124939, "learning_rate": 5.5289999999999994e-05, "loss": 0.07, "step": 5530 }, { "epoch": 0.187295040400284, "grad_norm": 0.3704143762588501, "learning_rate": 5.5389999999999997e-05, "loss": 0.0634, "step": 5540 }, { "epoch": 0.1876331180905372, "grad_norm": 0.29803791642189026, "learning_rate": 5.549e-05, "loss": 0.069, "step": 5550 }, { "epoch": 0.1879711957807904, "grad_norm": 0.4349849820137024, "learning_rate": 5.5589999999999995e-05, "loss": 0.0735, "step": 5560 }, { "epoch": 0.18830927347104365, "grad_norm": 0.2973686456680298, "learning_rate": 5.569e-05, "loss": 0.0785, "step": 5570 }, { "epoch": 0.18864735116129686, "grad_norm": 0.4325537085533142, "learning_rate": 5.579e-05, "loss": 0.0667, "step": 5580 }, { "epoch": 0.1889854288515501, "grad_norm": 0.3357708156108856, "learning_rate": 5.5889999999999995e-05, "loss": 0.0617, "step": 5590 }, { "epoch": 0.1893235065418033, "grad_norm": 0.5214833617210388, "learning_rate": 5.599e-05, "loss": 0.0701, "step": 5600 }, { "epoch": 0.18966158423205653, "grad_norm": 0.24141763150691986, "learning_rate": 5.609e-05, "loss": 0.0718, "step": 5610 }, { "epoch": 0.18999966192230974, "grad_norm": 0.4673295021057129, "learning_rate": 5.6189999999999996e-05, "loss": 0.0703, "step": 5620 }, { "epoch": 0.19033773961256298, "grad_norm": 0.3186624050140381, "learning_rate": 5.629e-05, "loss": 0.0744, "step": 5630 }, { "epoch": 0.1906758173028162, "grad_norm": 0.3105831444263458, "learning_rate": 5.639e-05, "loss": 0.0659, "step": 5640 }, { "epoch": 0.1910138949930694, "grad_norm": 0.5084529519081116, "learning_rate": 5.6489999999999996e-05, "loss": 0.0633, "step": 5650 }, { "epoch": 0.19135197268332263, "grad_norm": 0.3277716338634491, "learning_rate": 5.659e-05, "loss": 0.0737, "step": 5660 }, { "epoch": 0.19169005037357584, "grad_norm": 0.3383154571056366, "learning_rate": 5.669e-05, "loss": 0.0708, "step": 5670 }, { "epoch": 0.19202812806382907, "grad_norm": 0.30705806612968445, "learning_rate": 5.679e-05, "loss": 0.0708, "step": 5680 }, { "epoch": 0.19236620575408228, "grad_norm": 0.34646129608154297, "learning_rate": 5.689e-05, "loss": 0.072, "step": 5690 }, { "epoch": 0.19270428344433552, "grad_norm": 0.3713158071041107, "learning_rate": 5.699e-05, "loss": 0.0664, "step": 5700 }, { "epoch": 0.19304236113458872, "grad_norm": 0.3073630928993225, "learning_rate": 5.709e-05, "loss": 0.0655, "step": 5710 }, { "epoch": 0.19338043882484196, "grad_norm": 0.32416611909866333, "learning_rate": 5.719e-05, "loss": 0.0693, "step": 5720 }, { "epoch": 0.19371851651509517, "grad_norm": 0.3505132794380188, "learning_rate": 5.729e-05, "loss": 0.0706, "step": 5730 }, { "epoch": 0.19405659420534838, "grad_norm": 0.6839906573295593, "learning_rate": 5.739e-05, "loss": 0.07, "step": 5740 }, { "epoch": 0.1943946718956016, "grad_norm": 0.3532968759536743, "learning_rate": 5.749e-05, "loss": 0.0705, "step": 5750 }, { "epoch": 0.19473274958585482, "grad_norm": 0.4514351487159729, "learning_rate": 5.759e-05, "loss": 0.0741, "step": 5760 }, { "epoch": 0.19507082727610806, "grad_norm": 0.4244040548801422, "learning_rate": 5.769e-05, "loss": 0.0706, "step": 5770 }, { "epoch": 0.19540890496636126, "grad_norm": 0.5127183794975281, "learning_rate": 5.779e-05, "loss": 0.0677, "step": 5780 }, { "epoch": 0.1957469826566145, "grad_norm": 0.33200979232788086, "learning_rate": 5.789e-05, "loss": 0.0707, "step": 5790 }, { "epoch": 0.1960850603468677, "grad_norm": 0.4184166193008423, "learning_rate": 5.799e-05, "loss": 0.0632, "step": 5800 }, { "epoch": 0.19642313803712094, "grad_norm": 0.42406463623046875, "learning_rate": 5.809e-05, "loss": 0.0688, "step": 5810 }, { "epoch": 0.19676121572737415, "grad_norm": 0.33916836977005005, "learning_rate": 5.819e-05, "loss": 0.0705, "step": 5820 }, { "epoch": 0.19709929341762736, "grad_norm": 0.27014368772506714, "learning_rate": 5.829e-05, "loss": 0.0644, "step": 5830 }, { "epoch": 0.1974373711078806, "grad_norm": 0.28067851066589355, "learning_rate": 5.839e-05, "loss": 0.0586, "step": 5840 }, { "epoch": 0.1977754487981338, "grad_norm": 0.3322197198867798, "learning_rate": 5.849e-05, "loss": 0.0704, "step": 5850 }, { "epoch": 0.19811352648838704, "grad_norm": 0.3846915364265442, "learning_rate": 5.859e-05, "loss": 0.0691, "step": 5860 }, { "epoch": 0.19845160417864025, "grad_norm": 0.4424373507499695, "learning_rate": 5.869e-05, "loss": 0.067, "step": 5870 }, { "epoch": 0.19878968186889348, "grad_norm": 0.3311385214328766, "learning_rate": 5.879e-05, "loss": 0.0659, "step": 5880 }, { "epoch": 0.1991277595591467, "grad_norm": 0.5282202959060669, "learning_rate": 5.889e-05, "loss": 0.0642, "step": 5890 }, { "epoch": 0.19946583724939992, "grad_norm": 0.38063573837280273, "learning_rate": 5.899e-05, "loss": 0.0652, "step": 5900 }, { "epoch": 0.19980391493965313, "grad_norm": 0.6936233639717102, "learning_rate": 5.909e-05, "loss": 0.0715, "step": 5910 }, { "epoch": 0.20014199262990634, "grad_norm": 0.5400716066360474, "learning_rate": 5.919e-05, "loss": 0.0697, "step": 5920 }, { "epoch": 0.20048007032015958, "grad_norm": 0.3162354528903961, "learning_rate": 5.929e-05, "loss": 0.0704, "step": 5930 }, { "epoch": 0.20081814801041278, "grad_norm": 0.24136967957019806, "learning_rate": 5.939e-05, "loss": 0.0728, "step": 5940 }, { "epoch": 0.20115622570066602, "grad_norm": 0.26192858815193176, "learning_rate": 5.949e-05, "loss": 0.0714, "step": 5950 }, { "epoch": 0.20149430339091923, "grad_norm": 0.43517005443573, "learning_rate": 5.9590000000000004e-05, "loss": 0.0641, "step": 5960 }, { "epoch": 0.20183238108117246, "grad_norm": 0.505982518196106, "learning_rate": 5.969e-05, "loss": 0.0643, "step": 5970 }, { "epoch": 0.20217045877142567, "grad_norm": 0.2502892017364502, "learning_rate": 5.979e-05, "loss": 0.0756, "step": 5980 }, { "epoch": 0.2025085364616789, "grad_norm": 0.2877435088157654, "learning_rate": 5.9890000000000004e-05, "loss": 0.0663, "step": 5990 }, { "epoch": 0.20284661415193211, "grad_norm": 0.34232816100120544, "learning_rate": 5.999e-05, "loss": 0.0656, "step": 6000 }, { "epoch": 0.20318469184218532, "grad_norm": 0.3635973334312439, "learning_rate": 6.009e-05, "loss": 0.0715, "step": 6010 }, { "epoch": 0.20352276953243856, "grad_norm": 0.28784045577049255, "learning_rate": 6.0190000000000005e-05, "loss": 0.072, "step": 6020 }, { "epoch": 0.20386084722269177, "grad_norm": 0.28504639863967896, "learning_rate": 6.029e-05, "loss": 0.0714, "step": 6030 }, { "epoch": 0.204198924912945, "grad_norm": 0.25487926602363586, "learning_rate": 6.039e-05, "loss": 0.0672, "step": 6040 }, { "epoch": 0.2045370026031982, "grad_norm": 0.27338287234306335, "learning_rate": 6.0490000000000005e-05, "loss": 0.0697, "step": 6050 }, { "epoch": 0.20487508029345144, "grad_norm": 0.41891151666641235, "learning_rate": 6.059e-05, "loss": 0.0722, "step": 6060 }, { "epoch": 0.20521315798370465, "grad_norm": 0.2604564130306244, "learning_rate": 6.069e-05, "loss": 0.0691, "step": 6070 }, { "epoch": 0.2055512356739579, "grad_norm": 0.29332461953163147, "learning_rate": 6.0790000000000006e-05, "loss": 0.0764, "step": 6080 }, { "epoch": 0.2058893133642111, "grad_norm": 0.2684028446674347, "learning_rate": 6.089e-05, "loss": 0.0731, "step": 6090 }, { "epoch": 0.2062273910544643, "grad_norm": 0.31647229194641113, "learning_rate": 6.0990000000000004e-05, "loss": 0.0648, "step": 6100 }, { "epoch": 0.20656546874471754, "grad_norm": 0.3063628375530243, "learning_rate": 6.109e-05, "loss": 0.0691, "step": 6110 }, { "epoch": 0.20690354643497075, "grad_norm": 0.32492172718048096, "learning_rate": 6.119e-05, "loss": 0.0662, "step": 6120 }, { "epoch": 0.20724162412522398, "grad_norm": 0.41072994470596313, "learning_rate": 6.129e-05, "loss": 0.0698, "step": 6130 }, { "epoch": 0.2075797018154772, "grad_norm": 0.31822535395622253, "learning_rate": 6.139000000000001e-05, "loss": 0.0668, "step": 6140 }, { "epoch": 0.20791777950573043, "grad_norm": 0.3648895025253296, "learning_rate": 6.149000000000001e-05, "loss": 0.065, "step": 6150 }, { "epoch": 0.20825585719598363, "grad_norm": 0.42366987466812134, "learning_rate": 6.159e-05, "loss": 0.0746, "step": 6160 }, { "epoch": 0.20859393488623687, "grad_norm": 0.3146710991859436, "learning_rate": 6.169e-05, "loss": 0.0666, "step": 6170 }, { "epoch": 0.20893201257649008, "grad_norm": 0.457878977060318, "learning_rate": 6.179e-05, "loss": 0.0679, "step": 6180 }, { "epoch": 0.20927009026674329, "grad_norm": 0.4586460590362549, "learning_rate": 6.189e-05, "loss": 0.0726, "step": 6190 }, { "epoch": 0.20960816795699652, "grad_norm": 0.3523881137371063, "learning_rate": 6.199000000000001e-05, "loss": 0.0787, "step": 6200 }, { "epoch": 0.20994624564724973, "grad_norm": 0.2624172270298004, "learning_rate": 6.209000000000001e-05, "loss": 0.0649, "step": 6210 }, { "epoch": 0.21028432333750297, "grad_norm": 0.27350884675979614, "learning_rate": 6.219e-05, "loss": 0.0729, "step": 6220 }, { "epoch": 0.21062240102775617, "grad_norm": 0.32634133100509644, "learning_rate": 6.229e-05, "loss": 0.0704, "step": 6230 }, { "epoch": 0.2109604787180094, "grad_norm": 0.2962538003921509, "learning_rate": 6.239e-05, "loss": 0.0698, "step": 6240 }, { "epoch": 0.21129855640826262, "grad_norm": 0.2855449914932251, "learning_rate": 6.249e-05, "loss": 0.0731, "step": 6250 }, { "epoch": 0.21163663409851585, "grad_norm": 0.2639908194541931, "learning_rate": 6.259000000000001e-05, "loss": 0.0685, "step": 6260 }, { "epoch": 0.21197471178876906, "grad_norm": 0.5626896023750305, "learning_rate": 6.269e-05, "loss": 0.066, "step": 6270 }, { "epoch": 0.21231278947902227, "grad_norm": 0.2542450726032257, "learning_rate": 6.279e-05, "loss": 0.062, "step": 6280 }, { "epoch": 0.2126508671692755, "grad_norm": 0.27140364050865173, "learning_rate": 6.289e-05, "loss": 0.0684, "step": 6290 }, { "epoch": 0.2129889448595287, "grad_norm": 0.3111289441585541, "learning_rate": 6.299e-05, "loss": 0.0697, "step": 6300 }, { "epoch": 0.21332702254978195, "grad_norm": 0.32026198506355286, "learning_rate": 6.309000000000001e-05, "loss": 0.0704, "step": 6310 }, { "epoch": 0.21366510024003516, "grad_norm": 0.39440420269966125, "learning_rate": 6.319000000000001e-05, "loss": 0.0662, "step": 6320 }, { "epoch": 0.2140031779302884, "grad_norm": 0.3738342523574829, "learning_rate": 6.329e-05, "loss": 0.0705, "step": 6330 }, { "epoch": 0.2143412556205416, "grad_norm": 0.4869997203350067, "learning_rate": 6.339e-05, "loss": 0.0729, "step": 6340 }, { "epoch": 0.21467933331079483, "grad_norm": 0.33531689643859863, "learning_rate": 6.349e-05, "loss": 0.0738, "step": 6350 }, { "epoch": 0.21501741100104804, "grad_norm": 0.3285125494003296, "learning_rate": 6.359e-05, "loss": 0.0721, "step": 6360 }, { "epoch": 0.21535548869130125, "grad_norm": 0.3603455722332001, "learning_rate": 6.369000000000001e-05, "loss": 0.0714, "step": 6370 }, { "epoch": 0.21569356638155449, "grad_norm": 0.3449251055717468, "learning_rate": 6.379000000000001e-05, "loss": 0.0642, "step": 6380 }, { "epoch": 0.2160316440718077, "grad_norm": 0.35618093609809875, "learning_rate": 6.389e-05, "loss": 0.0751, "step": 6390 }, { "epoch": 0.21636972176206093, "grad_norm": 0.3669542968273163, "learning_rate": 6.399e-05, "loss": 0.0694, "step": 6400 }, { "epoch": 0.21670779945231414, "grad_norm": 0.3004254400730133, "learning_rate": 6.409e-05, "loss": 0.0631, "step": 6410 }, { "epoch": 0.21704587714256737, "grad_norm": 0.2571595311164856, "learning_rate": 6.419000000000001e-05, "loss": 0.0633, "step": 6420 }, { "epoch": 0.21738395483282058, "grad_norm": 0.35646477341651917, "learning_rate": 6.429000000000001e-05, "loss": 0.062, "step": 6430 }, { "epoch": 0.2177220325230738, "grad_norm": 0.35990244150161743, "learning_rate": 6.439000000000001e-05, "loss": 0.0628, "step": 6440 }, { "epoch": 0.21806011021332702, "grad_norm": 0.47548428177833557, "learning_rate": 6.449e-05, "loss": 0.0698, "step": 6450 }, { "epoch": 0.21839818790358023, "grad_norm": 0.3960473835468292, "learning_rate": 6.459e-05, "loss": 0.0688, "step": 6460 }, { "epoch": 0.21873626559383347, "grad_norm": 0.5407641530036926, "learning_rate": 6.469e-05, "loss": 0.0732, "step": 6470 }, { "epoch": 0.21907434328408668, "grad_norm": 0.3572998642921448, "learning_rate": 6.479000000000001e-05, "loss": 0.0718, "step": 6480 }, { "epoch": 0.2194124209743399, "grad_norm": 0.27968263626098633, "learning_rate": 6.489000000000001e-05, "loss": 0.0689, "step": 6490 }, { "epoch": 0.21975049866459312, "grad_norm": 0.24907104671001434, "learning_rate": 6.499000000000001e-05, "loss": 0.0705, "step": 6500 }, { "epoch": 0.22008857635484635, "grad_norm": 0.44178786873817444, "learning_rate": 6.509e-05, "loss": 0.0703, "step": 6510 }, { "epoch": 0.22042665404509956, "grad_norm": 0.34264519810676575, "learning_rate": 6.519e-05, "loss": 0.0759, "step": 6520 }, { "epoch": 0.22076473173535277, "grad_norm": 0.28655529022216797, "learning_rate": 6.529e-05, "loss": 0.0646, "step": 6530 }, { "epoch": 0.221102809425606, "grad_norm": 0.41998180747032166, "learning_rate": 6.539000000000001e-05, "loss": 0.0664, "step": 6540 }, { "epoch": 0.22144088711585921, "grad_norm": 0.28742510080337524, "learning_rate": 6.549000000000001e-05, "loss": 0.0679, "step": 6550 }, { "epoch": 0.22177896480611245, "grad_norm": 0.232443168759346, "learning_rate": 6.559e-05, "loss": 0.0676, "step": 6560 }, { "epoch": 0.22211704249636566, "grad_norm": 0.3477000892162323, "learning_rate": 6.569e-05, "loss": 0.0658, "step": 6570 }, { "epoch": 0.2224551201866189, "grad_norm": 0.22928251326084137, "learning_rate": 6.579e-05, "loss": 0.0676, "step": 6580 }, { "epoch": 0.2227931978768721, "grad_norm": 0.3809927999973297, "learning_rate": 6.589000000000001e-05, "loss": 0.0649, "step": 6590 }, { "epoch": 0.22313127556712534, "grad_norm": 0.2779715657234192, "learning_rate": 6.599000000000001e-05, "loss": 0.0726, "step": 6600 }, { "epoch": 0.22346935325737854, "grad_norm": 0.2837430238723755, "learning_rate": 6.609000000000001e-05, "loss": 0.0675, "step": 6610 }, { "epoch": 0.22380743094763175, "grad_norm": 0.4974120259284973, "learning_rate": 6.619e-05, "loss": 0.0697, "step": 6620 }, { "epoch": 0.224145508637885, "grad_norm": 0.44487327337265015, "learning_rate": 6.629e-05, "loss": 0.0722, "step": 6630 }, { "epoch": 0.2244835863281382, "grad_norm": 0.2856789231300354, "learning_rate": 6.639e-05, "loss": 0.0679, "step": 6640 }, { "epoch": 0.22482166401839143, "grad_norm": 0.28159263730049133, "learning_rate": 6.649000000000001e-05, "loss": 0.0657, "step": 6650 }, { "epoch": 0.22515974170864464, "grad_norm": 0.329919695854187, "learning_rate": 6.659000000000001e-05, "loss": 0.0691, "step": 6660 }, { "epoch": 0.22549781939889788, "grad_norm": 0.2908368706703186, "learning_rate": 6.669000000000001e-05, "loss": 0.0674, "step": 6670 }, { "epoch": 0.22583589708915108, "grad_norm": 0.27326858043670654, "learning_rate": 6.679e-05, "loss": 0.0682, "step": 6680 }, { "epoch": 0.22617397477940432, "grad_norm": 0.2633392810821533, "learning_rate": 6.689e-05, "loss": 0.0741, "step": 6690 }, { "epoch": 0.22651205246965753, "grad_norm": 0.35240671038627625, "learning_rate": 6.699000000000001e-05, "loss": 0.0785, "step": 6700 }, { "epoch": 0.22685013015991073, "grad_norm": 0.3151485323905945, "learning_rate": 6.709000000000001e-05, "loss": 0.0744, "step": 6710 }, { "epoch": 0.22718820785016397, "grad_norm": 0.2941705584526062, "learning_rate": 6.719000000000001e-05, "loss": 0.068, "step": 6720 }, { "epoch": 0.22752628554041718, "grad_norm": 0.29235053062438965, "learning_rate": 6.729000000000001e-05, "loss": 0.0627, "step": 6730 }, { "epoch": 0.2278643632306704, "grad_norm": 0.28827518224716187, "learning_rate": 6.739e-05, "loss": 0.0697, "step": 6740 }, { "epoch": 0.22820244092092362, "grad_norm": 0.5200936198234558, "learning_rate": 6.749e-05, "loss": 0.0721, "step": 6750 }, { "epoch": 0.22854051861117686, "grad_norm": 0.3257015645503998, "learning_rate": 6.759e-05, "loss": 0.068, "step": 6760 }, { "epoch": 0.22887859630143007, "grad_norm": 0.3212600350379944, "learning_rate": 6.769e-05, "loss": 0.0744, "step": 6770 }, { "epoch": 0.2292166739916833, "grad_norm": 0.4814274311065674, "learning_rate": 6.779e-05, "loss": 0.0678, "step": 6780 }, { "epoch": 0.2295547516819365, "grad_norm": 0.39149340987205505, "learning_rate": 6.789e-05, "loss": 0.0716, "step": 6790 }, { "epoch": 0.22989282937218972, "grad_norm": 0.38954925537109375, "learning_rate": 6.799e-05, "loss": 0.0675, "step": 6800 }, { "epoch": 0.23023090706244295, "grad_norm": 0.35550981760025024, "learning_rate": 6.809e-05, "loss": 0.0713, "step": 6810 }, { "epoch": 0.23056898475269616, "grad_norm": 0.24274340271949768, "learning_rate": 6.819e-05, "loss": 0.0668, "step": 6820 }, { "epoch": 0.2309070624429494, "grad_norm": 0.27658358216285706, "learning_rate": 6.829e-05, "loss": 0.0774, "step": 6830 }, { "epoch": 0.2312451401332026, "grad_norm": 0.28544706106185913, "learning_rate": 6.839e-05, "loss": 0.0621, "step": 6840 }, { "epoch": 0.23158321782345584, "grad_norm": 0.44639384746551514, "learning_rate": 6.849e-05, "loss": 0.0608, "step": 6850 }, { "epoch": 0.23192129551370905, "grad_norm": 0.30973637104034424, "learning_rate": 6.858999999999999e-05, "loss": 0.0614, "step": 6860 }, { "epoch": 0.23225937320396228, "grad_norm": 0.31611305475234985, "learning_rate": 6.869e-05, "loss": 0.0603, "step": 6870 }, { "epoch": 0.2325974508942155, "grad_norm": 0.40065327286720276, "learning_rate": 6.879e-05, "loss": 0.0743, "step": 6880 }, { "epoch": 0.2329355285844687, "grad_norm": 0.4938322901725769, "learning_rate": 6.889e-05, "loss": 0.0622, "step": 6890 }, { "epoch": 0.23327360627472193, "grad_norm": 0.44809800386428833, "learning_rate": 6.899e-05, "loss": 0.071, "step": 6900 }, { "epoch": 0.23361168396497514, "grad_norm": 0.2597593069076538, "learning_rate": 6.909e-05, "loss": 0.0695, "step": 6910 }, { "epoch": 0.23394976165522838, "grad_norm": 0.5095857381820679, "learning_rate": 6.918999999999999e-05, "loss": 0.0751, "step": 6920 }, { "epoch": 0.23428783934548159, "grad_norm": 0.327078253030777, "learning_rate": 6.929e-05, "loss": 0.0747, "step": 6930 }, { "epoch": 0.23462591703573482, "grad_norm": 0.23605452477931976, "learning_rate": 6.939e-05, "loss": 0.0606, "step": 6940 }, { "epoch": 0.23496399472598803, "grad_norm": 0.34553924202919006, "learning_rate": 6.949e-05, "loss": 0.0654, "step": 6950 }, { "epoch": 0.23530207241624126, "grad_norm": 0.30776187777519226, "learning_rate": 6.959e-05, "loss": 0.0608, "step": 6960 }, { "epoch": 0.23564015010649447, "grad_norm": 0.6907650232315063, "learning_rate": 6.969e-05, "loss": 0.0745, "step": 6970 }, { "epoch": 0.23597822779674768, "grad_norm": 0.27309688925743103, "learning_rate": 6.979e-05, "loss": 0.0665, "step": 6980 }, { "epoch": 0.23631630548700092, "grad_norm": 0.22861769795417786, "learning_rate": 6.989e-05, "loss": 0.0641, "step": 6990 }, { "epoch": 0.23665438317725412, "grad_norm": 0.3194275498390198, "learning_rate": 6.999e-05, "loss": 0.0658, "step": 7000 }, { "epoch": 0.23699246086750736, "grad_norm": 0.26254838705062866, "learning_rate": 7.009e-05, "loss": 0.0717, "step": 7010 }, { "epoch": 0.23733053855776057, "grad_norm": 0.27872398495674133, "learning_rate": 7.019e-05, "loss": 0.0681, "step": 7020 }, { "epoch": 0.2376686162480138, "grad_norm": 0.35713207721710205, "learning_rate": 7.029e-05, "loss": 0.0687, "step": 7030 }, { "epoch": 0.238006693938267, "grad_norm": 0.30805280804634094, "learning_rate": 7.039e-05, "loss": 0.0651, "step": 7040 }, { "epoch": 0.23834477162852025, "grad_norm": 0.35623472929000854, "learning_rate": 7.049e-05, "loss": 0.0711, "step": 7050 }, { "epoch": 0.23868284931877345, "grad_norm": 0.3197704553604126, "learning_rate": 7.059e-05, "loss": 0.0676, "step": 7060 }, { "epoch": 0.23902092700902666, "grad_norm": 0.320995956659317, "learning_rate": 7.069e-05, "loss": 0.0731, "step": 7070 }, { "epoch": 0.2393590046992799, "grad_norm": 0.27918705344200134, "learning_rate": 7.079e-05, "loss": 0.0653, "step": 7080 }, { "epoch": 0.2396970823895331, "grad_norm": 0.3738672137260437, "learning_rate": 7.089000000000001e-05, "loss": 0.066, "step": 7090 }, { "epoch": 0.24003516007978634, "grad_norm": 0.2564426362514496, "learning_rate": 7.099e-05, "loss": 0.0665, "step": 7100 }, { "epoch": 0.24037323777003955, "grad_norm": 0.5205355882644653, "learning_rate": 7.109e-05, "loss": 0.0742, "step": 7110 }, { "epoch": 0.24071131546029279, "grad_norm": 0.29816943407058716, "learning_rate": 7.119e-05, "loss": 0.0712, "step": 7120 }, { "epoch": 0.241049393150546, "grad_norm": 0.47140613198280334, "learning_rate": 7.129e-05, "loss": 0.0682, "step": 7130 }, { "epoch": 0.24138747084079923, "grad_norm": 0.32024866342544556, "learning_rate": 7.139e-05, "loss": 0.0733, "step": 7140 }, { "epoch": 0.24172554853105244, "grad_norm": 0.18781711161136627, "learning_rate": 7.149e-05, "loss": 0.0686, "step": 7150 }, { "epoch": 0.24206362622130564, "grad_norm": 0.37728652358055115, "learning_rate": 7.159e-05, "loss": 0.0692, "step": 7160 }, { "epoch": 0.24240170391155888, "grad_norm": 0.40541893243789673, "learning_rate": 7.169e-05, "loss": 0.0643, "step": 7170 }, { "epoch": 0.2427397816018121, "grad_norm": 0.3078837990760803, "learning_rate": 7.179e-05, "loss": 0.0687, "step": 7180 }, { "epoch": 0.24307785929206532, "grad_norm": 0.320976585149765, "learning_rate": 7.189e-05, "loss": 0.0629, "step": 7190 }, { "epoch": 0.24341593698231853, "grad_norm": 0.36311808228492737, "learning_rate": 7.199000000000001e-05, "loss": 0.0747, "step": 7200 }, { "epoch": 0.24375401467257177, "grad_norm": 0.29171377420425415, "learning_rate": 7.209e-05, "loss": 0.0671, "step": 7210 }, { "epoch": 0.24409209236282498, "grad_norm": 0.2846105992794037, "learning_rate": 7.219e-05, "loss": 0.0736, "step": 7220 }, { "epoch": 0.2444301700530782, "grad_norm": 0.3945176303386688, "learning_rate": 7.229e-05, "loss": 0.0719, "step": 7230 }, { "epoch": 0.24476824774333142, "grad_norm": 0.39325809478759766, "learning_rate": 7.239e-05, "loss": 0.0746, "step": 7240 }, { "epoch": 0.24510632543358463, "grad_norm": 0.32509684562683105, "learning_rate": 7.249e-05, "loss": 0.0689, "step": 7250 }, { "epoch": 0.24544440312383786, "grad_norm": 0.2982785105705261, "learning_rate": 7.259000000000001e-05, "loss": 0.0728, "step": 7260 }, { "epoch": 0.24578248081409107, "grad_norm": 0.2781285345554352, "learning_rate": 7.269e-05, "loss": 0.0723, "step": 7270 }, { "epoch": 0.2461205585043443, "grad_norm": 0.3988015353679657, "learning_rate": 7.279e-05, "loss": 0.067, "step": 7280 }, { "epoch": 0.2464586361945975, "grad_norm": 0.4154343605041504, "learning_rate": 7.289e-05, "loss": 0.0703, "step": 7290 }, { "epoch": 0.24679671388485075, "grad_norm": 0.2923842966556549, "learning_rate": 7.299e-05, "loss": 0.0699, "step": 7300 }, { "epoch": 0.24713479157510396, "grad_norm": 0.25413474440574646, "learning_rate": 7.309e-05, "loss": 0.0695, "step": 7310 }, { "epoch": 0.2474728692653572, "grad_norm": 0.3077188730239868, "learning_rate": 7.319000000000001e-05, "loss": 0.0694, "step": 7320 }, { "epoch": 0.2478109469556104, "grad_norm": 0.34239986538887024, "learning_rate": 7.329e-05, "loss": 0.0635, "step": 7330 }, { "epoch": 0.2481490246458636, "grad_norm": 0.2327524721622467, "learning_rate": 7.339e-05, "loss": 0.0691, "step": 7340 }, { "epoch": 0.24848710233611684, "grad_norm": 0.31548014283180237, "learning_rate": 7.349e-05, "loss": 0.07, "step": 7350 }, { "epoch": 0.24882518002637005, "grad_norm": 0.37256065011024475, "learning_rate": 7.359e-05, "loss": 0.073, "step": 7360 }, { "epoch": 0.2491632577166233, "grad_norm": 0.36576929688453674, "learning_rate": 7.369000000000001e-05, "loss": 0.0725, "step": 7370 }, { "epoch": 0.2495013354068765, "grad_norm": 0.3891449272632599, "learning_rate": 7.379000000000001e-05, "loss": 0.0667, "step": 7380 }, { "epoch": 0.24983941309712973, "grad_norm": 0.32940050959587097, "learning_rate": 7.389e-05, "loss": 0.074, "step": 7390 }, { "epoch": 0.25017749078738294, "grad_norm": 0.3043605089187622, "learning_rate": 7.399e-05, "loss": 0.0694, "step": 7400 }, { "epoch": 0.2505155684776362, "grad_norm": 0.2763773500919342, "learning_rate": 7.409e-05, "loss": 0.0653, "step": 7410 }, { "epoch": 0.25085364616788935, "grad_norm": 0.30281588435173035, "learning_rate": 7.419e-05, "loss": 0.0639, "step": 7420 }, { "epoch": 0.2511917238581426, "grad_norm": 0.4993593990802765, "learning_rate": 7.429000000000001e-05, "loss": 0.0659, "step": 7430 }, { "epoch": 0.2515298015483958, "grad_norm": 0.2821089029312134, "learning_rate": 7.439e-05, "loss": 0.0706, "step": 7440 }, { "epoch": 0.25186787923864906, "grad_norm": 0.26409825682640076, "learning_rate": 7.449e-05, "loss": 0.0654, "step": 7450 }, { "epoch": 0.25220595692890224, "grad_norm": 0.44840696454048157, "learning_rate": 7.459e-05, "loss": 0.0769, "step": 7460 }, { "epoch": 0.2525440346191555, "grad_norm": 0.4126066267490387, "learning_rate": 7.469e-05, "loss": 0.0693, "step": 7470 }, { "epoch": 0.2528821123094087, "grad_norm": 0.43710076808929443, "learning_rate": 7.479000000000001e-05, "loss": 0.0635, "step": 7480 }, { "epoch": 0.25322018999966195, "grad_norm": 0.5248191356658936, "learning_rate": 7.489000000000001e-05, "loss": 0.0623, "step": 7490 }, { "epoch": 0.25355826768991513, "grad_norm": 0.3156410753726959, "learning_rate": 7.499e-05, "loss": 0.0665, "step": 7500 }, { "epoch": 0.25389634538016836, "grad_norm": 0.3067738115787506, "learning_rate": 7.509e-05, "loss": 0.0683, "step": 7510 }, { "epoch": 0.2542344230704216, "grad_norm": 0.4044918417930603, "learning_rate": 7.519e-05, "loss": 0.0719, "step": 7520 }, { "epoch": 0.2545725007606748, "grad_norm": 0.2621394693851471, "learning_rate": 7.529e-05, "loss": 0.0678, "step": 7530 }, { "epoch": 0.254910578450928, "grad_norm": 0.3368181586265564, "learning_rate": 7.539000000000001e-05, "loss": 0.0638, "step": 7540 }, { "epoch": 0.25524865614118125, "grad_norm": 0.2719883620738983, "learning_rate": 7.549000000000001e-05, "loss": 0.0636, "step": 7550 }, { "epoch": 0.2555867338314345, "grad_norm": 0.2867761552333832, "learning_rate": 7.559e-05, "loss": 0.0805, "step": 7560 }, { "epoch": 0.25592481152168767, "grad_norm": 0.29021498560905457, "learning_rate": 7.569e-05, "loss": 0.0691, "step": 7570 }, { "epoch": 0.2562628892119409, "grad_norm": 0.20315654575824738, "learning_rate": 7.579e-05, "loss": 0.0706, "step": 7580 }, { "epoch": 0.25660096690219414, "grad_norm": 0.2738615870475769, "learning_rate": 7.589000000000001e-05, "loss": 0.0716, "step": 7590 }, { "epoch": 0.2569390445924473, "grad_norm": 0.21146629750728607, "learning_rate": 7.599000000000001e-05, "loss": 0.0731, "step": 7600 }, { "epoch": 0.25727712228270055, "grad_norm": 0.3927782475948334, "learning_rate": 7.609000000000001e-05, "loss": 0.067, "step": 7610 }, { "epoch": 0.2576151999729538, "grad_norm": 0.3974152207374573, "learning_rate": 7.619e-05, "loss": 0.0651, "step": 7620 }, { "epoch": 0.257953277663207, "grad_norm": 0.4025062620639801, "learning_rate": 7.629e-05, "loss": 0.0728, "step": 7630 }, { "epoch": 0.2582913553534602, "grad_norm": 0.252300888299942, "learning_rate": 7.639e-05, "loss": 0.0625, "step": 7640 }, { "epoch": 0.25862943304371344, "grad_norm": 0.27880004048347473, "learning_rate": 7.649000000000001e-05, "loss": 0.0705, "step": 7650 }, { "epoch": 0.2589675107339667, "grad_norm": 0.36054909229278564, "learning_rate": 7.659000000000001e-05, "loss": 0.0665, "step": 7660 }, { "epoch": 0.2593055884242199, "grad_norm": 0.2660887837409973, "learning_rate": 7.669000000000001e-05, "loss": 0.0653, "step": 7670 }, { "epoch": 0.2596436661144731, "grad_norm": 0.32438603043556213, "learning_rate": 7.679e-05, "loss": 0.0705, "step": 7680 }, { "epoch": 0.25998174380472633, "grad_norm": 0.3285495340824127, "learning_rate": 7.689e-05, "loss": 0.0689, "step": 7690 }, { "epoch": 0.26031982149497956, "grad_norm": 0.3651605248451233, "learning_rate": 7.699e-05, "loss": 0.0701, "step": 7700 }, { "epoch": 0.26065789918523274, "grad_norm": 0.26248911023139954, "learning_rate": 7.709000000000001e-05, "loss": 0.0727, "step": 7710 }, { "epoch": 0.260995976875486, "grad_norm": 0.3138735890388489, "learning_rate": 7.719000000000001e-05, "loss": 0.0662, "step": 7720 }, { "epoch": 0.2613340545657392, "grad_norm": 0.43455392122268677, "learning_rate": 7.729e-05, "loss": 0.0659, "step": 7730 }, { "epoch": 0.26167213225599245, "grad_norm": 0.3326144516468048, "learning_rate": 7.739e-05, "loss": 0.0675, "step": 7740 }, { "epoch": 0.26201020994624563, "grad_norm": 0.34104111790657043, "learning_rate": 7.749e-05, "loss": 0.0699, "step": 7750 }, { "epoch": 0.26234828763649887, "grad_norm": 0.34505560994148254, "learning_rate": 7.759000000000001e-05, "loss": 0.0666, "step": 7760 }, { "epoch": 0.2626863653267521, "grad_norm": 0.3364699184894562, "learning_rate": 7.769000000000001e-05, "loss": 0.0738, "step": 7770 }, { "epoch": 0.2630244430170053, "grad_norm": 0.295213907957077, "learning_rate": 7.779000000000001e-05, "loss": 0.0648, "step": 7780 }, { "epoch": 0.2633625207072585, "grad_norm": 0.24595396220684052, "learning_rate": 7.789e-05, "loss": 0.065, "step": 7790 }, { "epoch": 0.26370059839751175, "grad_norm": 0.379081666469574, "learning_rate": 7.799e-05, "loss": 0.0697, "step": 7800 }, { "epoch": 0.264038676087765, "grad_norm": 0.2680777311325073, "learning_rate": 7.809e-05, "loss": 0.0694, "step": 7810 }, { "epoch": 0.26437675377801817, "grad_norm": 0.23461279273033142, "learning_rate": 7.819000000000001e-05, "loss": 0.0723, "step": 7820 }, { "epoch": 0.2647148314682714, "grad_norm": 0.2570856511592865, "learning_rate": 7.829000000000001e-05, "loss": 0.0701, "step": 7830 }, { "epoch": 0.26505290915852464, "grad_norm": 0.206537127494812, "learning_rate": 7.839000000000001e-05, "loss": 0.0682, "step": 7840 }, { "epoch": 0.2653909868487779, "grad_norm": 0.2627721130847931, "learning_rate": 7.849e-05, "loss": 0.07, "step": 7850 }, { "epoch": 0.26572906453903106, "grad_norm": 0.4095202684402466, "learning_rate": 7.859e-05, "loss": 0.0666, "step": 7860 }, { "epoch": 0.2660671422292843, "grad_norm": 0.20941218733787537, "learning_rate": 7.869000000000001e-05, "loss": 0.0727, "step": 7870 }, { "epoch": 0.26640521991953753, "grad_norm": 0.27622470259666443, "learning_rate": 7.879000000000001e-05, "loss": 0.0599, "step": 7880 }, { "epoch": 0.2667432976097907, "grad_norm": 0.3582231402397156, "learning_rate": 7.889000000000001e-05, "loss": 0.0724, "step": 7890 }, { "epoch": 0.26708137530004394, "grad_norm": 0.3218744695186615, "learning_rate": 7.899000000000001e-05, "loss": 0.0663, "step": 7900 }, { "epoch": 0.2674194529902972, "grad_norm": 0.40544936060905457, "learning_rate": 7.909e-05, "loss": 0.082, "step": 7910 }, { "epoch": 0.2677575306805504, "grad_norm": 0.697718620300293, "learning_rate": 7.919e-05, "loss": 0.0718, "step": 7920 }, { "epoch": 0.2680956083708036, "grad_norm": 0.41850176453590393, "learning_rate": 7.929000000000001e-05, "loss": 0.0749, "step": 7930 }, { "epoch": 0.26843368606105683, "grad_norm": 0.34039461612701416, "learning_rate": 7.939000000000001e-05, "loss": 0.0717, "step": 7940 }, { "epoch": 0.26877176375131007, "grad_norm": 0.3782693147659302, "learning_rate": 7.949000000000001e-05, "loss": 0.0765, "step": 7950 }, { "epoch": 0.26910984144156325, "grad_norm": 0.5013105273246765, "learning_rate": 7.959000000000002e-05, "loss": 0.0743, "step": 7960 }, { "epoch": 0.2694479191318165, "grad_norm": 0.29532310366630554, "learning_rate": 7.969e-05, "loss": 0.0681, "step": 7970 }, { "epoch": 0.2697859968220697, "grad_norm": 0.3333141803741455, "learning_rate": 7.979000000000001e-05, "loss": 0.0741, "step": 7980 }, { "epoch": 0.27012407451232295, "grad_norm": 0.26732656359672546, "learning_rate": 7.989000000000001e-05, "loss": 0.0643, "step": 7990 }, { "epoch": 0.27046215220257613, "grad_norm": 0.3030105233192444, "learning_rate": 7.999000000000001e-05, "loss": 0.0568, "step": 8000 }, { "epoch": 0.27080022989282937, "grad_norm": 0.40968945622444153, "learning_rate": 8.009e-05, "loss": 0.0675, "step": 8010 }, { "epoch": 0.2711383075830826, "grad_norm": 0.30035388469696045, "learning_rate": 8.019e-05, "loss": 0.067, "step": 8020 }, { "epoch": 0.2714763852733358, "grad_norm": 0.4618229269981384, "learning_rate": 8.028999999999999e-05, "loss": 0.0721, "step": 8030 }, { "epoch": 0.271814462963589, "grad_norm": 0.2635803818702698, "learning_rate": 8.039e-05, "loss": 0.0715, "step": 8040 }, { "epoch": 0.27215254065384226, "grad_norm": 0.31167617440223694, "learning_rate": 8.049e-05, "loss": 0.0687, "step": 8050 }, { "epoch": 0.2724906183440955, "grad_norm": 0.32386308908462524, "learning_rate": 8.059e-05, "loss": 0.0686, "step": 8060 }, { "epoch": 0.27282869603434867, "grad_norm": 0.2702082693576813, "learning_rate": 8.069e-05, "loss": 0.0623, "step": 8070 }, { "epoch": 0.2731667737246019, "grad_norm": 0.30954623222351074, "learning_rate": 8.079e-05, "loss": 0.0755, "step": 8080 }, { "epoch": 0.27350485141485514, "grad_norm": 0.3286321461200714, "learning_rate": 8.088999999999999e-05, "loss": 0.0677, "step": 8090 }, { "epoch": 0.2738429291051084, "grad_norm": 0.35454556345939636, "learning_rate": 8.099e-05, "loss": 0.0672, "step": 8100 }, { "epoch": 0.27418100679536156, "grad_norm": 0.29060155153274536, "learning_rate": 8.109e-05, "loss": 0.0734, "step": 8110 }, { "epoch": 0.2745190844856148, "grad_norm": 0.28590765595436096, "learning_rate": 8.119e-05, "loss": 0.0644, "step": 8120 }, { "epoch": 0.27485716217586803, "grad_norm": 0.28829628229141235, "learning_rate": 8.129e-05, "loss": 0.066, "step": 8130 }, { "epoch": 0.2751952398661212, "grad_norm": 0.34524381160736084, "learning_rate": 8.139e-05, "loss": 0.067, "step": 8140 }, { "epoch": 0.27553331755637445, "grad_norm": 0.3556766211986542, "learning_rate": 8.149e-05, "loss": 0.0684, "step": 8150 }, { "epoch": 0.2758713952466277, "grad_norm": 0.36136457324028015, "learning_rate": 8.159e-05, "loss": 0.0702, "step": 8160 }, { "epoch": 0.2762094729368809, "grad_norm": 0.4080522656440735, "learning_rate": 8.169e-05, "loss": 0.0679, "step": 8170 }, { "epoch": 0.2765475506271341, "grad_norm": 0.22762510180473328, "learning_rate": 8.179e-05, "loss": 0.0683, "step": 8180 }, { "epoch": 0.27688562831738733, "grad_norm": 0.4325878918170929, "learning_rate": 8.189e-05, "loss": 0.0661, "step": 8190 }, { "epoch": 0.27722370600764057, "grad_norm": 0.439698725938797, "learning_rate": 8.199e-05, "loss": 0.0645, "step": 8200 }, { "epoch": 0.27756178369789375, "grad_norm": 0.45239847898483276, "learning_rate": 8.209e-05, "loss": 0.0624, "step": 8210 }, { "epoch": 0.277899861388147, "grad_norm": 0.3857632279396057, "learning_rate": 8.219e-05, "loss": 0.0637, "step": 8220 }, { "epoch": 0.2782379390784002, "grad_norm": 0.23765477538108826, "learning_rate": 8.229e-05, "loss": 0.0753, "step": 8230 }, { "epoch": 0.27857601676865346, "grad_norm": 0.2808893322944641, "learning_rate": 8.239e-05, "loss": 0.0755, "step": 8240 }, { "epoch": 0.27891409445890664, "grad_norm": 0.4180174469947815, "learning_rate": 8.249e-05, "loss": 0.0863, "step": 8250 }, { "epoch": 0.27925217214915987, "grad_norm": 0.2552604079246521, "learning_rate": 8.259000000000001e-05, "loss": 0.0735, "step": 8260 }, { "epoch": 0.2795902498394131, "grad_norm": 0.33300554752349854, "learning_rate": 8.269e-05, "loss": 0.0669, "step": 8270 }, { "epoch": 0.27992832752966634, "grad_norm": 0.32194697856903076, "learning_rate": 8.279e-05, "loss": 0.0681, "step": 8280 }, { "epoch": 0.2802664052199195, "grad_norm": 0.38149547576904297, "learning_rate": 8.289e-05, "loss": 0.0709, "step": 8290 }, { "epoch": 0.28060448291017276, "grad_norm": 0.23168662190437317, "learning_rate": 8.299e-05, "loss": 0.0629, "step": 8300 }, { "epoch": 0.280942560600426, "grad_norm": 0.42463773488998413, "learning_rate": 8.309e-05, "loss": 0.0729, "step": 8310 }, { "epoch": 0.2812806382906792, "grad_norm": 0.2275780737400055, "learning_rate": 8.319e-05, "loss": 0.0742, "step": 8320 }, { "epoch": 0.2816187159809324, "grad_norm": 0.2858680188655853, "learning_rate": 8.329e-05, "loss": 0.0729, "step": 8330 }, { "epoch": 0.28195679367118565, "grad_norm": 0.23265355825424194, "learning_rate": 8.339e-05, "loss": 0.0684, "step": 8340 }, { "epoch": 0.2822948713614389, "grad_norm": 0.4039391279220581, "learning_rate": 8.349e-05, "loss": 0.072, "step": 8350 }, { "epoch": 0.28263294905169206, "grad_norm": 0.3360477089881897, "learning_rate": 8.359e-05, "loss": 0.0751, "step": 8360 }, { "epoch": 0.2829710267419453, "grad_norm": 0.24867786467075348, "learning_rate": 8.369000000000001e-05, "loss": 0.076, "step": 8370 }, { "epoch": 0.28330910443219853, "grad_norm": 0.283092737197876, "learning_rate": 8.379e-05, "loss": 0.0663, "step": 8380 }, { "epoch": 0.2836471821224517, "grad_norm": 0.32940396666526794, "learning_rate": 8.389e-05, "loss": 0.0683, "step": 8390 }, { "epoch": 0.28398525981270495, "grad_norm": 0.5569860339164734, "learning_rate": 8.399e-05, "loss": 0.0679, "step": 8400 }, { "epoch": 0.2843233375029582, "grad_norm": 0.29096540808677673, "learning_rate": 8.409e-05, "loss": 0.0673, "step": 8410 }, { "epoch": 0.2846614151932114, "grad_norm": 0.39714276790618896, "learning_rate": 8.419e-05, "loss": 0.0636, "step": 8420 }, { "epoch": 0.2849994928834646, "grad_norm": 0.30451124906539917, "learning_rate": 8.429000000000001e-05, "loss": 0.0625, "step": 8430 }, { "epoch": 0.28533757057371784, "grad_norm": 0.22238442301750183, "learning_rate": 8.439e-05, "loss": 0.0649, "step": 8440 }, { "epoch": 0.28567564826397107, "grad_norm": 0.33830371499061584, "learning_rate": 8.449e-05, "loss": 0.0763, "step": 8450 }, { "epoch": 0.2860137259542243, "grad_norm": 0.3885985314846039, "learning_rate": 8.459e-05, "loss": 0.0685, "step": 8460 }, { "epoch": 0.2863518036444775, "grad_norm": 0.22599542140960693, "learning_rate": 8.469e-05, "loss": 0.066, "step": 8470 }, { "epoch": 0.2866898813347307, "grad_norm": 0.3364439010620117, "learning_rate": 8.479e-05, "loss": 0.0674, "step": 8480 }, { "epoch": 0.28702795902498396, "grad_norm": 0.27713045477867126, "learning_rate": 8.489000000000001e-05, "loss": 0.0671, "step": 8490 }, { "epoch": 0.28736603671523714, "grad_norm": 0.4968343675136566, "learning_rate": 8.499e-05, "loss": 0.0687, "step": 8500 }, { "epoch": 0.2877041144054904, "grad_norm": 0.3354155123233795, "learning_rate": 8.509e-05, "loss": 0.0668, "step": 8510 }, { "epoch": 0.2880421920957436, "grad_norm": 0.26619088649749756, "learning_rate": 8.519e-05, "loss": 0.0694, "step": 8520 }, { "epoch": 0.28838026978599685, "grad_norm": 0.318217933177948, "learning_rate": 8.529e-05, "loss": 0.0641, "step": 8530 }, { "epoch": 0.28871834747625, "grad_norm": 0.3245730400085449, "learning_rate": 8.539000000000001e-05, "loss": 0.0685, "step": 8540 }, { "epoch": 0.28905642516650326, "grad_norm": 0.3404865860939026, "learning_rate": 8.549000000000001e-05, "loss": 0.0649, "step": 8550 }, { "epoch": 0.2893945028567565, "grad_norm": 0.23281866312026978, "learning_rate": 8.559e-05, "loss": 0.0735, "step": 8560 }, { "epoch": 0.2897325805470097, "grad_norm": 0.6995651125907898, "learning_rate": 8.569e-05, "loss": 0.0795, "step": 8570 }, { "epoch": 0.2900706582372629, "grad_norm": 0.24809429049491882, "learning_rate": 8.579e-05, "loss": 0.0722, "step": 8580 }, { "epoch": 0.29040873592751615, "grad_norm": 0.2584843337535858, "learning_rate": 8.589e-05, "loss": 0.0705, "step": 8590 }, { "epoch": 0.2907468136177694, "grad_norm": 0.3336651027202606, "learning_rate": 8.599000000000001e-05, "loss": 0.0628, "step": 8600 }, { "epoch": 0.29108489130802256, "grad_norm": 0.32893243432044983, "learning_rate": 8.609e-05, "loss": 0.0718, "step": 8610 }, { "epoch": 0.2914229689982758, "grad_norm": 0.37621694803237915, "learning_rate": 8.619e-05, "loss": 0.0687, "step": 8620 }, { "epoch": 0.29176104668852904, "grad_norm": 0.3196501135826111, "learning_rate": 8.629e-05, "loss": 0.068, "step": 8630 }, { "epoch": 0.29209912437878227, "grad_norm": 0.27913302183151245, "learning_rate": 8.639e-05, "loss": 0.0698, "step": 8640 }, { "epoch": 0.29243720206903545, "grad_norm": 0.5410332679748535, "learning_rate": 8.649000000000001e-05, "loss": 0.0725, "step": 8650 }, { "epoch": 0.2927752797592887, "grad_norm": 0.34034785628318787, "learning_rate": 8.659000000000001e-05, "loss": 0.0762, "step": 8660 }, { "epoch": 0.2931133574495419, "grad_norm": 0.2913448214530945, "learning_rate": 8.669e-05, "loss": 0.0716, "step": 8670 }, { "epoch": 0.2934514351397951, "grad_norm": 0.44594353437423706, "learning_rate": 8.679e-05, "loss": 0.0706, "step": 8680 }, { "epoch": 0.29378951283004834, "grad_norm": 0.3610019385814667, "learning_rate": 8.689e-05, "loss": 0.0622, "step": 8690 }, { "epoch": 0.2941275905203016, "grad_norm": 0.30438023805618286, "learning_rate": 8.699e-05, "loss": 0.0699, "step": 8700 }, { "epoch": 0.2944656682105548, "grad_norm": 0.25995349884033203, "learning_rate": 8.709000000000001e-05, "loss": 0.0681, "step": 8710 }, { "epoch": 0.294803745900808, "grad_norm": 0.28912079334259033, "learning_rate": 8.719000000000001e-05, "loss": 0.0714, "step": 8720 }, { "epoch": 0.2951418235910612, "grad_norm": 0.33167964220046997, "learning_rate": 8.729e-05, "loss": 0.0685, "step": 8730 }, { "epoch": 0.29547990128131446, "grad_norm": 0.3273722231388092, "learning_rate": 8.739e-05, "loss": 0.0644, "step": 8740 }, { "epoch": 0.29581797897156764, "grad_norm": 0.3132167458534241, "learning_rate": 8.749e-05, "loss": 0.0638, "step": 8750 }, { "epoch": 0.2961560566618209, "grad_norm": 0.26827940344810486, "learning_rate": 8.759e-05, "loss": 0.0667, "step": 8760 }, { "epoch": 0.2964941343520741, "grad_norm": 0.40028417110443115, "learning_rate": 8.769000000000001e-05, "loss": 0.0693, "step": 8770 }, { "epoch": 0.29683221204232735, "grad_norm": 0.3169361352920532, "learning_rate": 8.779000000000001e-05, "loss": 0.071, "step": 8780 }, { "epoch": 0.29717028973258053, "grad_norm": 0.34033095836639404, "learning_rate": 8.789e-05, "loss": 0.0672, "step": 8790 }, { "epoch": 0.29750836742283376, "grad_norm": 0.21894773840904236, "learning_rate": 8.799e-05, "loss": 0.0583, "step": 8800 }, { "epoch": 0.297846445113087, "grad_norm": 0.33015134930610657, "learning_rate": 8.809e-05, "loss": 0.0686, "step": 8810 }, { "epoch": 0.29818452280334024, "grad_norm": 0.3059764504432678, "learning_rate": 8.819000000000001e-05, "loss": 0.0716, "step": 8820 }, { "epoch": 0.2985226004935934, "grad_norm": 0.23112836480140686, "learning_rate": 8.829000000000001e-05, "loss": 0.0665, "step": 8830 }, { "epoch": 0.29886067818384665, "grad_norm": 0.298313170671463, "learning_rate": 8.839000000000001e-05, "loss": 0.0717, "step": 8840 }, { "epoch": 0.2991987558740999, "grad_norm": 0.31452152132987976, "learning_rate": 8.849e-05, "loss": 0.0689, "step": 8850 }, { "epoch": 0.29953683356435307, "grad_norm": 0.39079439640045166, "learning_rate": 8.859e-05, "loss": 0.0706, "step": 8860 }, { "epoch": 0.2998749112546063, "grad_norm": 0.2616451680660248, "learning_rate": 8.869e-05, "loss": 0.066, "step": 8870 }, { "epoch": 0.30021298894485954, "grad_norm": 0.22371743619441986, "learning_rate": 8.879000000000001e-05, "loss": 0.0676, "step": 8880 }, { "epoch": 0.3005510666351128, "grad_norm": 0.2149435132741928, "learning_rate": 8.889000000000001e-05, "loss": 0.0679, "step": 8890 }, { "epoch": 0.30088914432536595, "grad_norm": 0.37787923216819763, "learning_rate": 8.899e-05, "loss": 0.0635, "step": 8900 }, { "epoch": 0.3012272220156192, "grad_norm": 0.30486005544662476, "learning_rate": 8.909e-05, "loss": 0.0643, "step": 8910 }, { "epoch": 0.3015652997058724, "grad_norm": 0.28861165046691895, "learning_rate": 8.919e-05, "loss": 0.0699, "step": 8920 }, { "epoch": 0.3019033773961256, "grad_norm": 0.2383137345314026, "learning_rate": 8.929000000000001e-05, "loss": 0.0699, "step": 8930 }, { "epoch": 0.30224145508637884, "grad_norm": 0.3541710078716278, "learning_rate": 8.939000000000001e-05, "loss": 0.0654, "step": 8940 }, { "epoch": 0.3025795327766321, "grad_norm": 0.2483447790145874, "learning_rate": 8.949000000000001e-05, "loss": 0.0597, "step": 8950 }, { "epoch": 0.3029176104668853, "grad_norm": 0.31865012645721436, "learning_rate": 8.959e-05, "loss": 0.0665, "step": 8960 }, { "epoch": 0.3032556881571385, "grad_norm": 0.4610338509082794, "learning_rate": 8.969e-05, "loss": 0.0742, "step": 8970 }, { "epoch": 0.30359376584739173, "grad_norm": 0.4005155861377716, "learning_rate": 8.979e-05, "loss": 0.0667, "step": 8980 }, { "epoch": 0.30393184353764496, "grad_norm": 0.23349949717521667, "learning_rate": 8.989000000000001e-05, "loss": 0.0669, "step": 8990 }, { "epoch": 0.30426992122789814, "grad_norm": 0.40851765871047974, "learning_rate": 8.999000000000001e-05, "loss": 0.0651, "step": 9000 }, { "epoch": 0.3046079989181514, "grad_norm": 0.47475510835647583, "learning_rate": 9.009000000000001e-05, "loss": 0.0673, "step": 9010 }, { "epoch": 0.3049460766084046, "grad_norm": 0.4119532108306885, "learning_rate": 9.019e-05, "loss": 0.0753, "step": 9020 }, { "epoch": 0.30528415429865785, "grad_norm": 0.33090198040008545, "learning_rate": 9.029e-05, "loss": 0.0689, "step": 9030 }, { "epoch": 0.30562223198891103, "grad_norm": 0.4332920014858246, "learning_rate": 9.039000000000001e-05, "loss": 0.0681, "step": 9040 }, { "epoch": 0.30596030967916427, "grad_norm": 0.2922382652759552, "learning_rate": 9.049000000000001e-05, "loss": 0.0755, "step": 9050 }, { "epoch": 0.3062983873694175, "grad_norm": 0.349718302488327, "learning_rate": 9.059000000000001e-05, "loss": 0.0635, "step": 9060 }, { "epoch": 0.30663646505967074, "grad_norm": 0.30649712681770325, "learning_rate": 9.069000000000001e-05, "loss": 0.0686, "step": 9070 }, { "epoch": 0.3069745427499239, "grad_norm": 0.33038339018821716, "learning_rate": 9.079e-05, "loss": 0.0685, "step": 9080 }, { "epoch": 0.30731262044017715, "grad_norm": 0.2973896563053131, "learning_rate": 9.089e-05, "loss": 0.0683, "step": 9090 }, { "epoch": 0.3076506981304304, "grad_norm": 0.23122082650661469, "learning_rate": 9.099000000000001e-05, "loss": 0.0701, "step": 9100 }, { "epoch": 0.30798877582068357, "grad_norm": 0.3072817623615265, "learning_rate": 9.109000000000001e-05, "loss": 0.0608, "step": 9110 }, { "epoch": 0.3083268535109368, "grad_norm": 0.5837863087654114, "learning_rate": 9.119000000000001e-05, "loss": 0.0745, "step": 9120 }, { "epoch": 0.30866493120119004, "grad_norm": 0.31724730134010315, "learning_rate": 9.129000000000002e-05, "loss": 0.0676, "step": 9130 }, { "epoch": 0.3090030088914433, "grad_norm": 0.3243112862110138, "learning_rate": 9.139e-05, "loss": 0.061, "step": 9140 }, { "epoch": 0.30934108658169646, "grad_norm": 0.32939770817756653, "learning_rate": 9.149e-05, "loss": 0.0673, "step": 9150 }, { "epoch": 0.3096791642719497, "grad_norm": 0.4169028699398041, "learning_rate": 9.159000000000001e-05, "loss": 0.0655, "step": 9160 }, { "epoch": 0.3100172419622029, "grad_norm": 0.28733301162719727, "learning_rate": 9.169000000000001e-05, "loss": 0.0702, "step": 9170 }, { "epoch": 0.3103553196524561, "grad_norm": 0.16336579620838165, "learning_rate": 9.179000000000001e-05, "loss": 0.0704, "step": 9180 }, { "epoch": 0.31069339734270934, "grad_norm": 0.2996024191379547, "learning_rate": 9.189e-05, "loss": 0.0681, "step": 9190 }, { "epoch": 0.3110314750329626, "grad_norm": 0.30453792214393616, "learning_rate": 9.199e-05, "loss": 0.0691, "step": 9200 }, { "epoch": 0.3113695527232158, "grad_norm": 0.39805346727371216, "learning_rate": 9.209000000000001e-05, "loss": 0.0737, "step": 9210 }, { "epoch": 0.311707630413469, "grad_norm": 0.31673887372016907, "learning_rate": 9.219000000000001e-05, "loss": 0.0645, "step": 9220 }, { "epoch": 0.31204570810372223, "grad_norm": 0.2978845536708832, "learning_rate": 9.229000000000001e-05, "loss": 0.0661, "step": 9230 }, { "epoch": 0.31238378579397547, "grad_norm": 0.35443148016929626, "learning_rate": 9.239000000000001e-05, "loss": 0.068, "step": 9240 }, { "epoch": 0.3127218634842287, "grad_norm": 0.3034539222717285, "learning_rate": 9.249e-05, "loss": 0.0825, "step": 9250 }, { "epoch": 0.3130599411744819, "grad_norm": 0.3850052058696747, "learning_rate": 9.258999999999999e-05, "loss": 0.0751, "step": 9260 }, { "epoch": 0.3133980188647351, "grad_norm": 0.37645700573921204, "learning_rate": 9.269e-05, "loss": 0.0744, "step": 9270 }, { "epoch": 0.31373609655498835, "grad_norm": 0.29464560747146606, "learning_rate": 9.279e-05, "loss": 0.0647, "step": 9280 }, { "epoch": 0.31407417424524153, "grad_norm": 0.2612125277519226, "learning_rate": 9.289e-05, "loss": 0.0693, "step": 9290 }, { "epoch": 0.31441225193549477, "grad_norm": 0.22448718547821045, "learning_rate": 9.299e-05, "loss": 0.0699, "step": 9300 }, { "epoch": 0.314750329625748, "grad_norm": 0.4281945824623108, "learning_rate": 9.309e-05, "loss": 0.0724, "step": 9310 }, { "epoch": 0.31508840731600124, "grad_norm": 0.2777540981769562, "learning_rate": 9.319e-05, "loss": 0.0673, "step": 9320 }, { "epoch": 0.3154264850062544, "grad_norm": 0.25459522008895874, "learning_rate": 9.329e-05, "loss": 0.0711, "step": 9330 }, { "epoch": 0.31576456269650766, "grad_norm": 0.4329175353050232, "learning_rate": 9.339e-05, "loss": 0.0728, "step": 9340 }, { "epoch": 0.3161026403867609, "grad_norm": 0.5475261807441711, "learning_rate": 9.349e-05, "loss": 0.0699, "step": 9350 }, { "epoch": 0.31644071807701407, "grad_norm": 0.3281181752681732, "learning_rate": 9.359e-05, "loss": 0.0691, "step": 9360 }, { "epoch": 0.3167787957672673, "grad_norm": 0.4091145992279053, "learning_rate": 9.369e-05, "loss": 0.0694, "step": 9370 }, { "epoch": 0.31711687345752054, "grad_norm": 0.2761368155479431, "learning_rate": 9.379e-05, "loss": 0.0614, "step": 9380 }, { "epoch": 0.3174549511477738, "grad_norm": 0.22302541136741638, "learning_rate": 9.389e-05, "loss": 0.0642, "step": 9390 }, { "epoch": 0.31779302883802696, "grad_norm": 0.39143866300582886, "learning_rate": 9.399e-05, "loss": 0.072, "step": 9400 }, { "epoch": 0.3181311065282802, "grad_norm": 0.24242763221263885, "learning_rate": 9.409e-05, "loss": 0.0681, "step": 9410 }, { "epoch": 0.31846918421853343, "grad_norm": 0.23688091337680817, "learning_rate": 9.419e-05, "loss": 0.0618, "step": 9420 }, { "epoch": 0.31880726190878667, "grad_norm": 0.23550380766391754, "learning_rate": 9.429000000000001e-05, "loss": 0.0621, "step": 9430 }, { "epoch": 0.31914533959903985, "grad_norm": 0.27469727396965027, "learning_rate": 9.439e-05, "loss": 0.0617, "step": 9440 }, { "epoch": 0.3194834172892931, "grad_norm": 0.37763306498527527, "learning_rate": 9.449e-05, "loss": 0.0729, "step": 9450 }, { "epoch": 0.3198214949795463, "grad_norm": 0.25825950503349304, "learning_rate": 9.459e-05, "loss": 0.0681, "step": 9460 }, { "epoch": 0.3201595726697995, "grad_norm": 0.3224400579929352, "learning_rate": 9.469e-05, "loss": 0.0676, "step": 9470 }, { "epoch": 0.32049765036005273, "grad_norm": 0.343975692987442, "learning_rate": 9.479e-05, "loss": 0.0675, "step": 9480 }, { "epoch": 0.32083572805030597, "grad_norm": 0.25197160243988037, "learning_rate": 9.489e-05, "loss": 0.0679, "step": 9490 }, { "epoch": 0.3211738057405592, "grad_norm": 0.3240128457546234, "learning_rate": 9.499e-05, "loss": 0.064, "step": 9500 }, { "epoch": 0.3215118834308124, "grad_norm": 0.5212570428848267, "learning_rate": 9.509e-05, "loss": 0.0723, "step": 9510 }, { "epoch": 0.3218499611210656, "grad_norm": 0.28149741888046265, "learning_rate": 9.519e-05, "loss": 0.0601, "step": 9520 }, { "epoch": 0.32218803881131886, "grad_norm": 0.27982354164123535, "learning_rate": 9.529e-05, "loss": 0.0691, "step": 9530 }, { "epoch": 0.32252611650157204, "grad_norm": 0.4243796467781067, "learning_rate": 9.539e-05, "loss": 0.0652, "step": 9540 }, { "epoch": 0.32286419419182527, "grad_norm": 0.3372665047645569, "learning_rate": 9.549e-05, "loss": 0.0711, "step": 9550 }, { "epoch": 0.3232022718820785, "grad_norm": 0.22867542505264282, "learning_rate": 9.559e-05, "loss": 0.0657, "step": 9560 }, { "epoch": 0.32354034957233174, "grad_norm": 0.34174251556396484, "learning_rate": 9.569e-05, "loss": 0.0638, "step": 9570 }, { "epoch": 0.3238784272625849, "grad_norm": 0.31420886516571045, "learning_rate": 9.579e-05, "loss": 0.0721, "step": 9580 }, { "epoch": 0.32421650495283816, "grad_norm": 0.29340675473213196, "learning_rate": 9.589e-05, "loss": 0.0724, "step": 9590 }, { "epoch": 0.3245545826430914, "grad_norm": 0.24417151510715485, "learning_rate": 9.599000000000001e-05, "loss": 0.0643, "step": 9600 }, { "epoch": 0.32489266033334463, "grad_norm": 0.4128023684024811, "learning_rate": 9.609e-05, "loss": 0.0691, "step": 9610 }, { "epoch": 0.3252307380235978, "grad_norm": 0.2373443990945816, "learning_rate": 9.619e-05, "loss": 0.0606, "step": 9620 }, { "epoch": 0.32556881571385105, "grad_norm": 0.25084203481674194, "learning_rate": 9.629e-05, "loss": 0.0744, "step": 9630 }, { "epoch": 0.3259068934041043, "grad_norm": 0.2674984335899353, "learning_rate": 9.639e-05, "loss": 0.0672, "step": 9640 }, { "epoch": 0.32624497109435746, "grad_norm": 0.47052761912345886, "learning_rate": 9.649e-05, "loss": 0.0758, "step": 9650 }, { "epoch": 0.3265830487846107, "grad_norm": 0.21554745733737946, "learning_rate": 9.659000000000001e-05, "loss": 0.0691, "step": 9660 }, { "epoch": 0.32692112647486393, "grad_norm": 0.24778173863887787, "learning_rate": 9.669e-05, "loss": 0.0736, "step": 9670 }, { "epoch": 0.32725920416511717, "grad_norm": 0.3240603804588318, "learning_rate": 9.679e-05, "loss": 0.0719, "step": 9680 }, { "epoch": 0.32759728185537035, "grad_norm": 0.24482251703739166, "learning_rate": 9.689e-05, "loss": 0.07, "step": 9690 }, { "epoch": 0.3279353595456236, "grad_norm": 0.3222423493862152, "learning_rate": 9.699e-05, "loss": 0.076, "step": 9700 }, { "epoch": 0.3282734372358768, "grad_norm": 0.2654714286327362, "learning_rate": 9.709000000000001e-05, "loss": 0.069, "step": 9710 }, { "epoch": 0.32861151492613, "grad_norm": 0.2808191478252411, "learning_rate": 9.719000000000001e-05, "loss": 0.0721, "step": 9720 }, { "epoch": 0.32894959261638324, "grad_norm": 0.3115299344062805, "learning_rate": 9.729e-05, "loss": 0.0776, "step": 9730 }, { "epoch": 0.32928767030663647, "grad_norm": 0.3665913939476013, "learning_rate": 9.739e-05, "loss": 0.0751, "step": 9740 }, { "epoch": 0.3296257479968897, "grad_norm": 0.28455182909965515, "learning_rate": 9.749e-05, "loss": 0.0698, "step": 9750 }, { "epoch": 0.3299638256871429, "grad_norm": 0.2189568728208542, "learning_rate": 9.759e-05, "loss": 0.0757, "step": 9760 }, { "epoch": 0.3303019033773961, "grad_norm": 0.3142379820346832, "learning_rate": 9.769000000000001e-05, "loss": 0.0743, "step": 9770 }, { "epoch": 0.33063998106764936, "grad_norm": 0.3893892467021942, "learning_rate": 9.779e-05, "loss": 0.0743, "step": 9780 }, { "epoch": 0.3309780587579026, "grad_norm": 0.23058345913887024, "learning_rate": 9.789e-05, "loss": 0.0746, "step": 9790 }, { "epoch": 0.3313161364481558, "grad_norm": 0.3097909688949585, "learning_rate": 9.799e-05, "loss": 0.0683, "step": 9800 }, { "epoch": 0.331654214138409, "grad_norm": 0.2886795103549957, "learning_rate": 9.809e-05, "loss": 0.0648, "step": 9810 }, { "epoch": 0.33199229182866224, "grad_norm": 0.36379653215408325, "learning_rate": 9.819000000000001e-05, "loss": 0.0643, "step": 9820 }, { "epoch": 0.3323303695189154, "grad_norm": 0.27226367592811584, "learning_rate": 9.829000000000001e-05, "loss": 0.0607, "step": 9830 }, { "epoch": 0.33266844720916866, "grad_norm": 0.3311687707901001, "learning_rate": 9.839e-05, "loss": 0.0708, "step": 9840 }, { "epoch": 0.3330065248994219, "grad_norm": 0.2073339968919754, "learning_rate": 9.849e-05, "loss": 0.0654, "step": 9850 }, { "epoch": 0.33334460258967513, "grad_norm": 0.29569435119628906, "learning_rate": 9.859e-05, "loss": 0.0683, "step": 9860 }, { "epoch": 0.3336826802799283, "grad_norm": 0.6210514307022095, "learning_rate": 9.869e-05, "loss": 0.0713, "step": 9870 }, { "epoch": 0.33402075797018155, "grad_norm": 0.4567168354988098, "learning_rate": 9.879000000000001e-05, "loss": 0.0684, "step": 9880 }, { "epoch": 0.3343588356604348, "grad_norm": 0.3006940186023712, "learning_rate": 9.889000000000001e-05, "loss": 0.0746, "step": 9890 }, { "epoch": 0.33469691335068796, "grad_norm": 0.6139218807220459, "learning_rate": 9.899e-05, "loss": 0.078, "step": 9900 }, { "epoch": 0.3350349910409412, "grad_norm": 0.3126339912414551, "learning_rate": 9.909e-05, "loss": 0.0727, "step": 9910 }, { "epoch": 0.33537306873119443, "grad_norm": 0.31969237327575684, "learning_rate": 9.919e-05, "loss": 0.0794, "step": 9920 }, { "epoch": 0.33571114642144767, "grad_norm": 0.37566784024238586, "learning_rate": 9.929e-05, "loss": 0.0757, "step": 9930 }, { "epoch": 0.33604922411170085, "grad_norm": 0.3679836392402649, "learning_rate": 9.939000000000001e-05, "loss": 0.0686, "step": 9940 }, { "epoch": 0.3363873018019541, "grad_norm": 0.2665485441684723, "learning_rate": 9.949000000000001e-05, "loss": 0.0688, "step": 9950 }, { "epoch": 0.3367253794922073, "grad_norm": 0.21993400156497955, "learning_rate": 9.959e-05, "loss": 0.0656, "step": 9960 }, { "epoch": 0.3370634571824605, "grad_norm": 0.19419653713703156, "learning_rate": 9.969e-05, "loss": 0.0682, "step": 9970 }, { "epoch": 0.33740153487271374, "grad_norm": 0.7352328300476074, "learning_rate": 9.979e-05, "loss": 0.0687, "step": 9980 }, { "epoch": 0.337739612562967, "grad_norm": 0.3196682929992676, "learning_rate": 9.989000000000001e-05, "loss": 0.0631, "step": 9990 }, { "epoch": 0.3380776902532202, "grad_norm": 0.4625927805900574, "learning_rate": 9.999000000000001e-05, "loss": 0.0748, "step": 10000 }, { "epoch": 0.3384157679434734, "grad_norm": 0.2726537883281708, "learning_rate": 9.999999944637262e-05, "loss": 0.0683, "step": 10010 }, { "epoch": 0.3387538456337266, "grad_norm": 0.41390496492385864, "learning_rate": 9.999999753259892e-05, "loss": 0.0672, "step": 10020 }, { "epoch": 0.33909192332397986, "grad_norm": 0.29301539063453674, "learning_rate": 9.999999425184409e-05, "loss": 0.0682, "step": 10030 }, { "epoch": 0.3394300010142331, "grad_norm": 0.21069525182247162, "learning_rate": 9.999998960410819e-05, "loss": 0.0651, "step": 10040 }, { "epoch": 0.3397680787044863, "grad_norm": 0.23346444964408875, "learning_rate": 9.999998358939136e-05, "loss": 0.068, "step": 10050 }, { "epoch": 0.3401061563947395, "grad_norm": 0.3571281135082245, "learning_rate": 9.999997620769377e-05, "loss": 0.0722, "step": 10060 }, { "epoch": 0.34044423408499275, "grad_norm": 0.3233571946620941, "learning_rate": 9.999996745901561e-05, "loss": 0.0643, "step": 10070 }, { "epoch": 0.3407823117752459, "grad_norm": 0.23534958064556122, "learning_rate": 9.999995734335712e-05, "loss": 0.0725, "step": 10080 }, { "epoch": 0.34112038946549916, "grad_norm": 0.46468180418014526, "learning_rate": 9.99999458607186e-05, "loss": 0.0689, "step": 10090 }, { "epoch": 0.3414584671557524, "grad_norm": 0.3562585711479187, "learning_rate": 9.999993301110033e-05, "loss": 0.0728, "step": 10100 }, { "epoch": 0.34179654484600563, "grad_norm": 0.26714110374450684, "learning_rate": 9.999991879450268e-05, "loss": 0.0686, "step": 10110 }, { "epoch": 0.3421346225362588, "grad_norm": 0.36091241240501404, "learning_rate": 9.999990321092603e-05, "loss": 0.0732, "step": 10120 }, { "epoch": 0.34247270022651205, "grad_norm": 0.24995340406894684, "learning_rate": 9.99998862603708e-05, "loss": 0.0698, "step": 10130 }, { "epoch": 0.3428107779167653, "grad_norm": 0.1828518956899643, "learning_rate": 9.999986794283748e-05, "loss": 0.0649, "step": 10140 }, { "epoch": 0.34314885560701847, "grad_norm": 0.3981345593929291, "learning_rate": 9.999984825832654e-05, "loss": 0.0628, "step": 10150 }, { "epoch": 0.3434869332972717, "grad_norm": 0.5429570078849792, "learning_rate": 9.999982720683854e-05, "loss": 0.0691, "step": 10160 }, { "epoch": 0.34382501098752494, "grad_norm": 0.3591170012950897, "learning_rate": 9.999980478837405e-05, "loss": 0.0728, "step": 10170 }, { "epoch": 0.3441630886777782, "grad_norm": 0.28045764565467834, "learning_rate": 9.999978100293367e-05, "loss": 0.0755, "step": 10180 }, { "epoch": 0.34450116636803135, "grad_norm": 0.29588741064071655, "learning_rate": 9.999975585051808e-05, "loss": 0.0637, "step": 10190 }, { "epoch": 0.3448392440582846, "grad_norm": 0.29934608936309814, "learning_rate": 9.999972933112795e-05, "loss": 0.0714, "step": 10200 }, { "epoch": 0.3451773217485378, "grad_norm": 0.24687162041664124, "learning_rate": 9.999970144476398e-05, "loss": 0.0675, "step": 10210 }, { "epoch": 0.34551539943879106, "grad_norm": 0.27828362584114075, "learning_rate": 9.999967219142698e-05, "loss": 0.0739, "step": 10220 }, { "epoch": 0.34585347712904424, "grad_norm": 0.2876240313053131, "learning_rate": 9.999964157111771e-05, "loss": 0.0721, "step": 10230 }, { "epoch": 0.3461915548192975, "grad_norm": 0.3007817268371582, "learning_rate": 9.999960958383704e-05, "loss": 0.0712, "step": 10240 }, { "epoch": 0.3465296325095507, "grad_norm": 0.23205500841140747, "learning_rate": 9.999957622958582e-05, "loss": 0.0696, "step": 10250 }, { "epoch": 0.3468677101998039, "grad_norm": 0.33976230025291443, "learning_rate": 9.999954150836497e-05, "loss": 0.0744, "step": 10260 }, { "epoch": 0.3472057878900571, "grad_norm": 0.21184463798999786, "learning_rate": 9.999950542017544e-05, "loss": 0.068, "step": 10270 }, { "epoch": 0.34754386558031036, "grad_norm": 0.23649689555168152, "learning_rate": 9.99994679650182e-05, "loss": 0.0651, "step": 10280 }, { "epoch": 0.3478819432705636, "grad_norm": 0.2525469660758972, "learning_rate": 9.99994291428943e-05, "loss": 0.0676, "step": 10290 }, { "epoch": 0.3482200209608168, "grad_norm": 0.42259514331817627, "learning_rate": 9.999938895380481e-05, "loss": 0.0724, "step": 10300 }, { "epoch": 0.34855809865107, "grad_norm": 0.2908109724521637, "learning_rate": 9.999934739775079e-05, "loss": 0.0755, "step": 10310 }, { "epoch": 0.34889617634132325, "grad_norm": 0.3986474871635437, "learning_rate": 9.99993044747334e-05, "loss": 0.0662, "step": 10320 }, { "epoch": 0.34923425403157643, "grad_norm": 0.3247448205947876, "learning_rate": 9.99992601847538e-05, "loss": 0.0728, "step": 10330 }, { "epoch": 0.34957233172182967, "grad_norm": 0.32184678316116333, "learning_rate": 9.999921452781324e-05, "loss": 0.0648, "step": 10340 }, { "epoch": 0.3499104094120829, "grad_norm": 0.28398874402046204, "learning_rate": 9.999916750391291e-05, "loss": 0.0646, "step": 10350 }, { "epoch": 0.35024848710233614, "grad_norm": 0.29339027404785156, "learning_rate": 9.999911911305413e-05, "loss": 0.0712, "step": 10360 }, { "epoch": 0.3505865647925893, "grad_norm": 0.2635957598686218, "learning_rate": 9.999906935523822e-05, "loss": 0.0692, "step": 10370 }, { "epoch": 0.35092464248284255, "grad_norm": 0.357182115316391, "learning_rate": 9.999901823046655e-05, "loss": 0.0682, "step": 10380 }, { "epoch": 0.3512627201730958, "grad_norm": 0.3671891987323761, "learning_rate": 9.999896573874048e-05, "loss": 0.0732, "step": 10390 }, { "epoch": 0.351600797863349, "grad_norm": 0.3336332440376282, "learning_rate": 9.999891188006149e-05, "loss": 0.0621, "step": 10400 }, { "epoch": 0.3519388755536022, "grad_norm": 0.3567427694797516, "learning_rate": 9.999885665443103e-05, "loss": 0.0684, "step": 10410 }, { "epoch": 0.35227695324385544, "grad_norm": 0.2625023424625397, "learning_rate": 9.99988000618506e-05, "loss": 0.0714, "step": 10420 }, { "epoch": 0.3526150309341087, "grad_norm": 0.26078087091445923, "learning_rate": 9.999874210232179e-05, "loss": 0.0611, "step": 10430 }, { "epoch": 0.35295310862436186, "grad_norm": 0.2992513179779053, "learning_rate": 9.999868277584612e-05, "loss": 0.0629, "step": 10440 }, { "epoch": 0.3532911863146151, "grad_norm": 0.40346601605415344, "learning_rate": 9.999862208242526e-05, "loss": 0.0701, "step": 10450 }, { "epoch": 0.3536292640048683, "grad_norm": 0.24247021973133087, "learning_rate": 9.999856002206086e-05, "loss": 0.0832, "step": 10460 }, { "epoch": 0.35396734169512156, "grad_norm": 0.3168807327747345, "learning_rate": 9.99984965947546e-05, "loss": 0.0654, "step": 10470 }, { "epoch": 0.35430541938537474, "grad_norm": 0.313511461019516, "learning_rate": 9.999843180050825e-05, "loss": 0.0731, "step": 10480 }, { "epoch": 0.354643497075628, "grad_norm": 0.3286755681037903, "learning_rate": 9.999836563932355e-05, "loss": 0.0682, "step": 10490 }, { "epoch": 0.3549815747658812, "grad_norm": 0.27513137459754944, "learning_rate": 9.99982981112023e-05, "loss": 0.0716, "step": 10500 }, { "epoch": 0.3553196524561344, "grad_norm": 0.22605907917022705, "learning_rate": 9.999822921614636e-05, "loss": 0.0607, "step": 10510 }, { "epoch": 0.35565773014638763, "grad_norm": 0.2835770547389984, "learning_rate": 9.999815895415764e-05, "loss": 0.0658, "step": 10520 }, { "epoch": 0.35599580783664087, "grad_norm": 0.2627856135368347, "learning_rate": 9.999808732523804e-05, "loss": 0.0675, "step": 10530 }, { "epoch": 0.3563338855268941, "grad_norm": 0.2319875955581665, "learning_rate": 9.99980143293895e-05, "loss": 0.0676, "step": 10540 }, { "epoch": 0.3566719632171473, "grad_norm": 0.24927280843257904, "learning_rate": 9.999793996661402e-05, "loss": 0.0665, "step": 10550 }, { "epoch": 0.3570100409074005, "grad_norm": 0.22487251460552216, "learning_rate": 9.999786423691368e-05, "loss": 0.068, "step": 10560 }, { "epoch": 0.35734811859765375, "grad_norm": 0.23944982886314392, "learning_rate": 9.999778714029049e-05, "loss": 0.0694, "step": 10570 }, { "epoch": 0.357686196287907, "grad_norm": 0.5058746337890625, "learning_rate": 9.999770867674658e-05, "loss": 0.0649, "step": 10580 }, { "epoch": 0.35802427397816017, "grad_norm": 0.22981855273246765, "learning_rate": 9.999762884628413e-05, "loss": 0.0619, "step": 10590 }, { "epoch": 0.3583623516684134, "grad_norm": 0.26924076676368713, "learning_rate": 9.999754764890525e-05, "loss": 0.0728, "step": 10600 }, { "epoch": 0.35870042935866664, "grad_norm": 0.2595076858997345, "learning_rate": 9.999746508461222e-05, "loss": 0.0632, "step": 10610 }, { "epoch": 0.3590385070489198, "grad_norm": 0.23314888775348663, "learning_rate": 9.999738115340727e-05, "loss": 0.0676, "step": 10620 }, { "epoch": 0.35937658473917306, "grad_norm": 0.31406015157699585, "learning_rate": 9.999729585529271e-05, "loss": 0.067, "step": 10630 }, { "epoch": 0.3597146624294263, "grad_norm": 0.26352888345718384, "learning_rate": 9.999720919027087e-05, "loss": 0.0702, "step": 10640 }, { "epoch": 0.3600527401196795, "grad_norm": 0.3566734194755554, "learning_rate": 9.99971211583441e-05, "loss": 0.0673, "step": 10650 }, { "epoch": 0.3603908178099327, "grad_norm": 0.34269553422927856, "learning_rate": 9.999703175951483e-05, "loss": 0.0652, "step": 10660 }, { "epoch": 0.36072889550018594, "grad_norm": 0.2534126937389374, "learning_rate": 9.999694099378549e-05, "loss": 0.0757, "step": 10670 }, { "epoch": 0.3610669731904392, "grad_norm": 0.30217477679252625, "learning_rate": 9.999684886115857e-05, "loss": 0.0711, "step": 10680 }, { "epoch": 0.36140505088069236, "grad_norm": 0.2671118974685669, "learning_rate": 9.999675536163657e-05, "loss": 0.0732, "step": 10690 }, { "epoch": 0.3617431285709456, "grad_norm": 0.22177456319332123, "learning_rate": 9.999666049522208e-05, "loss": 0.0648, "step": 10700 }, { "epoch": 0.36208120626119883, "grad_norm": 0.2546204626560211, "learning_rate": 9.999656426191766e-05, "loss": 0.0691, "step": 10710 }, { "epoch": 0.36241928395145206, "grad_norm": 0.3277845084667206, "learning_rate": 9.999646666172597e-05, "loss": 0.0632, "step": 10720 }, { "epoch": 0.36275736164170524, "grad_norm": 0.3144679665565491, "learning_rate": 9.999636769464967e-05, "loss": 0.0652, "step": 10730 }, { "epoch": 0.3630954393319585, "grad_norm": 0.2537784278392792, "learning_rate": 9.999626736069144e-05, "loss": 0.07, "step": 10740 }, { "epoch": 0.3634335170222117, "grad_norm": 0.25308334827423096, "learning_rate": 9.999616565985406e-05, "loss": 0.0701, "step": 10750 }, { "epoch": 0.36377159471246495, "grad_norm": 0.3108534514904022, "learning_rate": 9.999606259214029e-05, "loss": 0.0714, "step": 10760 }, { "epoch": 0.36410967240271813, "grad_norm": 0.175559401512146, "learning_rate": 9.999595815755294e-05, "loss": 0.0639, "step": 10770 }, { "epoch": 0.36444775009297137, "grad_norm": 0.29798460006713867, "learning_rate": 9.999585235609488e-05, "loss": 0.064, "step": 10780 }, { "epoch": 0.3647858277832246, "grad_norm": 0.231851726770401, "learning_rate": 9.9995745187769e-05, "loss": 0.0683, "step": 10790 }, { "epoch": 0.3651239054734778, "grad_norm": 0.3189156651496887, "learning_rate": 9.999563665257823e-05, "loss": 0.0659, "step": 10800 }, { "epoch": 0.365461983163731, "grad_norm": 0.23067976534366608, "learning_rate": 9.999552675052554e-05, "loss": 0.0694, "step": 10810 }, { "epoch": 0.36580006085398425, "grad_norm": 0.23071566224098206, "learning_rate": 9.999541548161392e-05, "loss": 0.0672, "step": 10820 }, { "epoch": 0.3661381385442375, "grad_norm": 0.23002545535564423, "learning_rate": 9.999530284584644e-05, "loss": 0.0703, "step": 10830 }, { "epoch": 0.36647621623449067, "grad_norm": 0.3188582956790924, "learning_rate": 9.999518884322617e-05, "loss": 0.0648, "step": 10840 }, { "epoch": 0.3668142939247439, "grad_norm": 0.2703190743923187, "learning_rate": 9.99950734737562e-05, "loss": 0.0753, "step": 10850 }, { "epoch": 0.36715237161499714, "grad_norm": 0.3051704168319702, "learning_rate": 9.99949567374397e-05, "loss": 0.0624, "step": 10860 }, { "epoch": 0.3674904493052503, "grad_norm": 0.27207231521606445, "learning_rate": 9.999483863427988e-05, "loss": 0.072, "step": 10870 }, { "epoch": 0.36782852699550356, "grad_norm": 0.4492802321910858, "learning_rate": 9.999471916427993e-05, "loss": 0.0715, "step": 10880 }, { "epoch": 0.3681666046857568, "grad_norm": 0.2961169183254242, "learning_rate": 9.999459832744318e-05, "loss": 0.0644, "step": 10890 }, { "epoch": 0.36850468237601003, "grad_norm": 0.285459965467453, "learning_rate": 9.999447612377286e-05, "loss": 0.0688, "step": 10900 }, { "epoch": 0.3688427600662632, "grad_norm": 0.22746039927005768, "learning_rate": 9.999435255327236e-05, "loss": 0.0627, "step": 10910 }, { "epoch": 0.36918083775651644, "grad_norm": 0.2650723457336426, "learning_rate": 9.999422761594503e-05, "loss": 0.0622, "step": 10920 }, { "epoch": 0.3695189154467697, "grad_norm": 0.2991267442703247, "learning_rate": 9.999410131179431e-05, "loss": 0.0653, "step": 10930 }, { "epoch": 0.36985699313702286, "grad_norm": 0.30523234605789185, "learning_rate": 9.999397364082365e-05, "loss": 0.0694, "step": 10940 }, { "epoch": 0.3701950708272761, "grad_norm": 0.2718335688114166, "learning_rate": 9.99938446030365e-05, "loss": 0.0629, "step": 10950 }, { "epoch": 0.37053314851752933, "grad_norm": 0.20968124270439148, "learning_rate": 9.999371419843646e-05, "loss": 0.0611, "step": 10960 }, { "epoch": 0.37087122620778257, "grad_norm": 0.3373114764690399, "learning_rate": 9.999358242702703e-05, "loss": 0.0617, "step": 10970 }, { "epoch": 0.37120930389803575, "grad_norm": 0.21274466812610626, "learning_rate": 9.999344928881185e-05, "loss": 0.0678, "step": 10980 }, { "epoch": 0.371547381588289, "grad_norm": 0.38452082872390747, "learning_rate": 9.999331478379455e-05, "loss": 0.0686, "step": 10990 }, { "epoch": 0.3718854592785422, "grad_norm": 0.37159019708633423, "learning_rate": 9.999317891197881e-05, "loss": 0.0692, "step": 11000 }, { "epoch": 0.37222353696879545, "grad_norm": 0.2831898331642151, "learning_rate": 9.999304167336832e-05, "loss": 0.0689, "step": 11010 }, { "epoch": 0.37256161465904863, "grad_norm": 0.3538314700126648, "learning_rate": 9.999290306796687e-05, "loss": 0.0701, "step": 11020 }, { "epoch": 0.37289969234930187, "grad_norm": 0.35443365573883057, "learning_rate": 9.99927630957782e-05, "loss": 0.0724, "step": 11030 }, { "epoch": 0.3732377700395551, "grad_norm": 0.19397981464862823, "learning_rate": 9.999262175680621e-05, "loss": 0.0615, "step": 11040 }, { "epoch": 0.3735758477298083, "grad_norm": 0.3967226445674896, "learning_rate": 9.999247905105471e-05, "loss": 0.0627, "step": 11050 }, { "epoch": 0.3739139254200615, "grad_norm": 0.4636348485946655, "learning_rate": 9.99923349785276e-05, "loss": 0.0717, "step": 11060 }, { "epoch": 0.37425200311031476, "grad_norm": 0.341144323348999, "learning_rate": 9.999218953922884e-05, "loss": 0.0684, "step": 11070 }, { "epoch": 0.374590080800568, "grad_norm": 0.31872597336769104, "learning_rate": 9.99920427331624e-05, "loss": 0.0632, "step": 11080 }, { "epoch": 0.3749281584908212, "grad_norm": 0.30253997445106506, "learning_rate": 9.99918945603323e-05, "loss": 0.0627, "step": 11090 }, { "epoch": 0.3752662361810744, "grad_norm": 0.25989851355552673, "learning_rate": 9.999174502074258e-05, "loss": 0.0602, "step": 11100 }, { "epoch": 0.37560431387132764, "grad_norm": 0.31268370151519775, "learning_rate": 9.999159411439733e-05, "loss": 0.0681, "step": 11110 }, { "epoch": 0.3759423915615808, "grad_norm": 0.2316466122865677, "learning_rate": 9.999144184130067e-05, "loss": 0.0669, "step": 11120 }, { "epoch": 0.37628046925183406, "grad_norm": 0.24150820076465607, "learning_rate": 9.999128820145678e-05, "loss": 0.0637, "step": 11130 }, { "epoch": 0.3766185469420873, "grad_norm": 0.29583680629730225, "learning_rate": 9.999113319486986e-05, "loss": 0.0646, "step": 11140 }, { "epoch": 0.37695662463234053, "grad_norm": 0.1769183725118637, "learning_rate": 9.999097682154413e-05, "loss": 0.0625, "step": 11150 }, { "epoch": 0.3772947023225937, "grad_norm": 0.19455017149448395, "learning_rate": 9.999081908148387e-05, "loss": 0.0679, "step": 11160 }, { "epoch": 0.37763278001284695, "grad_norm": 0.2495660036802292, "learning_rate": 9.99906599746934e-05, "loss": 0.0707, "step": 11170 }, { "epoch": 0.3779708577031002, "grad_norm": 0.25217723846435547, "learning_rate": 9.999049950117707e-05, "loss": 0.0661, "step": 11180 }, { "epoch": 0.3783089353933534, "grad_norm": 0.19508647918701172, "learning_rate": 9.999033766093926e-05, "loss": 0.0654, "step": 11190 }, { "epoch": 0.3786470130836066, "grad_norm": 0.2330135703086853, "learning_rate": 9.99901744539844e-05, "loss": 0.0702, "step": 11200 }, { "epoch": 0.37898509077385983, "grad_norm": 0.23574787378311157, "learning_rate": 9.999000988031695e-05, "loss": 0.0623, "step": 11210 }, { "epoch": 0.37932316846411307, "grad_norm": 0.25444266200065613, "learning_rate": 9.99898439399414e-05, "loss": 0.0657, "step": 11220 }, { "epoch": 0.37966124615436625, "grad_norm": 0.22675691545009613, "learning_rate": 9.998967663286232e-05, "loss": 0.0705, "step": 11230 }, { "epoch": 0.3799993238446195, "grad_norm": 0.2748886048793793, "learning_rate": 9.998950795908425e-05, "loss": 0.0702, "step": 11240 }, { "epoch": 0.3803374015348727, "grad_norm": 0.27950751781463623, "learning_rate": 9.99893379186118e-05, "loss": 0.0684, "step": 11250 }, { "epoch": 0.38067547922512596, "grad_norm": 0.2346382737159729, "learning_rate": 9.998916651144966e-05, "loss": 0.0587, "step": 11260 }, { "epoch": 0.38101355691537914, "grad_norm": 0.35532140731811523, "learning_rate": 9.998899373760247e-05, "loss": 0.0702, "step": 11270 }, { "epoch": 0.3813516346056324, "grad_norm": 0.2381218820810318, "learning_rate": 9.998881959707497e-05, "loss": 0.0645, "step": 11280 }, { "epoch": 0.3816897122958856, "grad_norm": 0.40650975704193115, "learning_rate": 9.998864408987192e-05, "loss": 0.0717, "step": 11290 }, { "epoch": 0.3820277899861388, "grad_norm": 0.43510574102401733, "learning_rate": 9.998846721599813e-05, "loss": 0.0711, "step": 11300 }, { "epoch": 0.382365867676392, "grad_norm": 0.35735899209976196, "learning_rate": 9.998828897545842e-05, "loss": 0.071, "step": 11310 }, { "epoch": 0.38270394536664526, "grad_norm": 0.23968233168125153, "learning_rate": 9.998810936825766e-05, "loss": 0.0706, "step": 11320 }, { "epoch": 0.3830420230568985, "grad_norm": 0.3751976490020752, "learning_rate": 9.998792839440079e-05, "loss": 0.0622, "step": 11330 }, { "epoch": 0.3833801007471517, "grad_norm": 0.2354944944381714, "learning_rate": 9.998774605389273e-05, "loss": 0.0672, "step": 11340 }, { "epoch": 0.3837181784374049, "grad_norm": 0.20341138541698456, "learning_rate": 9.998756234673847e-05, "loss": 0.0631, "step": 11350 }, { "epoch": 0.38405625612765815, "grad_norm": 0.3479911684989929, "learning_rate": 9.998737727294303e-05, "loss": 0.0751, "step": 11360 }, { "epoch": 0.3843943338179114, "grad_norm": 0.35696253180503845, "learning_rate": 9.998719083251149e-05, "loss": 0.0657, "step": 11370 }, { "epoch": 0.38473241150816456, "grad_norm": 0.26188957691192627, "learning_rate": 9.998700302544893e-05, "loss": 0.0686, "step": 11380 }, { "epoch": 0.3850704891984178, "grad_norm": 0.23301026225090027, "learning_rate": 9.998681385176047e-05, "loss": 0.0725, "step": 11390 }, { "epoch": 0.38540856688867103, "grad_norm": 0.24793878197669983, "learning_rate": 9.99866233114513e-05, "loss": 0.0732, "step": 11400 }, { "epoch": 0.3857466445789242, "grad_norm": 0.27319613099098206, "learning_rate": 9.998643140452663e-05, "loss": 0.0754, "step": 11410 }, { "epoch": 0.38608472226917745, "grad_norm": 0.2932403087615967, "learning_rate": 9.99862381309917e-05, "loss": 0.0612, "step": 11420 }, { "epoch": 0.3864227999594307, "grad_norm": 0.3742261230945587, "learning_rate": 9.998604349085182e-05, "loss": 0.0691, "step": 11430 }, { "epoch": 0.3867608776496839, "grad_norm": 0.45497825741767883, "learning_rate": 9.998584748411228e-05, "loss": 0.0671, "step": 11440 }, { "epoch": 0.3870989553399371, "grad_norm": 0.372010201215744, "learning_rate": 9.998565011077844e-05, "loss": 0.0621, "step": 11450 }, { "epoch": 0.38743703303019034, "grad_norm": 0.32726332545280457, "learning_rate": 9.99854513708557e-05, "loss": 0.0707, "step": 11460 }, { "epoch": 0.38777511072044357, "grad_norm": 0.3276490569114685, "learning_rate": 9.998525126434952e-05, "loss": 0.064, "step": 11470 }, { "epoch": 0.38811318841069675, "grad_norm": 0.42563340067863464, "learning_rate": 9.998504979126533e-05, "loss": 0.0688, "step": 11480 }, { "epoch": 0.38845126610095, "grad_norm": 0.22005529701709747, "learning_rate": 9.998484695160868e-05, "loss": 0.0657, "step": 11490 }, { "epoch": 0.3887893437912032, "grad_norm": 0.3476591110229492, "learning_rate": 9.998464274538507e-05, "loss": 0.0653, "step": 11500 }, { "epoch": 0.38912742148145646, "grad_norm": 0.2653920352458954, "learning_rate": 9.99844371726001e-05, "loss": 0.0665, "step": 11510 }, { "epoch": 0.38946549917170964, "grad_norm": 0.3099082410335541, "learning_rate": 9.998423023325943e-05, "loss": 0.065, "step": 11520 }, { "epoch": 0.3898035768619629, "grad_norm": 0.4939483106136322, "learning_rate": 9.998402192736865e-05, "loss": 0.0673, "step": 11530 }, { "epoch": 0.3901416545522161, "grad_norm": 0.32421332597732544, "learning_rate": 9.99838122549335e-05, "loss": 0.0704, "step": 11540 }, { "epoch": 0.39047973224246935, "grad_norm": 0.3229186534881592, "learning_rate": 9.998360121595972e-05, "loss": 0.0692, "step": 11550 }, { "epoch": 0.3908178099327225, "grad_norm": 0.19318069517612457, "learning_rate": 9.998338881045304e-05, "loss": 0.0673, "step": 11560 }, { "epoch": 0.39115588762297576, "grad_norm": 0.32702702283859253, "learning_rate": 9.998317503841927e-05, "loss": 0.0709, "step": 11570 }, { "epoch": 0.391493965313229, "grad_norm": 0.4143475294113159, "learning_rate": 9.99829598998643e-05, "loss": 0.0677, "step": 11580 }, { "epoch": 0.3918320430034822, "grad_norm": 0.36416149139404297, "learning_rate": 9.998274339479398e-05, "loss": 0.0664, "step": 11590 }, { "epoch": 0.3921701206937354, "grad_norm": 0.2756890058517456, "learning_rate": 9.998252552321423e-05, "loss": 0.0691, "step": 11600 }, { "epoch": 0.39250819838398865, "grad_norm": 0.36155760288238525, "learning_rate": 9.998230628513101e-05, "loss": 0.072, "step": 11610 }, { "epoch": 0.3928462760742419, "grad_norm": 0.3259612023830414, "learning_rate": 9.998208568055032e-05, "loss": 0.0743, "step": 11620 }, { "epoch": 0.39318435376449506, "grad_norm": 0.38280200958251953, "learning_rate": 9.998186370947817e-05, "loss": 0.0628, "step": 11630 }, { "epoch": 0.3935224314547483, "grad_norm": 0.31330406665802, "learning_rate": 9.998164037192066e-05, "loss": 0.0736, "step": 11640 }, { "epoch": 0.39386050914500154, "grad_norm": 0.21599464118480682, "learning_rate": 9.998141566788387e-05, "loss": 0.064, "step": 11650 }, { "epoch": 0.3941985868352547, "grad_norm": 0.2907182574272156, "learning_rate": 9.998118959737395e-05, "loss": 0.0671, "step": 11660 }, { "epoch": 0.39453666452550795, "grad_norm": 0.4032227694988251, "learning_rate": 9.998096216039709e-05, "loss": 0.0684, "step": 11670 }, { "epoch": 0.3948747422157612, "grad_norm": 0.3709966540336609, "learning_rate": 9.99807333569595e-05, "loss": 0.0716, "step": 11680 }, { "epoch": 0.3952128199060144, "grad_norm": 0.2332250028848648, "learning_rate": 9.998050318706744e-05, "loss": 0.0656, "step": 11690 }, { "epoch": 0.3955508975962676, "grad_norm": 0.24790741503238678, "learning_rate": 9.998027165072718e-05, "loss": 0.0716, "step": 11700 }, { "epoch": 0.39588897528652084, "grad_norm": 0.2929638922214508, "learning_rate": 9.998003874794511e-05, "loss": 0.0642, "step": 11710 }, { "epoch": 0.3962270529767741, "grad_norm": 0.26947444677352905, "learning_rate": 9.997980447872753e-05, "loss": 0.0632, "step": 11720 }, { "epoch": 0.3965651306670273, "grad_norm": 0.28129446506500244, "learning_rate": 9.997956884308086e-05, "loss": 0.0684, "step": 11730 }, { "epoch": 0.3969032083572805, "grad_norm": 0.2933569848537445, "learning_rate": 9.997933184101158e-05, "loss": 0.0751, "step": 11740 }, { "epoch": 0.3972412860475337, "grad_norm": 0.373108446598053, "learning_rate": 9.997909347252613e-05, "loss": 0.0691, "step": 11750 }, { "epoch": 0.39757936373778696, "grad_norm": 0.2855663001537323, "learning_rate": 9.997885373763106e-05, "loss": 0.0678, "step": 11760 }, { "epoch": 0.39791744142804014, "grad_norm": 0.2901618778705597, "learning_rate": 9.997861263633288e-05, "loss": 0.0747, "step": 11770 }, { "epoch": 0.3982555191182934, "grad_norm": 0.2707326412200928, "learning_rate": 9.99783701686382e-05, "loss": 0.0681, "step": 11780 }, { "epoch": 0.3985935968085466, "grad_norm": 0.2604125142097473, "learning_rate": 9.997812633455368e-05, "loss": 0.066, "step": 11790 }, { "epoch": 0.39893167449879985, "grad_norm": 0.2456633448600769, "learning_rate": 9.997788113408595e-05, "loss": 0.0615, "step": 11800 }, { "epoch": 0.39926975218905303, "grad_norm": 0.397151917219162, "learning_rate": 9.997763456724172e-05, "loss": 0.0702, "step": 11810 }, { "epoch": 0.39960782987930626, "grad_norm": 0.3737679123878479, "learning_rate": 9.997738663402775e-05, "loss": 0.0617, "step": 11820 }, { "epoch": 0.3999459075695595, "grad_norm": 0.2901252806186676, "learning_rate": 9.99771373344508e-05, "loss": 0.0727, "step": 11830 }, { "epoch": 0.4002839852598127, "grad_norm": 0.2861708700656891, "learning_rate": 9.99768866685177e-05, "loss": 0.0652, "step": 11840 }, { "epoch": 0.4006220629500659, "grad_norm": 0.37788084149360657, "learning_rate": 9.997663463623528e-05, "loss": 0.0661, "step": 11850 }, { "epoch": 0.40096014064031915, "grad_norm": 0.32851624488830566, "learning_rate": 9.997638123761046e-05, "loss": 0.0659, "step": 11860 }, { "epoch": 0.4012982183305724, "grad_norm": 0.33713769912719727, "learning_rate": 9.997612647265014e-05, "loss": 0.0697, "step": 11870 }, { "epoch": 0.40163629602082557, "grad_norm": 0.2085457295179367, "learning_rate": 9.99758703413613e-05, "loss": 0.0649, "step": 11880 }, { "epoch": 0.4019743737110788, "grad_norm": 0.28258663415908813, "learning_rate": 9.997561284375093e-05, "loss": 0.0692, "step": 11890 }, { "epoch": 0.40231245140133204, "grad_norm": 0.30641475319862366, "learning_rate": 9.997535397982609e-05, "loss": 0.0681, "step": 11900 }, { "epoch": 0.4026505290915852, "grad_norm": 0.42980486154556274, "learning_rate": 9.997509374959384e-05, "loss": 0.0703, "step": 11910 }, { "epoch": 0.40298860678183845, "grad_norm": 0.25520288944244385, "learning_rate": 9.997483215306132e-05, "loss": 0.0676, "step": 11920 }, { "epoch": 0.4033266844720917, "grad_norm": 0.36420387029647827, "learning_rate": 9.997456919023565e-05, "loss": 0.0738, "step": 11930 }, { "epoch": 0.4036647621623449, "grad_norm": 0.307278573513031, "learning_rate": 9.997430486112402e-05, "loss": 0.0667, "step": 11940 }, { "epoch": 0.4040028398525981, "grad_norm": 0.2929416596889496, "learning_rate": 9.997403916573369e-05, "loss": 0.0639, "step": 11950 }, { "epoch": 0.40434091754285134, "grad_norm": 0.26682037115097046, "learning_rate": 9.997377210407191e-05, "loss": 0.0674, "step": 11960 }, { "epoch": 0.4046789952331046, "grad_norm": 0.24743759632110596, "learning_rate": 9.997350367614596e-05, "loss": 0.0654, "step": 11970 }, { "epoch": 0.4050170729233578, "grad_norm": 0.25093430280685425, "learning_rate": 9.99732338819632e-05, "loss": 0.0705, "step": 11980 }, { "epoch": 0.405355150613611, "grad_norm": 0.2987595796585083, "learning_rate": 9.997296272153098e-05, "loss": 0.0654, "step": 11990 }, { "epoch": 0.40569322830386423, "grad_norm": 0.27158451080322266, "learning_rate": 9.997269019485675e-05, "loss": 0.0721, "step": 12000 }, { "epoch": 0.40603130599411746, "grad_norm": 0.23317395150661469, "learning_rate": 9.997241630194795e-05, "loss": 0.0655, "step": 12010 }, { "epoch": 0.40636938368437064, "grad_norm": 0.29577264189720154, "learning_rate": 9.997214104281206e-05, "loss": 0.0736, "step": 12020 }, { "epoch": 0.4067074613746239, "grad_norm": 0.2771465480327606, "learning_rate": 9.99718644174566e-05, "loss": 0.0626, "step": 12030 }, { "epoch": 0.4070455390648771, "grad_norm": 0.30925002694129944, "learning_rate": 9.997158642588913e-05, "loss": 0.0609, "step": 12040 }, { "epoch": 0.40738361675513035, "grad_norm": 0.2623407542705536, "learning_rate": 9.997130706811728e-05, "loss": 0.0685, "step": 12050 }, { "epoch": 0.40772169444538353, "grad_norm": 0.2378334403038025, "learning_rate": 9.997102634414866e-05, "loss": 0.0683, "step": 12060 }, { "epoch": 0.40805977213563677, "grad_norm": 0.3006080389022827, "learning_rate": 9.997074425399095e-05, "loss": 0.0591, "step": 12070 }, { "epoch": 0.40839784982589, "grad_norm": 0.24826060235500336, "learning_rate": 9.997046079765187e-05, "loss": 0.0686, "step": 12080 }, { "epoch": 0.4087359275161432, "grad_norm": 0.22244301438331604, "learning_rate": 9.997017597513917e-05, "loss": 0.0642, "step": 12090 }, { "epoch": 0.4090740052063964, "grad_norm": 0.3399873375892639, "learning_rate": 9.996988978646062e-05, "loss": 0.0657, "step": 12100 }, { "epoch": 0.40941208289664965, "grad_norm": 0.2209894359111786, "learning_rate": 9.996960223162406e-05, "loss": 0.0645, "step": 12110 }, { "epoch": 0.4097501605869029, "grad_norm": 0.19075152277946472, "learning_rate": 9.996931331063735e-05, "loss": 0.0606, "step": 12120 }, { "epoch": 0.41008823827715607, "grad_norm": 0.3069913983345032, "learning_rate": 9.996902302350838e-05, "loss": 0.069, "step": 12130 }, { "epoch": 0.4104263159674093, "grad_norm": 0.2751665711402893, "learning_rate": 9.996873137024512e-05, "loss": 0.0637, "step": 12140 }, { "epoch": 0.41076439365766254, "grad_norm": 0.2346903383731842, "learning_rate": 9.99684383508555e-05, "loss": 0.0673, "step": 12150 }, { "epoch": 0.4111024713479158, "grad_norm": 0.317053884267807, "learning_rate": 9.996814396534753e-05, "loss": 0.0719, "step": 12160 }, { "epoch": 0.41144054903816896, "grad_norm": 0.2318786382675171, "learning_rate": 9.99678482137293e-05, "loss": 0.063, "step": 12170 }, { "epoch": 0.4117786267284222, "grad_norm": 0.1841597706079483, "learning_rate": 9.996755109600886e-05, "loss": 0.0669, "step": 12180 }, { "epoch": 0.41211670441867543, "grad_norm": 0.3951021730899811, "learning_rate": 9.996725261219435e-05, "loss": 0.0716, "step": 12190 }, { "epoch": 0.4124547821089286, "grad_norm": 0.26641735434532166, "learning_rate": 9.996695276229392e-05, "loss": 0.0644, "step": 12200 }, { "epoch": 0.41279285979918184, "grad_norm": 0.3263845145702362, "learning_rate": 9.996665154631577e-05, "loss": 0.0693, "step": 12210 }, { "epoch": 0.4131309374894351, "grad_norm": 0.26326224207878113, "learning_rate": 9.996634896426815e-05, "loss": 0.0652, "step": 12220 }, { "epoch": 0.4134690151796883, "grad_norm": 0.3118642568588257, "learning_rate": 9.996604501615931e-05, "loss": 0.0712, "step": 12230 }, { "epoch": 0.4138070928699415, "grad_norm": 0.3845488727092743, "learning_rate": 9.996573970199758e-05, "loss": 0.0681, "step": 12240 }, { "epoch": 0.41414517056019473, "grad_norm": 0.4873294234275818, "learning_rate": 9.996543302179128e-05, "loss": 0.0613, "step": 12250 }, { "epoch": 0.41448324825044797, "grad_norm": 0.35329872369766235, "learning_rate": 9.996512497554884e-05, "loss": 0.0639, "step": 12260 }, { "epoch": 0.41482132594070115, "grad_norm": 0.3573746681213379, "learning_rate": 9.996481556327863e-05, "loss": 0.0648, "step": 12270 }, { "epoch": 0.4151594036309544, "grad_norm": 0.23798976838588715, "learning_rate": 9.996450478498915e-05, "loss": 0.0751, "step": 12280 }, { "epoch": 0.4154974813212076, "grad_norm": 0.30972179770469666, "learning_rate": 9.996419264068888e-05, "loss": 0.0635, "step": 12290 }, { "epoch": 0.41583555901146085, "grad_norm": 0.24681951105594635, "learning_rate": 9.996387913038635e-05, "loss": 0.0613, "step": 12300 }, { "epoch": 0.41617363670171403, "grad_norm": 0.2677389979362488, "learning_rate": 9.996356425409013e-05, "loss": 0.0756, "step": 12310 }, { "epoch": 0.41651171439196727, "grad_norm": 0.3687395453453064, "learning_rate": 9.996324801180884e-05, "loss": 0.0678, "step": 12320 }, { "epoch": 0.4168497920822205, "grad_norm": 0.2500505745410919, "learning_rate": 9.996293040355111e-05, "loss": 0.0637, "step": 12330 }, { "epoch": 0.41718786977247374, "grad_norm": 0.2697368860244751, "learning_rate": 9.996261142932564e-05, "loss": 0.0688, "step": 12340 }, { "epoch": 0.4175259474627269, "grad_norm": 0.3254317045211792, "learning_rate": 9.996229108914113e-05, "loss": 0.0638, "step": 12350 }, { "epoch": 0.41786402515298016, "grad_norm": 0.3552045226097107, "learning_rate": 9.996196938300638e-05, "loss": 0.071, "step": 12360 }, { "epoch": 0.4182021028432334, "grad_norm": 0.23405756056308746, "learning_rate": 9.996164631093012e-05, "loss": 0.0725, "step": 12370 }, { "epoch": 0.41854018053348657, "grad_norm": 0.21360373497009277, "learning_rate": 9.996132187292125e-05, "loss": 0.0689, "step": 12380 }, { "epoch": 0.4188782582237398, "grad_norm": 0.21805791556835175, "learning_rate": 9.99609960689886e-05, "loss": 0.062, "step": 12390 }, { "epoch": 0.41921633591399304, "grad_norm": 0.2790396213531494, "learning_rate": 9.996066889914109e-05, "loss": 0.0707, "step": 12400 }, { "epoch": 0.4195544136042463, "grad_norm": 0.279813289642334, "learning_rate": 9.996034036338766e-05, "loss": 0.0619, "step": 12410 }, { "epoch": 0.41989249129449946, "grad_norm": 0.2482043206691742, "learning_rate": 9.996001046173728e-05, "loss": 0.0623, "step": 12420 }, { "epoch": 0.4202305689847527, "grad_norm": 0.2649824023246765, "learning_rate": 9.9959679194199e-05, "loss": 0.0682, "step": 12430 }, { "epoch": 0.42056864667500593, "grad_norm": 0.29538869857788086, "learning_rate": 9.995934656078186e-05, "loss": 0.0661, "step": 12440 }, { "epoch": 0.4209067243652591, "grad_norm": 0.30649179220199585, "learning_rate": 9.995901256149494e-05, "loss": 0.0657, "step": 12450 }, { "epoch": 0.42124480205551235, "grad_norm": 0.23637324571609497, "learning_rate": 9.99586771963474e-05, "loss": 0.0701, "step": 12460 }, { "epoch": 0.4215828797457656, "grad_norm": 0.30418825149536133, "learning_rate": 9.995834046534839e-05, "loss": 0.0663, "step": 12470 }, { "epoch": 0.4219209574360188, "grad_norm": 0.2447231560945511, "learning_rate": 9.995800236850712e-05, "loss": 0.0662, "step": 12480 }, { "epoch": 0.422259035126272, "grad_norm": 0.16455776989459991, "learning_rate": 9.995766290583283e-05, "loss": 0.066, "step": 12490 }, { "epoch": 0.42259711281652523, "grad_norm": 0.31145134568214417, "learning_rate": 9.99573220773348e-05, "loss": 0.0627, "step": 12500 }, { "epoch": 0.42293519050677847, "grad_norm": 0.20428535342216492, "learning_rate": 9.995697988302237e-05, "loss": 0.0658, "step": 12510 }, { "epoch": 0.4232732681970317, "grad_norm": 0.33829206228256226, "learning_rate": 9.995663632290487e-05, "loss": 0.061, "step": 12520 }, { "epoch": 0.4236113458872849, "grad_norm": 0.3423236310482025, "learning_rate": 9.995629139699169e-05, "loss": 0.0584, "step": 12530 }, { "epoch": 0.4239494235775381, "grad_norm": 0.26081156730651855, "learning_rate": 9.995594510529227e-05, "loss": 0.0581, "step": 12540 }, { "epoch": 0.42428750126779136, "grad_norm": 0.3299981653690338, "learning_rate": 9.99555974478161e-05, "loss": 0.0694, "step": 12550 }, { "epoch": 0.42462557895804454, "grad_norm": 0.19886140525341034, "learning_rate": 9.995524842457265e-05, "loss": 0.067, "step": 12560 }, { "epoch": 0.42496365664829777, "grad_norm": 0.27776944637298584, "learning_rate": 9.995489803557148e-05, "loss": 0.0637, "step": 12570 }, { "epoch": 0.425301734338551, "grad_norm": 0.2988968789577484, "learning_rate": 9.995454628082214e-05, "loss": 0.065, "step": 12580 }, { "epoch": 0.42563981202880424, "grad_norm": 0.279916912317276, "learning_rate": 9.99541931603343e-05, "loss": 0.0603, "step": 12590 }, { "epoch": 0.4259778897190574, "grad_norm": 0.19169972836971283, "learning_rate": 9.995383867411758e-05, "loss": 0.0628, "step": 12600 }, { "epoch": 0.42631596740931066, "grad_norm": 0.34978416562080383, "learning_rate": 9.995348282218167e-05, "loss": 0.0645, "step": 12610 }, { "epoch": 0.4266540450995639, "grad_norm": 0.35248205065727234, "learning_rate": 9.995312560453632e-05, "loss": 0.0681, "step": 12620 }, { "epoch": 0.4269921227898171, "grad_norm": 0.40727871656417847, "learning_rate": 9.995276702119128e-05, "loss": 0.0672, "step": 12630 }, { "epoch": 0.4273302004800703, "grad_norm": 0.31385964155197144, "learning_rate": 9.995240707215635e-05, "loss": 0.0648, "step": 12640 }, { "epoch": 0.42766827817032355, "grad_norm": 0.34162259101867676, "learning_rate": 9.995204575744137e-05, "loss": 0.0653, "step": 12650 }, { "epoch": 0.4280063558605768, "grad_norm": 0.26822105050086975, "learning_rate": 9.995168307705624e-05, "loss": 0.0603, "step": 12660 }, { "epoch": 0.42834443355082996, "grad_norm": 0.3931901454925537, "learning_rate": 9.995131903101084e-05, "loss": 0.0756, "step": 12670 }, { "epoch": 0.4286825112410832, "grad_norm": 0.3191660940647125, "learning_rate": 9.995095361931516e-05, "loss": 0.0662, "step": 12680 }, { "epoch": 0.42902058893133643, "grad_norm": 0.2883807122707367, "learning_rate": 9.995058684197918e-05, "loss": 0.0705, "step": 12690 }, { "epoch": 0.42935866662158967, "grad_norm": 0.3303722143173218, "learning_rate": 9.995021869901291e-05, "loss": 0.0635, "step": 12700 }, { "epoch": 0.42969674431184285, "grad_norm": 0.23120583593845367, "learning_rate": 9.994984919042642e-05, "loss": 0.0652, "step": 12710 }, { "epoch": 0.4300348220020961, "grad_norm": 0.3038819134235382, "learning_rate": 9.994947831622983e-05, "loss": 0.0634, "step": 12720 }, { "epoch": 0.4303728996923493, "grad_norm": 0.3585040271282196, "learning_rate": 9.994910607643327e-05, "loss": 0.074, "step": 12730 }, { "epoch": 0.4307109773826025, "grad_norm": 0.16416670382022858, "learning_rate": 9.99487324710469e-05, "loss": 0.0567, "step": 12740 }, { "epoch": 0.43104905507285574, "grad_norm": 0.24785618484020233, "learning_rate": 9.994835750008096e-05, "loss": 0.0656, "step": 12750 }, { "epoch": 0.43138713276310897, "grad_norm": 0.2868116497993469, "learning_rate": 9.994798116354568e-05, "loss": 0.0714, "step": 12760 }, { "epoch": 0.4317252104533622, "grad_norm": 0.2998870015144348, "learning_rate": 9.994760346145138e-05, "loss": 0.0673, "step": 12770 }, { "epoch": 0.4320632881436154, "grad_norm": 0.21167506277561188, "learning_rate": 9.994722439380834e-05, "loss": 0.0683, "step": 12780 }, { "epoch": 0.4324013658338686, "grad_norm": 0.21022465825080872, "learning_rate": 9.994684396062696e-05, "loss": 0.0684, "step": 12790 }, { "epoch": 0.43273944352412186, "grad_norm": 0.2954005300998688, "learning_rate": 9.994646216191766e-05, "loss": 0.071, "step": 12800 }, { "epoch": 0.43307752121437504, "grad_norm": 0.2882833182811737, "learning_rate": 9.994607899769082e-05, "loss": 0.0693, "step": 12810 }, { "epoch": 0.4334155989046283, "grad_norm": 0.2734072208404541, "learning_rate": 9.994569446795694e-05, "loss": 0.0623, "step": 12820 }, { "epoch": 0.4337536765948815, "grad_norm": 0.2631424069404602, "learning_rate": 9.994530857272655e-05, "loss": 0.068, "step": 12830 }, { "epoch": 0.43409175428513475, "grad_norm": 0.2429913729429245, "learning_rate": 9.994492131201017e-05, "loss": 0.0698, "step": 12840 }, { "epoch": 0.4344298319753879, "grad_norm": 0.23075774312019348, "learning_rate": 9.994453268581842e-05, "loss": 0.0748, "step": 12850 }, { "epoch": 0.43476790966564116, "grad_norm": 0.29043394327163696, "learning_rate": 9.99441426941619e-05, "loss": 0.0701, "step": 12860 }, { "epoch": 0.4351059873558944, "grad_norm": 0.2524773180484772, "learning_rate": 9.994375133705129e-05, "loss": 0.0719, "step": 12870 }, { "epoch": 0.4354440650461476, "grad_norm": 0.2646607756614685, "learning_rate": 9.994335861449728e-05, "loss": 0.0672, "step": 12880 }, { "epoch": 0.4357821427364008, "grad_norm": 0.3008696436882019, "learning_rate": 9.99429645265106e-05, "loss": 0.0704, "step": 12890 }, { "epoch": 0.43612022042665405, "grad_norm": 0.2666872441768646, "learning_rate": 9.994256907310205e-05, "loss": 0.0636, "step": 12900 }, { "epoch": 0.4364582981169073, "grad_norm": 0.22257301211357117, "learning_rate": 9.994217225428241e-05, "loss": 0.0672, "step": 12910 }, { "epoch": 0.43679637580716046, "grad_norm": 0.2861250340938568, "learning_rate": 9.994177407006255e-05, "loss": 0.0764, "step": 12920 }, { "epoch": 0.4371344534974137, "grad_norm": 0.2757081985473633, "learning_rate": 9.994137452045333e-05, "loss": 0.0638, "step": 12930 }, { "epoch": 0.43747253118766694, "grad_norm": 0.1802682876586914, "learning_rate": 9.994097360546572e-05, "loss": 0.0631, "step": 12940 }, { "epoch": 0.43781060887792017, "grad_norm": 0.17232397198677063, "learning_rate": 9.994057132511063e-05, "loss": 0.0651, "step": 12950 }, { "epoch": 0.43814868656817335, "grad_norm": 0.19928057491779327, "learning_rate": 9.99401676793991e-05, "loss": 0.0642, "step": 12960 }, { "epoch": 0.4384867642584266, "grad_norm": 0.31933078169822693, "learning_rate": 9.993976266834214e-05, "loss": 0.0658, "step": 12970 }, { "epoch": 0.4388248419486798, "grad_norm": 0.2696523666381836, "learning_rate": 9.993935629195084e-05, "loss": 0.0711, "step": 12980 }, { "epoch": 0.439162919638933, "grad_norm": 0.33063438534736633, "learning_rate": 9.993894855023629e-05, "loss": 0.0621, "step": 12990 }, { "epoch": 0.43950099732918624, "grad_norm": 0.18768073618412018, "learning_rate": 9.993853944320964e-05, "loss": 0.0708, "step": 13000 }, { "epoch": 0.4398390750194395, "grad_norm": 0.2509642541408539, "learning_rate": 9.99381289708821e-05, "loss": 0.0669, "step": 13010 }, { "epoch": 0.4401771527096927, "grad_norm": 0.273671418428421, "learning_rate": 9.993771713326487e-05, "loss": 0.0682, "step": 13020 }, { "epoch": 0.4405152303999459, "grad_norm": 0.2776893377304077, "learning_rate": 9.99373039303692e-05, "loss": 0.07, "step": 13030 }, { "epoch": 0.4408533080901991, "grad_norm": 0.2134074568748474, "learning_rate": 9.993688936220642e-05, "loss": 0.0724, "step": 13040 }, { "epoch": 0.44119138578045236, "grad_norm": 0.29132354259490967, "learning_rate": 9.993647342878783e-05, "loss": 0.0711, "step": 13050 }, { "epoch": 0.44152946347070554, "grad_norm": 0.3323717713356018, "learning_rate": 9.993605613012483e-05, "loss": 0.0662, "step": 13060 }, { "epoch": 0.4418675411609588, "grad_norm": 0.21842509508132935, "learning_rate": 9.993563746622881e-05, "loss": 0.0633, "step": 13070 }, { "epoch": 0.442205618851212, "grad_norm": 0.26106441020965576, "learning_rate": 9.993521743711121e-05, "loss": 0.0647, "step": 13080 }, { "epoch": 0.44254369654146525, "grad_norm": 0.33138126134872437, "learning_rate": 9.993479604278354e-05, "loss": 0.067, "step": 13090 }, { "epoch": 0.44288177423171843, "grad_norm": 0.17634673416614532, "learning_rate": 9.99343732832573e-05, "loss": 0.0713, "step": 13100 }, { "epoch": 0.44321985192197166, "grad_norm": 0.29864954948425293, "learning_rate": 9.993394915854406e-05, "loss": 0.0701, "step": 13110 }, { "epoch": 0.4435579296122249, "grad_norm": 0.276274710893631, "learning_rate": 9.99335236686554e-05, "loss": 0.0669, "step": 13120 }, { "epoch": 0.44389600730247813, "grad_norm": 0.30800220370292664, "learning_rate": 9.993309681360297e-05, "loss": 0.0641, "step": 13130 }, { "epoch": 0.4442340849927313, "grad_norm": 0.41428300738334656, "learning_rate": 9.993266859339842e-05, "loss": 0.0669, "step": 13140 }, { "epoch": 0.44457216268298455, "grad_norm": 0.31375542283058167, "learning_rate": 9.993223900805348e-05, "loss": 0.0675, "step": 13150 }, { "epoch": 0.4449102403732378, "grad_norm": 0.2637532949447632, "learning_rate": 9.993180805757989e-05, "loss": 0.0667, "step": 13160 }, { "epoch": 0.44524831806349097, "grad_norm": 0.3249589800834656, "learning_rate": 9.993137574198943e-05, "loss": 0.0636, "step": 13170 }, { "epoch": 0.4455863957537442, "grad_norm": 0.20818910002708435, "learning_rate": 9.99309420612939e-05, "loss": 0.0662, "step": 13180 }, { "epoch": 0.44592447344399744, "grad_norm": 0.23057913780212402, "learning_rate": 9.993050701550517e-05, "loss": 0.0659, "step": 13190 }, { "epoch": 0.4462625511342507, "grad_norm": 0.3566499650478363, "learning_rate": 9.993007060463515e-05, "loss": 0.0634, "step": 13200 }, { "epoch": 0.44660062882450385, "grad_norm": 0.29271218180656433, "learning_rate": 9.992963282869575e-05, "loss": 0.0657, "step": 13210 }, { "epoch": 0.4469387065147571, "grad_norm": 0.34998592734336853, "learning_rate": 9.992919368769895e-05, "loss": 0.0677, "step": 13220 }, { "epoch": 0.4472767842050103, "grad_norm": 0.3148688077926636, "learning_rate": 9.992875318165676e-05, "loss": 0.066, "step": 13230 }, { "epoch": 0.4476148618952635, "grad_norm": 0.2655922472476959, "learning_rate": 9.992831131058119e-05, "loss": 0.0588, "step": 13240 }, { "epoch": 0.44795293958551674, "grad_norm": 0.20373718440532684, "learning_rate": 9.992786807448436e-05, "loss": 0.0656, "step": 13250 }, { "epoch": 0.44829101727577, "grad_norm": 0.26415395736694336, "learning_rate": 9.992742347337837e-05, "loss": 0.0675, "step": 13260 }, { "epoch": 0.4486290949660232, "grad_norm": 0.3387272357940674, "learning_rate": 9.992697750727539e-05, "loss": 0.0679, "step": 13270 }, { "epoch": 0.4489671726562764, "grad_norm": 0.2721927762031555, "learning_rate": 9.99265301761876e-05, "loss": 0.0636, "step": 13280 }, { "epoch": 0.4493052503465296, "grad_norm": 0.23051641881465912, "learning_rate": 9.992608148012722e-05, "loss": 0.0649, "step": 13290 }, { "epoch": 0.44964332803678286, "grad_norm": 0.2425537407398224, "learning_rate": 9.992563141910654e-05, "loss": 0.0603, "step": 13300 }, { "epoch": 0.4499814057270361, "grad_norm": 0.3817272484302521, "learning_rate": 9.992517999313784e-05, "loss": 0.065, "step": 13310 }, { "epoch": 0.4503194834172893, "grad_norm": 0.28327059745788574, "learning_rate": 9.992472720223349e-05, "loss": 0.0766, "step": 13320 }, { "epoch": 0.4506575611075425, "grad_norm": 0.25764408707618713, "learning_rate": 9.992427304640584e-05, "loss": 0.0646, "step": 13330 }, { "epoch": 0.45099563879779575, "grad_norm": 0.26377853751182556, "learning_rate": 9.992381752566733e-05, "loss": 0.074, "step": 13340 }, { "epoch": 0.45133371648804893, "grad_norm": 0.32816439867019653, "learning_rate": 9.992336064003041e-05, "loss": 0.0717, "step": 13350 }, { "epoch": 0.45167179417830217, "grad_norm": 0.2626783847808838, "learning_rate": 9.992290238950754e-05, "loss": 0.07, "step": 13360 }, { "epoch": 0.4520098718685554, "grad_norm": 0.4278288185596466, "learning_rate": 9.99224427741113e-05, "loss": 0.0705, "step": 13370 }, { "epoch": 0.45234794955880864, "grad_norm": 0.370523601770401, "learning_rate": 9.992198179385423e-05, "loss": 0.0645, "step": 13380 }, { "epoch": 0.4526860272490618, "grad_norm": 0.2531181573867798, "learning_rate": 9.992151944874893e-05, "loss": 0.0697, "step": 13390 }, { "epoch": 0.45302410493931505, "grad_norm": 0.29898667335510254, "learning_rate": 9.992105573880804e-05, "loss": 0.0625, "step": 13400 }, { "epoch": 0.4533621826295683, "grad_norm": 0.3449639678001404, "learning_rate": 9.992059066404422e-05, "loss": 0.0622, "step": 13410 }, { "epoch": 0.45370026031982147, "grad_norm": 0.1901799589395523, "learning_rate": 9.992012422447022e-05, "loss": 0.0601, "step": 13420 }, { "epoch": 0.4540383380100747, "grad_norm": 0.3326184153556824, "learning_rate": 9.99196564200988e-05, "loss": 0.0697, "step": 13430 }, { "epoch": 0.45437641570032794, "grad_norm": 0.24139095842838287, "learning_rate": 9.991918725094269e-05, "loss": 0.0671, "step": 13440 }, { "epoch": 0.4547144933905812, "grad_norm": 0.22008825838565826, "learning_rate": 9.991871671701479e-05, "loss": 0.0656, "step": 13450 }, { "epoch": 0.45505257108083436, "grad_norm": 0.20897981524467468, "learning_rate": 9.99182448183279e-05, "loss": 0.0667, "step": 13460 }, { "epoch": 0.4553906487710876, "grad_norm": 0.20945513248443604, "learning_rate": 9.991777155489496e-05, "loss": 0.0738, "step": 13470 }, { "epoch": 0.4557287264613408, "grad_norm": 0.3732315003871918, "learning_rate": 9.99172969267289e-05, "loss": 0.0616, "step": 13480 }, { "epoch": 0.45606680415159406, "grad_norm": 0.26957446336746216, "learning_rate": 9.99168209338427e-05, "loss": 0.0631, "step": 13490 }, { "epoch": 0.45640488184184724, "grad_norm": 0.24123626947402954, "learning_rate": 9.991634357624936e-05, "loss": 0.0653, "step": 13500 }, { "epoch": 0.4567429595321005, "grad_norm": 0.27054014801979065, "learning_rate": 9.991586485396193e-05, "loss": 0.0648, "step": 13510 }, { "epoch": 0.4570810372223537, "grad_norm": 0.27296504378318787, "learning_rate": 9.991538476699353e-05, "loss": 0.0554, "step": 13520 }, { "epoch": 0.4574191149126069, "grad_norm": 0.27045464515686035, "learning_rate": 9.991490331535725e-05, "loss": 0.0584, "step": 13530 }, { "epoch": 0.45775719260286013, "grad_norm": 0.3550007939338684, "learning_rate": 9.991442049906628e-05, "loss": 0.0643, "step": 13540 }, { "epoch": 0.45809527029311337, "grad_norm": 0.4102431833744049, "learning_rate": 9.991393631813379e-05, "loss": 0.0717, "step": 13550 }, { "epoch": 0.4584333479833666, "grad_norm": 0.2312033623456955, "learning_rate": 9.991345077257304e-05, "loss": 0.0621, "step": 13560 }, { "epoch": 0.4587714256736198, "grad_norm": 0.22174100577831268, "learning_rate": 9.991296386239729e-05, "loss": 0.0705, "step": 13570 }, { "epoch": 0.459109503363873, "grad_norm": 0.34359294176101685, "learning_rate": 9.991247558761987e-05, "loss": 0.0664, "step": 13580 }, { "epoch": 0.45944758105412625, "grad_norm": 0.27471643686294556, "learning_rate": 9.991198594825411e-05, "loss": 0.0644, "step": 13590 }, { "epoch": 0.45978565874437943, "grad_norm": 0.28386399149894714, "learning_rate": 9.99114949443134e-05, "loss": 0.0718, "step": 13600 }, { "epoch": 0.46012373643463267, "grad_norm": 0.2180245965719223, "learning_rate": 9.991100257581119e-05, "loss": 0.0689, "step": 13610 }, { "epoch": 0.4604618141248859, "grad_norm": 0.2542070150375366, "learning_rate": 9.991050884276089e-05, "loss": 0.0719, "step": 13620 }, { "epoch": 0.46079989181513914, "grad_norm": 0.25713393092155457, "learning_rate": 9.991001374517606e-05, "loss": 0.0688, "step": 13630 }, { "epoch": 0.4611379695053923, "grad_norm": 0.42017000913619995, "learning_rate": 9.990951728307019e-05, "loss": 0.0587, "step": 13640 }, { "epoch": 0.46147604719564556, "grad_norm": 0.20115728676319122, "learning_rate": 9.990901945645687e-05, "loss": 0.0675, "step": 13650 }, { "epoch": 0.4618141248858988, "grad_norm": 0.1744532436132431, "learning_rate": 9.990852026534969e-05, "loss": 0.0654, "step": 13660 }, { "epoch": 0.462152202576152, "grad_norm": 0.3547879159450531, "learning_rate": 9.990801970976234e-05, "loss": 0.0646, "step": 13670 }, { "epoch": 0.4624902802664052, "grad_norm": 0.26078787446022034, "learning_rate": 9.990751778970848e-05, "loss": 0.0619, "step": 13680 }, { "epoch": 0.46282835795665844, "grad_norm": 0.2683429419994354, "learning_rate": 9.990701450520181e-05, "loss": 0.0652, "step": 13690 }, { "epoch": 0.4631664356469117, "grad_norm": 0.3128945231437683, "learning_rate": 9.990650985625613e-05, "loss": 0.0669, "step": 13700 }, { "epoch": 0.46350451333716486, "grad_norm": 0.2967166006565094, "learning_rate": 9.990600384288523e-05, "loss": 0.0654, "step": 13710 }, { "epoch": 0.4638425910274181, "grad_norm": 0.25399842858314514, "learning_rate": 9.990549646510291e-05, "loss": 0.0625, "step": 13720 }, { "epoch": 0.46418066871767133, "grad_norm": 0.4291578233242035, "learning_rate": 9.990498772292309e-05, "loss": 0.0715, "step": 13730 }, { "epoch": 0.46451874640792457, "grad_norm": 0.24693726003170013, "learning_rate": 9.990447761635964e-05, "loss": 0.0673, "step": 13740 }, { "epoch": 0.46485682409817775, "grad_norm": 0.2422996163368225, "learning_rate": 9.990396614542652e-05, "loss": 0.0654, "step": 13750 }, { "epoch": 0.465194901788431, "grad_norm": 0.3792622983455658, "learning_rate": 9.990345331013771e-05, "loss": 0.0643, "step": 13760 }, { "epoch": 0.4655329794786842, "grad_norm": 0.38839006423950195, "learning_rate": 9.990293911050725e-05, "loss": 0.0619, "step": 13770 }, { "epoch": 0.4658710571689374, "grad_norm": 0.44221827387809753, "learning_rate": 9.990242354654916e-05, "loss": 0.0641, "step": 13780 }, { "epoch": 0.46620913485919063, "grad_norm": 0.3465974032878876, "learning_rate": 9.990190661827757e-05, "loss": 0.066, "step": 13790 }, { "epoch": 0.46654721254944387, "grad_norm": 0.3210327923297882, "learning_rate": 9.99013883257066e-05, "loss": 0.0625, "step": 13800 }, { "epoch": 0.4668852902396971, "grad_norm": 0.42188066244125366, "learning_rate": 9.990086866885042e-05, "loss": 0.067, "step": 13810 }, { "epoch": 0.4672233679299503, "grad_norm": 0.23691852390766144, "learning_rate": 9.990034764772323e-05, "loss": 0.0617, "step": 13820 }, { "epoch": 0.4675614456202035, "grad_norm": 0.2082611322402954, "learning_rate": 9.98998252623393e-05, "loss": 0.0677, "step": 13830 }, { "epoch": 0.46789952331045676, "grad_norm": 0.32898664474487305, "learning_rate": 9.989930151271287e-05, "loss": 0.0699, "step": 13840 }, { "epoch": 0.46823760100070994, "grad_norm": 0.28907865285873413, "learning_rate": 9.98987763988583e-05, "loss": 0.0641, "step": 13850 }, { "epoch": 0.46857567869096317, "grad_norm": 0.4207862913608551, "learning_rate": 9.989824992078993e-05, "loss": 0.069, "step": 13860 }, { "epoch": 0.4689137563812164, "grad_norm": 0.27988654375076294, "learning_rate": 9.989772207852215e-05, "loss": 0.0698, "step": 13870 }, { "epoch": 0.46925183407146964, "grad_norm": 0.25114941596984863, "learning_rate": 9.989719287206939e-05, "loss": 0.0673, "step": 13880 }, { "epoch": 0.4695899117617228, "grad_norm": 0.22309887409210205, "learning_rate": 9.989666230144612e-05, "loss": 0.0569, "step": 13890 }, { "epoch": 0.46992798945197606, "grad_norm": 0.2994719445705414, "learning_rate": 9.989613036666686e-05, "loss": 0.0639, "step": 13900 }, { "epoch": 0.4702660671422293, "grad_norm": 0.3425162136554718, "learning_rate": 9.989559706774613e-05, "loss": 0.073, "step": 13910 }, { "epoch": 0.47060414483248253, "grad_norm": 0.2703486680984497, "learning_rate": 9.989506240469852e-05, "loss": 0.0653, "step": 13920 }, { "epoch": 0.4709422225227357, "grad_norm": 0.21433821320533752, "learning_rate": 9.989452637753867e-05, "loss": 0.0665, "step": 13930 }, { "epoch": 0.47128030021298895, "grad_norm": 0.2295757383108139, "learning_rate": 9.989398898628119e-05, "loss": 0.0711, "step": 13940 }, { "epoch": 0.4716183779032422, "grad_norm": 0.24027176201343536, "learning_rate": 9.98934502309408e-05, "loss": 0.0659, "step": 13950 }, { "epoch": 0.47195645559349536, "grad_norm": 0.2613104581832886, "learning_rate": 9.989291011153224e-05, "loss": 0.07, "step": 13960 }, { "epoch": 0.4722945332837486, "grad_norm": 0.1933690756559372, "learning_rate": 9.989236862807023e-05, "loss": 0.0623, "step": 13970 }, { "epoch": 0.47263261097400183, "grad_norm": 0.35075104236602783, "learning_rate": 9.989182578056964e-05, "loss": 0.063, "step": 13980 }, { "epoch": 0.47297068866425507, "grad_norm": 0.33663299679756165, "learning_rate": 9.989128156904524e-05, "loss": 0.0601, "step": 13990 }, { "epoch": 0.47330876635450825, "grad_norm": 0.40002503991127014, "learning_rate": 9.989073599351198e-05, "loss": 0.0624, "step": 14000 }, { "epoch": 0.4736468440447615, "grad_norm": 0.33707869052886963, "learning_rate": 9.989018905398473e-05, "loss": 0.065, "step": 14010 }, { "epoch": 0.4739849217350147, "grad_norm": 0.19508403539657593, "learning_rate": 9.988964075047844e-05, "loss": 0.0697, "step": 14020 }, { "epoch": 0.4743229994252679, "grad_norm": 0.2566433548927307, "learning_rate": 9.988909108300813e-05, "loss": 0.0639, "step": 14030 }, { "epoch": 0.47466107711552114, "grad_norm": 0.21594251692295074, "learning_rate": 9.98885400515888e-05, "loss": 0.0615, "step": 14040 }, { "epoch": 0.47499915480577437, "grad_norm": 0.20603123307228088, "learning_rate": 9.988798765623555e-05, "loss": 0.0691, "step": 14050 }, { "epoch": 0.4753372324960276, "grad_norm": 0.19486874341964722, "learning_rate": 9.988743389696343e-05, "loss": 0.064, "step": 14060 }, { "epoch": 0.4756753101862808, "grad_norm": 0.3995839059352875, "learning_rate": 9.988687877378762e-05, "loss": 0.0642, "step": 14070 }, { "epoch": 0.476013387876534, "grad_norm": 0.18354667723178864, "learning_rate": 9.98863222867233e-05, "loss": 0.0635, "step": 14080 }, { "epoch": 0.47635146556678726, "grad_norm": 0.3121386468410492, "learning_rate": 9.988576443578565e-05, "loss": 0.061, "step": 14090 }, { "epoch": 0.4766895432570405, "grad_norm": 0.22032521665096283, "learning_rate": 9.988520522098996e-05, "loss": 0.0672, "step": 14100 }, { "epoch": 0.4770276209472937, "grad_norm": 0.3279229402542114, "learning_rate": 9.988464464235148e-05, "loss": 0.0591, "step": 14110 }, { "epoch": 0.4773656986375469, "grad_norm": 0.29394933581352234, "learning_rate": 9.988408269988559e-05, "loss": 0.0574, "step": 14120 }, { "epoch": 0.47770377632780014, "grad_norm": 0.31940409541130066, "learning_rate": 9.988351939360758e-05, "loss": 0.0655, "step": 14130 }, { "epoch": 0.4780418540180533, "grad_norm": 0.26790422201156616, "learning_rate": 9.98829547235329e-05, "loss": 0.0641, "step": 14140 }, { "epoch": 0.47837993170830656, "grad_norm": 0.26538214087486267, "learning_rate": 9.988238868967699e-05, "loss": 0.0681, "step": 14150 }, { "epoch": 0.4787180093985598, "grad_norm": 0.2687227427959442, "learning_rate": 9.988182129205532e-05, "loss": 0.0657, "step": 14160 }, { "epoch": 0.47905608708881303, "grad_norm": 0.24283301830291748, "learning_rate": 9.988125253068338e-05, "loss": 0.0635, "step": 14170 }, { "epoch": 0.4793941647790662, "grad_norm": 0.42432278394699097, "learning_rate": 9.988068240557673e-05, "loss": 0.0712, "step": 14180 }, { "epoch": 0.47973224246931945, "grad_norm": 0.2161034643650055, "learning_rate": 9.988011091675098e-05, "loss": 0.0599, "step": 14190 }, { "epoch": 0.4800703201595727, "grad_norm": 0.16422417759895325, "learning_rate": 9.987953806422173e-05, "loss": 0.0665, "step": 14200 }, { "epoch": 0.48040839784982586, "grad_norm": 0.3644326329231262, "learning_rate": 9.987896384800465e-05, "loss": 0.0693, "step": 14210 }, { "epoch": 0.4807464755400791, "grad_norm": 0.24836978316307068, "learning_rate": 9.987838826811543e-05, "loss": 0.0716, "step": 14220 }, { "epoch": 0.48108455323033233, "grad_norm": 0.2667579650878906, "learning_rate": 9.987781132456982e-05, "loss": 0.0682, "step": 14230 }, { "epoch": 0.48142263092058557, "grad_norm": 0.3760448098182678, "learning_rate": 9.987723301738358e-05, "loss": 0.0692, "step": 14240 }, { "epoch": 0.48176070861083875, "grad_norm": 0.2524842619895935, "learning_rate": 9.987665334657253e-05, "loss": 0.0651, "step": 14250 }, { "epoch": 0.482098786301092, "grad_norm": 0.178732231259346, "learning_rate": 9.987607231215251e-05, "loss": 0.0608, "step": 14260 }, { "epoch": 0.4824368639913452, "grad_norm": 0.28245705366134644, "learning_rate": 9.987548991413942e-05, "loss": 0.062, "step": 14270 }, { "epoch": 0.48277494168159846, "grad_norm": 0.18969948589801788, "learning_rate": 9.987490615254917e-05, "loss": 0.062, "step": 14280 }, { "epoch": 0.48311301937185164, "grad_norm": 0.2796260416507721, "learning_rate": 9.987432102739773e-05, "loss": 0.0628, "step": 14290 }, { "epoch": 0.4834510970621049, "grad_norm": 0.34377557039260864, "learning_rate": 9.987373453870107e-05, "loss": 0.067, "step": 14300 }, { "epoch": 0.4837891747523581, "grad_norm": 0.1986936628818512, "learning_rate": 9.987314668647526e-05, "loss": 0.0718, "step": 14310 }, { "epoch": 0.4841272524426113, "grad_norm": 0.3173161447048187, "learning_rate": 9.987255747073636e-05, "loss": 0.0665, "step": 14320 }, { "epoch": 0.4844653301328645, "grad_norm": 0.2867687940597534, "learning_rate": 9.987196689150046e-05, "loss": 0.0568, "step": 14330 }, { "epoch": 0.48480340782311776, "grad_norm": 0.2957778573036194, "learning_rate": 9.987137494878374e-05, "loss": 0.0667, "step": 14340 }, { "epoch": 0.485141485513371, "grad_norm": 0.25924214720726013, "learning_rate": 9.987078164260235e-05, "loss": 0.0629, "step": 14350 }, { "epoch": 0.4854795632036242, "grad_norm": 0.2779247462749481, "learning_rate": 9.987018697297252e-05, "loss": 0.062, "step": 14360 }, { "epoch": 0.4858176408938774, "grad_norm": 0.21051597595214844, "learning_rate": 9.986959093991053e-05, "loss": 0.0596, "step": 14370 }, { "epoch": 0.48615571858413065, "grad_norm": 0.2893418073654175, "learning_rate": 9.986899354343265e-05, "loss": 0.0606, "step": 14380 }, { "epoch": 0.4864937962743838, "grad_norm": 0.2550332248210907, "learning_rate": 9.986839478355523e-05, "loss": 0.0653, "step": 14390 }, { "epoch": 0.48683187396463706, "grad_norm": 0.2842765152454376, "learning_rate": 9.986779466029461e-05, "loss": 0.0596, "step": 14400 }, { "epoch": 0.4871699516548903, "grad_norm": 0.2085721343755722, "learning_rate": 9.986719317366725e-05, "loss": 0.0613, "step": 14410 }, { "epoch": 0.48750802934514353, "grad_norm": 0.2568814158439636, "learning_rate": 9.986659032368953e-05, "loss": 0.0697, "step": 14420 }, { "epoch": 0.4878461070353967, "grad_norm": 0.24331706762313843, "learning_rate": 9.986598611037799e-05, "loss": 0.0635, "step": 14430 }, { "epoch": 0.48818418472564995, "grad_norm": 0.1793292611837387, "learning_rate": 9.986538053374912e-05, "loss": 0.0646, "step": 14440 }, { "epoch": 0.4885222624159032, "grad_norm": 0.2765239477157593, "learning_rate": 9.986477359381947e-05, "loss": 0.0656, "step": 14450 }, { "epoch": 0.4888603401061564, "grad_norm": 0.25697842240333557, "learning_rate": 9.986416529060564e-05, "loss": 0.0643, "step": 14460 }, { "epoch": 0.4891984177964096, "grad_norm": 0.19346880912780762, "learning_rate": 9.986355562412427e-05, "loss": 0.0632, "step": 14470 }, { "epoch": 0.48953649548666284, "grad_norm": 0.2821345627307892, "learning_rate": 9.986294459439203e-05, "loss": 0.0642, "step": 14480 }, { "epoch": 0.4898745731769161, "grad_norm": 0.22432348132133484, "learning_rate": 9.98623322014256e-05, "loss": 0.0669, "step": 14490 }, { "epoch": 0.49021265086716925, "grad_norm": 0.20044377446174622, "learning_rate": 9.986171844524176e-05, "loss": 0.0633, "step": 14500 }, { "epoch": 0.4905507285574225, "grad_norm": 0.20403791964054108, "learning_rate": 9.986110332585725e-05, "loss": 0.0699, "step": 14510 }, { "epoch": 0.4908888062476757, "grad_norm": 0.33326205611228943, "learning_rate": 9.986048684328893e-05, "loss": 0.0646, "step": 14520 }, { "epoch": 0.49122688393792896, "grad_norm": 0.2573673129081726, "learning_rate": 9.985986899755362e-05, "loss": 0.0671, "step": 14530 }, { "epoch": 0.49156496162818214, "grad_norm": 0.3308497667312622, "learning_rate": 9.985924978866822e-05, "loss": 0.065, "step": 14540 }, { "epoch": 0.4919030393184354, "grad_norm": 0.2998906075954437, "learning_rate": 9.985862921664965e-05, "loss": 0.0664, "step": 14550 }, { "epoch": 0.4922411170086886, "grad_norm": 0.33204197883605957, "learning_rate": 9.985800728151491e-05, "loss": 0.063, "step": 14560 }, { "epoch": 0.4925791946989418, "grad_norm": 0.3038094639778137, "learning_rate": 9.985738398328097e-05, "loss": 0.0737, "step": 14570 }, { "epoch": 0.492917272389195, "grad_norm": 0.2879648208618164, "learning_rate": 9.98567593219649e-05, "loss": 0.0654, "step": 14580 }, { "epoch": 0.49325535007944826, "grad_norm": 0.4249301850795746, "learning_rate": 9.985613329758375e-05, "loss": 0.0617, "step": 14590 }, { "epoch": 0.4935934277697015, "grad_norm": 0.19323863089084625, "learning_rate": 9.985550591015463e-05, "loss": 0.0638, "step": 14600 }, { "epoch": 0.4939315054599547, "grad_norm": 0.2561129033565521, "learning_rate": 9.985487715969473e-05, "loss": 0.0698, "step": 14610 }, { "epoch": 0.4942695831502079, "grad_norm": 0.2018284946680069, "learning_rate": 9.985424704622119e-05, "loss": 0.0689, "step": 14620 }, { "epoch": 0.49460766084046115, "grad_norm": 0.2639164328575134, "learning_rate": 9.985361556975131e-05, "loss": 0.0664, "step": 14630 }, { "epoch": 0.4949457385307144, "grad_norm": 0.21796894073486328, "learning_rate": 9.985298273030226e-05, "loss": 0.0667, "step": 14640 }, { "epoch": 0.49528381622096757, "grad_norm": 0.3494921326637268, "learning_rate": 9.985234852789142e-05, "loss": 0.0673, "step": 14650 }, { "epoch": 0.4956218939112208, "grad_norm": 0.24313542246818542, "learning_rate": 9.98517129625361e-05, "loss": 0.0632, "step": 14660 }, { "epoch": 0.49595997160147404, "grad_norm": 0.29040858149528503, "learning_rate": 9.985107603425367e-05, "loss": 0.0628, "step": 14670 }, { "epoch": 0.4962980492917272, "grad_norm": 0.18936026096343994, "learning_rate": 9.985043774306155e-05, "loss": 0.0636, "step": 14680 }, { "epoch": 0.49663612698198045, "grad_norm": 0.4755830764770508, "learning_rate": 9.984979808897719e-05, "loss": 0.0676, "step": 14690 }, { "epoch": 0.4969742046722337, "grad_norm": 0.19494032859802246, "learning_rate": 9.984915707201807e-05, "loss": 0.0572, "step": 14700 }, { "epoch": 0.4973122823624869, "grad_norm": 0.23603007197380066, "learning_rate": 9.984851469220174e-05, "loss": 0.0642, "step": 14710 }, { "epoch": 0.4976503600527401, "grad_norm": 0.18325138092041016, "learning_rate": 9.984787094954573e-05, "loss": 0.0651, "step": 14720 }, { "epoch": 0.49798843774299334, "grad_norm": 0.3259008824825287, "learning_rate": 9.984722584406767e-05, "loss": 0.0651, "step": 14730 }, { "epoch": 0.4983265154332466, "grad_norm": 0.33778271079063416, "learning_rate": 9.984657937578516e-05, "loss": 0.0627, "step": 14740 }, { "epoch": 0.49866459312349976, "grad_norm": 0.3162662982940674, "learning_rate": 9.984593154471592e-05, "loss": 0.0662, "step": 14750 }, { "epoch": 0.499002670813753, "grad_norm": 0.24531182646751404, "learning_rate": 9.984528235087763e-05, "loss": 0.065, "step": 14760 }, { "epoch": 0.4993407485040062, "grad_norm": 0.23710697889328003, "learning_rate": 9.984463179428804e-05, "loss": 0.0644, "step": 14770 }, { "epoch": 0.49967882619425946, "grad_norm": 0.27748239040374756, "learning_rate": 9.984397987496494e-05, "loss": 0.0631, "step": 14780 }, { "epoch": 0.5000169038845127, "grad_norm": 0.17644034326076508, "learning_rate": 9.984332659292616e-05, "loss": 0.063, "step": 14790 }, { "epoch": 0.5003549815747659, "grad_norm": 0.27662307024002075, "learning_rate": 9.984267194818956e-05, "loss": 0.0643, "step": 14800 }, { "epoch": 0.5006930592650191, "grad_norm": 0.4204474985599518, "learning_rate": 9.984201594077304e-05, "loss": 0.0617, "step": 14810 }, { "epoch": 0.5010311369552723, "grad_norm": 0.36193138360977173, "learning_rate": 9.984135857069452e-05, "loss": 0.0606, "step": 14820 }, { "epoch": 0.5013692146455255, "grad_norm": 0.33228200674057007, "learning_rate": 9.984069983797199e-05, "loss": 0.0645, "step": 14830 }, { "epoch": 0.5017072923357787, "grad_norm": 0.29374879598617554, "learning_rate": 9.984003974262344e-05, "loss": 0.0622, "step": 14840 }, { "epoch": 0.502045370026032, "grad_norm": 0.19842693209648132, "learning_rate": 9.983937828466694e-05, "loss": 0.0664, "step": 14850 }, { "epoch": 0.5023834477162852, "grad_norm": 0.23629926145076752, "learning_rate": 9.983871546412054e-05, "loss": 0.0677, "step": 14860 }, { "epoch": 0.5027215254065385, "grad_norm": 0.29102975130081177, "learning_rate": 9.98380512810024e-05, "loss": 0.0646, "step": 14870 }, { "epoch": 0.5030596030967917, "grad_norm": 0.27075880765914917, "learning_rate": 9.983738573533065e-05, "loss": 0.0639, "step": 14880 }, { "epoch": 0.5033976807870448, "grad_norm": 0.2264859676361084, "learning_rate": 9.983671882712351e-05, "loss": 0.0696, "step": 14890 }, { "epoch": 0.5037357584772981, "grad_norm": 0.27164751291275024, "learning_rate": 9.983605055639919e-05, "loss": 0.0706, "step": 14900 }, { "epoch": 0.5040738361675513, "grad_norm": 0.24910157918930054, "learning_rate": 9.983538092317599e-05, "loss": 0.0627, "step": 14910 }, { "epoch": 0.5044119138578045, "grad_norm": 0.30848026275634766, "learning_rate": 9.983470992747218e-05, "loss": 0.0693, "step": 14920 }, { "epoch": 0.5047499915480578, "grad_norm": 0.3285810053348541, "learning_rate": 9.983403756930611e-05, "loss": 0.0708, "step": 14930 }, { "epoch": 0.505088069238311, "grad_norm": 0.23453210294246674, "learning_rate": 9.98333638486962e-05, "loss": 0.0657, "step": 14940 }, { "epoch": 0.5054261469285641, "grad_norm": 0.32704219222068787, "learning_rate": 9.983268876566083e-05, "loss": 0.0619, "step": 14950 }, { "epoch": 0.5057642246188174, "grad_norm": 0.2642788887023926, "learning_rate": 9.983201232021847e-05, "loss": 0.0617, "step": 14960 }, { "epoch": 0.5061023023090706, "grad_norm": 0.3250301778316498, "learning_rate": 9.983133451238762e-05, "loss": 0.0667, "step": 14970 }, { "epoch": 0.5064403799993239, "grad_norm": 0.2673223316669464, "learning_rate": 9.98306553421868e-05, "loss": 0.07, "step": 14980 }, { "epoch": 0.5067784576895771, "grad_norm": 0.29321759939193726, "learning_rate": 9.982997480963458e-05, "loss": 0.0637, "step": 14990 }, { "epoch": 0.5071165353798303, "grad_norm": 0.2277747392654419, "learning_rate": 9.982929291474958e-05, "loss": 0.0611, "step": 15000 }, { "epoch": 0.5074546130700835, "grad_norm": 0.351171612739563, "learning_rate": 9.982860965755043e-05, "loss": 0.0709, "step": 15010 }, { "epoch": 0.5077926907603367, "grad_norm": 0.27540209889411926, "learning_rate": 9.98279250380558e-05, "loss": 0.0668, "step": 15020 }, { "epoch": 0.5081307684505899, "grad_norm": 0.28690752387046814, "learning_rate": 9.982723905628444e-05, "loss": 0.0602, "step": 15030 }, { "epoch": 0.5084688461408432, "grad_norm": 0.3330153822898865, "learning_rate": 9.982655171225507e-05, "loss": 0.0638, "step": 15040 }, { "epoch": 0.5088069238310964, "grad_norm": 0.3244529962539673, "learning_rate": 9.982586300598648e-05, "loss": 0.062, "step": 15050 }, { "epoch": 0.5091450015213496, "grad_norm": 0.21078094840049744, "learning_rate": 9.982517293749755e-05, "loss": 0.0645, "step": 15060 }, { "epoch": 0.5094830792116029, "grad_norm": 0.3135485053062439, "learning_rate": 9.98244815068071e-05, "loss": 0.0684, "step": 15070 }, { "epoch": 0.509821156901856, "grad_norm": 0.26427963376045227, "learning_rate": 9.982378871393402e-05, "loss": 0.0658, "step": 15080 }, { "epoch": 0.5101592345921092, "grad_norm": 0.2971445620059967, "learning_rate": 9.982309455889729e-05, "loss": 0.0687, "step": 15090 }, { "epoch": 0.5104973122823625, "grad_norm": 0.27060696482658386, "learning_rate": 9.982239904171587e-05, "loss": 0.0646, "step": 15100 }, { "epoch": 0.5108353899726157, "grad_norm": 0.23709343373775482, "learning_rate": 9.982170216240877e-05, "loss": 0.0702, "step": 15110 }, { "epoch": 0.511173467662869, "grad_norm": 0.28531280159950256, "learning_rate": 9.982100392099507e-05, "loss": 0.0682, "step": 15120 }, { "epoch": 0.5115115453531222, "grad_norm": 0.33657893538475037, "learning_rate": 9.982030431749383e-05, "loss": 0.068, "step": 15130 }, { "epoch": 0.5118496230433753, "grad_norm": 0.27469751238822937, "learning_rate": 9.981960335192418e-05, "loss": 0.0638, "step": 15140 }, { "epoch": 0.5121877007336286, "grad_norm": 0.3119707405567169, "learning_rate": 9.98189010243053e-05, "loss": 0.0668, "step": 15150 }, { "epoch": 0.5125257784238818, "grad_norm": 0.24530602991580963, "learning_rate": 9.981819733465639e-05, "loss": 0.0647, "step": 15160 }, { "epoch": 0.512863856114135, "grad_norm": 0.22370091080665588, "learning_rate": 9.981749228299666e-05, "loss": 0.0667, "step": 15170 }, { "epoch": 0.5132019338043883, "grad_norm": 0.18020951747894287, "learning_rate": 9.981678586934541e-05, "loss": 0.0627, "step": 15180 }, { "epoch": 0.5135400114946415, "grad_norm": 0.22811463475227356, "learning_rate": 9.981607809372195e-05, "loss": 0.0643, "step": 15190 }, { "epoch": 0.5138780891848946, "grad_norm": 0.39137327671051025, "learning_rate": 9.981536895614564e-05, "loss": 0.0646, "step": 15200 }, { "epoch": 0.5142161668751479, "grad_norm": 0.22575944662094116, "learning_rate": 9.981465845663584e-05, "loss": 0.0648, "step": 15210 }, { "epoch": 0.5145542445654011, "grad_norm": 0.35739269852638245, "learning_rate": 9.981394659521202e-05, "loss": 0.0588, "step": 15220 }, { "epoch": 0.5148923222556544, "grad_norm": 0.22549591958522797, "learning_rate": 9.98132333718936e-05, "loss": 0.0647, "step": 15230 }, { "epoch": 0.5152303999459076, "grad_norm": 0.23473916947841644, "learning_rate": 9.981251878670008e-05, "loss": 0.0654, "step": 15240 }, { "epoch": 0.5155684776361608, "grad_norm": 0.22827908396720886, "learning_rate": 9.981180283965104e-05, "loss": 0.0568, "step": 15250 }, { "epoch": 0.515906555326414, "grad_norm": 0.2588604688644409, "learning_rate": 9.981108553076599e-05, "loss": 0.0615, "step": 15260 }, { "epoch": 0.5162446330166672, "grad_norm": 0.45579564571380615, "learning_rate": 9.981036686006459e-05, "loss": 0.0669, "step": 15270 }, { "epoch": 0.5165827107069204, "grad_norm": 0.2596732974052429, "learning_rate": 9.980964682756648e-05, "loss": 0.0666, "step": 15280 }, { "epoch": 0.5169207883971737, "grad_norm": 0.20895175635814667, "learning_rate": 9.980892543329135e-05, "loss": 0.065, "step": 15290 }, { "epoch": 0.5172588660874269, "grad_norm": 0.2932634651660919, "learning_rate": 9.980820267725888e-05, "loss": 0.0711, "step": 15300 }, { "epoch": 0.5175969437776801, "grad_norm": 0.2623646855354309, "learning_rate": 9.980747855948887e-05, "loss": 0.0571, "step": 15310 }, { "epoch": 0.5179350214679334, "grad_norm": 0.29273152351379395, "learning_rate": 9.980675308000111e-05, "loss": 0.0608, "step": 15320 }, { "epoch": 0.5182730991581865, "grad_norm": 0.3178764283657074, "learning_rate": 9.980602623881544e-05, "loss": 0.0678, "step": 15330 }, { "epoch": 0.5186111768484398, "grad_norm": 0.5216978788375854, "learning_rate": 9.980529803595172e-05, "loss": 0.0597, "step": 15340 }, { "epoch": 0.518949254538693, "grad_norm": 0.27574920654296875, "learning_rate": 9.980456847142986e-05, "loss": 0.065, "step": 15350 }, { "epoch": 0.5192873322289462, "grad_norm": 0.5151643753051758, "learning_rate": 9.980383754526981e-05, "loss": 0.0687, "step": 15360 }, { "epoch": 0.5196254099191995, "grad_norm": 0.22341565787792206, "learning_rate": 9.980310525749154e-05, "loss": 0.0646, "step": 15370 }, { "epoch": 0.5199634876094527, "grad_norm": 0.3058348000049591, "learning_rate": 9.98023716081151e-05, "loss": 0.069, "step": 15380 }, { "epoch": 0.5203015652997058, "grad_norm": 0.3085953891277313, "learning_rate": 9.980163659716053e-05, "loss": 0.0694, "step": 15390 }, { "epoch": 0.5206396429899591, "grad_norm": 0.2741049826145172, "learning_rate": 9.980090022464792e-05, "loss": 0.0654, "step": 15400 }, { "epoch": 0.5209777206802123, "grad_norm": 0.2315359264612198, "learning_rate": 9.980016249059741e-05, "loss": 0.0628, "step": 15410 }, { "epoch": 0.5213157983704655, "grad_norm": 0.3198952078819275, "learning_rate": 9.979942339502917e-05, "loss": 0.0588, "step": 15420 }, { "epoch": 0.5216538760607188, "grad_norm": 0.23331983387470245, "learning_rate": 9.979868293796339e-05, "loss": 0.0657, "step": 15430 }, { "epoch": 0.521991953750972, "grad_norm": 0.19891776144504547, "learning_rate": 9.979794111942033e-05, "loss": 0.0646, "step": 15440 }, { "epoch": 0.5223300314412251, "grad_norm": 0.33102256059646606, "learning_rate": 9.979719793942027e-05, "loss": 0.067, "step": 15450 }, { "epoch": 0.5226681091314784, "grad_norm": 0.18174780905246735, "learning_rate": 9.979645339798354e-05, "loss": 0.0591, "step": 15460 }, { "epoch": 0.5230061868217316, "grad_norm": 0.20456469058990479, "learning_rate": 9.979570749513045e-05, "loss": 0.0638, "step": 15470 }, { "epoch": 0.5233442645119849, "grad_norm": 0.21242831647396088, "learning_rate": 9.979496023088145e-05, "loss": 0.0733, "step": 15480 }, { "epoch": 0.5236823422022381, "grad_norm": 0.3269086480140686, "learning_rate": 9.979421160525694e-05, "loss": 0.0598, "step": 15490 }, { "epoch": 0.5240204198924913, "grad_norm": 0.20283904671669006, "learning_rate": 9.979346161827739e-05, "loss": 0.0669, "step": 15500 }, { "epoch": 0.5243584975827446, "grad_norm": 0.23733848333358765, "learning_rate": 9.979271026996331e-05, "loss": 0.0574, "step": 15510 }, { "epoch": 0.5246965752729977, "grad_norm": 0.45683926343917847, "learning_rate": 9.979195756033523e-05, "loss": 0.0626, "step": 15520 }, { "epoch": 0.5250346529632509, "grad_norm": 0.23333211243152618, "learning_rate": 9.979120348941374e-05, "loss": 0.0632, "step": 15530 }, { "epoch": 0.5253727306535042, "grad_norm": 0.2338869720697403, "learning_rate": 9.979044805721944e-05, "loss": 0.0634, "step": 15540 }, { "epoch": 0.5257108083437574, "grad_norm": 0.21049043536186218, "learning_rate": 9.978969126377302e-05, "loss": 0.0637, "step": 15550 }, { "epoch": 0.5260488860340106, "grad_norm": 0.25200390815734863, "learning_rate": 9.978893310909513e-05, "loss": 0.063, "step": 15560 }, { "epoch": 0.5263869637242639, "grad_norm": 0.20405127108097076, "learning_rate": 9.978817359320651e-05, "loss": 0.0547, "step": 15570 }, { "epoch": 0.526725041414517, "grad_norm": 0.22704556584358215, "learning_rate": 9.978741271612795e-05, "loss": 0.0644, "step": 15580 }, { "epoch": 0.5270631191047703, "grad_norm": 0.3919825553894043, "learning_rate": 9.97866504778802e-05, "loss": 0.0705, "step": 15590 }, { "epoch": 0.5274011967950235, "grad_norm": 0.3978733718395233, "learning_rate": 9.978588687848415e-05, "loss": 0.0637, "step": 15600 }, { "epoch": 0.5277392744852767, "grad_norm": 0.23407261073589325, "learning_rate": 9.978512191796065e-05, "loss": 0.059, "step": 15610 }, { "epoch": 0.52807735217553, "grad_norm": 0.25549235939979553, "learning_rate": 9.978435559633064e-05, "loss": 0.06, "step": 15620 }, { "epoch": 0.5284154298657832, "grad_norm": 0.20849154889583588, "learning_rate": 9.978358791361501e-05, "loss": 0.0676, "step": 15630 }, { "epoch": 0.5287535075560363, "grad_norm": 0.3509672284126282, "learning_rate": 9.978281886983483e-05, "loss": 0.0617, "step": 15640 }, { "epoch": 0.5290915852462896, "grad_norm": 0.221439927816391, "learning_rate": 9.978204846501107e-05, "loss": 0.0588, "step": 15650 }, { "epoch": 0.5294296629365428, "grad_norm": 0.29249605536460876, "learning_rate": 9.97812766991648e-05, "loss": 0.0561, "step": 15660 }, { "epoch": 0.529767740626796, "grad_norm": 0.31124040484428406, "learning_rate": 9.978050357231713e-05, "loss": 0.0639, "step": 15670 }, { "epoch": 0.5301058183170493, "grad_norm": 0.27765554189682007, "learning_rate": 9.977972908448919e-05, "loss": 0.0652, "step": 15680 }, { "epoch": 0.5304438960073025, "grad_norm": 0.2148120403289795, "learning_rate": 9.977895323570217e-05, "loss": 0.0634, "step": 15690 }, { "epoch": 0.5307819736975558, "grad_norm": 0.2793019115924835, "learning_rate": 9.977817602597726e-05, "loss": 0.0623, "step": 15700 }, { "epoch": 0.5311200513878089, "grad_norm": 0.33366671204566956, "learning_rate": 9.977739745533572e-05, "loss": 0.0695, "step": 15710 }, { "epoch": 0.5314581290780621, "grad_norm": 0.26681071519851685, "learning_rate": 9.977661752379883e-05, "loss": 0.0641, "step": 15720 }, { "epoch": 0.5317962067683154, "grad_norm": 0.19598881900310516, "learning_rate": 9.977583623138793e-05, "loss": 0.0689, "step": 15730 }, { "epoch": 0.5321342844585686, "grad_norm": 0.4234233498573303, "learning_rate": 9.977505357812436e-05, "loss": 0.0613, "step": 15740 }, { "epoch": 0.5324723621488218, "grad_norm": 0.22945500910282135, "learning_rate": 9.977426956402953e-05, "loss": 0.0781, "step": 15750 }, { "epoch": 0.5328104398390751, "grad_norm": 0.24133343994617462, "learning_rate": 9.977348418912486e-05, "loss": 0.0619, "step": 15760 }, { "epoch": 0.5331485175293282, "grad_norm": 0.18251192569732666, "learning_rate": 9.977269745343183e-05, "loss": 0.0633, "step": 15770 }, { "epoch": 0.5334865952195814, "grad_norm": 0.2974420487880707, "learning_rate": 9.977190935697196e-05, "loss": 0.0617, "step": 15780 }, { "epoch": 0.5338246729098347, "grad_norm": 0.321062833070755, "learning_rate": 9.977111989976678e-05, "loss": 0.0629, "step": 15790 }, { "epoch": 0.5341627506000879, "grad_norm": 0.2615441381931305, "learning_rate": 9.977032908183789e-05, "loss": 0.0673, "step": 15800 }, { "epoch": 0.5345008282903411, "grad_norm": 0.32128992676734924, "learning_rate": 9.97695369032069e-05, "loss": 0.0734, "step": 15810 }, { "epoch": 0.5348389059805944, "grad_norm": 0.24726881086826324, "learning_rate": 9.976874336389546e-05, "loss": 0.0698, "step": 15820 }, { "epoch": 0.5351769836708475, "grad_norm": 0.3053199052810669, "learning_rate": 9.976794846392527e-05, "loss": 0.0649, "step": 15830 }, { "epoch": 0.5355150613611008, "grad_norm": 0.2548052966594696, "learning_rate": 9.976715220331807e-05, "loss": 0.0604, "step": 15840 }, { "epoch": 0.535853139051354, "grad_norm": 0.19669011235237122, "learning_rate": 9.976635458209563e-05, "loss": 0.0607, "step": 15850 }, { "epoch": 0.5361912167416072, "grad_norm": 0.3797494173049927, "learning_rate": 9.976555560027975e-05, "loss": 0.073, "step": 15860 }, { "epoch": 0.5365292944318605, "grad_norm": 0.18251991271972656, "learning_rate": 9.976475525789227e-05, "loss": 0.0592, "step": 15870 }, { "epoch": 0.5368673721221137, "grad_norm": 0.2521688640117645, "learning_rate": 9.976395355495509e-05, "loss": 0.0601, "step": 15880 }, { "epoch": 0.5372054498123668, "grad_norm": 0.33601030707359314, "learning_rate": 9.976315049149009e-05, "loss": 0.0649, "step": 15890 }, { "epoch": 0.5375435275026201, "grad_norm": 0.346666157245636, "learning_rate": 9.976234606751928e-05, "loss": 0.0659, "step": 15900 }, { "epoch": 0.5378816051928733, "grad_norm": 0.18434731662273407, "learning_rate": 9.976154028306461e-05, "loss": 0.0585, "step": 15910 }, { "epoch": 0.5382196828831265, "grad_norm": 0.2485870122909546, "learning_rate": 9.976073313814814e-05, "loss": 0.0623, "step": 15920 }, { "epoch": 0.5385577605733798, "grad_norm": 0.3976130187511444, "learning_rate": 9.975992463279189e-05, "loss": 0.0679, "step": 15930 }, { "epoch": 0.538895838263633, "grad_norm": 0.24570141732692719, "learning_rate": 9.975911476701801e-05, "loss": 0.0652, "step": 15940 }, { "epoch": 0.5392339159538863, "grad_norm": 0.27598896622657776, "learning_rate": 9.975830354084863e-05, "loss": 0.0656, "step": 15950 }, { "epoch": 0.5395719936441394, "grad_norm": 0.2415866255760193, "learning_rate": 9.975749095430591e-05, "loss": 0.0671, "step": 15960 }, { "epoch": 0.5399100713343926, "grad_norm": 0.4326671063899994, "learning_rate": 9.97566770074121e-05, "loss": 0.0675, "step": 15970 }, { "epoch": 0.5402481490246459, "grad_norm": 0.24821072816848755, "learning_rate": 9.975586170018943e-05, "loss": 0.0641, "step": 15980 }, { "epoch": 0.5405862267148991, "grad_norm": 0.2925005555152893, "learning_rate": 9.975504503266017e-05, "loss": 0.0589, "step": 15990 }, { "epoch": 0.5409243044051523, "grad_norm": 0.28232184052467346, "learning_rate": 9.975422700484669e-05, "loss": 0.0611, "step": 16000 }, { "epoch": 0.5412623820954056, "grad_norm": 0.30535486340522766, "learning_rate": 9.975340761677134e-05, "loss": 0.0635, "step": 16010 }, { "epoch": 0.5416004597856587, "grad_norm": 0.291109174489975, "learning_rate": 9.97525868684565e-05, "loss": 0.0607, "step": 16020 }, { "epoch": 0.5419385374759119, "grad_norm": 0.25257232785224915, "learning_rate": 9.975176475992465e-05, "loss": 0.0572, "step": 16030 }, { "epoch": 0.5422766151661652, "grad_norm": 0.26134878396987915, "learning_rate": 9.975094129119822e-05, "loss": 0.06, "step": 16040 }, { "epoch": 0.5426146928564184, "grad_norm": 0.22909830510616302, "learning_rate": 9.975011646229976e-05, "loss": 0.0599, "step": 16050 }, { "epoch": 0.5429527705466716, "grad_norm": 0.2246238738298416, "learning_rate": 9.974929027325179e-05, "loss": 0.0685, "step": 16060 }, { "epoch": 0.5432908482369249, "grad_norm": 0.24652430415153503, "learning_rate": 9.974846272407694e-05, "loss": 0.0626, "step": 16070 }, { "epoch": 0.543628925927178, "grad_norm": 0.19887204468250275, "learning_rate": 9.97476338147978e-05, "loss": 0.0705, "step": 16080 }, { "epoch": 0.5439670036174313, "grad_norm": 0.22814121842384338, "learning_rate": 9.974680354543702e-05, "loss": 0.0675, "step": 16090 }, { "epoch": 0.5443050813076845, "grad_norm": 0.2899927496910095, "learning_rate": 9.974597191601733e-05, "loss": 0.0616, "step": 16100 }, { "epoch": 0.5446431589979377, "grad_norm": 0.24342066049575806, "learning_rate": 9.974513892656147e-05, "loss": 0.0576, "step": 16110 }, { "epoch": 0.544981236688191, "grad_norm": 0.22099342942237854, "learning_rate": 9.974430457709218e-05, "loss": 0.0726, "step": 16120 }, { "epoch": 0.5453193143784442, "grad_norm": 0.32875728607177734, "learning_rate": 9.974346886763231e-05, "loss": 0.064, "step": 16130 }, { "epoch": 0.5456573920686973, "grad_norm": 0.400307297706604, "learning_rate": 9.974263179820469e-05, "loss": 0.0643, "step": 16140 }, { "epoch": 0.5459954697589506, "grad_norm": 0.20545603334903717, "learning_rate": 9.97417933688322e-05, "loss": 0.0655, "step": 16150 }, { "epoch": 0.5463335474492038, "grad_norm": 0.2499261349439621, "learning_rate": 9.974095357953774e-05, "loss": 0.0603, "step": 16160 }, { "epoch": 0.546671625139457, "grad_norm": 0.2648391127586365, "learning_rate": 9.974011243034434e-05, "loss": 0.0633, "step": 16170 }, { "epoch": 0.5470097028297103, "grad_norm": 0.34700846672058105, "learning_rate": 9.973926992127491e-05, "loss": 0.0706, "step": 16180 }, { "epoch": 0.5473477805199635, "grad_norm": 0.24321120977401733, "learning_rate": 9.973842605235255e-05, "loss": 0.0668, "step": 16190 }, { "epoch": 0.5476858582102168, "grad_norm": 0.22601909935474396, "learning_rate": 9.973758082360031e-05, "loss": 0.0726, "step": 16200 }, { "epoch": 0.5480239359004699, "grad_norm": 0.2540223300457001, "learning_rate": 9.973673423504128e-05, "loss": 0.0595, "step": 16210 }, { "epoch": 0.5483620135907231, "grad_norm": 0.21116675436496735, "learning_rate": 9.973588628669865e-05, "loss": 0.0603, "step": 16220 }, { "epoch": 0.5487000912809764, "grad_norm": 0.2093290388584137, "learning_rate": 9.973503697859554e-05, "loss": 0.065, "step": 16230 }, { "epoch": 0.5490381689712296, "grad_norm": 0.280485063791275, "learning_rate": 9.973418631075521e-05, "loss": 0.0632, "step": 16240 }, { "epoch": 0.5493762466614828, "grad_norm": 0.22660639882087708, "learning_rate": 9.973333428320094e-05, "loss": 0.0644, "step": 16250 }, { "epoch": 0.5497143243517361, "grad_norm": 0.25961437821388245, "learning_rate": 9.973248089595598e-05, "loss": 0.0675, "step": 16260 }, { "epoch": 0.5500524020419892, "grad_norm": 0.22672291100025177, "learning_rate": 9.973162614904365e-05, "loss": 0.0635, "step": 16270 }, { "epoch": 0.5503904797322424, "grad_norm": 0.15443523228168488, "learning_rate": 9.973077004248737e-05, "loss": 0.0605, "step": 16280 }, { "epoch": 0.5507285574224957, "grad_norm": 0.1876388043165207, "learning_rate": 9.972991257631052e-05, "loss": 0.0583, "step": 16290 }, { "epoch": 0.5510666351127489, "grad_norm": 0.22011397778987885, "learning_rate": 9.972905375053654e-05, "loss": 0.0671, "step": 16300 }, { "epoch": 0.5514047128030022, "grad_norm": 0.21135754883289337, "learning_rate": 9.972819356518889e-05, "loss": 0.0596, "step": 16310 }, { "epoch": 0.5517427904932554, "grad_norm": 0.28918391466140747, "learning_rate": 9.972733202029115e-05, "loss": 0.0558, "step": 16320 }, { "epoch": 0.5520808681835085, "grad_norm": 0.19357790052890778, "learning_rate": 9.97264691158668e-05, "loss": 0.0621, "step": 16330 }, { "epoch": 0.5524189458737618, "grad_norm": 0.2461680769920349, "learning_rate": 9.972560485193949e-05, "loss": 0.0612, "step": 16340 }, { "epoch": 0.552757023564015, "grad_norm": 0.3073681890964508, "learning_rate": 9.972473922853281e-05, "loss": 0.0621, "step": 16350 }, { "epoch": 0.5530951012542682, "grad_norm": 0.2138809710741043, "learning_rate": 9.972387224567045e-05, "loss": 0.062, "step": 16360 }, { "epoch": 0.5534331789445215, "grad_norm": 0.18018893897533417, "learning_rate": 9.97230039033761e-05, "loss": 0.0634, "step": 16370 }, { "epoch": 0.5537712566347747, "grad_norm": 0.2220856249332428, "learning_rate": 9.972213420167349e-05, "loss": 0.0635, "step": 16380 }, { "epoch": 0.5541093343250278, "grad_norm": 0.3247758746147156, "learning_rate": 9.972126314058643e-05, "loss": 0.0634, "step": 16390 }, { "epoch": 0.5544474120152811, "grad_norm": 0.22098079323768616, "learning_rate": 9.97203907201387e-05, "loss": 0.0633, "step": 16400 }, { "epoch": 0.5547854897055343, "grad_norm": 0.22850017249584198, "learning_rate": 9.971951694035417e-05, "loss": 0.0591, "step": 16410 }, { "epoch": 0.5551235673957875, "grad_norm": 0.23259808123111725, "learning_rate": 9.971864180125674e-05, "loss": 0.0637, "step": 16420 }, { "epoch": 0.5554616450860408, "grad_norm": 0.3060200810432434, "learning_rate": 9.97177653028703e-05, "loss": 0.0677, "step": 16430 }, { "epoch": 0.555799722776294, "grad_norm": 0.3969142735004425, "learning_rate": 9.971688744521885e-05, "loss": 0.0676, "step": 16440 }, { "epoch": 0.5561378004665473, "grad_norm": 0.20644597709178925, "learning_rate": 9.971600822832636e-05, "loss": 0.0595, "step": 16450 }, { "epoch": 0.5564758781568004, "grad_norm": 0.2926825284957886, "learning_rate": 9.971512765221691e-05, "loss": 0.0655, "step": 16460 }, { "epoch": 0.5568139558470536, "grad_norm": 0.2566647231578827, "learning_rate": 9.971424571691452e-05, "loss": 0.0607, "step": 16470 }, { "epoch": 0.5571520335373069, "grad_norm": 0.3298570215702057, "learning_rate": 9.971336242244333e-05, "loss": 0.0623, "step": 16480 }, { "epoch": 0.5574901112275601, "grad_norm": 0.20393678545951843, "learning_rate": 9.971247776882751e-05, "loss": 0.0647, "step": 16490 }, { "epoch": 0.5578281889178133, "grad_norm": 0.24818478524684906, "learning_rate": 9.971159175609121e-05, "loss": 0.0672, "step": 16500 }, { "epoch": 0.5581662666080666, "grad_norm": 0.2801789939403534, "learning_rate": 9.971070438425866e-05, "loss": 0.0622, "step": 16510 }, { "epoch": 0.5585043442983197, "grad_norm": 0.26988232135772705, "learning_rate": 9.970981565335415e-05, "loss": 0.0591, "step": 16520 }, { "epoch": 0.5588424219885729, "grad_norm": 0.2651515305042267, "learning_rate": 9.970892556340194e-05, "loss": 0.0594, "step": 16530 }, { "epoch": 0.5591804996788262, "grad_norm": 0.32923945784568787, "learning_rate": 9.970803411442637e-05, "loss": 0.0636, "step": 16540 }, { "epoch": 0.5595185773690794, "grad_norm": 0.16806641221046448, "learning_rate": 9.970714130645184e-05, "loss": 0.0672, "step": 16550 }, { "epoch": 0.5598566550593327, "grad_norm": 0.221955344080925, "learning_rate": 9.970624713950273e-05, "loss": 0.0651, "step": 16560 }, { "epoch": 0.5601947327495859, "grad_norm": 0.24267995357513428, "learning_rate": 9.97053516136035e-05, "loss": 0.0641, "step": 16570 }, { "epoch": 0.560532810439839, "grad_norm": 0.33105048537254333, "learning_rate": 9.970445472877863e-05, "loss": 0.0652, "step": 16580 }, { "epoch": 0.5608708881300923, "grad_norm": 0.31838923692703247, "learning_rate": 9.970355648505264e-05, "loss": 0.0627, "step": 16590 }, { "epoch": 0.5612089658203455, "grad_norm": 0.37404072284698486, "learning_rate": 9.970265688245009e-05, "loss": 0.0637, "step": 16600 }, { "epoch": 0.5615470435105987, "grad_norm": 0.2517707943916321, "learning_rate": 9.970175592099557e-05, "loss": 0.0595, "step": 16610 }, { "epoch": 0.561885121200852, "grad_norm": 0.28752174973487854, "learning_rate": 9.970085360071371e-05, "loss": 0.0615, "step": 16620 }, { "epoch": 0.5622231988911052, "grad_norm": 0.2824572026729584, "learning_rate": 9.969994992162921e-05, "loss": 0.0641, "step": 16630 }, { "epoch": 0.5625612765813583, "grad_norm": 0.30474475026130676, "learning_rate": 9.969904488376673e-05, "loss": 0.0629, "step": 16640 }, { "epoch": 0.5628993542716116, "grad_norm": 0.3423914611339569, "learning_rate": 9.969813848715103e-05, "loss": 0.0617, "step": 16650 }, { "epoch": 0.5632374319618648, "grad_norm": 0.33498644828796387, "learning_rate": 9.96972307318069e-05, "loss": 0.0624, "step": 16660 }, { "epoch": 0.5635755096521181, "grad_norm": 0.2359306514263153, "learning_rate": 9.969632161775916e-05, "loss": 0.0641, "step": 16670 }, { "epoch": 0.5639135873423713, "grad_norm": 0.19844785332679749, "learning_rate": 9.969541114503265e-05, "loss": 0.0615, "step": 16680 }, { "epoch": 0.5642516650326245, "grad_norm": 0.24437715113162994, "learning_rate": 9.969449931365227e-05, "loss": 0.0628, "step": 16690 }, { "epoch": 0.5645897427228778, "grad_norm": 0.35859888792037964, "learning_rate": 9.969358612364294e-05, "loss": 0.0621, "step": 16700 }, { "epoch": 0.5649278204131309, "grad_norm": 0.29759806394577026, "learning_rate": 9.969267157502964e-05, "loss": 0.0658, "step": 16710 }, { "epoch": 0.5652658981033841, "grad_norm": 0.3140183985233307, "learning_rate": 9.969175566783738e-05, "loss": 0.0626, "step": 16720 }, { "epoch": 0.5656039757936374, "grad_norm": 0.25864362716674805, "learning_rate": 9.969083840209116e-05, "loss": 0.0628, "step": 16730 }, { "epoch": 0.5659420534838906, "grad_norm": 0.39731261134147644, "learning_rate": 9.968991977781613e-05, "loss": 0.062, "step": 16740 }, { "epoch": 0.5662801311741438, "grad_norm": 0.21827749907970428, "learning_rate": 9.968899979503731e-05, "loss": 0.073, "step": 16750 }, { "epoch": 0.5666182088643971, "grad_norm": 0.24848692119121552, "learning_rate": 9.968807845377994e-05, "loss": 0.0711, "step": 16760 }, { "epoch": 0.5669562865546502, "grad_norm": 0.21794447302818298, "learning_rate": 9.968715575406916e-05, "loss": 0.0703, "step": 16770 }, { "epoch": 0.5672943642449034, "grad_norm": 0.17949190735816956, "learning_rate": 9.96862316959302e-05, "loss": 0.0589, "step": 16780 }, { "epoch": 0.5676324419351567, "grad_norm": 0.2552039921283722, "learning_rate": 9.968530627938834e-05, "loss": 0.0682, "step": 16790 }, { "epoch": 0.5679705196254099, "grad_norm": 0.16558857262134552, "learning_rate": 9.968437950446887e-05, "loss": 0.0568, "step": 16800 }, { "epoch": 0.5683085973156632, "grad_norm": 0.2840074896812439, "learning_rate": 9.968345137119715e-05, "loss": 0.0587, "step": 16810 }, { "epoch": 0.5686466750059164, "grad_norm": 0.19912366569042206, "learning_rate": 9.96825218795985e-05, "loss": 0.0643, "step": 16820 }, { "epoch": 0.5689847526961695, "grad_norm": 0.25368162989616394, "learning_rate": 9.968159102969839e-05, "loss": 0.0571, "step": 16830 }, { "epoch": 0.5693228303864228, "grad_norm": 0.21075177192687988, "learning_rate": 9.968065882152225e-05, "loss": 0.0641, "step": 16840 }, { "epoch": 0.569660908076676, "grad_norm": 0.16204573214054108, "learning_rate": 9.967972525509554e-05, "loss": 0.0651, "step": 16850 }, { "epoch": 0.5699989857669292, "grad_norm": 0.2904435098171234, "learning_rate": 9.967879033044382e-05, "loss": 0.0672, "step": 16860 }, { "epoch": 0.5703370634571825, "grad_norm": 0.3409374952316284, "learning_rate": 9.967785404759263e-05, "loss": 0.0661, "step": 16870 }, { "epoch": 0.5706751411474357, "grad_norm": 0.22134119272232056, "learning_rate": 9.967691640656757e-05, "loss": 0.0608, "step": 16880 }, { "epoch": 0.5710132188376889, "grad_norm": 0.2190474569797516, "learning_rate": 9.967597740739429e-05, "loss": 0.0608, "step": 16890 }, { "epoch": 0.5713512965279421, "grad_norm": 0.2565614879131317, "learning_rate": 9.967503705009845e-05, "loss": 0.0663, "step": 16900 }, { "epoch": 0.5716893742181953, "grad_norm": 0.2733725607395172, "learning_rate": 9.967409533470575e-05, "loss": 0.0567, "step": 16910 }, { "epoch": 0.5720274519084486, "grad_norm": 0.28861624002456665, "learning_rate": 9.967315226124194e-05, "loss": 0.0634, "step": 16920 }, { "epoch": 0.5723655295987018, "grad_norm": 0.7974069118499756, "learning_rate": 9.967220782973282e-05, "loss": 0.0657, "step": 16930 }, { "epoch": 0.572703607288955, "grad_norm": 0.1934712529182434, "learning_rate": 9.967126204020421e-05, "loss": 0.0596, "step": 16940 }, { "epoch": 0.5730416849792083, "grad_norm": 0.31189247965812683, "learning_rate": 9.967031489268193e-05, "loss": 0.0627, "step": 16950 }, { "epoch": 0.5733797626694614, "grad_norm": 0.29429593682289124, "learning_rate": 9.96693663871919e-05, "loss": 0.0599, "step": 16960 }, { "epoch": 0.5737178403597146, "grad_norm": 0.18510012328624725, "learning_rate": 9.966841652376009e-05, "loss": 0.0612, "step": 16970 }, { "epoch": 0.5740559180499679, "grad_norm": 0.23584116995334625, "learning_rate": 9.96674653024124e-05, "loss": 0.0601, "step": 16980 }, { "epoch": 0.5743939957402211, "grad_norm": 0.3039664626121521, "learning_rate": 9.966651272317487e-05, "loss": 0.0671, "step": 16990 }, { "epoch": 0.5747320734304743, "grad_norm": 0.3146785795688629, "learning_rate": 9.966555878607355e-05, "loss": 0.0623, "step": 17000 }, { "epoch": 0.5750701511207276, "grad_norm": 0.37045443058013916, "learning_rate": 9.96646034911345e-05, "loss": 0.0614, "step": 17010 }, { "epoch": 0.5754082288109807, "grad_norm": 0.2606818675994873, "learning_rate": 9.966364683838388e-05, "loss": 0.0667, "step": 17020 }, { "epoch": 0.5757463065012339, "grad_norm": 0.2768472135066986, "learning_rate": 9.966268882784777e-05, "loss": 0.0676, "step": 17030 }, { "epoch": 0.5760843841914872, "grad_norm": 0.2730010151863098, "learning_rate": 9.966172945955243e-05, "loss": 0.062, "step": 17040 }, { "epoch": 0.5764224618817404, "grad_norm": 0.28773170709609985, "learning_rate": 9.966076873352405e-05, "loss": 0.0691, "step": 17050 }, { "epoch": 0.5767605395719937, "grad_norm": 0.3748909533023834, "learning_rate": 9.965980664978893e-05, "loss": 0.0669, "step": 17060 }, { "epoch": 0.5770986172622469, "grad_norm": 0.34364473819732666, "learning_rate": 9.965884320837334e-05, "loss": 0.0621, "step": 17070 }, { "epoch": 0.5774366949525, "grad_norm": 0.36599573493003845, "learning_rate": 9.965787840930363e-05, "loss": 0.0672, "step": 17080 }, { "epoch": 0.5777747726427533, "grad_norm": 0.1967371106147766, "learning_rate": 9.965691225260619e-05, "loss": 0.0658, "step": 17090 }, { "epoch": 0.5781128503330065, "grad_norm": 0.26337090134620667, "learning_rate": 9.965594473830742e-05, "loss": 0.0637, "step": 17100 }, { "epoch": 0.5784509280232597, "grad_norm": 0.23741009831428528, "learning_rate": 9.965497586643377e-05, "loss": 0.07, "step": 17110 }, { "epoch": 0.578789005713513, "grad_norm": 0.2750290632247925, "learning_rate": 9.965400563701172e-05, "loss": 0.0617, "step": 17120 }, { "epoch": 0.5791270834037662, "grad_norm": 0.2760494351387024, "learning_rate": 9.965303405006784e-05, "loss": 0.0687, "step": 17130 }, { "epoch": 0.5794651610940194, "grad_norm": 0.2075916975736618, "learning_rate": 9.965206110562864e-05, "loss": 0.0612, "step": 17140 }, { "epoch": 0.5798032387842726, "grad_norm": 0.3773403465747833, "learning_rate": 9.965108680372076e-05, "loss": 0.0655, "step": 17150 }, { "epoch": 0.5801413164745258, "grad_norm": 0.28339484333992004, "learning_rate": 9.965011114437082e-05, "loss": 0.0632, "step": 17160 }, { "epoch": 0.5804793941647791, "grad_norm": 0.1784794181585312, "learning_rate": 9.964913412760549e-05, "loss": 0.0645, "step": 17170 }, { "epoch": 0.5808174718550323, "grad_norm": 0.28053438663482666, "learning_rate": 9.964815575345148e-05, "loss": 0.0546, "step": 17180 }, { "epoch": 0.5811555495452855, "grad_norm": 0.1904456615447998, "learning_rate": 9.964717602193554e-05, "loss": 0.0633, "step": 17190 }, { "epoch": 0.5814936272355388, "grad_norm": 0.23658674955368042, "learning_rate": 9.964619493308447e-05, "loss": 0.0624, "step": 17200 }, { "epoch": 0.581831704925792, "grad_norm": 0.32315099239349365, "learning_rate": 9.964521248692507e-05, "loss": 0.0671, "step": 17210 }, { "epoch": 0.5821697826160451, "grad_norm": 0.26552391052246094, "learning_rate": 9.964422868348421e-05, "loss": 0.0641, "step": 17220 }, { "epoch": 0.5825078603062984, "grad_norm": 0.20717692375183105, "learning_rate": 9.96432435227888e-05, "loss": 0.0591, "step": 17230 }, { "epoch": 0.5828459379965516, "grad_norm": 0.21346336603164673, "learning_rate": 9.964225700486576e-05, "loss": 0.058, "step": 17240 }, { "epoch": 0.5831840156868048, "grad_norm": 0.2737636864185333, "learning_rate": 9.964126912974207e-05, "loss": 0.0639, "step": 17250 }, { "epoch": 0.5835220933770581, "grad_norm": 0.2579597532749176, "learning_rate": 9.964027989744471e-05, "loss": 0.0596, "step": 17260 }, { "epoch": 0.5838601710673113, "grad_norm": 0.19462497532367706, "learning_rate": 9.963928930800077e-05, "loss": 0.0589, "step": 17270 }, { "epoch": 0.5841982487575645, "grad_norm": 0.24219685792922974, "learning_rate": 9.96382973614373e-05, "loss": 0.0535, "step": 17280 }, { "epoch": 0.5845363264478177, "grad_norm": 0.30969786643981934, "learning_rate": 9.963730405778142e-05, "loss": 0.0647, "step": 17290 }, { "epoch": 0.5848744041380709, "grad_norm": 0.22682306170463562, "learning_rate": 9.96363093970603e-05, "loss": 0.0633, "step": 17300 }, { "epoch": 0.5852124818283242, "grad_norm": 0.23142525553703308, "learning_rate": 9.963531337930114e-05, "loss": 0.0641, "step": 17310 }, { "epoch": 0.5855505595185774, "grad_norm": 0.22681774199008942, "learning_rate": 9.963431600453115e-05, "loss": 0.0625, "step": 17320 }, { "epoch": 0.5858886372088306, "grad_norm": 0.31341108679771423, "learning_rate": 9.96333172727776e-05, "loss": 0.0612, "step": 17330 }, { "epoch": 0.5862267148990838, "grad_norm": 0.2536013424396515, "learning_rate": 9.96323171840678e-05, "loss": 0.0583, "step": 17340 }, { "epoch": 0.586564792589337, "grad_norm": 0.16263510286808014, "learning_rate": 9.963131573842912e-05, "loss": 0.063, "step": 17350 }, { "epoch": 0.5869028702795902, "grad_norm": 0.37114766240119934, "learning_rate": 9.963031293588889e-05, "loss": 0.07, "step": 17360 }, { "epoch": 0.5872409479698435, "grad_norm": 0.33940041065216064, "learning_rate": 9.962930877647457e-05, "loss": 0.0666, "step": 17370 }, { "epoch": 0.5875790256600967, "grad_norm": 0.2629479467868805, "learning_rate": 9.962830326021357e-05, "loss": 0.0613, "step": 17380 }, { "epoch": 0.5879171033503499, "grad_norm": 0.2756689488887787, "learning_rate": 9.962729638713341e-05, "loss": 0.0594, "step": 17390 }, { "epoch": 0.5882551810406031, "grad_norm": 0.2733009159564972, "learning_rate": 9.96262881572616e-05, "loss": 0.0668, "step": 17400 }, { "epoch": 0.5885932587308563, "grad_norm": 0.22247660160064697, "learning_rate": 9.962527857062574e-05, "loss": 0.066, "step": 17410 }, { "epoch": 0.5889313364211096, "grad_norm": 0.25067007541656494, "learning_rate": 9.96242676272534e-05, "loss": 0.0665, "step": 17420 }, { "epoch": 0.5892694141113628, "grad_norm": 0.3185804784297943, "learning_rate": 9.962325532717222e-05, "loss": 0.072, "step": 17430 }, { "epoch": 0.589607491801616, "grad_norm": 0.3640611469745636, "learning_rate": 9.962224167040988e-05, "loss": 0.0668, "step": 17440 }, { "epoch": 0.5899455694918693, "grad_norm": 0.3530450165271759, "learning_rate": 9.96212266569941e-05, "loss": 0.0677, "step": 17450 }, { "epoch": 0.5902836471821225, "grad_norm": 0.20204751193523407, "learning_rate": 9.962021028695263e-05, "loss": 0.0648, "step": 17460 }, { "epoch": 0.5906217248723756, "grad_norm": 0.23764455318450928, "learning_rate": 9.961919256031324e-05, "loss": 0.0642, "step": 17470 }, { "epoch": 0.5909598025626289, "grad_norm": 0.21733315289020538, "learning_rate": 9.961817347710378e-05, "loss": 0.0652, "step": 17480 }, { "epoch": 0.5912978802528821, "grad_norm": 0.38809332251548767, "learning_rate": 9.961715303735208e-05, "loss": 0.0698, "step": 17490 }, { "epoch": 0.5916359579431353, "grad_norm": 0.3492497503757477, "learning_rate": 9.961613124108606e-05, "loss": 0.0622, "step": 17500 }, { "epoch": 0.5919740356333886, "grad_norm": 0.2521151602268219, "learning_rate": 9.961510808833368e-05, "loss": 0.0568, "step": 17510 }, { "epoch": 0.5923121133236418, "grad_norm": 0.2284328192472458, "learning_rate": 9.961408357912286e-05, "loss": 0.0676, "step": 17520 }, { "epoch": 0.592650191013895, "grad_norm": 0.29595521092414856, "learning_rate": 9.961305771348163e-05, "loss": 0.0613, "step": 17530 }, { "epoch": 0.5929882687041482, "grad_norm": 0.2608036398887634, "learning_rate": 9.961203049143805e-05, "loss": 0.0573, "step": 17540 }, { "epoch": 0.5933263463944014, "grad_norm": 0.27510523796081543, "learning_rate": 9.961100191302019e-05, "loss": 0.0665, "step": 17550 }, { "epoch": 0.5936644240846547, "grad_norm": 0.3180468678474426, "learning_rate": 9.960997197825618e-05, "loss": 0.0673, "step": 17560 }, { "epoch": 0.5940025017749079, "grad_norm": 0.19541962444782257, "learning_rate": 9.960894068717416e-05, "loss": 0.0686, "step": 17570 }, { "epoch": 0.5943405794651611, "grad_norm": 0.25391608476638794, "learning_rate": 9.960790803980235e-05, "loss": 0.0642, "step": 17580 }, { "epoch": 0.5946786571554143, "grad_norm": 0.3118875324726105, "learning_rate": 9.960687403616898e-05, "loss": 0.0642, "step": 17590 }, { "epoch": 0.5950167348456675, "grad_norm": 0.2530166506767273, "learning_rate": 9.960583867630231e-05, "loss": 0.0629, "step": 17600 }, { "epoch": 0.5953548125359207, "grad_norm": 0.22204025089740753, "learning_rate": 9.960480196023064e-05, "loss": 0.0568, "step": 17610 }, { "epoch": 0.595692890226174, "grad_norm": 0.3467594087123871, "learning_rate": 9.960376388798232e-05, "loss": 0.0644, "step": 17620 }, { "epoch": 0.5960309679164272, "grad_norm": 0.2509109675884247, "learning_rate": 9.960272445958572e-05, "loss": 0.0668, "step": 17630 }, { "epoch": 0.5963690456066805, "grad_norm": 0.24735622107982635, "learning_rate": 9.960168367506927e-05, "loss": 0.0655, "step": 17640 }, { "epoch": 0.5967071232969337, "grad_norm": 0.25806984305381775, "learning_rate": 9.960064153446145e-05, "loss": 0.0659, "step": 17650 }, { "epoch": 0.5970452009871868, "grad_norm": 0.29768431186676025, "learning_rate": 9.95995980377907e-05, "loss": 0.0637, "step": 17660 }, { "epoch": 0.5973832786774401, "grad_norm": 0.2802470028400421, "learning_rate": 9.959855318508558e-05, "loss": 0.0662, "step": 17670 }, { "epoch": 0.5977213563676933, "grad_norm": 0.32521697878837585, "learning_rate": 9.959750697637465e-05, "loss": 0.0575, "step": 17680 }, { "epoch": 0.5980594340579465, "grad_norm": 0.31460142135620117, "learning_rate": 9.95964594116865e-05, "loss": 0.067, "step": 17690 }, { "epoch": 0.5983975117481998, "grad_norm": 0.25524571537971497, "learning_rate": 9.959541049104979e-05, "loss": 0.0606, "step": 17700 }, { "epoch": 0.598735589438453, "grad_norm": 0.23194649815559387, "learning_rate": 9.95943602144932e-05, "loss": 0.0552, "step": 17710 }, { "epoch": 0.5990736671287061, "grad_norm": 0.3499494194984436, "learning_rate": 9.959330858204542e-05, "loss": 0.0633, "step": 17720 }, { "epoch": 0.5994117448189594, "grad_norm": 0.2532595694065094, "learning_rate": 9.959225559373521e-05, "loss": 0.0578, "step": 17730 }, { "epoch": 0.5997498225092126, "grad_norm": 0.40760332345962524, "learning_rate": 9.959120124959138e-05, "loss": 0.0593, "step": 17740 }, { "epoch": 0.6000879001994658, "grad_norm": 0.44537854194641113, "learning_rate": 9.959014554964273e-05, "loss": 0.0566, "step": 17750 }, { "epoch": 0.6004259778897191, "grad_norm": 0.30924683809280396, "learning_rate": 9.958908849391812e-05, "loss": 0.0665, "step": 17760 }, { "epoch": 0.6007640555799723, "grad_norm": 0.2902486026287079, "learning_rate": 9.958803008244649e-05, "loss": 0.0616, "step": 17770 }, { "epoch": 0.6011021332702255, "grad_norm": 0.2826058864593506, "learning_rate": 9.958697031525672e-05, "loss": 0.0608, "step": 17780 }, { "epoch": 0.6014402109604787, "grad_norm": 0.3465513288974762, "learning_rate": 9.958590919237783e-05, "loss": 0.0561, "step": 17790 }, { "epoch": 0.6017782886507319, "grad_norm": 0.3481718897819519, "learning_rate": 9.958484671383879e-05, "loss": 0.0635, "step": 17800 }, { "epoch": 0.6021163663409852, "grad_norm": 0.308514803647995, "learning_rate": 9.958378287966868e-05, "loss": 0.0582, "step": 17810 }, { "epoch": 0.6024544440312384, "grad_norm": 0.2645130455493927, "learning_rate": 9.958271768989658e-05, "loss": 0.0611, "step": 17820 }, { "epoch": 0.6027925217214916, "grad_norm": 0.3099738359451294, "learning_rate": 9.95816511445516e-05, "loss": 0.0591, "step": 17830 }, { "epoch": 0.6031305994117448, "grad_norm": 0.37272268533706665, "learning_rate": 9.958058324366291e-05, "loss": 0.0625, "step": 17840 }, { "epoch": 0.603468677101998, "grad_norm": 0.28042295575141907, "learning_rate": 9.957951398725969e-05, "loss": 0.0607, "step": 17850 }, { "epoch": 0.6038067547922512, "grad_norm": 0.19097204506397247, "learning_rate": 9.95784433753712e-05, "loss": 0.065, "step": 17860 }, { "epoch": 0.6041448324825045, "grad_norm": 0.17812910676002502, "learning_rate": 9.957737140802666e-05, "loss": 0.0633, "step": 17870 }, { "epoch": 0.6044829101727577, "grad_norm": 0.49751919507980347, "learning_rate": 9.957629808525544e-05, "loss": 0.0689, "step": 17880 }, { "epoch": 0.604820987863011, "grad_norm": 0.3866598308086395, "learning_rate": 9.957522340708685e-05, "loss": 0.0602, "step": 17890 }, { "epoch": 0.6051590655532642, "grad_norm": 0.17458631098270416, "learning_rate": 9.957414737355027e-05, "loss": 0.061, "step": 17900 }, { "epoch": 0.6054971432435173, "grad_norm": 0.24413469433784485, "learning_rate": 9.957306998467512e-05, "loss": 0.063, "step": 17910 }, { "epoch": 0.6058352209337706, "grad_norm": 0.3762401044368744, "learning_rate": 9.957199124049088e-05, "loss": 0.0657, "step": 17920 }, { "epoch": 0.6061732986240238, "grad_norm": 0.25732845067977905, "learning_rate": 9.957091114102702e-05, "loss": 0.0558, "step": 17930 }, { "epoch": 0.606511376314277, "grad_norm": 0.22488245368003845, "learning_rate": 9.956982968631306e-05, "loss": 0.0673, "step": 17940 }, { "epoch": 0.6068494540045303, "grad_norm": 0.3888118863105774, "learning_rate": 9.956874687637858e-05, "loss": 0.0633, "step": 17950 }, { "epoch": 0.6071875316947835, "grad_norm": 0.3257654905319214, "learning_rate": 9.95676627112532e-05, "loss": 0.0588, "step": 17960 }, { "epoch": 0.6075256093850366, "grad_norm": 0.2963639199733734, "learning_rate": 9.956657719096652e-05, "loss": 0.0607, "step": 17970 }, { "epoch": 0.6078636870752899, "grad_norm": 0.29223138093948364, "learning_rate": 9.956549031554825e-05, "loss": 0.0629, "step": 17980 }, { "epoch": 0.6082017647655431, "grad_norm": 0.2305477112531662, "learning_rate": 9.956440208502811e-05, "loss": 0.0574, "step": 17990 }, { "epoch": 0.6085398424557963, "grad_norm": 0.2870107889175415, "learning_rate": 9.956331249943583e-05, "loss": 0.0629, "step": 18000 }, { "epoch": 0.6088779201460496, "grad_norm": 0.4071638882160187, "learning_rate": 9.956222155880118e-05, "loss": 0.0643, "step": 18010 }, { "epoch": 0.6092159978363028, "grad_norm": 0.2894194722175598, "learning_rate": 9.956112926315405e-05, "loss": 0.0571, "step": 18020 }, { "epoch": 0.609554075526556, "grad_norm": 0.28837743401527405, "learning_rate": 9.956003561252424e-05, "loss": 0.0629, "step": 18030 }, { "epoch": 0.6098921532168092, "grad_norm": 0.24922451376914978, "learning_rate": 9.955894060694168e-05, "loss": 0.0592, "step": 18040 }, { "epoch": 0.6102302309070624, "grad_norm": 0.1942148655653, "learning_rate": 9.95578442464363e-05, "loss": 0.0627, "step": 18050 }, { "epoch": 0.6105683085973157, "grad_norm": 0.16205009818077087, "learning_rate": 9.955674653103808e-05, "loss": 0.0586, "step": 18060 }, { "epoch": 0.6109063862875689, "grad_norm": 0.3008301556110382, "learning_rate": 9.955564746077704e-05, "loss": 0.0595, "step": 18070 }, { "epoch": 0.6112444639778221, "grad_norm": 0.24136605858802795, "learning_rate": 9.955454703568319e-05, "loss": 0.0594, "step": 18080 }, { "epoch": 0.6115825416680754, "grad_norm": 0.35648760199546814, "learning_rate": 9.955344525578667e-05, "loss": 0.0647, "step": 18090 }, { "epoch": 0.6119206193583285, "grad_norm": 0.17383505403995514, "learning_rate": 9.955234212111755e-05, "loss": 0.0604, "step": 18100 }, { "epoch": 0.6122586970485817, "grad_norm": 0.21871116757392883, "learning_rate": 9.955123763170602e-05, "loss": 0.0635, "step": 18110 }, { "epoch": 0.612596774738835, "grad_norm": 0.36733976006507874, "learning_rate": 9.955013178758228e-05, "loss": 0.0612, "step": 18120 }, { "epoch": 0.6129348524290882, "grad_norm": 0.2550097107887268, "learning_rate": 9.954902458877654e-05, "loss": 0.0653, "step": 18130 }, { "epoch": 0.6132729301193415, "grad_norm": 0.2723497152328491, "learning_rate": 9.954791603531908e-05, "loss": 0.0648, "step": 18140 }, { "epoch": 0.6136110078095947, "grad_norm": 0.32904455065727234, "learning_rate": 9.954680612724021e-05, "loss": 0.0618, "step": 18150 }, { "epoch": 0.6139490854998478, "grad_norm": 0.22394700348377228, "learning_rate": 9.954569486457028e-05, "loss": 0.0567, "step": 18160 }, { "epoch": 0.6142871631901011, "grad_norm": 0.22875069081783295, "learning_rate": 9.954458224733966e-05, "loss": 0.0663, "step": 18170 }, { "epoch": 0.6146252408803543, "grad_norm": 0.20241527259349823, "learning_rate": 9.954346827557879e-05, "loss": 0.0632, "step": 18180 }, { "epoch": 0.6149633185706075, "grad_norm": 0.2676035463809967, "learning_rate": 9.954235294931809e-05, "loss": 0.06, "step": 18190 }, { "epoch": 0.6153013962608608, "grad_norm": 0.3108541965484619, "learning_rate": 9.954123626858808e-05, "loss": 0.0649, "step": 18200 }, { "epoch": 0.615639473951114, "grad_norm": 0.34441348910331726, "learning_rate": 9.954011823341928e-05, "loss": 0.0614, "step": 18210 }, { "epoch": 0.6159775516413671, "grad_norm": 0.3677418828010559, "learning_rate": 9.953899884384228e-05, "loss": 0.0631, "step": 18220 }, { "epoch": 0.6163156293316204, "grad_norm": 0.2270917445421219, "learning_rate": 9.953787809988764e-05, "loss": 0.0663, "step": 18230 }, { "epoch": 0.6166537070218736, "grad_norm": 0.2696094810962677, "learning_rate": 9.953675600158603e-05, "loss": 0.0667, "step": 18240 }, { "epoch": 0.6169917847121269, "grad_norm": 0.25197696685791016, "learning_rate": 9.953563254896812e-05, "loss": 0.057, "step": 18250 }, { "epoch": 0.6173298624023801, "grad_norm": 0.22683896124362946, "learning_rate": 9.953450774206462e-05, "loss": 0.0582, "step": 18260 }, { "epoch": 0.6176679400926333, "grad_norm": 0.2747174799442291, "learning_rate": 9.953338158090629e-05, "loss": 0.062, "step": 18270 }, { "epoch": 0.6180060177828866, "grad_norm": 0.28524264693260193, "learning_rate": 9.953225406552392e-05, "loss": 0.0651, "step": 18280 }, { "epoch": 0.6183440954731397, "grad_norm": 0.21700191497802734, "learning_rate": 9.953112519594834e-05, "loss": 0.0581, "step": 18290 }, { "epoch": 0.6186821731633929, "grad_norm": 0.2412644624710083, "learning_rate": 9.952999497221038e-05, "loss": 0.0693, "step": 18300 }, { "epoch": 0.6190202508536462, "grad_norm": 0.23885971307754517, "learning_rate": 9.9528863394341e-05, "loss": 0.061, "step": 18310 }, { "epoch": 0.6193583285438994, "grad_norm": 0.24295789003372192, "learning_rate": 9.952773046237108e-05, "loss": 0.0601, "step": 18320 }, { "epoch": 0.6196964062341526, "grad_norm": 0.3625904321670532, "learning_rate": 9.95265961763316e-05, "loss": 0.0636, "step": 18330 }, { "epoch": 0.6200344839244059, "grad_norm": 0.28263214230537415, "learning_rate": 9.952546053625362e-05, "loss": 0.0638, "step": 18340 }, { "epoch": 0.620372561614659, "grad_norm": 0.18825964629650116, "learning_rate": 9.952432354216815e-05, "loss": 0.0601, "step": 18350 }, { "epoch": 0.6207106393049122, "grad_norm": 0.3493225574493408, "learning_rate": 9.952318519410625e-05, "loss": 0.0578, "step": 18360 }, { "epoch": 0.6210487169951655, "grad_norm": 0.24821600317955017, "learning_rate": 9.952204549209909e-05, "loss": 0.0627, "step": 18370 }, { "epoch": 0.6213867946854187, "grad_norm": 0.22082674503326416, "learning_rate": 9.952090443617783e-05, "loss": 0.0653, "step": 18380 }, { "epoch": 0.621724872375672, "grad_norm": 0.38095325231552124, "learning_rate": 9.951976202637364e-05, "loss": 0.0629, "step": 18390 }, { "epoch": 0.6220629500659252, "grad_norm": 0.3609021008014679, "learning_rate": 9.951861826271773e-05, "loss": 0.054, "step": 18400 }, { "epoch": 0.6224010277561783, "grad_norm": 0.27912628650665283, "learning_rate": 9.951747314524143e-05, "loss": 0.0699, "step": 18410 }, { "epoch": 0.6227391054464316, "grad_norm": 0.3297361731529236, "learning_rate": 9.9516326673976e-05, "loss": 0.0605, "step": 18420 }, { "epoch": 0.6230771831366848, "grad_norm": 0.2160894125699997, "learning_rate": 9.95151788489528e-05, "loss": 0.0653, "step": 18430 }, { "epoch": 0.623415260826938, "grad_norm": 0.1894078552722931, "learning_rate": 9.951402967020322e-05, "loss": 0.0588, "step": 18440 }, { "epoch": 0.6237533385171913, "grad_norm": 0.21620984375476837, "learning_rate": 9.951287913775869e-05, "loss": 0.0627, "step": 18450 }, { "epoch": 0.6240914162074445, "grad_norm": 0.2398182898759842, "learning_rate": 9.951172725165061e-05, "loss": 0.0627, "step": 18460 }, { "epoch": 0.6244294938976976, "grad_norm": 0.23830173909664154, "learning_rate": 9.951057401191053e-05, "loss": 0.0623, "step": 18470 }, { "epoch": 0.6247675715879509, "grad_norm": 0.3026750087738037, "learning_rate": 9.950941941856996e-05, "loss": 0.059, "step": 18480 }, { "epoch": 0.6251056492782041, "grad_norm": 0.3057614862918854, "learning_rate": 9.950826347166045e-05, "loss": 0.0592, "step": 18490 }, { "epoch": 0.6254437269684574, "grad_norm": 0.18591976165771484, "learning_rate": 9.950710617121362e-05, "loss": 0.0566, "step": 18500 }, { "epoch": 0.6257818046587106, "grad_norm": 0.2761962413787842, "learning_rate": 9.95059475172611e-05, "loss": 0.0613, "step": 18510 }, { "epoch": 0.6261198823489638, "grad_norm": 0.29051345586776733, "learning_rate": 9.950478750983458e-05, "loss": 0.0664, "step": 18520 }, { "epoch": 0.626457960039217, "grad_norm": 0.21943338215351105, "learning_rate": 9.950362614896575e-05, "loss": 0.0608, "step": 18530 }, { "epoch": 0.6267960377294702, "grad_norm": 0.2161584198474884, "learning_rate": 9.950246343468642e-05, "loss": 0.0666, "step": 18540 }, { "epoch": 0.6271341154197234, "grad_norm": 0.18554037809371948, "learning_rate": 9.95012993670283e-05, "loss": 0.0632, "step": 18550 }, { "epoch": 0.6274721931099767, "grad_norm": 0.3359203040599823, "learning_rate": 9.950013394602327e-05, "loss": 0.0649, "step": 18560 }, { "epoch": 0.6278102708002299, "grad_norm": 0.20169074833393097, "learning_rate": 9.949896717170316e-05, "loss": 0.0705, "step": 18570 }, { "epoch": 0.6281483484904831, "grad_norm": 0.18292908370494843, "learning_rate": 9.949779904409989e-05, "loss": 0.058, "step": 18580 }, { "epoch": 0.6284864261807364, "grad_norm": 0.4914010167121887, "learning_rate": 9.949662956324539e-05, "loss": 0.063, "step": 18590 }, { "epoch": 0.6288245038709895, "grad_norm": 0.23361878097057343, "learning_rate": 9.949545872917164e-05, "loss": 0.0621, "step": 18600 }, { "epoch": 0.6291625815612428, "grad_norm": 0.22943824529647827, "learning_rate": 9.949428654191065e-05, "loss": 0.0618, "step": 18610 }, { "epoch": 0.629500659251496, "grad_norm": 0.33680659532546997, "learning_rate": 9.949311300149444e-05, "loss": 0.0707, "step": 18620 }, { "epoch": 0.6298387369417492, "grad_norm": 0.3618358373641968, "learning_rate": 9.949193810795511e-05, "loss": 0.0667, "step": 18630 }, { "epoch": 0.6301768146320025, "grad_norm": 0.2545361816883087, "learning_rate": 9.94907618613248e-05, "loss": 0.0657, "step": 18640 }, { "epoch": 0.6305148923222557, "grad_norm": 0.3319374620914459, "learning_rate": 9.948958426163566e-05, "loss": 0.0653, "step": 18650 }, { "epoch": 0.6308529700125088, "grad_norm": 0.28979483246803284, "learning_rate": 9.948840530891987e-05, "loss": 0.0625, "step": 18660 }, { "epoch": 0.6311910477027621, "grad_norm": 0.27547675371170044, "learning_rate": 9.948722500320966e-05, "loss": 0.0601, "step": 18670 }, { "epoch": 0.6315291253930153, "grad_norm": 0.23167820274829865, "learning_rate": 9.948604334453733e-05, "loss": 0.0637, "step": 18680 }, { "epoch": 0.6318672030832685, "grad_norm": 0.312624454498291, "learning_rate": 9.948486033293515e-05, "loss": 0.0598, "step": 18690 }, { "epoch": 0.6322052807735218, "grad_norm": 0.20749036967754364, "learning_rate": 9.948367596843549e-05, "loss": 0.06, "step": 18700 }, { "epoch": 0.632543358463775, "grad_norm": 0.3396869897842407, "learning_rate": 9.94824902510707e-05, "loss": 0.0593, "step": 18710 }, { "epoch": 0.6328814361540281, "grad_norm": 0.17498648166656494, "learning_rate": 9.948130318087325e-05, "loss": 0.0621, "step": 18720 }, { "epoch": 0.6332195138442814, "grad_norm": 0.26100704073905945, "learning_rate": 9.948011475787553e-05, "loss": 0.0589, "step": 18730 }, { "epoch": 0.6335575915345346, "grad_norm": 0.30935385823249817, "learning_rate": 9.947892498211007e-05, "loss": 0.066, "step": 18740 }, { "epoch": 0.6338956692247879, "grad_norm": 0.2096358686685562, "learning_rate": 9.947773385360939e-05, "loss": 0.0562, "step": 18750 }, { "epoch": 0.6342337469150411, "grad_norm": 0.207455575466156, "learning_rate": 9.947654137240606e-05, "loss": 0.0562, "step": 18760 }, { "epoch": 0.6345718246052943, "grad_norm": 0.19987282156944275, "learning_rate": 9.947534753853267e-05, "loss": 0.0587, "step": 18770 }, { "epoch": 0.6349099022955476, "grad_norm": 0.4514789581298828, "learning_rate": 9.947415235202187e-05, "loss": 0.0665, "step": 18780 }, { "epoch": 0.6352479799858007, "grad_norm": 0.40856850147247314, "learning_rate": 9.947295581290632e-05, "loss": 0.0602, "step": 18790 }, { "epoch": 0.6355860576760539, "grad_norm": 0.26729142665863037, "learning_rate": 9.947175792121877e-05, "loss": 0.0528, "step": 18800 }, { "epoch": 0.6359241353663072, "grad_norm": 0.37495866417884827, "learning_rate": 9.947055867699192e-05, "loss": 0.065, "step": 18810 }, { "epoch": 0.6362622130565604, "grad_norm": 0.328453004360199, "learning_rate": 9.946935808025859e-05, "loss": 0.0589, "step": 18820 }, { "epoch": 0.6366002907468136, "grad_norm": 0.2478933036327362, "learning_rate": 9.946815613105159e-05, "loss": 0.0685, "step": 18830 }, { "epoch": 0.6369383684370669, "grad_norm": 0.20722626149654388, "learning_rate": 9.94669528294038e-05, "loss": 0.0666, "step": 18840 }, { "epoch": 0.63727644612732, "grad_norm": 0.2709486484527588, "learning_rate": 9.946574817534809e-05, "loss": 0.0682, "step": 18850 }, { "epoch": 0.6376145238175733, "grad_norm": 0.22283503413200378, "learning_rate": 9.946454216891741e-05, "loss": 0.064, "step": 18860 }, { "epoch": 0.6379526015078265, "grad_norm": 0.25431597232818604, "learning_rate": 9.946333481014472e-05, "loss": 0.0579, "step": 18870 }, { "epoch": 0.6382906791980797, "grad_norm": 0.2845730185508728, "learning_rate": 9.946212609906305e-05, "loss": 0.0629, "step": 18880 }, { "epoch": 0.638628756888333, "grad_norm": 0.3851380944252014, "learning_rate": 9.946091603570545e-05, "loss": 0.0639, "step": 18890 }, { "epoch": 0.6389668345785862, "grad_norm": 0.20638208091259003, "learning_rate": 9.945970462010495e-05, "loss": 0.0613, "step": 18900 }, { "epoch": 0.6393049122688393, "grad_norm": 0.22709542512893677, "learning_rate": 9.945849185229473e-05, "loss": 0.0594, "step": 18910 }, { "epoch": 0.6396429899590926, "grad_norm": 0.23569630086421967, "learning_rate": 9.945727773230793e-05, "loss": 0.0666, "step": 18920 }, { "epoch": 0.6399810676493458, "grad_norm": 0.2047184258699417, "learning_rate": 9.945606226017772e-05, "loss": 0.0578, "step": 18930 }, { "epoch": 0.640319145339599, "grad_norm": 0.22650228440761566, "learning_rate": 9.945484543593736e-05, "loss": 0.067, "step": 18940 }, { "epoch": 0.6406572230298523, "grad_norm": 0.32825008034706116, "learning_rate": 9.945362725962011e-05, "loss": 0.0652, "step": 18950 }, { "epoch": 0.6409953007201055, "grad_norm": 0.17683112621307373, "learning_rate": 9.945240773125926e-05, "loss": 0.0624, "step": 18960 }, { "epoch": 0.6413333784103586, "grad_norm": 0.19800102710723877, "learning_rate": 9.945118685088815e-05, "loss": 0.0618, "step": 18970 }, { "epoch": 0.6416714561006119, "grad_norm": 0.2947453260421753, "learning_rate": 9.944996461854019e-05, "loss": 0.0603, "step": 18980 }, { "epoch": 0.6420095337908651, "grad_norm": 0.20457887649536133, "learning_rate": 9.944874103424876e-05, "loss": 0.062, "step": 18990 }, { "epoch": 0.6423476114811184, "grad_norm": 0.19489933550357819, "learning_rate": 9.944751609804733e-05, "loss": 0.0548, "step": 19000 }, { "epoch": 0.6426856891713716, "grad_norm": 0.376570463180542, "learning_rate": 9.94462898099694e-05, "loss": 0.0627, "step": 19010 }, { "epoch": 0.6430237668616248, "grad_norm": 0.21792110800743103, "learning_rate": 9.944506217004847e-05, "loss": 0.0568, "step": 19020 }, { "epoch": 0.6433618445518781, "grad_norm": 0.22475697100162506, "learning_rate": 9.944383317831812e-05, "loss": 0.0658, "step": 19030 }, { "epoch": 0.6436999222421312, "grad_norm": 0.2544727921485901, "learning_rate": 9.944260283481194e-05, "loss": 0.0596, "step": 19040 }, { "epoch": 0.6440379999323844, "grad_norm": 0.3921695053577423, "learning_rate": 9.944137113956358e-05, "loss": 0.0616, "step": 19050 }, { "epoch": 0.6443760776226377, "grad_norm": 0.28408321738243103, "learning_rate": 9.944013809260671e-05, "loss": 0.0717, "step": 19060 }, { "epoch": 0.6447141553128909, "grad_norm": 0.2842633128166199, "learning_rate": 9.943890369397502e-05, "loss": 0.065, "step": 19070 }, { "epoch": 0.6450522330031441, "grad_norm": 0.24898622930049896, "learning_rate": 9.943766794370228e-05, "loss": 0.0646, "step": 19080 }, { "epoch": 0.6453903106933974, "grad_norm": 0.2449505478143692, "learning_rate": 9.943643084182229e-05, "loss": 0.0653, "step": 19090 }, { "epoch": 0.6457283883836505, "grad_norm": 0.2962802052497864, "learning_rate": 9.943519238836885e-05, "loss": 0.0697, "step": 19100 }, { "epoch": 0.6460664660739038, "grad_norm": 0.24325785040855408, "learning_rate": 9.943395258337581e-05, "loss": 0.0632, "step": 19110 }, { "epoch": 0.646404543764157, "grad_norm": 0.3104466199874878, "learning_rate": 9.943271142687709e-05, "loss": 0.0651, "step": 19120 }, { "epoch": 0.6467426214544102, "grad_norm": 0.36370304226875305, "learning_rate": 9.94314689189066e-05, "loss": 0.0643, "step": 19130 }, { "epoch": 0.6470806991446635, "grad_norm": 0.34924206137657166, "learning_rate": 9.943022505949833e-05, "loss": 0.0661, "step": 19140 }, { "epoch": 0.6474187768349167, "grad_norm": 0.2044522613286972, "learning_rate": 9.942897984868626e-05, "loss": 0.0606, "step": 19150 }, { "epoch": 0.6477568545251698, "grad_norm": 0.3080368638038635, "learning_rate": 9.942773328650447e-05, "loss": 0.0596, "step": 19160 }, { "epoch": 0.6480949322154231, "grad_norm": 0.2678172290325165, "learning_rate": 9.942648537298701e-05, "loss": 0.0607, "step": 19170 }, { "epoch": 0.6484330099056763, "grad_norm": 0.2737983167171478, "learning_rate": 9.9425236108168e-05, "loss": 0.063, "step": 19180 }, { "epoch": 0.6487710875959295, "grad_norm": 0.23264767229557037, "learning_rate": 9.942398549208163e-05, "loss": 0.067, "step": 19190 }, { "epoch": 0.6491091652861828, "grad_norm": 0.22966712713241577, "learning_rate": 9.942273352476205e-05, "loss": 0.0582, "step": 19200 }, { "epoch": 0.649447242976436, "grad_norm": 0.2382517009973526, "learning_rate": 9.94214802062435e-05, "loss": 0.0629, "step": 19210 }, { "epoch": 0.6497853206666893, "grad_norm": 0.2886159121990204, "learning_rate": 9.942022553656025e-05, "loss": 0.0659, "step": 19220 }, { "epoch": 0.6501233983569424, "grad_norm": 0.29071107506752014, "learning_rate": 9.941896951574659e-05, "loss": 0.0648, "step": 19230 }, { "epoch": 0.6504614760471956, "grad_norm": 0.24014437198638916, "learning_rate": 9.941771214383689e-05, "loss": 0.0573, "step": 19240 }, { "epoch": 0.6507995537374489, "grad_norm": 0.3802126348018646, "learning_rate": 9.94164534208655e-05, "loss": 0.0626, "step": 19250 }, { "epoch": 0.6511376314277021, "grad_norm": 0.2885184586048126, "learning_rate": 9.941519334686682e-05, "loss": 0.0635, "step": 19260 }, { "epoch": 0.6514757091179553, "grad_norm": 0.34770917892456055, "learning_rate": 9.941393192187534e-05, "loss": 0.0573, "step": 19270 }, { "epoch": 0.6518137868082086, "grad_norm": 0.31455084681510925, "learning_rate": 9.941266914592551e-05, "loss": 0.0663, "step": 19280 }, { "epoch": 0.6521518644984617, "grad_norm": 0.3477047085762024, "learning_rate": 9.941140501905188e-05, "loss": 0.0636, "step": 19290 }, { "epoch": 0.6524899421887149, "grad_norm": 0.16996857523918152, "learning_rate": 9.941013954128899e-05, "loss": 0.0676, "step": 19300 }, { "epoch": 0.6528280198789682, "grad_norm": 0.24304117262363434, "learning_rate": 9.940887271267145e-05, "loss": 0.0604, "step": 19310 }, { "epoch": 0.6531660975692214, "grad_norm": 0.27502143383026123, "learning_rate": 9.94076045332339e-05, "loss": 0.0607, "step": 19320 }, { "epoch": 0.6535041752594746, "grad_norm": 0.2845669090747833, "learning_rate": 9.9406335003011e-05, "loss": 0.0573, "step": 19330 }, { "epoch": 0.6538422529497279, "grad_norm": 0.204901322722435, "learning_rate": 9.940506412203746e-05, "loss": 0.0623, "step": 19340 }, { "epoch": 0.654180330639981, "grad_norm": 0.17361359298229218, "learning_rate": 9.940379189034802e-05, "loss": 0.0573, "step": 19350 }, { "epoch": 0.6545184083302343, "grad_norm": 0.28859180212020874, "learning_rate": 9.940251830797749e-05, "loss": 0.0667, "step": 19360 }, { "epoch": 0.6548564860204875, "grad_norm": 0.23246252536773682, "learning_rate": 9.940124337496065e-05, "loss": 0.0718, "step": 19370 }, { "epoch": 0.6551945637107407, "grad_norm": 0.17955367267131805, "learning_rate": 9.939996709133239e-05, "loss": 0.0621, "step": 19380 }, { "epoch": 0.655532641400994, "grad_norm": 0.26934266090393066, "learning_rate": 9.939868945712758e-05, "loss": 0.0614, "step": 19390 }, { "epoch": 0.6558707190912472, "grad_norm": 0.3173200190067291, "learning_rate": 9.939741047238115e-05, "loss": 0.0655, "step": 19400 }, { "epoch": 0.6562087967815003, "grad_norm": 0.24572840332984924, "learning_rate": 9.939613013712809e-05, "loss": 0.0635, "step": 19410 }, { "epoch": 0.6565468744717536, "grad_norm": 0.31741783022880554, "learning_rate": 9.939484845140339e-05, "loss": 0.0599, "step": 19420 }, { "epoch": 0.6568849521620068, "grad_norm": 0.34694820642471313, "learning_rate": 9.939356541524207e-05, "loss": 0.0627, "step": 19430 }, { "epoch": 0.65722302985226, "grad_norm": 0.20386773347854614, "learning_rate": 9.939228102867926e-05, "loss": 0.0645, "step": 19440 }, { "epoch": 0.6575611075425133, "grad_norm": 0.19213855266571045, "learning_rate": 9.939099529175003e-05, "loss": 0.0587, "step": 19450 }, { "epoch": 0.6578991852327665, "grad_norm": 0.20066815614700317, "learning_rate": 9.938970820448951e-05, "loss": 0.0564, "step": 19460 }, { "epoch": 0.6582372629230198, "grad_norm": 0.182091623544693, "learning_rate": 9.938841976693296e-05, "loss": 0.0582, "step": 19470 }, { "epoch": 0.6585753406132729, "grad_norm": 0.2582418918609619, "learning_rate": 9.938712997911557e-05, "loss": 0.065, "step": 19480 }, { "epoch": 0.6589134183035261, "grad_norm": 0.3989869952201843, "learning_rate": 9.938583884107257e-05, "loss": 0.0635, "step": 19490 }, { "epoch": 0.6592514959937794, "grad_norm": 0.2920822501182556, "learning_rate": 9.938454635283932e-05, "loss": 0.064, "step": 19500 }, { "epoch": 0.6595895736840326, "grad_norm": 0.28245478868484497, "learning_rate": 9.938325251445112e-05, "loss": 0.0648, "step": 19510 }, { "epoch": 0.6599276513742858, "grad_norm": 0.2086271345615387, "learning_rate": 9.938195732594335e-05, "loss": 0.0574, "step": 19520 }, { "epoch": 0.6602657290645391, "grad_norm": 0.3799055218696594, "learning_rate": 9.93806607873514e-05, "loss": 0.0634, "step": 19530 }, { "epoch": 0.6606038067547922, "grad_norm": 0.40510088205337524, "learning_rate": 9.937936289871075e-05, "loss": 0.0554, "step": 19540 }, { "epoch": 0.6609418844450454, "grad_norm": 0.6236116886138916, "learning_rate": 9.937806366005686e-05, "loss": 0.0603, "step": 19550 }, { "epoch": 0.6612799621352987, "grad_norm": 0.23869630694389343, "learning_rate": 9.937676307142527e-05, "loss": 0.0621, "step": 19560 }, { "epoch": 0.6616180398255519, "grad_norm": 0.2971199154853821, "learning_rate": 9.937546113285152e-05, "loss": 0.0618, "step": 19570 }, { "epoch": 0.6619561175158052, "grad_norm": 0.20245224237442017, "learning_rate": 9.937415784437123e-05, "loss": 0.0556, "step": 19580 }, { "epoch": 0.6622941952060584, "grad_norm": 0.2164304107427597, "learning_rate": 9.937285320601999e-05, "loss": 0.0641, "step": 19590 }, { "epoch": 0.6626322728963115, "grad_norm": 0.22053338587284088, "learning_rate": 9.93715472178335e-05, "loss": 0.0647, "step": 19600 }, { "epoch": 0.6629703505865648, "grad_norm": 0.304663747549057, "learning_rate": 9.937023987984746e-05, "loss": 0.0669, "step": 19610 }, { "epoch": 0.663308428276818, "grad_norm": 0.35113778710365295, "learning_rate": 9.936893119209761e-05, "loss": 0.0579, "step": 19620 }, { "epoch": 0.6636465059670712, "grad_norm": 0.22640392184257507, "learning_rate": 9.936762115461972e-05, "loss": 0.0617, "step": 19630 }, { "epoch": 0.6639845836573245, "grad_norm": 0.23715001344680786, "learning_rate": 9.936630976744963e-05, "loss": 0.0536, "step": 19640 }, { "epoch": 0.6643226613475777, "grad_norm": 0.3353125751018524, "learning_rate": 9.936499703062316e-05, "loss": 0.0597, "step": 19650 }, { "epoch": 0.6646607390378308, "grad_norm": 0.22203519940376282, "learning_rate": 9.936368294417623e-05, "loss": 0.0689, "step": 19660 }, { "epoch": 0.6649988167280841, "grad_norm": 0.2971036434173584, "learning_rate": 9.936236750814475e-05, "loss": 0.0654, "step": 19670 }, { "epoch": 0.6653368944183373, "grad_norm": 0.454489529132843, "learning_rate": 9.936105072256467e-05, "loss": 0.0608, "step": 19680 }, { "epoch": 0.6656749721085905, "grad_norm": 0.21811555325984955, "learning_rate": 9.935973258747203e-05, "loss": 0.0595, "step": 19690 }, { "epoch": 0.6660130497988438, "grad_norm": 0.5677911639213562, "learning_rate": 9.935841310290283e-05, "loss": 0.0605, "step": 19700 }, { "epoch": 0.666351127489097, "grad_norm": 0.29794979095458984, "learning_rate": 9.935709226889319e-05, "loss": 0.057, "step": 19710 }, { "epoch": 0.6666892051793503, "grad_norm": 0.26399630308151245, "learning_rate": 9.935577008547915e-05, "loss": 0.0569, "step": 19720 }, { "epoch": 0.6670272828696034, "grad_norm": 0.27342432737350464, "learning_rate": 9.935444655269692e-05, "loss": 0.0546, "step": 19730 }, { "epoch": 0.6673653605598566, "grad_norm": 0.31471917033195496, "learning_rate": 9.935312167058265e-05, "loss": 0.0665, "step": 19740 }, { "epoch": 0.6677034382501099, "grad_norm": 0.2946819067001343, "learning_rate": 9.935179543917258e-05, "loss": 0.0696, "step": 19750 }, { "epoch": 0.6680415159403631, "grad_norm": 0.19358427822589874, "learning_rate": 9.935046785850297e-05, "loss": 0.0688, "step": 19760 }, { "epoch": 0.6683795936306163, "grad_norm": 0.19944456219673157, "learning_rate": 9.93491389286101e-05, "loss": 0.065, "step": 19770 }, { "epoch": 0.6687176713208696, "grad_norm": 0.3449261784553528, "learning_rate": 9.934780864953031e-05, "loss": 0.0537, "step": 19780 }, { "epoch": 0.6690557490111227, "grad_norm": 0.32291048765182495, "learning_rate": 9.934647702129998e-05, "loss": 0.069, "step": 19790 }, { "epoch": 0.6693938267013759, "grad_norm": 0.2964285910129547, "learning_rate": 9.934514404395549e-05, "loss": 0.057, "step": 19800 }, { "epoch": 0.6697319043916292, "grad_norm": 0.18707768619060516, "learning_rate": 9.934380971753333e-05, "loss": 0.0676, "step": 19810 }, { "epoch": 0.6700699820818824, "grad_norm": 0.22292523086071014, "learning_rate": 9.934247404206992e-05, "loss": 0.0626, "step": 19820 }, { "epoch": 0.6704080597721357, "grad_norm": 0.3305288851261139, "learning_rate": 9.934113701760181e-05, "loss": 0.0696, "step": 19830 }, { "epoch": 0.6707461374623889, "grad_norm": 0.21047332882881165, "learning_rate": 9.933979864416558e-05, "loss": 0.0632, "step": 19840 }, { "epoch": 0.671084215152642, "grad_norm": 0.23440873622894287, "learning_rate": 9.933845892179776e-05, "loss": 0.0591, "step": 19850 }, { "epoch": 0.6714222928428953, "grad_norm": 0.2579069137573242, "learning_rate": 9.933711785053502e-05, "loss": 0.0662, "step": 19860 }, { "epoch": 0.6717603705331485, "grad_norm": 0.2145482301712036, "learning_rate": 9.933577543041401e-05, "loss": 0.0647, "step": 19870 }, { "epoch": 0.6720984482234017, "grad_norm": 0.30641523003578186, "learning_rate": 9.933443166147144e-05, "loss": 0.059, "step": 19880 }, { "epoch": 0.672436525913655, "grad_norm": 0.2547534108161926, "learning_rate": 9.933308654374403e-05, "loss": 0.0599, "step": 19890 }, { "epoch": 0.6727746036039082, "grad_norm": 0.2390393316745758, "learning_rate": 9.933174007726858e-05, "loss": 0.06, "step": 19900 }, { "epoch": 0.6731126812941614, "grad_norm": 0.27033278346061707, "learning_rate": 9.933039226208189e-05, "loss": 0.0625, "step": 19910 }, { "epoch": 0.6734507589844146, "grad_norm": 0.5190185904502869, "learning_rate": 9.93290430982208e-05, "loss": 0.0642, "step": 19920 }, { "epoch": 0.6737888366746678, "grad_norm": 0.3295147716999054, "learning_rate": 9.93276925857222e-05, "loss": 0.0669, "step": 19930 }, { "epoch": 0.674126914364921, "grad_norm": 0.2243558019399643, "learning_rate": 9.932634072462304e-05, "loss": 0.0602, "step": 19940 }, { "epoch": 0.6744649920551743, "grad_norm": 0.2074432373046875, "learning_rate": 9.932498751496024e-05, "loss": 0.0612, "step": 19950 }, { "epoch": 0.6748030697454275, "grad_norm": 0.1918170005083084, "learning_rate": 9.93236329567708e-05, "loss": 0.0588, "step": 19960 }, { "epoch": 0.6751411474356808, "grad_norm": 0.2888266146183014, "learning_rate": 9.932227705009178e-05, "loss": 0.0639, "step": 19970 }, { "epoch": 0.675479225125934, "grad_norm": 0.3325073719024658, "learning_rate": 9.932091979496023e-05, "loss": 0.0651, "step": 19980 }, { "epoch": 0.6758173028161871, "grad_norm": 0.19314125180244446, "learning_rate": 9.931956119141326e-05, "loss": 0.0623, "step": 19990 }, { "epoch": 0.6761553805064404, "grad_norm": 0.305599182844162, "learning_rate": 9.931820123948803e-05, "loss": 0.0637, "step": 20000 }, { "epoch": 0.6764934581966936, "grad_norm": 0.37503334879875183, "learning_rate": 9.931683993922169e-05, "loss": 0.0621, "step": 20010 }, { "epoch": 0.6768315358869468, "grad_norm": 0.2874654233455658, "learning_rate": 9.931547729065147e-05, "loss": 0.0668, "step": 20020 }, { "epoch": 0.6771696135772001, "grad_norm": 0.381188303232193, "learning_rate": 9.931411329381464e-05, "loss": 0.0611, "step": 20030 }, { "epoch": 0.6775076912674532, "grad_norm": 0.22256462275981903, "learning_rate": 9.931274794874847e-05, "loss": 0.0626, "step": 20040 }, { "epoch": 0.6778457689577064, "grad_norm": 0.27457693219184875, "learning_rate": 9.931138125549031e-05, "loss": 0.0668, "step": 20050 }, { "epoch": 0.6781838466479597, "grad_norm": 0.37672707438468933, "learning_rate": 9.93100132140775e-05, "loss": 0.0678, "step": 20060 }, { "epoch": 0.6785219243382129, "grad_norm": 0.3512507379055023, "learning_rate": 9.930864382454746e-05, "loss": 0.0697, "step": 20070 }, { "epoch": 0.6788600020284662, "grad_norm": 0.2865961194038391, "learning_rate": 9.930727308693762e-05, "loss": 0.067, "step": 20080 }, { "epoch": 0.6791980797187194, "grad_norm": 0.20125019550323486, "learning_rate": 9.930590100128545e-05, "loss": 0.0554, "step": 20090 }, { "epoch": 0.6795361574089726, "grad_norm": 0.2546904683113098, "learning_rate": 9.930452756762848e-05, "loss": 0.0652, "step": 20100 }, { "epoch": 0.6798742350992258, "grad_norm": 0.4695655107498169, "learning_rate": 9.930315278600425e-05, "loss": 0.0578, "step": 20110 }, { "epoch": 0.680212312789479, "grad_norm": 0.31424272060394287, "learning_rate": 9.930177665645034e-05, "loss": 0.0586, "step": 20120 }, { "epoch": 0.6805503904797322, "grad_norm": 0.25751903653144836, "learning_rate": 9.930039917900437e-05, "loss": 0.0703, "step": 20130 }, { "epoch": 0.6808884681699855, "grad_norm": 0.27470651268959045, "learning_rate": 9.929902035370403e-05, "loss": 0.0574, "step": 20140 }, { "epoch": 0.6812265458602387, "grad_norm": 0.2215442806482315, "learning_rate": 9.929764018058698e-05, "loss": 0.0609, "step": 20150 }, { "epoch": 0.6815646235504919, "grad_norm": 0.23434387147426605, "learning_rate": 9.929625865969096e-05, "loss": 0.0647, "step": 20160 }, { "epoch": 0.6819027012407451, "grad_norm": 0.6857904195785522, "learning_rate": 9.929487579105376e-05, "loss": 0.0589, "step": 20170 }, { "epoch": 0.6822407789309983, "grad_norm": 0.2063659429550171, "learning_rate": 9.929349157471319e-05, "loss": 0.0639, "step": 20180 }, { "epoch": 0.6825788566212516, "grad_norm": 0.2896900475025177, "learning_rate": 9.929210601070706e-05, "loss": 0.0637, "step": 20190 }, { "epoch": 0.6829169343115048, "grad_norm": 0.248770073056221, "learning_rate": 9.929071909907326e-05, "loss": 0.0634, "step": 20200 }, { "epoch": 0.683255012001758, "grad_norm": 0.27904462814331055, "learning_rate": 9.928933083984973e-05, "loss": 0.0531, "step": 20210 }, { "epoch": 0.6835930896920113, "grad_norm": 0.2887120842933655, "learning_rate": 9.92879412330744e-05, "loss": 0.058, "step": 20220 }, { "epoch": 0.6839311673822644, "grad_norm": 0.25639158487319946, "learning_rate": 9.928655027878529e-05, "loss": 0.0549, "step": 20230 }, { "epoch": 0.6842692450725176, "grad_norm": 0.25660648941993713, "learning_rate": 9.92851579770204e-05, "loss": 0.0641, "step": 20240 }, { "epoch": 0.6846073227627709, "grad_norm": 0.2584003210067749, "learning_rate": 9.928376432781783e-05, "loss": 0.0602, "step": 20250 }, { "epoch": 0.6849454004530241, "grad_norm": 0.2708424925804138, "learning_rate": 9.928236933121564e-05, "loss": 0.0667, "step": 20260 }, { "epoch": 0.6852834781432773, "grad_norm": 0.30685675144195557, "learning_rate": 9.928097298725199e-05, "loss": 0.0673, "step": 20270 }, { "epoch": 0.6856215558335306, "grad_norm": 0.41798949241638184, "learning_rate": 9.927957529596505e-05, "loss": 0.0656, "step": 20280 }, { "epoch": 0.6859596335237838, "grad_norm": 0.2726134955883026, "learning_rate": 9.927817625739303e-05, "loss": 0.0668, "step": 20290 }, { "epoch": 0.6862977112140369, "grad_norm": 0.24389995634555817, "learning_rate": 9.927677587157418e-05, "loss": 0.0679, "step": 20300 }, { "epoch": 0.6866357889042902, "grad_norm": 0.2885638475418091, "learning_rate": 9.92753741385468e-05, "loss": 0.0613, "step": 20310 }, { "epoch": 0.6869738665945434, "grad_norm": 0.28536438941955566, "learning_rate": 9.927397105834922e-05, "loss": 0.0626, "step": 20320 }, { "epoch": 0.6873119442847967, "grad_norm": 0.25475457310676575, "learning_rate": 9.927256663101975e-05, "loss": 0.0689, "step": 20330 }, { "epoch": 0.6876500219750499, "grad_norm": 0.2511432468891144, "learning_rate": 9.927116085659683e-05, "loss": 0.0612, "step": 20340 }, { "epoch": 0.687988099665303, "grad_norm": 0.25990229845046997, "learning_rate": 9.926975373511889e-05, "loss": 0.0575, "step": 20350 }, { "epoch": 0.6883261773555563, "grad_norm": 0.36085399985313416, "learning_rate": 9.926834526662439e-05, "loss": 0.061, "step": 20360 }, { "epoch": 0.6886642550458095, "grad_norm": 0.34296050667762756, "learning_rate": 9.926693545115184e-05, "loss": 0.0664, "step": 20370 }, { "epoch": 0.6890023327360627, "grad_norm": 0.22869449853897095, "learning_rate": 9.926552428873979e-05, "loss": 0.0596, "step": 20380 }, { "epoch": 0.689340410426316, "grad_norm": 0.4676835536956787, "learning_rate": 9.926411177942681e-05, "loss": 0.0609, "step": 20390 }, { "epoch": 0.6896784881165692, "grad_norm": 0.4117128849029541, "learning_rate": 9.926269792325151e-05, "loss": 0.0614, "step": 20400 }, { "epoch": 0.6900165658068224, "grad_norm": 0.20906853675842285, "learning_rate": 9.926128272025257e-05, "loss": 0.0618, "step": 20410 }, { "epoch": 0.6903546434970756, "grad_norm": 0.20701800286769867, "learning_rate": 9.925986617046865e-05, "loss": 0.0628, "step": 20420 }, { "epoch": 0.6906927211873288, "grad_norm": 0.2586853802204132, "learning_rate": 9.925844827393851e-05, "loss": 0.0611, "step": 20430 }, { "epoch": 0.6910307988775821, "grad_norm": 0.1994394063949585, "learning_rate": 9.925702903070089e-05, "loss": 0.06, "step": 20440 }, { "epoch": 0.6913688765678353, "grad_norm": 0.2586667537689209, "learning_rate": 9.925560844079461e-05, "loss": 0.0622, "step": 20450 }, { "epoch": 0.6917069542580885, "grad_norm": 0.3211113214492798, "learning_rate": 9.92541865042585e-05, "loss": 0.0597, "step": 20460 }, { "epoch": 0.6920450319483418, "grad_norm": 0.3262970447540283, "learning_rate": 9.925276322113143e-05, "loss": 0.0604, "step": 20470 }, { "epoch": 0.692383109638595, "grad_norm": 0.2961803674697876, "learning_rate": 9.925133859145232e-05, "loss": 0.0641, "step": 20480 }, { "epoch": 0.6927211873288481, "grad_norm": 0.2098010629415512, "learning_rate": 9.924991261526011e-05, "loss": 0.0555, "step": 20490 }, { "epoch": 0.6930592650191014, "grad_norm": 0.23326613008975983, "learning_rate": 9.92484852925938e-05, "loss": 0.059, "step": 20500 }, { "epoch": 0.6933973427093546, "grad_norm": 0.18452462553977966, "learning_rate": 9.92470566234924e-05, "loss": 0.0578, "step": 20510 }, { "epoch": 0.6937354203996078, "grad_norm": 0.27195167541503906, "learning_rate": 9.924562660799497e-05, "loss": 0.0642, "step": 20520 }, { "epoch": 0.6940734980898611, "grad_norm": 0.23868031799793243, "learning_rate": 9.924419524614061e-05, "loss": 0.0634, "step": 20530 }, { "epoch": 0.6944115757801143, "grad_norm": 0.28563180565834045, "learning_rate": 9.924276253796845e-05, "loss": 0.0683, "step": 20540 }, { "epoch": 0.6947496534703675, "grad_norm": 0.2138938307762146, "learning_rate": 9.924132848351767e-05, "loss": 0.0692, "step": 20550 }, { "epoch": 0.6950877311606207, "grad_norm": 0.46472856402397156, "learning_rate": 9.923989308282746e-05, "loss": 0.0599, "step": 20560 }, { "epoch": 0.6954258088508739, "grad_norm": 0.30031585693359375, "learning_rate": 9.923845633593708e-05, "loss": 0.063, "step": 20570 }, { "epoch": 0.6957638865411272, "grad_norm": 0.2705516219139099, "learning_rate": 9.92370182428858e-05, "loss": 0.0693, "step": 20580 }, { "epoch": 0.6961019642313804, "grad_norm": 0.3510136902332306, "learning_rate": 9.923557880371293e-05, "loss": 0.0726, "step": 20590 }, { "epoch": 0.6964400419216336, "grad_norm": 0.16180498898029327, "learning_rate": 9.923413801845785e-05, "loss": 0.0557, "step": 20600 }, { "epoch": 0.6967781196118868, "grad_norm": 0.2544477880001068, "learning_rate": 9.923269588715991e-05, "loss": 0.0599, "step": 20610 }, { "epoch": 0.69711619730214, "grad_norm": 0.26875850558280945, "learning_rate": 9.923125240985857e-05, "loss": 0.0592, "step": 20620 }, { "epoch": 0.6974542749923932, "grad_norm": 0.48395290970802307, "learning_rate": 9.922980758659331e-05, "loss": 0.0676, "step": 20630 }, { "epoch": 0.6977923526826465, "grad_norm": 0.3673531115055084, "learning_rate": 9.922836141740357e-05, "loss": 0.0604, "step": 20640 }, { "epoch": 0.6981304303728997, "grad_norm": 0.25585079193115234, "learning_rate": 9.922691390232894e-05, "loss": 0.0545, "step": 20650 }, { "epoch": 0.6984685080631529, "grad_norm": 0.18349865078926086, "learning_rate": 9.922546504140897e-05, "loss": 0.0633, "step": 20660 }, { "epoch": 0.6988065857534062, "grad_norm": 0.24034102261066437, "learning_rate": 9.922401483468327e-05, "loss": 0.0614, "step": 20670 }, { "epoch": 0.6991446634436593, "grad_norm": 0.18373385071754456, "learning_rate": 9.922256328219151e-05, "loss": 0.0587, "step": 20680 }, { "epoch": 0.6994827411339126, "grad_norm": 0.2194625735282898, "learning_rate": 9.922111038397337e-05, "loss": 0.0609, "step": 20690 }, { "epoch": 0.6998208188241658, "grad_norm": 0.29780906438827515, "learning_rate": 9.921965614006855e-05, "loss": 0.0646, "step": 20700 }, { "epoch": 0.700158896514419, "grad_norm": 0.40774744749069214, "learning_rate": 9.921820055051682e-05, "loss": 0.0664, "step": 20710 }, { "epoch": 0.7004969742046723, "grad_norm": 0.3103221654891968, "learning_rate": 9.9216743615358e-05, "loss": 0.0593, "step": 20720 }, { "epoch": 0.7008350518949255, "grad_norm": 0.19792978465557098, "learning_rate": 9.921528533463186e-05, "loss": 0.0612, "step": 20730 }, { "epoch": 0.7011731295851786, "grad_norm": 0.5719030499458313, "learning_rate": 9.921382570837835e-05, "loss": 0.0568, "step": 20740 }, { "epoch": 0.7015112072754319, "grad_norm": 0.33452144265174866, "learning_rate": 9.921236473663731e-05, "loss": 0.0619, "step": 20750 }, { "epoch": 0.7018492849656851, "grad_norm": 0.2637845277786255, "learning_rate": 9.921090241944872e-05, "loss": 0.0625, "step": 20760 }, { "epoch": 0.7021873626559383, "grad_norm": 0.2920602560043335, "learning_rate": 9.920943875685254e-05, "loss": 0.065, "step": 20770 }, { "epoch": 0.7025254403461916, "grad_norm": 0.4026431441307068, "learning_rate": 9.920797374888879e-05, "loss": 0.07, "step": 20780 }, { "epoch": 0.7028635180364448, "grad_norm": 0.22391392290592194, "learning_rate": 9.920650739559751e-05, "loss": 0.063, "step": 20790 }, { "epoch": 0.703201595726698, "grad_norm": 0.3494378924369812, "learning_rate": 9.920503969701882e-05, "loss": 0.0601, "step": 20800 }, { "epoch": 0.7035396734169512, "grad_norm": 0.2925780713558197, "learning_rate": 9.920357065319284e-05, "loss": 0.0617, "step": 20810 }, { "epoch": 0.7038777511072044, "grad_norm": 0.30496132373809814, "learning_rate": 9.92021002641597e-05, "loss": 0.0607, "step": 20820 }, { "epoch": 0.7042158287974577, "grad_norm": 0.23376961052417755, "learning_rate": 9.920062852995964e-05, "loss": 0.0631, "step": 20830 }, { "epoch": 0.7045539064877109, "grad_norm": 0.4520890712738037, "learning_rate": 9.919915545063286e-05, "loss": 0.0604, "step": 20840 }, { "epoch": 0.7048919841779641, "grad_norm": 0.28967487812042236, "learning_rate": 9.919768102621966e-05, "loss": 0.0623, "step": 20850 }, { "epoch": 0.7052300618682174, "grad_norm": 0.520036518573761, "learning_rate": 9.919620525676036e-05, "loss": 0.0619, "step": 20860 }, { "epoch": 0.7055681395584705, "grad_norm": 0.2619536817073822, "learning_rate": 9.919472814229528e-05, "loss": 0.0646, "step": 20870 }, { "epoch": 0.7059062172487237, "grad_norm": 0.1743474304676056, "learning_rate": 9.919324968286481e-05, "loss": 0.0542, "step": 20880 }, { "epoch": 0.706244294938977, "grad_norm": 0.27430301904678345, "learning_rate": 9.919176987850938e-05, "loss": 0.0631, "step": 20890 }, { "epoch": 0.7065823726292302, "grad_norm": 0.25795236229896545, "learning_rate": 9.919028872926946e-05, "loss": 0.0627, "step": 20900 }, { "epoch": 0.7069204503194834, "grad_norm": 0.3918288052082062, "learning_rate": 9.918880623518549e-05, "loss": 0.0625, "step": 20910 }, { "epoch": 0.7072585280097367, "grad_norm": 0.3917545676231384, "learning_rate": 9.918732239629806e-05, "loss": 0.0574, "step": 20920 }, { "epoch": 0.7075966056999898, "grad_norm": 0.425252765417099, "learning_rate": 9.918583721264771e-05, "loss": 0.0586, "step": 20930 }, { "epoch": 0.7079346833902431, "grad_norm": 0.38320550322532654, "learning_rate": 9.918435068427505e-05, "loss": 0.0622, "step": 20940 }, { "epoch": 0.7082727610804963, "grad_norm": 0.30925995111465454, "learning_rate": 9.918286281122072e-05, "loss": 0.0613, "step": 20950 }, { "epoch": 0.7086108387707495, "grad_norm": 0.33450567722320557, "learning_rate": 9.918137359352539e-05, "loss": 0.058, "step": 20960 }, { "epoch": 0.7089489164610028, "grad_norm": 0.25966179370880127, "learning_rate": 9.917988303122978e-05, "loss": 0.0631, "step": 20970 }, { "epoch": 0.709286994151256, "grad_norm": 0.2822137475013733, "learning_rate": 9.917839112437466e-05, "loss": 0.0587, "step": 20980 }, { "epoch": 0.7096250718415091, "grad_norm": 0.15396030247211456, "learning_rate": 9.917689787300079e-05, "loss": 0.0619, "step": 20990 }, { "epoch": 0.7099631495317624, "grad_norm": 0.18847371637821198, "learning_rate": 9.917540327714902e-05, "loss": 0.068, "step": 21000 }, { "epoch": 0.7103012272220156, "grad_norm": 0.3205145299434662, "learning_rate": 9.917390733686018e-05, "loss": 0.0664, "step": 21010 }, { "epoch": 0.7106393049122688, "grad_norm": 0.189012810587883, "learning_rate": 9.91724100521752e-05, "loss": 0.0586, "step": 21020 }, { "epoch": 0.7109773826025221, "grad_norm": 0.34705910086631775, "learning_rate": 9.917091142313499e-05, "loss": 0.0612, "step": 21030 }, { "epoch": 0.7113154602927753, "grad_norm": 0.2482767254114151, "learning_rate": 9.916941144978056e-05, "loss": 0.0688, "step": 21040 }, { "epoch": 0.7116535379830286, "grad_norm": 0.21112586557865143, "learning_rate": 9.916791013215288e-05, "loss": 0.0569, "step": 21050 }, { "epoch": 0.7119916156732817, "grad_norm": 0.29766348004341125, "learning_rate": 9.9166407470293e-05, "loss": 0.0607, "step": 21060 }, { "epoch": 0.7123296933635349, "grad_norm": 0.42625612020492554, "learning_rate": 9.9164903464242e-05, "loss": 0.0627, "step": 21070 }, { "epoch": 0.7126677710537882, "grad_norm": 0.3826825022697449, "learning_rate": 9.916339811404104e-05, "loss": 0.0642, "step": 21080 }, { "epoch": 0.7130058487440414, "grad_norm": 0.4095350503921509, "learning_rate": 9.916189141973122e-05, "loss": 0.0671, "step": 21090 }, { "epoch": 0.7133439264342946, "grad_norm": 0.3799176514148712, "learning_rate": 9.916038338135378e-05, "loss": 0.0659, "step": 21100 }, { "epoch": 0.7136820041245479, "grad_norm": 0.5114386677742004, "learning_rate": 9.915887399894992e-05, "loss": 0.0591, "step": 21110 }, { "epoch": 0.714020081814801, "grad_norm": 0.24535703659057617, "learning_rate": 9.91573632725609e-05, "loss": 0.0603, "step": 21120 }, { "epoch": 0.7143581595050542, "grad_norm": 0.32069721817970276, "learning_rate": 9.915585120222806e-05, "loss": 0.063, "step": 21130 }, { "epoch": 0.7146962371953075, "grad_norm": 0.3311024606227875, "learning_rate": 9.915433778799269e-05, "loss": 0.0568, "step": 21140 }, { "epoch": 0.7150343148855607, "grad_norm": 0.22670191526412964, "learning_rate": 9.915282302989619e-05, "loss": 0.0597, "step": 21150 }, { "epoch": 0.715372392575814, "grad_norm": 0.3043304681777954, "learning_rate": 9.915130692798e-05, "loss": 0.0635, "step": 21160 }, { "epoch": 0.7157104702660672, "grad_norm": 0.24471431970596313, "learning_rate": 9.914978948228553e-05, "loss": 0.059, "step": 21170 }, { "epoch": 0.7160485479563203, "grad_norm": 0.2842797338962555, "learning_rate": 9.914827069285427e-05, "loss": 0.0588, "step": 21180 }, { "epoch": 0.7163866256465736, "grad_norm": 0.2610223591327667, "learning_rate": 9.914675055972777e-05, "loss": 0.0544, "step": 21190 }, { "epoch": 0.7167247033368268, "grad_norm": 0.4724378287792206, "learning_rate": 9.914522908294757e-05, "loss": 0.0568, "step": 21200 }, { "epoch": 0.71706278102708, "grad_norm": 0.3239700198173523, "learning_rate": 9.914370626255526e-05, "loss": 0.0532, "step": 21210 }, { "epoch": 0.7174008587173333, "grad_norm": 0.2614169418811798, "learning_rate": 9.914218209859249e-05, "loss": 0.0581, "step": 21220 }, { "epoch": 0.7177389364075865, "grad_norm": 0.26266753673553467, "learning_rate": 9.914065659110092e-05, "loss": 0.0571, "step": 21230 }, { "epoch": 0.7180770140978396, "grad_norm": 0.2003590166568756, "learning_rate": 9.913912974012228e-05, "loss": 0.0593, "step": 21240 }, { "epoch": 0.7184150917880929, "grad_norm": 0.21711453795433044, "learning_rate": 9.913760154569827e-05, "loss": 0.0599, "step": 21250 }, { "epoch": 0.7187531694783461, "grad_norm": 0.2721111476421356, "learning_rate": 9.913607200787069e-05, "loss": 0.061, "step": 21260 }, { "epoch": 0.7190912471685993, "grad_norm": 0.22581954300403595, "learning_rate": 9.913454112668136e-05, "loss": 0.0525, "step": 21270 }, { "epoch": 0.7194293248588526, "grad_norm": 0.38956019282341003, "learning_rate": 9.913300890217216e-05, "loss": 0.0595, "step": 21280 }, { "epoch": 0.7197674025491058, "grad_norm": 0.3163206875324249, "learning_rate": 9.913147533438494e-05, "loss": 0.055, "step": 21290 }, { "epoch": 0.720105480239359, "grad_norm": 0.3665933310985565, "learning_rate": 9.912994042336164e-05, "loss": 0.0603, "step": 21300 }, { "epoch": 0.7204435579296122, "grad_norm": 0.355698823928833, "learning_rate": 9.912840416914423e-05, "loss": 0.0664, "step": 21310 }, { "epoch": 0.7207816356198654, "grad_norm": 0.2944171726703644, "learning_rate": 9.91268665717747e-05, "loss": 0.0572, "step": 21320 }, { "epoch": 0.7211197133101187, "grad_norm": 0.26194143295288086, "learning_rate": 9.912532763129511e-05, "loss": 0.0645, "step": 21330 }, { "epoch": 0.7214577910003719, "grad_norm": 0.24555765092372894, "learning_rate": 9.91237873477475e-05, "loss": 0.0538, "step": 21340 }, { "epoch": 0.7217958686906251, "grad_norm": 0.24582739174365997, "learning_rate": 9.912224572117402e-05, "loss": 0.0655, "step": 21350 }, { "epoch": 0.7221339463808784, "grad_norm": 0.29200518131256104, "learning_rate": 9.912070275161677e-05, "loss": 0.0574, "step": 21360 }, { "epoch": 0.7224720240711315, "grad_norm": 0.2039177566766739, "learning_rate": 9.911915843911797e-05, "loss": 0.0587, "step": 21370 }, { "epoch": 0.7228101017613847, "grad_norm": 0.17131271958351135, "learning_rate": 9.911761278371985e-05, "loss": 0.0546, "step": 21380 }, { "epoch": 0.723148179451638, "grad_norm": 0.35508811473846436, "learning_rate": 9.911606578546462e-05, "loss": 0.0597, "step": 21390 }, { "epoch": 0.7234862571418912, "grad_norm": 0.4192199409008026, "learning_rate": 9.911451744439461e-05, "loss": 0.0614, "step": 21400 }, { "epoch": 0.7238243348321445, "grad_norm": 0.28428637981414795, "learning_rate": 9.911296776055217e-05, "loss": 0.0656, "step": 21410 }, { "epoch": 0.7241624125223977, "grad_norm": 0.25075194239616394, "learning_rate": 9.911141673397961e-05, "loss": 0.0553, "step": 21420 }, { "epoch": 0.7245004902126508, "grad_norm": 0.35661864280700684, "learning_rate": 9.910986436471938e-05, "loss": 0.0663, "step": 21430 }, { "epoch": 0.7248385679029041, "grad_norm": 0.34606269001960754, "learning_rate": 9.910831065281391e-05, "loss": 0.0621, "step": 21440 }, { "epoch": 0.7251766455931573, "grad_norm": 0.4895918369293213, "learning_rate": 9.91067555983057e-05, "loss": 0.0617, "step": 21450 }, { "epoch": 0.7255147232834105, "grad_norm": 0.3626318871974945, "learning_rate": 9.910519920123719e-05, "loss": 0.061, "step": 21460 }, { "epoch": 0.7258528009736638, "grad_norm": 0.3222164511680603, "learning_rate": 9.910364146165102e-05, "loss": 0.0633, "step": 21470 }, { "epoch": 0.726190878663917, "grad_norm": 0.29740774631500244, "learning_rate": 9.910208237958974e-05, "loss": 0.057, "step": 21480 }, { "epoch": 0.7265289563541701, "grad_norm": 0.34712105989456177, "learning_rate": 9.910052195509595e-05, "loss": 0.0572, "step": 21490 }, { "epoch": 0.7268670340444234, "grad_norm": 0.29136139154434204, "learning_rate": 9.909896018821234e-05, "loss": 0.0619, "step": 21500 }, { "epoch": 0.7272051117346766, "grad_norm": 0.32350295782089233, "learning_rate": 9.909739707898162e-05, "loss": 0.0547, "step": 21510 }, { "epoch": 0.7275431894249299, "grad_norm": 0.2733055651187897, "learning_rate": 9.909583262744651e-05, "loss": 0.0569, "step": 21520 }, { "epoch": 0.7278812671151831, "grad_norm": 0.3625321090221405, "learning_rate": 9.909426683364975e-05, "loss": 0.057, "step": 21530 }, { "epoch": 0.7282193448054363, "grad_norm": 0.24806687235832214, "learning_rate": 9.909269969763422e-05, "loss": 0.0598, "step": 21540 }, { "epoch": 0.7285574224956896, "grad_norm": 0.22708012163639069, "learning_rate": 9.909113121944269e-05, "loss": 0.0643, "step": 21550 }, { "epoch": 0.7288955001859427, "grad_norm": 0.2794658839702606, "learning_rate": 9.90895613991181e-05, "loss": 0.0622, "step": 21560 }, { "epoch": 0.7292335778761959, "grad_norm": 0.2341281771659851, "learning_rate": 9.908799023670333e-05, "loss": 0.0614, "step": 21570 }, { "epoch": 0.7295716555664492, "grad_norm": 0.24664545059204102, "learning_rate": 9.908641773224134e-05, "loss": 0.0608, "step": 21580 }, { "epoch": 0.7299097332567024, "grad_norm": 0.2113097459077835, "learning_rate": 9.908484388577514e-05, "loss": 0.0632, "step": 21590 }, { "epoch": 0.7302478109469556, "grad_norm": 0.17329661548137665, "learning_rate": 9.908326869734775e-05, "loss": 0.0574, "step": 21600 }, { "epoch": 0.7305858886372089, "grad_norm": 0.3266075551509857, "learning_rate": 9.908169216700223e-05, "loss": 0.0645, "step": 21610 }, { "epoch": 0.730923966327462, "grad_norm": 0.518523097038269, "learning_rate": 9.90801142947817e-05, "loss": 0.0605, "step": 21620 }, { "epoch": 0.7312620440177152, "grad_norm": 0.2752475440502167, "learning_rate": 9.907853508072927e-05, "loss": 0.0652, "step": 21630 }, { "epoch": 0.7316001217079685, "grad_norm": 0.250068336725235, "learning_rate": 9.907695452488812e-05, "loss": 0.0598, "step": 21640 }, { "epoch": 0.7319381993982217, "grad_norm": 0.2723868489265442, "learning_rate": 9.907537262730149e-05, "loss": 0.0598, "step": 21650 }, { "epoch": 0.732276277088475, "grad_norm": 0.296554297208786, "learning_rate": 9.907378938801259e-05, "loss": 0.0601, "step": 21660 }, { "epoch": 0.7326143547787282, "grad_norm": 0.2427532821893692, "learning_rate": 9.907220480706474e-05, "loss": 0.0631, "step": 21670 }, { "epoch": 0.7329524324689813, "grad_norm": 0.18149849772453308, "learning_rate": 9.907061888450125e-05, "loss": 0.0654, "step": 21680 }, { "epoch": 0.7332905101592346, "grad_norm": 0.20973505079746246, "learning_rate": 9.906903162036545e-05, "loss": 0.0598, "step": 21690 }, { "epoch": 0.7336285878494878, "grad_norm": 0.2334495484828949, "learning_rate": 9.906744301470079e-05, "loss": 0.0618, "step": 21700 }, { "epoch": 0.733966665539741, "grad_norm": 0.2675803601741791, "learning_rate": 9.906585306755066e-05, "loss": 0.0613, "step": 21710 }, { "epoch": 0.7343047432299943, "grad_norm": 0.29477593302726746, "learning_rate": 9.906426177895852e-05, "loss": 0.0634, "step": 21720 }, { "epoch": 0.7346428209202475, "grad_norm": 0.20785127580165863, "learning_rate": 9.90626691489679e-05, "loss": 0.0616, "step": 21730 }, { "epoch": 0.7349808986105006, "grad_norm": 0.3009190559387207, "learning_rate": 9.906107517762238e-05, "loss": 0.0624, "step": 21740 }, { "epoch": 0.7353189763007539, "grad_norm": 0.31093016266822815, "learning_rate": 9.905947986496546e-05, "loss": 0.0619, "step": 21750 }, { "epoch": 0.7356570539910071, "grad_norm": 0.2820194959640503, "learning_rate": 9.905788321104079e-05, "loss": 0.0624, "step": 21760 }, { "epoch": 0.7359951316812604, "grad_norm": 0.23957084119319916, "learning_rate": 9.905628521589204e-05, "loss": 0.0579, "step": 21770 }, { "epoch": 0.7363332093715136, "grad_norm": 0.18700887262821198, "learning_rate": 9.905468587956285e-05, "loss": 0.0703, "step": 21780 }, { "epoch": 0.7366712870617668, "grad_norm": 0.23893146216869354, "learning_rate": 9.905308520209702e-05, "loss": 0.0677, "step": 21790 }, { "epoch": 0.7370093647520201, "grad_norm": 0.226735457777977, "learning_rate": 9.905148318353824e-05, "loss": 0.0552, "step": 21800 }, { "epoch": 0.7373474424422732, "grad_norm": 0.24632492661476135, "learning_rate": 9.904987982393035e-05, "loss": 0.0573, "step": 21810 }, { "epoch": 0.7376855201325264, "grad_norm": 0.5281603336334229, "learning_rate": 9.904827512331717e-05, "loss": 0.0618, "step": 21820 }, { "epoch": 0.7380235978227797, "grad_norm": 0.3345308303833008, "learning_rate": 9.904666908174257e-05, "loss": 0.0652, "step": 21830 }, { "epoch": 0.7383616755130329, "grad_norm": 0.268923819065094, "learning_rate": 9.904506169925047e-05, "loss": 0.0557, "step": 21840 }, { "epoch": 0.7386997532032861, "grad_norm": 0.24518953263759613, "learning_rate": 9.90434529758848e-05, "loss": 0.0563, "step": 21850 }, { "epoch": 0.7390378308935394, "grad_norm": 0.2980830669403076, "learning_rate": 9.904184291168956e-05, "loss": 0.0609, "step": 21860 }, { "epoch": 0.7393759085837925, "grad_norm": 0.2012006640434265, "learning_rate": 9.904023150670876e-05, "loss": 0.0594, "step": 21870 }, { "epoch": 0.7397139862740457, "grad_norm": 0.3207605183124542, "learning_rate": 9.903861876098644e-05, "loss": 0.0708, "step": 21880 }, { "epoch": 0.740052063964299, "grad_norm": 0.26694387197494507, "learning_rate": 9.903700467456672e-05, "loss": 0.0596, "step": 21890 }, { "epoch": 0.7403901416545522, "grad_norm": 0.2715781331062317, "learning_rate": 9.903538924749371e-05, "loss": 0.0596, "step": 21900 }, { "epoch": 0.7407282193448055, "grad_norm": 0.25681254267692566, "learning_rate": 9.903377247981159e-05, "loss": 0.0676, "step": 21910 }, { "epoch": 0.7410662970350587, "grad_norm": 0.32321593165397644, "learning_rate": 9.903215437156452e-05, "loss": 0.0508, "step": 21920 }, { "epoch": 0.7414043747253118, "grad_norm": 0.2436198741197586, "learning_rate": 9.903053492279679e-05, "loss": 0.0616, "step": 21930 }, { "epoch": 0.7417424524155651, "grad_norm": 0.20898890495300293, "learning_rate": 9.902891413355267e-05, "loss": 0.0661, "step": 21940 }, { "epoch": 0.7420805301058183, "grad_norm": 0.30593782663345337, "learning_rate": 9.902729200387644e-05, "loss": 0.0649, "step": 21950 }, { "epoch": 0.7424186077960715, "grad_norm": 0.3063855469226837, "learning_rate": 9.902566853381246e-05, "loss": 0.0577, "step": 21960 }, { "epoch": 0.7427566854863248, "grad_norm": 0.23964188992977142, "learning_rate": 9.902404372340514e-05, "loss": 0.0645, "step": 21970 }, { "epoch": 0.743094763176578, "grad_norm": 0.20646393299102783, "learning_rate": 9.902241757269886e-05, "loss": 0.0595, "step": 21980 }, { "epoch": 0.7434328408668311, "grad_norm": 0.4535864293575287, "learning_rate": 9.902079008173811e-05, "loss": 0.0609, "step": 21990 }, { "epoch": 0.7437709185570844, "grad_norm": 0.44292905926704407, "learning_rate": 9.901916125056738e-05, "loss": 0.0646, "step": 22000 }, { "epoch": 0.7441089962473376, "grad_norm": 0.28210386633872986, "learning_rate": 9.90175310792312e-05, "loss": 0.0597, "step": 22010 }, { "epoch": 0.7444470739375909, "grad_norm": 0.22355793416500092, "learning_rate": 9.901589956777413e-05, "loss": 0.0703, "step": 22020 }, { "epoch": 0.7447851516278441, "grad_norm": 0.27109774947166443, "learning_rate": 9.901426671624077e-05, "loss": 0.0606, "step": 22030 }, { "epoch": 0.7451232293180973, "grad_norm": 0.25652047991752625, "learning_rate": 9.901263252467579e-05, "loss": 0.057, "step": 22040 }, { "epoch": 0.7454613070083506, "grad_norm": 0.29260772466659546, "learning_rate": 9.901099699312384e-05, "loss": 0.0575, "step": 22050 }, { "epoch": 0.7457993846986037, "grad_norm": 0.2562659978866577, "learning_rate": 9.900936012162965e-05, "loss": 0.0554, "step": 22060 }, { "epoch": 0.7461374623888569, "grad_norm": 0.3355458974838257, "learning_rate": 9.900772191023795e-05, "loss": 0.0562, "step": 22070 }, { "epoch": 0.7464755400791102, "grad_norm": 0.275166392326355, "learning_rate": 9.900608235899356e-05, "loss": 0.0593, "step": 22080 }, { "epoch": 0.7468136177693634, "grad_norm": 0.3275170624256134, "learning_rate": 9.900444146794128e-05, "loss": 0.0622, "step": 22090 }, { "epoch": 0.7471516954596166, "grad_norm": 0.32046809792518616, "learning_rate": 9.900279923712598e-05, "loss": 0.064, "step": 22100 }, { "epoch": 0.7474897731498699, "grad_norm": 0.24532216787338257, "learning_rate": 9.900115566659257e-05, "loss": 0.0604, "step": 22110 }, { "epoch": 0.747827850840123, "grad_norm": 0.2181469351053238, "learning_rate": 9.899951075638596e-05, "loss": 0.0572, "step": 22120 }, { "epoch": 0.7481659285303763, "grad_norm": 0.2673370838165283, "learning_rate": 9.899786450655116e-05, "loss": 0.058, "step": 22130 }, { "epoch": 0.7485040062206295, "grad_norm": 0.2790282666683197, "learning_rate": 9.899621691713313e-05, "loss": 0.0607, "step": 22140 }, { "epoch": 0.7488420839108827, "grad_norm": 0.25894826650619507, "learning_rate": 9.899456798817693e-05, "loss": 0.0608, "step": 22150 }, { "epoch": 0.749180161601136, "grad_norm": 0.21224497258663177, "learning_rate": 9.899291771972765e-05, "loss": 0.0643, "step": 22160 }, { "epoch": 0.7495182392913892, "grad_norm": 0.17676378786563873, "learning_rate": 9.899126611183042e-05, "loss": 0.0613, "step": 22170 }, { "epoch": 0.7498563169816423, "grad_norm": 0.24393559992313385, "learning_rate": 9.898961316453036e-05, "loss": 0.0622, "step": 22180 }, { "epoch": 0.7501943946718956, "grad_norm": 0.17276711761951447, "learning_rate": 9.898795887787271e-05, "loss": 0.0595, "step": 22190 }, { "epoch": 0.7505324723621488, "grad_norm": 0.21787266433238983, "learning_rate": 9.898630325190265e-05, "loss": 0.062, "step": 22200 }, { "epoch": 0.750870550052402, "grad_norm": 0.21776416897773743, "learning_rate": 9.898464628666546e-05, "loss": 0.0586, "step": 22210 }, { "epoch": 0.7512086277426553, "grad_norm": 0.28851813077926636, "learning_rate": 9.898298798220644e-05, "loss": 0.0629, "step": 22220 }, { "epoch": 0.7515467054329085, "grad_norm": 0.21925337612628937, "learning_rate": 9.898132833857094e-05, "loss": 0.0612, "step": 22230 }, { "epoch": 0.7518847831231616, "grad_norm": 0.3149351477622986, "learning_rate": 9.897966735580433e-05, "loss": 0.0633, "step": 22240 }, { "epoch": 0.7522228608134149, "grad_norm": 0.19818684458732605, "learning_rate": 9.8978005033952e-05, "loss": 0.06, "step": 22250 }, { "epoch": 0.7525609385036681, "grad_norm": 0.19432540237903595, "learning_rate": 9.897634137305943e-05, "loss": 0.0605, "step": 22260 }, { "epoch": 0.7528990161939214, "grad_norm": 0.2357977330684662, "learning_rate": 9.897467637317207e-05, "loss": 0.0556, "step": 22270 }, { "epoch": 0.7532370938841746, "grad_norm": 0.26061850786209106, "learning_rate": 9.897301003433545e-05, "loss": 0.0593, "step": 22280 }, { "epoch": 0.7535751715744278, "grad_norm": 0.32699665427207947, "learning_rate": 9.897134235659515e-05, "loss": 0.059, "step": 22290 }, { "epoch": 0.7539132492646811, "grad_norm": 0.30137544870376587, "learning_rate": 9.896967333999676e-05, "loss": 0.0597, "step": 22300 }, { "epoch": 0.7542513269549342, "grad_norm": 0.30223241448402405, "learning_rate": 9.896800298458588e-05, "loss": 0.0633, "step": 22310 }, { "epoch": 0.7545894046451874, "grad_norm": 0.2109425812959671, "learning_rate": 9.896633129040819e-05, "loss": 0.0603, "step": 22320 }, { "epoch": 0.7549274823354407, "grad_norm": 0.22253474593162537, "learning_rate": 9.89646582575094e-05, "loss": 0.0623, "step": 22330 }, { "epoch": 0.7552655600256939, "grad_norm": 0.22265523672103882, "learning_rate": 9.896298388593525e-05, "loss": 0.0625, "step": 22340 }, { "epoch": 0.7556036377159471, "grad_norm": 0.24153228104114532, "learning_rate": 9.896130817573153e-05, "loss": 0.0663, "step": 22350 }, { "epoch": 0.7559417154062004, "grad_norm": 0.3268776834011078, "learning_rate": 9.895963112694403e-05, "loss": 0.0536, "step": 22360 }, { "epoch": 0.7562797930964535, "grad_norm": 0.20823997259140015, "learning_rate": 9.895795273961861e-05, "loss": 0.064, "step": 22370 }, { "epoch": 0.7566178707867068, "grad_norm": 0.3369155526161194, "learning_rate": 9.895627301380115e-05, "loss": 0.0587, "step": 22380 }, { "epoch": 0.75695594847696, "grad_norm": 0.28307080268859863, "learning_rate": 9.895459194953758e-05, "loss": 0.0605, "step": 22390 }, { "epoch": 0.7572940261672132, "grad_norm": 0.15352082252502441, "learning_rate": 9.895290954687385e-05, "loss": 0.0628, "step": 22400 }, { "epoch": 0.7576321038574665, "grad_norm": 0.27070245146751404, "learning_rate": 9.895122580585597e-05, "loss": 0.0629, "step": 22410 }, { "epoch": 0.7579701815477197, "grad_norm": 0.23552358150482178, "learning_rate": 9.894954072652996e-05, "loss": 0.0601, "step": 22420 }, { "epoch": 0.7583082592379728, "grad_norm": 0.2279796004295349, "learning_rate": 9.894785430894191e-05, "loss": 0.0599, "step": 22430 }, { "epoch": 0.7586463369282261, "grad_norm": 0.214778870344162, "learning_rate": 9.894616655313791e-05, "loss": 0.0603, "step": 22440 }, { "epoch": 0.7589844146184793, "grad_norm": 0.2436802238225937, "learning_rate": 9.894447745916409e-05, "loss": 0.0639, "step": 22450 }, { "epoch": 0.7593224923087325, "grad_norm": 0.20102444291114807, "learning_rate": 9.894278702706665e-05, "loss": 0.057, "step": 22460 }, { "epoch": 0.7596605699989858, "grad_norm": 0.3025158941745758, "learning_rate": 9.89410952568918e-05, "loss": 0.0592, "step": 22470 }, { "epoch": 0.759998647689239, "grad_norm": 0.3012106716632843, "learning_rate": 9.89394021486858e-05, "loss": 0.0612, "step": 22480 }, { "epoch": 0.7603367253794923, "grad_norm": 0.3134859502315521, "learning_rate": 9.893770770249493e-05, "loss": 0.0554, "step": 22490 }, { "epoch": 0.7606748030697454, "grad_norm": 0.23360711336135864, "learning_rate": 9.89360119183655e-05, "loss": 0.0657, "step": 22500 }, { "epoch": 0.7610128807599986, "grad_norm": 0.3121790289878845, "learning_rate": 9.893431479634391e-05, "loss": 0.0586, "step": 22510 }, { "epoch": 0.7613509584502519, "grad_norm": 0.4341184198856354, "learning_rate": 9.893261633647652e-05, "loss": 0.0658, "step": 22520 }, { "epoch": 0.7616890361405051, "grad_norm": 0.3380875289440155, "learning_rate": 9.893091653880979e-05, "loss": 0.0544, "step": 22530 }, { "epoch": 0.7620271138307583, "grad_norm": 0.22834376990795135, "learning_rate": 9.89292154033902e-05, "loss": 0.065, "step": 22540 }, { "epoch": 0.7623651915210116, "grad_norm": 0.33373886346817017, "learning_rate": 9.89275129302642e-05, "loss": 0.0664, "step": 22550 }, { "epoch": 0.7627032692112647, "grad_norm": 0.1734601855278015, "learning_rate": 9.892580911947841e-05, "loss": 0.0635, "step": 22560 }, { "epoch": 0.7630413469015179, "grad_norm": 0.28800004720687866, "learning_rate": 9.892410397107939e-05, "loss": 0.0629, "step": 22570 }, { "epoch": 0.7633794245917712, "grad_norm": 0.26500940322875977, "learning_rate": 9.892239748511372e-05, "loss": 0.0571, "step": 22580 }, { "epoch": 0.7637175022820244, "grad_norm": 0.32821375131607056, "learning_rate": 9.89206896616281e-05, "loss": 0.0608, "step": 22590 }, { "epoch": 0.7640555799722776, "grad_norm": 0.290640652179718, "learning_rate": 9.891898050066919e-05, "loss": 0.0614, "step": 22600 }, { "epoch": 0.7643936576625309, "grad_norm": 0.43561550974845886, "learning_rate": 9.891727000228374e-05, "loss": 0.0666, "step": 22610 }, { "epoch": 0.764731735352784, "grad_norm": 0.5216332674026489, "learning_rate": 9.891555816651851e-05, "loss": 0.0591, "step": 22620 }, { "epoch": 0.7650698130430373, "grad_norm": 0.34867948293685913, "learning_rate": 9.89138449934203e-05, "loss": 0.0649, "step": 22630 }, { "epoch": 0.7654078907332905, "grad_norm": 0.28488561511039734, "learning_rate": 9.891213048303592e-05, "loss": 0.0634, "step": 22640 }, { "epoch": 0.7657459684235437, "grad_norm": 0.2695598006248474, "learning_rate": 9.891041463541228e-05, "loss": 0.0629, "step": 22650 }, { "epoch": 0.766084046113797, "grad_norm": 0.24141883850097656, "learning_rate": 9.890869745059628e-05, "loss": 0.0617, "step": 22660 }, { "epoch": 0.7664221238040502, "grad_norm": 0.35766181349754333, "learning_rate": 9.890697892863487e-05, "loss": 0.0614, "step": 22670 }, { "epoch": 0.7667602014943034, "grad_norm": 0.24179191887378693, "learning_rate": 9.890525906957501e-05, "loss": 0.0574, "step": 22680 }, { "epoch": 0.7670982791845566, "grad_norm": 0.2381327599287033, "learning_rate": 9.890353787346377e-05, "loss": 0.0595, "step": 22690 }, { "epoch": 0.7674363568748098, "grad_norm": 0.3225760757923126, "learning_rate": 9.890181534034817e-05, "loss": 0.0599, "step": 22700 }, { "epoch": 0.767774434565063, "grad_norm": 0.2594650387763977, "learning_rate": 9.89000914702753e-05, "loss": 0.068, "step": 22710 }, { "epoch": 0.7681125122553163, "grad_norm": 0.26427319645881653, "learning_rate": 9.889836626329229e-05, "loss": 0.062, "step": 22720 }, { "epoch": 0.7684505899455695, "grad_norm": 0.40834805369377136, "learning_rate": 9.889663971944634e-05, "loss": 0.064, "step": 22730 }, { "epoch": 0.7687886676358228, "grad_norm": 0.2841132581233978, "learning_rate": 9.88949118387846e-05, "loss": 0.0661, "step": 22740 }, { "epoch": 0.769126745326076, "grad_norm": 0.3252575695514679, "learning_rate": 9.889318262135438e-05, "loss": 0.0696, "step": 22750 }, { "epoch": 0.7694648230163291, "grad_norm": 0.24409501254558563, "learning_rate": 9.88914520672029e-05, "loss": 0.0603, "step": 22760 }, { "epoch": 0.7698029007065824, "grad_norm": 0.2921215295791626, "learning_rate": 9.888972017637747e-05, "loss": 0.0622, "step": 22770 }, { "epoch": 0.7701409783968356, "grad_norm": 0.27528873085975647, "learning_rate": 9.888798694892548e-05, "loss": 0.0573, "step": 22780 }, { "epoch": 0.7704790560870888, "grad_norm": 0.22370368242263794, "learning_rate": 9.888625238489428e-05, "loss": 0.0573, "step": 22790 }, { "epoch": 0.7708171337773421, "grad_norm": 0.25150179862976074, "learning_rate": 9.888451648433131e-05, "loss": 0.0632, "step": 22800 }, { "epoch": 0.7711552114675952, "grad_norm": 0.2536206841468811, "learning_rate": 9.888277924728402e-05, "loss": 0.0571, "step": 22810 }, { "epoch": 0.7714932891578484, "grad_norm": 0.32066553831100464, "learning_rate": 9.88810406737999e-05, "loss": 0.0613, "step": 22820 }, { "epoch": 0.7718313668481017, "grad_norm": 0.35491514205932617, "learning_rate": 9.887930076392651e-05, "loss": 0.0574, "step": 22830 }, { "epoch": 0.7721694445383549, "grad_norm": 0.40362808108329773, "learning_rate": 9.887755951771139e-05, "loss": 0.0642, "step": 22840 }, { "epoch": 0.7725075222286081, "grad_norm": 0.33522483706474304, "learning_rate": 9.887581693520216e-05, "loss": 0.0607, "step": 22850 }, { "epoch": 0.7728455999188614, "grad_norm": 0.31311431527137756, "learning_rate": 9.887407301644646e-05, "loss": 0.0534, "step": 22860 }, { "epoch": 0.7731836776091146, "grad_norm": 0.2418423891067505, "learning_rate": 9.887232776149195e-05, "loss": 0.0581, "step": 22870 }, { "epoch": 0.7735217552993678, "grad_norm": 0.23774488270282745, "learning_rate": 9.887058117038637e-05, "loss": 0.0639, "step": 22880 }, { "epoch": 0.773859832989621, "grad_norm": 0.21751560270786285, "learning_rate": 9.886883324317748e-05, "loss": 0.0581, "step": 22890 }, { "epoch": 0.7741979106798742, "grad_norm": 0.3147844672203064, "learning_rate": 9.886708397991302e-05, "loss": 0.0615, "step": 22900 }, { "epoch": 0.7745359883701275, "grad_norm": 0.19028452038764954, "learning_rate": 9.886533338064085e-05, "loss": 0.0578, "step": 22910 }, { "epoch": 0.7748740660603807, "grad_norm": 0.2932259440422058, "learning_rate": 9.886358144540882e-05, "loss": 0.0591, "step": 22920 }, { "epoch": 0.7752121437506339, "grad_norm": 0.17857114970684052, "learning_rate": 9.886182817426483e-05, "loss": 0.0601, "step": 22930 }, { "epoch": 0.7755502214408871, "grad_norm": 0.24937887489795685, "learning_rate": 9.886007356725685e-05, "loss": 0.0598, "step": 22940 }, { "epoch": 0.7758882991311403, "grad_norm": 0.25831955671310425, "learning_rate": 9.885831762443276e-05, "loss": 0.0643, "step": 22950 }, { "epoch": 0.7762263768213935, "grad_norm": 0.3743494749069214, "learning_rate": 9.885656034584065e-05, "loss": 0.0695, "step": 22960 }, { "epoch": 0.7765644545116468, "grad_norm": 0.26239824295043945, "learning_rate": 9.885480173152854e-05, "loss": 0.0663, "step": 22970 }, { "epoch": 0.7769025322019, "grad_norm": 0.2638348937034607, "learning_rate": 9.885304178154449e-05, "loss": 0.0578, "step": 22980 }, { "epoch": 0.7772406098921533, "grad_norm": 0.25213494896888733, "learning_rate": 9.885128049593664e-05, "loss": 0.0631, "step": 22990 }, { "epoch": 0.7775786875824064, "grad_norm": 0.285079687833786, "learning_rate": 9.884951787475313e-05, "loss": 0.0588, "step": 23000 }, { "epoch": 0.7779167652726596, "grad_norm": 0.21496985852718353, "learning_rate": 9.884775391804216e-05, "loss": 0.0619, "step": 23010 }, { "epoch": 0.7782548429629129, "grad_norm": 0.2930292785167694, "learning_rate": 9.884598862585193e-05, "loss": 0.0649, "step": 23020 }, { "epoch": 0.7785929206531661, "grad_norm": 0.3094651699066162, "learning_rate": 9.884422199823073e-05, "loss": 0.0588, "step": 23030 }, { "epoch": 0.7789309983434193, "grad_norm": 0.28878700733184814, "learning_rate": 9.884245403522686e-05, "loss": 0.0647, "step": 23040 }, { "epoch": 0.7792690760336726, "grad_norm": 0.4421848952770233, "learning_rate": 9.884068473688863e-05, "loss": 0.0593, "step": 23050 }, { "epoch": 0.7796071537239257, "grad_norm": 0.26753443479537964, "learning_rate": 9.883891410326444e-05, "loss": 0.0609, "step": 23060 }, { "epoch": 0.7799452314141789, "grad_norm": 0.17184771597385406, "learning_rate": 9.883714213440269e-05, "loss": 0.0543, "step": 23070 }, { "epoch": 0.7802833091044322, "grad_norm": 0.24429893493652344, "learning_rate": 9.883536883035181e-05, "loss": 0.0627, "step": 23080 }, { "epoch": 0.7806213867946854, "grad_norm": 0.2575129270553589, "learning_rate": 9.883359419116031e-05, "loss": 0.0606, "step": 23090 }, { "epoch": 0.7809594644849387, "grad_norm": 0.31540241837501526, "learning_rate": 9.883181821687669e-05, "loss": 0.0602, "step": 23100 }, { "epoch": 0.7812975421751919, "grad_norm": 0.19532890617847443, "learning_rate": 9.883004090754948e-05, "loss": 0.0627, "step": 23110 }, { "epoch": 0.781635619865445, "grad_norm": 0.367435097694397, "learning_rate": 9.882826226322731e-05, "loss": 0.0618, "step": 23120 }, { "epoch": 0.7819736975556983, "grad_norm": 0.24684295058250427, "learning_rate": 9.882648228395879e-05, "loss": 0.0558, "step": 23130 }, { "epoch": 0.7823117752459515, "grad_norm": 0.2148052304983139, "learning_rate": 9.882470096979259e-05, "loss": 0.0629, "step": 23140 }, { "epoch": 0.7826498529362047, "grad_norm": 0.35330668091773987, "learning_rate": 9.882291832077741e-05, "loss": 0.0704, "step": 23150 }, { "epoch": 0.782987930626458, "grad_norm": 0.21466641128063202, "learning_rate": 9.882113433696199e-05, "loss": 0.0646, "step": 23160 }, { "epoch": 0.7833260083167112, "grad_norm": 0.20954559743404388, "learning_rate": 9.88193490183951e-05, "loss": 0.0594, "step": 23170 }, { "epoch": 0.7836640860069644, "grad_norm": 0.26339906454086304, "learning_rate": 9.881756236512554e-05, "loss": 0.0594, "step": 23180 }, { "epoch": 0.7840021636972176, "grad_norm": 0.29276835918426514, "learning_rate": 9.881577437720217e-05, "loss": 0.0626, "step": 23190 }, { "epoch": 0.7843402413874708, "grad_norm": 0.26244786381721497, "learning_rate": 9.881398505467385e-05, "loss": 0.0595, "step": 23200 }, { "epoch": 0.784678319077724, "grad_norm": 0.36223840713500977, "learning_rate": 9.881219439758954e-05, "loss": 0.062, "step": 23210 }, { "epoch": 0.7850163967679773, "grad_norm": 0.36496227979660034, "learning_rate": 9.881040240599817e-05, "loss": 0.0573, "step": 23220 }, { "epoch": 0.7853544744582305, "grad_norm": 0.3490529954433441, "learning_rate": 9.880860907994873e-05, "loss": 0.0584, "step": 23230 }, { "epoch": 0.7856925521484838, "grad_norm": 0.24446582794189453, "learning_rate": 9.880681441949025e-05, "loss": 0.0562, "step": 23240 }, { "epoch": 0.786030629838737, "grad_norm": 0.7284011244773865, "learning_rate": 9.880501842467181e-05, "loss": 0.0619, "step": 23250 }, { "epoch": 0.7863687075289901, "grad_norm": 0.18870609998703003, "learning_rate": 9.880322109554249e-05, "loss": 0.0634, "step": 23260 }, { "epoch": 0.7867067852192434, "grad_norm": 0.22462069988250732, "learning_rate": 9.880142243215145e-05, "loss": 0.0596, "step": 23270 }, { "epoch": 0.7870448629094966, "grad_norm": 0.28809842467308044, "learning_rate": 9.879962243454783e-05, "loss": 0.0588, "step": 23280 }, { "epoch": 0.7873829405997498, "grad_norm": 0.2904588580131531, "learning_rate": 9.87978211027809e-05, "loss": 0.0581, "step": 23290 }, { "epoch": 0.7877210182900031, "grad_norm": 0.3230495750904083, "learning_rate": 9.879601843689986e-05, "loss": 0.0582, "step": 23300 }, { "epoch": 0.7880590959802563, "grad_norm": 0.21497666835784912, "learning_rate": 9.8794214436954e-05, "loss": 0.0691, "step": 23310 }, { "epoch": 0.7883971736705094, "grad_norm": 0.1943371444940567, "learning_rate": 9.879240910299265e-05, "loss": 0.0562, "step": 23320 }, { "epoch": 0.7887352513607627, "grad_norm": 0.3924698531627655, "learning_rate": 9.879060243506518e-05, "loss": 0.0651, "step": 23330 }, { "epoch": 0.7890733290510159, "grad_norm": 0.27657386660575867, "learning_rate": 9.878879443322095e-05, "loss": 0.0618, "step": 23340 }, { "epoch": 0.7894114067412692, "grad_norm": 0.209797665476799, "learning_rate": 9.878698509750944e-05, "loss": 0.0635, "step": 23350 }, { "epoch": 0.7897494844315224, "grad_norm": 0.34184202551841736, "learning_rate": 9.878517442798006e-05, "loss": 0.0593, "step": 23360 }, { "epoch": 0.7900875621217756, "grad_norm": 0.21981148421764374, "learning_rate": 9.878336242468234e-05, "loss": 0.0583, "step": 23370 }, { "epoch": 0.7904256398120288, "grad_norm": 0.20671716332435608, "learning_rate": 9.878154908766581e-05, "loss": 0.0556, "step": 23380 }, { "epoch": 0.790763717502282, "grad_norm": 0.3103969395160675, "learning_rate": 9.877973441698007e-05, "loss": 0.0536, "step": 23390 }, { "epoch": 0.7911017951925352, "grad_norm": 0.24196603894233704, "learning_rate": 9.877791841267472e-05, "loss": 0.0571, "step": 23400 }, { "epoch": 0.7914398728827885, "grad_norm": 0.18932348489761353, "learning_rate": 9.87761010747994e-05, "loss": 0.0545, "step": 23410 }, { "epoch": 0.7917779505730417, "grad_norm": 0.20667530596256256, "learning_rate": 9.87742824034038e-05, "loss": 0.0573, "step": 23420 }, { "epoch": 0.7921160282632949, "grad_norm": 0.31804534792900085, "learning_rate": 9.877246239853764e-05, "loss": 0.0584, "step": 23430 }, { "epoch": 0.7924541059535481, "grad_norm": 0.2607397437095642, "learning_rate": 9.877064106025069e-05, "loss": 0.0641, "step": 23440 }, { "epoch": 0.7927921836438013, "grad_norm": 0.23173269629478455, "learning_rate": 9.876881838859273e-05, "loss": 0.054, "step": 23450 }, { "epoch": 0.7931302613340546, "grad_norm": 0.24607904255390167, "learning_rate": 9.876699438361359e-05, "loss": 0.0601, "step": 23460 }, { "epoch": 0.7934683390243078, "grad_norm": 0.21325811743736267, "learning_rate": 9.876516904536315e-05, "loss": 0.0605, "step": 23470 }, { "epoch": 0.793806416714561, "grad_norm": 0.3347870707511902, "learning_rate": 9.876334237389131e-05, "loss": 0.0603, "step": 23480 }, { "epoch": 0.7941444944048143, "grad_norm": 0.3228316605091095, "learning_rate": 9.876151436924801e-05, "loss": 0.0552, "step": 23490 }, { "epoch": 0.7944825720950675, "grad_norm": 0.3503703474998474, "learning_rate": 9.875968503148323e-05, "loss": 0.0596, "step": 23500 }, { "epoch": 0.7948206497853206, "grad_norm": 0.4903656244277954, "learning_rate": 9.875785436064697e-05, "loss": 0.0746, "step": 23510 }, { "epoch": 0.7951587274755739, "grad_norm": 0.282796174287796, "learning_rate": 9.875602235678929e-05, "loss": 0.0586, "step": 23520 }, { "epoch": 0.7954968051658271, "grad_norm": 0.2114914208650589, "learning_rate": 9.875418901996027e-05, "loss": 0.0632, "step": 23530 }, { "epoch": 0.7958348828560803, "grad_norm": 0.30599960684776306, "learning_rate": 9.875235435021007e-05, "loss": 0.0683, "step": 23540 }, { "epoch": 0.7961729605463336, "grad_norm": 0.2808926999568939, "learning_rate": 9.875051834758878e-05, "loss": 0.056, "step": 23550 }, { "epoch": 0.7965110382365868, "grad_norm": 0.2537619471549988, "learning_rate": 9.874868101214664e-05, "loss": 0.0634, "step": 23560 }, { "epoch": 0.7968491159268399, "grad_norm": 0.41742706298828125, "learning_rate": 9.87468423439339e-05, "loss": 0.069, "step": 23570 }, { "epoch": 0.7971871936170932, "grad_norm": 0.3341688811779022, "learning_rate": 9.874500234300078e-05, "loss": 0.0615, "step": 23580 }, { "epoch": 0.7975252713073464, "grad_norm": 0.2688228189945221, "learning_rate": 9.874316100939763e-05, "loss": 0.0567, "step": 23590 }, { "epoch": 0.7978633489975997, "grad_norm": 0.3639688491821289, "learning_rate": 9.874131834317473e-05, "loss": 0.0596, "step": 23600 }, { "epoch": 0.7982014266878529, "grad_norm": 0.19523575901985168, "learning_rate": 9.873947434438253e-05, "loss": 0.0545, "step": 23610 }, { "epoch": 0.7985395043781061, "grad_norm": 0.26388776302337646, "learning_rate": 9.87376290130714e-05, "loss": 0.065, "step": 23620 }, { "epoch": 0.7988775820683593, "grad_norm": 0.3307257294654846, "learning_rate": 9.873578234929182e-05, "loss": 0.066, "step": 23630 }, { "epoch": 0.7992156597586125, "grad_norm": 0.2088618129491806, "learning_rate": 9.873393435309424e-05, "loss": 0.0627, "step": 23640 }, { "epoch": 0.7995537374488657, "grad_norm": 0.1708630472421646, "learning_rate": 9.873208502452921e-05, "loss": 0.0598, "step": 23650 }, { "epoch": 0.799891815139119, "grad_norm": 0.27491891384124756, "learning_rate": 9.873023436364728e-05, "loss": 0.061, "step": 23660 }, { "epoch": 0.8002298928293722, "grad_norm": 0.31214189529418945, "learning_rate": 9.872838237049907e-05, "loss": 0.058, "step": 23670 }, { "epoch": 0.8005679705196254, "grad_norm": 0.2542344927787781, "learning_rate": 9.872652904513516e-05, "loss": 0.0625, "step": 23680 }, { "epoch": 0.8009060482098787, "grad_norm": 0.32789769768714905, "learning_rate": 9.872467438760626e-05, "loss": 0.0631, "step": 23690 }, { "epoch": 0.8012441259001318, "grad_norm": 0.273617684841156, "learning_rate": 9.87228183979631e-05, "loss": 0.0656, "step": 23700 }, { "epoch": 0.8015822035903851, "grad_norm": 0.19645313918590546, "learning_rate": 9.872096107625634e-05, "loss": 0.0598, "step": 23710 }, { "epoch": 0.8019202812806383, "grad_norm": 0.20795482397079468, "learning_rate": 9.871910242253683e-05, "loss": 0.0579, "step": 23720 }, { "epoch": 0.8022583589708915, "grad_norm": 0.38675737380981445, "learning_rate": 9.871724243685538e-05, "loss": 0.0672, "step": 23730 }, { "epoch": 0.8025964366611448, "grad_norm": 0.25429773330688477, "learning_rate": 9.87153811192628e-05, "loss": 0.061, "step": 23740 }, { "epoch": 0.802934514351398, "grad_norm": 0.3041660189628601, "learning_rate": 9.871351846981e-05, "loss": 0.0654, "step": 23750 }, { "epoch": 0.8032725920416511, "grad_norm": 0.24265149235725403, "learning_rate": 9.871165448854794e-05, "loss": 0.0553, "step": 23760 }, { "epoch": 0.8036106697319044, "grad_norm": 0.20902518928050995, "learning_rate": 9.87097891755275e-05, "loss": 0.0581, "step": 23770 }, { "epoch": 0.8039487474221576, "grad_norm": 0.21608096361160278, "learning_rate": 9.870792253079976e-05, "loss": 0.0584, "step": 23780 }, { "epoch": 0.8042868251124108, "grad_norm": 0.21230855584144592, "learning_rate": 9.87060545544157e-05, "loss": 0.0512, "step": 23790 }, { "epoch": 0.8046249028026641, "grad_norm": 0.21055693924427032, "learning_rate": 9.87041852464264e-05, "loss": 0.0601, "step": 23800 }, { "epoch": 0.8049629804929173, "grad_norm": 0.28215929865837097, "learning_rate": 9.870231460688299e-05, "loss": 0.0622, "step": 23810 }, { "epoch": 0.8053010581831704, "grad_norm": 0.19541020691394806, "learning_rate": 9.870044263583658e-05, "loss": 0.0579, "step": 23820 }, { "epoch": 0.8056391358734237, "grad_norm": 0.3205251097679138, "learning_rate": 9.869856933333838e-05, "loss": 0.055, "step": 23830 }, { "epoch": 0.8059772135636769, "grad_norm": 0.2670902907848358, "learning_rate": 9.869669469943957e-05, "loss": 0.0588, "step": 23840 }, { "epoch": 0.8063152912539302, "grad_norm": 0.22091883420944214, "learning_rate": 9.869481873419145e-05, "loss": 0.0473, "step": 23850 }, { "epoch": 0.8066533689441834, "grad_norm": 0.44757765531539917, "learning_rate": 9.869294143764524e-05, "loss": 0.0608, "step": 23860 }, { "epoch": 0.8069914466344366, "grad_norm": 0.14018277823925018, "learning_rate": 9.869106280985234e-05, "loss": 0.0521, "step": 23870 }, { "epoch": 0.8073295243246899, "grad_norm": 0.27005115151405334, "learning_rate": 9.868918285086406e-05, "loss": 0.0587, "step": 23880 }, { "epoch": 0.807667602014943, "grad_norm": 0.22903189063072205, "learning_rate": 9.86873015607318e-05, "loss": 0.0682, "step": 23890 }, { "epoch": 0.8080056797051962, "grad_norm": 0.28753048181533813, "learning_rate": 9.868541893950702e-05, "loss": 0.0638, "step": 23900 }, { "epoch": 0.8083437573954495, "grad_norm": 0.4412683844566345, "learning_rate": 9.868353498724118e-05, "loss": 0.0585, "step": 23910 }, { "epoch": 0.8086818350857027, "grad_norm": 0.4264059066772461, "learning_rate": 9.868164970398579e-05, "loss": 0.0549, "step": 23920 }, { "epoch": 0.8090199127759559, "grad_norm": 0.26938483119010925, "learning_rate": 9.867976308979236e-05, "loss": 0.0587, "step": 23930 }, { "epoch": 0.8093579904662092, "grad_norm": 0.24315369129180908, "learning_rate": 9.867787514471252e-05, "loss": 0.054, "step": 23940 }, { "epoch": 0.8096960681564623, "grad_norm": 0.3087003827095032, "learning_rate": 9.867598586879785e-05, "loss": 0.0611, "step": 23950 }, { "epoch": 0.8100341458467156, "grad_norm": 0.22197942435741425, "learning_rate": 9.86740952621e-05, "loss": 0.0603, "step": 23960 }, { "epoch": 0.8103722235369688, "grad_norm": 0.3025505542755127, "learning_rate": 9.867220332467069e-05, "loss": 0.0615, "step": 23970 }, { "epoch": 0.810710301227222, "grad_norm": 0.18988722562789917, "learning_rate": 9.867031005656163e-05, "loss": 0.0657, "step": 23980 }, { "epoch": 0.8110483789174753, "grad_norm": 0.22907397150993347, "learning_rate": 9.866841545782457e-05, "loss": 0.0616, "step": 23990 }, { "epoch": 0.8113864566077285, "grad_norm": 0.27191439270973206, "learning_rate": 9.866651952851131e-05, "loss": 0.068, "step": 24000 }, { "epoch": 0.8117245342979816, "grad_norm": 0.2028915286064148, "learning_rate": 9.86646222686737e-05, "loss": 0.0633, "step": 24010 }, { "epoch": 0.8120626119882349, "grad_norm": 0.2517000436782837, "learning_rate": 9.86627236783636e-05, "loss": 0.0614, "step": 24020 }, { "epoch": 0.8124006896784881, "grad_norm": 0.3637053072452545, "learning_rate": 9.86608237576329e-05, "loss": 0.0624, "step": 24030 }, { "epoch": 0.8127387673687413, "grad_norm": 0.28057920932769775, "learning_rate": 9.865892250653357e-05, "loss": 0.0617, "step": 24040 }, { "epoch": 0.8130768450589946, "grad_norm": 0.32507434487342834, "learning_rate": 9.865701992511757e-05, "loss": 0.0592, "step": 24050 }, { "epoch": 0.8134149227492478, "grad_norm": 0.22207075357437134, "learning_rate": 9.865511601343696e-05, "loss": 0.0624, "step": 24060 }, { "epoch": 0.813753000439501, "grad_norm": 0.3717833459377289, "learning_rate": 9.865321077154372e-05, "loss": 0.0584, "step": 24070 }, { "epoch": 0.8140910781297542, "grad_norm": 0.18679110705852509, "learning_rate": 9.865130419948998e-05, "loss": 0.0625, "step": 24080 }, { "epoch": 0.8144291558200074, "grad_norm": 0.27947521209716797, "learning_rate": 9.864939629732788e-05, "loss": 0.0625, "step": 24090 }, { "epoch": 0.8147672335102607, "grad_norm": 0.9382670521736145, "learning_rate": 9.864748706510954e-05, "loss": 0.0588, "step": 24100 }, { "epoch": 0.8151053112005139, "grad_norm": 0.25211748480796814, "learning_rate": 9.864557650288719e-05, "loss": 0.0628, "step": 24110 }, { "epoch": 0.8154433888907671, "grad_norm": 0.203110933303833, "learning_rate": 9.864366461071306e-05, "loss": 0.0514, "step": 24120 }, { "epoch": 0.8157814665810204, "grad_norm": 0.29344186186790466, "learning_rate": 9.86417513886394e-05, "loss": 0.0628, "step": 24130 }, { "epoch": 0.8161195442712735, "grad_norm": 0.4524051249027252, "learning_rate": 9.863983683671853e-05, "loss": 0.0592, "step": 24140 }, { "epoch": 0.8164576219615267, "grad_norm": 0.3399488925933838, "learning_rate": 9.863792095500282e-05, "loss": 0.0622, "step": 24150 }, { "epoch": 0.81679569965178, "grad_norm": 0.2795988917350769, "learning_rate": 9.863600374354459e-05, "loss": 0.058, "step": 24160 }, { "epoch": 0.8171337773420332, "grad_norm": 0.22894205152988434, "learning_rate": 9.86340852023963e-05, "loss": 0.0579, "step": 24170 }, { "epoch": 0.8174718550322864, "grad_norm": 0.2000303417444229, "learning_rate": 9.863216533161041e-05, "loss": 0.0583, "step": 24180 }, { "epoch": 0.8178099327225397, "grad_norm": 0.255906879901886, "learning_rate": 9.863024413123936e-05, "loss": 0.0568, "step": 24190 }, { "epoch": 0.8181480104127928, "grad_norm": 0.24393732845783234, "learning_rate": 9.862832160133572e-05, "loss": 0.0654, "step": 24200 }, { "epoch": 0.8184860881030461, "grad_norm": 0.21989810466766357, "learning_rate": 9.862639774195205e-05, "loss": 0.0592, "step": 24210 }, { "epoch": 0.8188241657932993, "grad_norm": 0.2477506697177887, "learning_rate": 9.862447255314091e-05, "loss": 0.0572, "step": 24220 }, { "epoch": 0.8191622434835525, "grad_norm": 0.33821481466293335, "learning_rate": 9.862254603495498e-05, "loss": 0.0585, "step": 24230 }, { "epoch": 0.8195003211738058, "grad_norm": 0.21724693477153778, "learning_rate": 9.86206181874469e-05, "loss": 0.0599, "step": 24240 }, { "epoch": 0.819838398864059, "grad_norm": 0.3458164632320404, "learning_rate": 9.861868901066937e-05, "loss": 0.0536, "step": 24250 }, { "epoch": 0.8201764765543121, "grad_norm": 0.23034919798374176, "learning_rate": 9.861675850467517e-05, "loss": 0.0595, "step": 24260 }, { "epoch": 0.8205145542445654, "grad_norm": 0.21840858459472656, "learning_rate": 9.861482666951705e-05, "loss": 0.0611, "step": 24270 }, { "epoch": 0.8208526319348186, "grad_norm": 0.2596433162689209, "learning_rate": 9.861289350524781e-05, "loss": 0.0631, "step": 24280 }, { "epoch": 0.8211907096250718, "grad_norm": 0.23177917301654816, "learning_rate": 9.861095901192035e-05, "loss": 0.0592, "step": 24290 }, { "epoch": 0.8215287873153251, "grad_norm": 0.23940281569957733, "learning_rate": 9.860902318958753e-05, "loss": 0.0605, "step": 24300 }, { "epoch": 0.8218668650055783, "grad_norm": 1.272692322731018, "learning_rate": 9.860708603830227e-05, "loss": 0.0677, "step": 24310 }, { "epoch": 0.8222049426958316, "grad_norm": 0.40032222867012024, "learning_rate": 9.860514755811753e-05, "loss": 0.0623, "step": 24320 }, { "epoch": 0.8225430203860847, "grad_norm": 0.26579514145851135, "learning_rate": 9.860320774908633e-05, "loss": 0.0529, "step": 24330 }, { "epoch": 0.8228810980763379, "grad_norm": 0.249376118183136, "learning_rate": 9.86012666112617e-05, "loss": 0.06, "step": 24340 }, { "epoch": 0.8232191757665912, "grad_norm": 0.3250029981136322, "learning_rate": 9.859932414469668e-05, "loss": 0.0575, "step": 24350 }, { "epoch": 0.8235572534568444, "grad_norm": 0.31661325693130493, "learning_rate": 9.859738034944441e-05, "loss": 0.0564, "step": 24360 }, { "epoch": 0.8238953311470976, "grad_norm": 0.22803062200546265, "learning_rate": 9.859543522555801e-05, "loss": 0.0572, "step": 24370 }, { "epoch": 0.8242334088373509, "grad_norm": 0.21486641466617584, "learning_rate": 9.859348877309067e-05, "loss": 0.0602, "step": 24380 }, { "epoch": 0.824571486527604, "grad_norm": 0.20969413220882416, "learning_rate": 9.85915409920956e-05, "loss": 0.0647, "step": 24390 }, { "epoch": 0.8249095642178572, "grad_norm": 0.5942002534866333, "learning_rate": 9.858959188262607e-05, "loss": 0.0641, "step": 24400 }, { "epoch": 0.8252476419081105, "grad_norm": 0.32482099533081055, "learning_rate": 9.858764144473532e-05, "loss": 0.0603, "step": 24410 }, { "epoch": 0.8255857195983637, "grad_norm": 0.329282283782959, "learning_rate": 9.858568967847673e-05, "loss": 0.0648, "step": 24420 }, { "epoch": 0.825923797288617, "grad_norm": 0.27832233905792236, "learning_rate": 9.858373658390363e-05, "loss": 0.068, "step": 24430 }, { "epoch": 0.8262618749788702, "grad_norm": 0.20042023062705994, "learning_rate": 9.858178216106943e-05, "loss": 0.0619, "step": 24440 }, { "epoch": 0.8265999526691233, "grad_norm": 0.35508015751838684, "learning_rate": 9.857982641002755e-05, "loss": 0.0686, "step": 24450 }, { "epoch": 0.8269380303593766, "grad_norm": 0.24746611714363098, "learning_rate": 9.857786933083146e-05, "loss": 0.0562, "step": 24460 }, { "epoch": 0.8272761080496298, "grad_norm": 0.321000337600708, "learning_rate": 9.857591092353469e-05, "loss": 0.0582, "step": 24470 }, { "epoch": 0.827614185739883, "grad_norm": 0.2073824405670166, "learning_rate": 9.857395118819075e-05, "loss": 0.0613, "step": 24480 }, { "epoch": 0.8279522634301363, "grad_norm": 0.21487519145011902, "learning_rate": 9.857199012485324e-05, "loss": 0.0628, "step": 24490 }, { "epoch": 0.8282903411203895, "grad_norm": 0.19838924705982208, "learning_rate": 9.857002773357577e-05, "loss": 0.062, "step": 24500 }, { "epoch": 0.8286284188106426, "grad_norm": 0.25730422139167786, "learning_rate": 9.856806401441199e-05, "loss": 0.0632, "step": 24510 }, { "epoch": 0.8289664965008959, "grad_norm": 0.2706793546676636, "learning_rate": 9.856609896741558e-05, "loss": 0.0581, "step": 24520 }, { "epoch": 0.8293045741911491, "grad_norm": 0.2664664387702942, "learning_rate": 9.856413259264025e-05, "loss": 0.0656, "step": 24530 }, { "epoch": 0.8296426518814023, "grad_norm": 0.20752565562725067, "learning_rate": 9.856216489013981e-05, "loss": 0.0565, "step": 24540 }, { "epoch": 0.8299807295716556, "grad_norm": 0.32660019397735596, "learning_rate": 9.856019585996802e-05, "loss": 0.0619, "step": 24550 }, { "epoch": 0.8303188072619088, "grad_norm": 0.4457175135612488, "learning_rate": 9.855822550217871e-05, "loss": 0.0517, "step": 24560 }, { "epoch": 0.8306568849521621, "grad_norm": 0.27439025044441223, "learning_rate": 9.855625381682576e-05, "loss": 0.0568, "step": 24570 }, { "epoch": 0.8309949626424152, "grad_norm": 0.29532015323638916, "learning_rate": 9.855428080396307e-05, "loss": 0.0614, "step": 24580 }, { "epoch": 0.8313330403326684, "grad_norm": 0.3041818141937256, "learning_rate": 9.85523064636446e-05, "loss": 0.0605, "step": 24590 }, { "epoch": 0.8316711180229217, "grad_norm": 0.23040977120399475, "learning_rate": 9.85503307959243e-05, "loss": 0.0561, "step": 24600 }, { "epoch": 0.8320091957131749, "grad_norm": 0.17898470163345337, "learning_rate": 9.854835380085619e-05, "loss": 0.062, "step": 24610 }, { "epoch": 0.8323472734034281, "grad_norm": 0.24656721949577332, "learning_rate": 9.854637547849434e-05, "loss": 0.063, "step": 24620 }, { "epoch": 0.8326853510936814, "grad_norm": 0.2003496140241623, "learning_rate": 9.854439582889281e-05, "loss": 0.0592, "step": 24630 }, { "epoch": 0.8330234287839345, "grad_norm": 0.25979360938072205, "learning_rate": 9.854241485210573e-05, "loss": 0.0541, "step": 24640 }, { "epoch": 0.8333615064741877, "grad_norm": 0.3142815828323364, "learning_rate": 9.854043254818728e-05, "loss": 0.0606, "step": 24650 }, { "epoch": 0.833699584164441, "grad_norm": 0.19770881533622742, "learning_rate": 9.853844891719163e-05, "loss": 0.0604, "step": 24660 }, { "epoch": 0.8340376618546942, "grad_norm": 0.30316492915153503, "learning_rate": 9.853646395917302e-05, "loss": 0.0584, "step": 24670 }, { "epoch": 0.8343757395449475, "grad_norm": 0.2193150818347931, "learning_rate": 9.853447767418573e-05, "loss": 0.0628, "step": 24680 }, { "epoch": 0.8347138172352007, "grad_norm": 0.2555127441883087, "learning_rate": 9.853249006228405e-05, "loss": 0.0581, "step": 24690 }, { "epoch": 0.8350518949254538, "grad_norm": 0.22133496403694153, "learning_rate": 9.853050112352231e-05, "loss": 0.0563, "step": 24700 }, { "epoch": 0.8353899726157071, "grad_norm": 0.2147582471370697, "learning_rate": 9.852851085795493e-05, "loss": 0.0618, "step": 24710 }, { "epoch": 0.8357280503059603, "grad_norm": 0.32829001545906067, "learning_rate": 9.852651926563628e-05, "loss": 0.0566, "step": 24720 }, { "epoch": 0.8360661279962135, "grad_norm": 0.2641909122467041, "learning_rate": 9.852452634662081e-05, "loss": 0.0553, "step": 24730 }, { "epoch": 0.8364042056864668, "grad_norm": 0.2792264223098755, "learning_rate": 9.852253210096304e-05, "loss": 0.062, "step": 24740 }, { "epoch": 0.83674228337672, "grad_norm": 0.3706536889076233, "learning_rate": 9.852053652871746e-05, "loss": 0.0598, "step": 24750 }, { "epoch": 0.8370803610669731, "grad_norm": 0.35106945037841797, "learning_rate": 9.851853962993864e-05, "loss": 0.0613, "step": 24760 }, { "epoch": 0.8374184387572264, "grad_norm": 0.2126329094171524, "learning_rate": 9.851654140468118e-05, "loss": 0.0573, "step": 24770 }, { "epoch": 0.8377565164474796, "grad_norm": 0.4551897943019867, "learning_rate": 9.851454185299971e-05, "loss": 0.0562, "step": 24780 }, { "epoch": 0.8380945941377328, "grad_norm": 0.30703234672546387, "learning_rate": 9.851254097494887e-05, "loss": 0.0632, "step": 24790 }, { "epoch": 0.8384326718279861, "grad_norm": 0.24690717458724976, "learning_rate": 9.851053877058341e-05, "loss": 0.0603, "step": 24800 }, { "epoch": 0.8387707495182393, "grad_norm": 0.36201390624046326, "learning_rate": 9.850853523995803e-05, "loss": 0.0583, "step": 24810 }, { "epoch": 0.8391088272084926, "grad_norm": 0.2803959846496582, "learning_rate": 9.850653038312752e-05, "loss": 0.0568, "step": 24820 }, { "epoch": 0.8394469048987457, "grad_norm": 0.21852754056453705, "learning_rate": 9.85045242001467e-05, "loss": 0.0571, "step": 24830 }, { "epoch": 0.8397849825889989, "grad_norm": 0.24518832564353943, "learning_rate": 9.85025166910704e-05, "loss": 0.0561, "step": 24840 }, { "epoch": 0.8401230602792522, "grad_norm": 0.3361785113811493, "learning_rate": 9.850050785595352e-05, "loss": 0.0559, "step": 24850 }, { "epoch": 0.8404611379695054, "grad_norm": 0.30592265725135803, "learning_rate": 9.849849769485098e-05, "loss": 0.0725, "step": 24860 }, { "epoch": 0.8407992156597586, "grad_norm": 0.22331640124320984, "learning_rate": 9.849648620781773e-05, "loss": 0.0559, "step": 24870 }, { "epoch": 0.8411372933500119, "grad_norm": 0.25810766220092773, "learning_rate": 9.849447339490877e-05, "loss": 0.0638, "step": 24880 }, { "epoch": 0.841475371040265, "grad_norm": 0.23943640291690826, "learning_rate": 9.849245925617911e-05, "loss": 0.0646, "step": 24890 }, { "epoch": 0.8418134487305182, "grad_norm": 0.21761959791183472, "learning_rate": 9.849044379168386e-05, "loss": 0.0654, "step": 24900 }, { "epoch": 0.8421515264207715, "grad_norm": 0.34006357192993164, "learning_rate": 9.848842700147808e-05, "loss": 0.0559, "step": 24910 }, { "epoch": 0.8424896041110247, "grad_norm": 0.24309217929840088, "learning_rate": 9.848640888561691e-05, "loss": 0.0554, "step": 24920 }, { "epoch": 0.842827681801278, "grad_norm": 0.26276496052742004, "learning_rate": 9.848438944415555e-05, "loss": 0.061, "step": 24930 }, { "epoch": 0.8431657594915312, "grad_norm": 0.308880090713501, "learning_rate": 9.848236867714919e-05, "loss": 0.0561, "step": 24940 }, { "epoch": 0.8435038371817843, "grad_norm": 0.31704649329185486, "learning_rate": 9.84803465846531e-05, "loss": 0.0579, "step": 24950 }, { "epoch": 0.8438419148720376, "grad_norm": 0.44275882840156555, "learning_rate": 9.847832316672254e-05, "loss": 0.0619, "step": 24960 }, { "epoch": 0.8441799925622908, "grad_norm": 0.20849944651126862, "learning_rate": 9.847629842341282e-05, "loss": 0.0602, "step": 24970 }, { "epoch": 0.844518070252544, "grad_norm": 0.27057546377182007, "learning_rate": 9.847427235477933e-05, "loss": 0.0576, "step": 24980 }, { "epoch": 0.8448561479427973, "grad_norm": 0.25424647331237793, "learning_rate": 9.847224496087744e-05, "loss": 0.0638, "step": 24990 }, { "epoch": 0.8451942256330505, "grad_norm": 0.23291540145874023, "learning_rate": 9.84702162417626e-05, "loss": 0.0631, "step": 25000 }, { "epoch": 0.8455323033233036, "grad_norm": 0.19816558063030243, "learning_rate": 9.846818619749023e-05, "loss": 0.0576, "step": 25010 }, { "epoch": 0.8458703810135569, "grad_norm": 0.48824548721313477, "learning_rate": 9.846615482811588e-05, "loss": 0.0602, "step": 25020 }, { "epoch": 0.8462084587038101, "grad_norm": 0.3543594479560852, "learning_rate": 9.846412213369504e-05, "loss": 0.0567, "step": 25030 }, { "epoch": 0.8465465363940634, "grad_norm": 0.2397403120994568, "learning_rate": 9.846208811428332e-05, "loss": 0.0612, "step": 25040 }, { "epoch": 0.8468846140843166, "grad_norm": 0.26766303181648254, "learning_rate": 9.846005276993633e-05, "loss": 0.0628, "step": 25050 }, { "epoch": 0.8472226917745698, "grad_norm": 0.33567315340042114, "learning_rate": 9.845801610070969e-05, "loss": 0.0592, "step": 25060 }, { "epoch": 0.8475607694648231, "grad_norm": 0.21736419200897217, "learning_rate": 9.84559781066591e-05, "loss": 0.0618, "step": 25070 }, { "epoch": 0.8478988471550762, "grad_norm": 0.28665122389793396, "learning_rate": 9.845393878784026e-05, "loss": 0.0621, "step": 25080 }, { "epoch": 0.8482369248453294, "grad_norm": 0.22473762929439545, "learning_rate": 9.845189814430895e-05, "loss": 0.0522, "step": 25090 }, { "epoch": 0.8485750025355827, "grad_norm": 0.23540428280830383, "learning_rate": 9.844985617612095e-05, "loss": 0.0583, "step": 25100 }, { "epoch": 0.8489130802258359, "grad_norm": 0.19933216273784637, "learning_rate": 9.844781288333207e-05, "loss": 0.0602, "step": 25110 }, { "epoch": 0.8492511579160891, "grad_norm": 0.27992314100265503, "learning_rate": 9.84457682659982e-05, "loss": 0.0552, "step": 25120 }, { "epoch": 0.8495892356063424, "grad_norm": 0.19047918915748596, "learning_rate": 9.844372232417522e-05, "loss": 0.0586, "step": 25130 }, { "epoch": 0.8499273132965955, "grad_norm": 0.2133409082889557, "learning_rate": 9.844167505791907e-05, "loss": 0.061, "step": 25140 }, { "epoch": 0.8502653909868487, "grad_norm": 0.26870980858802795, "learning_rate": 9.843962646728573e-05, "loss": 0.0601, "step": 25150 }, { "epoch": 0.850603468677102, "grad_norm": 0.2382041960954666, "learning_rate": 9.843757655233119e-05, "loss": 0.0583, "step": 25160 }, { "epoch": 0.8509415463673552, "grad_norm": 0.18806658685207367, "learning_rate": 9.843552531311151e-05, "loss": 0.0577, "step": 25170 }, { "epoch": 0.8512796240576085, "grad_norm": 0.24365493655204773, "learning_rate": 9.843347274968275e-05, "loss": 0.0586, "step": 25180 }, { "epoch": 0.8516177017478617, "grad_norm": 0.34417593479156494, "learning_rate": 9.843141886210107e-05, "loss": 0.055, "step": 25190 }, { "epoch": 0.8519557794381148, "grad_norm": 0.29740333557128906, "learning_rate": 9.842936365042256e-05, "loss": 0.0594, "step": 25200 }, { "epoch": 0.8522938571283681, "grad_norm": 0.48944419622421265, "learning_rate": 9.842730711470347e-05, "loss": 0.0536, "step": 25210 }, { "epoch": 0.8526319348186213, "grad_norm": 0.3002474904060364, "learning_rate": 9.842524925499999e-05, "loss": 0.0586, "step": 25220 }, { "epoch": 0.8529700125088745, "grad_norm": 0.2608483135700226, "learning_rate": 9.842319007136838e-05, "loss": 0.0558, "step": 25230 }, { "epoch": 0.8533080901991278, "grad_norm": 0.25454461574554443, "learning_rate": 9.842112956386494e-05, "loss": 0.062, "step": 25240 }, { "epoch": 0.853646167889381, "grad_norm": 0.26344335079193115, "learning_rate": 9.841906773254602e-05, "loss": 0.0673, "step": 25250 }, { "epoch": 0.8539842455796341, "grad_norm": 0.20137864351272583, "learning_rate": 9.841700457746799e-05, "loss": 0.0658, "step": 25260 }, { "epoch": 0.8543223232698874, "grad_norm": 0.17586755752563477, "learning_rate": 9.841494009868722e-05, "loss": 0.0617, "step": 25270 }, { "epoch": 0.8546604009601406, "grad_norm": 0.21927762031555176, "learning_rate": 9.841287429626018e-05, "loss": 0.0631, "step": 25280 }, { "epoch": 0.8549984786503939, "grad_norm": 0.33049094676971436, "learning_rate": 9.841080717024335e-05, "loss": 0.0619, "step": 25290 }, { "epoch": 0.8553365563406471, "grad_norm": 0.25674742460250854, "learning_rate": 9.840873872069326e-05, "loss": 0.0642, "step": 25300 }, { "epoch": 0.8556746340309003, "grad_norm": 0.3737248480319977, "learning_rate": 9.840666894766642e-05, "loss": 0.056, "step": 25310 }, { "epoch": 0.8560127117211536, "grad_norm": 0.27586448192596436, "learning_rate": 9.840459785121943e-05, "loss": 0.0593, "step": 25320 }, { "epoch": 0.8563507894114067, "grad_norm": 0.2613130211830139, "learning_rate": 9.840252543140891e-05, "loss": 0.0615, "step": 25330 }, { "epoch": 0.8566888671016599, "grad_norm": 0.5564700961112976, "learning_rate": 9.840045168829155e-05, "loss": 0.0559, "step": 25340 }, { "epoch": 0.8570269447919132, "grad_norm": 0.36341720819473267, "learning_rate": 9.839837662192401e-05, "loss": 0.0641, "step": 25350 }, { "epoch": 0.8573650224821664, "grad_norm": 0.29337409138679504, "learning_rate": 9.839630023236306e-05, "loss": 0.0555, "step": 25360 }, { "epoch": 0.8577031001724196, "grad_norm": 0.3263542056083679, "learning_rate": 9.839422251966541e-05, "loss": 0.0629, "step": 25370 }, { "epoch": 0.8580411778626729, "grad_norm": 0.3463417887687683, "learning_rate": 9.839214348388793e-05, "loss": 0.0621, "step": 25380 }, { "epoch": 0.858379255552926, "grad_norm": 0.24616876244544983, "learning_rate": 9.83900631250874e-05, "loss": 0.0575, "step": 25390 }, { "epoch": 0.8587173332431793, "grad_norm": 0.21841245889663696, "learning_rate": 9.838798144332074e-05, "loss": 0.0566, "step": 25400 }, { "epoch": 0.8590554109334325, "grad_norm": 0.26289501786231995, "learning_rate": 9.838589843864484e-05, "loss": 0.0676, "step": 25410 }, { "epoch": 0.8593934886236857, "grad_norm": 0.2221917062997818, "learning_rate": 9.838381411111665e-05, "loss": 0.0606, "step": 25420 }, { "epoch": 0.859731566313939, "grad_norm": 0.21839649975299835, "learning_rate": 9.838172846079317e-05, "loss": 0.0644, "step": 25430 }, { "epoch": 0.8600696440041922, "grad_norm": 0.2361256182193756, "learning_rate": 9.83796414877314e-05, "loss": 0.059, "step": 25440 }, { "epoch": 0.8604077216944453, "grad_norm": 0.17451229691505432, "learning_rate": 9.837755319198843e-05, "loss": 0.0585, "step": 25450 }, { "epoch": 0.8607457993846986, "grad_norm": 0.34418582916259766, "learning_rate": 9.837546357362131e-05, "loss": 0.0622, "step": 25460 }, { "epoch": 0.8610838770749518, "grad_norm": 0.28428196907043457, "learning_rate": 9.837337263268717e-05, "loss": 0.0588, "step": 25470 }, { "epoch": 0.861421954765205, "grad_norm": 0.5063326954841614, "learning_rate": 9.837128036924324e-05, "loss": 0.0657, "step": 25480 }, { "epoch": 0.8617600324554583, "grad_norm": 0.26283031702041626, "learning_rate": 9.836918678334665e-05, "loss": 0.0607, "step": 25490 }, { "epoch": 0.8620981101457115, "grad_norm": 0.2708762288093567, "learning_rate": 9.836709187505466e-05, "loss": 0.0594, "step": 25500 }, { "epoch": 0.8624361878359647, "grad_norm": 0.24403247237205505, "learning_rate": 9.836499564442457e-05, "loss": 0.058, "step": 25510 }, { "epoch": 0.8627742655262179, "grad_norm": 0.6622843742370605, "learning_rate": 9.836289809151365e-05, "loss": 0.0646, "step": 25520 }, { "epoch": 0.8631123432164711, "grad_norm": 0.2616939842700958, "learning_rate": 9.836079921637926e-05, "loss": 0.0621, "step": 25530 }, { "epoch": 0.8634504209067244, "grad_norm": 0.23093046247959137, "learning_rate": 9.835869901907881e-05, "loss": 0.0567, "step": 25540 }, { "epoch": 0.8637884985969776, "grad_norm": 0.21277345716953278, "learning_rate": 9.835659749966967e-05, "loss": 0.0602, "step": 25550 }, { "epoch": 0.8641265762872308, "grad_norm": 0.313784658908844, "learning_rate": 9.835449465820935e-05, "loss": 0.0617, "step": 25560 }, { "epoch": 0.8644646539774841, "grad_norm": 0.28680405020713806, "learning_rate": 9.83523904947553e-05, "loss": 0.0579, "step": 25570 }, { "epoch": 0.8648027316677372, "grad_norm": 0.7300626635551453, "learning_rate": 9.835028500936505e-05, "loss": 0.0575, "step": 25580 }, { "epoch": 0.8651408093579904, "grad_norm": 0.21337643265724182, "learning_rate": 9.834817820209619e-05, "loss": 0.0609, "step": 25590 }, { "epoch": 0.8654788870482437, "grad_norm": 0.4364396035671234, "learning_rate": 9.834607007300629e-05, "loss": 0.056, "step": 25600 }, { "epoch": 0.8658169647384969, "grad_norm": 0.18770912289619446, "learning_rate": 9.8343960622153e-05, "loss": 0.0533, "step": 25610 }, { "epoch": 0.8661550424287501, "grad_norm": 0.3882684111595154, "learning_rate": 9.834184984959398e-05, "loss": 0.0569, "step": 25620 }, { "epoch": 0.8664931201190034, "grad_norm": 0.21112430095672607, "learning_rate": 9.833973775538697e-05, "loss": 0.0597, "step": 25630 }, { "epoch": 0.8668311978092565, "grad_norm": 0.2261687070131302, "learning_rate": 9.833762433958966e-05, "loss": 0.0609, "step": 25640 }, { "epoch": 0.8671692754995098, "grad_norm": 0.32324981689453125, "learning_rate": 9.833550960225989e-05, "loss": 0.0522, "step": 25650 }, { "epoch": 0.867507353189763, "grad_norm": 0.2671915888786316, "learning_rate": 9.833339354345541e-05, "loss": 0.056, "step": 25660 }, { "epoch": 0.8678454308800162, "grad_norm": 0.2792285084724426, "learning_rate": 9.833127616323413e-05, "loss": 0.0601, "step": 25670 }, { "epoch": 0.8681835085702695, "grad_norm": 0.3118078410625458, "learning_rate": 9.832915746165392e-05, "loss": 0.0619, "step": 25680 }, { "epoch": 0.8685215862605227, "grad_norm": 0.28038185834884644, "learning_rate": 9.83270374387727e-05, "loss": 0.0567, "step": 25690 }, { "epoch": 0.8688596639507759, "grad_norm": 0.2512909471988678, "learning_rate": 9.832491609464843e-05, "loss": 0.0647, "step": 25700 }, { "epoch": 0.8691977416410291, "grad_norm": 0.27313822507858276, "learning_rate": 9.83227934293391e-05, "loss": 0.0589, "step": 25710 }, { "epoch": 0.8695358193312823, "grad_norm": 0.26744794845581055, "learning_rate": 9.832066944290277e-05, "loss": 0.0635, "step": 25720 }, { "epoch": 0.8698738970215355, "grad_norm": 0.19256709516048431, "learning_rate": 9.831854413539747e-05, "loss": 0.0586, "step": 25730 }, { "epoch": 0.8702119747117888, "grad_norm": 0.2536051571369171, "learning_rate": 9.831641750688135e-05, "loss": 0.063, "step": 25740 }, { "epoch": 0.870550052402042, "grad_norm": 0.21618026494979858, "learning_rate": 9.83142895574125e-05, "loss": 0.0561, "step": 25750 }, { "epoch": 0.8708881300922952, "grad_norm": 0.18769283592700958, "learning_rate": 9.831216028704914e-05, "loss": 0.0629, "step": 25760 }, { "epoch": 0.8712262077825484, "grad_norm": 0.28206315636634827, "learning_rate": 9.831002969584946e-05, "loss": 0.0643, "step": 25770 }, { "epoch": 0.8715642854728016, "grad_norm": 0.22803601622581482, "learning_rate": 9.830789778387173e-05, "loss": 0.0549, "step": 25780 }, { "epoch": 0.8719023631630549, "grad_norm": 0.28694993257522583, "learning_rate": 9.830576455117422e-05, "loss": 0.062, "step": 25790 }, { "epoch": 0.8722404408533081, "grad_norm": 0.2464636117219925, "learning_rate": 9.830362999781524e-05, "loss": 0.0583, "step": 25800 }, { "epoch": 0.8725785185435613, "grad_norm": 0.2044299691915512, "learning_rate": 9.830149412385317e-05, "loss": 0.0656, "step": 25810 }, { "epoch": 0.8729165962338146, "grad_norm": 0.20389088988304138, "learning_rate": 9.829935692934639e-05, "loss": 0.0556, "step": 25820 }, { "epoch": 0.8732546739240677, "grad_norm": 0.23761633038520813, "learning_rate": 9.829721841435336e-05, "loss": 0.0605, "step": 25830 }, { "epoch": 0.8735927516143209, "grad_norm": 0.20116448402404785, "learning_rate": 9.829507857893251e-05, "loss": 0.0552, "step": 25840 }, { "epoch": 0.8739308293045742, "grad_norm": 0.39445698261260986, "learning_rate": 9.829293742314237e-05, "loss": 0.0597, "step": 25850 }, { "epoch": 0.8742689069948274, "grad_norm": 0.2516215741634369, "learning_rate": 9.829079494704143e-05, "loss": 0.0566, "step": 25860 }, { "epoch": 0.8746069846850806, "grad_norm": 0.3093113899230957, "learning_rate": 9.828865115068833e-05, "loss": 0.0593, "step": 25870 }, { "epoch": 0.8749450623753339, "grad_norm": 0.23711763322353363, "learning_rate": 9.828650603414162e-05, "loss": 0.0623, "step": 25880 }, { "epoch": 0.875283140065587, "grad_norm": 0.33091703057289124, "learning_rate": 9.828435959746e-05, "loss": 0.0564, "step": 25890 }, { "epoch": 0.8756212177558403, "grad_norm": 0.20788942277431488, "learning_rate": 9.828221184070212e-05, "loss": 0.0589, "step": 25900 }, { "epoch": 0.8759592954460935, "grad_norm": 0.24027495086193085, "learning_rate": 9.828006276392669e-05, "loss": 0.0583, "step": 25910 }, { "epoch": 0.8762973731363467, "grad_norm": 0.2285684496164322, "learning_rate": 9.82779123671925e-05, "loss": 0.0643, "step": 25920 }, { "epoch": 0.8766354508266, "grad_norm": 0.2710198760032654, "learning_rate": 9.827576065055832e-05, "loss": 0.0597, "step": 25930 }, { "epoch": 0.8769735285168532, "grad_norm": 0.364509642124176, "learning_rate": 9.8273607614083e-05, "loss": 0.0612, "step": 25940 }, { "epoch": 0.8773116062071064, "grad_norm": 0.24999991059303284, "learning_rate": 9.827145325782535e-05, "loss": 0.0624, "step": 25950 }, { "epoch": 0.8776496838973596, "grad_norm": 0.2415352761745453, "learning_rate": 9.826929758184432e-05, "loss": 0.0617, "step": 25960 }, { "epoch": 0.8779877615876128, "grad_norm": 0.3660169839859009, "learning_rate": 9.826714058619882e-05, "loss": 0.0602, "step": 25970 }, { "epoch": 0.878325839277866, "grad_norm": 0.2862337827682495, "learning_rate": 9.826498227094784e-05, "loss": 0.0547, "step": 25980 }, { "epoch": 0.8786639169681193, "grad_norm": 0.22786769270896912, "learning_rate": 9.826282263615036e-05, "loss": 0.0646, "step": 25990 }, { "epoch": 0.8790019946583725, "grad_norm": 0.2668466866016388, "learning_rate": 9.826066168186547e-05, "loss": 0.0569, "step": 26000 }, { "epoch": 0.8793400723486258, "grad_norm": 0.2703624665737152, "learning_rate": 9.82584994081522e-05, "loss": 0.0603, "step": 26010 }, { "epoch": 0.879678150038879, "grad_norm": 0.3644154667854309, "learning_rate": 9.825633581506968e-05, "loss": 0.0613, "step": 26020 }, { "epoch": 0.8800162277291321, "grad_norm": 0.20964692533016205, "learning_rate": 9.825417090267707e-05, "loss": 0.0605, "step": 26030 }, { "epoch": 0.8803543054193854, "grad_norm": 0.21143034100532532, "learning_rate": 9.825200467103357e-05, "loss": 0.0584, "step": 26040 }, { "epoch": 0.8806923831096386, "grad_norm": 0.22004009783267975, "learning_rate": 9.824983712019838e-05, "loss": 0.0546, "step": 26050 }, { "epoch": 0.8810304607998918, "grad_norm": 0.18627724051475525, "learning_rate": 9.824766825023077e-05, "loss": 0.0579, "step": 26060 }, { "epoch": 0.8813685384901451, "grad_norm": 0.2281564325094223, "learning_rate": 9.824549806119005e-05, "loss": 0.0574, "step": 26070 }, { "epoch": 0.8817066161803983, "grad_norm": 0.14801743626594543, "learning_rate": 9.824332655313553e-05, "loss": 0.0561, "step": 26080 }, { "epoch": 0.8820446938706514, "grad_norm": 2.6431331634521484, "learning_rate": 9.824115372612657e-05, "loss": 0.0585, "step": 26090 }, { "epoch": 0.8823827715609047, "grad_norm": 0.18148083984851837, "learning_rate": 9.82389795802226e-05, "loss": 0.0571, "step": 26100 }, { "epoch": 0.8827208492511579, "grad_norm": 0.24951894581317902, "learning_rate": 9.823680411548306e-05, "loss": 0.06, "step": 26110 }, { "epoch": 0.8830589269414111, "grad_norm": 0.2518949806690216, "learning_rate": 9.823462733196741e-05, "loss": 0.0562, "step": 26120 }, { "epoch": 0.8833970046316644, "grad_norm": 0.31193721294403076, "learning_rate": 9.823244922973518e-05, "loss": 0.0569, "step": 26130 }, { "epoch": 0.8837350823219176, "grad_norm": 0.2817668318748474, "learning_rate": 9.823026980884589e-05, "loss": 0.0555, "step": 26140 }, { "epoch": 0.8840731600121708, "grad_norm": 0.25491851568222046, "learning_rate": 9.822808906935914e-05, "loss": 0.0655, "step": 26150 }, { "epoch": 0.884411237702424, "grad_norm": 0.30235567688941956, "learning_rate": 9.822590701133457e-05, "loss": 0.0589, "step": 26160 }, { "epoch": 0.8847493153926772, "grad_norm": 0.3204825520515442, "learning_rate": 9.822372363483181e-05, "loss": 0.0653, "step": 26170 }, { "epoch": 0.8850873930829305, "grad_norm": 0.2442537546157837, "learning_rate": 9.822153893991057e-05, "loss": 0.0611, "step": 26180 }, { "epoch": 0.8854254707731837, "grad_norm": 0.3772072196006775, "learning_rate": 9.821935292663056e-05, "loss": 0.0565, "step": 26190 }, { "epoch": 0.8857635484634369, "grad_norm": 0.21102623641490936, "learning_rate": 9.821716559505157e-05, "loss": 0.0636, "step": 26200 }, { "epoch": 0.8861016261536901, "grad_norm": 0.2333020567893982, "learning_rate": 9.821497694523339e-05, "loss": 0.0588, "step": 26210 }, { "epoch": 0.8864397038439433, "grad_norm": 0.2256849706172943, "learning_rate": 9.821278697723583e-05, "loss": 0.0595, "step": 26220 }, { "epoch": 0.8867777815341965, "grad_norm": 0.27719685435295105, "learning_rate": 9.821059569111881e-05, "loss": 0.0625, "step": 26230 }, { "epoch": 0.8871158592244498, "grad_norm": 0.28092190623283386, "learning_rate": 9.820840308694221e-05, "loss": 0.0569, "step": 26240 }, { "epoch": 0.887453936914703, "grad_norm": 0.4472864270210266, "learning_rate": 9.820620916476597e-05, "loss": 0.0598, "step": 26250 }, { "epoch": 0.8877920146049563, "grad_norm": 0.4634694755077362, "learning_rate": 9.820401392465011e-05, "loss": 0.066, "step": 26260 }, { "epoch": 0.8881300922952095, "grad_norm": 0.2330649197101593, "learning_rate": 9.820181736665459e-05, "loss": 0.0643, "step": 26270 }, { "epoch": 0.8884681699854626, "grad_norm": 0.24349769949913025, "learning_rate": 9.81996194908395e-05, "loss": 0.0602, "step": 26280 }, { "epoch": 0.8888062476757159, "grad_norm": 0.39191603660583496, "learning_rate": 9.819742029726492e-05, "loss": 0.0606, "step": 26290 }, { "epoch": 0.8891443253659691, "grad_norm": 0.27045050263404846, "learning_rate": 9.819521978599098e-05, "loss": 0.0558, "step": 26300 }, { "epoch": 0.8894824030562223, "grad_norm": 0.30171769857406616, "learning_rate": 9.819301795707783e-05, "loss": 0.0685, "step": 26310 }, { "epoch": 0.8898204807464756, "grad_norm": 0.25252220034599304, "learning_rate": 9.819081481058567e-05, "loss": 0.0635, "step": 26320 }, { "epoch": 0.8901585584367288, "grad_norm": 0.5659950375556946, "learning_rate": 9.818861034657476e-05, "loss": 0.0602, "step": 26330 }, { "epoch": 0.8904966361269819, "grad_norm": 0.43113401532173157, "learning_rate": 9.818640456510533e-05, "loss": 0.065, "step": 26340 }, { "epoch": 0.8908347138172352, "grad_norm": 0.25412678718566895, "learning_rate": 9.81841974662377e-05, "loss": 0.0636, "step": 26350 }, { "epoch": 0.8911727915074884, "grad_norm": 0.32657575607299805, "learning_rate": 9.818198905003222e-05, "loss": 0.0605, "step": 26360 }, { "epoch": 0.8915108691977417, "grad_norm": 0.2967219352722168, "learning_rate": 9.817977931654926e-05, "loss": 0.0597, "step": 26370 }, { "epoch": 0.8918489468879949, "grad_norm": 0.6129677891731262, "learning_rate": 9.817756826584923e-05, "loss": 0.0603, "step": 26380 }, { "epoch": 0.8921870245782481, "grad_norm": 0.36960870027542114, "learning_rate": 9.817535589799258e-05, "loss": 0.062, "step": 26390 }, { "epoch": 0.8925251022685013, "grad_norm": 0.3313728868961334, "learning_rate": 9.817314221303982e-05, "loss": 0.0633, "step": 26400 }, { "epoch": 0.8928631799587545, "grad_norm": 0.23027485609054565, "learning_rate": 9.817092721105143e-05, "loss": 0.0597, "step": 26410 }, { "epoch": 0.8932012576490077, "grad_norm": 0.25532689690589905, "learning_rate": 9.8168710892088e-05, "loss": 0.0599, "step": 26420 }, { "epoch": 0.893539335339261, "grad_norm": 0.2916722595691681, "learning_rate": 9.81664932562101e-05, "loss": 0.0598, "step": 26430 }, { "epoch": 0.8938774130295142, "grad_norm": 0.30110791325569153, "learning_rate": 9.816427430347836e-05, "loss": 0.0633, "step": 26440 }, { "epoch": 0.8942154907197674, "grad_norm": 0.19469264149665833, "learning_rate": 9.816205403395347e-05, "loss": 0.0565, "step": 26450 }, { "epoch": 0.8945535684100206, "grad_norm": 0.25441882014274597, "learning_rate": 9.815983244769611e-05, "loss": 0.0642, "step": 26460 }, { "epoch": 0.8948916461002738, "grad_norm": 0.21226835250854492, "learning_rate": 9.815760954476703e-05, "loss": 0.0654, "step": 26470 }, { "epoch": 0.895229723790527, "grad_norm": 0.17893917858600616, "learning_rate": 9.815538532522699e-05, "loss": 0.0591, "step": 26480 }, { "epoch": 0.8955678014807803, "grad_norm": 0.32474932074546814, "learning_rate": 9.81531597891368e-05, "loss": 0.0606, "step": 26490 }, { "epoch": 0.8959058791710335, "grad_norm": 0.398303359746933, "learning_rate": 9.815093293655733e-05, "loss": 0.0575, "step": 26500 }, { "epoch": 0.8962439568612868, "grad_norm": 0.2883043587207794, "learning_rate": 9.814870476754942e-05, "loss": 0.066, "step": 26510 }, { "epoch": 0.89658203455154, "grad_norm": 0.3132454454898834, "learning_rate": 9.814647528217403e-05, "loss": 0.0591, "step": 26520 }, { "epoch": 0.8969201122417931, "grad_norm": 0.1983809620141983, "learning_rate": 9.814424448049209e-05, "loss": 0.0569, "step": 26530 }, { "epoch": 0.8972581899320464, "grad_norm": 0.21201570332050323, "learning_rate": 9.81420123625646e-05, "loss": 0.0649, "step": 26540 }, { "epoch": 0.8975962676222996, "grad_norm": 0.29221251606941223, "learning_rate": 9.813977892845256e-05, "loss": 0.0592, "step": 26550 }, { "epoch": 0.8979343453125528, "grad_norm": 0.37110018730163574, "learning_rate": 9.813754417821705e-05, "loss": 0.0518, "step": 26560 }, { "epoch": 0.8982724230028061, "grad_norm": 0.39792129397392273, "learning_rate": 9.813530811191918e-05, "loss": 0.0617, "step": 26570 }, { "epoch": 0.8986105006930593, "grad_norm": 0.3823128044605255, "learning_rate": 9.813307072962007e-05, "loss": 0.0612, "step": 26580 }, { "epoch": 0.8989485783833124, "grad_norm": 0.19045349955558777, "learning_rate": 9.813083203138088e-05, "loss": 0.0548, "step": 26590 }, { "epoch": 0.8992866560735657, "grad_norm": 0.19368867576122284, "learning_rate": 9.812859201726283e-05, "loss": 0.0593, "step": 26600 }, { "epoch": 0.8996247337638189, "grad_norm": 0.27829688787460327, "learning_rate": 9.812635068732715e-05, "loss": 0.0653, "step": 26610 }, { "epoch": 0.8999628114540722, "grad_norm": 0.3147909343242645, "learning_rate": 9.812410804163513e-05, "loss": 0.0623, "step": 26620 }, { "epoch": 0.9003008891443254, "grad_norm": 0.2586056590080261, "learning_rate": 9.812186408024806e-05, "loss": 0.0583, "step": 26630 }, { "epoch": 0.9006389668345786, "grad_norm": 0.279592365026474, "learning_rate": 9.811961880322731e-05, "loss": 0.058, "step": 26640 }, { "epoch": 0.9009770445248318, "grad_norm": 0.3089207708835602, "learning_rate": 9.811737221063429e-05, "loss": 0.0565, "step": 26650 }, { "epoch": 0.901315122215085, "grad_norm": 0.2220047414302826, "learning_rate": 9.811512430253036e-05, "loss": 0.0592, "step": 26660 }, { "epoch": 0.9016531999053382, "grad_norm": 0.18743112683296204, "learning_rate": 9.811287507897701e-05, "loss": 0.0559, "step": 26670 }, { "epoch": 0.9019912775955915, "grad_norm": 0.35047248005867004, "learning_rate": 9.811062454003572e-05, "loss": 0.061, "step": 26680 }, { "epoch": 0.9023293552858447, "grad_norm": 0.2534694969654083, "learning_rate": 9.810837268576805e-05, "loss": 0.0562, "step": 26690 }, { "epoch": 0.9026674329760979, "grad_norm": 0.26988711953163147, "learning_rate": 9.810611951623554e-05, "loss": 0.0603, "step": 26700 }, { "epoch": 0.9030055106663512, "grad_norm": 0.2802980840206146, "learning_rate": 9.810386503149979e-05, "loss": 0.0568, "step": 26710 }, { "epoch": 0.9033435883566043, "grad_norm": 0.19396154582500458, "learning_rate": 9.810160923162245e-05, "loss": 0.0637, "step": 26720 }, { "epoch": 0.9036816660468575, "grad_norm": 0.33554577827453613, "learning_rate": 9.809935211666518e-05, "loss": 0.0638, "step": 26730 }, { "epoch": 0.9040197437371108, "grad_norm": 0.18222591280937195, "learning_rate": 9.809709368668969e-05, "loss": 0.0565, "step": 26740 }, { "epoch": 0.904357821427364, "grad_norm": 0.28939539194107056, "learning_rate": 9.809483394175773e-05, "loss": 0.0611, "step": 26750 }, { "epoch": 0.9046958991176173, "grad_norm": 0.4651142954826355, "learning_rate": 9.809257288193107e-05, "loss": 0.0637, "step": 26760 }, { "epoch": 0.9050339768078705, "grad_norm": 0.21554584801197052, "learning_rate": 9.809031050727155e-05, "loss": 0.0604, "step": 26770 }, { "epoch": 0.9053720544981236, "grad_norm": 0.2504948675632477, "learning_rate": 9.8088046817841e-05, "loss": 0.0569, "step": 26780 }, { "epoch": 0.9057101321883769, "grad_norm": 0.329632043838501, "learning_rate": 9.808578181370131e-05, "loss": 0.0612, "step": 26790 }, { "epoch": 0.9060482098786301, "grad_norm": 0.349531352519989, "learning_rate": 9.808351549491443e-05, "loss": 0.0591, "step": 26800 }, { "epoch": 0.9063862875688833, "grad_norm": 0.2538350820541382, "learning_rate": 9.808124786154228e-05, "loss": 0.0622, "step": 26810 }, { "epoch": 0.9067243652591366, "grad_norm": 0.40314528346061707, "learning_rate": 9.807897891364688e-05, "loss": 0.0671, "step": 26820 }, { "epoch": 0.9070624429493898, "grad_norm": 0.27456653118133545, "learning_rate": 9.807670865129026e-05, "loss": 0.057, "step": 26830 }, { "epoch": 0.9074005206396429, "grad_norm": 0.31979742646217346, "learning_rate": 9.80744370745345e-05, "loss": 0.0598, "step": 26840 }, { "epoch": 0.9077385983298962, "grad_norm": 0.18548358976840973, "learning_rate": 9.807216418344167e-05, "loss": 0.0581, "step": 26850 }, { "epoch": 0.9080766760201494, "grad_norm": 0.31388044357299805, "learning_rate": 9.806988997807396e-05, "loss": 0.0591, "step": 26860 }, { "epoch": 0.9084147537104027, "grad_norm": 0.2966422140598297, "learning_rate": 9.806761445849348e-05, "loss": 0.0553, "step": 26870 }, { "epoch": 0.9087528314006559, "grad_norm": 0.23237378895282745, "learning_rate": 9.806533762476252e-05, "loss": 0.0583, "step": 26880 }, { "epoch": 0.9090909090909091, "grad_norm": 0.26291951537132263, "learning_rate": 9.806305947694325e-05, "loss": 0.0624, "step": 26890 }, { "epoch": 0.9094289867811624, "grad_norm": 0.2814169228076935, "learning_rate": 9.806078001509801e-05, "loss": 0.0601, "step": 26900 }, { "epoch": 0.9097670644714155, "grad_norm": 0.22373446822166443, "learning_rate": 9.80584992392891e-05, "loss": 0.0588, "step": 26910 }, { "epoch": 0.9101051421616687, "grad_norm": 0.39022698998451233, "learning_rate": 9.805621714957886e-05, "loss": 0.0582, "step": 26920 }, { "epoch": 0.910443219851922, "grad_norm": 0.2857377529144287, "learning_rate": 9.805393374602972e-05, "loss": 0.0604, "step": 26930 }, { "epoch": 0.9107812975421752, "grad_norm": 0.30799567699432373, "learning_rate": 9.805164902870406e-05, "loss": 0.059, "step": 26940 }, { "epoch": 0.9111193752324284, "grad_norm": 0.18003366887569427, "learning_rate": 9.804936299766438e-05, "loss": 0.0566, "step": 26950 }, { "epoch": 0.9114574529226817, "grad_norm": 0.28295251727104187, "learning_rate": 9.804707565297318e-05, "loss": 0.0624, "step": 26960 }, { "epoch": 0.9117955306129348, "grad_norm": 0.24338871240615845, "learning_rate": 9.804478699469296e-05, "loss": 0.0589, "step": 26970 }, { "epoch": 0.9121336083031881, "grad_norm": 0.35386645793914795, "learning_rate": 9.804249702288633e-05, "loss": 0.0645, "step": 26980 }, { "epoch": 0.9124716859934413, "grad_norm": 0.17457586526870728, "learning_rate": 9.804020573761587e-05, "loss": 0.0605, "step": 26990 }, { "epoch": 0.9128097636836945, "grad_norm": 0.2100502848625183, "learning_rate": 9.803791313894423e-05, "loss": 0.0561, "step": 27000 }, { "epoch": 0.9131478413739478, "grad_norm": 0.26955646276474, "learning_rate": 9.80356192269341e-05, "loss": 0.0594, "step": 27010 }, { "epoch": 0.913485919064201, "grad_norm": 0.2504948377609253, "learning_rate": 9.803332400164817e-05, "loss": 0.0585, "step": 27020 }, { "epoch": 0.9138239967544541, "grad_norm": 0.20446385443210602, "learning_rate": 9.803102746314922e-05, "loss": 0.0594, "step": 27030 }, { "epoch": 0.9141620744447074, "grad_norm": 0.22462798655033112, "learning_rate": 9.802872961150003e-05, "loss": 0.0548, "step": 27040 }, { "epoch": 0.9145001521349606, "grad_norm": 0.14790566265583038, "learning_rate": 9.802643044676341e-05, "loss": 0.0565, "step": 27050 }, { "epoch": 0.9148382298252138, "grad_norm": 0.22676268219947815, "learning_rate": 9.802412996900223e-05, "loss": 0.0552, "step": 27060 }, { "epoch": 0.9151763075154671, "grad_norm": 0.21869394183158875, "learning_rate": 9.802182817827936e-05, "loss": 0.0609, "step": 27070 }, { "epoch": 0.9155143852057203, "grad_norm": 0.24581731855869293, "learning_rate": 9.801952507465777e-05, "loss": 0.0582, "step": 27080 }, { "epoch": 0.9158524628959734, "grad_norm": 0.22636982798576355, "learning_rate": 9.80172206582004e-05, "loss": 0.0566, "step": 27090 }, { "epoch": 0.9161905405862267, "grad_norm": 0.21881476044654846, "learning_rate": 9.801491492897025e-05, "loss": 0.0594, "step": 27100 }, { "epoch": 0.9165286182764799, "grad_norm": 0.2287275195121765, "learning_rate": 9.801260788703037e-05, "loss": 0.0563, "step": 27110 }, { "epoch": 0.9168666959667332, "grad_norm": 0.3160817623138428, "learning_rate": 9.801029953244383e-05, "loss": 0.0631, "step": 27120 }, { "epoch": 0.9172047736569864, "grad_norm": 0.3246334493160248, "learning_rate": 9.800798986527372e-05, "loss": 0.0549, "step": 27130 }, { "epoch": 0.9175428513472396, "grad_norm": 0.27594730257987976, "learning_rate": 9.800567888558321e-05, "loss": 0.0624, "step": 27140 }, { "epoch": 0.9178809290374929, "grad_norm": 0.23644289374351501, "learning_rate": 9.800336659343549e-05, "loss": 0.0575, "step": 27150 }, { "epoch": 0.918219006727746, "grad_norm": 0.2784785032272339, "learning_rate": 9.800105298889375e-05, "loss": 0.0603, "step": 27160 }, { "epoch": 0.9185570844179992, "grad_norm": 0.26253974437713623, "learning_rate": 9.799873807202124e-05, "loss": 0.0581, "step": 27170 }, { "epoch": 0.9188951621082525, "grad_norm": 0.3516509532928467, "learning_rate": 9.799642184288129e-05, "loss": 0.0616, "step": 27180 }, { "epoch": 0.9192332397985057, "grad_norm": 0.21526937186717987, "learning_rate": 9.799410430153719e-05, "loss": 0.0597, "step": 27190 }, { "epoch": 0.9195713174887589, "grad_norm": 0.27443772554397583, "learning_rate": 9.799178544805229e-05, "loss": 0.0629, "step": 27200 }, { "epoch": 0.9199093951790122, "grad_norm": 0.3294036388397217, "learning_rate": 9.798946528249003e-05, "loss": 0.0639, "step": 27210 }, { "epoch": 0.9202474728692653, "grad_norm": 0.35306161642074585, "learning_rate": 9.79871438049138e-05, "loss": 0.0562, "step": 27220 }, { "epoch": 0.9205855505595186, "grad_norm": 0.2239813208580017, "learning_rate": 9.798482101538711e-05, "loss": 0.0616, "step": 27230 }, { "epoch": 0.9209236282497718, "grad_norm": 0.22300618886947632, "learning_rate": 9.798249691397343e-05, "loss": 0.0586, "step": 27240 }, { "epoch": 0.921261705940025, "grad_norm": 0.28342491388320923, "learning_rate": 9.79801715007363e-05, "loss": 0.0568, "step": 27250 }, { "epoch": 0.9215997836302783, "grad_norm": 0.20084121823310852, "learning_rate": 9.79778447757393e-05, "loss": 0.0571, "step": 27260 }, { "epoch": 0.9219378613205315, "grad_norm": 0.26400554180145264, "learning_rate": 9.797551673904607e-05, "loss": 0.0593, "step": 27270 }, { "epoch": 0.9222759390107846, "grad_norm": 0.2587924301624298, "learning_rate": 9.797318739072023e-05, "loss": 0.0631, "step": 27280 }, { "epoch": 0.9226140167010379, "grad_norm": 0.20002113282680511, "learning_rate": 9.797085673082547e-05, "loss": 0.0599, "step": 27290 }, { "epoch": 0.9229520943912911, "grad_norm": 0.32644328474998474, "learning_rate": 9.79685247594255e-05, "loss": 0.0607, "step": 27300 }, { "epoch": 0.9232901720815443, "grad_norm": 0.3083655834197998, "learning_rate": 9.796619147658408e-05, "loss": 0.0567, "step": 27310 }, { "epoch": 0.9236282497717976, "grad_norm": 0.38587668538093567, "learning_rate": 9.796385688236503e-05, "loss": 0.0605, "step": 27320 }, { "epoch": 0.9239663274620508, "grad_norm": 0.3407217264175415, "learning_rate": 9.796152097683213e-05, "loss": 0.0633, "step": 27330 }, { "epoch": 0.924304405152304, "grad_norm": 0.45725148916244507, "learning_rate": 9.795918376004927e-05, "loss": 0.0579, "step": 27340 }, { "epoch": 0.9246424828425572, "grad_norm": 0.42436283826828003, "learning_rate": 9.795684523208034e-05, "loss": 0.0558, "step": 27350 }, { "epoch": 0.9249805605328104, "grad_norm": 0.2674354314804077, "learning_rate": 9.795450539298928e-05, "loss": 0.0631, "step": 27360 }, { "epoch": 0.9253186382230637, "grad_norm": 0.3319926857948303, "learning_rate": 9.795216424284006e-05, "loss": 0.0669, "step": 27370 }, { "epoch": 0.9256567159133169, "grad_norm": 0.2248079925775528, "learning_rate": 9.79498217816967e-05, "loss": 0.0617, "step": 27380 }, { "epoch": 0.9259947936035701, "grad_norm": 0.31562256813049316, "learning_rate": 9.79474780096232e-05, "loss": 0.0604, "step": 27390 }, { "epoch": 0.9263328712938234, "grad_norm": 0.305186927318573, "learning_rate": 9.794513292668368e-05, "loss": 0.0535, "step": 27400 }, { "epoch": 0.9266709489840765, "grad_norm": 0.2588707506656647, "learning_rate": 9.794278653294222e-05, "loss": 0.0576, "step": 27410 }, { "epoch": 0.9270090266743297, "grad_norm": 0.36680474877357483, "learning_rate": 9.794043882846302e-05, "loss": 0.055, "step": 27420 }, { "epoch": 0.927347104364583, "grad_norm": 0.2308063954114914, "learning_rate": 9.793808981331021e-05, "loss": 0.0605, "step": 27430 }, { "epoch": 0.9276851820548362, "grad_norm": 0.2668234407901764, "learning_rate": 9.793573948754802e-05, "loss": 0.0641, "step": 27440 }, { "epoch": 0.9280232597450894, "grad_norm": 0.21228472888469696, "learning_rate": 9.793338785124076e-05, "loss": 0.0607, "step": 27450 }, { "epoch": 0.9283613374353427, "grad_norm": 0.33486855030059814, "learning_rate": 9.793103490445265e-05, "loss": 0.0611, "step": 27460 }, { "epoch": 0.9286994151255958, "grad_norm": 0.3543325364589691, "learning_rate": 9.792868064724808e-05, "loss": 0.0609, "step": 27470 }, { "epoch": 0.9290374928158491, "grad_norm": 0.2944825291633606, "learning_rate": 9.792632507969135e-05, "loss": 0.0518, "step": 27480 }, { "epoch": 0.9293755705061023, "grad_norm": 0.238104909658432, "learning_rate": 9.792396820184693e-05, "loss": 0.0636, "step": 27490 }, { "epoch": 0.9297136481963555, "grad_norm": 0.22459135949611664, "learning_rate": 9.792161001377921e-05, "loss": 0.0584, "step": 27500 }, { "epoch": 0.9300517258866088, "grad_norm": 0.31562894582748413, "learning_rate": 9.791925051555269e-05, "loss": 0.057, "step": 27510 }, { "epoch": 0.930389803576862, "grad_norm": 0.3136160671710968, "learning_rate": 9.791688970723184e-05, "loss": 0.0648, "step": 27520 }, { "epoch": 0.9307278812671151, "grad_norm": 0.26126760244369507, "learning_rate": 9.791452758888125e-05, "loss": 0.0608, "step": 27530 }, { "epoch": 0.9310659589573684, "grad_norm": 0.3245609700679779, "learning_rate": 9.791216416056545e-05, "loss": 0.0598, "step": 27540 }, { "epoch": 0.9314040366476216, "grad_norm": 0.20883752405643463, "learning_rate": 9.79097994223491e-05, "loss": 0.0606, "step": 27550 }, { "epoch": 0.9317421143378748, "grad_norm": 0.22428667545318604, "learning_rate": 9.790743337429683e-05, "loss": 0.0603, "step": 27560 }, { "epoch": 0.9320801920281281, "grad_norm": 0.2111411988735199, "learning_rate": 9.790506601647334e-05, "loss": 0.0627, "step": 27570 }, { "epoch": 0.9324182697183813, "grad_norm": 0.2751721143722534, "learning_rate": 9.790269734894333e-05, "loss": 0.0581, "step": 27580 }, { "epoch": 0.9327563474086346, "grad_norm": 0.27453866600990295, "learning_rate": 9.790032737177157e-05, "loss": 0.0634, "step": 27590 }, { "epoch": 0.9330944250988877, "grad_norm": 0.3414202928543091, "learning_rate": 9.789795608502285e-05, "loss": 0.0562, "step": 27600 }, { "epoch": 0.9334325027891409, "grad_norm": 0.18103277683258057, "learning_rate": 9.789558348876201e-05, "loss": 0.0591, "step": 27610 }, { "epoch": 0.9337705804793942, "grad_norm": 0.30386826395988464, "learning_rate": 9.789320958305392e-05, "loss": 0.0635, "step": 27620 }, { "epoch": 0.9341086581696474, "grad_norm": 0.3388933539390564, "learning_rate": 9.789083436796347e-05, "loss": 0.0555, "step": 27630 }, { "epoch": 0.9344467358599006, "grad_norm": 0.3914779722690582, "learning_rate": 9.788845784355558e-05, "loss": 0.0583, "step": 27640 }, { "epoch": 0.9347848135501539, "grad_norm": 0.3730839788913727, "learning_rate": 9.788608000989528e-05, "loss": 0.0561, "step": 27650 }, { "epoch": 0.935122891240407, "grad_norm": 0.5343735814094543, "learning_rate": 9.78837008670475e-05, "loss": 0.0565, "step": 27660 }, { "epoch": 0.9354609689306602, "grad_norm": 0.25693976879119873, "learning_rate": 9.788132041507736e-05, "loss": 0.065, "step": 27670 }, { "epoch": 0.9357990466209135, "grad_norm": 0.2391936182975769, "learning_rate": 9.787893865404989e-05, "loss": 0.0648, "step": 27680 }, { "epoch": 0.9361371243111667, "grad_norm": 0.3114472031593323, "learning_rate": 9.787655558403023e-05, "loss": 0.0605, "step": 27690 }, { "epoch": 0.9364752020014199, "grad_norm": 0.3215681314468384, "learning_rate": 9.787417120508354e-05, "loss": 0.058, "step": 27700 }, { "epoch": 0.9368132796916732, "grad_norm": 0.2425711452960968, "learning_rate": 9.787178551727498e-05, "loss": 0.0592, "step": 27710 }, { "epoch": 0.9371513573819263, "grad_norm": 0.25258132815361023, "learning_rate": 9.786939852066979e-05, "loss": 0.0648, "step": 27720 }, { "epoch": 0.9374894350721796, "grad_norm": 0.28505703806877136, "learning_rate": 9.786701021533322e-05, "loss": 0.0676, "step": 27730 }, { "epoch": 0.9378275127624328, "grad_norm": 0.18032942712306976, "learning_rate": 9.786462060133059e-05, "loss": 0.0545, "step": 27740 }, { "epoch": 0.938165590452686, "grad_norm": 0.3386473059654236, "learning_rate": 9.78622296787272e-05, "loss": 0.0577, "step": 27750 }, { "epoch": 0.9385036681429393, "grad_norm": 0.21776323020458221, "learning_rate": 9.785983744758845e-05, "loss": 0.0566, "step": 27760 }, { "epoch": 0.9388417458331925, "grad_norm": 0.31865015625953674, "learning_rate": 9.78574439079797e-05, "loss": 0.0602, "step": 27770 }, { "epoch": 0.9391798235234456, "grad_norm": 0.3210708498954773, "learning_rate": 9.785504905996645e-05, "loss": 0.065, "step": 27780 }, { "epoch": 0.9395179012136989, "grad_norm": 0.21199624240398407, "learning_rate": 9.78526529036141e-05, "loss": 0.0629, "step": 27790 }, { "epoch": 0.9398559789039521, "grad_norm": 0.2916504442691803, "learning_rate": 9.785025543898822e-05, "loss": 0.0593, "step": 27800 }, { "epoch": 0.9401940565942053, "grad_norm": 0.3597277104854584, "learning_rate": 9.784785666615433e-05, "loss": 0.0546, "step": 27810 }, { "epoch": 0.9405321342844586, "grad_norm": 0.38831689953804016, "learning_rate": 9.7845456585178e-05, "loss": 0.0594, "step": 27820 }, { "epoch": 0.9408702119747118, "grad_norm": 0.2241044044494629, "learning_rate": 9.784305519612487e-05, "loss": 0.0646, "step": 27830 }, { "epoch": 0.9412082896649651, "grad_norm": 0.2697252035140991, "learning_rate": 9.784065249906059e-05, "loss": 0.0549, "step": 27840 }, { "epoch": 0.9415463673552182, "grad_norm": 0.3673935532569885, "learning_rate": 9.783824849405084e-05, "loss": 0.0535, "step": 27850 }, { "epoch": 0.9418844450454714, "grad_norm": 0.25311794877052307, "learning_rate": 9.783584318116133e-05, "loss": 0.0577, "step": 27860 }, { "epoch": 0.9422225227357247, "grad_norm": 0.17183522880077362, "learning_rate": 9.783343656045786e-05, "loss": 0.0591, "step": 27870 }, { "epoch": 0.9425606004259779, "grad_norm": 0.27661630511283875, "learning_rate": 9.78310286320062e-05, "loss": 0.0679, "step": 27880 }, { "epoch": 0.9428986781162311, "grad_norm": 0.2817843556404114, "learning_rate": 9.782861939587218e-05, "loss": 0.0563, "step": 27890 }, { "epoch": 0.9432367558064844, "grad_norm": 0.41824647784233093, "learning_rate": 9.782620885212168e-05, "loss": 0.0668, "step": 27900 }, { "epoch": 0.9435748334967375, "grad_norm": 0.27162662148475647, "learning_rate": 9.782379700082061e-05, "loss": 0.0669, "step": 27910 }, { "epoch": 0.9439129111869907, "grad_norm": 0.2675575017929077, "learning_rate": 9.782138384203487e-05, "loss": 0.0561, "step": 27920 }, { "epoch": 0.944250988877244, "grad_norm": 0.29989519715309143, "learning_rate": 9.781896937583048e-05, "loss": 0.0609, "step": 27930 }, { "epoch": 0.9445890665674972, "grad_norm": 0.2064027041196823, "learning_rate": 9.781655360227343e-05, "loss": 0.062, "step": 27940 }, { "epoch": 0.9449271442577505, "grad_norm": 0.2760336399078369, "learning_rate": 9.781413652142976e-05, "loss": 0.0572, "step": 27950 }, { "epoch": 0.9452652219480037, "grad_norm": 0.24078814685344696, "learning_rate": 9.781171813336556e-05, "loss": 0.0594, "step": 27960 }, { "epoch": 0.9456032996382568, "grad_norm": 0.26180142164230347, "learning_rate": 9.780929843814696e-05, "loss": 0.0592, "step": 27970 }, { "epoch": 0.9459413773285101, "grad_norm": 0.3001522421836853, "learning_rate": 9.78068774358401e-05, "loss": 0.0574, "step": 27980 }, { "epoch": 0.9462794550187633, "grad_norm": 0.21982334554195404, "learning_rate": 9.780445512651116e-05, "loss": 0.0628, "step": 27990 }, { "epoch": 0.9466175327090165, "grad_norm": 0.2922931909561157, "learning_rate": 9.780203151022639e-05, "loss": 0.0624, "step": 28000 }, { "epoch": 0.9469556103992698, "grad_norm": 0.19785615801811218, "learning_rate": 9.779960658705203e-05, "loss": 0.0605, "step": 28010 }, { "epoch": 0.947293688089523, "grad_norm": 0.2789371907711029, "learning_rate": 9.779718035705439e-05, "loss": 0.0639, "step": 28020 }, { "epoch": 0.9476317657797761, "grad_norm": 0.15410763025283813, "learning_rate": 9.779475282029977e-05, "loss": 0.0583, "step": 28030 }, { "epoch": 0.9479698434700294, "grad_norm": 0.20576031506061554, "learning_rate": 9.779232397685458e-05, "loss": 0.0613, "step": 28040 }, { "epoch": 0.9483079211602826, "grad_norm": 0.2959049940109253, "learning_rate": 9.778989382678523e-05, "loss": 0.055, "step": 28050 }, { "epoch": 0.9486459988505358, "grad_norm": 0.6072283387184143, "learning_rate": 9.77874623701581e-05, "loss": 0.0604, "step": 28060 }, { "epoch": 0.9489840765407891, "grad_norm": 0.23860087990760803, "learning_rate": 9.778502960703972e-05, "loss": 0.0654, "step": 28070 }, { "epoch": 0.9493221542310423, "grad_norm": 0.30548033118247986, "learning_rate": 9.778259553749658e-05, "loss": 0.0601, "step": 28080 }, { "epoch": 0.9496602319212956, "grad_norm": 0.26677367091178894, "learning_rate": 9.778016016159524e-05, "loss": 0.0533, "step": 28090 }, { "epoch": 0.9499983096115487, "grad_norm": 0.24898117780685425, "learning_rate": 9.777772347940226e-05, "loss": 0.0563, "step": 28100 }, { "epoch": 0.9503363873018019, "grad_norm": 0.3363621234893799, "learning_rate": 9.777528549098427e-05, "loss": 0.059, "step": 28110 }, { "epoch": 0.9506744649920552, "grad_norm": 0.4471493661403656, "learning_rate": 9.777284619640792e-05, "loss": 0.0592, "step": 28120 }, { "epoch": 0.9510125426823084, "grad_norm": 0.27234306931495667, "learning_rate": 9.777040559573992e-05, "loss": 0.0538, "step": 28130 }, { "epoch": 0.9513506203725616, "grad_norm": 0.2740347981452942, "learning_rate": 9.776796368904695e-05, "loss": 0.0607, "step": 28140 }, { "epoch": 0.9516886980628149, "grad_norm": 0.29221010208129883, "learning_rate": 9.776552047639581e-05, "loss": 0.0642, "step": 28150 }, { "epoch": 0.952026775753068, "grad_norm": 0.4670029282569885, "learning_rate": 9.77630759578533e-05, "loss": 0.0553, "step": 28160 }, { "epoch": 0.9523648534433212, "grad_norm": 0.24953043460845947, "learning_rate": 9.776063013348622e-05, "loss": 0.0591, "step": 28170 }, { "epoch": 0.9527029311335745, "grad_norm": 0.29538196325302124, "learning_rate": 9.775818300336146e-05, "loss": 0.0612, "step": 28180 }, { "epoch": 0.9530410088238277, "grad_norm": 0.36629754304885864, "learning_rate": 9.775573456754591e-05, "loss": 0.0565, "step": 28190 }, { "epoch": 0.953379086514081, "grad_norm": 0.1814892292022705, "learning_rate": 9.775328482610653e-05, "loss": 0.0613, "step": 28200 }, { "epoch": 0.9537171642043342, "grad_norm": 0.2809857726097107, "learning_rate": 9.775083377911028e-05, "loss": 0.0593, "step": 28210 }, { "epoch": 0.9540552418945873, "grad_norm": 0.2631179094314575, "learning_rate": 9.774838142662417e-05, "loss": 0.0562, "step": 28220 }, { "epoch": 0.9543933195848406, "grad_norm": 0.38152986764907837, "learning_rate": 9.774592776871525e-05, "loss": 0.064, "step": 28230 }, { "epoch": 0.9547313972750938, "grad_norm": 0.3597586154937744, "learning_rate": 9.77434728054506e-05, "loss": 0.0552, "step": 28240 }, { "epoch": 0.955069474965347, "grad_norm": 0.4242722690105438, "learning_rate": 9.774101653689735e-05, "loss": 0.0602, "step": 28250 }, { "epoch": 0.9554075526556003, "grad_norm": 0.23980151116847992, "learning_rate": 9.773855896312263e-05, "loss": 0.0592, "step": 28260 }, { "epoch": 0.9557456303458535, "grad_norm": 0.24049028754234314, "learning_rate": 9.773610008419365e-05, "loss": 0.065, "step": 28270 }, { "epoch": 0.9560837080361066, "grad_norm": 0.3095865845680237, "learning_rate": 9.773363990017763e-05, "loss": 0.0625, "step": 28280 }, { "epoch": 0.9564217857263599, "grad_norm": 0.2318384200334549, "learning_rate": 9.773117841114182e-05, "loss": 0.0614, "step": 28290 }, { "epoch": 0.9567598634166131, "grad_norm": 0.21453124284744263, "learning_rate": 9.772871561715353e-05, "loss": 0.0552, "step": 28300 }, { "epoch": 0.9570979411068664, "grad_norm": 0.2282780557870865, "learning_rate": 9.772625151828007e-05, "loss": 0.0628, "step": 28310 }, { "epoch": 0.9574360187971196, "grad_norm": 0.15896441042423248, "learning_rate": 9.772378611458886e-05, "loss": 0.0638, "step": 28320 }, { "epoch": 0.9577740964873728, "grad_norm": 0.3189753293991089, "learning_rate": 9.772131940614724e-05, "loss": 0.0584, "step": 28330 }, { "epoch": 0.9581121741776261, "grad_norm": 0.3478779196739197, "learning_rate": 9.771885139302268e-05, "loss": 0.0618, "step": 28340 }, { "epoch": 0.9584502518678792, "grad_norm": 0.29647180438041687, "learning_rate": 9.771638207528264e-05, "loss": 0.0615, "step": 28350 }, { "epoch": 0.9587883295581324, "grad_norm": 0.37807413935661316, "learning_rate": 9.771391145299465e-05, "loss": 0.0632, "step": 28360 }, { "epoch": 0.9591264072483857, "grad_norm": 0.28857412934303284, "learning_rate": 9.771143952622626e-05, "loss": 0.0521, "step": 28370 }, { "epoch": 0.9594644849386389, "grad_norm": 0.2155407816171646, "learning_rate": 9.770896629504502e-05, "loss": 0.0537, "step": 28380 }, { "epoch": 0.9598025626288921, "grad_norm": 0.20986001193523407, "learning_rate": 9.770649175951857e-05, "loss": 0.057, "step": 28390 }, { "epoch": 0.9601406403191454, "grad_norm": 0.27315959334373474, "learning_rate": 9.770401591971455e-05, "loss": 0.059, "step": 28400 }, { "epoch": 0.9604787180093985, "grad_norm": 0.28660133481025696, "learning_rate": 9.770153877570067e-05, "loss": 0.0652, "step": 28410 }, { "epoch": 0.9608167956996517, "grad_norm": 0.23158951103687286, "learning_rate": 9.769906032754464e-05, "loss": 0.0595, "step": 28420 }, { "epoch": 0.961154873389905, "grad_norm": 0.26737502217292786, "learning_rate": 9.769658057531422e-05, "loss": 0.0614, "step": 28430 }, { "epoch": 0.9614929510801582, "grad_norm": 0.1844329684972763, "learning_rate": 9.769409951907719e-05, "loss": 0.0552, "step": 28440 }, { "epoch": 0.9618310287704115, "grad_norm": 0.21116121113300323, "learning_rate": 9.76916171589014e-05, "loss": 0.0586, "step": 28450 }, { "epoch": 0.9621691064606647, "grad_norm": 0.13420474529266357, "learning_rate": 9.768913349485472e-05, "loss": 0.0585, "step": 28460 }, { "epoch": 0.9625071841509178, "grad_norm": 0.2977113723754883, "learning_rate": 9.768664852700505e-05, "loss": 0.0579, "step": 28470 }, { "epoch": 0.9628452618411711, "grad_norm": 0.22625617682933807, "learning_rate": 9.768416225542032e-05, "loss": 0.0646, "step": 28480 }, { "epoch": 0.9631833395314243, "grad_norm": 0.2552014887332916, "learning_rate": 9.76816746801685e-05, "loss": 0.0574, "step": 28490 }, { "epoch": 0.9635214172216775, "grad_norm": 0.5514748692512512, "learning_rate": 9.767918580131763e-05, "loss": 0.061, "step": 28500 }, { "epoch": 0.9638594949119308, "grad_norm": 0.2735542356967926, "learning_rate": 9.767669561893572e-05, "loss": 0.0563, "step": 28510 }, { "epoch": 0.964197572602184, "grad_norm": 0.5220099687576294, "learning_rate": 9.767420413309086e-05, "loss": 0.0571, "step": 28520 }, { "epoch": 0.9645356502924372, "grad_norm": 0.3911762535572052, "learning_rate": 9.767171134385118e-05, "loss": 0.0536, "step": 28530 }, { "epoch": 0.9648737279826904, "grad_norm": 0.22137169539928436, "learning_rate": 9.766921725128481e-05, "loss": 0.0555, "step": 28540 }, { "epoch": 0.9652118056729436, "grad_norm": 0.40290001034736633, "learning_rate": 9.766672185545995e-05, "loss": 0.0634, "step": 28550 }, { "epoch": 0.9655498833631969, "grad_norm": 0.21122561395168304, "learning_rate": 9.766422515644482e-05, "loss": 0.0631, "step": 28560 }, { "epoch": 0.9658879610534501, "grad_norm": 0.32551059126853943, "learning_rate": 9.766172715430769e-05, "loss": 0.0607, "step": 28570 }, { "epoch": 0.9662260387437033, "grad_norm": 0.2339930534362793, "learning_rate": 9.765922784911683e-05, "loss": 0.0629, "step": 28580 }, { "epoch": 0.9665641164339566, "grad_norm": 0.15823134779930115, "learning_rate": 9.76567272409406e-05, "loss": 0.0549, "step": 28590 }, { "epoch": 0.9669021941242097, "grad_norm": 0.3594124913215637, "learning_rate": 9.765422532984735e-05, "loss": 0.0655, "step": 28600 }, { "epoch": 0.9672402718144629, "grad_norm": 0.28982409834861755, "learning_rate": 9.765172211590547e-05, "loss": 0.0573, "step": 28610 }, { "epoch": 0.9675783495047162, "grad_norm": 0.2535153329372406, "learning_rate": 9.764921759918342e-05, "loss": 0.0593, "step": 28620 }, { "epoch": 0.9679164271949694, "grad_norm": 0.31573665142059326, "learning_rate": 9.764671177974966e-05, "loss": 0.0584, "step": 28630 }, { "epoch": 0.9682545048852226, "grad_norm": 0.20109403133392334, "learning_rate": 9.76442046576727e-05, "loss": 0.0605, "step": 28640 }, { "epoch": 0.9685925825754759, "grad_norm": 0.369261234998703, "learning_rate": 9.764169623302109e-05, "loss": 0.0562, "step": 28650 }, { "epoch": 0.968930660265729, "grad_norm": 0.2889997959136963, "learning_rate": 9.763918650586338e-05, "loss": 0.0586, "step": 28660 }, { "epoch": 0.9692687379559823, "grad_norm": 0.3187502324581146, "learning_rate": 9.763667547626825e-05, "loss": 0.0601, "step": 28670 }, { "epoch": 0.9696068156462355, "grad_norm": 0.3089054226875305, "learning_rate": 9.763416314430427e-05, "loss": 0.0617, "step": 28680 }, { "epoch": 0.9699448933364887, "grad_norm": 0.30551496148109436, "learning_rate": 9.763164951004018e-05, "loss": 0.0654, "step": 28690 }, { "epoch": 0.970282971026742, "grad_norm": 0.24013492465019226, "learning_rate": 9.762913457354469e-05, "loss": 0.0584, "step": 28700 }, { "epoch": 0.9706210487169952, "grad_norm": 0.199468731880188, "learning_rate": 9.762661833488654e-05, "loss": 0.0571, "step": 28710 }, { "epoch": 0.9709591264072484, "grad_norm": 0.19406673312187195, "learning_rate": 9.762410079413455e-05, "loss": 0.0532, "step": 28720 }, { "epoch": 0.9712972040975016, "grad_norm": 0.3021693229675293, "learning_rate": 9.762158195135754e-05, "loss": 0.0582, "step": 28730 }, { "epoch": 0.9716352817877548, "grad_norm": 0.2567639946937561, "learning_rate": 9.761906180662435e-05, "loss": 0.0621, "step": 28740 }, { "epoch": 0.971973359478008, "grad_norm": 0.38053151965141296, "learning_rate": 9.761654036000392e-05, "loss": 0.0535, "step": 28750 }, { "epoch": 0.9723114371682613, "grad_norm": 0.181413933634758, "learning_rate": 9.761401761156516e-05, "loss": 0.0583, "step": 28760 }, { "epoch": 0.9726495148585145, "grad_norm": 0.263590544462204, "learning_rate": 9.761149356137704e-05, "loss": 0.0597, "step": 28770 }, { "epoch": 0.9729875925487677, "grad_norm": 0.24922268092632294, "learning_rate": 9.760896820950858e-05, "loss": 0.06, "step": 28780 }, { "epoch": 0.973325670239021, "grad_norm": 0.2534850239753723, "learning_rate": 9.76064415560288e-05, "loss": 0.0586, "step": 28790 }, { "epoch": 0.9736637479292741, "grad_norm": 0.6221825480461121, "learning_rate": 9.760391360100681e-05, "loss": 0.0573, "step": 28800 }, { "epoch": 0.9740018256195274, "grad_norm": 0.2979000508785248, "learning_rate": 9.76013843445117e-05, "loss": 0.0615, "step": 28810 }, { "epoch": 0.9743399033097806, "grad_norm": 0.28475069999694824, "learning_rate": 9.759885378661263e-05, "loss": 0.0596, "step": 28820 }, { "epoch": 0.9746779810000338, "grad_norm": 0.28194659948349, "learning_rate": 9.759632192737878e-05, "loss": 0.0632, "step": 28830 }, { "epoch": 0.9750160586902871, "grad_norm": 0.34667718410491943, "learning_rate": 9.759378876687937e-05, "loss": 0.0633, "step": 28840 }, { "epoch": 0.9753541363805402, "grad_norm": 0.322090208530426, "learning_rate": 9.759125430518365e-05, "loss": 0.0596, "step": 28850 }, { "epoch": 0.9756922140707934, "grad_norm": 0.21247591078281403, "learning_rate": 9.758871854236091e-05, "loss": 0.0651, "step": 28860 }, { "epoch": 0.9760302917610467, "grad_norm": 0.28620532155036926, "learning_rate": 9.758618147848049e-05, "loss": 0.0601, "step": 28870 }, { "epoch": 0.9763683694512999, "grad_norm": 0.3549841344356537, "learning_rate": 9.758364311361174e-05, "loss": 0.0627, "step": 28880 }, { "epoch": 0.9767064471415531, "grad_norm": 0.2182544320821762, "learning_rate": 9.758110344782408e-05, "loss": 0.0557, "step": 28890 }, { "epoch": 0.9770445248318064, "grad_norm": 0.3031162917613983, "learning_rate": 9.75785624811869e-05, "loss": 0.0567, "step": 28900 }, { "epoch": 0.9773826025220596, "grad_norm": 0.49388617277145386, "learning_rate": 9.757602021376972e-05, "loss": 0.0687, "step": 28910 }, { "epoch": 0.9777206802123128, "grad_norm": 0.21869295835494995, "learning_rate": 9.757347664564202e-05, "loss": 0.0537, "step": 28920 }, { "epoch": 0.978058757902566, "grad_norm": 0.28445151448249817, "learning_rate": 9.757093177687333e-05, "loss": 0.0539, "step": 28930 }, { "epoch": 0.9783968355928192, "grad_norm": 0.21859951317310333, "learning_rate": 9.756838560753325e-05, "loss": 0.0574, "step": 28940 }, { "epoch": 0.9787349132830725, "grad_norm": 0.44055479764938354, "learning_rate": 9.756583813769136e-05, "loss": 0.0629, "step": 28950 }, { "epoch": 0.9790729909733257, "grad_norm": 0.31762588024139404, "learning_rate": 9.756328936741734e-05, "loss": 0.062, "step": 28960 }, { "epoch": 0.9794110686635789, "grad_norm": 0.24962589144706726, "learning_rate": 9.756073929678085e-05, "loss": 0.0639, "step": 28970 }, { "epoch": 0.9797491463538321, "grad_norm": 0.279347687959671, "learning_rate": 9.75581879258516e-05, "loss": 0.0553, "step": 28980 }, { "epoch": 0.9800872240440853, "grad_norm": 0.1886403113603592, "learning_rate": 9.75556352546994e-05, "loss": 0.0619, "step": 28990 }, { "epoch": 0.9804253017343385, "grad_norm": 0.25771158933639526, "learning_rate": 9.755308128339396e-05, "loss": 0.0576, "step": 29000 }, { "epoch": 0.9807633794245918, "grad_norm": 0.24462732672691345, "learning_rate": 9.755052601200517e-05, "loss": 0.0606, "step": 29010 }, { "epoch": 0.981101457114845, "grad_norm": 0.24634279310703278, "learning_rate": 9.754796944060285e-05, "loss": 0.0592, "step": 29020 }, { "epoch": 0.9814395348050982, "grad_norm": 0.35686740279197693, "learning_rate": 9.75454115692569e-05, "loss": 0.0594, "step": 29030 }, { "epoch": 0.9817776124953514, "grad_norm": 0.24329376220703125, "learning_rate": 9.754285239803727e-05, "loss": 0.065, "step": 29040 }, { "epoch": 0.9821156901856046, "grad_norm": 0.23629572987556458, "learning_rate": 9.75402919270139e-05, "loss": 0.0594, "step": 29050 }, { "epoch": 0.9824537678758579, "grad_norm": 0.27027708292007446, "learning_rate": 9.753773015625682e-05, "loss": 0.0656, "step": 29060 }, { "epoch": 0.9827918455661111, "grad_norm": 0.2550494372844696, "learning_rate": 9.753516708583607e-05, "loss": 0.0649, "step": 29070 }, { "epoch": 0.9831299232563643, "grad_norm": 0.1848481297492981, "learning_rate": 9.75326027158217e-05, "loss": 0.0553, "step": 29080 }, { "epoch": 0.9834680009466176, "grad_norm": 0.2419605851173401, "learning_rate": 9.753003704628383e-05, "loss": 0.0643, "step": 29090 }, { "epoch": 0.9838060786368708, "grad_norm": 0.3600636124610901, "learning_rate": 9.752747007729259e-05, "loss": 0.0581, "step": 29100 }, { "epoch": 0.9841441563271239, "grad_norm": 0.21131989359855652, "learning_rate": 9.75249018089182e-05, "loss": 0.0522, "step": 29110 }, { "epoch": 0.9844822340173772, "grad_norm": 0.30328017473220825, "learning_rate": 9.752233224123083e-05, "loss": 0.054, "step": 29120 }, { "epoch": 0.9848203117076304, "grad_norm": 0.28256502747535706, "learning_rate": 9.751976137430075e-05, "loss": 0.0651, "step": 29130 }, { "epoch": 0.9851583893978836, "grad_norm": 0.2449694722890854, "learning_rate": 9.751718920819826e-05, "loss": 0.0617, "step": 29140 }, { "epoch": 0.9854964670881369, "grad_norm": 0.33998504281044006, "learning_rate": 9.751461574299365e-05, "loss": 0.0638, "step": 29150 }, { "epoch": 0.98583454477839, "grad_norm": 0.491793155670166, "learning_rate": 9.75120409787573e-05, "loss": 0.0622, "step": 29160 }, { "epoch": 0.9861726224686433, "grad_norm": 0.22300399839878082, "learning_rate": 9.750946491555961e-05, "loss": 0.0579, "step": 29170 }, { "epoch": 0.9865107001588965, "grad_norm": 0.3142044246196747, "learning_rate": 9.7506887553471e-05, "loss": 0.0603, "step": 29180 }, { "epoch": 0.9868487778491497, "grad_norm": 0.22727487981319427, "learning_rate": 9.75043088925619e-05, "loss": 0.0645, "step": 29190 }, { "epoch": 0.987186855539403, "grad_norm": 0.2869369089603424, "learning_rate": 9.750172893290288e-05, "loss": 0.0571, "step": 29200 }, { "epoch": 0.9875249332296562, "grad_norm": 0.23125790059566498, "learning_rate": 9.749914767456441e-05, "loss": 0.0648, "step": 29210 }, { "epoch": 0.9878630109199094, "grad_norm": 0.390281617641449, "learning_rate": 9.74965651176171e-05, "loss": 0.0556, "step": 29220 }, { "epoch": 0.9882010886101626, "grad_norm": 0.2878570854663849, "learning_rate": 9.749398126213154e-05, "loss": 0.0563, "step": 29230 }, { "epoch": 0.9885391663004158, "grad_norm": 0.3304772675037384, "learning_rate": 9.749139610817838e-05, "loss": 0.0574, "step": 29240 }, { "epoch": 0.988877243990669, "grad_norm": 0.20867837965488434, "learning_rate": 9.748880965582829e-05, "loss": 0.0557, "step": 29250 }, { "epoch": 0.9892153216809223, "grad_norm": 0.48182544112205505, "learning_rate": 9.748622190515198e-05, "loss": 0.0618, "step": 29260 }, { "epoch": 0.9895533993711755, "grad_norm": 0.2855322062969208, "learning_rate": 9.74836328562202e-05, "loss": 0.0615, "step": 29270 }, { "epoch": 0.9898914770614288, "grad_norm": 0.5429818034172058, "learning_rate": 9.748104250910375e-05, "loss": 0.0631, "step": 29280 }, { "epoch": 0.990229554751682, "grad_norm": 0.2002713680267334, "learning_rate": 9.747845086387342e-05, "loss": 0.0625, "step": 29290 }, { "epoch": 0.9905676324419351, "grad_norm": 1.2610019445419312, "learning_rate": 9.74758579206001e-05, "loss": 0.0676, "step": 29300 }, { "epoch": 0.9909057101321884, "grad_norm": 0.3818122446537018, "learning_rate": 9.747326367935464e-05, "loss": 0.0614, "step": 29310 }, { "epoch": 0.9912437878224416, "grad_norm": 0.2543768882751465, "learning_rate": 9.7470668140208e-05, "loss": 0.0609, "step": 29320 }, { "epoch": 0.9915818655126948, "grad_norm": 0.19343429803848267, "learning_rate": 9.746807130323112e-05, "loss": 0.0568, "step": 29330 }, { "epoch": 0.9919199432029481, "grad_norm": 0.2608703076839447, "learning_rate": 9.746547316849502e-05, "loss": 0.0579, "step": 29340 }, { "epoch": 0.9922580208932013, "grad_norm": 0.24792329967021942, "learning_rate": 9.746287373607069e-05, "loss": 0.0606, "step": 29350 }, { "epoch": 0.9925960985834544, "grad_norm": 0.26770123839378357, "learning_rate": 9.746027300602922e-05, "loss": 0.064, "step": 29360 }, { "epoch": 0.9929341762737077, "grad_norm": 0.2165563404560089, "learning_rate": 9.745767097844174e-05, "loss": 0.0611, "step": 29370 }, { "epoch": 0.9932722539639609, "grad_norm": 0.2822762727737427, "learning_rate": 9.745506765337935e-05, "loss": 0.0615, "step": 29380 }, { "epoch": 0.9936103316542141, "grad_norm": 0.5907112956047058, "learning_rate": 9.745246303091324e-05, "loss": 0.0609, "step": 29390 }, { "epoch": 0.9939484093444674, "grad_norm": 0.2313842475414276, "learning_rate": 9.744985711111464e-05, "loss": 0.0514, "step": 29400 }, { "epoch": 0.9942864870347206, "grad_norm": 0.2432263344526291, "learning_rate": 9.744724989405474e-05, "loss": 0.0603, "step": 29410 }, { "epoch": 0.9946245647249738, "grad_norm": 0.3516450524330139, "learning_rate": 9.744464137980487e-05, "loss": 0.0581, "step": 29420 }, { "epoch": 0.994962642415227, "grad_norm": 0.17307697236537933, "learning_rate": 9.744203156843632e-05, "loss": 0.0596, "step": 29430 }, { "epoch": 0.9953007201054802, "grad_norm": 0.28372031450271606, "learning_rate": 9.743942046002047e-05, "loss": 0.0597, "step": 29440 }, { "epoch": 0.9956387977957335, "grad_norm": 0.2542682886123657, "learning_rate": 9.743680805462864e-05, "loss": 0.0595, "step": 29450 }, { "epoch": 0.9959768754859867, "grad_norm": 0.3302063047885895, "learning_rate": 9.743419435233234e-05, "loss": 0.0583, "step": 29460 }, { "epoch": 0.9963149531762399, "grad_norm": 0.3772037923336029, "learning_rate": 9.743157935320297e-05, "loss": 0.0589, "step": 29470 }, { "epoch": 0.9966530308664932, "grad_norm": 0.26432400941848755, "learning_rate": 9.742896305731206e-05, "loss": 0.0608, "step": 29480 }, { "epoch": 0.9969911085567463, "grad_norm": 0.31055355072021484, "learning_rate": 9.742634546473111e-05, "loss": 0.0546, "step": 29490 }, { "epoch": 0.9973291862469995, "grad_norm": 0.21483148634433746, "learning_rate": 9.742372657553167e-05, "loss": 0.0593, "step": 29500 }, { "epoch": 0.9976672639372528, "grad_norm": 0.3364061117172241, "learning_rate": 9.742110638978538e-05, "loss": 0.0618, "step": 29510 }, { "epoch": 0.998005341627506, "grad_norm": 0.15980133414268494, "learning_rate": 9.741848490756387e-05, "loss": 0.0587, "step": 29520 }, { "epoch": 0.9983434193177593, "grad_norm": 0.24126353859901428, "learning_rate": 9.741586212893877e-05, "loss": 0.0586, "step": 29530 }, { "epoch": 0.9986814970080125, "grad_norm": 0.3649369776248932, "learning_rate": 9.741323805398184e-05, "loss": 0.0646, "step": 29540 }, { "epoch": 0.9990195746982656, "grad_norm": 0.3026980459690094, "learning_rate": 9.741061268276477e-05, "loss": 0.0597, "step": 29550 }, { "epoch": 0.9993576523885189, "grad_norm": 0.2973880469799042, "learning_rate": 9.740798601535937e-05, "loss": 0.0594, "step": 29560 }, { "epoch": 0.9996957300787721, "grad_norm": 0.22633664309978485, "learning_rate": 9.740535805183744e-05, "loss": 0.0617, "step": 29570 }, { "epoch": 1.0000338077690254, "grad_norm": 0.4684033989906311, "learning_rate": 9.740272879227084e-05, "loss": 0.0555, "step": 29580 }, { "epoch": 1.0003718854592785, "grad_norm": 0.464397132396698, "learning_rate": 9.740009823673144e-05, "loss": 0.0616, "step": 29590 }, { "epoch": 1.0007099631495318, "grad_norm": 0.2541770935058594, "learning_rate": 9.739746638529115e-05, "loss": 0.0544, "step": 29600 }, { "epoch": 1.001048040839785, "grad_norm": 0.38231417536735535, "learning_rate": 9.739483323802194e-05, "loss": 0.0585, "step": 29610 }, { "epoch": 1.0013861185300381, "grad_norm": 0.19339890778064728, "learning_rate": 9.73921987949958e-05, "loss": 0.0548, "step": 29620 }, { "epoch": 1.0017241962202914, "grad_norm": 0.27721112966537476, "learning_rate": 9.738956305628476e-05, "loss": 0.0549, "step": 29630 }, { "epoch": 1.0020622739105447, "grad_norm": 0.34240707755088806, "learning_rate": 9.738692602196085e-05, "loss": 0.0534, "step": 29640 }, { "epoch": 1.0024003516007978, "grad_norm": 0.19869299232959747, "learning_rate": 9.738428769209618e-05, "loss": 0.0612, "step": 29650 }, { "epoch": 1.002738429291051, "grad_norm": 0.22563382983207703, "learning_rate": 9.73816480667629e-05, "loss": 0.0568, "step": 29660 }, { "epoch": 1.0030765069813043, "grad_norm": 0.20840151607990265, "learning_rate": 9.737900714603316e-05, "loss": 0.0603, "step": 29670 }, { "epoch": 1.0034145846715574, "grad_norm": 0.2836461663246155, "learning_rate": 9.737636492997916e-05, "loss": 0.0638, "step": 29680 }, { "epoch": 1.0037526623618107, "grad_norm": 0.22182928025722504, "learning_rate": 9.737372141867315e-05, "loss": 0.0556, "step": 29690 }, { "epoch": 1.004090740052064, "grad_norm": 0.2313470095396042, "learning_rate": 9.737107661218738e-05, "loss": 0.0553, "step": 29700 }, { "epoch": 1.0044288177423173, "grad_norm": 0.28361281752586365, "learning_rate": 9.73684305105942e-05, "loss": 0.0592, "step": 29710 }, { "epoch": 1.0047668954325704, "grad_norm": 0.2404135763645172, "learning_rate": 9.736578311396589e-05, "loss": 0.0527, "step": 29720 }, { "epoch": 1.0051049731228237, "grad_norm": 0.2563896179199219, "learning_rate": 9.736313442237488e-05, "loss": 0.0567, "step": 29730 }, { "epoch": 1.005443050813077, "grad_norm": 0.19488562643527985, "learning_rate": 9.736048443589356e-05, "loss": 0.0561, "step": 29740 }, { "epoch": 1.00578112850333, "grad_norm": 0.4045695960521698, "learning_rate": 9.735783315459441e-05, "loss": 0.0623, "step": 29750 }, { "epoch": 1.0061192061935833, "grad_norm": 0.2752428948879242, "learning_rate": 9.735518057854988e-05, "loss": 0.0613, "step": 29760 }, { "epoch": 1.0064572838838366, "grad_norm": 0.3476431667804718, "learning_rate": 9.73525267078325e-05, "loss": 0.0621, "step": 29770 }, { "epoch": 1.0067953615740897, "grad_norm": 0.38055622577667236, "learning_rate": 9.734987154251483e-05, "loss": 0.064, "step": 29780 }, { "epoch": 1.007133439264343, "grad_norm": 0.4726589322090149, "learning_rate": 9.734721508266946e-05, "loss": 0.0526, "step": 29790 }, { "epoch": 1.0074715169545962, "grad_norm": 0.30469846725463867, "learning_rate": 9.734455732836902e-05, "loss": 0.0619, "step": 29800 }, { "epoch": 1.0078095946448493, "grad_norm": 0.3159236013889313, "learning_rate": 9.734189827968618e-05, "loss": 0.0608, "step": 29810 }, { "epoch": 1.0081476723351026, "grad_norm": 0.30510076880455017, "learning_rate": 9.733923793669362e-05, "loss": 0.0605, "step": 29820 }, { "epoch": 1.008485750025356, "grad_norm": 0.2623688876628876, "learning_rate": 9.733657629946406e-05, "loss": 0.0578, "step": 29830 }, { "epoch": 1.008823827715609, "grad_norm": 0.2039112150669098, "learning_rate": 9.73339133680703e-05, "loss": 0.0608, "step": 29840 }, { "epoch": 1.0091619054058623, "grad_norm": 0.2929064631462097, "learning_rate": 9.733124914258514e-05, "loss": 0.0569, "step": 29850 }, { "epoch": 1.0094999830961155, "grad_norm": 0.17448081076145172, "learning_rate": 9.732858362308141e-05, "loss": 0.0548, "step": 29860 }, { "epoch": 1.0098380607863686, "grad_norm": 0.3799864947795868, "learning_rate": 9.732591680963196e-05, "loss": 0.0568, "step": 29870 }, { "epoch": 1.010176138476622, "grad_norm": 0.1957504153251648, "learning_rate": 9.732324870230975e-05, "loss": 0.0627, "step": 29880 }, { "epoch": 1.0105142161668752, "grad_norm": 0.2619172930717468, "learning_rate": 9.732057930118769e-05, "loss": 0.0604, "step": 29890 }, { "epoch": 1.0108522938571283, "grad_norm": 0.21521535515785217, "learning_rate": 9.731790860633878e-05, "loss": 0.0561, "step": 29900 }, { "epoch": 1.0111903715473816, "grad_norm": 0.5087517499923706, "learning_rate": 9.731523661783601e-05, "loss": 0.054, "step": 29910 }, { "epoch": 1.0115284492376349, "grad_norm": 0.2895056903362274, "learning_rate": 9.731256333575246e-05, "loss": 0.0677, "step": 29920 }, { "epoch": 1.011866526927888, "grad_norm": 0.24538539350032806, "learning_rate": 9.730988876016118e-05, "loss": 0.0616, "step": 29930 }, { "epoch": 1.0122046046181412, "grad_norm": 0.2613603174686432, "learning_rate": 9.730721289113534e-05, "loss": 0.0549, "step": 29940 }, { "epoch": 1.0125426823083945, "grad_norm": 0.5094873905181885, "learning_rate": 9.730453572874805e-05, "loss": 0.0564, "step": 29950 }, { "epoch": 1.0128807599986478, "grad_norm": 0.2578490674495697, "learning_rate": 9.730185727307255e-05, "loss": 0.056, "step": 29960 }, { "epoch": 1.0132188376889009, "grad_norm": 0.2702632546424866, "learning_rate": 9.729917752418203e-05, "loss": 0.0603, "step": 29970 }, { "epoch": 1.0135569153791542, "grad_norm": 0.17687886953353882, "learning_rate": 9.729649648214976e-05, "loss": 0.0604, "step": 29980 }, { "epoch": 1.0138949930694074, "grad_norm": 0.3551025688648224, "learning_rate": 9.729381414704906e-05, "loss": 0.0602, "step": 29990 }, { "epoch": 1.0142330707596605, "grad_norm": 0.24540656805038452, "learning_rate": 9.729113051895322e-05, "loss": 0.0603, "step": 30000 }, { "epoch": 1.0145711484499138, "grad_norm": 0.3634001314640045, "learning_rate": 9.728844559793566e-05, "loss": 0.0584, "step": 30010 }, { "epoch": 1.014909226140167, "grad_norm": 0.2953491508960724, "learning_rate": 9.728575938406976e-05, "loss": 0.0636, "step": 30020 }, { "epoch": 1.0152473038304202, "grad_norm": 0.37206152081489563, "learning_rate": 9.728307187742897e-05, "loss": 0.0656, "step": 30030 }, { "epoch": 1.0155853815206735, "grad_norm": 0.2286759912967682, "learning_rate": 9.728038307808674e-05, "loss": 0.062, "step": 30040 }, { "epoch": 1.0159234592109267, "grad_norm": 0.7046418190002441, "learning_rate": 9.727769298611662e-05, "loss": 0.0589, "step": 30050 }, { "epoch": 1.0162615369011798, "grad_norm": 0.4789946377277374, "learning_rate": 9.727500160159212e-05, "loss": 0.059, "step": 30060 }, { "epoch": 1.016599614591433, "grad_norm": 0.19706609845161438, "learning_rate": 9.727230892458684e-05, "loss": 0.0618, "step": 30070 }, { "epoch": 1.0169376922816864, "grad_norm": 0.23099493980407715, "learning_rate": 9.72696149551744e-05, "loss": 0.0645, "step": 30080 }, { "epoch": 1.0172757699719395, "grad_norm": 0.369169682264328, "learning_rate": 9.726691969342844e-05, "loss": 0.0603, "step": 30090 }, { "epoch": 1.0176138476621928, "grad_norm": 0.43223676085472107, "learning_rate": 9.726422313942265e-05, "loss": 0.0675, "step": 30100 }, { "epoch": 1.017951925352446, "grad_norm": 0.2577108144760132, "learning_rate": 9.726152529323076e-05, "loss": 0.0573, "step": 30110 }, { "epoch": 1.0182900030426991, "grad_norm": 0.3530677258968353, "learning_rate": 9.725882615492652e-05, "loss": 0.056, "step": 30120 }, { "epoch": 1.0186280807329524, "grad_norm": 0.2446477711200714, "learning_rate": 9.725612572458373e-05, "loss": 0.059, "step": 30130 }, { "epoch": 1.0189661584232057, "grad_norm": 0.34821629524230957, "learning_rate": 9.725342400227621e-05, "loss": 0.057, "step": 30140 }, { "epoch": 1.0193042361134588, "grad_norm": 0.3505595624446869, "learning_rate": 9.725072098807784e-05, "loss": 0.0611, "step": 30150 }, { "epoch": 1.019642313803712, "grad_norm": 0.23762856423854828, "learning_rate": 9.72480166820625e-05, "loss": 0.0629, "step": 30160 }, { "epoch": 1.0199803914939654, "grad_norm": 0.28823933005332947, "learning_rate": 9.724531108430414e-05, "loss": 0.0592, "step": 30170 }, { "epoch": 1.0203184691842186, "grad_norm": 0.2934286594390869, "learning_rate": 9.724260419487674e-05, "loss": 0.059, "step": 30180 }, { "epoch": 1.0206565468744717, "grad_norm": 0.348868727684021, "learning_rate": 9.723989601385425e-05, "loss": 0.0519, "step": 30190 }, { "epoch": 1.020994624564725, "grad_norm": 0.1992371678352356, "learning_rate": 9.723718654131078e-05, "loss": 0.0637, "step": 30200 }, { "epoch": 1.0213327022549783, "grad_norm": 0.23710107803344727, "learning_rate": 9.723447577732037e-05, "loss": 0.0632, "step": 30210 }, { "epoch": 1.0216707799452314, "grad_norm": 0.38175156712532043, "learning_rate": 9.723176372195715e-05, "loss": 0.0541, "step": 30220 }, { "epoch": 1.0220088576354847, "grad_norm": 0.37683773040771484, "learning_rate": 9.722905037529524e-05, "loss": 0.0622, "step": 30230 }, { "epoch": 1.022346935325738, "grad_norm": 0.4181617498397827, "learning_rate": 9.722633573740883e-05, "loss": 0.0598, "step": 30240 }, { "epoch": 1.022685013015991, "grad_norm": 0.20205843448638916, "learning_rate": 9.722361980837215e-05, "loss": 0.0589, "step": 30250 }, { "epoch": 1.0230230907062443, "grad_norm": 0.1934085488319397, "learning_rate": 9.722090258825944e-05, "loss": 0.0572, "step": 30260 }, { "epoch": 1.0233611683964976, "grad_norm": 0.28337666392326355, "learning_rate": 9.7218184077145e-05, "loss": 0.0588, "step": 30270 }, { "epoch": 1.0236992460867507, "grad_norm": 0.24995338916778564, "learning_rate": 9.721546427510315e-05, "loss": 0.0543, "step": 30280 }, { "epoch": 1.024037323777004, "grad_norm": 0.5915443301200867, "learning_rate": 9.721274318220823e-05, "loss": 0.0597, "step": 30290 }, { "epoch": 1.0243754014672573, "grad_norm": 0.2544187903404236, "learning_rate": 9.721002079853467e-05, "loss": 0.068, "step": 30300 }, { "epoch": 1.0247134791575103, "grad_norm": 0.34424927830696106, "learning_rate": 9.720729712415685e-05, "loss": 0.0668, "step": 30310 }, { "epoch": 1.0250515568477636, "grad_norm": 0.44392696022987366, "learning_rate": 9.720457215914928e-05, "loss": 0.06, "step": 30320 }, { "epoch": 1.025389634538017, "grad_norm": 0.21565641462802887, "learning_rate": 9.720184590358646e-05, "loss": 0.059, "step": 30330 }, { "epoch": 1.02572771222827, "grad_norm": 0.28229257464408875, "learning_rate": 9.719911835754287e-05, "loss": 0.0537, "step": 30340 }, { "epoch": 1.0260657899185233, "grad_norm": 0.22884239256381989, "learning_rate": 9.719638952109314e-05, "loss": 0.0567, "step": 30350 }, { "epoch": 1.0264038676087766, "grad_norm": 0.24886393547058105, "learning_rate": 9.719365939431183e-05, "loss": 0.0584, "step": 30360 }, { "epoch": 1.0267419452990296, "grad_norm": 0.20729579031467438, "learning_rate": 9.719092797727362e-05, "loss": 0.0638, "step": 30370 }, { "epoch": 1.027080022989283, "grad_norm": 0.227373406291008, "learning_rate": 9.718819527005316e-05, "loss": 0.0591, "step": 30380 }, { "epoch": 1.0274181006795362, "grad_norm": 0.2678352892398834, "learning_rate": 9.718546127272516e-05, "loss": 0.0626, "step": 30390 }, { "epoch": 1.0277561783697893, "grad_norm": 0.37624499201774597, "learning_rate": 9.71827259853644e-05, "loss": 0.0608, "step": 30400 }, { "epoch": 1.0280942560600426, "grad_norm": 0.21066869795322418, "learning_rate": 9.717998940804562e-05, "loss": 0.0599, "step": 30410 }, { "epoch": 1.0284323337502959, "grad_norm": 0.286929190158844, "learning_rate": 9.717725154084365e-05, "loss": 0.0574, "step": 30420 }, { "epoch": 1.028770411440549, "grad_norm": 0.2955387234687805, "learning_rate": 9.717451238383335e-05, "loss": 0.0609, "step": 30430 }, { "epoch": 1.0291084891308022, "grad_norm": 0.35787805914878845, "learning_rate": 9.717177193708962e-05, "loss": 0.0555, "step": 30440 }, { "epoch": 1.0294465668210555, "grad_norm": 0.3914480209350586, "learning_rate": 9.716903020068735e-05, "loss": 0.0561, "step": 30450 }, { "epoch": 1.0297846445113088, "grad_norm": 0.5462701916694641, "learning_rate": 9.716628717470152e-05, "loss": 0.0546, "step": 30460 }, { "epoch": 1.0301227222015619, "grad_norm": 0.31889647245407104, "learning_rate": 9.716354285920712e-05, "loss": 0.0577, "step": 30470 }, { "epoch": 1.0304607998918152, "grad_norm": 0.30281683802604675, "learning_rate": 9.716079725427918e-05, "loss": 0.0583, "step": 30480 }, { "epoch": 1.0307988775820685, "grad_norm": 0.23378397524356842, "learning_rate": 9.715805035999275e-05, "loss": 0.0549, "step": 30490 }, { "epoch": 1.0311369552723215, "grad_norm": 0.2690887153148651, "learning_rate": 9.715530217642295e-05, "loss": 0.0672, "step": 30500 }, { "epoch": 1.0314750329625748, "grad_norm": 0.20001737773418427, "learning_rate": 9.715255270364489e-05, "loss": 0.0588, "step": 30510 }, { "epoch": 1.031813110652828, "grad_norm": 0.27222442626953125, "learning_rate": 9.714980194173376e-05, "loss": 0.0585, "step": 30520 }, { "epoch": 1.0321511883430812, "grad_norm": 0.23265261948108673, "learning_rate": 9.714704989076477e-05, "loss": 0.0573, "step": 30530 }, { "epoch": 1.0324892660333345, "grad_norm": 0.4319738745689392, "learning_rate": 9.714429655081314e-05, "loss": 0.0591, "step": 30540 }, { "epoch": 1.0328273437235878, "grad_norm": 0.22795850038528442, "learning_rate": 9.714154192195415e-05, "loss": 0.0582, "step": 30550 }, { "epoch": 1.0331654214138408, "grad_norm": 0.26036736369132996, "learning_rate": 9.713878600426312e-05, "loss": 0.0554, "step": 30560 }, { "epoch": 1.0335034991040941, "grad_norm": 0.31053751707077026, "learning_rate": 9.713602879781538e-05, "loss": 0.0613, "step": 30570 }, { "epoch": 1.0338415767943474, "grad_norm": 0.27740830183029175, "learning_rate": 9.713327030268633e-05, "loss": 0.0577, "step": 30580 }, { "epoch": 1.0341796544846005, "grad_norm": 0.30823931097984314, "learning_rate": 9.713051051895137e-05, "loss": 0.0618, "step": 30590 }, { "epoch": 1.0345177321748538, "grad_norm": 0.2389252781867981, "learning_rate": 9.712774944668596e-05, "loss": 0.0646, "step": 30600 }, { "epoch": 1.034855809865107, "grad_norm": 0.2639399468898773, "learning_rate": 9.712498708596559e-05, "loss": 0.0547, "step": 30610 }, { "epoch": 1.0351938875553601, "grad_norm": 0.35294878482818604, "learning_rate": 9.712222343686578e-05, "loss": 0.0636, "step": 30620 }, { "epoch": 1.0355319652456134, "grad_norm": 0.27660831809043884, "learning_rate": 9.711945849946207e-05, "loss": 0.0516, "step": 30630 }, { "epoch": 1.0358700429358667, "grad_norm": 0.3617748022079468, "learning_rate": 9.71166922738301e-05, "loss": 0.0597, "step": 30640 }, { "epoch": 1.0362081206261198, "grad_norm": 0.32925301790237427, "learning_rate": 9.711392476004543e-05, "loss": 0.0574, "step": 30650 }, { "epoch": 1.036546198316373, "grad_norm": 0.2759394645690918, "learning_rate": 9.711115595818376e-05, "loss": 0.0582, "step": 30660 }, { "epoch": 1.0368842760066264, "grad_norm": 0.36131101846694946, "learning_rate": 9.71083858683208e-05, "loss": 0.0646, "step": 30670 }, { "epoch": 1.0372223536968797, "grad_norm": 0.3236643671989441, "learning_rate": 9.710561449053226e-05, "loss": 0.0547, "step": 30680 }, { "epoch": 1.0375604313871327, "grad_norm": 0.29060545563697815, "learning_rate": 9.710284182489391e-05, "loss": 0.0591, "step": 30690 }, { "epoch": 1.037898509077386, "grad_norm": 0.2488906979560852, "learning_rate": 9.710006787148158e-05, "loss": 0.0582, "step": 30700 }, { "epoch": 1.0382365867676393, "grad_norm": 0.29451659321784973, "learning_rate": 9.709729263037106e-05, "loss": 0.0566, "step": 30710 }, { "epoch": 1.0385746644578924, "grad_norm": 0.2431429922580719, "learning_rate": 9.709451610163826e-05, "loss": 0.0645, "step": 30720 }, { "epoch": 1.0389127421481457, "grad_norm": 0.24406972527503967, "learning_rate": 9.709173828535911e-05, "loss": 0.062, "step": 30730 }, { "epoch": 1.039250819838399, "grad_norm": 0.36878764629364014, "learning_rate": 9.708895918160951e-05, "loss": 0.0592, "step": 30740 }, { "epoch": 1.039588897528652, "grad_norm": 0.23109523952007294, "learning_rate": 9.708617879046544e-05, "loss": 0.061, "step": 30750 }, { "epoch": 1.0399269752189053, "grad_norm": 0.25753840804100037, "learning_rate": 9.708339711200295e-05, "loss": 0.0612, "step": 30760 }, { "epoch": 1.0402650529091586, "grad_norm": 0.25952082872390747, "learning_rate": 9.708061414629808e-05, "loss": 0.0617, "step": 30770 }, { "epoch": 1.0406031305994117, "grad_norm": 0.2513565421104431, "learning_rate": 9.707782989342688e-05, "loss": 0.0535, "step": 30780 }, { "epoch": 1.040941208289665, "grad_norm": 0.1749599277973175, "learning_rate": 9.707504435346552e-05, "loss": 0.0545, "step": 30790 }, { "epoch": 1.0412792859799183, "grad_norm": 0.34054243564605713, "learning_rate": 9.707225752649012e-05, "loss": 0.0577, "step": 30800 }, { "epoch": 1.0416173636701713, "grad_norm": 0.25968289375305176, "learning_rate": 9.706946941257689e-05, "loss": 0.055, "step": 30810 }, { "epoch": 1.0419554413604246, "grad_norm": 0.3469538688659668, "learning_rate": 9.706668001180204e-05, "loss": 0.0587, "step": 30820 }, { "epoch": 1.042293519050678, "grad_norm": 0.32608476281166077, "learning_rate": 9.706388932424185e-05, "loss": 0.0557, "step": 30830 }, { "epoch": 1.042631596740931, "grad_norm": 0.26124894618988037, "learning_rate": 9.70610973499726e-05, "loss": 0.0572, "step": 30840 }, { "epoch": 1.0429696744311843, "grad_norm": 0.22752217948436737, "learning_rate": 9.705830408907063e-05, "loss": 0.0593, "step": 30850 }, { "epoch": 1.0433077521214376, "grad_norm": 0.30140623450279236, "learning_rate": 9.705550954161233e-05, "loss": 0.0548, "step": 30860 }, { "epoch": 1.0436458298116906, "grad_norm": 0.21117496490478516, "learning_rate": 9.705271370767403e-05, "loss": 0.0606, "step": 30870 }, { "epoch": 1.043983907501944, "grad_norm": 0.251069575548172, "learning_rate": 9.704991658733225e-05, "loss": 0.0604, "step": 30880 }, { "epoch": 1.0443219851921972, "grad_norm": 0.565182089805603, "learning_rate": 9.704711818066342e-05, "loss": 0.0607, "step": 30890 }, { "epoch": 1.0446600628824503, "grad_norm": 0.2400330752134323, "learning_rate": 9.704431848774405e-05, "loss": 0.066, "step": 30900 }, { "epoch": 1.0449981405727036, "grad_norm": 0.262085884809494, "learning_rate": 9.704151750865069e-05, "loss": 0.0598, "step": 30910 }, { "epoch": 1.0453362182629569, "grad_norm": 0.25119441747665405, "learning_rate": 9.70387152434599e-05, "loss": 0.0556, "step": 30920 }, { "epoch": 1.0456742959532102, "grad_norm": 0.2348570078611374, "learning_rate": 9.703591169224833e-05, "loss": 0.057, "step": 30930 }, { "epoch": 1.0460123736434632, "grad_norm": 0.41174715757369995, "learning_rate": 9.703310685509258e-05, "loss": 0.0569, "step": 30940 }, { "epoch": 1.0463504513337165, "grad_norm": 0.3513069748878479, "learning_rate": 9.703030073206939e-05, "loss": 0.0601, "step": 30950 }, { "epoch": 1.0466885290239698, "grad_norm": 0.3077208995819092, "learning_rate": 9.70274933232554e-05, "loss": 0.053, "step": 30960 }, { "epoch": 1.0470266067142229, "grad_norm": 0.22183550894260406, "learning_rate": 9.702468462872745e-05, "loss": 0.052, "step": 30970 }, { "epoch": 1.0473646844044762, "grad_norm": 1.427620768547058, "learning_rate": 9.702187464856227e-05, "loss": 0.0575, "step": 30980 }, { "epoch": 1.0477027620947295, "grad_norm": 1.317177414894104, "learning_rate": 9.70190633828367e-05, "loss": 0.0547, "step": 30990 }, { "epoch": 1.0480408397849825, "grad_norm": 0.3252991735935211, "learning_rate": 9.701625083162763e-05, "loss": 0.0567, "step": 31000 }, { "epoch": 1.0483789174752358, "grad_norm": 0.3616145849227905, "learning_rate": 9.701343699501189e-05, "loss": 0.0576, "step": 31010 }, { "epoch": 1.048716995165489, "grad_norm": 0.2792626619338989, "learning_rate": 9.701062187306645e-05, "loss": 0.0612, "step": 31020 }, { "epoch": 1.0490550728557422, "grad_norm": 0.26738235354423523, "learning_rate": 9.700780546586827e-05, "loss": 0.0608, "step": 31030 }, { "epoch": 1.0493931505459955, "grad_norm": 0.23779569566249847, "learning_rate": 9.700498777349435e-05, "loss": 0.06, "step": 31040 }, { "epoch": 1.0497312282362488, "grad_norm": 0.19225649535655975, "learning_rate": 9.700216879602175e-05, "loss": 0.0584, "step": 31050 }, { "epoch": 1.0500693059265018, "grad_norm": 0.21956056356430054, "learning_rate": 9.699934853352749e-05, "loss": 0.0655, "step": 31060 }, { "epoch": 1.0504073836167551, "grad_norm": 0.27302810549736023, "learning_rate": 9.69965269860887e-05, "loss": 0.0554, "step": 31070 }, { "epoch": 1.0507454613070084, "grad_norm": 0.20601709187030792, "learning_rate": 9.699370415378252e-05, "loss": 0.0586, "step": 31080 }, { "epoch": 1.0510835389972615, "grad_norm": 0.16998597979545593, "learning_rate": 9.699088003668612e-05, "loss": 0.0533, "step": 31090 }, { "epoch": 1.0514216166875148, "grad_norm": 0.2958008944988251, "learning_rate": 9.698805463487671e-05, "loss": 0.0527, "step": 31100 }, { "epoch": 1.051759694377768, "grad_norm": 0.33213135600090027, "learning_rate": 9.698522794843154e-05, "loss": 0.0545, "step": 31110 }, { "epoch": 1.0520977720680211, "grad_norm": 0.249151811003685, "learning_rate": 9.69823999774279e-05, "loss": 0.0565, "step": 31120 }, { "epoch": 1.0524358497582744, "grad_norm": 0.28631526231765747, "learning_rate": 9.697957072194307e-05, "loss": 0.0594, "step": 31130 }, { "epoch": 1.0527739274485277, "grad_norm": 0.2901904881000519, "learning_rate": 9.697674018205446e-05, "loss": 0.0594, "step": 31140 }, { "epoch": 1.053112005138781, "grad_norm": 0.24130316078662872, "learning_rate": 9.697390835783941e-05, "loss": 0.0557, "step": 31150 }, { "epoch": 1.053450082829034, "grad_norm": 0.5774691700935364, "learning_rate": 9.697107524937534e-05, "loss": 0.0649, "step": 31160 }, { "epoch": 1.0537881605192874, "grad_norm": 0.2292492538690567, "learning_rate": 9.696824085673973e-05, "loss": 0.0586, "step": 31170 }, { "epoch": 1.0541262382095407, "grad_norm": 0.23233580589294434, "learning_rate": 9.696540518001006e-05, "loss": 0.057, "step": 31180 }, { "epoch": 1.0544643158997937, "grad_norm": 0.39799678325653076, "learning_rate": 9.696256821926386e-05, "loss": 0.0645, "step": 31190 }, { "epoch": 1.054802393590047, "grad_norm": 0.9218637347221375, "learning_rate": 9.695972997457867e-05, "loss": 0.0639, "step": 31200 }, { "epoch": 1.0551404712803003, "grad_norm": 0.2830340564250946, "learning_rate": 9.695689044603212e-05, "loss": 0.0577, "step": 31210 }, { "epoch": 1.0554785489705534, "grad_norm": 0.37669530510902405, "learning_rate": 9.695404963370183e-05, "loss": 0.0594, "step": 31220 }, { "epoch": 1.0558166266608067, "grad_norm": 0.4653826653957367, "learning_rate": 9.695120753766544e-05, "loss": 0.0635, "step": 31230 }, { "epoch": 1.05615470435106, "grad_norm": 0.7953742742538452, "learning_rate": 9.69483641580007e-05, "loss": 0.0525, "step": 31240 }, { "epoch": 1.056492782041313, "grad_norm": 0.30068352818489075, "learning_rate": 9.694551949478533e-05, "loss": 0.057, "step": 31250 }, { "epoch": 1.0568308597315663, "grad_norm": 0.2880414128303528, "learning_rate": 9.694267354809705e-05, "loss": 0.0545, "step": 31260 }, { "epoch": 1.0571689374218196, "grad_norm": 0.2575841546058655, "learning_rate": 9.693982631801374e-05, "loss": 0.0592, "step": 31270 }, { "epoch": 1.0575070151120727, "grad_norm": 0.26653268933296204, "learning_rate": 9.693697780461322e-05, "loss": 0.0521, "step": 31280 }, { "epoch": 1.057845092802326, "grad_norm": 0.20952776074409485, "learning_rate": 9.693412800797335e-05, "loss": 0.0605, "step": 31290 }, { "epoch": 1.0581831704925793, "grad_norm": 0.189285010099411, "learning_rate": 9.693127692817204e-05, "loss": 0.0643, "step": 31300 }, { "epoch": 1.0585212481828323, "grad_norm": 0.22213582694530487, "learning_rate": 9.692842456528726e-05, "loss": 0.0643, "step": 31310 }, { "epoch": 1.0588593258730856, "grad_norm": 0.8406102657318115, "learning_rate": 9.692557091939697e-05, "loss": 0.0579, "step": 31320 }, { "epoch": 1.059197403563339, "grad_norm": 0.2735289931297302, "learning_rate": 9.692271599057922e-05, "loss": 0.0621, "step": 31330 }, { "epoch": 1.059535481253592, "grad_norm": 0.3314771056175232, "learning_rate": 9.691985977891205e-05, "loss": 0.0596, "step": 31340 }, { "epoch": 1.0598735589438453, "grad_norm": 0.3433828353881836, "learning_rate": 9.691700228447353e-05, "loss": 0.0587, "step": 31350 }, { "epoch": 1.0602116366340986, "grad_norm": 0.36201342940330505, "learning_rate": 9.691414350734178e-05, "loss": 0.0624, "step": 31360 }, { "epoch": 1.0605497143243516, "grad_norm": 0.2535654604434967, "learning_rate": 9.691128344759499e-05, "loss": 0.0574, "step": 31370 }, { "epoch": 1.060887792014605, "grad_norm": 0.265342116355896, "learning_rate": 9.690842210531133e-05, "loss": 0.0553, "step": 31380 }, { "epoch": 1.0612258697048582, "grad_norm": 0.22371487319469452, "learning_rate": 9.690555948056903e-05, "loss": 0.06, "step": 31390 }, { "epoch": 1.0615639473951113, "grad_norm": 0.30396679043769836, "learning_rate": 9.690269557344635e-05, "loss": 0.056, "step": 31400 }, { "epoch": 1.0619020250853646, "grad_norm": 0.3057507872581482, "learning_rate": 9.689983038402159e-05, "loss": 0.0719, "step": 31410 }, { "epoch": 1.0622401027756179, "grad_norm": 0.2997560501098633, "learning_rate": 9.689696391237311e-05, "loss": 0.0561, "step": 31420 }, { "epoch": 1.0625781804658712, "grad_norm": 0.46235910058021545, "learning_rate": 9.689409615857923e-05, "loss": 0.0603, "step": 31430 }, { "epoch": 1.0629162581561242, "grad_norm": 0.256775826215744, "learning_rate": 9.689122712271838e-05, "loss": 0.0541, "step": 31440 }, { "epoch": 1.0632543358463775, "grad_norm": 0.3040803074836731, "learning_rate": 9.688835680486901e-05, "loss": 0.0517, "step": 31450 }, { "epoch": 1.0635924135366308, "grad_norm": 0.2291562855243683, "learning_rate": 9.688548520510959e-05, "loss": 0.0598, "step": 31460 }, { "epoch": 1.0639304912268839, "grad_norm": 0.28143811225891113, "learning_rate": 9.68826123235186e-05, "loss": 0.061, "step": 31470 }, { "epoch": 1.0642685689171372, "grad_norm": 0.29671716690063477, "learning_rate": 9.68797381601746e-05, "loss": 0.0605, "step": 31480 }, { "epoch": 1.0646066466073905, "grad_norm": 0.2751860022544861, "learning_rate": 9.687686271515618e-05, "loss": 0.0563, "step": 31490 }, { "epoch": 1.0649447242976435, "grad_norm": 0.3146272599697113, "learning_rate": 9.687398598854194e-05, "loss": 0.0553, "step": 31500 }, { "epoch": 1.0652828019878968, "grad_norm": 0.22120636701583862, "learning_rate": 9.687110798041053e-05, "loss": 0.0601, "step": 31510 }, { "epoch": 1.0656208796781501, "grad_norm": 0.7794344425201416, "learning_rate": 9.686822869084065e-05, "loss": 0.0609, "step": 31520 }, { "epoch": 1.0659589573684032, "grad_norm": 0.23269027471542358, "learning_rate": 9.686534811991099e-05, "loss": 0.0666, "step": 31530 }, { "epoch": 1.0662970350586565, "grad_norm": 0.15919265151023865, "learning_rate": 9.686246626770032e-05, "loss": 0.0581, "step": 31540 }, { "epoch": 1.0666351127489098, "grad_norm": 0.18829593062400818, "learning_rate": 9.685958313428744e-05, "loss": 0.0633, "step": 31550 }, { "epoch": 1.0669731904391628, "grad_norm": 0.21884682774543762, "learning_rate": 9.685669871975113e-05, "loss": 0.0574, "step": 31560 }, { "epoch": 1.0673112681294161, "grad_norm": 0.5830971598625183, "learning_rate": 9.685381302417032e-05, "loss": 0.0594, "step": 31570 }, { "epoch": 1.0676493458196694, "grad_norm": 0.22102463245391846, "learning_rate": 9.685092604762385e-05, "loss": 0.0558, "step": 31580 }, { "epoch": 1.0679874235099225, "grad_norm": 0.17789965867996216, "learning_rate": 9.684803779019065e-05, "loss": 0.0636, "step": 31590 }, { "epoch": 1.0683255012001758, "grad_norm": 0.2394639253616333, "learning_rate": 9.684514825194971e-05, "loss": 0.0515, "step": 31600 }, { "epoch": 1.068663578890429, "grad_norm": 0.21805928647518158, "learning_rate": 9.684225743298001e-05, "loss": 0.0602, "step": 31610 }, { "epoch": 1.0690016565806824, "grad_norm": 0.2391745001077652, "learning_rate": 9.68393653333606e-05, "loss": 0.0537, "step": 31620 }, { "epoch": 1.0693397342709354, "grad_norm": 0.17358830571174622, "learning_rate": 9.683647195317053e-05, "loss": 0.0539, "step": 31630 }, { "epoch": 1.0696778119611887, "grad_norm": 0.17509548366069794, "learning_rate": 9.683357729248892e-05, "loss": 0.0583, "step": 31640 }, { "epoch": 1.070015889651442, "grad_norm": 0.33832690119743347, "learning_rate": 9.683068135139488e-05, "loss": 0.0599, "step": 31650 }, { "epoch": 1.070353967341695, "grad_norm": 0.2193458080291748, "learning_rate": 9.682778412996762e-05, "loss": 0.0546, "step": 31660 }, { "epoch": 1.0706920450319484, "grad_norm": 0.2691827714443207, "learning_rate": 9.682488562828634e-05, "loss": 0.0606, "step": 31670 }, { "epoch": 1.0710301227222017, "grad_norm": 0.5362676978111267, "learning_rate": 9.682198584643027e-05, "loss": 0.053, "step": 31680 }, { "epoch": 1.0713682004124547, "grad_norm": 0.2738834619522095, "learning_rate": 9.68190847844787e-05, "loss": 0.0608, "step": 31690 }, { "epoch": 1.071706278102708, "grad_norm": 0.1911444365978241, "learning_rate": 9.681618244251094e-05, "loss": 0.0644, "step": 31700 }, { "epoch": 1.0720443557929613, "grad_norm": 0.436871737241745, "learning_rate": 9.681327882060634e-05, "loss": 0.0603, "step": 31710 }, { "epoch": 1.0723824334832144, "grad_norm": 0.21576468646526337, "learning_rate": 9.681037391884428e-05, "loss": 0.0572, "step": 31720 }, { "epoch": 1.0727205111734677, "grad_norm": 0.23023875057697296, "learning_rate": 9.680746773730421e-05, "loss": 0.0617, "step": 31730 }, { "epoch": 1.073058588863721, "grad_norm": 0.42067018151283264, "learning_rate": 9.680456027606552e-05, "loss": 0.0655, "step": 31740 }, { "epoch": 1.073396666553974, "grad_norm": 0.2415880262851715, "learning_rate": 9.680165153520775e-05, "loss": 0.0576, "step": 31750 }, { "epoch": 1.0737347442442273, "grad_norm": 0.22920367121696472, "learning_rate": 9.679874151481041e-05, "loss": 0.0549, "step": 31760 }, { "epoch": 1.0740728219344806, "grad_norm": 0.2810083329677582, "learning_rate": 9.679583021495307e-05, "loss": 0.063, "step": 31770 }, { "epoch": 1.0744108996247337, "grad_norm": 0.28908446431159973, "learning_rate": 9.67929176357153e-05, "loss": 0.0587, "step": 31780 }, { "epoch": 1.074748977314987, "grad_norm": 0.3270852267742157, "learning_rate": 9.679000377717673e-05, "loss": 0.058, "step": 31790 }, { "epoch": 1.0750870550052403, "grad_norm": 0.35478442907333374, "learning_rate": 9.678708863941705e-05, "loss": 0.0627, "step": 31800 }, { "epoch": 1.0754251326954933, "grad_norm": 0.3900524973869324, "learning_rate": 9.678417222251594e-05, "loss": 0.0505, "step": 31810 }, { "epoch": 1.0757632103857466, "grad_norm": 0.2553800642490387, "learning_rate": 9.678125452655313e-05, "loss": 0.0594, "step": 31820 }, { "epoch": 1.076101288076, "grad_norm": 0.2230110764503479, "learning_rate": 9.67783355516084e-05, "loss": 0.0651, "step": 31830 }, { "epoch": 1.076439365766253, "grad_norm": 0.23158515989780426, "learning_rate": 9.677541529776156e-05, "loss": 0.0587, "step": 31840 }, { "epoch": 1.0767774434565063, "grad_norm": 0.22778968513011932, "learning_rate": 9.677249376509243e-05, "loss": 0.0603, "step": 31850 }, { "epoch": 1.0771155211467596, "grad_norm": 0.3919815123081207, "learning_rate": 9.67695709536809e-05, "loss": 0.0603, "step": 31860 }, { "epoch": 1.0774535988370126, "grad_norm": 0.2331584244966507, "learning_rate": 9.676664686360686e-05, "loss": 0.061, "step": 31870 }, { "epoch": 1.077791676527266, "grad_norm": 0.23930059373378754, "learning_rate": 9.676372149495027e-05, "loss": 0.0567, "step": 31880 }, { "epoch": 1.0781297542175192, "grad_norm": 0.3995511829853058, "learning_rate": 9.67607948477911e-05, "loss": 0.0605, "step": 31890 }, { "epoch": 1.0784678319077723, "grad_norm": 0.31844767928123474, "learning_rate": 9.675786692220938e-05, "loss": 0.0553, "step": 31900 }, { "epoch": 1.0788059095980256, "grad_norm": 0.2769261300563812, "learning_rate": 9.675493771828512e-05, "loss": 0.0537, "step": 31910 }, { "epoch": 1.0791439872882789, "grad_norm": 0.21815939247608185, "learning_rate": 9.675200723609845e-05, "loss": 0.0551, "step": 31920 }, { "epoch": 1.0794820649785322, "grad_norm": 0.4581466317176819, "learning_rate": 9.674907547572944e-05, "loss": 0.0559, "step": 31930 }, { "epoch": 1.0798201426687852, "grad_norm": 0.34458404779434204, "learning_rate": 9.67461424372583e-05, "loss": 0.0578, "step": 31940 }, { "epoch": 1.0801582203590385, "grad_norm": 0.24704399704933167, "learning_rate": 9.674320812076516e-05, "loss": 0.064, "step": 31950 }, { "epoch": 1.0804962980492918, "grad_norm": 0.2266072928905487, "learning_rate": 9.674027252633029e-05, "loss": 0.0584, "step": 31960 }, { "epoch": 1.0808343757395449, "grad_norm": 0.23778490722179413, "learning_rate": 9.673733565403392e-05, "loss": 0.0594, "step": 31970 }, { "epoch": 1.0811724534297982, "grad_norm": 0.19565759599208832, "learning_rate": 9.673439750395636e-05, "loss": 0.0524, "step": 31980 }, { "epoch": 1.0815105311200515, "grad_norm": 0.24012291431427002, "learning_rate": 9.673145807617792e-05, "loss": 0.0629, "step": 31990 }, { "epoch": 1.0818486088103045, "grad_norm": 0.27924010157585144, "learning_rate": 9.672851737077898e-05, "loss": 0.0564, "step": 32000 }, { "epoch": 1.0821866865005578, "grad_norm": 0.35547536611557007, "learning_rate": 9.672557538783993e-05, "loss": 0.0545, "step": 32010 }, { "epoch": 1.0825247641908111, "grad_norm": 0.28657034039497375, "learning_rate": 9.67226321274412e-05, "loss": 0.0599, "step": 32020 }, { "epoch": 1.0828628418810642, "grad_norm": 0.28603485226631165, "learning_rate": 9.671968758966326e-05, "loss": 0.0581, "step": 32030 }, { "epoch": 1.0832009195713175, "grad_norm": 0.2689402401447296, "learning_rate": 9.67167417745866e-05, "loss": 0.058, "step": 32040 }, { "epoch": 1.0835389972615708, "grad_norm": 0.27816319465637207, "learning_rate": 9.671379468229178e-05, "loss": 0.0614, "step": 32050 }, { "epoch": 1.0838770749518238, "grad_norm": 0.28386184573173523, "learning_rate": 9.671084631285937e-05, "loss": 0.0602, "step": 32060 }, { "epoch": 1.0842151526420771, "grad_norm": 0.2961484491825104, "learning_rate": 9.670789666636997e-05, "loss": 0.052, "step": 32070 }, { "epoch": 1.0845532303323304, "grad_norm": 0.2002166211605072, "learning_rate": 9.670494574290421e-05, "loss": 0.0568, "step": 32080 }, { "epoch": 1.0848913080225835, "grad_norm": 0.3126868009567261, "learning_rate": 9.670199354254279e-05, "loss": 0.0546, "step": 32090 }, { "epoch": 1.0852293857128368, "grad_norm": 0.31567323207855225, "learning_rate": 9.66990400653664e-05, "loss": 0.0573, "step": 32100 }, { "epoch": 1.08556746340309, "grad_norm": 0.23740804195404053, "learning_rate": 9.669608531145581e-05, "loss": 0.0624, "step": 32110 }, { "epoch": 1.0859055410933434, "grad_norm": 0.23660849034786224, "learning_rate": 9.669312928089181e-05, "loss": 0.0566, "step": 32120 }, { "epoch": 1.0862436187835964, "grad_norm": 0.5200197696685791, "learning_rate": 9.669017197375517e-05, "loss": 0.0563, "step": 32130 }, { "epoch": 1.0865816964738497, "grad_norm": 0.23433871567249298, "learning_rate": 9.668721339012677e-05, "loss": 0.0548, "step": 32140 }, { "epoch": 1.086919774164103, "grad_norm": 0.26166877150535583, "learning_rate": 9.668425353008749e-05, "loss": 0.0576, "step": 32150 }, { "epoch": 1.087257851854356, "grad_norm": 0.27867093682289124, "learning_rate": 9.668129239371828e-05, "loss": 0.0586, "step": 32160 }, { "epoch": 1.0875959295446094, "grad_norm": 0.25016704201698303, "learning_rate": 9.667832998110006e-05, "loss": 0.055, "step": 32170 }, { "epoch": 1.0879340072348627, "grad_norm": 0.24702541530132294, "learning_rate": 9.667536629231384e-05, "loss": 0.0617, "step": 32180 }, { "epoch": 1.0882720849251157, "grad_norm": 0.37868672609329224, "learning_rate": 9.667240132744063e-05, "loss": 0.0552, "step": 32190 }, { "epoch": 1.088610162615369, "grad_norm": 0.21884681284427643, "learning_rate": 9.66694350865615e-05, "loss": 0.0567, "step": 32200 }, { "epoch": 1.0889482403056223, "grad_norm": 0.2886156141757965, "learning_rate": 9.666646756975757e-05, "loss": 0.0537, "step": 32210 }, { "epoch": 1.0892863179958754, "grad_norm": 0.3626093864440918, "learning_rate": 9.666349877710991e-05, "loss": 0.0557, "step": 32220 }, { "epoch": 1.0896243956861287, "grad_norm": 0.2696119546890259, "learning_rate": 9.666052870869977e-05, "loss": 0.0494, "step": 32230 }, { "epoch": 1.089962473376382, "grad_norm": 0.23980207741260529, "learning_rate": 9.665755736460828e-05, "loss": 0.0645, "step": 32240 }, { "epoch": 1.090300551066635, "grad_norm": 0.3041541278362274, "learning_rate": 9.66545847449167e-05, "loss": 0.0566, "step": 32250 }, { "epoch": 1.0906386287568883, "grad_norm": 0.37420913577079773, "learning_rate": 9.665161084970631e-05, "loss": 0.0566, "step": 32260 }, { "epoch": 1.0909767064471416, "grad_norm": 0.24355237185955048, "learning_rate": 9.66486356790584e-05, "loss": 0.0589, "step": 32270 }, { "epoch": 1.0913147841373947, "grad_norm": 0.31896692514419556, "learning_rate": 9.664565923305432e-05, "loss": 0.0607, "step": 32280 }, { "epoch": 1.091652861827648, "grad_norm": 0.24239638447761536, "learning_rate": 9.664268151177543e-05, "loss": 0.0549, "step": 32290 }, { "epoch": 1.0919909395179013, "grad_norm": 0.37432318925857544, "learning_rate": 9.663970251530314e-05, "loss": 0.0554, "step": 32300 }, { "epoch": 1.0923290172081543, "grad_norm": 0.29765039682388306, "learning_rate": 9.663672224371892e-05, "loss": 0.0635, "step": 32310 }, { "epoch": 1.0926670948984076, "grad_norm": 0.2850472033023834, "learning_rate": 9.663374069710423e-05, "loss": 0.0567, "step": 32320 }, { "epoch": 1.093005172588661, "grad_norm": 0.23249292373657227, "learning_rate": 9.66307578755406e-05, "loss": 0.0581, "step": 32330 }, { "epoch": 1.093343250278914, "grad_norm": 0.1932269185781479, "learning_rate": 9.662777377910956e-05, "loss": 0.0585, "step": 32340 }, { "epoch": 1.0936813279691673, "grad_norm": 0.23994389176368713, "learning_rate": 9.662478840789271e-05, "loss": 0.0596, "step": 32350 }, { "epoch": 1.0940194056594206, "grad_norm": 0.2808725833892822, "learning_rate": 9.662180176197164e-05, "loss": 0.0564, "step": 32360 }, { "epoch": 1.0943574833496736, "grad_norm": 0.2693062722682953, "learning_rate": 9.661881384142805e-05, "loss": 0.056, "step": 32370 }, { "epoch": 1.094695561039927, "grad_norm": 0.23766884207725525, "learning_rate": 9.66158246463436e-05, "loss": 0.061, "step": 32380 }, { "epoch": 1.0950336387301802, "grad_norm": 1.4898083209991455, "learning_rate": 9.66128341768e-05, "loss": 0.0556, "step": 32390 }, { "epoch": 1.0953717164204335, "grad_norm": 0.2606273889541626, "learning_rate": 9.660984243287903e-05, "loss": 0.0615, "step": 32400 }, { "epoch": 1.0957097941106866, "grad_norm": 0.367117315530777, "learning_rate": 9.66068494146625e-05, "loss": 0.0625, "step": 32410 }, { "epoch": 1.0960478718009399, "grad_norm": 0.28189945220947266, "learning_rate": 9.660385512223219e-05, "loss": 0.0646, "step": 32420 }, { "epoch": 1.0963859494911932, "grad_norm": 0.29257872700691223, "learning_rate": 9.660085955567e-05, "loss": 0.0585, "step": 32430 }, { "epoch": 1.0967240271814462, "grad_norm": 0.26608309149742126, "learning_rate": 9.659786271505782e-05, "loss": 0.0567, "step": 32440 }, { "epoch": 1.0970621048716995, "grad_norm": 0.2781253755092621, "learning_rate": 9.659486460047757e-05, "loss": 0.0561, "step": 32450 }, { "epoch": 1.0974001825619528, "grad_norm": 0.28707155585289, "learning_rate": 9.659186521201122e-05, "loss": 0.0572, "step": 32460 }, { "epoch": 1.0977382602522059, "grad_norm": 0.2923785150051117, "learning_rate": 9.65888645497408e-05, "loss": 0.0577, "step": 32470 }, { "epoch": 1.0980763379424592, "grad_norm": 0.2474852055311203, "learning_rate": 9.658586261374832e-05, "loss": 0.0603, "step": 32480 }, { "epoch": 1.0984144156327125, "grad_norm": 0.3511246144771576, "learning_rate": 9.658285940411584e-05, "loss": 0.0601, "step": 32490 }, { "epoch": 1.0987524933229655, "grad_norm": 0.5754612684249878, "learning_rate": 9.657985492092551e-05, "loss": 0.061, "step": 32500 }, { "epoch": 1.0990905710132188, "grad_norm": 0.2272210270166397, "learning_rate": 9.657684916425944e-05, "loss": 0.0593, "step": 32510 }, { "epoch": 1.0994286487034721, "grad_norm": 0.2898885905742645, "learning_rate": 9.65738421341998e-05, "loss": 0.0539, "step": 32520 }, { "epoch": 1.0997667263937252, "grad_norm": 0.24826237559318542, "learning_rate": 9.657083383082881e-05, "loss": 0.0595, "step": 32530 }, { "epoch": 1.1001048040839785, "grad_norm": 0.20462806522846222, "learning_rate": 9.656782425422874e-05, "loss": 0.0608, "step": 32540 }, { "epoch": 1.1004428817742318, "grad_norm": 0.1732936054468155, "learning_rate": 9.656481340448184e-05, "loss": 0.0504, "step": 32550 }, { "epoch": 1.1007809594644848, "grad_norm": 0.17703105509281158, "learning_rate": 9.656180128167042e-05, "loss": 0.0607, "step": 32560 }, { "epoch": 1.1011190371547381, "grad_norm": 0.21976083517074585, "learning_rate": 9.655878788587687e-05, "loss": 0.0632, "step": 32570 }, { "epoch": 1.1014571148449914, "grad_norm": 0.3670908808708191, "learning_rate": 9.655577321718353e-05, "loss": 0.0581, "step": 32580 }, { "epoch": 1.1017951925352447, "grad_norm": 0.22553564608097076, "learning_rate": 9.655275727567285e-05, "loss": 0.0539, "step": 32590 }, { "epoch": 1.1021332702254978, "grad_norm": 0.26414069533348083, "learning_rate": 9.654974006142727e-05, "loss": 0.0597, "step": 32600 }, { "epoch": 1.102471347915751, "grad_norm": 0.27421340346336365, "learning_rate": 9.654672157452929e-05, "loss": 0.0529, "step": 32610 }, { "epoch": 1.1028094256060044, "grad_norm": 0.39794325828552246, "learning_rate": 9.654370181506143e-05, "loss": 0.06, "step": 32620 }, { "epoch": 1.1031475032962574, "grad_norm": 0.5050188899040222, "learning_rate": 9.654068078310625e-05, "loss": 0.0643, "step": 32630 }, { "epoch": 1.1034855809865107, "grad_norm": 0.24323244392871857, "learning_rate": 9.653765847874634e-05, "loss": 0.0572, "step": 32640 }, { "epoch": 1.103823658676764, "grad_norm": 0.31684210896492004, "learning_rate": 9.653463490206432e-05, "loss": 0.059, "step": 32650 }, { "epoch": 1.104161736367017, "grad_norm": 0.3406434953212738, "learning_rate": 9.653161005314288e-05, "loss": 0.0555, "step": 32660 }, { "epoch": 1.1044998140572704, "grad_norm": 0.2962169945240021, "learning_rate": 9.65285839320647e-05, "loss": 0.0657, "step": 32670 }, { "epoch": 1.1048378917475237, "grad_norm": 0.3397774398326874, "learning_rate": 9.65255565389125e-05, "loss": 0.0592, "step": 32680 }, { "epoch": 1.1051759694377767, "grad_norm": 0.2609814405441284, "learning_rate": 9.652252787376908e-05, "loss": 0.0569, "step": 32690 }, { "epoch": 1.10551404712803, "grad_norm": 0.23512624204158783, "learning_rate": 9.651949793671722e-05, "loss": 0.0545, "step": 32700 }, { "epoch": 1.1058521248182833, "grad_norm": 0.5010001063346863, "learning_rate": 9.651646672783976e-05, "loss": 0.0646, "step": 32710 }, { "epoch": 1.1061902025085364, "grad_norm": 0.38665446639060974, "learning_rate": 9.651343424721957e-05, "loss": 0.0628, "step": 32720 }, { "epoch": 1.1065282801987897, "grad_norm": 0.25653207302093506, "learning_rate": 9.651040049493956e-05, "loss": 0.0593, "step": 32730 }, { "epoch": 1.106866357889043, "grad_norm": 0.3251650035381317, "learning_rate": 9.65073654710827e-05, "loss": 0.06, "step": 32740 }, { "epoch": 1.107204435579296, "grad_norm": 0.2884964942932129, "learning_rate": 9.650432917573192e-05, "loss": 0.0586, "step": 32750 }, { "epoch": 1.1075425132695493, "grad_norm": 0.2814955711364746, "learning_rate": 9.650129160897025e-05, "loss": 0.0514, "step": 32760 }, { "epoch": 1.1078805909598026, "grad_norm": 0.31174200773239136, "learning_rate": 9.649825277088073e-05, "loss": 0.0595, "step": 32770 }, { "epoch": 1.1082186686500557, "grad_norm": 0.828022301197052, "learning_rate": 9.649521266154646e-05, "loss": 0.056, "step": 32780 }, { "epoch": 1.108556746340309, "grad_norm": 0.29226526618003845, "learning_rate": 9.649217128105055e-05, "loss": 0.058, "step": 32790 }, { "epoch": 1.1088948240305623, "grad_norm": 0.2532488703727722, "learning_rate": 9.648912862947613e-05, "loss": 0.0509, "step": 32800 }, { "epoch": 1.1092329017208153, "grad_norm": 0.22431789338588715, "learning_rate": 9.64860847069064e-05, "loss": 0.0596, "step": 32810 }, { "epoch": 1.1095709794110686, "grad_norm": 0.24088937044143677, "learning_rate": 9.648303951342459e-05, "loss": 0.0584, "step": 32820 }, { "epoch": 1.109909057101322, "grad_norm": 0.4769197702407837, "learning_rate": 9.647999304911392e-05, "loss": 0.0601, "step": 32830 }, { "epoch": 1.110247134791575, "grad_norm": 0.34593692421913147, "learning_rate": 9.647694531405773e-05, "loss": 0.0636, "step": 32840 }, { "epoch": 1.1105852124818283, "grad_norm": 0.30619382858276367, "learning_rate": 9.64738963083393e-05, "loss": 0.0616, "step": 32850 }, { "epoch": 1.1109232901720816, "grad_norm": 0.2412242889404297, "learning_rate": 9.6470846032042e-05, "loss": 0.0586, "step": 32860 }, { "epoch": 1.1112613678623346, "grad_norm": 0.25585201382637024, "learning_rate": 9.646779448524923e-05, "loss": 0.0579, "step": 32870 }, { "epoch": 1.111599445552588, "grad_norm": 0.30249911546707153, "learning_rate": 9.646474166804442e-05, "loss": 0.0572, "step": 32880 }, { "epoch": 1.1119375232428412, "grad_norm": 0.2573614716529846, "learning_rate": 9.646168758051104e-05, "loss": 0.0548, "step": 32890 }, { "epoch": 1.1122756009330945, "grad_norm": 0.21661429107189178, "learning_rate": 9.645863222273256e-05, "loss": 0.0657, "step": 32900 }, { "epoch": 1.1126136786233476, "grad_norm": 0.39221200346946716, "learning_rate": 9.645557559479253e-05, "loss": 0.0575, "step": 32910 }, { "epoch": 1.1129517563136009, "grad_norm": 0.30713340640068054, "learning_rate": 9.645251769677452e-05, "loss": 0.062, "step": 32920 }, { "epoch": 1.1132898340038542, "grad_norm": 0.4606127440929413, "learning_rate": 9.644945852876213e-05, "loss": 0.0615, "step": 32930 }, { "epoch": 1.1136279116941072, "grad_norm": 0.24532125890254974, "learning_rate": 9.6446398090839e-05, "loss": 0.0558, "step": 32940 }, { "epoch": 1.1139659893843605, "grad_norm": 0.34204018115997314, "learning_rate": 9.644333638308878e-05, "loss": 0.06, "step": 32950 }, { "epoch": 1.1143040670746138, "grad_norm": 0.21701133251190186, "learning_rate": 9.644027340559519e-05, "loss": 0.0647, "step": 32960 }, { "epoch": 1.114642144764867, "grad_norm": 0.2793455123901367, "learning_rate": 9.643720915844199e-05, "loss": 0.0539, "step": 32970 }, { "epoch": 1.1149802224551202, "grad_norm": 0.2134568840265274, "learning_rate": 9.643414364171292e-05, "loss": 0.0599, "step": 32980 }, { "epoch": 1.1153183001453735, "grad_norm": 0.3551085889339447, "learning_rate": 9.643107685549182e-05, "loss": 0.057, "step": 32990 }, { "epoch": 1.1156563778356265, "grad_norm": 0.32601287961006165, "learning_rate": 9.642800879986251e-05, "loss": 0.062, "step": 33000 }, { "epoch": 1.1159944555258798, "grad_norm": 0.3008292615413666, "learning_rate": 9.642493947490889e-05, "loss": 0.0512, "step": 33010 }, { "epoch": 1.1163325332161331, "grad_norm": 0.23528335988521576, "learning_rate": 9.642186888071486e-05, "loss": 0.0499, "step": 33020 }, { "epoch": 1.1166706109063862, "grad_norm": 0.31222233176231384, "learning_rate": 9.641879701736438e-05, "loss": 0.06, "step": 33030 }, { "epoch": 1.1170086885966395, "grad_norm": 0.384847491979599, "learning_rate": 9.641572388494143e-05, "loss": 0.0545, "step": 33040 }, { "epoch": 1.1173467662868928, "grad_norm": 0.366374671459198, "learning_rate": 9.641264948353001e-05, "loss": 0.0516, "step": 33050 }, { "epoch": 1.1176848439771458, "grad_norm": 0.3567977547645569, "learning_rate": 9.640957381321422e-05, "loss": 0.0561, "step": 33060 }, { "epoch": 1.1180229216673991, "grad_norm": 0.3021322190761566, "learning_rate": 9.64064968740781e-05, "loss": 0.051, "step": 33070 }, { "epoch": 1.1183609993576524, "grad_norm": 0.3985152840614319, "learning_rate": 9.64034186662058e-05, "loss": 0.0578, "step": 33080 }, { "epoch": 1.1186990770479057, "grad_norm": 0.23492272198200226, "learning_rate": 9.640033918968143e-05, "loss": 0.0635, "step": 33090 }, { "epoch": 1.1190371547381588, "grad_norm": 0.23347151279449463, "learning_rate": 9.639725844458926e-05, "loss": 0.0547, "step": 33100 }, { "epoch": 1.119375232428412, "grad_norm": 0.3741791546344757, "learning_rate": 9.639417643101345e-05, "loss": 0.0581, "step": 33110 }, { "epoch": 1.1197133101186654, "grad_norm": 0.365445613861084, "learning_rate": 9.63910931490383e-05, "loss": 0.0618, "step": 33120 }, { "epoch": 1.1200513878089184, "grad_norm": 0.22848589718341827, "learning_rate": 9.638800859874808e-05, "loss": 0.0617, "step": 33130 }, { "epoch": 1.1203894654991717, "grad_norm": 0.2492479532957077, "learning_rate": 9.638492278022715e-05, "loss": 0.053, "step": 33140 }, { "epoch": 1.120727543189425, "grad_norm": 0.351685494184494, "learning_rate": 9.638183569355984e-05, "loss": 0.0645, "step": 33150 }, { "epoch": 1.121065620879678, "grad_norm": 0.22603270411491394, "learning_rate": 9.637874733883057e-05, "loss": 0.053, "step": 33160 }, { "epoch": 1.1214036985699314, "grad_norm": 0.28535425662994385, "learning_rate": 9.637565771612378e-05, "loss": 0.0574, "step": 33170 }, { "epoch": 1.1217417762601847, "grad_norm": 0.21809250116348267, "learning_rate": 9.637256682552391e-05, "loss": 0.0566, "step": 33180 }, { "epoch": 1.1220798539504377, "grad_norm": 0.19990628957748413, "learning_rate": 9.636947466711551e-05, "loss": 0.0671, "step": 33190 }, { "epoch": 1.122417931640691, "grad_norm": 0.22132636606693268, "learning_rate": 9.636638124098308e-05, "loss": 0.0599, "step": 33200 }, { "epoch": 1.1227560093309443, "grad_norm": 0.29064974188804626, "learning_rate": 9.636328654721121e-05, "loss": 0.0618, "step": 33210 }, { "epoch": 1.1230940870211974, "grad_norm": 0.2044263482093811, "learning_rate": 9.63601905858845e-05, "loss": 0.0588, "step": 33220 }, { "epoch": 1.1234321647114507, "grad_norm": 0.2420269101858139, "learning_rate": 9.63570933570876e-05, "loss": 0.0601, "step": 33230 }, { "epoch": 1.123770242401704, "grad_norm": 0.2947719395160675, "learning_rate": 9.635399486090518e-05, "loss": 0.0539, "step": 33240 }, { "epoch": 1.124108320091957, "grad_norm": 0.29531845450401306, "learning_rate": 9.635089509742196e-05, "loss": 0.0615, "step": 33250 }, { "epoch": 1.1244463977822103, "grad_norm": 0.26350486278533936, "learning_rate": 9.634779406672267e-05, "loss": 0.0547, "step": 33260 }, { "epoch": 1.1247844754724636, "grad_norm": 0.3279115855693817, "learning_rate": 9.634469176889211e-05, "loss": 0.0564, "step": 33270 }, { "epoch": 1.1251225531627167, "grad_norm": 0.6890733242034912, "learning_rate": 9.634158820401508e-05, "loss": 0.0548, "step": 33280 }, { "epoch": 1.12546063085297, "grad_norm": 0.3292968273162842, "learning_rate": 9.633848337217646e-05, "loss": 0.0608, "step": 33290 }, { "epoch": 1.1257987085432233, "grad_norm": 0.4733574092388153, "learning_rate": 9.63353772734611e-05, "loss": 0.0529, "step": 33300 }, { "epoch": 1.1261367862334764, "grad_norm": 0.28832364082336426, "learning_rate": 9.633226990795392e-05, "loss": 0.0516, "step": 33310 }, { "epoch": 1.1264748639237296, "grad_norm": 0.36945706605911255, "learning_rate": 9.632916127573989e-05, "loss": 0.0664, "step": 33320 }, { "epoch": 1.126812941613983, "grad_norm": 0.24931225180625916, "learning_rate": 9.632605137690402e-05, "loss": 0.0635, "step": 33330 }, { "epoch": 1.127151019304236, "grad_norm": 0.21698719263076782, "learning_rate": 9.63229402115313e-05, "loss": 0.0582, "step": 33340 }, { "epoch": 1.1274890969944893, "grad_norm": 0.25475186109542847, "learning_rate": 9.631982777970678e-05, "loss": 0.0602, "step": 33350 }, { "epoch": 1.1278271746847426, "grad_norm": 0.31452521681785583, "learning_rate": 9.631671408151559e-05, "loss": 0.0523, "step": 33360 }, { "epoch": 1.1281652523749957, "grad_norm": 0.5594089031219482, "learning_rate": 9.631359911704281e-05, "loss": 0.0517, "step": 33370 }, { "epoch": 1.128503330065249, "grad_norm": 0.2056441605091095, "learning_rate": 9.631048288637366e-05, "loss": 0.0567, "step": 33380 }, { "epoch": 1.1288414077555022, "grad_norm": 0.4665636122226715, "learning_rate": 9.630736538959328e-05, "loss": 0.0596, "step": 33390 }, { "epoch": 1.1291794854457555, "grad_norm": 0.27286773920059204, "learning_rate": 9.630424662678695e-05, "loss": 0.0556, "step": 33400 }, { "epoch": 1.1295175631360086, "grad_norm": 0.17027819156646729, "learning_rate": 9.630112659803991e-05, "loss": 0.053, "step": 33410 }, { "epoch": 1.1298556408262619, "grad_norm": 0.29866528511047363, "learning_rate": 9.629800530343745e-05, "loss": 0.054, "step": 33420 }, { "epoch": 1.1301937185165152, "grad_norm": 0.24548497796058655, "learning_rate": 9.629488274306493e-05, "loss": 0.0624, "step": 33430 }, { "epoch": 1.1305317962067682, "grad_norm": 0.18488748371601105, "learning_rate": 9.62917589170077e-05, "loss": 0.0605, "step": 33440 }, { "epoch": 1.1308698738970215, "grad_norm": 0.30133309960365295, "learning_rate": 9.628863382535118e-05, "loss": 0.0592, "step": 33450 }, { "epoch": 1.1312079515872748, "grad_norm": 0.2889573276042938, "learning_rate": 9.628550746818079e-05, "loss": 0.0593, "step": 33460 }, { "epoch": 1.131546029277528, "grad_norm": 0.2685140371322632, "learning_rate": 9.628237984558203e-05, "loss": 0.0592, "step": 33470 }, { "epoch": 1.1318841069677812, "grad_norm": 0.47402679920196533, "learning_rate": 9.627925095764037e-05, "loss": 0.0583, "step": 33480 }, { "epoch": 1.1322221846580345, "grad_norm": 0.25560736656188965, "learning_rate": 9.62761208044414e-05, "loss": 0.0598, "step": 33490 }, { "epoch": 1.1325602623482875, "grad_norm": 0.26853108406066895, "learning_rate": 9.627298938607066e-05, "loss": 0.0533, "step": 33500 }, { "epoch": 1.1328983400385408, "grad_norm": 0.26585501432418823, "learning_rate": 9.626985670261377e-05, "loss": 0.0534, "step": 33510 }, { "epoch": 1.1332364177287941, "grad_norm": 0.2920114994049072, "learning_rate": 9.626672275415639e-05, "loss": 0.0535, "step": 33520 }, { "epoch": 1.1335744954190472, "grad_norm": 0.21500571072101593, "learning_rate": 9.626358754078418e-05, "loss": 0.0525, "step": 33530 }, { "epoch": 1.1339125731093005, "grad_norm": 0.16352249681949615, "learning_rate": 9.626045106258288e-05, "loss": 0.0551, "step": 33540 }, { "epoch": 1.1342506507995538, "grad_norm": 0.42305585741996765, "learning_rate": 9.625731331963822e-05, "loss": 0.0632, "step": 33550 }, { "epoch": 1.134588728489807, "grad_norm": 0.1797967404127121, "learning_rate": 9.6254174312036e-05, "loss": 0.0618, "step": 33560 }, { "epoch": 1.1349268061800601, "grad_norm": 0.24761910736560822, "learning_rate": 9.625103403986202e-05, "loss": 0.0587, "step": 33570 }, { "epoch": 1.1352648838703134, "grad_norm": 0.6767323613166809, "learning_rate": 9.624789250320214e-05, "loss": 0.0578, "step": 33580 }, { "epoch": 1.1356029615605667, "grad_norm": 0.22795166075229645, "learning_rate": 9.624474970214227e-05, "loss": 0.055, "step": 33590 }, { "epoch": 1.1359410392508198, "grad_norm": 0.27907902002334595, "learning_rate": 9.62416056367683e-05, "loss": 0.0535, "step": 33600 }, { "epoch": 1.136279116941073, "grad_norm": 0.288259357213974, "learning_rate": 9.623846030716623e-05, "loss": 0.0558, "step": 33610 }, { "epoch": 1.1366171946313264, "grad_norm": 0.26965251564979553, "learning_rate": 9.623531371342201e-05, "loss": 0.058, "step": 33620 }, { "epoch": 1.1369552723215794, "grad_norm": 0.2816045582294464, "learning_rate": 9.623216585562167e-05, "loss": 0.0532, "step": 33630 }, { "epoch": 1.1372933500118327, "grad_norm": 0.3716954290866852, "learning_rate": 9.62290167338513e-05, "loss": 0.0614, "step": 33640 }, { "epoch": 1.137631427702086, "grad_norm": 0.2718285918235779, "learning_rate": 9.622586634819697e-05, "loss": 0.0631, "step": 33650 }, { "epoch": 1.137969505392339, "grad_norm": 0.3703029155731201, "learning_rate": 9.622271469874484e-05, "loss": 0.0564, "step": 33660 }, { "epoch": 1.1383075830825924, "grad_norm": 0.25519299507141113, "learning_rate": 9.621956178558104e-05, "loss": 0.0544, "step": 33670 }, { "epoch": 1.1386456607728457, "grad_norm": 0.2901768088340759, "learning_rate": 9.621640760879179e-05, "loss": 0.0604, "step": 33680 }, { "epoch": 1.1389837384630987, "grad_norm": 0.26099589467048645, "learning_rate": 9.621325216846331e-05, "loss": 0.0616, "step": 33690 }, { "epoch": 1.139321816153352, "grad_norm": 0.2080443650484085, "learning_rate": 9.621009546468188e-05, "loss": 0.0552, "step": 33700 }, { "epoch": 1.1396598938436053, "grad_norm": 0.253112256526947, "learning_rate": 9.62069374975338e-05, "loss": 0.0532, "step": 33710 }, { "epoch": 1.1399979715338584, "grad_norm": 0.19288966059684753, "learning_rate": 9.620377826710541e-05, "loss": 0.0533, "step": 33720 }, { "epoch": 1.1403360492241117, "grad_norm": 0.31607404351234436, "learning_rate": 9.620061777348308e-05, "loss": 0.0558, "step": 33730 }, { "epoch": 1.140674126914365, "grad_norm": 0.30935442447662354, "learning_rate": 9.619745601675324e-05, "loss": 0.0593, "step": 33740 }, { "epoch": 1.141012204604618, "grad_norm": 0.3298223912715912, "learning_rate": 9.619429299700226e-05, "loss": 0.0576, "step": 33750 }, { "epoch": 1.1413502822948713, "grad_norm": 0.23758676648139954, "learning_rate": 9.61911287143167e-05, "loss": 0.0592, "step": 33760 }, { "epoch": 1.1416883599851246, "grad_norm": 0.24149160087108612, "learning_rate": 9.618796316878302e-05, "loss": 0.0504, "step": 33770 }, { "epoch": 1.1420264376753777, "grad_norm": 0.20410165190696716, "learning_rate": 9.618479636048778e-05, "loss": 0.0586, "step": 33780 }, { "epoch": 1.142364515365631, "grad_norm": 0.43201032280921936, "learning_rate": 9.618162828951757e-05, "loss": 0.0617, "step": 33790 }, { "epoch": 1.1427025930558843, "grad_norm": 0.2839832305908203, "learning_rate": 9.617845895595898e-05, "loss": 0.0596, "step": 33800 }, { "epoch": 1.1430406707461374, "grad_norm": 0.21613039076328278, "learning_rate": 9.617528835989867e-05, "loss": 0.0534, "step": 33810 }, { "epoch": 1.1433787484363906, "grad_norm": 0.27792972326278687, "learning_rate": 9.617211650142332e-05, "loss": 0.0596, "step": 33820 }, { "epoch": 1.143716826126644, "grad_norm": 0.26905587315559387, "learning_rate": 9.616894338061967e-05, "loss": 0.0591, "step": 33830 }, { "epoch": 1.144054903816897, "grad_norm": 0.2462191879749298, "learning_rate": 9.616576899757444e-05, "loss": 0.0567, "step": 33840 }, { "epoch": 1.1443929815071503, "grad_norm": 0.26556435227394104, "learning_rate": 9.616259335237443e-05, "loss": 0.0514, "step": 33850 }, { "epoch": 1.1447310591974036, "grad_norm": 0.19680622220039368, "learning_rate": 9.615941644510647e-05, "loss": 0.0569, "step": 33860 }, { "epoch": 1.1450691368876569, "grad_norm": 0.5309991836547852, "learning_rate": 9.615623827585738e-05, "loss": 0.0585, "step": 33870 }, { "epoch": 1.14540721457791, "grad_norm": 0.2638305425643921, "learning_rate": 9.615305884471411e-05, "loss": 0.0531, "step": 33880 }, { "epoch": 1.1457452922681632, "grad_norm": 0.17844261229038239, "learning_rate": 9.614987815176352e-05, "loss": 0.0554, "step": 33890 }, { "epoch": 1.1460833699584165, "grad_norm": 0.18152883648872375, "learning_rate": 9.614669619709262e-05, "loss": 0.0609, "step": 33900 }, { "epoch": 1.1464214476486696, "grad_norm": 0.31256672739982605, "learning_rate": 9.614351298078836e-05, "loss": 0.058, "step": 33910 }, { "epoch": 1.146759525338923, "grad_norm": 0.17580044269561768, "learning_rate": 9.614032850293782e-05, "loss": 0.0565, "step": 33920 }, { "epoch": 1.1470976030291762, "grad_norm": 0.24028880894184113, "learning_rate": 9.613714276362802e-05, "loss": 0.0601, "step": 33930 }, { "epoch": 1.1474356807194293, "grad_norm": 0.26411497592926025, "learning_rate": 9.613395576294606e-05, "loss": 0.0612, "step": 33940 }, { "epoch": 1.1477737584096825, "grad_norm": 0.3014588952064514, "learning_rate": 9.613076750097909e-05, "loss": 0.0575, "step": 33950 }, { "epoch": 1.1481118360999358, "grad_norm": 0.24953924119472504, "learning_rate": 9.612757797781426e-05, "loss": 0.0569, "step": 33960 }, { "epoch": 1.148449913790189, "grad_norm": 0.284837007522583, "learning_rate": 9.612438719353878e-05, "loss": 0.0709, "step": 33970 }, { "epoch": 1.1487879914804422, "grad_norm": 0.20381273329257965, "learning_rate": 9.612119514823989e-05, "loss": 0.06, "step": 33980 }, { "epoch": 1.1491260691706955, "grad_norm": 0.19244150817394257, "learning_rate": 9.611800184200484e-05, "loss": 0.0592, "step": 33990 }, { "epoch": 1.1494641468609486, "grad_norm": 0.27823328971862793, "learning_rate": 9.611480727492095e-05, "loss": 0.0585, "step": 34000 }, { "epoch": 1.1498022245512018, "grad_norm": 0.23347294330596924, "learning_rate": 9.611161144707555e-05, "loss": 0.0542, "step": 34010 }, { "epoch": 1.1501403022414551, "grad_norm": 0.46530479192733765, "learning_rate": 9.610841435855603e-05, "loss": 0.0525, "step": 34020 }, { "epoch": 1.1504783799317084, "grad_norm": 0.21147556602954865, "learning_rate": 9.610521600944978e-05, "loss": 0.0612, "step": 34030 }, { "epoch": 1.1508164576219615, "grad_norm": 0.33674344420433044, "learning_rate": 9.610201639984425e-05, "loss": 0.0588, "step": 34040 }, { "epoch": 1.1511545353122148, "grad_norm": 0.29306188225746155, "learning_rate": 9.60988155298269e-05, "loss": 0.065, "step": 34050 }, { "epoch": 1.151492613002468, "grad_norm": 0.3889126777648926, "learning_rate": 9.609561339948524e-05, "loss": 0.0598, "step": 34060 }, { "epoch": 1.1518306906927211, "grad_norm": 0.22052201628684998, "learning_rate": 9.609241000890685e-05, "loss": 0.0563, "step": 34070 }, { "epoch": 1.1521687683829744, "grad_norm": 0.25277480483055115, "learning_rate": 9.608920535817927e-05, "loss": 0.0608, "step": 34080 }, { "epoch": 1.1525068460732277, "grad_norm": 0.2565675377845764, "learning_rate": 9.608599944739014e-05, "loss": 0.0632, "step": 34090 }, { "epoch": 1.1528449237634808, "grad_norm": 0.3189801871776581, "learning_rate": 9.608279227662708e-05, "loss": 0.063, "step": 34100 }, { "epoch": 1.153183001453734, "grad_norm": 2.2032854557037354, "learning_rate": 9.607958384597782e-05, "loss": 0.0607, "step": 34110 }, { "epoch": 1.1535210791439874, "grad_norm": 0.5112471580505371, "learning_rate": 9.607637415553004e-05, "loss": 0.0586, "step": 34120 }, { "epoch": 1.1538591568342405, "grad_norm": 0.22091956436634064, "learning_rate": 9.60731632053715e-05, "loss": 0.0548, "step": 34130 }, { "epoch": 1.1541972345244937, "grad_norm": 0.483533650636673, "learning_rate": 9.606995099558998e-05, "loss": 0.0629, "step": 34140 }, { "epoch": 1.154535312214747, "grad_norm": 0.3287438452243805, "learning_rate": 9.606673752627331e-05, "loss": 0.0523, "step": 34150 }, { "epoch": 1.154873389905, "grad_norm": 0.34263476729393005, "learning_rate": 9.606352279750934e-05, "loss": 0.0504, "step": 34160 }, { "epoch": 1.1552114675952534, "grad_norm": 0.34871238470077515, "learning_rate": 9.606030680938597e-05, "loss": 0.0595, "step": 34170 }, { "epoch": 1.1555495452855067, "grad_norm": 0.28101062774658203, "learning_rate": 9.60570895619911e-05, "loss": 0.0534, "step": 34180 }, { "epoch": 1.1558876229757598, "grad_norm": 0.5068068504333496, "learning_rate": 9.605387105541272e-05, "loss": 0.0555, "step": 34190 }, { "epoch": 1.156225700666013, "grad_norm": 0.22095218300819397, "learning_rate": 9.605065128973882e-05, "loss": 0.0612, "step": 34200 }, { "epoch": 1.1565637783562663, "grad_norm": 0.30551430583000183, "learning_rate": 9.60474302650574e-05, "loss": 0.0617, "step": 34210 }, { "epoch": 1.1569018560465194, "grad_norm": 0.3252962529659271, "learning_rate": 9.604420798145652e-05, "loss": 0.0569, "step": 34220 }, { "epoch": 1.1572399337367727, "grad_norm": 0.28480568528175354, "learning_rate": 9.604098443902431e-05, "loss": 0.0553, "step": 34230 }, { "epoch": 1.157578011427026, "grad_norm": 0.31041067838668823, "learning_rate": 9.603775963784888e-05, "loss": 0.0552, "step": 34240 }, { "epoch": 1.157916089117279, "grad_norm": 0.30789250135421753, "learning_rate": 9.60345335780184e-05, "loss": 0.0573, "step": 34250 }, { "epoch": 1.1582541668075323, "grad_norm": 0.39186757802963257, "learning_rate": 9.603130625962105e-05, "loss": 0.0566, "step": 34260 }, { "epoch": 1.1585922444977856, "grad_norm": 0.7177406549453735, "learning_rate": 9.602807768274509e-05, "loss": 0.0556, "step": 34270 }, { "epoch": 1.1589303221880387, "grad_norm": 0.21195855736732483, "learning_rate": 9.602484784747879e-05, "loss": 0.0613, "step": 34280 }, { "epoch": 1.159268399878292, "grad_norm": 0.251671701669693, "learning_rate": 9.602161675391042e-05, "loss": 0.0643, "step": 34290 }, { "epoch": 1.1596064775685453, "grad_norm": 0.35062161087989807, "learning_rate": 9.601838440212835e-05, "loss": 0.0627, "step": 34300 }, { "epoch": 1.1599445552587984, "grad_norm": 0.3272019028663635, "learning_rate": 9.601515079222094e-05, "loss": 0.061, "step": 34310 }, { "epoch": 1.1602826329490517, "grad_norm": 0.33764222264289856, "learning_rate": 9.601191592427658e-05, "loss": 0.0632, "step": 34320 }, { "epoch": 1.160620710639305, "grad_norm": 0.4223364591598511, "learning_rate": 9.600867979838374e-05, "loss": 0.0581, "step": 34330 }, { "epoch": 1.160958788329558, "grad_norm": 0.2449866384267807, "learning_rate": 9.600544241463088e-05, "loss": 0.0573, "step": 34340 }, { "epoch": 1.1612968660198113, "grad_norm": 0.28745806217193604, "learning_rate": 9.60022037731065e-05, "loss": 0.0606, "step": 34350 }, { "epoch": 1.1616349437100646, "grad_norm": 0.3310845196247101, "learning_rate": 9.599896387389915e-05, "loss": 0.0595, "step": 34360 }, { "epoch": 1.1619730214003179, "grad_norm": 0.3695114254951477, "learning_rate": 9.599572271709741e-05, "loss": 0.0572, "step": 34370 }, { "epoch": 1.162311099090571, "grad_norm": 0.3004867136478424, "learning_rate": 9.599248030278989e-05, "loss": 0.0643, "step": 34380 }, { "epoch": 1.1626491767808242, "grad_norm": 0.26666200160980225, "learning_rate": 9.598923663106524e-05, "loss": 0.0624, "step": 34390 }, { "epoch": 1.1629872544710775, "grad_norm": 0.5813899636268616, "learning_rate": 9.598599170201213e-05, "loss": 0.0626, "step": 34400 }, { "epoch": 1.1633253321613306, "grad_norm": 0.33837395906448364, "learning_rate": 9.59827455157193e-05, "loss": 0.059, "step": 34410 }, { "epoch": 1.163663409851584, "grad_norm": 0.331827312707901, "learning_rate": 9.597949807227546e-05, "loss": 0.0565, "step": 34420 }, { "epoch": 1.1640014875418372, "grad_norm": 0.24007092416286469, "learning_rate": 9.597624937176943e-05, "loss": 0.0558, "step": 34430 }, { "epoch": 1.1643395652320903, "grad_norm": 0.2891419529914856, "learning_rate": 9.597299941429002e-05, "loss": 0.0637, "step": 34440 }, { "epoch": 1.1646776429223435, "grad_norm": 0.32633715867996216, "learning_rate": 9.596974819992606e-05, "loss": 0.0566, "step": 34450 }, { "epoch": 1.1650157206125968, "grad_norm": 0.2411215603351593, "learning_rate": 9.596649572876648e-05, "loss": 0.0526, "step": 34460 }, { "epoch": 1.16535379830285, "grad_norm": 0.24233990907669067, "learning_rate": 9.596324200090016e-05, "loss": 0.057, "step": 34470 }, { "epoch": 1.1656918759931032, "grad_norm": 0.27006980776786804, "learning_rate": 9.595998701641606e-05, "loss": 0.0596, "step": 34480 }, { "epoch": 1.1660299536833565, "grad_norm": 0.39965808391571045, "learning_rate": 9.595673077540319e-05, "loss": 0.0565, "step": 34490 }, { "epoch": 1.1663680313736096, "grad_norm": 0.33104604482650757, "learning_rate": 9.595347327795058e-05, "loss": 0.0547, "step": 34500 }, { "epoch": 1.1667061090638629, "grad_norm": 0.700973391532898, "learning_rate": 9.595021452414727e-05, "loss": 0.0595, "step": 34510 }, { "epoch": 1.1670441867541161, "grad_norm": 0.4116021394729614, "learning_rate": 9.594695451408235e-05, "loss": 0.0585, "step": 34520 }, { "epoch": 1.1673822644443694, "grad_norm": 0.3603658676147461, "learning_rate": 9.594369324784495e-05, "loss": 0.0564, "step": 34530 }, { "epoch": 1.1677203421346225, "grad_norm": 0.2814156115055084, "learning_rate": 9.594043072552425e-05, "loss": 0.0536, "step": 34540 }, { "epoch": 1.1680584198248758, "grad_norm": 0.29186052083969116, "learning_rate": 9.593716694720942e-05, "loss": 0.0596, "step": 34550 }, { "epoch": 1.168396497515129, "grad_norm": 0.2920096814632416, "learning_rate": 9.59339019129897e-05, "loss": 0.0549, "step": 34560 }, { "epoch": 1.1687345752053822, "grad_norm": 0.3224203586578369, "learning_rate": 9.593063562295437e-05, "loss": 0.0558, "step": 34570 }, { "epoch": 1.1690726528956354, "grad_norm": 0.41581326723098755, "learning_rate": 9.592736807719271e-05, "loss": 0.0532, "step": 34580 }, { "epoch": 1.1694107305858887, "grad_norm": 0.28106117248535156, "learning_rate": 9.592409927579407e-05, "loss": 0.0623, "step": 34590 }, { "epoch": 1.1697488082761418, "grad_norm": 0.32932034134864807, "learning_rate": 9.592082921884782e-05, "loss": 0.0604, "step": 34600 }, { "epoch": 1.170086885966395, "grad_norm": 0.3437102735042572, "learning_rate": 9.591755790644332e-05, "loss": 0.0554, "step": 34610 }, { "epoch": 1.1704249636566484, "grad_norm": 0.24264514446258545, "learning_rate": 9.591428533867007e-05, "loss": 0.0598, "step": 34620 }, { "epoch": 1.1707630413469015, "grad_norm": 0.21527910232543945, "learning_rate": 9.591101151561748e-05, "loss": 0.0554, "step": 34630 }, { "epoch": 1.1711011190371547, "grad_norm": 0.2318674921989441, "learning_rate": 9.59077364373751e-05, "loss": 0.0544, "step": 34640 }, { "epoch": 1.171439196727408, "grad_norm": 0.3703412413597107, "learning_rate": 9.590446010403245e-05, "loss": 0.0593, "step": 34650 }, { "epoch": 1.171777274417661, "grad_norm": 0.6496084928512573, "learning_rate": 9.59011825156791e-05, "loss": 0.0527, "step": 34660 }, { "epoch": 1.1721153521079144, "grad_norm": 0.21646283566951752, "learning_rate": 9.589790367240465e-05, "loss": 0.0519, "step": 34670 }, { "epoch": 1.1724534297981677, "grad_norm": 0.3277036249637604, "learning_rate": 9.589462357429878e-05, "loss": 0.0611, "step": 34680 }, { "epoch": 1.1727915074884208, "grad_norm": 0.41378161311149597, "learning_rate": 9.589134222145114e-05, "loss": 0.0567, "step": 34690 }, { "epoch": 1.173129585178674, "grad_norm": 0.24443532526493073, "learning_rate": 9.588805961395144e-05, "loss": 0.0553, "step": 34700 }, { "epoch": 1.1734676628689273, "grad_norm": 0.21569690108299255, "learning_rate": 9.588477575188943e-05, "loss": 0.0614, "step": 34710 }, { "epoch": 1.1738057405591804, "grad_norm": 0.2062235176563263, "learning_rate": 9.58814906353549e-05, "loss": 0.0548, "step": 34720 }, { "epoch": 1.1741438182494337, "grad_norm": 0.29149875044822693, "learning_rate": 9.587820426443763e-05, "loss": 0.057, "step": 34730 }, { "epoch": 1.174481895939687, "grad_norm": 0.2530285120010376, "learning_rate": 9.58749166392275e-05, "loss": 0.0557, "step": 34740 }, { "epoch": 1.17481997362994, "grad_norm": 0.23019906878471375, "learning_rate": 9.587162775981439e-05, "loss": 0.0615, "step": 34750 }, { "epoch": 1.1751580513201934, "grad_norm": 0.2747979760169983, "learning_rate": 9.586833762628821e-05, "loss": 0.0593, "step": 34760 }, { "epoch": 1.1754961290104466, "grad_norm": 0.28163471817970276, "learning_rate": 9.58650462387389e-05, "loss": 0.0579, "step": 34770 }, { "epoch": 1.1758342067006997, "grad_norm": 0.31446129083633423, "learning_rate": 9.586175359725648e-05, "loss": 0.0581, "step": 34780 }, { "epoch": 1.176172284390953, "grad_norm": 0.3319644033908844, "learning_rate": 9.585845970193093e-05, "loss": 0.0574, "step": 34790 }, { "epoch": 1.1765103620812063, "grad_norm": 0.27348774671554565, "learning_rate": 9.585516455285231e-05, "loss": 0.0621, "step": 34800 }, { "epoch": 1.1768484397714594, "grad_norm": 0.28398633003234863, "learning_rate": 9.585186815011073e-05, "loss": 0.0637, "step": 34810 }, { "epoch": 1.1771865174617127, "grad_norm": 0.266220360994339, "learning_rate": 9.584857049379631e-05, "loss": 0.0625, "step": 34820 }, { "epoch": 1.177524595151966, "grad_norm": 0.22908850014209747, "learning_rate": 9.584527158399919e-05, "loss": 0.054, "step": 34830 }, { "epoch": 1.1778626728422192, "grad_norm": 0.28387200832366943, "learning_rate": 9.584197142080955e-05, "loss": 0.0607, "step": 34840 }, { "epoch": 1.1782007505324723, "grad_norm": 0.30583637952804565, "learning_rate": 9.583867000431765e-05, "loss": 0.0579, "step": 34850 }, { "epoch": 1.1785388282227256, "grad_norm": 0.36765313148498535, "learning_rate": 9.583536733461373e-05, "loss": 0.0639, "step": 34860 }, { "epoch": 1.178876905912979, "grad_norm": 0.2269747406244278, "learning_rate": 9.58320634117881e-05, "loss": 0.0543, "step": 34870 }, { "epoch": 1.179214983603232, "grad_norm": 0.4079711437225342, "learning_rate": 9.582875823593106e-05, "loss": 0.0567, "step": 34880 }, { "epoch": 1.1795530612934852, "grad_norm": 0.2688472867012024, "learning_rate": 9.582545180713298e-05, "loss": 0.0516, "step": 34890 }, { "epoch": 1.1798911389837385, "grad_norm": 0.2384854555130005, "learning_rate": 9.582214412548428e-05, "loss": 0.0522, "step": 34900 }, { "epoch": 1.1802292166739916, "grad_norm": 0.33350345492362976, "learning_rate": 9.581883519107538e-05, "loss": 0.0565, "step": 34910 }, { "epoch": 1.180567294364245, "grad_norm": 0.4602130055427551, "learning_rate": 9.581552500399671e-05, "loss": 0.0574, "step": 34920 }, { "epoch": 1.1809053720544982, "grad_norm": 0.2536221742630005, "learning_rate": 9.581221356433882e-05, "loss": 0.0556, "step": 34930 }, { "epoch": 1.1812434497447513, "grad_norm": 0.18539194762706757, "learning_rate": 9.580890087219221e-05, "loss": 0.0531, "step": 34940 }, { "epoch": 1.1815815274350046, "grad_norm": 0.30441489815711975, "learning_rate": 9.580558692764747e-05, "loss": 0.0545, "step": 34950 }, { "epoch": 1.1819196051252578, "grad_norm": 0.3497132360935211, "learning_rate": 9.580227173079519e-05, "loss": 0.0575, "step": 34960 }, { "epoch": 1.182257682815511, "grad_norm": 0.32982882857322693, "learning_rate": 9.5798955281726e-05, "loss": 0.0621, "step": 34970 }, { "epoch": 1.1825957605057642, "grad_norm": 0.2606242299079895, "learning_rate": 9.579563758053058e-05, "loss": 0.0535, "step": 34980 }, { "epoch": 1.1829338381960175, "grad_norm": 0.25398337841033936, "learning_rate": 9.579231862729965e-05, "loss": 0.0569, "step": 34990 }, { "epoch": 1.1832719158862708, "grad_norm": 0.2685031294822693, "learning_rate": 9.578899842212392e-05, "loss": 0.0584, "step": 35000 }, { "epoch": 1.1836099935765239, "grad_norm": 0.3834598660469055, "learning_rate": 9.578567696509418e-05, "loss": 0.0601, "step": 35010 }, { "epoch": 1.1839480712667771, "grad_norm": 0.26304492354393005, "learning_rate": 9.578235425630123e-05, "loss": 0.0599, "step": 35020 }, { "epoch": 1.1842861489570304, "grad_norm": 0.256717324256897, "learning_rate": 9.57790302958359e-05, "loss": 0.0618, "step": 35030 }, { "epoch": 1.1846242266472835, "grad_norm": 0.19830265641212463, "learning_rate": 9.577570508378911e-05, "loss": 0.0583, "step": 35040 }, { "epoch": 1.1849623043375368, "grad_norm": 0.43515831232070923, "learning_rate": 9.57723786202517e-05, "loss": 0.0558, "step": 35050 }, { "epoch": 1.18530038202779, "grad_norm": 0.4591422975063324, "learning_rate": 9.576905090531468e-05, "loss": 0.0525, "step": 35060 }, { "epoch": 1.1856384597180432, "grad_norm": 0.278950572013855, "learning_rate": 9.576572193906901e-05, "loss": 0.0551, "step": 35070 }, { "epoch": 1.1859765374082964, "grad_norm": 0.3147999048233032, "learning_rate": 9.576239172160568e-05, "loss": 0.06, "step": 35080 }, { "epoch": 1.1863146150985497, "grad_norm": 0.2830778956413269, "learning_rate": 9.575906025301575e-05, "loss": 0.0603, "step": 35090 }, { "epoch": 1.1866526927888028, "grad_norm": 0.21710219979286194, "learning_rate": 9.575572753339032e-05, "loss": 0.0597, "step": 35100 }, { "epoch": 1.186990770479056, "grad_norm": 0.2775120437145233, "learning_rate": 9.575239356282049e-05, "loss": 0.0609, "step": 35110 }, { "epoch": 1.1873288481693094, "grad_norm": 0.28734368085861206, "learning_rate": 9.57490583413974e-05, "loss": 0.0537, "step": 35120 }, { "epoch": 1.1876669258595625, "grad_norm": 0.2659660577774048, "learning_rate": 9.574572186921222e-05, "loss": 0.0504, "step": 35130 }, { "epoch": 1.1880050035498158, "grad_norm": 0.19626328349113464, "learning_rate": 9.574238414635621e-05, "loss": 0.0488, "step": 35140 }, { "epoch": 1.188343081240069, "grad_norm": 0.5344597101211548, "learning_rate": 9.57390451729206e-05, "loss": 0.0606, "step": 35150 }, { "epoch": 1.1886811589303221, "grad_norm": 0.2887278199195862, "learning_rate": 9.573570494899668e-05, "loss": 0.066, "step": 35160 }, { "epoch": 1.1890192366205754, "grad_norm": 0.22616246342658997, "learning_rate": 9.573236347467578e-05, "loss": 0.0553, "step": 35170 }, { "epoch": 1.1893573143108287, "grad_norm": 0.3045574426651001, "learning_rate": 9.572902075004924e-05, "loss": 0.0593, "step": 35180 }, { "epoch": 1.1896953920010818, "grad_norm": 0.26916444301605225, "learning_rate": 9.572567677520843e-05, "loss": 0.0511, "step": 35190 }, { "epoch": 1.190033469691335, "grad_norm": 0.3132772445678711, "learning_rate": 9.572233155024482e-05, "loss": 0.0573, "step": 35200 }, { "epoch": 1.1903715473815883, "grad_norm": 0.3608984053134918, "learning_rate": 9.571898507524984e-05, "loss": 0.055, "step": 35210 }, { "epoch": 1.1907096250718414, "grad_norm": 0.5712829828262329, "learning_rate": 9.571563735031497e-05, "loss": 0.0568, "step": 35220 }, { "epoch": 1.1910477027620947, "grad_norm": 0.18830668926239014, "learning_rate": 9.571228837553177e-05, "loss": 0.0613, "step": 35230 }, { "epoch": 1.191385780452348, "grad_norm": 0.29993242025375366, "learning_rate": 9.570893815099176e-05, "loss": 0.0575, "step": 35240 }, { "epoch": 1.191723858142601, "grad_norm": 0.21801573038101196, "learning_rate": 9.570558667678656e-05, "loss": 0.0509, "step": 35250 }, { "epoch": 1.1920619358328544, "grad_norm": 0.477451354265213, "learning_rate": 9.570223395300779e-05, "loss": 0.0529, "step": 35260 }, { "epoch": 1.1924000135231076, "grad_norm": 0.3037533760070801, "learning_rate": 9.569887997974712e-05, "loss": 0.0601, "step": 35270 }, { "epoch": 1.1927380912133607, "grad_norm": 0.37493371963500977, "learning_rate": 9.569552475709623e-05, "loss": 0.0612, "step": 35280 }, { "epoch": 1.193076168903614, "grad_norm": 0.27168720960617065, "learning_rate": 9.569216828514685e-05, "loss": 0.0526, "step": 35290 }, { "epoch": 1.1934142465938673, "grad_norm": 0.29839447140693665, "learning_rate": 9.568881056399077e-05, "loss": 0.0571, "step": 35300 }, { "epoch": 1.1937523242841204, "grad_norm": 0.2423780858516693, "learning_rate": 9.568545159371978e-05, "loss": 0.0505, "step": 35310 }, { "epoch": 1.1940904019743737, "grad_norm": 0.388430118560791, "learning_rate": 9.56820913744257e-05, "loss": 0.0516, "step": 35320 }, { "epoch": 1.194428479664627, "grad_norm": 0.3943400979042053, "learning_rate": 9.56787299062004e-05, "loss": 0.0575, "step": 35330 }, { "epoch": 1.1947665573548802, "grad_norm": 0.19079019129276276, "learning_rate": 9.567536718913578e-05, "loss": 0.0491, "step": 35340 }, { "epoch": 1.1951046350451333, "grad_norm": 0.27585965394973755, "learning_rate": 9.567200322332379e-05, "loss": 0.0632, "step": 35350 }, { "epoch": 1.1954427127353866, "grad_norm": 0.3130986988544464, "learning_rate": 9.566863800885638e-05, "loss": 0.0601, "step": 35360 }, { "epoch": 1.19578079042564, "grad_norm": 0.27861925959587097, "learning_rate": 9.566527154582558e-05, "loss": 0.0601, "step": 35370 }, { "epoch": 1.196118868115893, "grad_norm": 0.3429776132106781, "learning_rate": 9.56619038343234e-05, "loss": 0.0581, "step": 35380 }, { "epoch": 1.1964569458061463, "grad_norm": 0.26484447717666626, "learning_rate": 9.565853487444192e-05, "loss": 0.0584, "step": 35390 }, { "epoch": 1.1967950234963995, "grad_norm": 0.37738901376724243, "learning_rate": 9.565516466627327e-05, "loss": 0.0565, "step": 35400 }, { "epoch": 1.1971331011866526, "grad_norm": 0.32070574164390564, "learning_rate": 9.565179320990955e-05, "loss": 0.0568, "step": 35410 }, { "epoch": 1.197471178876906, "grad_norm": 0.2714320421218872, "learning_rate": 9.564842050544296e-05, "loss": 0.0587, "step": 35420 }, { "epoch": 1.1978092565671592, "grad_norm": 0.26368796825408936, "learning_rate": 9.564504655296569e-05, "loss": 0.0575, "step": 35430 }, { "epoch": 1.1981473342574123, "grad_norm": 0.3562394976615906, "learning_rate": 9.564167135257001e-05, "loss": 0.0613, "step": 35440 }, { "epoch": 1.1984854119476656, "grad_norm": 0.3905876576900482, "learning_rate": 9.56382949043482e-05, "loss": 0.0614, "step": 35450 }, { "epoch": 1.1988234896379188, "grad_norm": 0.37624862790107727, "learning_rate": 9.563491720839253e-05, "loss": 0.065, "step": 35460 }, { "epoch": 1.199161567328172, "grad_norm": 0.42630064487457275, "learning_rate": 9.563153826479536e-05, "loss": 0.0578, "step": 35470 }, { "epoch": 1.1994996450184252, "grad_norm": 0.26353299617767334, "learning_rate": 9.562815807364909e-05, "loss": 0.061, "step": 35480 }, { "epoch": 1.1998377227086785, "grad_norm": 0.2894956171512604, "learning_rate": 9.562477663504611e-05, "loss": 0.0523, "step": 35490 }, { "epoch": 1.2001758003989318, "grad_norm": 0.2546267509460449, "learning_rate": 9.562139394907888e-05, "loss": 0.0557, "step": 35500 }, { "epoch": 1.2005138780891849, "grad_norm": 0.2398775815963745, "learning_rate": 9.56180100158399e-05, "loss": 0.0564, "step": 35510 }, { "epoch": 1.2008519557794382, "grad_norm": 0.18618184328079224, "learning_rate": 9.561462483542164e-05, "loss": 0.0558, "step": 35520 }, { "epoch": 1.2011900334696914, "grad_norm": 0.33678555488586426, "learning_rate": 9.561123840791668e-05, "loss": 0.0559, "step": 35530 }, { "epoch": 1.2015281111599445, "grad_norm": 0.2697972059249878, "learning_rate": 9.56078507334176e-05, "loss": 0.0553, "step": 35540 }, { "epoch": 1.2018661888501978, "grad_norm": 0.27658867835998535, "learning_rate": 9.560446181201701e-05, "loss": 0.0535, "step": 35550 }, { "epoch": 1.202204266540451, "grad_norm": 0.33308109641075134, "learning_rate": 9.560107164380757e-05, "loss": 0.06, "step": 35560 }, { "epoch": 1.2025423442307042, "grad_norm": 0.2239331305027008, "learning_rate": 9.559768022888196e-05, "loss": 0.0504, "step": 35570 }, { "epoch": 1.2028804219209575, "grad_norm": 0.19796545803546906, "learning_rate": 9.559428756733291e-05, "loss": 0.0586, "step": 35580 }, { "epoch": 1.2032184996112107, "grad_norm": 0.2494743913412094, "learning_rate": 9.559089365925315e-05, "loss": 0.0565, "step": 35590 }, { "epoch": 1.2035565773014638, "grad_norm": 0.33348873257637024, "learning_rate": 9.55874985047355e-05, "loss": 0.0624, "step": 35600 }, { "epoch": 1.203894654991717, "grad_norm": 0.28265705704689026, "learning_rate": 9.558410210387277e-05, "loss": 0.0555, "step": 35610 }, { "epoch": 1.2042327326819704, "grad_norm": 0.21416500210762024, "learning_rate": 9.558070445675782e-05, "loss": 0.0533, "step": 35620 }, { "epoch": 1.2045708103722235, "grad_norm": 0.2351299375295639, "learning_rate": 9.557730556348352e-05, "loss": 0.0546, "step": 35630 }, { "epoch": 1.2049088880624768, "grad_norm": 0.3358142077922821, "learning_rate": 9.557390542414281e-05, "loss": 0.0555, "step": 35640 }, { "epoch": 1.20524696575273, "grad_norm": 0.2913195490837097, "learning_rate": 9.557050403882864e-05, "loss": 0.0544, "step": 35650 }, { "epoch": 1.2055850434429831, "grad_norm": 0.35277947783470154, "learning_rate": 9.556710140763401e-05, "loss": 0.0582, "step": 35660 }, { "epoch": 1.2059231211332364, "grad_norm": 0.26132264733314514, "learning_rate": 9.556369753065197e-05, "loss": 0.0621, "step": 35670 }, { "epoch": 1.2062611988234897, "grad_norm": 0.3298356831073761, "learning_rate": 9.556029240797553e-05, "loss": 0.0554, "step": 35680 }, { "epoch": 1.2065992765137428, "grad_norm": 0.19752025604248047, "learning_rate": 9.555688603969783e-05, "loss": 0.0586, "step": 35690 }, { "epoch": 1.206937354203996, "grad_norm": 0.4011576175689697, "learning_rate": 9.555347842591198e-05, "loss": 0.0542, "step": 35700 }, { "epoch": 1.2072754318942494, "grad_norm": 0.2306644469499588, "learning_rate": 9.555006956671113e-05, "loss": 0.0562, "step": 35710 }, { "epoch": 1.2076135095845024, "grad_norm": 0.22429241240024567, "learning_rate": 9.55466594621885e-05, "loss": 0.0574, "step": 35720 }, { "epoch": 1.2079515872747557, "grad_norm": 0.25659796595573425, "learning_rate": 9.554324811243731e-05, "loss": 0.0526, "step": 35730 }, { "epoch": 1.208289664965009, "grad_norm": 0.3174155056476593, "learning_rate": 9.553983551755081e-05, "loss": 0.0585, "step": 35740 }, { "epoch": 1.208627742655262, "grad_norm": 0.3058999180793762, "learning_rate": 9.553642167762233e-05, "loss": 0.0574, "step": 35750 }, { "epoch": 1.2089658203455154, "grad_norm": 0.29023006558418274, "learning_rate": 9.55330065927452e-05, "loss": 0.0574, "step": 35760 }, { "epoch": 1.2093038980357687, "grad_norm": 0.24237550795078278, "learning_rate": 9.552959026301276e-05, "loss": 0.0494, "step": 35770 }, { "epoch": 1.2096419757260217, "grad_norm": 0.26423609256744385, "learning_rate": 9.552617268851844e-05, "loss": 0.0602, "step": 35780 }, { "epoch": 1.209980053416275, "grad_norm": 0.3288435935974121, "learning_rate": 9.552275386935565e-05, "loss": 0.0562, "step": 35790 }, { "epoch": 1.2103181311065283, "grad_norm": 0.27663305401802063, "learning_rate": 9.551933380561788e-05, "loss": 0.06, "step": 35800 }, { "epoch": 1.2106562087967816, "grad_norm": 0.46368804574012756, "learning_rate": 9.551591249739862e-05, "loss": 0.0561, "step": 35810 }, { "epoch": 1.2109942864870347, "grad_norm": 0.2741178274154663, "learning_rate": 9.55124899447914e-05, "loss": 0.0618, "step": 35820 }, { "epoch": 1.211332364177288, "grad_norm": 0.264067143201828, "learning_rate": 9.550906614788982e-05, "loss": 0.0548, "step": 35830 }, { "epoch": 1.2116704418675412, "grad_norm": 0.25167882442474365, "learning_rate": 9.550564110678746e-05, "loss": 0.0585, "step": 35840 }, { "epoch": 1.2120085195577943, "grad_norm": 0.30589529871940613, "learning_rate": 9.550221482157797e-05, "loss": 0.0563, "step": 35850 }, { "epoch": 1.2123465972480476, "grad_norm": 0.2676481306552887, "learning_rate": 9.549878729235502e-05, "loss": 0.0564, "step": 35860 }, { "epoch": 1.212684674938301, "grad_norm": 0.27731698751449585, "learning_rate": 9.549535851921232e-05, "loss": 0.0594, "step": 35870 }, { "epoch": 1.213022752628554, "grad_norm": 0.2726663053035736, "learning_rate": 9.549192850224362e-05, "loss": 0.0551, "step": 35880 }, { "epoch": 1.2133608303188073, "grad_norm": 0.22920717298984528, "learning_rate": 9.548849724154268e-05, "loss": 0.0542, "step": 35890 }, { "epoch": 1.2136989080090606, "grad_norm": 0.18584400415420532, "learning_rate": 9.54850647372033e-05, "loss": 0.0566, "step": 35900 }, { "epoch": 1.2140369856993136, "grad_norm": 0.24276083707809448, "learning_rate": 9.548163098931935e-05, "loss": 0.054, "step": 35910 }, { "epoch": 1.214375063389567, "grad_norm": 0.4580259323120117, "learning_rate": 9.547819599798468e-05, "loss": 0.0591, "step": 35920 }, { "epoch": 1.2147131410798202, "grad_norm": 0.22206172347068787, "learning_rate": 9.547475976329325e-05, "loss": 0.0599, "step": 35930 }, { "epoch": 1.2150512187700733, "grad_norm": 0.25925177335739136, "learning_rate": 9.547132228533895e-05, "loss": 0.0636, "step": 35940 }, { "epoch": 1.2153892964603266, "grad_norm": 0.2317526787519455, "learning_rate": 9.546788356421578e-05, "loss": 0.0584, "step": 35950 }, { "epoch": 1.2157273741505799, "grad_norm": 0.2139112800359726, "learning_rate": 9.546444360001775e-05, "loss": 0.0563, "step": 35960 }, { "epoch": 1.2160654518408331, "grad_norm": 0.24710367619991302, "learning_rate": 9.546100239283892e-05, "loss": 0.0529, "step": 35970 }, { "epoch": 1.2164035295310862, "grad_norm": 0.39623039960861206, "learning_rate": 9.545755994277337e-05, "loss": 0.056, "step": 35980 }, { "epoch": 1.2167416072213395, "grad_norm": 0.31126028299331665, "learning_rate": 9.545411624991521e-05, "loss": 0.051, "step": 35990 }, { "epoch": 1.2170796849115928, "grad_norm": 0.23599942028522491, "learning_rate": 9.545067131435858e-05, "loss": 0.0569, "step": 36000 }, { "epoch": 1.2174177626018459, "grad_norm": 0.35455986857414246, "learning_rate": 9.544722513619769e-05, "loss": 0.0516, "step": 36010 }, { "epoch": 1.2177558402920992, "grad_norm": 0.2820851802825928, "learning_rate": 9.54437777155267e-05, "loss": 0.0494, "step": 36020 }, { "epoch": 1.2180939179823524, "grad_norm": 0.23169292509555817, "learning_rate": 9.544032905243994e-05, "loss": 0.0576, "step": 36030 }, { "epoch": 1.2184319956726055, "grad_norm": 0.24395766854286194, "learning_rate": 9.543687914703162e-05, "loss": 0.0561, "step": 36040 }, { "epoch": 1.2187700733628588, "grad_norm": 0.2634219527244568, "learning_rate": 9.543342799939612e-05, "loss": 0.0591, "step": 36050 }, { "epoch": 1.219108151053112, "grad_norm": 0.6792777180671692, "learning_rate": 9.542997560962775e-05, "loss": 0.0557, "step": 36060 }, { "epoch": 1.2194462287433652, "grad_norm": 0.25558191537857056, "learning_rate": 9.542652197782092e-05, "loss": 0.0577, "step": 36070 }, { "epoch": 1.2197843064336185, "grad_norm": 0.2723236083984375, "learning_rate": 9.542306710407005e-05, "loss": 0.0543, "step": 36080 }, { "epoch": 1.2201223841238718, "grad_norm": 0.3135468065738678, "learning_rate": 9.541961098846959e-05, "loss": 0.0558, "step": 36090 }, { "epoch": 1.2204604618141248, "grad_norm": 0.5519605278968811, "learning_rate": 9.541615363111402e-05, "loss": 0.0555, "step": 36100 }, { "epoch": 1.220798539504378, "grad_norm": 0.24980396032333374, "learning_rate": 9.541269503209787e-05, "loss": 0.0522, "step": 36110 }, { "epoch": 1.2211366171946314, "grad_norm": 0.2448996752500534, "learning_rate": 9.540923519151571e-05, "loss": 0.053, "step": 36120 }, { "epoch": 1.2214746948848845, "grad_norm": 0.203586146235466, "learning_rate": 9.540577410946212e-05, "loss": 0.0563, "step": 36130 }, { "epoch": 1.2218127725751378, "grad_norm": 0.9368453025817871, "learning_rate": 9.540231178603172e-05, "loss": 0.0563, "step": 36140 }, { "epoch": 1.222150850265391, "grad_norm": 0.22015231847763062, "learning_rate": 9.539884822131918e-05, "loss": 0.0612, "step": 36150 }, { "epoch": 1.2224889279556441, "grad_norm": 0.3027173578739166, "learning_rate": 9.539538341541916e-05, "loss": 0.0585, "step": 36160 }, { "epoch": 1.2228270056458974, "grad_norm": 0.20858865976333618, "learning_rate": 9.539191736842643e-05, "loss": 0.0558, "step": 36170 }, { "epoch": 1.2231650833361507, "grad_norm": 0.20078973472118378, "learning_rate": 9.538845008043573e-05, "loss": 0.0572, "step": 36180 }, { "epoch": 1.2235031610264038, "grad_norm": 0.1849302053451538, "learning_rate": 9.538498155154185e-05, "loss": 0.0562, "step": 36190 }, { "epoch": 1.223841238716657, "grad_norm": 0.23463766276836395, "learning_rate": 9.538151178183965e-05, "loss": 0.058, "step": 36200 }, { "epoch": 1.2241793164069104, "grad_norm": 0.34980469942092896, "learning_rate": 9.537804077142394e-05, "loss": 0.0547, "step": 36210 }, { "epoch": 1.2245173940971634, "grad_norm": 0.3193244934082031, "learning_rate": 9.537456852038965e-05, "loss": 0.0614, "step": 36220 }, { "epoch": 1.2248554717874167, "grad_norm": 0.32651954889297485, "learning_rate": 9.53710950288317e-05, "loss": 0.0633, "step": 36230 }, { "epoch": 1.22519354947767, "grad_norm": 0.3323945999145508, "learning_rate": 9.536762029684507e-05, "loss": 0.063, "step": 36240 }, { "epoch": 1.225531627167923, "grad_norm": 0.2722984850406647, "learning_rate": 9.536414432452473e-05, "loss": 0.0515, "step": 36250 }, { "epoch": 1.2258697048581764, "grad_norm": 0.2529185116291046, "learning_rate": 9.536066711196573e-05, "loss": 0.0575, "step": 36260 }, { "epoch": 1.2262077825484297, "grad_norm": 0.21919170022010803, "learning_rate": 9.535718865926314e-05, "loss": 0.0581, "step": 36270 }, { "epoch": 1.2265458602386827, "grad_norm": 0.46891167759895325, "learning_rate": 9.535370896651205e-05, "loss": 0.0632, "step": 36280 }, { "epoch": 1.226883937928936, "grad_norm": 0.23872634768486023, "learning_rate": 9.535022803380759e-05, "loss": 0.0605, "step": 36290 }, { "epoch": 1.2272220156191893, "grad_norm": 0.2964627742767334, "learning_rate": 9.534674586124494e-05, "loss": 0.0592, "step": 36300 }, { "epoch": 1.2275600933094426, "grad_norm": 0.29028189182281494, "learning_rate": 9.534326244891928e-05, "loss": 0.0565, "step": 36310 }, { "epoch": 1.2278981709996957, "grad_norm": 0.2981182038784027, "learning_rate": 9.533977779692588e-05, "loss": 0.0628, "step": 36320 }, { "epoch": 1.228236248689949, "grad_norm": 0.22987695038318634, "learning_rate": 9.533629190535997e-05, "loss": 0.0655, "step": 36330 }, { "epoch": 1.2285743263802023, "grad_norm": 0.33497878909111023, "learning_rate": 9.533280477431688e-05, "loss": 0.0578, "step": 36340 }, { "epoch": 1.2289124040704553, "grad_norm": 0.2711617350578308, "learning_rate": 9.532931640389193e-05, "loss": 0.0631, "step": 36350 }, { "epoch": 1.2292504817607086, "grad_norm": 0.28582561016082764, "learning_rate": 9.532582679418052e-05, "loss": 0.0618, "step": 36360 }, { "epoch": 1.229588559450962, "grad_norm": 0.3803589344024658, "learning_rate": 9.532233594527802e-05, "loss": 0.0615, "step": 36370 }, { "epoch": 1.229926637141215, "grad_norm": 0.3904787302017212, "learning_rate": 9.531884385727987e-05, "loss": 0.06, "step": 36380 }, { "epoch": 1.2302647148314683, "grad_norm": 0.31606170535087585, "learning_rate": 9.531535053028158e-05, "loss": 0.0592, "step": 36390 }, { "epoch": 1.2306027925217216, "grad_norm": 0.26432785391807556, "learning_rate": 9.531185596437861e-05, "loss": 0.0598, "step": 36400 }, { "epoch": 1.2309408702119746, "grad_norm": 0.33929017186164856, "learning_rate": 9.530836015966654e-05, "loss": 0.0625, "step": 36410 }, { "epoch": 1.231278947902228, "grad_norm": 0.37820708751678467, "learning_rate": 9.53048631162409e-05, "loss": 0.0603, "step": 36420 }, { "epoch": 1.2316170255924812, "grad_norm": 0.36609211564064026, "learning_rate": 9.530136483419733e-05, "loss": 0.0598, "step": 36430 }, { "epoch": 1.2319551032827343, "grad_norm": 0.6142157316207886, "learning_rate": 9.529786531363148e-05, "loss": 0.0545, "step": 36440 }, { "epoch": 1.2322931809729876, "grad_norm": 0.2054520547389984, "learning_rate": 9.5294364554639e-05, "loss": 0.0546, "step": 36450 }, { "epoch": 1.2326312586632409, "grad_norm": 0.3291938900947571, "learning_rate": 9.52908625573156e-05, "loss": 0.065, "step": 36460 }, { "epoch": 1.2329693363534941, "grad_norm": 0.3186419904232025, "learning_rate": 9.528735932175704e-05, "loss": 0.0534, "step": 36470 }, { "epoch": 1.2333074140437472, "grad_norm": 0.26404011249542236, "learning_rate": 9.52838548480591e-05, "loss": 0.0622, "step": 36480 }, { "epoch": 1.2336454917340005, "grad_norm": 0.48045656085014343, "learning_rate": 9.528034913631756e-05, "loss": 0.0613, "step": 36490 }, { "epoch": 1.2339835694242538, "grad_norm": 0.31406542658805847, "learning_rate": 9.527684218662831e-05, "loss": 0.0616, "step": 36500 }, { "epoch": 1.2343216471145069, "grad_norm": 0.19710107147693634, "learning_rate": 9.527333399908718e-05, "loss": 0.0494, "step": 36510 }, { "epoch": 1.2346597248047602, "grad_norm": 0.26293978095054626, "learning_rate": 9.526982457379012e-05, "loss": 0.0574, "step": 36520 }, { "epoch": 1.2349978024950135, "grad_norm": 0.25516775250434875, "learning_rate": 9.526631391083307e-05, "loss": 0.0554, "step": 36530 }, { "epoch": 1.2353358801852665, "grad_norm": 0.31003403663635254, "learning_rate": 9.526280201031199e-05, "loss": 0.0548, "step": 36540 }, { "epoch": 1.2356739578755198, "grad_norm": 0.3247447907924652, "learning_rate": 9.525928887232292e-05, "loss": 0.0577, "step": 36550 }, { "epoch": 1.236012035565773, "grad_norm": 0.2193087935447693, "learning_rate": 9.525577449696188e-05, "loss": 0.0575, "step": 36560 }, { "epoch": 1.2363501132560262, "grad_norm": 0.3954932987689972, "learning_rate": 9.525225888432499e-05, "loss": 0.0544, "step": 36570 }, { "epoch": 1.2366881909462795, "grad_norm": 0.2879047691822052, "learning_rate": 9.524874203450832e-05, "loss": 0.0565, "step": 36580 }, { "epoch": 1.2370262686365328, "grad_norm": 0.2637042701244354, "learning_rate": 9.524522394760804e-05, "loss": 0.061, "step": 36590 }, { "epoch": 1.2373643463267858, "grad_norm": 0.225764662027359, "learning_rate": 9.524170462372035e-05, "loss": 0.0543, "step": 36600 }, { "epoch": 1.2377024240170391, "grad_norm": 0.18661554157733917, "learning_rate": 9.523818406294145e-05, "loss": 0.0586, "step": 36610 }, { "epoch": 1.2380405017072924, "grad_norm": 0.3777570128440857, "learning_rate": 9.523466226536759e-05, "loss": 0.0615, "step": 36620 }, { "epoch": 1.2383785793975455, "grad_norm": 0.42210203409194946, "learning_rate": 9.523113923109506e-05, "loss": 0.0604, "step": 36630 }, { "epoch": 1.2387166570877988, "grad_norm": 0.3158259987831116, "learning_rate": 9.522761496022017e-05, "loss": 0.063, "step": 36640 }, { "epoch": 1.239054734778052, "grad_norm": 0.21655884385108948, "learning_rate": 9.52240894528393e-05, "loss": 0.0603, "step": 36650 }, { "epoch": 1.2393928124683051, "grad_norm": 0.2833675444126129, "learning_rate": 9.522056270904878e-05, "loss": 0.0502, "step": 36660 }, { "epoch": 1.2397308901585584, "grad_norm": 0.5545035004615784, "learning_rate": 9.521703472894508e-05, "loss": 0.0621, "step": 36670 }, { "epoch": 1.2400689678488117, "grad_norm": 0.462471067905426, "learning_rate": 9.521350551262465e-05, "loss": 0.0599, "step": 36680 }, { "epoch": 1.2404070455390648, "grad_norm": 0.3286322355270386, "learning_rate": 9.520997506018394e-05, "loss": 0.0572, "step": 36690 }, { "epoch": 1.240745123229318, "grad_norm": 0.19536089897155762, "learning_rate": 9.520644337171952e-05, "loss": 0.0608, "step": 36700 }, { "epoch": 1.2410832009195714, "grad_norm": 0.2653295397758484, "learning_rate": 9.52029104473279e-05, "loss": 0.0594, "step": 36710 }, { "epoch": 1.2414212786098244, "grad_norm": 0.23220765590667725, "learning_rate": 9.519937628710572e-05, "loss": 0.0535, "step": 36720 }, { "epoch": 1.2417593563000777, "grad_norm": 0.607427716255188, "learning_rate": 9.519584089114954e-05, "loss": 0.0541, "step": 36730 }, { "epoch": 1.242097433990331, "grad_norm": 0.21251623332500458, "learning_rate": 9.519230425955606e-05, "loss": 0.0556, "step": 36740 }, { "epoch": 1.242435511680584, "grad_norm": 0.22234714031219482, "learning_rate": 9.518876639242197e-05, "loss": 0.0571, "step": 36750 }, { "epoch": 1.2427735893708374, "grad_norm": 0.3247356116771698, "learning_rate": 9.518522728984399e-05, "loss": 0.0571, "step": 36760 }, { "epoch": 1.2431116670610907, "grad_norm": 0.2673957943916321, "learning_rate": 9.518168695191885e-05, "loss": 0.0617, "step": 36770 }, { "epoch": 1.243449744751344, "grad_norm": 0.2676803767681122, "learning_rate": 9.517814537874336e-05, "loss": 0.055, "step": 36780 }, { "epoch": 1.243787822441597, "grad_norm": 0.9901505708694458, "learning_rate": 9.517460257041436e-05, "loss": 0.0588, "step": 36790 }, { "epoch": 1.2441259001318503, "grad_norm": 0.7449643015861511, "learning_rate": 9.51710585270287e-05, "loss": 0.052, "step": 36800 }, { "epoch": 1.2444639778221036, "grad_norm": 0.22972966730594635, "learning_rate": 9.516751324868326e-05, "loss": 0.0586, "step": 36810 }, { "epoch": 1.2448020555123567, "grad_norm": 0.5112426280975342, "learning_rate": 9.516396673547498e-05, "loss": 0.0579, "step": 36820 }, { "epoch": 1.24514013320261, "grad_norm": 0.22329077124595642, "learning_rate": 9.516041898750083e-05, "loss": 0.0514, "step": 36830 }, { "epoch": 1.2454782108928633, "grad_norm": 0.22410309314727783, "learning_rate": 9.515687000485778e-05, "loss": 0.0576, "step": 36840 }, { "epoch": 1.2458162885831163, "grad_norm": 0.25830939412117004, "learning_rate": 9.515331978764286e-05, "loss": 0.0515, "step": 36850 }, { "epoch": 1.2461543662733696, "grad_norm": 0.2790699303150177, "learning_rate": 9.514976833595315e-05, "loss": 0.0604, "step": 36860 }, { "epoch": 1.246492443963623, "grad_norm": 0.18594925105571747, "learning_rate": 9.514621564988573e-05, "loss": 0.0581, "step": 36870 }, { "epoch": 1.246830521653876, "grad_norm": 0.2548982799053192, "learning_rate": 9.514266172953775e-05, "loss": 0.0594, "step": 36880 }, { "epoch": 1.2471685993441293, "grad_norm": 0.23358707129955292, "learning_rate": 9.513910657500635e-05, "loss": 0.0595, "step": 36890 }, { "epoch": 1.2475066770343826, "grad_norm": 0.3635452389717102, "learning_rate": 9.513555018638872e-05, "loss": 0.067, "step": 36900 }, { "epoch": 1.2478447547246356, "grad_norm": 0.22603128850460052, "learning_rate": 9.513199256378211e-05, "loss": 0.0557, "step": 36910 }, { "epoch": 1.248182832414889, "grad_norm": 0.4239748418331146, "learning_rate": 9.512843370728379e-05, "loss": 0.058, "step": 36920 }, { "epoch": 1.2485209101051422, "grad_norm": 0.3253561854362488, "learning_rate": 9.512487361699104e-05, "loss": 0.0596, "step": 36930 }, { "epoch": 1.2488589877953955, "grad_norm": 0.3685761094093323, "learning_rate": 9.51213122930012e-05, "loss": 0.0581, "step": 36940 }, { "epoch": 1.2491970654856486, "grad_norm": 0.18330363929271698, "learning_rate": 9.511774973541163e-05, "loss": 0.0518, "step": 36950 }, { "epoch": 1.2495351431759019, "grad_norm": 0.4726754128932953, "learning_rate": 9.511418594431974e-05, "loss": 0.0623, "step": 36960 }, { "epoch": 1.2498732208661552, "grad_norm": 0.23833231627941132, "learning_rate": 9.511062091982295e-05, "loss": 0.0613, "step": 36970 }, { "epoch": 1.2502112985564082, "grad_norm": 0.23926083743572235, "learning_rate": 9.510705466201872e-05, "loss": 0.0586, "step": 36980 }, { "epoch": 1.2505493762466615, "grad_norm": 0.4436635673046112, "learning_rate": 9.510348717100457e-05, "loss": 0.0566, "step": 36990 }, { "epoch": 1.2508874539369148, "grad_norm": 0.19728191196918488, "learning_rate": 9.509991844687803e-05, "loss": 0.0561, "step": 37000 }, { "epoch": 1.2512255316271679, "grad_norm": 0.2276071012020111, "learning_rate": 9.509634848973666e-05, "loss": 0.0649, "step": 37010 }, { "epoch": 1.2515636093174212, "grad_norm": 0.20598101615905762, "learning_rate": 9.509277729967808e-05, "loss": 0.0531, "step": 37020 }, { "epoch": 1.2519016870076745, "grad_norm": 0.3447912037372589, "learning_rate": 9.508920487679989e-05, "loss": 0.0578, "step": 37030 }, { "epoch": 1.2522397646979275, "grad_norm": 0.3585319221019745, "learning_rate": 9.508563122119978e-05, "loss": 0.062, "step": 37040 }, { "epoch": 1.2525778423881808, "grad_norm": 0.3936111330986023, "learning_rate": 9.508205633297544e-05, "loss": 0.0557, "step": 37050 }, { "epoch": 1.252915920078434, "grad_norm": 0.28080835938453674, "learning_rate": 9.507848021222463e-05, "loss": 0.062, "step": 37060 }, { "epoch": 1.2532539977686872, "grad_norm": 0.20483142137527466, "learning_rate": 9.50749028590451e-05, "loss": 0.0606, "step": 37070 }, { "epoch": 1.2535920754589405, "grad_norm": 0.4048832356929779, "learning_rate": 9.507132427353467e-05, "loss": 0.0563, "step": 37080 }, { "epoch": 1.2539301531491938, "grad_norm": 0.25957173109054565, "learning_rate": 9.506774445579116e-05, "loss": 0.0542, "step": 37090 }, { "epoch": 1.2542682308394468, "grad_norm": 0.2811545431613922, "learning_rate": 9.506416340591244e-05, "loss": 0.0552, "step": 37100 }, { "epoch": 1.2546063085297001, "grad_norm": 0.6214828491210938, "learning_rate": 9.506058112399643e-05, "loss": 0.0535, "step": 37110 }, { "epoch": 1.2549443862199534, "grad_norm": 0.1840749830007553, "learning_rate": 9.505699761014107e-05, "loss": 0.0564, "step": 37120 }, { "epoch": 1.2552824639102065, "grad_norm": 0.23955586552619934, "learning_rate": 9.505341286444432e-05, "loss": 0.0567, "step": 37130 }, { "epoch": 1.2556205416004598, "grad_norm": 0.2272670567035675, "learning_rate": 9.504982688700417e-05, "loss": 0.0586, "step": 37140 }, { "epoch": 1.255958619290713, "grad_norm": 0.24044035375118256, "learning_rate": 9.504623967791869e-05, "loss": 0.0489, "step": 37150 }, { "epoch": 1.2562966969809661, "grad_norm": 0.3058665990829468, "learning_rate": 9.504265123728594e-05, "loss": 0.0564, "step": 37160 }, { "epoch": 1.2566347746712194, "grad_norm": 0.2787776589393616, "learning_rate": 9.503906156520403e-05, "loss": 0.0557, "step": 37170 }, { "epoch": 1.2569728523614727, "grad_norm": 0.2062700390815735, "learning_rate": 9.503547066177109e-05, "loss": 0.0594, "step": 37180 }, { "epoch": 1.2573109300517258, "grad_norm": 0.5092566013336182, "learning_rate": 9.50318785270853e-05, "loss": 0.0609, "step": 37190 }, { "epoch": 1.257649007741979, "grad_norm": 0.2180127054452896, "learning_rate": 9.502828516124488e-05, "loss": 0.0558, "step": 37200 }, { "epoch": 1.2579870854322324, "grad_norm": 0.3507996201515198, "learning_rate": 9.502469056434805e-05, "loss": 0.0578, "step": 37210 }, { "epoch": 1.2583251631224854, "grad_norm": 0.34653112292289734, "learning_rate": 9.50210947364931e-05, "loss": 0.0554, "step": 37220 }, { "epoch": 1.2586632408127387, "grad_norm": 0.4112243056297302, "learning_rate": 9.501749767777832e-05, "loss": 0.0587, "step": 37230 }, { "epoch": 1.259001318502992, "grad_norm": 0.24278514087200165, "learning_rate": 9.50138993883021e-05, "loss": 0.057, "step": 37240 }, { "epoch": 1.259339396193245, "grad_norm": 0.49317002296447754, "learning_rate": 9.501029986816274e-05, "loss": 0.0641, "step": 37250 }, { "epoch": 1.2596774738834984, "grad_norm": 0.21551275253295898, "learning_rate": 9.500669911745871e-05, "loss": 0.0584, "step": 37260 }, { "epoch": 1.2600155515737517, "grad_norm": 0.4726448953151703, "learning_rate": 9.500309713628845e-05, "loss": 0.0543, "step": 37270 }, { "epoch": 1.2603536292640047, "grad_norm": 0.37461331486701965, "learning_rate": 9.49994939247504e-05, "loss": 0.0597, "step": 37280 }, { "epoch": 1.260691706954258, "grad_norm": 0.26321977376937866, "learning_rate": 9.499588948294309e-05, "loss": 0.0595, "step": 37290 }, { "epoch": 1.2610297846445113, "grad_norm": 0.24872058629989624, "learning_rate": 9.499228381096507e-05, "loss": 0.0541, "step": 37300 }, { "epoch": 1.2613678623347646, "grad_norm": 0.34574151039123535, "learning_rate": 9.498867690891493e-05, "loss": 0.058, "step": 37310 }, { "epoch": 1.2617059400250177, "grad_norm": 0.2671336233615875, "learning_rate": 9.498506877689125e-05, "loss": 0.0545, "step": 37320 }, { "epoch": 1.262044017715271, "grad_norm": 0.3735843300819397, "learning_rate": 9.498145941499269e-05, "loss": 0.0602, "step": 37330 }, { "epoch": 1.2623820954055243, "grad_norm": 0.23484918475151062, "learning_rate": 9.497784882331793e-05, "loss": 0.053, "step": 37340 }, { "epoch": 1.2627201730957773, "grad_norm": 0.25833621621131897, "learning_rate": 9.497423700196568e-05, "loss": 0.0529, "step": 37350 }, { "epoch": 1.2630582507860306, "grad_norm": 0.24701955914497375, "learning_rate": 9.497062395103469e-05, "loss": 0.0584, "step": 37360 }, { "epoch": 1.263396328476284, "grad_norm": 0.38474905490875244, "learning_rate": 9.496700967062373e-05, "loss": 0.0574, "step": 37370 }, { "epoch": 1.2637344061665372, "grad_norm": 0.32404825091362, "learning_rate": 9.496339416083162e-05, "loss": 0.0548, "step": 37380 }, { "epoch": 1.2640724838567903, "grad_norm": 0.26675477623939514, "learning_rate": 9.495977742175722e-05, "loss": 0.0573, "step": 37390 }, { "epoch": 1.2644105615470436, "grad_norm": 0.39166662096977234, "learning_rate": 9.495615945349937e-05, "loss": 0.0589, "step": 37400 }, { "epoch": 1.2647486392372969, "grad_norm": 0.2279069572687149, "learning_rate": 9.495254025615701e-05, "loss": 0.0545, "step": 37410 }, { "epoch": 1.26508671692755, "grad_norm": 0.40101727843284607, "learning_rate": 9.494891982982911e-05, "loss": 0.0578, "step": 37420 }, { "epoch": 1.2654247946178032, "grad_norm": 0.5148099064826965, "learning_rate": 9.494529817461462e-05, "loss": 0.056, "step": 37430 }, { "epoch": 1.2657628723080565, "grad_norm": 1.263156533241272, "learning_rate": 9.494167529061256e-05, "loss": 0.0537, "step": 37440 }, { "epoch": 1.2661009499983096, "grad_norm": 0.26052531599998474, "learning_rate": 9.493805117792199e-05, "loss": 0.0535, "step": 37450 }, { "epoch": 1.2664390276885629, "grad_norm": 0.40224385261535645, "learning_rate": 9.493442583664196e-05, "loss": 0.0541, "step": 37460 }, { "epoch": 1.2667771053788162, "grad_norm": 0.3336508274078369, "learning_rate": 9.493079926687163e-05, "loss": 0.0592, "step": 37470 }, { "epoch": 1.2671151830690692, "grad_norm": 0.3124580383300781, "learning_rate": 9.492717146871013e-05, "loss": 0.0562, "step": 37480 }, { "epoch": 1.2674532607593225, "grad_norm": 0.20650961995124817, "learning_rate": 9.492354244225663e-05, "loss": 0.0564, "step": 37490 }, { "epoch": 1.2677913384495758, "grad_norm": 0.25813764333724976, "learning_rate": 9.491991218761036e-05, "loss": 0.0528, "step": 37500 }, { "epoch": 1.2681294161398289, "grad_norm": 0.3352694511413574, "learning_rate": 9.491628070487056e-05, "loss": 0.0682, "step": 37510 }, { "epoch": 1.2684674938300822, "grad_norm": 0.25944870710372925, "learning_rate": 9.491264799413653e-05, "loss": 0.0578, "step": 37520 }, { "epoch": 1.2688055715203355, "grad_norm": 0.2476460188627243, "learning_rate": 9.490901405550757e-05, "loss": 0.0567, "step": 37530 }, { "epoch": 1.2691436492105885, "grad_norm": 0.3122856318950653, "learning_rate": 9.490537888908305e-05, "loss": 0.0573, "step": 37540 }, { "epoch": 1.2694817269008418, "grad_norm": 0.3345428705215454, "learning_rate": 9.490174249496231e-05, "loss": 0.0539, "step": 37550 }, { "epoch": 1.2698198045910951, "grad_norm": 0.4281669557094574, "learning_rate": 9.489810487324483e-05, "loss": 0.0584, "step": 37560 }, { "epoch": 1.2701578822813482, "grad_norm": 0.29891008138656616, "learning_rate": 9.489446602403002e-05, "loss": 0.0599, "step": 37570 }, { "epoch": 1.2704959599716015, "grad_norm": 0.3133098781108856, "learning_rate": 9.489082594741739e-05, "loss": 0.0589, "step": 37580 }, { "epoch": 1.2708340376618548, "grad_norm": 0.3147125244140625, "learning_rate": 9.48871846435064e-05, "loss": 0.0633, "step": 37590 }, { "epoch": 1.2711721153521078, "grad_norm": 0.524387001991272, "learning_rate": 9.48835421123967e-05, "loss": 0.0597, "step": 37600 }, { "epoch": 1.2715101930423611, "grad_norm": 0.3920101523399353, "learning_rate": 9.487989835418778e-05, "loss": 0.0574, "step": 37610 }, { "epoch": 1.2718482707326144, "grad_norm": 0.29134535789489746, "learning_rate": 9.487625336897931e-05, "loss": 0.059, "step": 37620 }, { "epoch": 1.2721863484228675, "grad_norm": 0.2706206440925598, "learning_rate": 9.487260715687095e-05, "loss": 0.0616, "step": 37630 }, { "epoch": 1.2725244261131208, "grad_norm": 0.2911831736564636, "learning_rate": 9.486895971796234e-05, "loss": 0.057, "step": 37640 }, { "epoch": 1.272862503803374, "grad_norm": 0.2771049737930298, "learning_rate": 9.486531105235326e-05, "loss": 0.0551, "step": 37650 }, { "epoch": 1.2732005814936271, "grad_norm": 0.36252692341804504, "learning_rate": 9.48616611601434e-05, "loss": 0.0559, "step": 37660 }, { "epoch": 1.2735386591838804, "grad_norm": 0.29231229424476624, "learning_rate": 9.48580100414326e-05, "loss": 0.0561, "step": 37670 }, { "epoch": 1.2738767368741337, "grad_norm": 0.31653323769569397, "learning_rate": 9.485435769632064e-05, "loss": 0.0539, "step": 37680 }, { "epoch": 1.2742148145643868, "grad_norm": 0.37751635909080505, "learning_rate": 9.48507041249074e-05, "loss": 0.0638, "step": 37690 }, { "epoch": 1.27455289225464, "grad_norm": 0.8191001415252686, "learning_rate": 9.484704932729276e-05, "loss": 0.0616, "step": 37700 }, { "epoch": 1.2748909699448934, "grad_norm": 0.32144808769226074, "learning_rate": 9.484339330357664e-05, "loss": 0.0648, "step": 37710 }, { "epoch": 1.2752290476351464, "grad_norm": 0.23479968309402466, "learning_rate": 9.483973605385898e-05, "loss": 0.0537, "step": 37720 }, { "epoch": 1.2755671253253997, "grad_norm": 0.3440322279930115, "learning_rate": 9.48360775782398e-05, "loss": 0.0602, "step": 37730 }, { "epoch": 1.275905203015653, "grad_norm": 0.256502628326416, "learning_rate": 9.48324178768191e-05, "loss": 0.0609, "step": 37740 }, { "epoch": 1.276243280705906, "grad_norm": 0.3323057293891907, "learning_rate": 9.482875694969693e-05, "loss": 0.05, "step": 37750 }, { "epoch": 1.2765813583961594, "grad_norm": 0.2115819901227951, "learning_rate": 9.482509479697339e-05, "loss": 0.0578, "step": 37760 }, { "epoch": 1.2769194360864127, "grad_norm": 0.45000964403152466, "learning_rate": 9.48214314187486e-05, "loss": 0.0599, "step": 37770 }, { "epoch": 1.2772575137766657, "grad_norm": 0.35768386721611023, "learning_rate": 9.481776681512272e-05, "loss": 0.0648, "step": 37780 }, { "epoch": 1.277595591466919, "grad_norm": 0.4181382954120636, "learning_rate": 9.481410098619592e-05, "loss": 0.0577, "step": 37790 }, { "epoch": 1.2779336691571723, "grad_norm": 0.27728337049484253, "learning_rate": 9.481043393206845e-05, "loss": 0.0586, "step": 37800 }, { "epoch": 1.2782717468474256, "grad_norm": 0.26741352677345276, "learning_rate": 9.480676565284054e-05, "loss": 0.0633, "step": 37810 }, { "epoch": 1.2786098245376787, "grad_norm": 0.4476354420185089, "learning_rate": 9.480309614861249e-05, "loss": 0.0621, "step": 37820 }, { "epoch": 1.278947902227932, "grad_norm": 0.4952319264411926, "learning_rate": 9.479942541948463e-05, "loss": 0.0635, "step": 37830 }, { "epoch": 1.2792859799181853, "grad_norm": 0.5012025237083435, "learning_rate": 9.47957534655573e-05, "loss": 0.0602, "step": 37840 }, { "epoch": 1.2796240576084383, "grad_norm": 0.3138914704322815, "learning_rate": 9.47920802869309e-05, "loss": 0.06, "step": 37850 }, { "epoch": 1.2799621352986916, "grad_norm": 0.2748347818851471, "learning_rate": 9.478840588370585e-05, "loss": 0.056, "step": 37860 }, { "epoch": 1.280300212988945, "grad_norm": 0.2518673539161682, "learning_rate": 9.478473025598263e-05, "loss": 0.0615, "step": 37870 }, { "epoch": 1.2806382906791982, "grad_norm": 0.3921326696872711, "learning_rate": 9.47810534038617e-05, "loss": 0.0626, "step": 37880 }, { "epoch": 1.2809763683694513, "grad_norm": 0.1904856115579605, "learning_rate": 9.477737532744357e-05, "loss": 0.0483, "step": 37890 }, { "epoch": 1.2813144460597046, "grad_norm": 0.7002211213111877, "learning_rate": 9.477369602682885e-05, "loss": 0.0601, "step": 37900 }, { "epoch": 1.2816525237499579, "grad_norm": 0.2489643543958664, "learning_rate": 9.477001550211808e-05, "loss": 0.0623, "step": 37910 }, { "epoch": 1.281990601440211, "grad_norm": 0.40447625517845154, "learning_rate": 9.476633375341193e-05, "loss": 0.0601, "step": 37920 }, { "epoch": 1.2823286791304642, "grad_norm": 0.33317846059799194, "learning_rate": 9.4762650780811e-05, "loss": 0.0566, "step": 37930 }, { "epoch": 1.2826667568207175, "grad_norm": 0.34613120555877686, "learning_rate": 9.475896658441605e-05, "loss": 0.0651, "step": 37940 }, { "epoch": 1.2830048345109706, "grad_norm": 0.43229275941848755, "learning_rate": 9.475528116432775e-05, "loss": 0.059, "step": 37950 }, { "epoch": 1.2833429122012239, "grad_norm": 0.538637638092041, "learning_rate": 9.475159452064688e-05, "loss": 0.0611, "step": 37960 }, { "epoch": 1.2836809898914772, "grad_norm": 0.2506363093852997, "learning_rate": 9.474790665347423e-05, "loss": 0.0602, "step": 37970 }, { "epoch": 1.2840190675817302, "grad_norm": 0.356136292219162, "learning_rate": 9.474421756291061e-05, "loss": 0.0552, "step": 37980 }, { "epoch": 1.2843571452719835, "grad_norm": 0.2776748538017273, "learning_rate": 9.47405272490569e-05, "loss": 0.056, "step": 37990 }, { "epoch": 1.2846952229622368, "grad_norm": 0.19682824611663818, "learning_rate": 9.473683571201399e-05, "loss": 0.063, "step": 38000 }, { "epoch": 1.2850333006524899, "grad_norm": 0.2471410483121872, "learning_rate": 9.47331429518828e-05, "loss": 0.0574, "step": 38010 }, { "epoch": 1.2853713783427432, "grad_norm": 0.3106364905834198, "learning_rate": 9.472944896876427e-05, "loss": 0.06, "step": 38020 }, { "epoch": 1.2857094560329965, "grad_norm": 0.21313172578811646, "learning_rate": 9.472575376275943e-05, "loss": 0.0555, "step": 38030 }, { "epoch": 1.2860475337232495, "grad_norm": 0.3268834352493286, "learning_rate": 9.472205733396926e-05, "loss": 0.0581, "step": 38040 }, { "epoch": 1.2863856114135028, "grad_norm": 0.29569461941719055, "learning_rate": 9.471835968249487e-05, "loss": 0.0542, "step": 38050 }, { "epoch": 1.2867236891037561, "grad_norm": 0.38287732005119324, "learning_rate": 9.471466080843732e-05, "loss": 0.0557, "step": 38060 }, { "epoch": 1.2870617667940092, "grad_norm": 0.24639253318309784, "learning_rate": 9.471096071189774e-05, "loss": 0.0505, "step": 38070 }, { "epoch": 1.2873998444842625, "grad_norm": 0.39430132508277893, "learning_rate": 9.470725939297727e-05, "loss": 0.0616, "step": 38080 }, { "epoch": 1.2877379221745158, "grad_norm": 0.42201659083366394, "learning_rate": 9.470355685177714e-05, "loss": 0.0559, "step": 38090 }, { "epoch": 1.2880759998647688, "grad_norm": 0.2515283524990082, "learning_rate": 9.469985308839857e-05, "loss": 0.0581, "step": 38100 }, { "epoch": 1.2884140775550221, "grad_norm": 0.26216158270835876, "learning_rate": 9.469614810294281e-05, "loss": 0.0652, "step": 38110 }, { "epoch": 1.2887521552452754, "grad_norm": 0.2998580038547516, "learning_rate": 9.469244189551114e-05, "loss": 0.0594, "step": 38120 }, { "epoch": 1.2890902329355285, "grad_norm": 0.18376216292381287, "learning_rate": 9.46887344662049e-05, "loss": 0.0554, "step": 38130 }, { "epoch": 1.2894283106257818, "grad_norm": 0.26626071333885193, "learning_rate": 9.468502581512543e-05, "loss": 0.0576, "step": 38140 }, { "epoch": 1.289766388316035, "grad_norm": 0.2293814718723297, "learning_rate": 9.468131594237419e-05, "loss": 0.0517, "step": 38150 }, { "epoch": 1.2901044660062881, "grad_norm": 0.4262256920337677, "learning_rate": 9.467760484805253e-05, "loss": 0.0645, "step": 38160 }, { "epoch": 1.2904425436965414, "grad_norm": 0.426312655210495, "learning_rate": 9.467389253226193e-05, "loss": 0.0556, "step": 38170 }, { "epoch": 1.2907806213867947, "grad_norm": 0.29588618874549866, "learning_rate": 9.46701789951039e-05, "loss": 0.0597, "step": 38180 }, { "epoch": 1.2911186990770478, "grad_norm": 0.28652071952819824, "learning_rate": 9.466646423667996e-05, "loss": 0.0611, "step": 38190 }, { "epoch": 1.291456776767301, "grad_norm": 0.2815987765789032, "learning_rate": 9.466274825709166e-05, "loss": 0.056, "step": 38200 }, { "epoch": 1.2917948544575544, "grad_norm": 0.2149786651134491, "learning_rate": 9.465903105644061e-05, "loss": 0.0544, "step": 38210 }, { "epoch": 1.2921329321478074, "grad_norm": 0.2869543433189392, "learning_rate": 9.465531263482843e-05, "loss": 0.0626, "step": 38220 }, { "epoch": 1.2924710098380607, "grad_norm": 0.31891268491744995, "learning_rate": 9.465159299235677e-05, "loss": 0.0587, "step": 38230 }, { "epoch": 1.292809087528314, "grad_norm": 0.7660833597183228, "learning_rate": 9.464787212912733e-05, "loss": 0.0558, "step": 38240 }, { "epoch": 1.293147165218567, "grad_norm": 0.5515622496604919, "learning_rate": 9.464415004524185e-05, "loss": 0.0536, "step": 38250 }, { "epoch": 1.2934852429088204, "grad_norm": 0.4027012884616852, "learning_rate": 9.46404267408021e-05, "loss": 0.0572, "step": 38260 }, { "epoch": 1.2938233205990737, "grad_norm": 0.238433837890625, "learning_rate": 9.463670221590982e-05, "loss": 0.054, "step": 38270 }, { "epoch": 1.294161398289327, "grad_norm": 0.3660764694213867, "learning_rate": 9.463297647066687e-05, "loss": 0.0568, "step": 38280 }, { "epoch": 1.29449947597958, "grad_norm": 0.40744930505752563, "learning_rate": 9.462924950517514e-05, "loss": 0.0565, "step": 38290 }, { "epoch": 1.2948375536698333, "grad_norm": 0.3374882936477661, "learning_rate": 9.462552131953646e-05, "loss": 0.058, "step": 38300 }, { "epoch": 1.2951756313600866, "grad_norm": 0.36303627490997314, "learning_rate": 9.462179191385281e-05, "loss": 0.0552, "step": 38310 }, { "epoch": 1.2955137090503397, "grad_norm": 0.3381247818470001, "learning_rate": 9.461806128822614e-05, "loss": 0.0549, "step": 38320 }, { "epoch": 1.295851786740593, "grad_norm": 0.2696290910243988, "learning_rate": 9.461432944275843e-05, "loss": 0.0625, "step": 38330 }, { "epoch": 1.2961898644308463, "grad_norm": 0.31061244010925293, "learning_rate": 9.46105963775517e-05, "loss": 0.0599, "step": 38340 }, { "epoch": 1.2965279421210996, "grad_norm": 0.3024449646472931, "learning_rate": 9.460686209270803e-05, "loss": 0.0582, "step": 38350 }, { "epoch": 1.2968660198113526, "grad_norm": 0.2688199579715729, "learning_rate": 9.460312658832953e-05, "loss": 0.0521, "step": 38360 }, { "epoch": 1.297204097501606, "grad_norm": 0.3981223702430725, "learning_rate": 9.459938986451828e-05, "loss": 0.056, "step": 38370 }, { "epoch": 1.2975421751918592, "grad_norm": 0.3967120945453644, "learning_rate": 9.45956519213765e-05, "loss": 0.0568, "step": 38380 }, { "epoch": 1.2978802528821123, "grad_norm": 0.3126355707645416, "learning_rate": 9.459191275900632e-05, "loss": 0.0568, "step": 38390 }, { "epoch": 1.2982183305723656, "grad_norm": 0.33035409450531006, "learning_rate": 9.458817237751e-05, "loss": 0.058, "step": 38400 }, { "epoch": 1.2985564082626189, "grad_norm": 0.28116467595100403, "learning_rate": 9.458443077698979e-05, "loss": 0.0595, "step": 38410 }, { "epoch": 1.298894485952872, "grad_norm": 0.3233063817024231, "learning_rate": 9.458068795754802e-05, "loss": 0.055, "step": 38420 }, { "epoch": 1.2992325636431252, "grad_norm": 0.3599449396133423, "learning_rate": 9.457694391928698e-05, "loss": 0.0579, "step": 38430 }, { "epoch": 1.2995706413333785, "grad_norm": 0.27225446701049805, "learning_rate": 9.457319866230902e-05, "loss": 0.0579, "step": 38440 }, { "epoch": 1.2999087190236316, "grad_norm": 0.28416359424591064, "learning_rate": 9.456945218671657e-05, "loss": 0.0627, "step": 38450 }, { "epoch": 1.3002467967138849, "grad_norm": 0.3594249486923218, "learning_rate": 9.456570449261203e-05, "loss": 0.0587, "step": 38460 }, { "epoch": 1.3005848744041382, "grad_norm": 0.22801747918128967, "learning_rate": 9.456195558009791e-05, "loss": 0.0626, "step": 38470 }, { "epoch": 1.3009229520943912, "grad_norm": 0.3386763334274292, "learning_rate": 9.455820544927662e-05, "loss": 0.0608, "step": 38480 }, { "epoch": 1.3012610297846445, "grad_norm": 0.20355696976184845, "learning_rate": 9.455445410025076e-05, "loss": 0.0552, "step": 38490 }, { "epoch": 1.3015991074748978, "grad_norm": 0.2507134675979614, "learning_rate": 9.455070153312285e-05, "loss": 0.057, "step": 38500 }, { "epoch": 1.301937185165151, "grad_norm": 0.3402285575866699, "learning_rate": 9.45469477479955e-05, "loss": 0.0586, "step": 38510 }, { "epoch": 1.3022752628554042, "grad_norm": 0.3381960093975067, "learning_rate": 9.454319274497134e-05, "loss": 0.0582, "step": 38520 }, { "epoch": 1.3026133405456575, "grad_norm": 0.27371945977211, "learning_rate": 9.453943652415302e-05, "loss": 0.0505, "step": 38530 }, { "epoch": 1.3029514182359105, "grad_norm": 0.21554087102413177, "learning_rate": 9.453567908564323e-05, "loss": 0.0539, "step": 38540 }, { "epoch": 1.3032894959261638, "grad_norm": 0.2660341262817383, "learning_rate": 9.453192042954471e-05, "loss": 0.066, "step": 38550 }, { "epoch": 1.3036275736164171, "grad_norm": 0.21675299108028412, "learning_rate": 9.452816055596022e-05, "loss": 0.058, "step": 38560 }, { "epoch": 1.3039656513066702, "grad_norm": 0.24253085255622864, "learning_rate": 9.452439946499254e-05, "loss": 0.0585, "step": 38570 }, { "epoch": 1.3043037289969235, "grad_norm": 0.4275231957435608, "learning_rate": 9.452063715674453e-05, "loss": 0.0556, "step": 38580 }, { "epoch": 1.3046418066871768, "grad_norm": 0.25916141271591187, "learning_rate": 9.451687363131901e-05, "loss": 0.0586, "step": 38590 }, { "epoch": 1.3049798843774298, "grad_norm": 0.24097879230976105, "learning_rate": 9.451310888881888e-05, "loss": 0.0546, "step": 38600 }, { "epoch": 1.3053179620676831, "grad_norm": 0.203591987490654, "learning_rate": 9.450934292934708e-05, "loss": 0.0529, "step": 38610 }, { "epoch": 1.3056560397579364, "grad_norm": 0.2962130308151245, "learning_rate": 9.450557575300655e-05, "loss": 0.0581, "step": 38620 }, { "epoch": 1.3059941174481895, "grad_norm": 0.3115135431289673, "learning_rate": 9.450180735990032e-05, "loss": 0.0604, "step": 38630 }, { "epoch": 1.3063321951384428, "grad_norm": 0.27123481035232544, "learning_rate": 9.44980377501314e-05, "loss": 0.0587, "step": 38640 }, { "epoch": 1.306670272828696, "grad_norm": 0.3331413269042969, "learning_rate": 9.449426692380284e-05, "loss": 0.058, "step": 38650 }, { "epoch": 1.3070083505189491, "grad_norm": 0.16045641899108887, "learning_rate": 9.449049488101773e-05, "loss": 0.0517, "step": 38660 }, { "epoch": 1.3073464282092024, "grad_norm": 0.2981153428554535, "learning_rate": 9.448672162187921e-05, "loss": 0.0517, "step": 38670 }, { "epoch": 1.3076845058994557, "grad_norm": 0.3400546908378601, "learning_rate": 9.448294714649044e-05, "loss": 0.0616, "step": 38680 }, { "epoch": 1.3080225835897088, "grad_norm": 0.234282448887825, "learning_rate": 9.447917145495459e-05, "loss": 0.0537, "step": 38690 }, { "epoch": 1.308360661279962, "grad_norm": 0.2597588896751404, "learning_rate": 9.447539454737493e-05, "loss": 0.0594, "step": 38700 }, { "epoch": 1.3086987389702154, "grad_norm": 0.243505597114563, "learning_rate": 9.447161642385467e-05, "loss": 0.0566, "step": 38710 }, { "epoch": 1.3090368166604684, "grad_norm": 0.24403069913387299, "learning_rate": 9.446783708449714e-05, "loss": 0.062, "step": 38720 }, { "epoch": 1.3093748943507217, "grad_norm": 0.8710570931434631, "learning_rate": 9.446405652940562e-05, "loss": 0.0657, "step": 38730 }, { "epoch": 1.309712972040975, "grad_norm": 0.2151639759540558, "learning_rate": 9.446027475868352e-05, "loss": 0.0531, "step": 38740 }, { "epoch": 1.310051049731228, "grad_norm": 0.29574331641197205, "learning_rate": 9.445649177243421e-05, "loss": 0.0584, "step": 38750 }, { "epoch": 1.3103891274214814, "grad_norm": 0.3054894208908081, "learning_rate": 9.445270757076112e-05, "loss": 0.0574, "step": 38760 }, { "epoch": 1.3107272051117347, "grad_norm": 0.3022472560405731, "learning_rate": 9.44489221537677e-05, "loss": 0.0657, "step": 38770 }, { "epoch": 1.311065282801988, "grad_norm": 0.28632521629333496, "learning_rate": 9.444513552155744e-05, "loss": 0.0584, "step": 38780 }, { "epoch": 1.311403360492241, "grad_norm": 0.24899645149707794, "learning_rate": 9.444134767423388e-05, "loss": 0.0553, "step": 38790 }, { "epoch": 1.3117414381824943, "grad_norm": 0.2126113921403885, "learning_rate": 9.443755861190058e-05, "loss": 0.054, "step": 38800 }, { "epoch": 1.3120795158727476, "grad_norm": 0.39090919494628906, "learning_rate": 9.44337683346611e-05, "loss": 0.0619, "step": 38810 }, { "epoch": 1.3124175935630007, "grad_norm": 0.18460752069950104, "learning_rate": 9.44299768426191e-05, "loss": 0.0556, "step": 38820 }, { "epoch": 1.312755671253254, "grad_norm": 0.2440101057291031, "learning_rate": 9.442618413587822e-05, "loss": 0.053, "step": 38830 }, { "epoch": 1.3130937489435073, "grad_norm": 0.22927768528461456, "learning_rate": 9.442239021454215e-05, "loss": 0.0583, "step": 38840 }, { "epoch": 1.3134318266337606, "grad_norm": 0.2035011202096939, "learning_rate": 9.441859507871465e-05, "loss": 0.0566, "step": 38850 }, { "epoch": 1.3137699043240136, "grad_norm": 0.19522689282894135, "learning_rate": 9.44147987284994e-05, "loss": 0.052, "step": 38860 }, { "epoch": 1.314107982014267, "grad_norm": 0.39717456698417664, "learning_rate": 9.441100116400028e-05, "loss": 0.0603, "step": 38870 }, { "epoch": 1.3144460597045202, "grad_norm": 0.3779821991920471, "learning_rate": 9.440720238532106e-05, "loss": 0.0611, "step": 38880 }, { "epoch": 1.3147841373947733, "grad_norm": 0.373574435710907, "learning_rate": 9.440340239256561e-05, "loss": 0.0633, "step": 38890 }, { "epoch": 1.3151222150850266, "grad_norm": 0.17038054764270782, "learning_rate": 9.439960118583782e-05, "loss": 0.0618, "step": 38900 }, { "epoch": 1.3154602927752799, "grad_norm": 0.7250527143478394, "learning_rate": 9.43957987652416e-05, "loss": 0.0532, "step": 38910 }, { "epoch": 1.315798370465533, "grad_norm": 0.23487746715545654, "learning_rate": 9.439199513088094e-05, "loss": 0.0613, "step": 38920 }, { "epoch": 1.3161364481557862, "grad_norm": 0.4264785647392273, "learning_rate": 9.43881902828598e-05, "loss": 0.0576, "step": 38930 }, { "epoch": 1.3164745258460395, "grad_norm": 0.20964257419109344, "learning_rate": 9.438438422128221e-05, "loss": 0.0537, "step": 38940 }, { "epoch": 1.3168126035362926, "grad_norm": 0.3201746642589569, "learning_rate": 9.438057694625224e-05, "loss": 0.0562, "step": 38950 }, { "epoch": 1.3171506812265459, "grad_norm": 0.25210386514663696, "learning_rate": 9.437676845787397e-05, "loss": 0.0575, "step": 38960 }, { "epoch": 1.3174887589167992, "grad_norm": 0.46636390686035156, "learning_rate": 9.437295875625151e-05, "loss": 0.0549, "step": 38970 }, { "epoch": 1.3178268366070522, "grad_norm": 0.2613075375556946, "learning_rate": 9.436914784148904e-05, "loss": 0.0555, "step": 38980 }, { "epoch": 1.3181649142973055, "grad_norm": 0.3078833818435669, "learning_rate": 9.436533571369073e-05, "loss": 0.0559, "step": 38990 }, { "epoch": 1.3185029919875588, "grad_norm": 0.3641972243785858, "learning_rate": 9.43615223729608e-05, "loss": 0.0552, "step": 39000 }, { "epoch": 1.318841069677812, "grad_norm": 0.3276596665382385, "learning_rate": 9.435770781940354e-05, "loss": 0.0601, "step": 39010 }, { "epoch": 1.3191791473680652, "grad_norm": 0.28296101093292236, "learning_rate": 9.435389205312319e-05, "loss": 0.0583, "step": 39020 }, { "epoch": 1.3195172250583185, "grad_norm": 0.2030395120382309, "learning_rate": 9.43500750742241e-05, "loss": 0.0603, "step": 39030 }, { "epoch": 1.3198553027485715, "grad_norm": 0.28509801626205444, "learning_rate": 9.434625688281063e-05, "loss": 0.0594, "step": 39040 }, { "epoch": 1.3201933804388248, "grad_norm": 0.3368168771266937, "learning_rate": 9.434243747898714e-05, "loss": 0.0578, "step": 39050 }, { "epoch": 1.3205314581290781, "grad_norm": 0.3098374307155609, "learning_rate": 9.433861686285808e-05, "loss": 0.0525, "step": 39060 }, { "epoch": 1.3208695358193312, "grad_norm": 0.35902541875839233, "learning_rate": 9.433479503452789e-05, "loss": 0.0566, "step": 39070 }, { "epoch": 1.3212076135095845, "grad_norm": 0.2966240644454956, "learning_rate": 9.433097199410106e-05, "loss": 0.0557, "step": 39080 }, { "epoch": 1.3215456911998378, "grad_norm": 0.3260064423084259, "learning_rate": 9.43271477416821e-05, "loss": 0.0578, "step": 39090 }, { "epoch": 1.3218837688900908, "grad_norm": 0.37620747089385986, "learning_rate": 9.432332227737559e-05, "loss": 0.0541, "step": 39100 }, { "epoch": 1.3222218465803441, "grad_norm": 0.26559001207351685, "learning_rate": 9.431949560128607e-05, "loss": 0.0616, "step": 39110 }, { "epoch": 1.3225599242705974, "grad_norm": 0.26763856410980225, "learning_rate": 9.431566771351823e-05, "loss": 0.0545, "step": 39120 }, { "epoch": 1.3228980019608505, "grad_norm": 0.23049227893352509, "learning_rate": 9.431183861417666e-05, "loss": 0.0482, "step": 39130 }, { "epoch": 1.3232360796511038, "grad_norm": 0.2967974543571472, "learning_rate": 9.430800830336608e-05, "loss": 0.0555, "step": 39140 }, { "epoch": 1.323574157341357, "grad_norm": 0.2220972329378128, "learning_rate": 9.430417678119119e-05, "loss": 0.06, "step": 39150 }, { "epoch": 1.3239122350316102, "grad_norm": 0.2261386662721634, "learning_rate": 9.430034404775676e-05, "loss": 0.0564, "step": 39160 }, { "epoch": 1.3242503127218634, "grad_norm": 0.22600509226322174, "learning_rate": 9.429651010316756e-05, "loss": 0.0553, "step": 39170 }, { "epoch": 1.3245883904121167, "grad_norm": 0.24527204036712646, "learning_rate": 9.429267494752842e-05, "loss": 0.0597, "step": 39180 }, { "epoch": 1.3249264681023698, "grad_norm": 0.2805189788341522, "learning_rate": 9.428883858094418e-05, "loss": 0.0624, "step": 39190 }, { "epoch": 1.325264545792623, "grad_norm": 0.6019434332847595, "learning_rate": 9.428500100351973e-05, "loss": 0.0533, "step": 39200 }, { "epoch": 1.3256026234828764, "grad_norm": 0.16758139431476593, "learning_rate": 9.428116221536e-05, "loss": 0.055, "step": 39210 }, { "epoch": 1.3259407011731295, "grad_norm": 0.40990713238716125, "learning_rate": 9.427732221656992e-05, "loss": 0.0528, "step": 39220 }, { "epoch": 1.3262787788633827, "grad_norm": 0.27787744998931885, "learning_rate": 9.427348100725448e-05, "loss": 0.0581, "step": 39230 }, { "epoch": 1.326616856553636, "grad_norm": 0.3965572714805603, "learning_rate": 9.426963858751872e-05, "loss": 0.0564, "step": 39240 }, { "epoch": 1.3269549342438893, "grad_norm": 0.315885454416275, "learning_rate": 9.426579495746767e-05, "loss": 0.0699, "step": 39250 }, { "epoch": 1.3272930119341424, "grad_norm": 0.1977500468492508, "learning_rate": 9.42619501172064e-05, "loss": 0.0611, "step": 39260 }, { "epoch": 1.3276310896243957, "grad_norm": 0.2904472053050995, "learning_rate": 9.425810406684006e-05, "loss": 0.0637, "step": 39270 }, { "epoch": 1.327969167314649, "grad_norm": 0.47392961382865906, "learning_rate": 9.425425680647377e-05, "loss": 0.0617, "step": 39280 }, { "epoch": 1.328307245004902, "grad_norm": 0.6936495304107666, "learning_rate": 9.425040833621273e-05, "loss": 0.0521, "step": 39290 }, { "epoch": 1.3286453226951553, "grad_norm": 0.4524950087070465, "learning_rate": 9.424655865616215e-05, "loss": 0.0575, "step": 39300 }, { "epoch": 1.3289834003854086, "grad_norm": 0.3147839903831482, "learning_rate": 9.424270776642728e-05, "loss": 0.0613, "step": 39310 }, { "epoch": 1.329321478075662, "grad_norm": 0.5209781527519226, "learning_rate": 9.423885566711339e-05, "loss": 0.0584, "step": 39320 }, { "epoch": 1.329659555765915, "grad_norm": 0.29998305439949036, "learning_rate": 9.423500235832584e-05, "loss": 0.0604, "step": 39330 }, { "epoch": 1.3299976334561683, "grad_norm": 0.29218789935112, "learning_rate": 9.42311478401699e-05, "loss": 0.0526, "step": 39340 }, { "epoch": 1.3303357111464216, "grad_norm": 0.4248081147670746, "learning_rate": 9.422729211275101e-05, "loss": 0.0583, "step": 39350 }, { "epoch": 1.3306737888366746, "grad_norm": 0.9396543502807617, "learning_rate": 9.422343517617458e-05, "loss": 0.055, "step": 39360 }, { "epoch": 1.331011866526928, "grad_norm": 0.4020707309246063, "learning_rate": 9.421957703054604e-05, "loss": 0.0546, "step": 39370 }, { "epoch": 1.3313499442171812, "grad_norm": 0.23553459346294403, "learning_rate": 9.421571767597086e-05, "loss": 0.0571, "step": 39380 }, { "epoch": 1.3316880219074343, "grad_norm": 0.3022969663143158, "learning_rate": 9.42118571125546e-05, "loss": 0.0515, "step": 39390 }, { "epoch": 1.3320260995976876, "grad_norm": 0.2998531758785248, "learning_rate": 9.420799534040275e-05, "loss": 0.0558, "step": 39400 }, { "epoch": 1.3323641772879409, "grad_norm": 0.1647636890411377, "learning_rate": 9.420413235962094e-05, "loss": 0.0498, "step": 39410 }, { "epoch": 1.332702254978194, "grad_norm": 0.27819523215293884, "learning_rate": 9.420026817031473e-05, "loss": 0.0673, "step": 39420 }, { "epoch": 1.3330403326684472, "grad_norm": 0.2554807662963867, "learning_rate": 9.419640277258981e-05, "loss": 0.058, "step": 39430 }, { "epoch": 1.3333784103587005, "grad_norm": 0.278929740190506, "learning_rate": 9.419253616655185e-05, "loss": 0.0565, "step": 39440 }, { "epoch": 1.3337164880489536, "grad_norm": 0.9332268238067627, "learning_rate": 9.418866835230654e-05, "loss": 0.0577, "step": 39450 }, { "epoch": 1.3340545657392069, "grad_norm": 0.27927836775779724, "learning_rate": 9.418479932995965e-05, "loss": 0.0592, "step": 39460 }, { "epoch": 1.3343926434294602, "grad_norm": 0.29357558488845825, "learning_rate": 9.418092909961693e-05, "loss": 0.0603, "step": 39470 }, { "epoch": 1.3347307211197132, "grad_norm": 0.5220485329627991, "learning_rate": 9.417705766138422e-05, "loss": 0.0589, "step": 39480 }, { "epoch": 1.3350687988099665, "grad_norm": 0.226174995303154, "learning_rate": 9.417318501536733e-05, "loss": 0.0531, "step": 39490 }, { "epoch": 1.3354068765002198, "grad_norm": 0.26341357827186584, "learning_rate": 9.416931116167218e-05, "loss": 0.0567, "step": 39500 }, { "epoch": 1.335744954190473, "grad_norm": 0.2510470449924469, "learning_rate": 9.416543610040465e-05, "loss": 0.0592, "step": 39510 }, { "epoch": 1.3360830318807262, "grad_norm": 0.2988089621067047, "learning_rate": 9.416155983167068e-05, "loss": 0.0553, "step": 39520 }, { "epoch": 1.3364211095709795, "grad_norm": 0.2414190024137497, "learning_rate": 9.415768235557627e-05, "loss": 0.057, "step": 39530 }, { "epoch": 1.3367591872612326, "grad_norm": 0.21146470308303833, "learning_rate": 9.415380367222739e-05, "loss": 0.058, "step": 39540 }, { "epoch": 1.3370972649514858, "grad_norm": 0.30777567625045776, "learning_rate": 9.414992378173013e-05, "loss": 0.0633, "step": 39550 }, { "epoch": 1.3374353426417391, "grad_norm": 0.5645135045051575, "learning_rate": 9.414604268419054e-05, "loss": 0.0572, "step": 39560 }, { "epoch": 1.3377734203319922, "grad_norm": 0.29577550292015076, "learning_rate": 9.414216037971471e-05, "loss": 0.0574, "step": 39570 }, { "epoch": 1.3381114980222455, "grad_norm": 0.46588414907455444, "learning_rate": 9.413827686840881e-05, "loss": 0.0645, "step": 39580 }, { "epoch": 1.3384495757124988, "grad_norm": 0.3151242434978485, "learning_rate": 9.4134392150379e-05, "loss": 0.0535, "step": 39590 }, { "epoch": 1.3387876534027519, "grad_norm": 0.3092568516731262, "learning_rate": 9.413050622573148e-05, "loss": 0.0589, "step": 39600 }, { "epoch": 1.3391257310930051, "grad_norm": 0.3750821352005005, "learning_rate": 9.412661909457252e-05, "loss": 0.0634, "step": 39610 }, { "epoch": 1.3394638087832584, "grad_norm": 0.2407669574022293, "learning_rate": 9.412273075700835e-05, "loss": 0.052, "step": 39620 }, { "epoch": 1.3398018864735115, "grad_norm": 0.19518810510635376, "learning_rate": 9.411884121314532e-05, "loss": 0.062, "step": 39630 }, { "epoch": 1.3401399641637648, "grad_norm": 0.19954080879688263, "learning_rate": 9.411495046308973e-05, "loss": 0.053, "step": 39640 }, { "epoch": 1.340478041854018, "grad_norm": 0.2664835751056671, "learning_rate": 9.411105850694796e-05, "loss": 0.0578, "step": 39650 }, { "epoch": 1.3408161195442712, "grad_norm": 1.0157616138458252, "learning_rate": 9.410716534482644e-05, "loss": 0.0575, "step": 39660 }, { "epoch": 1.3411541972345244, "grad_norm": 0.26073241233825684, "learning_rate": 9.410327097683158e-05, "loss": 0.0578, "step": 39670 }, { "epoch": 1.3414922749247777, "grad_norm": 0.42807185649871826, "learning_rate": 9.409937540306986e-05, "loss": 0.0561, "step": 39680 }, { "epoch": 1.3418303526150308, "grad_norm": 0.27707961201667786, "learning_rate": 9.409547862364778e-05, "loss": 0.0649, "step": 39690 }, { "epoch": 1.342168430305284, "grad_norm": 0.39025038480758667, "learning_rate": 9.40915806386719e-05, "loss": 0.0584, "step": 39700 }, { "epoch": 1.3425065079955374, "grad_norm": 0.3281348943710327, "learning_rate": 9.408768144824875e-05, "loss": 0.0541, "step": 39710 }, { "epoch": 1.3428445856857905, "grad_norm": 0.2739267945289612, "learning_rate": 9.408378105248496e-05, "loss": 0.058, "step": 39720 }, { "epoch": 1.3431826633760438, "grad_norm": 0.3128276765346527, "learning_rate": 9.407987945148716e-05, "loss": 0.0557, "step": 39730 }, { "epoch": 1.343520741066297, "grad_norm": 0.19504773616790771, "learning_rate": 9.4075976645362e-05, "loss": 0.0571, "step": 39740 }, { "epoch": 1.3438588187565503, "grad_norm": 0.31111592054367065, "learning_rate": 9.407207263421622e-05, "loss": 0.062, "step": 39750 }, { "epoch": 1.3441968964468034, "grad_norm": 0.25839415192604065, "learning_rate": 9.406816741815652e-05, "loss": 0.0597, "step": 39760 }, { "epoch": 1.3445349741370567, "grad_norm": 0.3910481333732605, "learning_rate": 9.406426099728967e-05, "loss": 0.0635, "step": 39770 }, { "epoch": 1.34487305182731, "grad_norm": 0.32225844264030457, "learning_rate": 9.40603533717225e-05, "loss": 0.0581, "step": 39780 }, { "epoch": 1.345211129517563, "grad_norm": 0.28217703104019165, "learning_rate": 9.40564445415618e-05, "loss": 0.0629, "step": 39790 }, { "epoch": 1.3455492072078163, "grad_norm": 0.2125421166419983, "learning_rate": 9.405253450691446e-05, "loss": 0.0571, "step": 39800 }, { "epoch": 1.3458872848980696, "grad_norm": 0.26747384667396545, "learning_rate": 9.404862326788739e-05, "loss": 0.0582, "step": 39810 }, { "epoch": 1.346225362588323, "grad_norm": 0.3593255579471588, "learning_rate": 9.40447108245875e-05, "loss": 0.065, "step": 39820 }, { "epoch": 1.346563440278576, "grad_norm": 0.482174813747406, "learning_rate": 9.404079717712176e-05, "loss": 0.0583, "step": 39830 }, { "epoch": 1.3469015179688293, "grad_norm": 0.32048290967941284, "learning_rate": 9.403688232559718e-05, "loss": 0.053, "step": 39840 }, { "epoch": 1.3472395956590826, "grad_norm": 0.27546656131744385, "learning_rate": 9.403296627012077e-05, "loss": 0.0596, "step": 39850 }, { "epoch": 1.3475776733493356, "grad_norm": 0.408174067735672, "learning_rate": 9.402904901079961e-05, "loss": 0.0594, "step": 39860 }, { "epoch": 1.347915751039589, "grad_norm": 0.42110374569892883, "learning_rate": 9.40251305477408e-05, "loss": 0.0537, "step": 39870 }, { "epoch": 1.3482538287298422, "grad_norm": 0.2893964648246765, "learning_rate": 9.402121088105144e-05, "loss": 0.0532, "step": 39880 }, { "epoch": 1.3485919064200953, "grad_norm": 0.49463164806365967, "learning_rate": 9.401729001083873e-05, "loss": 0.0642, "step": 39890 }, { "epoch": 1.3489299841103486, "grad_norm": 0.40281441807746887, "learning_rate": 9.401336793720983e-05, "loss": 0.0599, "step": 39900 }, { "epoch": 1.3492680618006019, "grad_norm": 0.3677949905395508, "learning_rate": 9.400944466027201e-05, "loss": 0.0586, "step": 39910 }, { "epoch": 1.349606139490855, "grad_norm": 0.27382221817970276, "learning_rate": 9.40055201801325e-05, "loss": 0.0513, "step": 39920 }, { "epoch": 1.3499442171811082, "grad_norm": 0.2957743704319, "learning_rate": 9.400159449689858e-05, "loss": 0.0564, "step": 39930 }, { "epoch": 1.3502822948713615, "grad_norm": 0.24147090315818787, "learning_rate": 9.399766761067761e-05, "loss": 0.0538, "step": 39940 }, { "epoch": 1.3506203725616146, "grad_norm": 0.40576696395874023, "learning_rate": 9.399373952157694e-05, "loss": 0.0528, "step": 39950 }, { "epoch": 1.350958450251868, "grad_norm": 0.2765306234359741, "learning_rate": 9.398981022970395e-05, "loss": 0.056, "step": 39960 }, { "epoch": 1.3512965279421212, "grad_norm": 0.34026071429252625, "learning_rate": 9.398587973516608e-05, "loss": 0.0576, "step": 39970 }, { "epoch": 1.3516346056323743, "grad_norm": 0.25355473160743713, "learning_rate": 9.398194803807078e-05, "loss": 0.0515, "step": 39980 }, { "epoch": 1.3519726833226275, "grad_norm": 0.31280866265296936, "learning_rate": 9.397801513852555e-05, "loss": 0.0573, "step": 39990 }, { "epoch": 1.3523107610128808, "grad_norm": 0.4093915522098541, "learning_rate": 9.397408103663792e-05, "loss": 0.0556, "step": 40000 }, { "epoch": 1.352648838703134, "grad_norm": 0.5673363208770752, "learning_rate": 9.397014573251541e-05, "loss": 0.0694, "step": 40010 }, { "epoch": 1.3529869163933872, "grad_norm": 0.29580044746398926, "learning_rate": 9.396620922626564e-05, "loss": 0.0586, "step": 40020 }, { "epoch": 1.3533249940836405, "grad_norm": 0.28987249732017517, "learning_rate": 9.396227151799623e-05, "loss": 0.0611, "step": 40030 }, { "epoch": 1.3536630717738936, "grad_norm": 0.17967508733272552, "learning_rate": 9.395833260781485e-05, "loss": 0.0577, "step": 40040 }, { "epoch": 1.3540011494641468, "grad_norm": 0.40224239230155945, "learning_rate": 9.395439249582916e-05, "loss": 0.0633, "step": 40050 }, { "epoch": 1.3543392271544001, "grad_norm": 0.2871398627758026, "learning_rate": 9.395045118214688e-05, "loss": 0.0597, "step": 40060 }, { "epoch": 1.3546773048446532, "grad_norm": 0.2810930907726288, "learning_rate": 9.394650866687578e-05, "loss": 0.0554, "step": 40070 }, { "epoch": 1.3550153825349065, "grad_norm": 0.3077181577682495, "learning_rate": 9.394256495012366e-05, "loss": 0.062, "step": 40080 }, { "epoch": 1.3553534602251598, "grad_norm": 0.23734904825687408, "learning_rate": 9.393862003199829e-05, "loss": 0.0549, "step": 40090 }, { "epoch": 1.3556915379154129, "grad_norm": 0.5452119708061218, "learning_rate": 9.393467391260759e-05, "loss": 0.0654, "step": 40100 }, { "epoch": 1.3560296156056661, "grad_norm": 0.26705777645111084, "learning_rate": 9.393072659205939e-05, "loss": 0.0582, "step": 40110 }, { "epoch": 1.3563676932959194, "grad_norm": 0.2236013412475586, "learning_rate": 9.392677807046164e-05, "loss": 0.0554, "step": 40120 }, { "epoch": 1.3567057709861725, "grad_norm": 0.3748600482940674, "learning_rate": 9.392282834792228e-05, "loss": 0.0623, "step": 40130 }, { "epoch": 1.3570438486764258, "grad_norm": 0.21578000485897064, "learning_rate": 9.391887742454928e-05, "loss": 0.0501, "step": 40140 }, { "epoch": 1.357381926366679, "grad_norm": 0.29760056734085083, "learning_rate": 9.391492530045068e-05, "loss": 0.057, "step": 40150 }, { "epoch": 1.3577200040569322, "grad_norm": 0.33311858773231506, "learning_rate": 9.391097197573452e-05, "loss": 0.0567, "step": 40160 }, { "epoch": 1.3580580817471855, "grad_norm": 0.33745694160461426, "learning_rate": 9.390701745050889e-05, "loss": 0.0556, "step": 40170 }, { "epoch": 1.3583961594374387, "grad_norm": 0.3198787569999695, "learning_rate": 9.390306172488188e-05, "loss": 0.052, "step": 40180 }, { "epoch": 1.3587342371276918, "grad_norm": 0.3261849284172058, "learning_rate": 9.389910479896167e-05, "loss": 0.0586, "step": 40190 }, { "epoch": 1.359072314817945, "grad_norm": 0.5221911072731018, "learning_rate": 9.389514667285642e-05, "loss": 0.0509, "step": 40200 }, { "epoch": 1.3594103925081984, "grad_norm": 0.293951153755188, "learning_rate": 9.389118734667435e-05, "loss": 0.0574, "step": 40210 }, { "epoch": 1.3597484701984517, "grad_norm": 0.2860569655895233, "learning_rate": 9.388722682052371e-05, "loss": 0.0617, "step": 40220 }, { "epoch": 1.3600865478887048, "grad_norm": 0.4197373390197754, "learning_rate": 9.388326509451278e-05, "loss": 0.0582, "step": 40230 }, { "epoch": 1.360424625578958, "grad_norm": 0.35807356238365173, "learning_rate": 9.387930216874987e-05, "loss": 0.0586, "step": 40240 }, { "epoch": 1.3607627032692113, "grad_norm": 0.5533141493797302, "learning_rate": 9.387533804334332e-05, "loss": 0.0541, "step": 40250 }, { "epoch": 1.3611007809594644, "grad_norm": 0.31021955609321594, "learning_rate": 9.387137271840151e-05, "loss": 0.059, "step": 40260 }, { "epoch": 1.3614388586497177, "grad_norm": 0.32976847887039185, "learning_rate": 9.386740619403285e-05, "loss": 0.0568, "step": 40270 }, { "epoch": 1.361776936339971, "grad_norm": 0.38634222745895386, "learning_rate": 9.386343847034578e-05, "loss": 0.0572, "step": 40280 }, { "epoch": 1.3621150140302243, "grad_norm": 0.20485630631446838, "learning_rate": 9.38594695474488e-05, "loss": 0.0604, "step": 40290 }, { "epoch": 1.3624530917204773, "grad_norm": 0.6393812894821167, "learning_rate": 9.385549942545038e-05, "loss": 0.0581, "step": 40300 }, { "epoch": 1.3627911694107306, "grad_norm": 0.4510592818260193, "learning_rate": 9.38515281044591e-05, "loss": 0.0583, "step": 40310 }, { "epoch": 1.363129247100984, "grad_norm": 0.27832651138305664, "learning_rate": 9.384755558458349e-05, "loss": 0.06, "step": 40320 }, { "epoch": 1.363467324791237, "grad_norm": 0.3450110852718353, "learning_rate": 9.38435818659322e-05, "loss": 0.0592, "step": 40330 }, { "epoch": 1.3638054024814903, "grad_norm": 0.38907405734062195, "learning_rate": 9.383960694861385e-05, "loss": 0.0528, "step": 40340 }, { "epoch": 1.3641434801717436, "grad_norm": 0.2674896717071533, "learning_rate": 9.383563083273712e-05, "loss": 0.0589, "step": 40350 }, { "epoch": 1.3644815578619967, "grad_norm": 0.27556872367858887, "learning_rate": 9.383165351841071e-05, "loss": 0.0572, "step": 40360 }, { "epoch": 1.36481963555225, "grad_norm": 0.38569507002830505, "learning_rate": 9.382767500574337e-05, "loss": 0.0587, "step": 40370 }, { "epoch": 1.3651577132425032, "grad_norm": 0.27388256788253784, "learning_rate": 9.382369529484383e-05, "loss": 0.0568, "step": 40380 }, { "epoch": 1.3654957909327563, "grad_norm": 0.38370078802108765, "learning_rate": 9.381971438582096e-05, "loss": 0.0621, "step": 40390 }, { "epoch": 1.3658338686230096, "grad_norm": 0.30806291103363037, "learning_rate": 9.381573227878353e-05, "loss": 0.0584, "step": 40400 }, { "epoch": 1.3661719463132629, "grad_norm": 0.26921555399894714, "learning_rate": 9.381174897384047e-05, "loss": 0.0607, "step": 40410 }, { "epoch": 1.366510024003516, "grad_norm": 0.3499610424041748, "learning_rate": 9.380776447110062e-05, "loss": 0.0582, "step": 40420 }, { "epoch": 1.3668481016937692, "grad_norm": 0.34475573897361755, "learning_rate": 9.380377877067297e-05, "loss": 0.0546, "step": 40430 }, { "epoch": 1.3671861793840225, "grad_norm": 0.25124043226242065, "learning_rate": 9.379979187266645e-05, "loss": 0.0586, "step": 40440 }, { "epoch": 1.3675242570742756, "grad_norm": 0.2587268054485321, "learning_rate": 9.379580377719009e-05, "loss": 0.0535, "step": 40450 }, { "epoch": 1.367862334764529, "grad_norm": 0.2623836398124695, "learning_rate": 9.37918144843529e-05, "loss": 0.056, "step": 40460 }, { "epoch": 1.3682004124547822, "grad_norm": 0.3373706340789795, "learning_rate": 9.378782399426396e-05, "loss": 0.0546, "step": 40470 }, { "epoch": 1.3685384901450353, "grad_norm": 0.40499842166900635, "learning_rate": 9.378383230703236e-05, "loss": 0.0588, "step": 40480 }, { "epoch": 1.3688765678352885, "grad_norm": 0.3613041937351227, "learning_rate": 9.377983942276723e-05, "loss": 0.0568, "step": 40490 }, { "epoch": 1.3692146455255418, "grad_norm": 0.30442014336586, "learning_rate": 9.377584534157773e-05, "loss": 0.0628, "step": 40500 }, { "epoch": 1.369552723215795, "grad_norm": 0.4696437120437622, "learning_rate": 9.377185006357307e-05, "loss": 0.0603, "step": 40510 }, { "epoch": 1.3698908009060482, "grad_norm": 0.3198702931404114, "learning_rate": 9.376785358886249e-05, "loss": 0.062, "step": 40520 }, { "epoch": 1.3702288785963015, "grad_norm": 0.2770669758319855, "learning_rate": 9.37638559175552e-05, "loss": 0.0615, "step": 40530 }, { "epoch": 1.3705669562865546, "grad_norm": 0.2904481291770935, "learning_rate": 9.375985704976057e-05, "loss": 0.0608, "step": 40540 }, { "epoch": 1.3709050339768079, "grad_norm": 0.2782052159309387, "learning_rate": 9.375585698558787e-05, "loss": 0.0596, "step": 40550 }, { "epoch": 1.3712431116670611, "grad_norm": 0.281046986579895, "learning_rate": 9.375185572514648e-05, "loss": 0.0554, "step": 40560 }, { "epoch": 1.3715811893573142, "grad_norm": 0.283798485994339, "learning_rate": 9.374785326854578e-05, "loss": 0.0535, "step": 40570 }, { "epoch": 1.3719192670475675, "grad_norm": 0.45528560876846313, "learning_rate": 9.374384961589523e-05, "loss": 0.0582, "step": 40580 }, { "epoch": 1.3722573447378208, "grad_norm": 0.4796608090400696, "learning_rate": 9.373984476730424e-05, "loss": 0.0558, "step": 40590 }, { "epoch": 1.3725954224280739, "grad_norm": 0.3627055883407593, "learning_rate": 9.373583872288235e-05, "loss": 0.0559, "step": 40600 }, { "epoch": 1.3729335001183272, "grad_norm": 0.22945547103881836, "learning_rate": 9.373183148273905e-05, "loss": 0.0587, "step": 40610 }, { "epoch": 1.3732715778085804, "grad_norm": 0.4456107020378113, "learning_rate": 9.372782304698391e-05, "loss": 0.0574, "step": 40620 }, { "epoch": 1.3736096554988335, "grad_norm": 0.3022717237472534, "learning_rate": 9.372381341572651e-05, "loss": 0.0554, "step": 40630 }, { "epoch": 1.3739477331890868, "grad_norm": 0.3366296887397766, "learning_rate": 9.371980258907647e-05, "loss": 0.0625, "step": 40640 }, { "epoch": 1.37428581087934, "grad_norm": 0.3480244278907776, "learning_rate": 9.371579056714345e-05, "loss": 0.0571, "step": 40650 }, { "epoch": 1.3746238885695932, "grad_norm": 0.21617491543293, "learning_rate": 9.371177735003716e-05, "loss": 0.0506, "step": 40660 }, { "epoch": 1.3749619662598465, "grad_norm": 0.3721858859062195, "learning_rate": 9.370776293786729e-05, "loss": 0.0535, "step": 40670 }, { "epoch": 1.3753000439500997, "grad_norm": 0.300916463136673, "learning_rate": 9.37037473307436e-05, "loss": 0.0579, "step": 40680 }, { "epoch": 1.3756381216403528, "grad_norm": 0.2879877984523773, "learning_rate": 9.369973052877587e-05, "loss": 0.058, "step": 40690 }, { "epoch": 1.375976199330606, "grad_norm": 0.32265686988830566, "learning_rate": 9.369571253207394e-05, "loss": 0.059, "step": 40700 }, { "epoch": 1.3763142770208594, "grad_norm": 0.30992981791496277, "learning_rate": 9.369169334074763e-05, "loss": 0.061, "step": 40710 }, { "epoch": 1.3766523547111127, "grad_norm": 0.2820926606655121, "learning_rate": 9.368767295490686e-05, "loss": 0.0624, "step": 40720 }, { "epoch": 1.3769904324013658, "grad_norm": 0.4577222466468811, "learning_rate": 9.36836513746615e-05, "loss": 0.0532, "step": 40730 }, { "epoch": 1.377328510091619, "grad_norm": 0.30003461241722107, "learning_rate": 9.367962860012154e-05, "loss": 0.0543, "step": 40740 }, { "epoch": 1.3776665877818723, "grad_norm": 0.2250499278306961, "learning_rate": 9.367560463139692e-05, "loss": 0.0518, "step": 40750 }, { "epoch": 1.3780046654721254, "grad_norm": 0.28430694341659546, "learning_rate": 9.36715794685977e-05, "loss": 0.0582, "step": 40760 }, { "epoch": 1.3783427431623787, "grad_norm": 0.39632993936538696, "learning_rate": 9.36675531118339e-05, "loss": 0.0551, "step": 40770 }, { "epoch": 1.378680820852632, "grad_norm": 0.3878566324710846, "learning_rate": 9.36635255612156e-05, "loss": 0.0599, "step": 40780 }, { "epoch": 1.3790188985428853, "grad_norm": 0.31880342960357666, "learning_rate": 9.365949681685291e-05, "loss": 0.0582, "step": 40790 }, { "epoch": 1.3793569762331384, "grad_norm": 0.24094364047050476, "learning_rate": 9.3655466878856e-05, "loss": 0.0619, "step": 40800 }, { "epoch": 1.3796950539233916, "grad_norm": 0.248020738363266, "learning_rate": 9.365143574733499e-05, "loss": 0.0568, "step": 40810 }, { "epoch": 1.380033131613645, "grad_norm": 0.32875797152519226, "learning_rate": 9.364740342240016e-05, "loss": 0.0524, "step": 40820 }, { "epoch": 1.380371209303898, "grad_norm": 0.30366814136505127, "learning_rate": 9.36433699041617e-05, "loss": 0.0655, "step": 40830 }, { "epoch": 1.3807092869941513, "grad_norm": 0.5339227318763733, "learning_rate": 9.36393351927299e-05, "loss": 0.0562, "step": 40840 }, { "epoch": 1.3810473646844046, "grad_norm": 0.2489510178565979, "learning_rate": 9.363529928821508e-05, "loss": 0.0608, "step": 40850 }, { "epoch": 1.3813854423746577, "grad_norm": 0.27350854873657227, "learning_rate": 9.363126219072757e-05, "loss": 0.0638, "step": 40860 }, { "epoch": 1.381723520064911, "grad_norm": 0.30448007583618164, "learning_rate": 9.362722390037775e-05, "loss": 0.0607, "step": 40870 }, { "epoch": 1.3820615977551642, "grad_norm": 0.31584295630455017, "learning_rate": 9.3623184417276e-05, "loss": 0.0576, "step": 40880 }, { "epoch": 1.3823996754454173, "grad_norm": 0.3285577893257141, "learning_rate": 9.36191437415328e-05, "loss": 0.0585, "step": 40890 }, { "epoch": 1.3827377531356706, "grad_norm": 0.3858788311481476, "learning_rate": 9.361510187325857e-05, "loss": 0.0555, "step": 40900 }, { "epoch": 1.383075830825924, "grad_norm": 0.25679588317871094, "learning_rate": 9.361105881256385e-05, "loss": 0.0513, "step": 40910 }, { "epoch": 1.383413908516177, "grad_norm": 0.39430543780326843, "learning_rate": 9.360701455955918e-05, "loss": 0.0521, "step": 40920 }, { "epoch": 1.3837519862064303, "grad_norm": 0.3034725785255432, "learning_rate": 9.360296911435509e-05, "loss": 0.0529, "step": 40930 }, { "epoch": 1.3840900638966835, "grad_norm": 0.34478145837783813, "learning_rate": 9.359892247706222e-05, "loss": 0.0573, "step": 40940 }, { "epoch": 1.3844281415869366, "grad_norm": 0.33622226119041443, "learning_rate": 9.35948746477912e-05, "loss": 0.0553, "step": 40950 }, { "epoch": 1.38476621927719, "grad_norm": 0.2417556792497635, "learning_rate": 9.359082562665266e-05, "loss": 0.0545, "step": 40960 }, { "epoch": 1.3851042969674432, "grad_norm": 0.36560872197151184, "learning_rate": 9.358677541375734e-05, "loss": 0.0564, "step": 40970 }, { "epoch": 1.3854423746576963, "grad_norm": 0.2899607717990875, "learning_rate": 9.358272400921596e-05, "loss": 0.0612, "step": 40980 }, { "epoch": 1.3857804523479496, "grad_norm": 0.2915334701538086, "learning_rate": 9.357867141313927e-05, "loss": 0.0597, "step": 40990 }, { "epoch": 1.3861185300382028, "grad_norm": 0.8530759215354919, "learning_rate": 9.357461762563807e-05, "loss": 0.0557, "step": 41000 }, { "epoch": 1.386456607728456, "grad_norm": 0.30811965465545654, "learning_rate": 9.35705626468232e-05, "loss": 0.0592, "step": 41010 }, { "epoch": 1.3867946854187092, "grad_norm": 0.3144611716270447, "learning_rate": 9.356650647680551e-05, "loss": 0.0545, "step": 41020 }, { "epoch": 1.3871327631089625, "grad_norm": 0.23284536600112915, "learning_rate": 9.356244911569591e-05, "loss": 0.0579, "step": 41030 }, { "epoch": 1.3874708407992156, "grad_norm": 0.5372380614280701, "learning_rate": 9.355839056360532e-05, "loss": 0.0611, "step": 41040 }, { "epoch": 1.3878089184894689, "grad_norm": 0.3777783215045929, "learning_rate": 9.355433082064469e-05, "loss": 0.0685, "step": 41050 }, { "epoch": 1.3881469961797221, "grad_norm": 0.39769113063812256, "learning_rate": 9.355026988692501e-05, "loss": 0.0536, "step": 41060 }, { "epoch": 1.3884850738699752, "grad_norm": 0.2853434383869171, "learning_rate": 9.354620776255731e-05, "loss": 0.0497, "step": 41070 }, { "epoch": 1.3888231515602285, "grad_norm": 0.33166414499282837, "learning_rate": 9.354214444765267e-05, "loss": 0.0621, "step": 41080 }, { "epoch": 1.3891612292504818, "grad_norm": 0.2321113497018814, "learning_rate": 9.353807994232214e-05, "loss": 0.063, "step": 41090 }, { "epoch": 1.3894993069407349, "grad_norm": 0.4228195548057556, "learning_rate": 9.353401424667687e-05, "loss": 0.0575, "step": 41100 }, { "epoch": 1.3898373846309882, "grad_norm": 0.28176045417785645, "learning_rate": 9.3529947360828e-05, "loss": 0.0606, "step": 41110 }, { "epoch": 1.3901754623212415, "grad_norm": 0.4826943278312683, "learning_rate": 9.352587928488673e-05, "loss": 0.0643, "step": 41120 }, { "epoch": 1.3905135400114945, "grad_norm": 0.32472941279411316, "learning_rate": 9.352181001896427e-05, "loss": 0.0571, "step": 41130 }, { "epoch": 1.3908516177017478, "grad_norm": 0.3416029214859009, "learning_rate": 9.351773956317187e-05, "loss": 0.0541, "step": 41140 }, { "epoch": 1.391189695392001, "grad_norm": 0.23656301200389862, "learning_rate": 9.351366791762081e-05, "loss": 0.0502, "step": 41150 }, { "epoch": 1.3915277730822542, "grad_norm": 0.440762996673584, "learning_rate": 9.350959508242244e-05, "loss": 0.0567, "step": 41160 }, { "epoch": 1.3918658507725075, "grad_norm": 0.22762104868888855, "learning_rate": 9.350552105768808e-05, "loss": 0.0561, "step": 41170 }, { "epoch": 1.3922039284627608, "grad_norm": 0.5317723751068115, "learning_rate": 9.350144584352911e-05, "loss": 0.0571, "step": 41180 }, { "epoch": 1.392542006153014, "grad_norm": 0.2642121911048889, "learning_rate": 9.349736944005696e-05, "loss": 0.0601, "step": 41190 }, { "epoch": 1.3928800838432671, "grad_norm": 0.27765703201293945, "learning_rate": 9.349329184738307e-05, "loss": 0.0644, "step": 41200 }, { "epoch": 1.3932181615335204, "grad_norm": 0.26462680101394653, "learning_rate": 9.348921306561891e-05, "loss": 0.0597, "step": 41210 }, { "epoch": 1.3935562392237737, "grad_norm": 0.24409331381320953, "learning_rate": 9.348513309487601e-05, "loss": 0.0573, "step": 41220 }, { "epoch": 1.3938943169140268, "grad_norm": 0.18320535123348236, "learning_rate": 9.348105193526591e-05, "loss": 0.0538, "step": 41230 }, { "epoch": 1.39423239460428, "grad_norm": 0.17277665436267853, "learning_rate": 9.347696958690019e-05, "loss": 0.0581, "step": 41240 }, { "epoch": 1.3945704722945333, "grad_norm": 0.2556116580963135, "learning_rate": 9.347288604989044e-05, "loss": 0.056, "step": 41250 }, { "epoch": 1.3949085499847866, "grad_norm": 0.3097032904624939, "learning_rate": 9.346880132434833e-05, "loss": 0.0685, "step": 41260 }, { "epoch": 1.3952466276750397, "grad_norm": 0.4279870092868805, "learning_rate": 9.346471541038553e-05, "loss": 0.0591, "step": 41270 }, { "epoch": 1.395584705365293, "grad_norm": 0.4080820679664612, "learning_rate": 9.346062830811372e-05, "loss": 0.0514, "step": 41280 }, { "epoch": 1.3959227830555463, "grad_norm": 0.2301596850156784, "learning_rate": 9.345654001764466e-05, "loss": 0.0575, "step": 41290 }, { "epoch": 1.3962608607457994, "grad_norm": 0.5035497546195984, "learning_rate": 9.345245053909012e-05, "loss": 0.0596, "step": 41300 }, { "epoch": 1.3965989384360527, "grad_norm": 0.48896536231040955, "learning_rate": 9.344835987256193e-05, "loss": 0.0592, "step": 41310 }, { "epoch": 1.396937016126306, "grad_norm": 0.5318601131439209, "learning_rate": 9.34442680181719e-05, "loss": 0.058, "step": 41320 }, { "epoch": 1.397275093816559, "grad_norm": 0.16763906180858612, "learning_rate": 9.344017497603189e-05, "loss": 0.0592, "step": 41330 }, { "epoch": 1.3976131715068123, "grad_norm": 0.22522865235805511, "learning_rate": 9.34360807462538e-05, "loss": 0.0547, "step": 41340 }, { "epoch": 1.3979512491970656, "grad_norm": 0.27378180623054504, "learning_rate": 9.343198532894961e-05, "loss": 0.0527, "step": 41350 }, { "epoch": 1.3982893268873187, "grad_norm": 0.2062675654888153, "learning_rate": 9.342788872423125e-05, "loss": 0.0586, "step": 41360 }, { "epoch": 1.398627404577572, "grad_norm": 0.4078320264816284, "learning_rate": 9.342379093221072e-05, "loss": 0.0626, "step": 41370 }, { "epoch": 1.3989654822678252, "grad_norm": 0.3031124174594879, "learning_rate": 9.341969195300006e-05, "loss": 0.059, "step": 41380 }, { "epoch": 1.3993035599580783, "grad_norm": 0.4338073432445526, "learning_rate": 9.341559178671135e-05, "loss": 0.0597, "step": 41390 }, { "epoch": 1.3996416376483316, "grad_norm": 0.3891245126724243, "learning_rate": 9.341149043345667e-05, "loss": 0.0562, "step": 41400 }, { "epoch": 1.399979715338585, "grad_norm": 0.38999131321907043, "learning_rate": 9.340738789334813e-05, "loss": 0.0584, "step": 41410 }, { "epoch": 1.400317793028838, "grad_norm": 0.2566107511520386, "learning_rate": 9.340328416649792e-05, "loss": 0.0572, "step": 41420 }, { "epoch": 1.4006558707190913, "grad_norm": 0.33974722027778625, "learning_rate": 9.339917925301824e-05, "loss": 0.0579, "step": 41430 }, { "epoch": 1.4009939484093445, "grad_norm": 0.23457200825214386, "learning_rate": 9.339507315302129e-05, "loss": 0.0568, "step": 41440 }, { "epoch": 1.4013320260995976, "grad_norm": 0.5830670595169067, "learning_rate": 9.339096586661934e-05, "loss": 0.0592, "step": 41450 }, { "epoch": 1.401670103789851, "grad_norm": 0.5004342794418335, "learning_rate": 9.338685739392469e-05, "loss": 0.0592, "step": 41460 }, { "epoch": 1.4020081814801042, "grad_norm": 0.27840667963027954, "learning_rate": 9.338274773504965e-05, "loss": 0.0588, "step": 41470 }, { "epoch": 1.4023462591703573, "grad_norm": 0.4442311227321625, "learning_rate": 9.33786368901066e-05, "loss": 0.0598, "step": 41480 }, { "epoch": 1.4026843368606106, "grad_norm": 0.26361945271492004, "learning_rate": 9.33745248592079e-05, "loss": 0.058, "step": 41490 }, { "epoch": 1.4030224145508639, "grad_norm": 0.3989883363246918, "learning_rate": 9.3370411642466e-05, "loss": 0.058, "step": 41500 }, { "epoch": 1.403360492241117, "grad_norm": 0.2603178918361664, "learning_rate": 9.336629723999333e-05, "loss": 0.061, "step": 41510 }, { "epoch": 1.4036985699313702, "grad_norm": 0.41155171394348145, "learning_rate": 9.336218165190239e-05, "loss": 0.0529, "step": 41520 }, { "epoch": 1.4040366476216235, "grad_norm": 0.38933682441711426, "learning_rate": 9.33580648783057e-05, "loss": 0.0581, "step": 41530 }, { "epoch": 1.4043747253118766, "grad_norm": 0.2833101153373718, "learning_rate": 9.33539469193158e-05, "loss": 0.0584, "step": 41540 }, { "epoch": 1.4047128030021299, "grad_norm": 0.2679831087589264, "learning_rate": 9.334982777504526e-05, "loss": 0.0584, "step": 41550 }, { "epoch": 1.4050508806923832, "grad_norm": 0.28091058135032654, "learning_rate": 9.334570744560673e-05, "loss": 0.0549, "step": 41560 }, { "epoch": 1.4053889583826362, "grad_norm": 0.28015974164009094, "learning_rate": 9.334158593111285e-05, "loss": 0.0592, "step": 41570 }, { "epoch": 1.4057270360728895, "grad_norm": 0.2791847586631775, "learning_rate": 9.333746323167629e-05, "loss": 0.0562, "step": 41580 }, { "epoch": 1.4060651137631428, "grad_norm": 0.30640581250190735, "learning_rate": 9.333333934740976e-05, "loss": 0.0559, "step": 41590 }, { "epoch": 1.4064031914533959, "grad_norm": 0.22004202008247375, "learning_rate": 9.3329214278426e-05, "loss": 0.0564, "step": 41600 }, { "epoch": 1.4067412691436492, "grad_norm": 0.387880802154541, "learning_rate": 9.332508802483782e-05, "loss": 0.0553, "step": 41610 }, { "epoch": 1.4070793468339025, "grad_norm": 0.42746809124946594, "learning_rate": 9.332096058675801e-05, "loss": 0.0566, "step": 41620 }, { "epoch": 1.4074174245241555, "grad_norm": 0.21551011502742767, "learning_rate": 9.33168319642994e-05, "loss": 0.0555, "step": 41630 }, { "epoch": 1.4077555022144088, "grad_norm": 0.23850655555725098, "learning_rate": 9.331270215757488e-05, "loss": 0.0544, "step": 41640 }, { "epoch": 1.408093579904662, "grad_norm": 0.38593217730522156, "learning_rate": 9.330857116669736e-05, "loss": 0.059, "step": 41650 }, { "epoch": 1.4084316575949152, "grad_norm": 0.33470091223716736, "learning_rate": 9.330443899177978e-05, "loss": 0.053, "step": 41660 }, { "epoch": 1.4087697352851685, "grad_norm": 0.22448480129241943, "learning_rate": 9.330030563293509e-05, "loss": 0.0548, "step": 41670 }, { "epoch": 1.4091078129754218, "grad_norm": 0.45519545674324036, "learning_rate": 9.329617109027633e-05, "loss": 0.0571, "step": 41680 }, { "epoch": 1.409445890665675, "grad_norm": 0.45112085342407227, "learning_rate": 9.329203536391651e-05, "loss": 0.0562, "step": 41690 }, { "epoch": 1.4097839683559281, "grad_norm": 0.3335437476634979, "learning_rate": 9.32878984539687e-05, "loss": 0.062, "step": 41700 }, { "epoch": 1.4101220460461814, "grad_norm": 0.2235681414604187, "learning_rate": 9.328376036054602e-05, "loss": 0.0559, "step": 41710 }, { "epoch": 1.4104601237364347, "grad_norm": 0.323276549577713, "learning_rate": 9.327962108376158e-05, "loss": 0.062, "step": 41720 }, { "epoch": 1.4107982014266878, "grad_norm": 0.3696430027484894, "learning_rate": 9.327548062372858e-05, "loss": 0.0604, "step": 41730 }, { "epoch": 1.411136279116941, "grad_norm": 0.18612805008888245, "learning_rate": 9.327133898056017e-05, "loss": 0.0607, "step": 41740 }, { "epoch": 1.4114743568071944, "grad_norm": 0.2610185146331787, "learning_rate": 9.326719615436963e-05, "loss": 0.0554, "step": 41750 }, { "epoch": 1.4118124344974476, "grad_norm": 0.2052338570356369, "learning_rate": 9.326305214527019e-05, "loss": 0.0572, "step": 41760 }, { "epoch": 1.4121505121877007, "grad_norm": 0.27669796347618103, "learning_rate": 9.325890695337516e-05, "loss": 0.0577, "step": 41770 }, { "epoch": 1.412488589877954, "grad_norm": 0.4122221767902374, "learning_rate": 9.325476057879787e-05, "loss": 0.0651, "step": 41780 }, { "epoch": 1.4128266675682073, "grad_norm": 0.517970085144043, "learning_rate": 9.325061302165167e-05, "loss": 0.056, "step": 41790 }, { "epoch": 1.4131647452584604, "grad_norm": 0.43091514706611633, "learning_rate": 9.324646428204997e-05, "loss": 0.0518, "step": 41800 }, { "epoch": 1.4135028229487137, "grad_norm": 0.29956626892089844, "learning_rate": 9.324231436010617e-05, "loss": 0.057, "step": 41810 }, { "epoch": 1.413840900638967, "grad_norm": 0.3045814037322998, "learning_rate": 9.323816325593375e-05, "loss": 0.0572, "step": 41820 }, { "epoch": 1.41417897832922, "grad_norm": 0.41872867941856384, "learning_rate": 9.323401096964618e-05, "loss": 0.06, "step": 41830 }, { "epoch": 1.4145170560194733, "grad_norm": 0.2901758849620819, "learning_rate": 9.3229857501357e-05, "loss": 0.0564, "step": 41840 }, { "epoch": 1.4148551337097266, "grad_norm": 0.35846641659736633, "learning_rate": 9.322570285117975e-05, "loss": 0.0599, "step": 41850 }, { "epoch": 1.4151932113999797, "grad_norm": 0.34106364846229553, "learning_rate": 9.322154701922803e-05, "loss": 0.0568, "step": 41860 }, { "epoch": 1.415531289090233, "grad_norm": 0.4083669185638428, "learning_rate": 9.321739000561545e-05, "loss": 0.0571, "step": 41870 }, { "epoch": 1.4158693667804862, "grad_norm": 0.48694711923599243, "learning_rate": 9.321323181045567e-05, "loss": 0.059, "step": 41880 }, { "epoch": 1.4162074444707393, "grad_norm": 0.24179323017597198, "learning_rate": 9.320907243386237e-05, "loss": 0.0635, "step": 41890 }, { "epoch": 1.4165455221609926, "grad_norm": 0.44656574726104736, "learning_rate": 9.320491187594925e-05, "loss": 0.0633, "step": 41900 }, { "epoch": 1.416883599851246, "grad_norm": 0.5779715776443481, "learning_rate": 9.320075013683007e-05, "loss": 0.0609, "step": 41910 }, { "epoch": 1.417221677541499, "grad_norm": 0.33756911754608154, "learning_rate": 9.319658721661861e-05, "loss": 0.0582, "step": 41920 }, { "epoch": 1.4175597552317523, "grad_norm": 0.3795984387397766, "learning_rate": 9.319242311542868e-05, "loss": 0.0531, "step": 41930 }, { "epoch": 1.4178978329220056, "grad_norm": 0.2560952305793762, "learning_rate": 9.318825783337415e-05, "loss": 0.0597, "step": 41940 }, { "epoch": 1.4182359106122586, "grad_norm": 0.5290001630783081, "learning_rate": 9.318409137056886e-05, "loss": 0.0582, "step": 41950 }, { "epoch": 1.418573988302512, "grad_norm": 0.332684725522995, "learning_rate": 9.317992372712674e-05, "loss": 0.0632, "step": 41960 }, { "epoch": 1.4189120659927652, "grad_norm": 0.26692232489585876, "learning_rate": 9.317575490316173e-05, "loss": 0.0584, "step": 41970 }, { "epoch": 1.4192501436830183, "grad_norm": 0.43408071994781494, "learning_rate": 9.317158489878779e-05, "loss": 0.0558, "step": 41980 }, { "epoch": 1.4195882213732716, "grad_norm": 0.23705363273620605, "learning_rate": 9.316741371411895e-05, "loss": 0.0602, "step": 41990 }, { "epoch": 1.4199262990635249, "grad_norm": 0.28001222014427185, "learning_rate": 9.316324134926923e-05, "loss": 0.0626, "step": 42000 }, { "epoch": 1.420264376753778, "grad_norm": 0.37647244334220886, "learning_rate": 9.31590678043527e-05, "loss": 0.0566, "step": 42010 }, { "epoch": 1.4206024544440312, "grad_norm": 0.28821444511413574, "learning_rate": 9.315489307948349e-05, "loss": 0.0553, "step": 42020 }, { "epoch": 1.4209405321342845, "grad_norm": 0.3135441839694977, "learning_rate": 9.31507171747757e-05, "loss": 0.0611, "step": 42030 }, { "epoch": 1.4212786098245376, "grad_norm": 0.4265105128288269, "learning_rate": 9.314654009034352e-05, "loss": 0.0495, "step": 42040 }, { "epoch": 1.4216166875147909, "grad_norm": 0.32113561034202576, "learning_rate": 9.314236182630114e-05, "loss": 0.0527, "step": 42050 }, { "epoch": 1.4219547652050442, "grad_norm": 0.4198494553565979, "learning_rate": 9.31381823827628e-05, "loss": 0.0585, "step": 42060 }, { "epoch": 1.4222928428952972, "grad_norm": 0.4609922468662262, "learning_rate": 9.313400175984275e-05, "loss": 0.0664, "step": 42070 }, { "epoch": 1.4226309205855505, "grad_norm": 0.32903483510017395, "learning_rate": 9.31298199576553e-05, "loss": 0.06, "step": 42080 }, { "epoch": 1.4229689982758038, "grad_norm": 0.6411097645759583, "learning_rate": 9.312563697631478e-05, "loss": 0.0653, "step": 42090 }, { "epoch": 1.4233070759660569, "grad_norm": 0.2967287302017212, "learning_rate": 9.312145281593554e-05, "loss": 0.0597, "step": 42100 }, { "epoch": 1.4236451536563102, "grad_norm": 0.28452831506729126, "learning_rate": 9.311726747663199e-05, "loss": 0.0544, "step": 42110 }, { "epoch": 1.4239832313465635, "grad_norm": 0.330131471157074, "learning_rate": 9.311308095851854e-05, "loss": 0.0552, "step": 42120 }, { "epoch": 1.4243213090368165, "grad_norm": 0.23987729847431183, "learning_rate": 9.310889326170967e-05, "loss": 0.0566, "step": 42130 }, { "epoch": 1.4246593867270698, "grad_norm": 0.32831183075904846, "learning_rate": 9.310470438631982e-05, "loss": 0.0603, "step": 42140 }, { "epoch": 1.4249974644173231, "grad_norm": 0.19908872246742249, "learning_rate": 9.310051433246357e-05, "loss": 0.061, "step": 42150 }, { "epoch": 1.4253355421075764, "grad_norm": 0.28302767872810364, "learning_rate": 9.309632310025544e-05, "loss": 0.0594, "step": 42160 }, { "epoch": 1.4256736197978295, "grad_norm": 0.2702932357788086, "learning_rate": 9.309213068981003e-05, "loss": 0.0528, "step": 42170 }, { "epoch": 1.4260116974880828, "grad_norm": 0.33200594782829285, "learning_rate": 9.308793710124195e-05, "loss": 0.065, "step": 42180 }, { "epoch": 1.426349775178336, "grad_norm": 0.30862778425216675, "learning_rate": 9.308374233466587e-05, "loss": 0.0562, "step": 42190 }, { "epoch": 1.4266878528685891, "grad_norm": 0.438567191362381, "learning_rate": 9.307954639019644e-05, "loss": 0.0593, "step": 42200 }, { "epoch": 1.4270259305588424, "grad_norm": 0.35736197233200073, "learning_rate": 9.307534926794842e-05, "loss": 0.0562, "step": 42210 }, { "epoch": 1.4273640082490957, "grad_norm": 0.6929407715797424, "learning_rate": 9.307115096803654e-05, "loss": 0.0607, "step": 42220 }, { "epoch": 1.427702085939349, "grad_norm": 0.43407344818115234, "learning_rate": 9.306695149057554e-05, "loss": 0.0538, "step": 42230 }, { "epoch": 1.428040163629602, "grad_norm": 0.3340182900428772, "learning_rate": 9.306275083568029e-05, "loss": 0.0582, "step": 42240 }, { "epoch": 1.4283782413198554, "grad_norm": 0.3708557188510895, "learning_rate": 9.30585490034656e-05, "loss": 0.0519, "step": 42250 }, { "epoch": 1.4287163190101086, "grad_norm": 0.4275110065937042, "learning_rate": 9.305434599404637e-05, "loss": 0.0596, "step": 42260 }, { "epoch": 1.4290543967003617, "grad_norm": 0.23463109135627747, "learning_rate": 9.305014180753749e-05, "loss": 0.0581, "step": 42270 }, { "epoch": 1.429392474390615, "grad_norm": 0.31576964259147644, "learning_rate": 9.304593644405388e-05, "loss": 0.0578, "step": 42280 }, { "epoch": 1.4297305520808683, "grad_norm": 0.35014426708221436, "learning_rate": 9.304172990371056e-05, "loss": 0.0617, "step": 42290 }, { "epoch": 1.4300686297711214, "grad_norm": 0.3862515687942505, "learning_rate": 9.303752218662251e-05, "loss": 0.0555, "step": 42300 }, { "epoch": 1.4304067074613747, "grad_norm": 0.33770519495010376, "learning_rate": 9.303331329290479e-05, "loss": 0.0552, "step": 42310 }, { "epoch": 1.430744785151628, "grad_norm": 0.4288913607597351, "learning_rate": 9.302910322267244e-05, "loss": 0.0572, "step": 42320 }, { "epoch": 1.431082862841881, "grad_norm": 0.40077170729637146, "learning_rate": 9.302489197604056e-05, "loss": 0.0619, "step": 42330 }, { "epoch": 1.4314209405321343, "grad_norm": 0.31307390332221985, "learning_rate": 9.302067955312432e-05, "loss": 0.0519, "step": 42340 }, { "epoch": 1.4317590182223876, "grad_norm": 0.5369732975959778, "learning_rate": 9.301646595403884e-05, "loss": 0.0619, "step": 42350 }, { "epoch": 1.4320970959126407, "grad_norm": 0.31376200914382935, "learning_rate": 9.301225117889936e-05, "loss": 0.0594, "step": 42360 }, { "epoch": 1.432435173602894, "grad_norm": 0.31710126996040344, "learning_rate": 9.300803522782108e-05, "loss": 0.0592, "step": 42370 }, { "epoch": 1.4327732512931473, "grad_norm": 0.32741624116897583, "learning_rate": 9.300381810091929e-05, "loss": 0.0605, "step": 42380 }, { "epoch": 1.4331113289834003, "grad_norm": 0.3326476812362671, "learning_rate": 9.299959979830925e-05, "loss": 0.0592, "step": 42390 }, { "epoch": 1.4334494066736536, "grad_norm": 0.24599449336528778, "learning_rate": 9.299538032010631e-05, "loss": 0.0604, "step": 42400 }, { "epoch": 1.433787484363907, "grad_norm": 0.2255924642086029, "learning_rate": 9.299115966642584e-05, "loss": 0.0621, "step": 42410 }, { "epoch": 1.43412556205416, "grad_norm": 0.2506251633167267, "learning_rate": 9.298693783738319e-05, "loss": 0.0588, "step": 42420 }, { "epoch": 1.4344636397444133, "grad_norm": 0.2407923936843872, "learning_rate": 9.298271483309383e-05, "loss": 0.063, "step": 42430 }, { "epoch": 1.4348017174346666, "grad_norm": 0.5521117448806763, "learning_rate": 9.297849065367318e-05, "loss": 0.0592, "step": 42440 }, { "epoch": 1.4351397951249196, "grad_norm": 0.3471984565258026, "learning_rate": 9.297426529923676e-05, "loss": 0.0531, "step": 42450 }, { "epoch": 1.435477872815173, "grad_norm": 0.3368596136569977, "learning_rate": 9.297003876990006e-05, "loss": 0.0542, "step": 42460 }, { "epoch": 1.4358159505054262, "grad_norm": 0.37904489040374756, "learning_rate": 9.296581106577864e-05, "loss": 0.0569, "step": 42470 }, { "epoch": 1.4361540281956793, "grad_norm": 0.3088332712650299, "learning_rate": 9.296158218698809e-05, "loss": 0.0568, "step": 42480 }, { "epoch": 1.4364921058859326, "grad_norm": 0.34688708186149597, "learning_rate": 9.295735213364403e-05, "loss": 0.0548, "step": 42490 }, { "epoch": 1.4368301835761859, "grad_norm": 0.4680502712726593, "learning_rate": 9.295312090586208e-05, "loss": 0.0616, "step": 42500 }, { "epoch": 1.437168261266439, "grad_norm": 0.2509571611881256, "learning_rate": 9.294888850375796e-05, "loss": 0.0558, "step": 42510 }, { "epoch": 1.4375063389566922, "grad_norm": 0.5717397332191467, "learning_rate": 9.294465492744737e-05, "loss": 0.0577, "step": 42520 }, { "epoch": 1.4378444166469455, "grad_norm": 0.4385621249675751, "learning_rate": 9.294042017704603e-05, "loss": 0.0663, "step": 42530 }, { "epoch": 1.4381824943371986, "grad_norm": 0.5482307076454163, "learning_rate": 9.293618425266974e-05, "loss": 0.0638, "step": 42540 }, { "epoch": 1.4385205720274519, "grad_norm": 0.4451143741607666, "learning_rate": 9.293194715443429e-05, "loss": 0.0629, "step": 42550 }, { "epoch": 1.4388586497177052, "grad_norm": 0.23434455692768097, "learning_rate": 9.292770888245557e-05, "loss": 0.0649, "step": 42560 }, { "epoch": 1.4391967274079582, "grad_norm": 0.4022148847579956, "learning_rate": 9.292346943684937e-05, "loss": 0.0585, "step": 42570 }, { "epoch": 1.4395348050982115, "grad_norm": 0.33464834094047546, "learning_rate": 9.291922881773167e-05, "loss": 0.0575, "step": 42580 }, { "epoch": 1.4398728827884648, "grad_norm": 0.3480454981327057, "learning_rate": 9.291498702521838e-05, "loss": 0.0695, "step": 42590 }, { "epoch": 1.4402109604787179, "grad_norm": 0.4497571587562561, "learning_rate": 9.291074405942546e-05, "loss": 0.0606, "step": 42600 }, { "epoch": 1.4405490381689712, "grad_norm": 0.28137028217315674, "learning_rate": 9.290649992046892e-05, "loss": 0.0598, "step": 42610 }, { "epoch": 1.4408871158592245, "grad_norm": 0.6475501656532288, "learning_rate": 9.29022546084648e-05, "loss": 0.0571, "step": 42620 }, { "epoch": 1.4412251935494775, "grad_norm": 0.3367249071598053, "learning_rate": 9.289800812352913e-05, "loss": 0.0572, "step": 42630 }, { "epoch": 1.4415632712397308, "grad_norm": 0.3189000189304352, "learning_rate": 9.289376046577806e-05, "loss": 0.0622, "step": 42640 }, { "epoch": 1.4419013489299841, "grad_norm": 0.2791796326637268, "learning_rate": 9.288951163532769e-05, "loss": 0.0633, "step": 42650 }, { "epoch": 1.4422394266202374, "grad_norm": 0.2685999572277069, "learning_rate": 9.288526163229416e-05, "loss": 0.0637, "step": 42660 }, { "epoch": 1.4425775043104905, "grad_norm": 0.260123074054718, "learning_rate": 9.288101045679371e-05, "loss": 0.0549, "step": 42670 }, { "epoch": 1.4429155820007438, "grad_norm": 0.3473772406578064, "learning_rate": 9.287675810894255e-05, "loss": 0.0569, "step": 42680 }, { "epoch": 1.443253659690997, "grad_norm": 0.22688063979148865, "learning_rate": 9.287250458885692e-05, "loss": 0.063, "step": 42690 }, { "epoch": 1.4435917373812501, "grad_norm": 0.3380509912967682, "learning_rate": 9.286824989665313e-05, "loss": 0.0567, "step": 42700 }, { "epoch": 1.4439298150715034, "grad_norm": 0.2913333475589752, "learning_rate": 9.286399403244748e-05, "loss": 0.0625, "step": 42710 }, { "epoch": 1.4442678927617567, "grad_norm": 0.290663480758667, "learning_rate": 9.285973699635634e-05, "loss": 0.0612, "step": 42720 }, { "epoch": 1.44460597045201, "grad_norm": 0.2422359436750412, "learning_rate": 9.285547878849609e-05, "loss": 0.0589, "step": 42730 }, { "epoch": 1.444944048142263, "grad_norm": 0.6117493510246277, "learning_rate": 9.285121940898317e-05, "loss": 0.0542, "step": 42740 }, { "epoch": 1.4452821258325164, "grad_norm": 0.8643693923950195, "learning_rate": 9.284695885793398e-05, "loss": 0.0583, "step": 42750 }, { "epoch": 1.4456202035227697, "grad_norm": 0.7207810878753662, "learning_rate": 9.284269713546505e-05, "loss": 0.0618, "step": 42760 }, { "epoch": 1.4459582812130227, "grad_norm": 0.2780471742153168, "learning_rate": 9.283843424169288e-05, "loss": 0.0613, "step": 42770 }, { "epoch": 1.446296358903276, "grad_norm": 0.30113884806632996, "learning_rate": 9.283417017673399e-05, "loss": 0.053, "step": 42780 }, { "epoch": 1.4466344365935293, "grad_norm": 0.33488085865974426, "learning_rate": 9.282990494070499e-05, "loss": 0.0571, "step": 42790 }, { "epoch": 1.4469725142837824, "grad_norm": 0.5097066164016724, "learning_rate": 9.282563853372249e-05, "loss": 0.0576, "step": 42800 }, { "epoch": 1.4473105919740357, "grad_norm": 0.22010737657546997, "learning_rate": 9.28213709559031e-05, "loss": 0.0623, "step": 42810 }, { "epoch": 1.447648669664289, "grad_norm": 0.2834562063217163, "learning_rate": 9.281710220736353e-05, "loss": 0.0606, "step": 42820 }, { "epoch": 1.447986747354542, "grad_norm": 0.2858451008796692, "learning_rate": 9.281283228822047e-05, "loss": 0.0592, "step": 42830 }, { "epoch": 1.4483248250447953, "grad_norm": 0.5122247338294983, "learning_rate": 9.280856119859067e-05, "loss": 0.0561, "step": 42840 }, { "epoch": 1.4486629027350486, "grad_norm": 0.35608306527137756, "learning_rate": 9.280428893859088e-05, "loss": 0.0595, "step": 42850 }, { "epoch": 1.4490009804253017, "grad_norm": 0.2696366310119629, "learning_rate": 9.28000155083379e-05, "loss": 0.0574, "step": 42860 }, { "epoch": 1.449339058115555, "grad_norm": 0.3266957104206085, "learning_rate": 9.279574090794858e-05, "loss": 0.0585, "step": 42870 }, { "epoch": 1.4496771358058083, "grad_norm": 0.27938130497932434, "learning_rate": 9.279146513753979e-05, "loss": 0.0565, "step": 42880 }, { "epoch": 1.4500152134960613, "grad_norm": 0.2646132707595825, "learning_rate": 9.278718819722841e-05, "loss": 0.0569, "step": 42890 }, { "epoch": 1.4503532911863146, "grad_norm": 0.24656695127487183, "learning_rate": 9.27829100871314e-05, "loss": 0.0544, "step": 42900 }, { "epoch": 1.450691368876568, "grad_norm": 0.33779996633529663, "learning_rate": 9.277863080736569e-05, "loss": 0.0549, "step": 42910 }, { "epoch": 1.451029446566821, "grad_norm": 0.3929242789745331, "learning_rate": 9.277435035804827e-05, "loss": 0.0626, "step": 42920 }, { "epoch": 1.4513675242570743, "grad_norm": 0.2816644012928009, "learning_rate": 9.277006873929619e-05, "loss": 0.0602, "step": 42930 }, { "epoch": 1.4517056019473276, "grad_norm": 0.38788706064224243, "learning_rate": 9.276578595122649e-05, "loss": 0.0561, "step": 42940 }, { "epoch": 1.4520436796375806, "grad_norm": 0.2804467976093292, "learning_rate": 9.276150199395628e-05, "loss": 0.0555, "step": 42950 }, { "epoch": 1.452381757327834, "grad_norm": 0.3183920681476593, "learning_rate": 9.275721686760267e-05, "loss": 0.058, "step": 42960 }, { "epoch": 1.4527198350180872, "grad_norm": 0.2260732799768448, "learning_rate": 9.275293057228281e-05, "loss": 0.0551, "step": 42970 }, { "epoch": 1.4530579127083403, "grad_norm": 0.35382822155952454, "learning_rate": 9.274864310811388e-05, "loss": 0.0549, "step": 42980 }, { "epoch": 1.4533959903985936, "grad_norm": 0.42027971148490906, "learning_rate": 9.274435447521313e-05, "loss": 0.0576, "step": 42990 }, { "epoch": 1.4537340680888469, "grad_norm": 0.2681684195995331, "learning_rate": 9.274006467369778e-05, "loss": 0.0556, "step": 43000 }, { "epoch": 1.4540721457791, "grad_norm": 0.42466625571250916, "learning_rate": 9.273577370368512e-05, "loss": 0.0546, "step": 43010 }, { "epoch": 1.4544102234693532, "grad_norm": 0.28080257773399353, "learning_rate": 9.273148156529246e-05, "loss": 0.0584, "step": 43020 }, { "epoch": 1.4547483011596065, "grad_norm": 0.2518968880176544, "learning_rate": 9.272718825863714e-05, "loss": 0.0523, "step": 43030 }, { "epoch": 1.4550863788498596, "grad_norm": 1.0029106140136719, "learning_rate": 9.272289378383654e-05, "loss": 0.0588, "step": 43040 }, { "epoch": 1.4554244565401129, "grad_norm": 0.3176569640636444, "learning_rate": 9.271859814100808e-05, "loss": 0.0559, "step": 43050 }, { "epoch": 1.4557625342303662, "grad_norm": 0.8395082950592041, "learning_rate": 9.27143013302692e-05, "loss": 0.0522, "step": 43060 }, { "epoch": 1.4561006119206192, "grad_norm": 0.26753637194633484, "learning_rate": 9.271000335173736e-05, "loss": 0.0617, "step": 43070 }, { "epoch": 1.4564386896108725, "grad_norm": 0.3474726974964142, "learning_rate": 9.270570420553008e-05, "loss": 0.0591, "step": 43080 }, { "epoch": 1.4567767673011258, "grad_norm": 0.32190579175949097, "learning_rate": 9.27014038917649e-05, "loss": 0.0589, "step": 43090 }, { "epoch": 1.4571148449913789, "grad_norm": 0.31261757016181946, "learning_rate": 9.269710241055937e-05, "loss": 0.0598, "step": 43100 }, { "epoch": 1.4574529226816322, "grad_norm": 0.31318002939224243, "learning_rate": 9.26927997620311e-05, "loss": 0.0523, "step": 43110 }, { "epoch": 1.4577910003718855, "grad_norm": 0.6569692492485046, "learning_rate": 9.268849594629773e-05, "loss": 0.0607, "step": 43120 }, { "epoch": 1.4581290780621388, "grad_norm": 0.32359516620635986, "learning_rate": 9.268419096347692e-05, "loss": 0.0574, "step": 43130 }, { "epoch": 1.4584671557523918, "grad_norm": 0.590825617313385, "learning_rate": 9.267988481368635e-05, "loss": 0.0574, "step": 43140 }, { "epoch": 1.4588052334426451, "grad_norm": 0.3161502778530121, "learning_rate": 9.267557749704377e-05, "loss": 0.0623, "step": 43150 }, { "epoch": 1.4591433111328984, "grad_norm": 0.655062735080719, "learning_rate": 9.267126901366694e-05, "loss": 0.0579, "step": 43160 }, { "epoch": 1.4594813888231515, "grad_norm": 0.21417362987995148, "learning_rate": 9.266695936367365e-05, "loss": 0.0573, "step": 43170 }, { "epoch": 1.4598194665134048, "grad_norm": 1.0099385976791382, "learning_rate": 9.26626485471817e-05, "loss": 0.0568, "step": 43180 }, { "epoch": 1.460157544203658, "grad_norm": 0.34694281220436096, "learning_rate": 9.265833656430898e-05, "loss": 0.0571, "step": 43190 }, { "epoch": 1.4604956218939114, "grad_norm": 0.31992024183273315, "learning_rate": 9.265402341517336e-05, "loss": 0.0581, "step": 43200 }, { "epoch": 1.4608336995841644, "grad_norm": 0.2777320444583893, "learning_rate": 9.264970909989276e-05, "loss": 0.0535, "step": 43210 }, { "epoch": 1.4611717772744177, "grad_norm": 0.26013460755348206, "learning_rate": 9.264539361858515e-05, "loss": 0.0564, "step": 43220 }, { "epoch": 1.461509854964671, "grad_norm": 0.2136276811361313, "learning_rate": 9.26410769713685e-05, "loss": 0.0581, "step": 43230 }, { "epoch": 1.461847932654924, "grad_norm": 0.28452616930007935, "learning_rate": 9.263675915836081e-05, "loss": 0.0557, "step": 43240 }, { "epoch": 1.4621860103451774, "grad_norm": 0.21206048130989075, "learning_rate": 9.263244017968014e-05, "loss": 0.0607, "step": 43250 }, { "epoch": 1.4625240880354307, "grad_norm": 0.3355487585067749, "learning_rate": 9.262812003544459e-05, "loss": 0.0594, "step": 43260 }, { "epoch": 1.4628621657256837, "grad_norm": 0.31886574625968933, "learning_rate": 9.262379872577225e-05, "loss": 0.0631, "step": 43270 }, { "epoch": 1.463200243415937, "grad_norm": 0.2794983386993408, "learning_rate": 9.261947625078126e-05, "loss": 0.0551, "step": 43280 }, { "epoch": 1.4635383211061903, "grad_norm": 0.316980242729187, "learning_rate": 9.26151526105898e-05, "loss": 0.0558, "step": 43290 }, { "epoch": 1.4638763987964434, "grad_norm": 0.3835516571998596, "learning_rate": 9.261082780531607e-05, "loss": 0.0553, "step": 43300 }, { "epoch": 1.4642144764866967, "grad_norm": 0.3713396191596985, "learning_rate": 9.260650183507834e-05, "loss": 0.0593, "step": 43310 }, { "epoch": 1.46455255417695, "grad_norm": 0.3862476646900177, "learning_rate": 9.260217469999483e-05, "loss": 0.058, "step": 43320 }, { "epoch": 1.464890631867203, "grad_norm": 0.3201952576637268, "learning_rate": 9.259784640018388e-05, "loss": 0.0659, "step": 43330 }, { "epoch": 1.4652287095574563, "grad_norm": 0.5318747758865356, "learning_rate": 9.25935169357638e-05, "loss": 0.067, "step": 43340 }, { "epoch": 1.4655667872477096, "grad_norm": 0.29812946915626526, "learning_rate": 9.258918630685299e-05, "loss": 0.0646, "step": 43350 }, { "epoch": 1.4659048649379627, "grad_norm": 0.27780526876449585, "learning_rate": 9.258485451356981e-05, "loss": 0.0602, "step": 43360 }, { "epoch": 1.466242942628216, "grad_norm": 0.24731166660785675, "learning_rate": 9.258052155603272e-05, "loss": 0.059, "step": 43370 }, { "epoch": 1.4665810203184693, "grad_norm": 0.36007171869277954, "learning_rate": 9.257618743436016e-05, "loss": 0.0532, "step": 43380 }, { "epoch": 1.4669190980087223, "grad_norm": 0.4210556745529175, "learning_rate": 9.257185214867063e-05, "loss": 0.0607, "step": 43390 }, { "epoch": 1.4672571756989756, "grad_norm": 0.31081634759902954, "learning_rate": 9.256751569908266e-05, "loss": 0.0571, "step": 43400 }, { "epoch": 1.467595253389229, "grad_norm": 0.33555343747138977, "learning_rate": 9.25631780857148e-05, "loss": 0.0587, "step": 43410 }, { "epoch": 1.467933331079482, "grad_norm": 0.242055281996727, "learning_rate": 9.255883930868563e-05, "loss": 0.0606, "step": 43420 }, { "epoch": 1.4682714087697353, "grad_norm": 0.2841531038284302, "learning_rate": 9.25544993681138e-05, "loss": 0.0581, "step": 43430 }, { "epoch": 1.4686094864599886, "grad_norm": 0.42823469638824463, "learning_rate": 9.255015826411794e-05, "loss": 0.0585, "step": 43440 }, { "epoch": 1.4689475641502416, "grad_norm": 0.4251706004142761, "learning_rate": 9.254581599681674e-05, "loss": 0.0617, "step": 43450 }, { "epoch": 1.469285641840495, "grad_norm": 0.28981101512908936, "learning_rate": 9.25414725663289e-05, "loss": 0.0509, "step": 43460 }, { "epoch": 1.4696237195307482, "grad_norm": 0.30820316076278687, "learning_rate": 9.253712797277319e-05, "loss": 0.0572, "step": 43470 }, { "epoch": 1.4699617972210013, "grad_norm": 0.2430974245071411, "learning_rate": 9.253278221626839e-05, "loss": 0.0567, "step": 43480 }, { "epoch": 1.4702998749112546, "grad_norm": 0.3515567481517792, "learning_rate": 9.25284352969333e-05, "loss": 0.0668, "step": 43490 }, { "epoch": 1.4706379526015079, "grad_norm": 0.5211124420166016, "learning_rate": 9.252408721488676e-05, "loss": 0.0591, "step": 43500 }, { "epoch": 1.470976030291761, "grad_norm": 0.6566750407218933, "learning_rate": 9.251973797024764e-05, "loss": 0.0542, "step": 43510 }, { "epoch": 1.4713141079820142, "grad_norm": 0.4069455564022064, "learning_rate": 9.251538756313488e-05, "loss": 0.0556, "step": 43520 }, { "epoch": 1.4716521856722675, "grad_norm": 0.16497719287872314, "learning_rate": 9.251103599366739e-05, "loss": 0.0554, "step": 43530 }, { "epoch": 1.4719902633625206, "grad_norm": 0.30831173062324524, "learning_rate": 9.250668326196415e-05, "loss": 0.0527, "step": 43540 }, { "epoch": 1.4723283410527739, "grad_norm": 0.23256143927574158, "learning_rate": 9.250232936814413e-05, "loss": 0.063, "step": 43550 }, { "epoch": 1.4726664187430272, "grad_norm": 0.39081257581710815, "learning_rate": 9.249797431232643e-05, "loss": 0.0599, "step": 43560 }, { "epoch": 1.4730044964332802, "grad_norm": 0.2755786180496216, "learning_rate": 9.249361809463007e-05, "loss": 0.0576, "step": 43570 }, { "epoch": 1.4733425741235335, "grad_norm": 0.313089519739151, "learning_rate": 9.248926071517414e-05, "loss": 0.053, "step": 43580 }, { "epoch": 1.4736806518137868, "grad_norm": 0.3968653380870819, "learning_rate": 9.24849021740778e-05, "loss": 0.0569, "step": 43590 }, { "epoch": 1.47401872950404, "grad_norm": 0.30220943689346313, "learning_rate": 9.248054247146019e-05, "loss": 0.0556, "step": 43600 }, { "epoch": 1.4743568071942932, "grad_norm": 0.7358960509300232, "learning_rate": 9.24761816074405e-05, "loss": 0.0631, "step": 43610 }, { "epoch": 1.4746948848845465, "grad_norm": 0.211166113615036, "learning_rate": 9.247181958213797e-05, "loss": 0.0579, "step": 43620 }, { "epoch": 1.4750329625747998, "grad_norm": 0.5254862904548645, "learning_rate": 9.246745639567185e-05, "loss": 0.0633, "step": 43630 }, { "epoch": 1.4753710402650528, "grad_norm": 0.34729301929473877, "learning_rate": 9.246309204816142e-05, "loss": 0.0538, "step": 43640 }, { "epoch": 1.4757091179553061, "grad_norm": 0.2621476948261261, "learning_rate": 9.245872653972602e-05, "loss": 0.0591, "step": 43650 }, { "epoch": 1.4760471956455594, "grad_norm": 0.2729272246360779, "learning_rate": 9.245435987048498e-05, "loss": 0.0557, "step": 43660 }, { "epoch": 1.4763852733358125, "grad_norm": 0.22601917386054993, "learning_rate": 9.244999204055769e-05, "loss": 0.0559, "step": 43670 }, { "epoch": 1.4767233510260658, "grad_norm": 0.3887157440185547, "learning_rate": 9.244562305006354e-05, "loss": 0.0543, "step": 43680 }, { "epoch": 1.477061428716319, "grad_norm": 0.2718469202518463, "learning_rate": 9.244125289912205e-05, "loss": 0.0536, "step": 43690 }, { "epoch": 1.4773995064065724, "grad_norm": 0.29856404662132263, "learning_rate": 9.243688158785263e-05, "loss": 0.0612, "step": 43700 }, { "epoch": 1.4777375840968254, "grad_norm": 0.38740676641464233, "learning_rate": 9.243250911637479e-05, "loss": 0.0525, "step": 43710 }, { "epoch": 1.4780756617870787, "grad_norm": 0.36175939440727234, "learning_rate": 9.242813548480811e-05, "loss": 0.0503, "step": 43720 }, { "epoch": 1.478413739477332, "grad_norm": 0.31379440426826477, "learning_rate": 9.242376069327215e-05, "loss": 0.0521, "step": 43730 }, { "epoch": 1.478751817167585, "grad_norm": 0.2949581444263458, "learning_rate": 9.24193847418865e-05, "loss": 0.0518, "step": 43740 }, { "epoch": 1.4790898948578384, "grad_norm": 0.413610577583313, "learning_rate": 9.24150076307708e-05, "loss": 0.0602, "step": 43750 }, { "epoch": 1.4794279725480917, "grad_norm": 0.2558327317237854, "learning_rate": 9.241062936004474e-05, "loss": 0.0539, "step": 43760 }, { "epoch": 1.4797660502383447, "grad_norm": 0.25042879581451416, "learning_rate": 9.240624992982801e-05, "loss": 0.0528, "step": 43770 }, { "epoch": 1.480104127928598, "grad_norm": 0.32038557529449463, "learning_rate": 9.240186934024034e-05, "loss": 0.0573, "step": 43780 }, { "epoch": 1.4804422056188513, "grad_norm": 0.5835912823677063, "learning_rate": 9.239748759140148e-05, "loss": 0.0607, "step": 43790 }, { "epoch": 1.4807802833091044, "grad_norm": 0.415348619222641, "learning_rate": 9.239310468343123e-05, "loss": 0.0554, "step": 43800 }, { "epoch": 1.4811183609993577, "grad_norm": 0.33575138449668884, "learning_rate": 9.238872061644943e-05, "loss": 0.0576, "step": 43810 }, { "epoch": 1.481456438689611, "grad_norm": 0.3573479950428009, "learning_rate": 9.238433539057594e-05, "loss": 0.0504, "step": 43820 }, { "epoch": 1.481794516379864, "grad_norm": 0.30492648482322693, "learning_rate": 9.237994900593065e-05, "loss": 0.0563, "step": 43830 }, { "epoch": 1.4821325940701173, "grad_norm": 0.2751052677631378, "learning_rate": 9.237556146263346e-05, "loss": 0.0535, "step": 43840 }, { "epoch": 1.4824706717603706, "grad_norm": 0.27049803733825684, "learning_rate": 9.237117276080435e-05, "loss": 0.0569, "step": 43850 }, { "epoch": 1.4828087494506237, "grad_norm": 0.32936882972717285, "learning_rate": 9.236678290056328e-05, "loss": 0.0589, "step": 43860 }, { "epoch": 1.483146827140877, "grad_norm": 0.35767877101898193, "learning_rate": 9.23623918820303e-05, "loss": 0.0593, "step": 43870 }, { "epoch": 1.4834849048311303, "grad_norm": 0.555709183216095, "learning_rate": 9.235799970532544e-05, "loss": 0.0575, "step": 43880 }, { "epoch": 1.4838229825213833, "grad_norm": 0.37518227100372314, "learning_rate": 9.235360637056878e-05, "loss": 0.0657, "step": 43890 }, { "epoch": 1.4841610602116366, "grad_norm": 0.20369885861873627, "learning_rate": 9.234921187788041e-05, "loss": 0.0515, "step": 43900 }, { "epoch": 1.48449913790189, "grad_norm": 0.3128197491168976, "learning_rate": 9.234481622738052e-05, "loss": 0.051, "step": 43910 }, { "epoch": 1.484837215592143, "grad_norm": 0.3165725767612457, "learning_rate": 9.234041941918927e-05, "loss": 0.0607, "step": 43920 }, { "epoch": 1.4851752932823963, "grad_norm": 0.212570458650589, "learning_rate": 9.233602145342684e-05, "loss": 0.0561, "step": 43930 }, { "epoch": 1.4855133709726496, "grad_norm": 0.34179818630218506, "learning_rate": 9.233162233021349e-05, "loss": 0.0586, "step": 43940 }, { "epoch": 1.4858514486629026, "grad_norm": 0.3177427649497986, "learning_rate": 9.23272220496695e-05, "loss": 0.0561, "step": 43950 }, { "epoch": 1.486189526353156, "grad_norm": 0.43098506331443787, "learning_rate": 9.232282061191515e-05, "loss": 0.0571, "step": 43960 }, { "epoch": 1.4865276040434092, "grad_norm": 0.539863646030426, "learning_rate": 9.231841801707079e-05, "loss": 0.0533, "step": 43970 }, { "epoch": 1.4868656817336623, "grad_norm": 0.5358266234397888, "learning_rate": 9.231401426525679e-05, "loss": 0.0616, "step": 43980 }, { "epoch": 1.4872037594239156, "grad_norm": 0.28736263513565063, "learning_rate": 9.230960935659352e-05, "loss": 0.0567, "step": 43990 }, { "epoch": 1.4875418371141689, "grad_norm": 0.2126256674528122, "learning_rate": 9.230520329120142e-05, "loss": 0.0576, "step": 44000 }, { "epoch": 1.487879914804422, "grad_norm": 0.23836031556129456, "learning_rate": 9.230079606920098e-05, "loss": 0.0564, "step": 44010 }, { "epoch": 1.4882179924946752, "grad_norm": 0.3511931598186493, "learning_rate": 9.229638769071265e-05, "loss": 0.0554, "step": 44020 }, { "epoch": 1.4885560701849285, "grad_norm": 0.6226288676261902, "learning_rate": 9.229197815585697e-05, "loss": 0.0593, "step": 44030 }, { "epoch": 1.4888941478751816, "grad_norm": 0.2561703026294708, "learning_rate": 9.228756746475451e-05, "loss": 0.0551, "step": 44040 }, { "epoch": 1.4892322255654349, "grad_norm": 0.2788468897342682, "learning_rate": 9.228315561752583e-05, "loss": 0.0599, "step": 44050 }, { "epoch": 1.4895703032556882, "grad_norm": 0.25621524453163147, "learning_rate": 9.227874261429157e-05, "loss": 0.0571, "step": 44060 }, { "epoch": 1.4899083809459412, "grad_norm": 1.0733312368392944, "learning_rate": 9.227432845517235e-05, "loss": 0.0583, "step": 44070 }, { "epoch": 1.4902464586361945, "grad_norm": 0.3071325421333313, "learning_rate": 9.22699131402889e-05, "loss": 0.0531, "step": 44080 }, { "epoch": 1.4905845363264478, "grad_norm": 0.3582649827003479, "learning_rate": 9.226549666976188e-05, "loss": 0.0618, "step": 44090 }, { "epoch": 1.4909226140167011, "grad_norm": 0.3947519361972809, "learning_rate": 9.226107904371208e-05, "loss": 0.0557, "step": 44100 }, { "epoch": 1.4912606917069542, "grad_norm": 0.5988633036613464, "learning_rate": 9.225666026226024e-05, "loss": 0.064, "step": 44110 }, { "epoch": 1.4915987693972075, "grad_norm": 0.4939824640750885, "learning_rate": 9.225224032552719e-05, "loss": 0.0589, "step": 44120 }, { "epoch": 1.4919368470874608, "grad_norm": 0.3685997426509857, "learning_rate": 9.224781923363374e-05, "loss": 0.0609, "step": 44130 }, { "epoch": 1.4922749247777138, "grad_norm": 0.2837047874927521, "learning_rate": 9.224339698670081e-05, "loss": 0.0578, "step": 44140 }, { "epoch": 1.4926130024679671, "grad_norm": 0.3096831738948822, "learning_rate": 9.223897358484929e-05, "loss": 0.0651, "step": 44150 }, { "epoch": 1.4929510801582204, "grad_norm": 0.4587409794330597, "learning_rate": 9.223454902820008e-05, "loss": 0.0614, "step": 44160 }, { "epoch": 1.4932891578484737, "grad_norm": 0.3897596597671509, "learning_rate": 9.223012331687416e-05, "loss": 0.0616, "step": 44170 }, { "epoch": 1.4936272355387268, "grad_norm": 0.2153462916612625, "learning_rate": 9.222569645099254e-05, "loss": 0.063, "step": 44180 }, { "epoch": 1.49396531322898, "grad_norm": 0.3146866261959076, "learning_rate": 9.222126843067624e-05, "loss": 0.0634, "step": 44190 }, { "epoch": 1.4943033909192334, "grad_norm": 0.20174190402030945, "learning_rate": 9.221683925604634e-05, "loss": 0.0574, "step": 44200 }, { "epoch": 1.4946414686094864, "grad_norm": 0.25065991282463074, "learning_rate": 9.221240892722389e-05, "loss": 0.0572, "step": 44210 }, { "epoch": 1.4949795462997397, "grad_norm": 0.26809439063072205, "learning_rate": 9.220797744433006e-05, "loss": 0.0583, "step": 44220 }, { "epoch": 1.495317623989993, "grad_norm": 0.2641662061214447, "learning_rate": 9.220354480748597e-05, "loss": 0.0621, "step": 44230 }, { "epoch": 1.495655701680246, "grad_norm": 0.2756577432155609, "learning_rate": 9.219911101681282e-05, "loss": 0.0565, "step": 44240 }, { "epoch": 1.4959937793704994, "grad_norm": 0.22937871515750885, "learning_rate": 9.219467607243183e-05, "loss": 0.0641, "step": 44250 }, { "epoch": 1.4963318570607527, "grad_norm": 0.2372129261493683, "learning_rate": 9.219023997446426e-05, "loss": 0.0497, "step": 44260 }, { "epoch": 1.4966699347510057, "grad_norm": 0.40961894392967224, "learning_rate": 9.218580272303137e-05, "loss": 0.0625, "step": 44270 }, { "epoch": 1.497008012441259, "grad_norm": 0.3188805878162384, "learning_rate": 9.218136431825448e-05, "loss": 0.0581, "step": 44280 }, { "epoch": 1.4973460901315123, "grad_norm": 0.2824525237083435, "learning_rate": 9.217692476025495e-05, "loss": 0.0521, "step": 44290 }, { "epoch": 1.4976841678217654, "grad_norm": 0.326537162065506, "learning_rate": 9.217248404915412e-05, "loss": 0.0586, "step": 44300 }, { "epoch": 1.4980222455120187, "grad_norm": 0.3305254280567169, "learning_rate": 9.216804218507343e-05, "loss": 0.0537, "step": 44310 }, { "epoch": 1.498360323202272, "grad_norm": 0.2737215757369995, "learning_rate": 9.21635991681343e-05, "loss": 0.0578, "step": 44320 }, { "epoch": 1.498698400892525, "grad_norm": 0.3525022864341736, "learning_rate": 9.215915499845822e-05, "loss": 0.0555, "step": 44330 }, { "epoch": 1.4990364785827783, "grad_norm": 0.3110436797142029, "learning_rate": 9.215470967616668e-05, "loss": 0.0619, "step": 44340 }, { "epoch": 1.4993745562730316, "grad_norm": 0.35465002059936523, "learning_rate": 9.215026320138123e-05, "loss": 0.0588, "step": 44350 }, { "epoch": 1.4997126339632847, "grad_norm": 0.644681990146637, "learning_rate": 9.21458155742234e-05, "loss": 0.06, "step": 44360 }, { "epoch": 1.500050711653538, "grad_norm": 0.37587639689445496, "learning_rate": 9.21413667948148e-05, "loss": 0.0535, "step": 44370 }, { "epoch": 1.5003887893437913, "grad_norm": 0.28062009811401367, "learning_rate": 9.213691686327707e-05, "loss": 0.052, "step": 44380 }, { "epoch": 1.5007268670340443, "grad_norm": 0.3053947389125824, "learning_rate": 9.213246577973187e-05, "loss": 0.0599, "step": 44390 }, { "epoch": 1.5010649447242976, "grad_norm": 0.3524883985519409, "learning_rate": 9.212801354430088e-05, "loss": 0.0609, "step": 44400 }, { "epoch": 1.501403022414551, "grad_norm": 0.2849745750427246, "learning_rate": 9.212356015710581e-05, "loss": 0.0563, "step": 44410 }, { "epoch": 1.501741100104804, "grad_norm": 0.3366580009460449, "learning_rate": 9.211910561826845e-05, "loss": 0.0603, "step": 44420 }, { "epoch": 1.5020791777950573, "grad_norm": 0.31174346804618835, "learning_rate": 9.211464992791057e-05, "loss": 0.0592, "step": 44430 }, { "epoch": 1.5024172554853106, "grad_norm": 0.44563958048820496, "learning_rate": 9.211019308615396e-05, "loss": 0.0606, "step": 44440 }, { "epoch": 1.5027553331755636, "grad_norm": 0.36093801259994507, "learning_rate": 9.21057350931205e-05, "loss": 0.0549, "step": 44450 }, { "epoch": 1.503093410865817, "grad_norm": 0.23104314506053925, "learning_rate": 9.210127594893207e-05, "loss": 0.0531, "step": 44460 }, { "epoch": 1.5034314885560702, "grad_norm": 0.2778114080429077, "learning_rate": 9.209681565371058e-05, "loss": 0.0541, "step": 44470 }, { "epoch": 1.5037695662463233, "grad_norm": 0.4091501235961914, "learning_rate": 9.209235420757794e-05, "loss": 0.0643, "step": 44480 }, { "epoch": 1.5041076439365766, "grad_norm": 0.32554692029953003, "learning_rate": 9.208789161065616e-05, "loss": 0.0503, "step": 44490 }, { "epoch": 1.5044457216268299, "grad_norm": 0.3380142152309418, "learning_rate": 9.208342786306725e-05, "loss": 0.0552, "step": 44500 }, { "epoch": 1.504783799317083, "grad_norm": 0.37739184498786926, "learning_rate": 9.207896296493321e-05, "loss": 0.0535, "step": 44510 }, { "epoch": 1.5051218770073362, "grad_norm": 0.2509155869483948, "learning_rate": 9.207449691637616e-05, "loss": 0.0565, "step": 44520 }, { "epoch": 1.5054599546975895, "grad_norm": 0.6016585826873779, "learning_rate": 9.207002971751814e-05, "loss": 0.0621, "step": 44530 }, { "epoch": 1.5057980323878426, "grad_norm": 0.19695404171943665, "learning_rate": 9.206556136848133e-05, "loss": 0.0542, "step": 44540 }, { "epoch": 1.506136110078096, "grad_norm": 0.36219877004623413, "learning_rate": 9.206109186938785e-05, "loss": 0.057, "step": 44550 }, { "epoch": 1.5064741877683492, "grad_norm": 0.3928464651107788, "learning_rate": 9.205662122035994e-05, "loss": 0.0578, "step": 44560 }, { "epoch": 1.5068122654586023, "grad_norm": 0.21502597630023956, "learning_rate": 9.20521494215198e-05, "loss": 0.0634, "step": 44570 }, { "epoch": 1.5071503431488558, "grad_norm": 0.37501242756843567, "learning_rate": 9.204767647298971e-05, "loss": 0.0579, "step": 44580 }, { "epoch": 1.5074884208391088, "grad_norm": 0.33807337284088135, "learning_rate": 9.204320237489192e-05, "loss": 0.0627, "step": 44590 }, { "epoch": 1.507826498529362, "grad_norm": 0.36603426933288574, "learning_rate": 9.203872712734879e-05, "loss": 0.0619, "step": 44600 }, { "epoch": 1.5081645762196154, "grad_norm": 0.2694862186908722, "learning_rate": 9.203425073048263e-05, "loss": 0.0551, "step": 44610 }, { "epoch": 1.5085026539098685, "grad_norm": 0.22315488755702972, "learning_rate": 9.202977318441587e-05, "loss": 0.052, "step": 44620 }, { "epoch": 1.5088407316001216, "grad_norm": 0.2238471359014511, "learning_rate": 9.202529448927091e-05, "loss": 0.0538, "step": 44630 }, { "epoch": 1.509178809290375, "grad_norm": 0.20607145130634308, "learning_rate": 9.202081464517016e-05, "loss": 0.0523, "step": 44640 }, { "epoch": 1.5095168869806281, "grad_norm": 0.3022650182247162, "learning_rate": 9.201633365223614e-05, "loss": 0.0624, "step": 44650 }, { "epoch": 1.5098549646708812, "grad_norm": 0.19240552186965942, "learning_rate": 9.201185151059132e-05, "loss": 0.0578, "step": 44660 }, { "epoch": 1.5101930423611347, "grad_norm": 0.31516170501708984, "learning_rate": 9.200736822035829e-05, "loss": 0.0545, "step": 44670 }, { "epoch": 1.5105311200513878, "grad_norm": 0.3808117210865021, "learning_rate": 9.200288378165958e-05, "loss": 0.0614, "step": 44680 }, { "epoch": 1.510869197741641, "grad_norm": 0.2939152121543884, "learning_rate": 9.199839819461782e-05, "loss": 0.0579, "step": 44690 }, { "epoch": 1.5112072754318944, "grad_norm": 0.284306138753891, "learning_rate": 9.199391145935561e-05, "loss": 0.0586, "step": 44700 }, { "epoch": 1.5115453531221474, "grad_norm": 0.40352991223335266, "learning_rate": 9.198942357599565e-05, "loss": 0.0576, "step": 44710 }, { "epoch": 1.5118834308124007, "grad_norm": 0.314222514629364, "learning_rate": 9.198493454466063e-05, "loss": 0.0574, "step": 44720 }, { "epoch": 1.512221508502654, "grad_norm": 0.3192752003669739, "learning_rate": 9.198044436547326e-05, "loss": 0.0524, "step": 44730 }, { "epoch": 1.512559586192907, "grad_norm": 0.41593965888023376, "learning_rate": 9.197595303855633e-05, "loss": 0.0558, "step": 44740 }, { "epoch": 1.5128976638831604, "grad_norm": 0.25715044140815735, "learning_rate": 9.19714605640326e-05, "loss": 0.0553, "step": 44750 }, { "epoch": 1.5132357415734137, "grad_norm": 0.37019696831703186, "learning_rate": 9.19669669420249e-05, "loss": 0.0537, "step": 44760 }, { "epoch": 1.5135738192636667, "grad_norm": 0.2650265693664551, "learning_rate": 9.19624721726561e-05, "loss": 0.0505, "step": 44770 }, { "epoch": 1.51391189695392, "grad_norm": 0.7394532561302185, "learning_rate": 9.195797625604908e-05, "loss": 0.0577, "step": 44780 }, { "epoch": 1.5142499746441733, "grad_norm": 0.5193639993667603, "learning_rate": 9.195347919232673e-05, "loss": 0.0558, "step": 44790 }, { "epoch": 1.5145880523344264, "grad_norm": 0.2844461500644684, "learning_rate": 9.194898098161204e-05, "loss": 0.059, "step": 44800 }, { "epoch": 1.5149261300246797, "grad_norm": 0.40250363945961, "learning_rate": 9.194448162402797e-05, "loss": 0.055, "step": 44810 }, { "epoch": 1.515264207714933, "grad_norm": 0.26630160212516785, "learning_rate": 9.193998111969753e-05, "loss": 0.0545, "step": 44820 }, { "epoch": 1.515602285405186, "grad_norm": 0.22598868608474731, "learning_rate": 9.193547946874374e-05, "loss": 0.0553, "step": 44830 }, { "epoch": 1.5159403630954393, "grad_norm": 0.3356873393058777, "learning_rate": 9.193097667128972e-05, "loss": 0.0565, "step": 44840 }, { "epoch": 1.5162784407856926, "grad_norm": 0.34827888011932373, "learning_rate": 9.192647272745855e-05, "loss": 0.0559, "step": 44850 }, { "epoch": 1.5166165184759457, "grad_norm": 0.2659357190132141, "learning_rate": 9.192196763737335e-05, "loss": 0.0605, "step": 44860 }, { "epoch": 1.516954596166199, "grad_norm": 0.322635680437088, "learning_rate": 9.191746140115732e-05, "loss": 0.0621, "step": 44870 }, { "epoch": 1.5172926738564523, "grad_norm": 0.44645634293556213, "learning_rate": 9.191295401893365e-05, "loss": 0.0583, "step": 44880 }, { "epoch": 1.5176307515467053, "grad_norm": 0.199236661195755, "learning_rate": 9.190844549082553e-05, "loss": 0.0571, "step": 44890 }, { "epoch": 1.5179688292369586, "grad_norm": 0.5335314869880676, "learning_rate": 9.190393581695629e-05, "loss": 0.0604, "step": 44900 }, { "epoch": 1.518306906927212, "grad_norm": 0.34253713488578796, "learning_rate": 9.189942499744916e-05, "loss": 0.0545, "step": 44910 }, { "epoch": 1.518644984617465, "grad_norm": 0.3630400598049164, "learning_rate": 9.189491303242752e-05, "loss": 0.0578, "step": 44920 }, { "epoch": 1.5189830623077183, "grad_norm": 0.29950088262557983, "learning_rate": 9.189039992201467e-05, "loss": 0.0532, "step": 44930 }, { "epoch": 1.5193211399979716, "grad_norm": 0.36577126383781433, "learning_rate": 9.188588566633403e-05, "loss": 0.0555, "step": 44940 }, { "epoch": 1.5196592176882247, "grad_norm": 0.38999417424201965, "learning_rate": 9.188137026550903e-05, "loss": 0.0605, "step": 44950 }, { "epoch": 1.519997295378478, "grad_norm": 0.3256262242794037, "learning_rate": 9.187685371966308e-05, "loss": 0.0598, "step": 44960 }, { "epoch": 1.5203353730687312, "grad_norm": 0.5550006628036499, "learning_rate": 9.187233602891969e-05, "loss": 0.0587, "step": 44970 }, { "epoch": 1.5206734507589843, "grad_norm": 0.29239213466644287, "learning_rate": 9.186781719340236e-05, "loss": 0.0584, "step": 44980 }, { "epoch": 1.5210115284492376, "grad_norm": 0.2787850499153137, "learning_rate": 9.186329721323464e-05, "loss": 0.0562, "step": 44990 }, { "epoch": 1.5213496061394909, "grad_norm": 0.32175105810165405, "learning_rate": 9.18587760885401e-05, "loss": 0.0558, "step": 45000 }, { "epoch": 1.521687683829744, "grad_norm": 0.32674312591552734, "learning_rate": 9.185425381944234e-05, "loss": 0.0594, "step": 45010 }, { "epoch": 1.5220257615199972, "grad_norm": 0.40018990635871887, "learning_rate": 9.184973040606502e-05, "loss": 0.0576, "step": 45020 }, { "epoch": 1.5223638392102505, "grad_norm": 0.44698774814605713, "learning_rate": 9.18452058485318e-05, "loss": 0.0549, "step": 45030 }, { "epoch": 1.5227019169005036, "grad_norm": 0.21995128691196442, "learning_rate": 9.184068014696635e-05, "loss": 0.0591, "step": 45040 }, { "epoch": 1.5230399945907571, "grad_norm": 0.37476786971092224, "learning_rate": 9.183615330149242e-05, "loss": 0.0643, "step": 45050 }, { "epoch": 1.5233780722810102, "grad_norm": 0.21653605997562408, "learning_rate": 9.183162531223379e-05, "loss": 0.0602, "step": 45060 }, { "epoch": 1.5237161499712633, "grad_norm": 0.1982632279396057, "learning_rate": 9.182709617931424e-05, "loss": 0.056, "step": 45070 }, { "epoch": 1.5240542276615168, "grad_norm": 0.24273018538951874, "learning_rate": 9.18225659028576e-05, "loss": 0.0609, "step": 45080 }, { "epoch": 1.5243923053517698, "grad_norm": 0.36313799023628235, "learning_rate": 9.181803448298771e-05, "loss": 0.0594, "step": 45090 }, { "epoch": 1.524730383042023, "grad_norm": 0.39138495922088623, "learning_rate": 9.181350191982848e-05, "loss": 0.0549, "step": 45100 }, { "epoch": 1.5250684607322764, "grad_norm": 0.42011475563049316, "learning_rate": 9.18089682135038e-05, "loss": 0.0569, "step": 45110 }, { "epoch": 1.5254065384225295, "grad_norm": 0.2887287735939026, "learning_rate": 9.180443336413764e-05, "loss": 0.0568, "step": 45120 }, { "epoch": 1.5257446161127826, "grad_norm": 0.32081276178359985, "learning_rate": 9.179989737185397e-05, "loss": 0.06, "step": 45130 }, { "epoch": 1.526082693803036, "grad_norm": 0.35651278495788574, "learning_rate": 9.179536023677682e-05, "loss": 0.0579, "step": 45140 }, { "epoch": 1.5264207714932891, "grad_norm": 0.4156666398048401, "learning_rate": 9.179082195903022e-05, "loss": 0.0506, "step": 45150 }, { "epoch": 1.5267588491835424, "grad_norm": 0.2965644598007202, "learning_rate": 9.178628253873825e-05, "loss": 0.0548, "step": 45160 }, { "epoch": 1.5270969268737957, "grad_norm": 0.217654287815094, "learning_rate": 9.178174197602501e-05, "loss": 0.0525, "step": 45170 }, { "epoch": 1.5274350045640488, "grad_norm": 0.35803481936454773, "learning_rate": 9.177720027101464e-05, "loss": 0.0589, "step": 45180 }, { "epoch": 1.527773082254302, "grad_norm": 0.2969190776348114, "learning_rate": 9.177265742383132e-05, "loss": 0.0622, "step": 45190 }, { "epoch": 1.5281111599445554, "grad_norm": 0.20462937653064728, "learning_rate": 9.176811343459923e-05, "loss": 0.0536, "step": 45200 }, { "epoch": 1.5284492376348084, "grad_norm": 0.2163236290216446, "learning_rate": 9.176356830344262e-05, "loss": 0.0561, "step": 45210 }, { "epoch": 1.5287873153250617, "grad_norm": 0.35520297288894653, "learning_rate": 9.175902203048573e-05, "loss": 0.0539, "step": 45220 }, { "epoch": 1.529125393015315, "grad_norm": 0.2023800015449524, "learning_rate": 9.175447461585287e-05, "loss": 0.0544, "step": 45230 }, { "epoch": 1.529463470705568, "grad_norm": 0.29822003841400146, "learning_rate": 9.174992605966835e-05, "loss": 0.0559, "step": 45240 }, { "epoch": 1.5298015483958214, "grad_norm": 0.21230943500995636, "learning_rate": 9.174537636205656e-05, "loss": 0.0564, "step": 45250 }, { "epoch": 1.5301396260860747, "grad_norm": 0.295522004365921, "learning_rate": 9.174082552314186e-05, "loss": 0.0547, "step": 45260 }, { "epoch": 1.5304777037763277, "grad_norm": 0.3295157551765442, "learning_rate": 9.173627354304865e-05, "loss": 0.0607, "step": 45270 }, { "epoch": 1.530815781466581, "grad_norm": 1.7221570014953613, "learning_rate": 9.173172042190141e-05, "loss": 0.0597, "step": 45280 }, { "epoch": 1.5311538591568343, "grad_norm": 0.24042236804962158, "learning_rate": 9.172716615982461e-05, "loss": 0.0554, "step": 45290 }, { "epoch": 1.5314919368470874, "grad_norm": 0.5404477119445801, "learning_rate": 9.172261075694277e-05, "loss": 0.0495, "step": 45300 }, { "epoch": 1.5318300145373407, "grad_norm": 0.3692869246006012, "learning_rate": 9.171805421338043e-05, "loss": 0.0552, "step": 45310 }, { "epoch": 1.532168092227594, "grad_norm": 0.9438931345939636, "learning_rate": 9.171349652926216e-05, "loss": 0.0597, "step": 45320 }, { "epoch": 1.532506169917847, "grad_norm": 0.35426491498947144, "learning_rate": 9.170893770471255e-05, "loss": 0.0521, "step": 45330 }, { "epoch": 1.5328442476081003, "grad_norm": 0.3180460035800934, "learning_rate": 9.170437773985627e-05, "loss": 0.0557, "step": 45340 }, { "epoch": 1.5331823252983536, "grad_norm": 0.3461400270462036, "learning_rate": 9.169981663481795e-05, "loss": 0.0517, "step": 45350 }, { "epoch": 1.5335204029886067, "grad_norm": 0.22805766761302948, "learning_rate": 9.169525438972234e-05, "loss": 0.0548, "step": 45360 }, { "epoch": 1.53385848067886, "grad_norm": 0.25953832268714905, "learning_rate": 9.169069100469411e-05, "loss": 0.0554, "step": 45370 }, { "epoch": 1.5341965583691133, "grad_norm": 0.47816595435142517, "learning_rate": 9.168612647985806e-05, "loss": 0.0605, "step": 45380 }, { "epoch": 1.5345346360593664, "grad_norm": 0.354104608297348, "learning_rate": 9.168156081533898e-05, "loss": 0.0564, "step": 45390 }, { "epoch": 1.5348727137496196, "grad_norm": 0.3720772862434387, "learning_rate": 9.167699401126167e-05, "loss": 0.0548, "step": 45400 }, { "epoch": 1.535210791439873, "grad_norm": 0.3806977868080139, "learning_rate": 9.167242606775101e-05, "loss": 0.0551, "step": 45410 }, { "epoch": 1.535548869130126, "grad_norm": 0.2504129409790039, "learning_rate": 9.166785698493189e-05, "loss": 0.0597, "step": 45420 }, { "epoch": 1.5358869468203793, "grad_norm": 0.3449966311454773, "learning_rate": 9.166328676292919e-05, "loss": 0.0571, "step": 45430 }, { "epoch": 1.5362250245106326, "grad_norm": 0.30036842823028564, "learning_rate": 9.16587154018679e-05, "loss": 0.0511, "step": 45440 }, { "epoch": 1.5365631022008857, "grad_norm": 0.284682035446167, "learning_rate": 9.165414290187298e-05, "loss": 0.0606, "step": 45450 }, { "epoch": 1.536901179891139, "grad_norm": 0.2578237056732178, "learning_rate": 9.164956926306944e-05, "loss": 0.0648, "step": 45460 }, { "epoch": 1.5372392575813922, "grad_norm": 0.3256244957447052, "learning_rate": 9.164499448558232e-05, "loss": 0.0606, "step": 45470 }, { "epoch": 1.5375773352716453, "grad_norm": 0.45412978529930115, "learning_rate": 9.16404185695367e-05, "loss": 0.0559, "step": 45480 }, { "epoch": 1.5379154129618986, "grad_norm": 0.29298076033592224, "learning_rate": 9.163584151505767e-05, "loss": 0.0563, "step": 45490 }, { "epoch": 1.538253490652152, "grad_norm": 0.26829710602760315, "learning_rate": 9.163126332227039e-05, "loss": 0.052, "step": 45500 }, { "epoch": 1.538591568342405, "grad_norm": 0.33172866702079773, "learning_rate": 9.16266839913e-05, "loss": 0.0519, "step": 45510 }, { "epoch": 1.5389296460326582, "grad_norm": 0.29978591203689575, "learning_rate": 9.162210352227171e-05, "loss": 0.0618, "step": 45520 }, { "epoch": 1.5392677237229115, "grad_norm": 0.3230612277984619, "learning_rate": 9.161752191531075e-05, "loss": 0.0622, "step": 45530 }, { "epoch": 1.5396058014131646, "grad_norm": 0.2311132550239563, "learning_rate": 9.161293917054236e-05, "loss": 0.0523, "step": 45540 }, { "epoch": 1.5399438791034181, "grad_norm": 0.3089430332183838, "learning_rate": 9.160835528809187e-05, "loss": 0.0625, "step": 45550 }, { "epoch": 1.5402819567936712, "grad_norm": 0.2321566343307495, "learning_rate": 9.160377026808457e-05, "loss": 0.0555, "step": 45560 }, { "epoch": 1.5406200344839243, "grad_norm": 0.3043546676635742, "learning_rate": 9.159918411064583e-05, "loss": 0.0621, "step": 45570 }, { "epoch": 1.5409581121741778, "grad_norm": 3.466294288635254, "learning_rate": 9.1594596815901e-05, "loss": 0.0561, "step": 45580 }, { "epoch": 1.5412961898644308, "grad_norm": 0.2542620599269867, "learning_rate": 9.159000838397553e-05, "loss": 0.0621, "step": 45590 }, { "epoch": 1.541634267554684, "grad_norm": 0.350452184677124, "learning_rate": 9.158541881499486e-05, "loss": 0.0588, "step": 45600 }, { "epoch": 1.5419723452449374, "grad_norm": 0.2610608637332916, "learning_rate": 9.158082810908447e-05, "loss": 0.0513, "step": 45610 }, { "epoch": 1.5423104229351905, "grad_norm": 0.3618064522743225, "learning_rate": 9.157623626636984e-05, "loss": 0.0489, "step": 45620 }, { "epoch": 1.5426485006254436, "grad_norm": 0.19081000983715057, "learning_rate": 9.157164328697654e-05, "loss": 0.0584, "step": 45630 }, { "epoch": 1.542986578315697, "grad_norm": 0.3334667682647705, "learning_rate": 9.156704917103013e-05, "loss": 0.0557, "step": 45640 }, { "epoch": 1.5433246560059501, "grad_norm": 0.30898845195770264, "learning_rate": 9.15624539186562e-05, "loss": 0.0566, "step": 45650 }, { "epoch": 1.5436627336962034, "grad_norm": 0.4600141942501068, "learning_rate": 9.155785752998043e-05, "loss": 0.0545, "step": 45660 }, { "epoch": 1.5440008113864567, "grad_norm": 0.2226000428199768, "learning_rate": 9.155326000512841e-05, "loss": 0.0561, "step": 45670 }, { "epoch": 1.5443388890767098, "grad_norm": 0.20264527201652527, "learning_rate": 9.15486613442259e-05, "loss": 0.052, "step": 45680 }, { "epoch": 1.544676966766963, "grad_norm": 0.4508977234363556, "learning_rate": 9.154406154739858e-05, "loss": 0.0574, "step": 45690 }, { "epoch": 1.5450150444572164, "grad_norm": 0.2038630098104477, "learning_rate": 9.153946061477223e-05, "loss": 0.0548, "step": 45700 }, { "epoch": 1.5453531221474694, "grad_norm": 0.407351553440094, "learning_rate": 9.153485854647264e-05, "loss": 0.0623, "step": 45710 }, { "epoch": 1.5456911998377227, "grad_norm": 0.2911006510257721, "learning_rate": 9.153025534262561e-05, "loss": 0.0551, "step": 45720 }, { "epoch": 1.546029277527976, "grad_norm": 0.23430578410625458, "learning_rate": 9.152565100335699e-05, "loss": 0.0604, "step": 45730 }, { "epoch": 1.546367355218229, "grad_norm": 0.18055343627929688, "learning_rate": 9.15210455287927e-05, "loss": 0.0569, "step": 45740 }, { "epoch": 1.5467054329084824, "grad_norm": 0.501251220703125, "learning_rate": 9.15164389190586e-05, "loss": 0.0618, "step": 45750 }, { "epoch": 1.5470435105987357, "grad_norm": 0.4677075147628784, "learning_rate": 9.151183117428067e-05, "loss": 0.0555, "step": 45760 }, { "epoch": 1.5473815882889888, "grad_norm": 0.25254517793655396, "learning_rate": 9.150722229458487e-05, "loss": 0.0551, "step": 45770 }, { "epoch": 1.547719665979242, "grad_norm": 0.3919758200645447, "learning_rate": 9.15026122800972e-05, "loss": 0.057, "step": 45780 }, { "epoch": 1.5480577436694953, "grad_norm": 0.2807323932647705, "learning_rate": 9.149800113094371e-05, "loss": 0.0622, "step": 45790 }, { "epoch": 1.5483958213597484, "grad_norm": 0.36543768644332886, "learning_rate": 9.149338884725045e-05, "loss": 0.0578, "step": 45800 }, { "epoch": 1.5487338990500017, "grad_norm": 0.23203293979167938, "learning_rate": 9.148877542914353e-05, "loss": 0.0545, "step": 45810 }, { "epoch": 1.549071976740255, "grad_norm": 0.30667856335639954, "learning_rate": 9.148416087674909e-05, "loss": 0.0533, "step": 45820 }, { "epoch": 1.549410054430508, "grad_norm": 0.30176517367362976, "learning_rate": 9.147954519019327e-05, "loss": 0.0517, "step": 45830 }, { "epoch": 1.5497481321207613, "grad_norm": 0.3519397974014282, "learning_rate": 9.147492836960225e-05, "loss": 0.0597, "step": 45840 }, { "epoch": 1.5500862098110146, "grad_norm": 0.36097535490989685, "learning_rate": 9.147031041510229e-05, "loss": 0.0562, "step": 45850 }, { "epoch": 1.5504242875012677, "grad_norm": 0.3617238402366638, "learning_rate": 9.146569132681961e-05, "loss": 0.054, "step": 45860 }, { "epoch": 1.550762365191521, "grad_norm": 0.3785706162452698, "learning_rate": 9.146107110488051e-05, "loss": 0.0532, "step": 45870 }, { "epoch": 1.5511004428817743, "grad_norm": 0.27999386191368103, "learning_rate": 9.145644974941132e-05, "loss": 0.0573, "step": 45880 }, { "epoch": 1.5514385205720274, "grad_norm": 0.2559223175048828, "learning_rate": 9.145182726053835e-05, "loss": 0.0568, "step": 45890 }, { "epoch": 1.5517765982622806, "grad_norm": 0.3618946969509125, "learning_rate": 9.144720363838799e-05, "loss": 0.0571, "step": 45900 }, { "epoch": 1.552114675952534, "grad_norm": 0.2601381242275238, "learning_rate": 9.144257888308667e-05, "loss": 0.062, "step": 45910 }, { "epoch": 1.552452753642787, "grad_norm": 0.3854595422744751, "learning_rate": 9.143795299476081e-05, "loss": 0.0633, "step": 45920 }, { "epoch": 1.5527908313330403, "grad_norm": 0.27282869815826416, "learning_rate": 9.143332597353686e-05, "loss": 0.0551, "step": 45930 }, { "epoch": 1.5531289090232936, "grad_norm": 0.3850380778312683, "learning_rate": 9.142869781954137e-05, "loss": 0.0571, "step": 45940 }, { "epoch": 1.5534669867135467, "grad_norm": 0.32476964592933655, "learning_rate": 9.142406853290084e-05, "loss": 0.0623, "step": 45950 }, { "epoch": 1.5538050644038, "grad_norm": 0.2606852054595947, "learning_rate": 9.141943811374183e-05, "loss": 0.0577, "step": 45960 }, { "epoch": 1.5541431420940532, "grad_norm": 0.35298478603363037, "learning_rate": 9.141480656219094e-05, "loss": 0.0559, "step": 45970 }, { "epoch": 1.5544812197843063, "grad_norm": 0.34266480803489685, "learning_rate": 9.141017387837483e-05, "loss": 0.0578, "step": 45980 }, { "epoch": 1.5548192974745596, "grad_norm": 0.32472115755081177, "learning_rate": 9.140554006242009e-05, "loss": 0.0526, "step": 45990 }, { "epoch": 1.555157375164813, "grad_norm": 0.41762298345565796, "learning_rate": 9.140090511445345e-05, "loss": 0.0563, "step": 46000 }, { "epoch": 1.555495452855066, "grad_norm": 0.31860196590423584, "learning_rate": 9.139626903460161e-05, "loss": 0.0573, "step": 46010 }, { "epoch": 1.5558335305453195, "grad_norm": 0.27050459384918213, "learning_rate": 9.139163182299133e-05, "loss": 0.0584, "step": 46020 }, { "epoch": 1.5561716082355725, "grad_norm": 0.31958961486816406, "learning_rate": 9.138699347974939e-05, "loss": 0.0563, "step": 46030 }, { "epoch": 1.5565096859258256, "grad_norm": 0.25320130586624146, "learning_rate": 9.13823540050026e-05, "loss": 0.0606, "step": 46040 }, { "epoch": 1.5568477636160791, "grad_norm": 0.275341272354126, "learning_rate": 9.137771339887779e-05, "loss": 0.0609, "step": 46050 }, { "epoch": 1.5571858413063322, "grad_norm": 0.267025887966156, "learning_rate": 9.137307166150185e-05, "loss": 0.0552, "step": 46060 }, { "epoch": 1.5575239189965853, "grad_norm": 0.2745276689529419, "learning_rate": 9.136842879300167e-05, "loss": 0.0559, "step": 46070 }, { "epoch": 1.5578619966868388, "grad_norm": 0.24487164616584778, "learning_rate": 9.136378479350418e-05, "loss": 0.0571, "step": 46080 }, { "epoch": 1.5582000743770918, "grad_norm": 0.40164613723754883, "learning_rate": 9.135913966313637e-05, "loss": 0.0516, "step": 46090 }, { "epoch": 1.558538152067345, "grad_norm": 0.2518320381641388, "learning_rate": 9.13544934020252e-05, "loss": 0.0587, "step": 46100 }, { "epoch": 1.5588762297575984, "grad_norm": 0.23354943096637726, "learning_rate": 9.134984601029773e-05, "loss": 0.0561, "step": 46110 }, { "epoch": 1.5592143074478515, "grad_norm": 0.4024699926376343, "learning_rate": 9.1345197488081e-05, "loss": 0.0592, "step": 46120 }, { "epoch": 1.5595523851381048, "grad_norm": 0.24613825976848602, "learning_rate": 9.13405478355021e-05, "loss": 0.0569, "step": 46130 }, { "epoch": 1.559890462828358, "grad_norm": 0.3232359290122986, "learning_rate": 9.133589705268817e-05, "loss": 0.0544, "step": 46140 }, { "epoch": 1.5602285405186112, "grad_norm": 0.2400941550731659, "learning_rate": 9.133124513976633e-05, "loss": 0.0521, "step": 46150 }, { "epoch": 1.5605666182088644, "grad_norm": 0.24441197514533997, "learning_rate": 9.132659209686377e-05, "loss": 0.0598, "step": 46160 }, { "epoch": 1.5609046958991177, "grad_norm": 0.3195306658744812, "learning_rate": 9.132193792410771e-05, "loss": 0.0566, "step": 46170 }, { "epoch": 1.5612427735893708, "grad_norm": 0.2509234547615051, "learning_rate": 9.13172826216254e-05, "loss": 0.0609, "step": 46180 }, { "epoch": 1.561580851279624, "grad_norm": 0.2833360433578491, "learning_rate": 9.13126261895441e-05, "loss": 0.0628, "step": 46190 }, { "epoch": 1.5619189289698774, "grad_norm": 0.4404905140399933, "learning_rate": 9.130796862799111e-05, "loss": 0.0547, "step": 46200 }, { "epoch": 1.5622570066601305, "grad_norm": 0.4278721213340759, "learning_rate": 9.13033099370938e-05, "loss": 0.0578, "step": 46210 }, { "epoch": 1.5625950843503837, "grad_norm": 0.24068264663219452, "learning_rate": 9.12986501169795e-05, "loss": 0.0532, "step": 46220 }, { "epoch": 1.562933162040637, "grad_norm": 0.3114660680294037, "learning_rate": 9.129398916777561e-05, "loss": 0.0558, "step": 46230 }, { "epoch": 1.56327123973089, "grad_norm": 0.22707104682922363, "learning_rate": 9.128932708960959e-05, "loss": 0.0595, "step": 46240 }, { "epoch": 1.5636093174211434, "grad_norm": 0.32939639687538147, "learning_rate": 9.128466388260888e-05, "loss": 0.0571, "step": 46250 }, { "epoch": 1.5639473951113967, "grad_norm": 0.33517178893089294, "learning_rate": 9.127999954690094e-05, "loss": 0.0562, "step": 46260 }, { "epoch": 1.5642854728016498, "grad_norm": 0.26169219613075256, "learning_rate": 9.127533408261335e-05, "loss": 0.0556, "step": 46270 }, { "epoch": 1.564623550491903, "grad_norm": 0.3318728506565094, "learning_rate": 9.127066748987363e-05, "loss": 0.0563, "step": 46280 }, { "epoch": 1.5649616281821563, "grad_norm": 0.26978495717048645, "learning_rate": 9.126599976880934e-05, "loss": 0.0567, "step": 46290 }, { "epoch": 1.5652997058724094, "grad_norm": 0.361056387424469, "learning_rate": 9.126133091954814e-05, "loss": 0.0577, "step": 46300 }, { "epoch": 1.5656377835626627, "grad_norm": 0.3019954264163971, "learning_rate": 9.125666094221766e-05, "loss": 0.0541, "step": 46310 }, { "epoch": 1.565975861252916, "grad_norm": 0.20564134418964386, "learning_rate": 9.125198983694556e-05, "loss": 0.0582, "step": 46320 }, { "epoch": 1.566313938943169, "grad_norm": 0.3892620801925659, "learning_rate": 9.124731760385956e-05, "loss": 0.0603, "step": 46330 }, { "epoch": 1.5666520166334224, "grad_norm": 0.2219911515712738, "learning_rate": 9.124264424308738e-05, "loss": 0.0521, "step": 46340 }, { "epoch": 1.5669900943236756, "grad_norm": 0.38188427686691284, "learning_rate": 9.12379697547568e-05, "loss": 0.0577, "step": 46350 }, { "epoch": 1.5673281720139287, "grad_norm": 0.2989346981048584, "learning_rate": 9.123329413899564e-05, "loss": 0.0576, "step": 46360 }, { "epoch": 1.567666249704182, "grad_norm": 0.26151183247566223, "learning_rate": 9.122861739593169e-05, "loss": 0.0564, "step": 46370 }, { "epoch": 1.5680043273944353, "grad_norm": 0.47680994868278503, "learning_rate": 9.122393952569283e-05, "loss": 0.057, "step": 46380 }, { "epoch": 1.5683424050846884, "grad_norm": 0.31757256388664246, "learning_rate": 9.121926052840698e-05, "loss": 0.0625, "step": 46390 }, { "epoch": 1.5686804827749417, "grad_norm": 0.5212836861610413, "learning_rate": 9.121458040420199e-05, "loss": 0.0578, "step": 46400 }, { "epoch": 1.569018560465195, "grad_norm": 0.1464037448167801, "learning_rate": 9.120989915320588e-05, "loss": 0.0543, "step": 46410 }, { "epoch": 1.569356638155448, "grad_norm": 0.27925145626068115, "learning_rate": 9.120521677554661e-05, "loss": 0.0642, "step": 46420 }, { "epoch": 1.5696947158457013, "grad_norm": 0.38481003046035767, "learning_rate": 9.120053327135219e-05, "loss": 0.0562, "step": 46430 }, { "epoch": 1.5700327935359546, "grad_norm": 0.3339003026485443, "learning_rate": 9.119584864075066e-05, "loss": 0.06, "step": 46440 }, { "epoch": 1.5703708712262077, "grad_norm": 0.28989478945732117, "learning_rate": 9.11911628838701e-05, "loss": 0.0543, "step": 46450 }, { "epoch": 1.570708948916461, "grad_norm": 0.2086474746465683, "learning_rate": 9.118647600083864e-05, "loss": 0.0563, "step": 46460 }, { "epoch": 1.5710470266067142, "grad_norm": 0.2725008726119995, "learning_rate": 9.118178799178439e-05, "loss": 0.0575, "step": 46470 }, { "epoch": 1.5713851042969673, "grad_norm": 0.4234864115715027, "learning_rate": 9.117709885683553e-05, "loss": 0.0581, "step": 46480 }, { "epoch": 1.5717231819872208, "grad_norm": 0.25103917717933655, "learning_rate": 9.117240859612026e-05, "loss": 0.0636, "step": 46490 }, { "epoch": 1.572061259677474, "grad_norm": 0.30189594626426697, "learning_rate": 9.11677172097668e-05, "loss": 0.0543, "step": 46500 }, { "epoch": 1.572399337367727, "grad_norm": 0.4453308582305908, "learning_rate": 9.116302469790341e-05, "loss": 0.0602, "step": 46510 }, { "epoch": 1.5727374150579805, "grad_norm": 0.34673628211021423, "learning_rate": 9.11583310606584e-05, "loss": 0.0568, "step": 46520 }, { "epoch": 1.5730754927482336, "grad_norm": 0.2586427330970764, "learning_rate": 9.115363629816009e-05, "loss": 0.0564, "step": 46530 }, { "epoch": 1.5734135704384866, "grad_norm": 0.317892462015152, "learning_rate": 9.114894041053681e-05, "loss": 0.0584, "step": 46540 }, { "epoch": 1.5737516481287401, "grad_norm": 0.2372204065322876, "learning_rate": 9.114424339791697e-05, "loss": 0.059, "step": 46550 }, { "epoch": 1.5740897258189932, "grad_norm": 0.24390794336795807, "learning_rate": 9.113954526042895e-05, "loss": 0.048, "step": 46560 }, { "epoch": 1.5744278035092463, "grad_norm": 0.3664223551750183, "learning_rate": 9.113484599820125e-05, "loss": 0.0544, "step": 46570 }, { "epoch": 1.5747658811994998, "grad_norm": 0.25170034170150757, "learning_rate": 9.113014561136229e-05, "loss": 0.0529, "step": 46580 }, { "epoch": 1.5751039588897529, "grad_norm": 0.34913891553878784, "learning_rate": 9.11254441000406e-05, "loss": 0.0589, "step": 46590 }, { "epoch": 1.575442036580006, "grad_norm": 0.49058303236961365, "learning_rate": 9.112074146436473e-05, "loss": 0.0551, "step": 46600 }, { "epoch": 1.5757801142702594, "grad_norm": 0.21571776270866394, "learning_rate": 9.111603770446324e-05, "loss": 0.0562, "step": 46610 }, { "epoch": 1.5761181919605125, "grad_norm": 0.20281799137592316, "learning_rate": 9.111133282046471e-05, "loss": 0.0556, "step": 46620 }, { "epoch": 1.5764562696507658, "grad_norm": 0.46363365650177, "learning_rate": 9.11066268124978e-05, "loss": 0.0554, "step": 46630 }, { "epoch": 1.576794347341019, "grad_norm": 0.3205604553222656, "learning_rate": 9.110191968069115e-05, "loss": 0.0615, "step": 46640 }, { "epoch": 1.5771324250312722, "grad_norm": 0.400907427072525, "learning_rate": 9.109721142517346e-05, "loss": 0.0578, "step": 46650 }, { "epoch": 1.5774705027215254, "grad_norm": 1.0008434057235718, "learning_rate": 9.109250204607344e-05, "loss": 0.0581, "step": 46660 }, { "epoch": 1.5778085804117787, "grad_norm": 0.6954232454299927, "learning_rate": 9.108779154351986e-05, "loss": 0.0576, "step": 46670 }, { "epoch": 1.5781466581020318, "grad_norm": 0.24554668366909027, "learning_rate": 9.108307991764148e-05, "loss": 0.0606, "step": 46680 }, { "epoch": 1.578484735792285, "grad_norm": 0.36052843928337097, "learning_rate": 9.107836716856715e-05, "loss": 0.062, "step": 46690 }, { "epoch": 1.5788228134825384, "grad_norm": 0.21481125056743622, "learning_rate": 9.107365329642569e-05, "loss": 0.056, "step": 46700 }, { "epoch": 1.5791608911727915, "grad_norm": 0.33052030205726624, "learning_rate": 9.106893830134598e-05, "loss": 0.0641, "step": 46710 }, { "epoch": 1.5794989688630447, "grad_norm": 0.3235275447368622, "learning_rate": 9.106422218345693e-05, "loss": 0.057, "step": 46720 }, { "epoch": 1.579837046553298, "grad_norm": 0.31025537848472595, "learning_rate": 9.105950494288747e-05, "loss": 0.0632, "step": 46730 }, { "epoch": 1.580175124243551, "grad_norm": 0.33381080627441406, "learning_rate": 9.105478657976656e-05, "loss": 0.0591, "step": 46740 }, { "epoch": 1.5805132019338044, "grad_norm": 0.2960173487663269, "learning_rate": 9.105006709422321e-05, "loss": 0.0543, "step": 46750 }, { "epoch": 1.5808512796240577, "grad_norm": 0.2571140229701996, "learning_rate": 9.104534648638645e-05, "loss": 0.0579, "step": 46760 }, { "epoch": 1.5811893573143108, "grad_norm": 0.24833433330059052, "learning_rate": 9.104062475638533e-05, "loss": 0.059, "step": 46770 }, { "epoch": 1.581527435004564, "grad_norm": 0.2942209839820862, "learning_rate": 9.103590190434896e-05, "loss": 0.0576, "step": 46780 }, { "epoch": 1.5818655126948173, "grad_norm": 0.3252277672290802, "learning_rate": 9.103117793040646e-05, "loss": 0.0593, "step": 46790 }, { "epoch": 1.5822035903850704, "grad_norm": 0.2644997537136078, "learning_rate": 9.102645283468694e-05, "loss": 0.057, "step": 46800 }, { "epoch": 1.5825416680753237, "grad_norm": 0.25603726506233215, "learning_rate": 9.102172661731963e-05, "loss": 0.0535, "step": 46810 }, { "epoch": 1.582879745765577, "grad_norm": 0.28524482250213623, "learning_rate": 9.101699927843372e-05, "loss": 0.0611, "step": 46820 }, { "epoch": 1.58321782345583, "grad_norm": 0.2844858467578888, "learning_rate": 9.101227081815847e-05, "loss": 0.0561, "step": 46830 }, { "epoch": 1.5835559011460834, "grad_norm": 0.43682241439819336, "learning_rate": 9.100754123662312e-05, "loss": 0.0559, "step": 46840 }, { "epoch": 1.5838939788363366, "grad_norm": 0.3181251287460327, "learning_rate": 9.100281053395702e-05, "loss": 0.0556, "step": 46850 }, { "epoch": 1.5842320565265897, "grad_norm": 0.25015977025032043, "learning_rate": 9.099807871028946e-05, "loss": 0.0582, "step": 46860 }, { "epoch": 1.584570134216843, "grad_norm": 0.2684362232685089, "learning_rate": 9.099334576574985e-05, "loss": 0.0596, "step": 46870 }, { "epoch": 1.5849082119070963, "grad_norm": 0.4428636431694031, "learning_rate": 9.098861170046757e-05, "loss": 0.0592, "step": 46880 }, { "epoch": 1.5852462895973494, "grad_norm": 0.2268752008676529, "learning_rate": 9.098387651457204e-05, "loss": 0.0554, "step": 46890 }, { "epoch": 1.5855843672876027, "grad_norm": 0.36294516921043396, "learning_rate": 9.097914020819272e-05, "loss": 0.0652, "step": 46900 }, { "epoch": 1.585922444977856, "grad_norm": 0.30381861329078674, "learning_rate": 9.09744027814591e-05, "loss": 0.0557, "step": 46910 }, { "epoch": 1.586260522668109, "grad_norm": 0.24823929369449615, "learning_rate": 9.096966423450071e-05, "loss": 0.056, "step": 46920 }, { "epoch": 1.5865986003583623, "grad_norm": 0.2407189905643463, "learning_rate": 9.096492456744709e-05, "loss": 0.0499, "step": 46930 }, { "epoch": 1.5869366780486156, "grad_norm": 0.4736238718032837, "learning_rate": 9.096018378042782e-05, "loss": 0.0585, "step": 46940 }, { "epoch": 1.5872747557388687, "grad_norm": 0.2577105462551117, "learning_rate": 9.095544187357252e-05, "loss": 0.0527, "step": 46950 }, { "epoch": 1.587612833429122, "grad_norm": 0.29464271664619446, "learning_rate": 9.095069884701082e-05, "loss": 0.054, "step": 46960 }, { "epoch": 1.5879509111193753, "grad_norm": 0.310400128364563, "learning_rate": 9.09459547008724e-05, "loss": 0.0529, "step": 46970 }, { "epoch": 1.5882889888096283, "grad_norm": 0.3236449658870697, "learning_rate": 9.094120943528694e-05, "loss": 0.0618, "step": 46980 }, { "epoch": 1.5886270664998818, "grad_norm": 0.3920484185218811, "learning_rate": 9.093646305038421e-05, "loss": 0.062, "step": 46990 }, { "epoch": 1.588965144190135, "grad_norm": 0.25386548042297363, "learning_rate": 9.093171554629396e-05, "loss": 0.0551, "step": 47000 }, { "epoch": 1.589303221880388, "grad_norm": 0.30812567472457886, "learning_rate": 9.092696692314599e-05, "loss": 0.0605, "step": 47010 }, { "epoch": 1.5896412995706415, "grad_norm": 0.33310896158218384, "learning_rate": 9.092221718107012e-05, "loss": 0.0598, "step": 47020 }, { "epoch": 1.5899793772608946, "grad_norm": 0.27457574009895325, "learning_rate": 9.09174663201962e-05, "loss": 0.0561, "step": 47030 }, { "epoch": 1.5903174549511476, "grad_norm": 0.24867862462997437, "learning_rate": 9.091271434065412e-05, "loss": 0.0538, "step": 47040 }, { "epoch": 1.5906555326414011, "grad_norm": 0.3468315899372101, "learning_rate": 9.090796124257381e-05, "loss": 0.0591, "step": 47050 }, { "epoch": 1.5909936103316542, "grad_norm": 0.38603296875953674, "learning_rate": 9.09032070260852e-05, "loss": 0.0571, "step": 47060 }, { "epoch": 1.5913316880219073, "grad_norm": 0.22063946723937988, "learning_rate": 9.089845169131828e-05, "loss": 0.0561, "step": 47070 }, { "epoch": 1.5916697657121608, "grad_norm": 0.38111183047294617, "learning_rate": 9.089369523840305e-05, "loss": 0.0593, "step": 47080 }, { "epoch": 1.5920078434024139, "grad_norm": 0.4110761880874634, "learning_rate": 9.088893766746956e-05, "loss": 0.0569, "step": 47090 }, { "epoch": 1.5923459210926671, "grad_norm": 0.3281371295452118, "learning_rate": 9.088417897864787e-05, "loss": 0.0562, "step": 47100 }, { "epoch": 1.5926839987829204, "grad_norm": 0.42832663655281067, "learning_rate": 9.08794191720681e-05, "loss": 0.0577, "step": 47110 }, { "epoch": 1.5930220764731735, "grad_norm": 0.20801393687725067, "learning_rate": 9.087465824786035e-05, "loss": 0.0537, "step": 47120 }, { "epoch": 1.5933601541634268, "grad_norm": 0.2309524565935135, "learning_rate": 9.086989620615482e-05, "loss": 0.0543, "step": 47130 }, { "epoch": 1.59369823185368, "grad_norm": 0.4034937918186188, "learning_rate": 9.086513304708166e-05, "loss": 0.0568, "step": 47140 }, { "epoch": 1.5940363095439332, "grad_norm": 0.27359965443611145, "learning_rate": 9.086036877077113e-05, "loss": 0.0577, "step": 47150 }, { "epoch": 1.5943743872341865, "grad_norm": 0.3968942165374756, "learning_rate": 9.085560337735346e-05, "loss": 0.0595, "step": 47160 }, { "epoch": 1.5947124649244397, "grad_norm": 0.32727035880088806, "learning_rate": 9.085083686695895e-05, "loss": 0.0625, "step": 47170 }, { "epoch": 1.5950505426146928, "grad_norm": 0.36252063512802124, "learning_rate": 9.08460692397179e-05, "loss": 0.0626, "step": 47180 }, { "epoch": 1.595388620304946, "grad_norm": 0.2118147313594818, "learning_rate": 9.084130049576067e-05, "loss": 0.0557, "step": 47190 }, { "epoch": 1.5957266979951994, "grad_norm": 0.2228725701570511, "learning_rate": 9.083653063521762e-05, "loss": 0.0569, "step": 47200 }, { "epoch": 1.5960647756854525, "grad_norm": 0.3908076286315918, "learning_rate": 9.083175965821917e-05, "loss": 0.0579, "step": 47210 }, { "epoch": 1.5964028533757058, "grad_norm": 0.2548825442790985, "learning_rate": 9.082698756489573e-05, "loss": 0.0571, "step": 47220 }, { "epoch": 1.596740931065959, "grad_norm": 0.3623880445957184, "learning_rate": 9.082221435537783e-05, "loss": 0.0562, "step": 47230 }, { "epoch": 1.5970790087562121, "grad_norm": 0.3013700842857361, "learning_rate": 9.081744002979588e-05, "loss": 0.0603, "step": 47240 }, { "epoch": 1.5974170864464654, "grad_norm": 0.38207629323005676, "learning_rate": 9.081266458828048e-05, "loss": 0.0561, "step": 47250 }, { "epoch": 1.5977551641367187, "grad_norm": 0.357036292552948, "learning_rate": 9.080788803096216e-05, "loss": 0.056, "step": 47260 }, { "epoch": 1.5980932418269718, "grad_norm": 0.33604079484939575, "learning_rate": 9.08031103579715e-05, "loss": 0.0549, "step": 47270 }, { "epoch": 1.598431319517225, "grad_norm": 0.3143388628959656, "learning_rate": 9.079833156943915e-05, "loss": 0.0585, "step": 47280 }, { "epoch": 1.5987693972074783, "grad_norm": 0.4992302656173706, "learning_rate": 9.079355166549575e-05, "loss": 0.0519, "step": 47290 }, { "epoch": 1.5991074748977314, "grad_norm": 0.4599037170410156, "learning_rate": 9.078877064627195e-05, "loss": 0.0554, "step": 47300 }, { "epoch": 1.5994455525879847, "grad_norm": 0.2693876326084137, "learning_rate": 9.078398851189849e-05, "loss": 0.0575, "step": 47310 }, { "epoch": 1.599783630278238, "grad_norm": 0.29410040378570557, "learning_rate": 9.077920526250611e-05, "loss": 0.0567, "step": 47320 }, { "epoch": 1.600121707968491, "grad_norm": 0.26828813552856445, "learning_rate": 9.077442089822557e-05, "loss": 0.0573, "step": 47330 }, { "epoch": 1.6004597856587444, "grad_norm": 0.3077784776687622, "learning_rate": 9.076963541918768e-05, "loss": 0.0522, "step": 47340 }, { "epoch": 1.6007978633489977, "grad_norm": 0.28324398398399353, "learning_rate": 9.076484882552327e-05, "loss": 0.0571, "step": 47350 }, { "epoch": 1.6011359410392507, "grad_norm": 0.34994834661483765, "learning_rate": 9.076006111736324e-05, "loss": 0.0568, "step": 47360 }, { "epoch": 1.601474018729504, "grad_norm": 0.2200196236371994, "learning_rate": 9.075527229483842e-05, "loss": 0.056, "step": 47370 }, { "epoch": 1.6018120964197573, "grad_norm": 0.1972540318965912, "learning_rate": 9.075048235807977e-05, "loss": 0.054, "step": 47380 }, { "epoch": 1.6021501741100104, "grad_norm": 0.2267405241727829, "learning_rate": 9.074569130721826e-05, "loss": 0.0578, "step": 47390 }, { "epoch": 1.6024882518002637, "grad_norm": 0.381113201379776, "learning_rate": 9.074089914238484e-05, "loss": 0.0584, "step": 47400 }, { "epoch": 1.602826329490517, "grad_norm": 0.48210635781288147, "learning_rate": 9.073610586371055e-05, "loss": 0.0544, "step": 47410 }, { "epoch": 1.60316440718077, "grad_norm": 0.20222370326519012, "learning_rate": 9.073131147132643e-05, "loss": 0.0509, "step": 47420 }, { "epoch": 1.6035024848710233, "grad_norm": 0.4149472713470459, "learning_rate": 9.072651596536355e-05, "loss": 0.0582, "step": 47430 }, { "epoch": 1.6038405625612766, "grad_norm": 0.2583896517753601, "learning_rate": 9.072171934595302e-05, "loss": 0.0561, "step": 47440 }, { "epoch": 1.6041786402515297, "grad_norm": 0.35969078540802, "learning_rate": 9.071692161322599e-05, "loss": 0.0628, "step": 47450 }, { "epoch": 1.6045167179417832, "grad_norm": 0.4093020260334015, "learning_rate": 9.071212276731362e-05, "loss": 0.0562, "step": 47460 }, { "epoch": 1.6048547956320363, "grad_norm": 0.3273966610431671, "learning_rate": 9.070732280834711e-05, "loss": 0.0604, "step": 47470 }, { "epoch": 1.6051928733222893, "grad_norm": 0.2571876347064972, "learning_rate": 9.070252173645768e-05, "loss": 0.0551, "step": 47480 }, { "epoch": 1.6055309510125428, "grad_norm": 0.2678815424442291, "learning_rate": 9.06977195517766e-05, "loss": 0.0586, "step": 47490 }, { "epoch": 1.605869028702796, "grad_norm": 0.3406779170036316, "learning_rate": 9.069291625443516e-05, "loss": 0.0551, "step": 47500 }, { "epoch": 1.606207106393049, "grad_norm": 0.2850443124771118, "learning_rate": 9.068811184456467e-05, "loss": 0.0583, "step": 47510 }, { "epoch": 1.6065451840833025, "grad_norm": 0.3076831102371216, "learning_rate": 9.06833063222965e-05, "loss": 0.0589, "step": 47520 }, { "epoch": 1.6068832617735556, "grad_norm": 0.3282538056373596, "learning_rate": 9.067849968776201e-05, "loss": 0.059, "step": 47530 }, { "epoch": 1.6072213394638086, "grad_norm": 0.3329192101955414, "learning_rate": 9.067369194109263e-05, "loss": 0.0638, "step": 47540 }, { "epoch": 1.6075594171540621, "grad_norm": 0.21818611025810242, "learning_rate": 9.066888308241979e-05, "loss": 0.0551, "step": 47550 }, { "epoch": 1.6078974948443152, "grad_norm": 0.24366168677806854, "learning_rate": 9.066407311187495e-05, "loss": 0.0579, "step": 47560 }, { "epoch": 1.6082355725345683, "grad_norm": 0.36084169149398804, "learning_rate": 9.065926202958964e-05, "loss": 0.0574, "step": 47570 }, { "epoch": 1.6085736502248218, "grad_norm": 0.3538455069065094, "learning_rate": 9.065444983569538e-05, "loss": 0.0575, "step": 47580 }, { "epoch": 1.6089117279150749, "grad_norm": 0.31587785482406616, "learning_rate": 9.064963653032375e-05, "loss": 0.0602, "step": 47590 }, { "epoch": 1.6092498056053282, "grad_norm": 0.3748188316822052, "learning_rate": 9.06448221136063e-05, "loss": 0.0551, "step": 47600 }, { "epoch": 1.6095878832955814, "grad_norm": 0.2659185230731964, "learning_rate": 9.06400065856747e-05, "loss": 0.0596, "step": 47610 }, { "epoch": 1.6099259609858345, "grad_norm": 0.44659197330474854, "learning_rate": 9.063518994666058e-05, "loss": 0.0588, "step": 47620 }, { "epoch": 1.6102640386760878, "grad_norm": 0.16818194091320038, "learning_rate": 9.063037219669563e-05, "loss": 0.0574, "step": 47630 }, { "epoch": 1.610602116366341, "grad_norm": 0.19883988797664642, "learning_rate": 9.062555333591159e-05, "loss": 0.0518, "step": 47640 }, { "epoch": 1.6109401940565942, "grad_norm": 0.3160466253757477, "learning_rate": 9.062073336444016e-05, "loss": 0.0562, "step": 47650 }, { "epoch": 1.6112782717468475, "grad_norm": 0.4370010495185852, "learning_rate": 9.061591228241314e-05, "loss": 0.0561, "step": 47660 }, { "epoch": 1.6116163494371007, "grad_norm": 0.2504914402961731, "learning_rate": 9.061109008996236e-05, "loss": 0.0656, "step": 47670 }, { "epoch": 1.6119544271273538, "grad_norm": 0.34339284896850586, "learning_rate": 9.060626678721961e-05, "loss": 0.0556, "step": 47680 }, { "epoch": 1.612292504817607, "grad_norm": 0.3791494369506836, "learning_rate": 9.06014423743168e-05, "loss": 0.0562, "step": 47690 }, { "epoch": 1.6126305825078604, "grad_norm": 0.663579523563385, "learning_rate": 9.05966168513858e-05, "loss": 0.0598, "step": 47700 }, { "epoch": 1.6129686601981135, "grad_norm": 0.28010880947113037, "learning_rate": 9.059179021855854e-05, "loss": 0.0629, "step": 47710 }, { "epoch": 1.6133067378883668, "grad_norm": 0.2444843202829361, "learning_rate": 9.058696247596701e-05, "loss": 0.0603, "step": 47720 }, { "epoch": 1.61364481557862, "grad_norm": 0.34668561816215515, "learning_rate": 9.058213362374315e-05, "loss": 0.0593, "step": 47730 }, { "epoch": 1.6139828932688731, "grad_norm": 0.2705889940261841, "learning_rate": 9.0577303662019e-05, "loss": 0.0616, "step": 47740 }, { "epoch": 1.6143209709591264, "grad_norm": 0.5190644860267639, "learning_rate": 9.057247259092662e-05, "loss": 0.0578, "step": 47750 }, { "epoch": 1.6146590486493797, "grad_norm": 0.19059684872627258, "learning_rate": 9.056764041059809e-05, "loss": 0.0557, "step": 47760 }, { "epoch": 1.6149971263396328, "grad_norm": 0.45304086804389954, "learning_rate": 9.056280712116551e-05, "loss": 0.058, "step": 47770 }, { "epoch": 1.615335204029886, "grad_norm": 0.2414131462574005, "learning_rate": 9.0557972722761e-05, "loss": 0.0526, "step": 47780 }, { "epoch": 1.6156732817201394, "grad_norm": 0.27504733204841614, "learning_rate": 9.055313721551677e-05, "loss": 0.0534, "step": 47790 }, { "epoch": 1.6160113594103924, "grad_norm": 0.21705962717533112, "learning_rate": 9.0548300599565e-05, "loss": 0.0527, "step": 47800 }, { "epoch": 1.6163494371006457, "grad_norm": 0.519312858581543, "learning_rate": 9.054346287503794e-05, "loss": 0.0603, "step": 47810 }, { "epoch": 1.616687514790899, "grad_norm": 0.35971248149871826, "learning_rate": 9.053862404206782e-05, "loss": 0.0604, "step": 47820 }, { "epoch": 1.617025592481152, "grad_norm": 0.4549260437488556, "learning_rate": 9.053378410078694e-05, "loss": 0.0612, "step": 47830 }, { "epoch": 1.6173636701714054, "grad_norm": 0.21816283464431763, "learning_rate": 9.052894305132765e-05, "loss": 0.0637, "step": 47840 }, { "epoch": 1.6177017478616587, "grad_norm": 0.30990561842918396, "learning_rate": 9.052410089382228e-05, "loss": 0.0471, "step": 47850 }, { "epoch": 1.6180398255519117, "grad_norm": 0.7657391428947449, "learning_rate": 9.051925762840321e-05, "loss": 0.0604, "step": 47860 }, { "epoch": 1.618377903242165, "grad_norm": 0.2560625672340393, "learning_rate": 9.051441325520287e-05, "loss": 0.0567, "step": 47870 }, { "epoch": 1.6187159809324183, "grad_norm": 0.3237692415714264, "learning_rate": 9.050956777435367e-05, "loss": 0.0563, "step": 47880 }, { "epoch": 1.6190540586226714, "grad_norm": 0.5888381600379944, "learning_rate": 9.050472118598811e-05, "loss": 0.0582, "step": 47890 }, { "epoch": 1.6193921363129247, "grad_norm": 0.3708884119987488, "learning_rate": 9.04998734902387e-05, "loss": 0.0581, "step": 47900 }, { "epoch": 1.619730214003178, "grad_norm": 0.28630417585372925, "learning_rate": 9.049502468723796e-05, "loss": 0.0584, "step": 47910 }, { "epoch": 1.620068291693431, "grad_norm": 0.24261708557605743, "learning_rate": 9.049017477711847e-05, "loss": 0.0533, "step": 47920 }, { "epoch": 1.6204063693836843, "grad_norm": 0.43137112259864807, "learning_rate": 9.04853237600128e-05, "loss": 0.0641, "step": 47930 }, { "epoch": 1.6207444470739376, "grad_norm": 0.9954841136932373, "learning_rate": 9.04804716360536e-05, "loss": 0.0618, "step": 47940 }, { "epoch": 1.6210825247641907, "grad_norm": 1.6738766431808472, "learning_rate": 9.04756184053735e-05, "loss": 0.0587, "step": 47950 }, { "epoch": 1.6214206024544442, "grad_norm": 0.8962857723236084, "learning_rate": 9.047076406810519e-05, "loss": 0.0561, "step": 47960 }, { "epoch": 1.6217586801446973, "grad_norm": 0.2149147242307663, "learning_rate": 9.046590862438142e-05, "loss": 0.0573, "step": 47970 }, { "epoch": 1.6220967578349503, "grad_norm": 0.45691508054733276, "learning_rate": 9.046105207433489e-05, "loss": 0.0587, "step": 47980 }, { "epoch": 1.6224348355252038, "grad_norm": 0.3023468852043152, "learning_rate": 9.04561944180984e-05, "loss": 0.0551, "step": 47990 }, { "epoch": 1.622772913215457, "grad_norm": 0.3906439542770386, "learning_rate": 9.045133565580476e-05, "loss": 0.0576, "step": 48000 }, { "epoch": 1.62311099090571, "grad_norm": 0.3336273431777954, "learning_rate": 9.04464757875868e-05, "loss": 0.0583, "step": 48010 }, { "epoch": 1.6234490685959635, "grad_norm": 0.2397904098033905, "learning_rate": 9.04416148135774e-05, "loss": 0.0686, "step": 48020 }, { "epoch": 1.6237871462862166, "grad_norm": 0.40302175283432007, "learning_rate": 9.043675273390942e-05, "loss": 0.0586, "step": 48030 }, { "epoch": 1.6241252239764696, "grad_norm": 0.4034986197948456, "learning_rate": 9.043188954871582e-05, "loss": 0.062, "step": 48040 }, { "epoch": 1.6244633016667231, "grad_norm": 0.4518544375896454, "learning_rate": 9.042702525812957e-05, "loss": 0.0564, "step": 48050 }, { "epoch": 1.6248013793569762, "grad_norm": 0.44763293862342834, "learning_rate": 9.042215986228359e-05, "loss": 0.0579, "step": 48060 }, { "epoch": 1.6251394570472295, "grad_norm": 0.30826428532600403, "learning_rate": 9.041729336131099e-05, "loss": 0.0629, "step": 48070 }, { "epoch": 1.6254775347374828, "grad_norm": 0.20646435022354126, "learning_rate": 9.041242575534474e-05, "loss": 0.0559, "step": 48080 }, { "epoch": 1.6258156124277359, "grad_norm": 0.31794828176498413, "learning_rate": 9.040755704451798e-05, "loss": 0.0544, "step": 48090 }, { "epoch": 1.6261536901179892, "grad_norm": 0.44122081995010376, "learning_rate": 9.040268722896378e-05, "loss": 0.0584, "step": 48100 }, { "epoch": 1.6264917678082425, "grad_norm": 0.43884676694869995, "learning_rate": 9.039781630881528e-05, "loss": 0.0509, "step": 48110 }, { "epoch": 1.6268298454984955, "grad_norm": 0.38784900307655334, "learning_rate": 9.039294428420568e-05, "loss": 0.0556, "step": 48120 }, { "epoch": 1.6271679231887488, "grad_norm": 0.31961262226104736, "learning_rate": 9.038807115526815e-05, "loss": 0.0531, "step": 48130 }, { "epoch": 1.627506000879002, "grad_norm": 0.2562243342399597, "learning_rate": 9.038319692213591e-05, "loss": 0.0597, "step": 48140 }, { "epoch": 1.6278440785692552, "grad_norm": 0.5645451545715332, "learning_rate": 9.037832158494226e-05, "loss": 0.0566, "step": 48150 }, { "epoch": 1.6281821562595085, "grad_norm": 0.2966507077217102, "learning_rate": 9.037344514382046e-05, "loss": 0.0545, "step": 48160 }, { "epoch": 1.6285202339497618, "grad_norm": 0.3382749557495117, "learning_rate": 9.036856759890383e-05, "loss": 0.0538, "step": 48170 }, { "epoch": 1.6288583116400148, "grad_norm": 0.2805957794189453, "learning_rate": 9.036368895032575e-05, "loss": 0.0608, "step": 48180 }, { "epoch": 1.6291963893302681, "grad_norm": 0.30613625049591064, "learning_rate": 9.035880919821956e-05, "loss": 0.0591, "step": 48190 }, { "epoch": 1.6295344670205214, "grad_norm": 0.4460150897502899, "learning_rate": 9.035392834271868e-05, "loss": 0.0615, "step": 48200 }, { "epoch": 1.6298725447107745, "grad_norm": 0.33815479278564453, "learning_rate": 9.034904638395656e-05, "loss": 0.0573, "step": 48210 }, { "epoch": 1.6302106224010278, "grad_norm": 0.24985960125923157, "learning_rate": 9.034416332206669e-05, "loss": 0.0565, "step": 48220 }, { "epoch": 1.630548700091281, "grad_norm": 0.31351426243782043, "learning_rate": 9.033927915718255e-05, "loss": 0.0597, "step": 48230 }, { "epoch": 1.6308867777815341, "grad_norm": 0.2774665653705597, "learning_rate": 9.033439388943765e-05, "loss": 0.0544, "step": 48240 }, { "epoch": 1.6312248554717874, "grad_norm": 0.3079177439212799, "learning_rate": 9.032950751896558e-05, "loss": 0.0586, "step": 48250 }, { "epoch": 1.6315629331620407, "grad_norm": 0.31059530377388, "learning_rate": 9.032462004589994e-05, "loss": 0.0583, "step": 48260 }, { "epoch": 1.6319010108522938, "grad_norm": 0.5296486616134644, "learning_rate": 9.031973147037433e-05, "loss": 0.0662, "step": 48270 }, { "epoch": 1.632239088542547, "grad_norm": 0.4071795344352722, "learning_rate": 9.031484179252241e-05, "loss": 0.0582, "step": 48280 }, { "epoch": 1.6325771662328004, "grad_norm": 0.4212479591369629, "learning_rate": 9.030995101247785e-05, "loss": 0.068, "step": 48290 }, { "epoch": 1.6329152439230534, "grad_norm": 0.2687501311302185, "learning_rate": 9.03050591303744e-05, "loss": 0.0615, "step": 48300 }, { "epoch": 1.6332533216133067, "grad_norm": 0.3375544548034668, "learning_rate": 9.030016614634576e-05, "loss": 0.0558, "step": 48310 }, { "epoch": 1.63359139930356, "grad_norm": 0.24706394970417023, "learning_rate": 9.029527206052571e-05, "loss": 0.0543, "step": 48320 }, { "epoch": 1.633929476993813, "grad_norm": 0.6417818069458008, "learning_rate": 9.029037687304806e-05, "loss": 0.0538, "step": 48330 }, { "epoch": 1.6342675546840664, "grad_norm": 0.3245295584201813, "learning_rate": 9.028548058404666e-05, "loss": 0.0576, "step": 48340 }, { "epoch": 1.6346056323743197, "grad_norm": 0.435421884059906, "learning_rate": 9.028058319365534e-05, "loss": 0.0577, "step": 48350 }, { "epoch": 1.6349437100645727, "grad_norm": 0.5313596129417419, "learning_rate": 9.027568470200801e-05, "loss": 0.0591, "step": 48360 }, { "epoch": 1.635281787754826, "grad_norm": 0.3148859143257141, "learning_rate": 9.02707851092386e-05, "loss": 0.0578, "step": 48370 }, { "epoch": 1.6356198654450793, "grad_norm": 0.2844279408454895, "learning_rate": 9.026588441548105e-05, "loss": 0.0648, "step": 48380 }, { "epoch": 1.6359579431353324, "grad_norm": 0.3752603530883789, "learning_rate": 9.026098262086934e-05, "loss": 0.0559, "step": 48390 }, { "epoch": 1.6362960208255857, "grad_norm": 0.672545850276947, "learning_rate": 9.025607972553749e-05, "loss": 0.0572, "step": 48400 }, { "epoch": 1.636634098515839, "grad_norm": 0.2980428636074066, "learning_rate": 9.025117572961956e-05, "loss": 0.0559, "step": 48410 }, { "epoch": 1.636972176206092, "grad_norm": 0.5716789960861206, "learning_rate": 9.024627063324959e-05, "loss": 0.0617, "step": 48420 }, { "epoch": 1.6373102538963455, "grad_norm": 0.46525466442108154, "learning_rate": 9.024136443656172e-05, "loss": 0.0552, "step": 48430 }, { "epoch": 1.6376483315865986, "grad_norm": 0.4942668378353119, "learning_rate": 9.023645713969004e-05, "loss": 0.0574, "step": 48440 }, { "epoch": 1.6379864092768517, "grad_norm": 0.33813315629959106, "learning_rate": 9.023154874276875e-05, "loss": 0.0527, "step": 48450 }, { "epoch": 1.6383244869671052, "grad_norm": 0.24941520392894745, "learning_rate": 9.022663924593204e-05, "loss": 0.0552, "step": 48460 }, { "epoch": 1.6386625646573583, "grad_norm": 0.29824867844581604, "learning_rate": 9.02217286493141e-05, "loss": 0.0538, "step": 48470 }, { "epoch": 1.6390006423476113, "grad_norm": 0.20940278470516205, "learning_rate": 9.021681695304924e-05, "loss": 0.0569, "step": 48480 }, { "epoch": 1.6393387200378648, "grad_norm": 0.22087650001049042, "learning_rate": 9.02119041572717e-05, "loss": 0.0574, "step": 48490 }, { "epoch": 1.639676797728118, "grad_norm": 0.3405373692512512, "learning_rate": 9.020699026211581e-05, "loss": 0.0547, "step": 48500 }, { "epoch": 1.640014875418371, "grad_norm": 0.5010370016098022, "learning_rate": 9.02020752677159e-05, "loss": 0.055, "step": 48510 }, { "epoch": 1.6403529531086245, "grad_norm": 0.44102367758750916, "learning_rate": 9.019715917420637e-05, "loss": 0.0556, "step": 48520 }, { "epoch": 1.6406910307988776, "grad_norm": 0.25662514567375183, "learning_rate": 9.01922419817216e-05, "loss": 0.0527, "step": 48530 }, { "epoch": 1.6410291084891306, "grad_norm": 0.25474318861961365, "learning_rate": 9.018732369039604e-05, "loss": 0.059, "step": 48540 }, { "epoch": 1.6413671861793842, "grad_norm": 0.2607620060443878, "learning_rate": 9.018240430036413e-05, "loss": 0.0599, "step": 48550 }, { "epoch": 1.6417052638696372, "grad_norm": 0.8120848536491394, "learning_rate": 9.017748381176042e-05, "loss": 0.0597, "step": 48560 }, { "epoch": 1.6420433415598905, "grad_norm": 0.41252365708351135, "learning_rate": 9.017256222471937e-05, "loss": 0.0583, "step": 48570 }, { "epoch": 1.6423814192501438, "grad_norm": 0.5030207633972168, "learning_rate": 9.016763953937555e-05, "loss": 0.0599, "step": 48580 }, { "epoch": 1.6427194969403969, "grad_norm": 0.2713949680328369, "learning_rate": 9.016271575586356e-05, "loss": 0.0503, "step": 48590 }, { "epoch": 1.6430575746306502, "grad_norm": 0.686367928981781, "learning_rate": 9.015779087431803e-05, "loss": 0.0592, "step": 48600 }, { "epoch": 1.6433956523209035, "grad_norm": 0.1946505457162857, "learning_rate": 9.015286489487358e-05, "loss": 0.0547, "step": 48610 }, { "epoch": 1.6437337300111565, "grad_norm": 0.22461764514446259, "learning_rate": 9.014793781766489e-05, "loss": 0.0545, "step": 48620 }, { "epoch": 1.6440718077014098, "grad_norm": 0.23212353885173798, "learning_rate": 9.014300964282665e-05, "loss": 0.0563, "step": 48630 }, { "epoch": 1.644409885391663, "grad_norm": 0.24732713401317596, "learning_rate": 9.013808037049361e-05, "loss": 0.0584, "step": 48640 }, { "epoch": 1.6447479630819162, "grad_norm": 0.3232957124710083, "learning_rate": 9.013315000080053e-05, "loss": 0.0562, "step": 48650 }, { "epoch": 1.6450860407721695, "grad_norm": 0.2742669880390167, "learning_rate": 9.012821853388221e-05, "loss": 0.0526, "step": 48660 }, { "epoch": 1.6454241184624228, "grad_norm": 0.25652217864990234, "learning_rate": 9.012328596987348e-05, "loss": 0.0563, "step": 48670 }, { "epoch": 1.6457621961526758, "grad_norm": 0.3026528060436249, "learning_rate": 9.011835230890917e-05, "loss": 0.0558, "step": 48680 }, { "epoch": 1.6461002738429291, "grad_norm": 0.22108888626098633, "learning_rate": 9.011341755112419e-05, "loss": 0.0623, "step": 48690 }, { "epoch": 1.6464383515331824, "grad_norm": 0.2493453323841095, "learning_rate": 9.010848169665344e-05, "loss": 0.061, "step": 48700 }, { "epoch": 1.6467764292234355, "grad_norm": 0.17219984531402588, "learning_rate": 9.010354474563188e-05, "loss": 0.0583, "step": 48710 }, { "epoch": 1.6471145069136888, "grad_norm": 0.3379688858985901, "learning_rate": 9.009860669819445e-05, "loss": 0.0593, "step": 48720 }, { "epoch": 1.647452584603942, "grad_norm": 0.31716564297676086, "learning_rate": 9.00936675544762e-05, "loss": 0.0603, "step": 48730 }, { "epoch": 1.6477906622941951, "grad_norm": 0.2036120742559433, "learning_rate": 9.008872731461211e-05, "loss": 0.0576, "step": 48740 }, { "epoch": 1.6481287399844484, "grad_norm": 0.35168424248695374, "learning_rate": 9.00837859787373e-05, "loss": 0.0515, "step": 48750 }, { "epoch": 1.6484668176747017, "grad_norm": 0.22232069075107574, "learning_rate": 9.007884354698683e-05, "loss": 0.0556, "step": 48760 }, { "epoch": 1.6488048953649548, "grad_norm": 0.5872870087623596, "learning_rate": 9.007390001949585e-05, "loss": 0.0556, "step": 48770 }, { "epoch": 1.649142973055208, "grad_norm": 0.2930389642715454, "learning_rate": 9.006895539639948e-05, "loss": 0.0562, "step": 48780 }, { "epoch": 1.6494810507454614, "grad_norm": 0.27053606510162354, "learning_rate": 9.006400967783294e-05, "loss": 0.0533, "step": 48790 }, { "epoch": 1.6498191284357144, "grad_norm": 0.2635691463947296, "learning_rate": 9.005906286393143e-05, "loss": 0.0638, "step": 48800 }, { "epoch": 1.6501572061259677, "grad_norm": 0.2813931405544281, "learning_rate": 9.005411495483016e-05, "loss": 0.0597, "step": 48810 }, { "epoch": 1.650495283816221, "grad_norm": 0.37145286798477173, "learning_rate": 9.004916595066446e-05, "loss": 0.0565, "step": 48820 }, { "epoch": 1.650833361506474, "grad_norm": 0.2710428237915039, "learning_rate": 9.004421585156962e-05, "loss": 0.0639, "step": 48830 }, { "epoch": 1.6511714391967274, "grad_norm": 0.2777254283428192, "learning_rate": 9.003926465768094e-05, "loss": 0.0583, "step": 48840 }, { "epoch": 1.6515095168869807, "grad_norm": 0.34091508388519287, "learning_rate": 9.00343123691338e-05, "loss": 0.0612, "step": 48850 }, { "epoch": 1.6518475945772337, "grad_norm": 0.9350189566612244, "learning_rate": 9.002935898606361e-05, "loss": 0.0566, "step": 48860 }, { "epoch": 1.652185672267487, "grad_norm": 0.40757042169570923, "learning_rate": 9.002440450860579e-05, "loss": 0.0523, "step": 48870 }, { "epoch": 1.6525237499577403, "grad_norm": 0.21934378147125244, "learning_rate": 9.001944893689578e-05, "loss": 0.0603, "step": 48880 }, { "epoch": 1.6528618276479934, "grad_norm": 0.3383789658546448, "learning_rate": 9.001449227106908e-05, "loss": 0.0517, "step": 48890 }, { "epoch": 1.6531999053382467, "grad_norm": 0.27293750643730164, "learning_rate": 9.000953451126118e-05, "loss": 0.0527, "step": 48900 }, { "epoch": 1.6535379830285, "grad_norm": 0.22732046246528625, "learning_rate": 9.000457565760764e-05, "loss": 0.0551, "step": 48910 }, { "epoch": 1.653876060718753, "grad_norm": 0.282499223947525, "learning_rate": 8.999961571024405e-05, "loss": 0.0626, "step": 48920 }, { "epoch": 1.6542141384090066, "grad_norm": 0.4105234146118164, "learning_rate": 8.999465466930597e-05, "loss": 0.0605, "step": 48930 }, { "epoch": 1.6545522160992596, "grad_norm": 0.3311312198638916, "learning_rate": 8.998969253492906e-05, "loss": 0.056, "step": 48940 }, { "epoch": 1.6548902937895127, "grad_norm": 0.2528083622455597, "learning_rate": 8.998472930724899e-05, "loss": 0.0604, "step": 48950 }, { "epoch": 1.6552283714797662, "grad_norm": 0.4725129306316376, "learning_rate": 8.997976498640144e-05, "loss": 0.0576, "step": 48960 }, { "epoch": 1.6555664491700193, "grad_norm": 0.17974533140659332, "learning_rate": 8.997479957252213e-05, "loss": 0.0609, "step": 48970 }, { "epoch": 1.6559045268602723, "grad_norm": 1.2710870504379272, "learning_rate": 8.996983306574682e-05, "loss": 0.0585, "step": 48980 }, { "epoch": 1.6562426045505259, "grad_norm": 0.23713919520378113, "learning_rate": 8.996486546621129e-05, "loss": 0.0577, "step": 48990 }, { "epoch": 1.656580682240779, "grad_norm": 0.585432767868042, "learning_rate": 8.995989677405135e-05, "loss": 0.053, "step": 49000 }, { "epoch": 1.656918759931032, "grad_norm": 0.29408884048461914, "learning_rate": 8.995492698940284e-05, "loss": 0.0583, "step": 49010 }, { "epoch": 1.6572568376212855, "grad_norm": 0.3104366362094879, "learning_rate": 8.994995611240164e-05, "loss": 0.0642, "step": 49020 }, { "epoch": 1.6575949153115386, "grad_norm": 0.26537254452705383, "learning_rate": 8.994498414318366e-05, "loss": 0.0581, "step": 49030 }, { "epoch": 1.6579329930017919, "grad_norm": 0.3064890205860138, "learning_rate": 8.994001108188481e-05, "loss": 0.0582, "step": 49040 }, { "epoch": 1.6582710706920452, "grad_norm": 0.25726044178009033, "learning_rate": 8.993503692864107e-05, "loss": 0.0549, "step": 49050 }, { "epoch": 1.6586091483822982, "grad_norm": 0.26543891429901123, "learning_rate": 8.993006168358841e-05, "loss": 0.0622, "step": 49060 }, { "epoch": 1.6589472260725515, "grad_norm": 1.3873568773269653, "learning_rate": 8.992508534686287e-05, "loss": 0.066, "step": 49070 }, { "epoch": 1.6592853037628048, "grad_norm": 0.27388790249824524, "learning_rate": 8.992010791860051e-05, "loss": 0.0591, "step": 49080 }, { "epoch": 1.6596233814530579, "grad_norm": 0.3005227744579315, "learning_rate": 8.991512939893737e-05, "loss": 0.0538, "step": 49090 }, { "epoch": 1.6599614591433112, "grad_norm": 0.35263293981552124, "learning_rate": 8.99101497880096e-05, "loss": 0.0595, "step": 49100 }, { "epoch": 1.6602995368335645, "grad_norm": 0.2840006947517395, "learning_rate": 8.990516908595333e-05, "loss": 0.0572, "step": 49110 }, { "epoch": 1.6606376145238175, "grad_norm": 0.30290794372558594, "learning_rate": 8.990018729290473e-05, "loss": 0.0548, "step": 49120 }, { "epoch": 1.6609756922140708, "grad_norm": 0.4004441797733307, "learning_rate": 8.9895204409e-05, "loss": 0.0575, "step": 49130 }, { "epoch": 1.6613137699043241, "grad_norm": 0.3809095025062561, "learning_rate": 8.989022043437536e-05, "loss": 0.0596, "step": 49140 }, { "epoch": 1.6616518475945772, "grad_norm": 0.3867391049861908, "learning_rate": 8.988523536916709e-05, "loss": 0.0555, "step": 49150 }, { "epoch": 1.6619899252848305, "grad_norm": 0.2766764163970947, "learning_rate": 8.988024921351147e-05, "loss": 0.0559, "step": 49160 }, { "epoch": 1.6623280029750838, "grad_norm": 0.3281806409358978, "learning_rate": 8.987526196754481e-05, "loss": 0.0598, "step": 49170 }, { "epoch": 1.6626660806653368, "grad_norm": 0.4682134687900543, "learning_rate": 8.987027363140348e-05, "loss": 0.0609, "step": 49180 }, { "epoch": 1.6630041583555901, "grad_norm": 0.6572383046150208, "learning_rate": 8.986528420522384e-05, "loss": 0.0603, "step": 49190 }, { "epoch": 1.6633422360458434, "grad_norm": 0.45699527859687805, "learning_rate": 8.986029368914231e-05, "loss": 0.0652, "step": 49200 }, { "epoch": 1.6636803137360965, "grad_norm": 0.7959378957748413, "learning_rate": 8.985530208329531e-05, "loss": 0.0628, "step": 49210 }, { "epoch": 1.6640183914263498, "grad_norm": 0.25168919563293457, "learning_rate": 8.985030938781935e-05, "loss": 0.0533, "step": 49220 }, { "epoch": 1.664356469116603, "grad_norm": 0.31790927052497864, "learning_rate": 8.984531560285088e-05, "loss": 0.0526, "step": 49230 }, { "epoch": 1.6646945468068561, "grad_norm": 0.3908686935901642, "learning_rate": 8.984032072852647e-05, "loss": 0.0639, "step": 49240 }, { "epoch": 1.6650326244971094, "grad_norm": 0.28377601504325867, "learning_rate": 8.983532476498264e-05, "loss": 0.0587, "step": 49250 }, { "epoch": 1.6653707021873627, "grad_norm": 0.394437313079834, "learning_rate": 8.9830327712356e-05, "loss": 0.0537, "step": 49260 }, { "epoch": 1.6657087798776158, "grad_norm": 0.6274127960205078, "learning_rate": 8.982532957078317e-05, "loss": 0.0603, "step": 49270 }, { "epoch": 1.666046857567869, "grad_norm": 0.42221519351005554, "learning_rate": 8.982033034040078e-05, "loss": 0.0565, "step": 49280 }, { "epoch": 1.6663849352581224, "grad_norm": 0.4312894344329834, "learning_rate": 8.981533002134553e-05, "loss": 0.0586, "step": 49290 }, { "epoch": 1.6667230129483754, "grad_norm": 0.39185377955436707, "learning_rate": 8.981032861375411e-05, "loss": 0.0588, "step": 49300 }, { "epoch": 1.6670610906386287, "grad_norm": 0.43158483505249023, "learning_rate": 8.980532611776327e-05, "loss": 0.0593, "step": 49310 }, { "epoch": 1.667399168328882, "grad_norm": 0.28587502241134644, "learning_rate": 8.980032253350976e-05, "loss": 0.0555, "step": 49320 }, { "epoch": 1.667737246019135, "grad_norm": 0.4047536849975586, "learning_rate": 8.979531786113039e-05, "loss": 0.0605, "step": 49330 }, { "epoch": 1.6680753237093884, "grad_norm": 0.2448943853378296, "learning_rate": 8.979031210076199e-05, "loss": 0.0573, "step": 49340 }, { "epoch": 1.6684134013996417, "grad_norm": 0.38889947533607483, "learning_rate": 8.978530525254138e-05, "loss": 0.0614, "step": 49350 }, { "epoch": 1.6687514790898947, "grad_norm": 0.289960116147995, "learning_rate": 8.978029731660548e-05, "loss": 0.0587, "step": 49360 }, { "epoch": 1.669089556780148, "grad_norm": 0.2747195065021515, "learning_rate": 8.977528829309121e-05, "loss": 0.0595, "step": 49370 }, { "epoch": 1.6694276344704013, "grad_norm": 0.2501336932182312, "learning_rate": 8.977027818213548e-05, "loss": 0.0602, "step": 49380 }, { "epoch": 1.6697657121606544, "grad_norm": 0.42404240369796753, "learning_rate": 8.976526698387531e-05, "loss": 0.0604, "step": 49390 }, { "epoch": 1.670103789850908, "grad_norm": 0.3346986472606659, "learning_rate": 8.976025469844766e-05, "loss": 0.0546, "step": 49400 }, { "epoch": 1.670441867541161, "grad_norm": 0.22659799456596375, "learning_rate": 8.975524132598958e-05, "loss": 0.06, "step": 49410 }, { "epoch": 1.670779945231414, "grad_norm": 0.24729853868484497, "learning_rate": 8.975022686663816e-05, "loss": 0.0592, "step": 49420 }, { "epoch": 1.6711180229216676, "grad_norm": 0.30667993426322937, "learning_rate": 8.974521132053045e-05, "loss": 0.0564, "step": 49430 }, { "epoch": 1.6714561006119206, "grad_norm": 0.38234859704971313, "learning_rate": 8.97401946878036e-05, "loss": 0.0555, "step": 49440 }, { "epoch": 1.6717941783021737, "grad_norm": 0.48064520955085754, "learning_rate": 8.973517696859476e-05, "loss": 0.0516, "step": 49450 }, { "epoch": 1.6721322559924272, "grad_norm": 0.443403035402298, "learning_rate": 8.97301581630411e-05, "loss": 0.0581, "step": 49460 }, { "epoch": 1.6724703336826803, "grad_norm": 0.37315627932548523, "learning_rate": 8.972513827127987e-05, "loss": 0.0611, "step": 49470 }, { "epoch": 1.6728084113729333, "grad_norm": 0.32263636589050293, "learning_rate": 8.972011729344824e-05, "loss": 0.0618, "step": 49480 }, { "epoch": 1.6731464890631869, "grad_norm": 0.31104543805122375, "learning_rate": 8.971509522968356e-05, "loss": 0.0573, "step": 49490 }, { "epoch": 1.67348456675344, "grad_norm": 0.42119377851486206, "learning_rate": 8.971007208012309e-05, "loss": 0.0603, "step": 49500 }, { "epoch": 1.673822644443693, "grad_norm": 0.28528547286987305, "learning_rate": 8.970504784490417e-05, "loss": 0.0562, "step": 49510 }, { "epoch": 1.6741607221339465, "grad_norm": 0.29786059260368347, "learning_rate": 8.970002252416415e-05, "loss": 0.0573, "step": 49520 }, { "epoch": 1.6744987998241996, "grad_norm": 0.4778680205345154, "learning_rate": 8.969499611804041e-05, "loss": 0.0568, "step": 49530 }, { "epoch": 1.6748368775144529, "grad_norm": 0.5544523596763611, "learning_rate": 8.968996862667042e-05, "loss": 0.0605, "step": 49540 }, { "epoch": 1.6751749552047062, "grad_norm": 0.3615993559360504, "learning_rate": 8.96849400501916e-05, "loss": 0.0639, "step": 49550 }, { "epoch": 1.6755130328949592, "grad_norm": 0.2627018392086029, "learning_rate": 8.967991038874142e-05, "loss": 0.0584, "step": 49560 }, { "epoch": 1.6758511105852125, "grad_norm": 0.3823406398296356, "learning_rate": 8.967487964245738e-05, "loss": 0.0565, "step": 49570 }, { "epoch": 1.6761891882754658, "grad_norm": 0.3334513306617737, "learning_rate": 8.966984781147705e-05, "loss": 0.054, "step": 49580 }, { "epoch": 1.6765272659657189, "grad_norm": 0.5538878440856934, "learning_rate": 8.966481489593799e-05, "loss": 0.0634, "step": 49590 }, { "epoch": 1.6768653436559722, "grad_norm": 0.33938243985176086, "learning_rate": 8.965978089597778e-05, "loss": 0.0567, "step": 49600 }, { "epoch": 1.6772034213462255, "grad_norm": 0.323199987411499, "learning_rate": 8.965474581173406e-05, "loss": 0.0596, "step": 49610 }, { "epoch": 1.6775414990364785, "grad_norm": 0.7308275103569031, "learning_rate": 8.964970964334449e-05, "loss": 0.0614, "step": 49620 }, { "epoch": 1.6778795767267318, "grad_norm": 0.26201367378234863, "learning_rate": 8.964467239094677e-05, "loss": 0.0626, "step": 49630 }, { "epoch": 1.6782176544169851, "grad_norm": 0.24283406138420105, "learning_rate": 8.963963405467858e-05, "loss": 0.0626, "step": 49640 }, { "epoch": 1.6785557321072382, "grad_norm": 0.2535593807697296, "learning_rate": 8.96345946346777e-05, "loss": 0.0566, "step": 49650 }, { "epoch": 1.6788938097974915, "grad_norm": 0.23897528648376465, "learning_rate": 8.962955413108187e-05, "loss": 0.0583, "step": 49660 }, { "epoch": 1.6792318874877448, "grad_norm": 0.2711528241634369, "learning_rate": 8.962451254402896e-05, "loss": 0.054, "step": 49670 }, { "epoch": 1.6795699651779978, "grad_norm": 0.27247971296310425, "learning_rate": 8.961946987365673e-05, "loss": 0.0551, "step": 49680 }, { "epoch": 1.6799080428682511, "grad_norm": 0.3320657014846802, "learning_rate": 8.961442612010308e-05, "loss": 0.0543, "step": 49690 }, { "epoch": 1.6802461205585044, "grad_norm": 0.39567193388938904, "learning_rate": 8.960938128350591e-05, "loss": 0.0597, "step": 49700 }, { "epoch": 1.6805841982487575, "grad_norm": 0.3288290500640869, "learning_rate": 8.960433536400313e-05, "loss": 0.0607, "step": 49710 }, { "epoch": 1.6809222759390108, "grad_norm": 0.4207441508769989, "learning_rate": 8.95992883617327e-05, "loss": 0.0599, "step": 49720 }, { "epoch": 1.681260353629264, "grad_norm": 0.40779033303260803, "learning_rate": 8.95942402768326e-05, "loss": 0.0627, "step": 49730 }, { "epoch": 1.6815984313195171, "grad_norm": 0.3111536204814911, "learning_rate": 8.958919110944087e-05, "loss": 0.0541, "step": 49740 }, { "epoch": 1.6819365090097704, "grad_norm": 0.39654842019081116, "learning_rate": 8.95841408596955e-05, "loss": 0.0588, "step": 49750 }, { "epoch": 1.6822745867000237, "grad_norm": 0.8377491235733032, "learning_rate": 8.957908952773461e-05, "loss": 0.0525, "step": 49760 }, { "epoch": 1.6826126643902768, "grad_norm": 0.5989631414413452, "learning_rate": 8.957403711369628e-05, "loss": 0.0587, "step": 49770 }, { "epoch": 1.68295074208053, "grad_norm": 0.316988468170166, "learning_rate": 8.956898361771863e-05, "loss": 0.0638, "step": 49780 }, { "epoch": 1.6832888197707834, "grad_norm": 0.3285139501094818, "learning_rate": 8.956392903993984e-05, "loss": 0.0532, "step": 49790 }, { "epoch": 1.6836268974610364, "grad_norm": 0.2745118737220764, "learning_rate": 8.955887338049809e-05, "loss": 0.0618, "step": 49800 }, { "epoch": 1.6839649751512897, "grad_norm": 0.3654307723045349, "learning_rate": 8.955381663953159e-05, "loss": 0.0605, "step": 49810 }, { "epoch": 1.684303052841543, "grad_norm": 0.6204471588134766, "learning_rate": 8.954875881717864e-05, "loss": 0.0576, "step": 49820 }, { "epoch": 1.684641130531796, "grad_norm": 0.37375494837760925, "learning_rate": 8.954369991357745e-05, "loss": 0.0573, "step": 49830 }, { "epoch": 1.6849792082220494, "grad_norm": 0.2410885989665985, "learning_rate": 8.953863992886638e-05, "loss": 0.0641, "step": 49840 }, { "epoch": 1.6853172859123027, "grad_norm": 0.4657418429851532, "learning_rate": 8.953357886318372e-05, "loss": 0.0605, "step": 49850 }, { "epoch": 1.6856553636025557, "grad_norm": 0.4260118305683136, "learning_rate": 8.952851671666788e-05, "loss": 0.0611, "step": 49860 }, { "epoch": 1.685993441292809, "grad_norm": 0.3193836808204651, "learning_rate": 8.952345348945725e-05, "loss": 0.0585, "step": 49870 }, { "epoch": 1.6863315189830623, "grad_norm": 0.37957093119621277, "learning_rate": 8.951838918169027e-05, "loss": 0.0607, "step": 49880 }, { "epoch": 1.6866695966733154, "grad_norm": 0.3526325225830078, "learning_rate": 8.951332379350535e-05, "loss": 0.0582, "step": 49890 }, { "epoch": 1.687007674363569, "grad_norm": 0.48011645674705505, "learning_rate": 8.9508257325041e-05, "loss": 0.0543, "step": 49900 }, { "epoch": 1.687345752053822, "grad_norm": 0.4563722312450409, "learning_rate": 8.950318977643574e-05, "loss": 0.063, "step": 49910 }, { "epoch": 1.687683829744075, "grad_norm": 0.2968568503856659, "learning_rate": 8.949812114782811e-05, "loss": 0.0549, "step": 49920 }, { "epoch": 1.6880219074343286, "grad_norm": 0.48532193899154663, "learning_rate": 8.949305143935669e-05, "loss": 0.0611, "step": 49930 }, { "epoch": 1.6883599851245816, "grad_norm": 0.3566323220729828, "learning_rate": 8.948798065116008e-05, "loss": 0.0641, "step": 49940 }, { "epoch": 1.6886980628148347, "grad_norm": 0.2675672471523285, "learning_rate": 8.948290878337692e-05, "loss": 0.0629, "step": 49950 }, { "epoch": 1.6890361405050882, "grad_norm": 0.31898820400238037, "learning_rate": 8.947783583614586e-05, "loss": 0.0596, "step": 49960 }, { "epoch": 1.6893742181953413, "grad_norm": 0.3198336660861969, "learning_rate": 8.947276180960559e-05, "loss": 0.0596, "step": 49970 }, { "epoch": 1.6897122958855944, "grad_norm": 0.29253697395324707, "learning_rate": 8.946768670389487e-05, "loss": 0.0553, "step": 49980 }, { "epoch": 1.6900503735758479, "grad_norm": 0.6374375820159912, "learning_rate": 8.94626105191524e-05, "loss": 0.0602, "step": 49990 }, { "epoch": 1.690388451266101, "grad_norm": 0.4131196141242981, "learning_rate": 8.9457533255517e-05, "loss": 0.0571, "step": 50000 }, { "epoch": 1.6907265289563542, "grad_norm": 0.2554936110973358, "learning_rate": 8.945245491312744e-05, "loss": 0.0579, "step": 50010 }, { "epoch": 1.6910646066466075, "grad_norm": 0.3297644257545471, "learning_rate": 8.944737549212261e-05, "loss": 0.0609, "step": 50020 }, { "epoch": 1.6914026843368606, "grad_norm": 0.3926493525505066, "learning_rate": 8.944229499264135e-05, "loss": 0.0535, "step": 50030 }, { "epoch": 1.6917407620271139, "grad_norm": 0.2878797650337219, "learning_rate": 8.943721341482253e-05, "loss": 0.0575, "step": 50040 }, { "epoch": 1.6920788397173672, "grad_norm": 0.35920169949531555, "learning_rate": 8.943213075880514e-05, "loss": 0.0545, "step": 50050 }, { "epoch": 1.6924169174076202, "grad_norm": 0.4388181269168854, "learning_rate": 8.942704702472812e-05, "loss": 0.0632, "step": 50060 }, { "epoch": 1.6927549950978735, "grad_norm": 0.48028311133384705, "learning_rate": 8.942196221273043e-05, "loss": 0.0627, "step": 50070 }, { "epoch": 1.6930930727881268, "grad_norm": 0.3490980267524719, "learning_rate": 8.94168763229511e-05, "loss": 0.0607, "step": 50080 }, { "epoch": 1.6934311504783799, "grad_norm": 0.3630174398422241, "learning_rate": 8.941178935552918e-05, "loss": 0.0619, "step": 50090 }, { "epoch": 1.6937692281686332, "grad_norm": 0.3793737590312958, "learning_rate": 8.940670131060375e-05, "loss": 0.0535, "step": 50100 }, { "epoch": 1.6941073058588865, "grad_norm": 0.5227362513542175, "learning_rate": 8.940161218831391e-05, "loss": 0.058, "step": 50110 }, { "epoch": 1.6944453835491395, "grad_norm": 0.2636306583881378, "learning_rate": 8.939652198879879e-05, "loss": 0.0611, "step": 50120 }, { "epoch": 1.6947834612393928, "grad_norm": 0.4054105281829834, "learning_rate": 8.939143071219758e-05, "loss": 0.0571, "step": 50130 }, { "epoch": 1.6951215389296461, "grad_norm": 0.35133519768714905, "learning_rate": 8.938633835864944e-05, "loss": 0.0634, "step": 50140 }, { "epoch": 1.6954596166198992, "grad_norm": 0.6816269755363464, "learning_rate": 8.93812449282936e-05, "loss": 0.0612, "step": 50150 }, { "epoch": 1.6957976943101525, "grad_norm": 0.3922114670276642, "learning_rate": 8.937615042126932e-05, "loss": 0.0527, "step": 50160 }, { "epoch": 1.6961357720004058, "grad_norm": 0.4716246724128723, "learning_rate": 8.937105483771588e-05, "loss": 0.0567, "step": 50170 }, { "epoch": 1.6964738496906588, "grad_norm": 0.3136149048805237, "learning_rate": 8.936595817777261e-05, "loss": 0.0618, "step": 50180 }, { "epoch": 1.6968119273809121, "grad_norm": 0.2449309080839157, "learning_rate": 8.936086044157881e-05, "loss": 0.06, "step": 50190 }, { "epoch": 1.6971500050711654, "grad_norm": 0.601662814617157, "learning_rate": 8.935576162927388e-05, "loss": 0.0622, "step": 50200 }, { "epoch": 1.6974880827614185, "grad_norm": 0.2939743101596832, "learning_rate": 8.935066174099721e-05, "loss": 0.0585, "step": 50210 }, { "epoch": 1.6978261604516718, "grad_norm": 0.27522990107536316, "learning_rate": 8.934556077688823e-05, "loss": 0.0567, "step": 50220 }, { "epoch": 1.698164238141925, "grad_norm": 0.6916036009788513, "learning_rate": 8.934045873708641e-05, "loss": 0.0535, "step": 50230 }, { "epoch": 1.6985023158321781, "grad_norm": 0.3134123384952545, "learning_rate": 8.933535562173121e-05, "loss": 0.0541, "step": 50240 }, { "epoch": 1.6988403935224314, "grad_norm": 0.3389943540096283, "learning_rate": 8.933025143096219e-05, "loss": 0.0599, "step": 50250 }, { "epoch": 1.6991784712126847, "grad_norm": 0.2674359679222107, "learning_rate": 8.932514616491885e-05, "loss": 0.0598, "step": 50260 }, { "epoch": 1.6995165489029378, "grad_norm": 0.31036072969436646, "learning_rate": 8.93200398237408e-05, "loss": 0.0584, "step": 50270 }, { "epoch": 1.699854626593191, "grad_norm": 0.294159859418869, "learning_rate": 8.931493240756762e-05, "loss": 0.0563, "step": 50280 }, { "epoch": 1.7001927042834444, "grad_norm": 0.3621813654899597, "learning_rate": 8.930982391653897e-05, "loss": 0.057, "step": 50290 }, { "epoch": 1.7005307819736974, "grad_norm": 0.3389167785644531, "learning_rate": 8.93047143507945e-05, "loss": 0.0596, "step": 50300 }, { "epoch": 1.7008688596639507, "grad_norm": 0.33719953894615173, "learning_rate": 8.92996037104739e-05, "loss": 0.0538, "step": 50310 }, { "epoch": 1.701206937354204, "grad_norm": 0.3446573317050934, "learning_rate": 8.929449199571692e-05, "loss": 0.0601, "step": 50320 }, { "epoch": 1.701545015044457, "grad_norm": 0.4387284219264984, "learning_rate": 8.928937920666325e-05, "loss": 0.0574, "step": 50330 }, { "epoch": 1.7018830927347104, "grad_norm": 0.34530603885650635, "learning_rate": 8.928426534345274e-05, "loss": 0.0574, "step": 50340 }, { "epoch": 1.7022211704249637, "grad_norm": 0.5948596000671387, "learning_rate": 8.927915040622518e-05, "loss": 0.0542, "step": 50350 }, { "epoch": 1.7025592481152168, "grad_norm": 0.39848169684410095, "learning_rate": 8.927403439512039e-05, "loss": 0.059, "step": 50360 }, { "epoch": 1.7028973258054703, "grad_norm": 0.3829577565193176, "learning_rate": 8.926891731027826e-05, "loss": 0.0594, "step": 50370 }, { "epoch": 1.7032354034957233, "grad_norm": 0.3231581449508667, "learning_rate": 8.926379915183869e-05, "loss": 0.0571, "step": 50380 }, { "epoch": 1.7035734811859764, "grad_norm": 0.4090569317340851, "learning_rate": 8.925867991994159e-05, "loss": 0.0574, "step": 50390 }, { "epoch": 1.70391155887623, "grad_norm": 0.4845467209815979, "learning_rate": 8.925355961472692e-05, "loss": 0.0599, "step": 50400 }, { "epoch": 1.704249636566483, "grad_norm": 0.6376245617866516, "learning_rate": 8.924843823633471e-05, "loss": 0.0582, "step": 50410 }, { "epoch": 1.704587714256736, "grad_norm": 0.5180490016937256, "learning_rate": 8.924331578490492e-05, "loss": 0.0529, "step": 50420 }, { "epoch": 1.7049257919469896, "grad_norm": 0.5152701139450073, "learning_rate": 8.923819226057762e-05, "loss": 0.0564, "step": 50430 }, { "epoch": 1.7052638696372426, "grad_norm": 0.34542757272720337, "learning_rate": 8.923306766349288e-05, "loss": 0.0539, "step": 50440 }, { "epoch": 1.7056019473274957, "grad_norm": 0.28297150135040283, "learning_rate": 8.922794199379084e-05, "loss": 0.0595, "step": 50450 }, { "epoch": 1.7059400250177492, "grad_norm": 0.590328574180603, "learning_rate": 8.922281525161157e-05, "loss": 0.0638, "step": 50460 }, { "epoch": 1.7062781027080023, "grad_norm": 0.5200374126434326, "learning_rate": 8.921768743709528e-05, "loss": 0.0709, "step": 50470 }, { "epoch": 1.7066161803982554, "grad_norm": 0.3605761229991913, "learning_rate": 8.921255855038215e-05, "loss": 0.0585, "step": 50480 }, { "epoch": 1.7069542580885089, "grad_norm": 0.42430147528648376, "learning_rate": 8.92074285916124e-05, "loss": 0.0548, "step": 50490 }, { "epoch": 1.707292335778762, "grad_norm": 0.3180597722530365, "learning_rate": 8.920229756092626e-05, "loss": 0.053, "step": 50500 }, { "epoch": 1.7076304134690152, "grad_norm": 0.305151104927063, "learning_rate": 8.919716545846406e-05, "loss": 0.0625, "step": 50510 }, { "epoch": 1.7079684911592685, "grad_norm": 0.49762916564941406, "learning_rate": 8.919203228436608e-05, "loss": 0.056, "step": 50520 }, { "epoch": 1.7083065688495216, "grad_norm": 0.31962957978248596, "learning_rate": 8.918689803877265e-05, "loss": 0.0514, "step": 50530 }, { "epoch": 1.7086446465397749, "grad_norm": 0.4877256155014038, "learning_rate": 8.918176272182415e-05, "loss": 0.0572, "step": 50540 }, { "epoch": 1.7089827242300282, "grad_norm": 0.31258705258369446, "learning_rate": 8.917662633366099e-05, "loss": 0.0558, "step": 50550 }, { "epoch": 1.7093208019202812, "grad_norm": 0.32782694697380066, "learning_rate": 8.917148887442357e-05, "loss": 0.0588, "step": 50560 }, { "epoch": 1.7096588796105345, "grad_norm": 0.4468458592891693, "learning_rate": 8.916635034425235e-05, "loss": 0.0627, "step": 50570 }, { "epoch": 1.7099969573007878, "grad_norm": 0.337340384721756, "learning_rate": 8.916121074328786e-05, "loss": 0.0595, "step": 50580 }, { "epoch": 1.710335034991041, "grad_norm": 0.30911165475845337, "learning_rate": 8.915607007167055e-05, "loss": 0.0583, "step": 50590 }, { "epoch": 1.7106731126812942, "grad_norm": 0.5917582511901855, "learning_rate": 8.915092832954102e-05, "loss": 0.0572, "step": 50600 }, { "epoch": 1.7110111903715475, "grad_norm": 0.3768131732940674, "learning_rate": 8.91457855170398e-05, "loss": 0.0571, "step": 50610 }, { "epoch": 1.7113492680618005, "grad_norm": 0.28604087233543396, "learning_rate": 8.914064163430751e-05, "loss": 0.0575, "step": 50620 }, { "epoch": 1.7116873457520538, "grad_norm": 0.31982138752937317, "learning_rate": 8.913549668148479e-05, "loss": 0.0584, "step": 50630 }, { "epoch": 1.7120254234423071, "grad_norm": 0.5198556780815125, "learning_rate": 8.91303506587123e-05, "loss": 0.0587, "step": 50640 }, { "epoch": 1.7123635011325602, "grad_norm": 0.475048303604126, "learning_rate": 8.91252035661307e-05, "loss": 0.0649, "step": 50650 }, { "epoch": 1.7127015788228135, "grad_norm": 0.2758937180042267, "learning_rate": 8.912005540388076e-05, "loss": 0.0548, "step": 50660 }, { "epoch": 1.7130396565130668, "grad_norm": 0.24513861536979675, "learning_rate": 8.911490617210319e-05, "loss": 0.0609, "step": 50670 }, { "epoch": 1.7133777342033198, "grad_norm": 0.43187418580055237, "learning_rate": 8.910975587093877e-05, "loss": 0.0603, "step": 50680 }, { "epoch": 1.7137158118935731, "grad_norm": 0.4410531520843506, "learning_rate": 8.910460450052832e-05, "loss": 0.0563, "step": 50690 }, { "epoch": 1.7140538895838264, "grad_norm": 0.3477708101272583, "learning_rate": 8.909945206101268e-05, "loss": 0.0571, "step": 50700 }, { "epoch": 1.7143919672740795, "grad_norm": 0.5314953923225403, "learning_rate": 8.909429855253271e-05, "loss": 0.0578, "step": 50710 }, { "epoch": 1.7147300449643328, "grad_norm": 0.37124279141426086, "learning_rate": 8.90891439752293e-05, "loss": 0.0596, "step": 50720 }, { "epoch": 1.715068122654586, "grad_norm": 0.38812142610549927, "learning_rate": 8.908398832924338e-05, "loss": 0.0599, "step": 50730 }, { "epoch": 1.7154062003448391, "grad_norm": 0.28603675961494446, "learning_rate": 8.907883161471591e-05, "loss": 0.0521, "step": 50740 }, { "epoch": 1.7157442780350924, "grad_norm": 0.3083973526954651, "learning_rate": 8.907367383178786e-05, "loss": 0.0588, "step": 50750 }, { "epoch": 1.7160823557253457, "grad_norm": 0.4125329256057739, "learning_rate": 8.906851498060025e-05, "loss": 0.0555, "step": 50760 }, { "epoch": 1.7164204334155988, "grad_norm": 0.17760498821735382, "learning_rate": 8.906335506129411e-05, "loss": 0.0558, "step": 50770 }, { "epoch": 1.716758511105852, "grad_norm": 0.5518366098403931, "learning_rate": 8.905819407401052e-05, "loss": 0.0613, "step": 50780 }, { "epoch": 1.7170965887961054, "grad_norm": 0.4009072184562683, "learning_rate": 8.905303201889058e-05, "loss": 0.0611, "step": 50790 }, { "epoch": 1.7174346664863585, "grad_norm": 0.24696461856365204, "learning_rate": 8.904786889607541e-05, "loss": 0.0587, "step": 50800 }, { "epoch": 1.7177727441766117, "grad_norm": 0.30586114525794983, "learning_rate": 8.904270470570619e-05, "loss": 0.0519, "step": 50810 }, { "epoch": 1.718110821866865, "grad_norm": 0.2977992594242096, "learning_rate": 8.903753944792408e-05, "loss": 0.0521, "step": 50820 }, { "epoch": 1.718448899557118, "grad_norm": 0.21364332735538483, "learning_rate": 8.903237312287031e-05, "loss": 0.0547, "step": 50830 }, { "epoch": 1.7187869772473714, "grad_norm": 0.39602130651474, "learning_rate": 8.902720573068612e-05, "loss": 0.0548, "step": 50840 }, { "epoch": 1.7191250549376247, "grad_norm": 0.45006901025772095, "learning_rate": 8.90220372715128e-05, "loss": 0.0614, "step": 50850 }, { "epoch": 1.7194631326278778, "grad_norm": 0.5535448789596558, "learning_rate": 8.901686774549162e-05, "loss": 0.0668, "step": 50860 }, { "epoch": 1.7198012103181313, "grad_norm": 0.47555792331695557, "learning_rate": 8.901169715276394e-05, "loss": 0.0567, "step": 50870 }, { "epoch": 1.7201392880083843, "grad_norm": 0.34949132800102234, "learning_rate": 8.900652549347113e-05, "loss": 0.0587, "step": 50880 }, { "epoch": 1.7204773656986374, "grad_norm": 0.26953086256980896, "learning_rate": 8.900135276775454e-05, "loss": 0.0604, "step": 50890 }, { "epoch": 1.720815443388891, "grad_norm": 0.26845747232437134, "learning_rate": 8.899617897575564e-05, "loss": 0.0533, "step": 50900 }, { "epoch": 1.721153521079144, "grad_norm": 0.32265275716781616, "learning_rate": 8.899100411761585e-05, "loss": 0.0582, "step": 50910 }, { "epoch": 1.721491598769397, "grad_norm": 0.19451816380023956, "learning_rate": 8.898582819347665e-05, "loss": 0.0518, "step": 50920 }, { "epoch": 1.7218296764596506, "grad_norm": 0.3155377507209778, "learning_rate": 8.898065120347956e-05, "loss": 0.0559, "step": 50930 }, { "epoch": 1.7221677541499036, "grad_norm": 0.3046512305736542, "learning_rate": 8.89754731477661e-05, "loss": 0.0582, "step": 50940 }, { "epoch": 1.7225058318401567, "grad_norm": 0.30958661437034607, "learning_rate": 8.897029402647787e-05, "loss": 0.0564, "step": 50950 }, { "epoch": 1.7228439095304102, "grad_norm": 0.3238644003868103, "learning_rate": 8.896511383975643e-05, "loss": 0.0491, "step": 50960 }, { "epoch": 1.7231819872206633, "grad_norm": 0.4245263636112213, "learning_rate": 8.89599325877434e-05, "loss": 0.0536, "step": 50970 }, { "epoch": 1.7235200649109166, "grad_norm": 0.33918124437332153, "learning_rate": 8.895475027058046e-05, "loss": 0.0591, "step": 50980 }, { "epoch": 1.7238581426011699, "grad_norm": 0.5333811044692993, "learning_rate": 8.89495668884093e-05, "loss": 0.0546, "step": 50990 }, { "epoch": 1.724196220291423, "grad_norm": 0.2822851240634918, "learning_rate": 8.89443824413716e-05, "loss": 0.0604, "step": 51000 }, { "epoch": 1.7245342979816762, "grad_norm": 0.6558526158332825, "learning_rate": 8.893919692960912e-05, "loss": 0.0568, "step": 51010 }, { "epoch": 1.7248723756719295, "grad_norm": 0.37012359499931335, "learning_rate": 8.89340103532636e-05, "loss": 0.0564, "step": 51020 }, { "epoch": 1.7252104533621826, "grad_norm": 0.2920953035354614, "learning_rate": 8.892882271247689e-05, "loss": 0.0594, "step": 51030 }, { "epoch": 1.7255485310524359, "grad_norm": 0.47622212767601013, "learning_rate": 8.892363400739079e-05, "loss": 0.0564, "step": 51040 }, { "epoch": 1.7258866087426892, "grad_norm": 0.3912917375564575, "learning_rate": 8.891844423814715e-05, "loss": 0.057, "step": 51050 }, { "epoch": 1.7262246864329422, "grad_norm": 0.5393288135528564, "learning_rate": 8.891325340488787e-05, "loss": 0.0595, "step": 51060 }, { "epoch": 1.7265627641231955, "grad_norm": 0.2698400914669037, "learning_rate": 8.890806150775486e-05, "loss": 0.0517, "step": 51070 }, { "epoch": 1.7269008418134488, "grad_norm": 0.8315902948379517, "learning_rate": 8.890286854689007e-05, "loss": 0.0568, "step": 51080 }, { "epoch": 1.727238919503702, "grad_norm": 0.21217851340770721, "learning_rate": 8.889767452243546e-05, "loss": 0.0579, "step": 51090 }, { "epoch": 1.7275769971939552, "grad_norm": 1.8518532514572144, "learning_rate": 8.889247943453305e-05, "loss": 0.0553, "step": 51100 }, { "epoch": 1.7279150748842085, "grad_norm": 0.32511165738105774, "learning_rate": 8.888728328332488e-05, "loss": 0.0647, "step": 51110 }, { "epoch": 1.7282531525744615, "grad_norm": 0.5628110766410828, "learning_rate": 8.888208606895296e-05, "loss": 0.0561, "step": 51120 }, { "epoch": 1.7285912302647148, "grad_norm": 0.33366480469703674, "learning_rate": 8.887688779155943e-05, "loss": 0.0542, "step": 51130 }, { "epoch": 1.7289293079549681, "grad_norm": 0.3687514364719391, "learning_rate": 8.88716884512864e-05, "loss": 0.0579, "step": 51140 }, { "epoch": 1.7292673856452212, "grad_norm": 0.27772822976112366, "learning_rate": 8.8866488048276e-05, "loss": 0.0605, "step": 51150 }, { "epoch": 1.7296054633354745, "grad_norm": 0.4443134665489197, "learning_rate": 8.886128658267041e-05, "loss": 0.0616, "step": 51160 }, { "epoch": 1.7299435410257278, "grad_norm": 0.25004762411117554, "learning_rate": 8.885608405461187e-05, "loss": 0.0587, "step": 51170 }, { "epoch": 1.7302816187159809, "grad_norm": 0.41067051887512207, "learning_rate": 8.885088046424258e-05, "loss": 0.0616, "step": 51180 }, { "epoch": 1.7306196964062341, "grad_norm": 0.5301944017410278, "learning_rate": 8.88456758117048e-05, "loss": 0.0603, "step": 51190 }, { "epoch": 1.7309577740964874, "grad_norm": 0.3586325943470001, "learning_rate": 8.884047009714084e-05, "loss": 0.0576, "step": 51200 }, { "epoch": 1.7312958517867405, "grad_norm": 0.5700859427452087, "learning_rate": 8.883526332069304e-05, "loss": 0.0537, "step": 51210 }, { "epoch": 1.7316339294769938, "grad_norm": 0.29435738921165466, "learning_rate": 8.88300554825037e-05, "loss": 0.0586, "step": 51220 }, { "epoch": 1.731972007167247, "grad_norm": 0.35739222168922424, "learning_rate": 8.882484658271526e-05, "loss": 0.0592, "step": 51230 }, { "epoch": 1.7323100848575002, "grad_norm": 0.5107804536819458, "learning_rate": 8.881963662147008e-05, "loss": 0.0561, "step": 51240 }, { "epoch": 1.7326481625477534, "grad_norm": 0.2420528531074524, "learning_rate": 8.881442559891063e-05, "loss": 0.0582, "step": 51250 }, { "epoch": 1.7329862402380067, "grad_norm": 0.4595950245857239, "learning_rate": 8.880921351517934e-05, "loss": 0.0565, "step": 51260 }, { "epoch": 1.7333243179282598, "grad_norm": 0.3372074067592621, "learning_rate": 8.880400037041876e-05, "loss": 0.0551, "step": 51270 }, { "epoch": 1.733662395618513, "grad_norm": 0.3527921736240387, "learning_rate": 8.87987861647714e-05, "loss": 0.057, "step": 51280 }, { "epoch": 1.7340004733087664, "grad_norm": 0.4035516381263733, "learning_rate": 8.879357089837977e-05, "loss": 0.0561, "step": 51290 }, { "epoch": 1.7343385509990195, "grad_norm": 0.43448489904403687, "learning_rate": 8.87883545713865e-05, "loss": 0.0538, "step": 51300 }, { "epoch": 1.7346766286892727, "grad_norm": 0.33729177713394165, "learning_rate": 8.87831371839342e-05, "loss": 0.0555, "step": 51310 }, { "epoch": 1.735014706379526, "grad_norm": 0.27036070823669434, "learning_rate": 8.877791873616548e-05, "loss": 0.0583, "step": 51320 }, { "epoch": 1.735352784069779, "grad_norm": 0.31148236989974976, "learning_rate": 8.877269922822305e-05, "loss": 0.0551, "step": 51330 }, { "epoch": 1.7356908617600326, "grad_norm": 0.32507339119911194, "learning_rate": 8.876747866024956e-05, "loss": 0.0602, "step": 51340 }, { "epoch": 1.7360289394502857, "grad_norm": 0.3494816720485687, "learning_rate": 8.876225703238778e-05, "loss": 0.0587, "step": 51350 }, { "epoch": 1.7363670171405388, "grad_norm": 0.2963871955871582, "learning_rate": 8.875703434478047e-05, "loss": 0.053, "step": 51360 }, { "epoch": 1.7367050948307923, "grad_norm": 0.35922548174858093, "learning_rate": 8.875181059757038e-05, "loss": 0.0541, "step": 51370 }, { "epoch": 1.7370431725210453, "grad_norm": 0.2381494641304016, "learning_rate": 8.874658579090036e-05, "loss": 0.0582, "step": 51380 }, { "epoch": 1.7373812502112984, "grad_norm": 0.3927482068538666, "learning_rate": 8.874135992491323e-05, "loss": 0.0583, "step": 51390 }, { "epoch": 1.737719327901552, "grad_norm": 0.4095551669597626, "learning_rate": 8.87361329997519e-05, "loss": 0.0539, "step": 51400 }, { "epoch": 1.738057405591805, "grad_norm": 0.6110865473747253, "learning_rate": 8.873090501555922e-05, "loss": 0.0549, "step": 51410 }, { "epoch": 1.738395483282058, "grad_norm": 0.40764275193214417, "learning_rate": 8.872567597247816e-05, "loss": 0.0586, "step": 51420 }, { "epoch": 1.7387335609723116, "grad_norm": 0.5358763933181763, "learning_rate": 8.872044587065167e-05, "loss": 0.0633, "step": 51430 }, { "epoch": 1.7390716386625646, "grad_norm": 0.40140071511268616, "learning_rate": 8.871521471022273e-05, "loss": 0.0575, "step": 51440 }, { "epoch": 1.7394097163528177, "grad_norm": 0.2175525724887848, "learning_rate": 8.870998249133435e-05, "loss": 0.0544, "step": 51450 }, { "epoch": 1.7397477940430712, "grad_norm": 0.372280478477478, "learning_rate": 8.870474921412962e-05, "loss": 0.0599, "step": 51460 }, { "epoch": 1.7400858717333243, "grad_norm": 0.4661327004432678, "learning_rate": 8.869951487875157e-05, "loss": 0.0525, "step": 51470 }, { "epoch": 1.7404239494235776, "grad_norm": 0.38185545802116394, "learning_rate": 8.869427948534332e-05, "loss": 0.0588, "step": 51480 }, { "epoch": 1.7407620271138309, "grad_norm": 0.22849209606647491, "learning_rate": 8.868904303404801e-05, "loss": 0.0559, "step": 51490 }, { "epoch": 1.741100104804084, "grad_norm": 0.2543376088142395, "learning_rate": 8.86838055250088e-05, "loss": 0.0572, "step": 51500 }, { "epoch": 1.7414381824943372, "grad_norm": 0.28583642840385437, "learning_rate": 8.867856695836888e-05, "loss": 0.0616, "step": 51510 }, { "epoch": 1.7417762601845905, "grad_norm": 0.2811681628227234, "learning_rate": 8.867332733427147e-05, "loss": 0.0523, "step": 51520 }, { "epoch": 1.7421143378748436, "grad_norm": 0.4847171902656555, "learning_rate": 8.866808665285981e-05, "loss": 0.0599, "step": 51530 }, { "epoch": 1.742452415565097, "grad_norm": 0.25221362709999084, "learning_rate": 8.866284491427721e-05, "loss": 0.0621, "step": 51540 }, { "epoch": 1.7427904932553502, "grad_norm": 0.8457407355308533, "learning_rate": 8.865760211866694e-05, "loss": 0.0579, "step": 51550 }, { "epoch": 1.7431285709456033, "grad_norm": 0.3097688853740692, "learning_rate": 8.865235826617235e-05, "loss": 0.0634, "step": 51560 }, { "epoch": 1.7434666486358565, "grad_norm": 0.3493639826774597, "learning_rate": 8.86471133569368e-05, "loss": 0.0587, "step": 51570 }, { "epoch": 1.7438047263261098, "grad_norm": 0.3347793519496918, "learning_rate": 8.86418673911037e-05, "loss": 0.0597, "step": 51580 }, { "epoch": 1.744142804016363, "grad_norm": 0.6658639311790466, "learning_rate": 8.863662036881647e-05, "loss": 0.0566, "step": 51590 }, { "epoch": 1.7444808817066162, "grad_norm": 0.37135496735572815, "learning_rate": 8.863137229021854e-05, "loss": 0.0654, "step": 51600 }, { "epoch": 1.7448189593968695, "grad_norm": 0.2988635003566742, "learning_rate": 8.862612315545341e-05, "loss": 0.0658, "step": 51610 }, { "epoch": 1.7451570370871226, "grad_norm": 0.3865065574645996, "learning_rate": 8.862087296466459e-05, "loss": 0.0571, "step": 51620 }, { "epoch": 1.7454951147773758, "grad_norm": 0.2860180735588074, "learning_rate": 8.86156217179956e-05, "loss": 0.0607, "step": 51630 }, { "epoch": 1.7458331924676291, "grad_norm": 0.2590595781803131, "learning_rate": 8.861036941559002e-05, "loss": 0.0616, "step": 51640 }, { "epoch": 1.7461712701578822, "grad_norm": 0.4128935635089874, "learning_rate": 8.860511605759146e-05, "loss": 0.0581, "step": 51650 }, { "epoch": 1.7465093478481355, "grad_norm": 0.49792197346687317, "learning_rate": 8.859986164414351e-05, "loss": 0.0604, "step": 51660 }, { "epoch": 1.7468474255383888, "grad_norm": 0.9443799257278442, "learning_rate": 8.859460617538986e-05, "loss": 0.0625, "step": 51670 }, { "epoch": 1.7471855032286419, "grad_norm": 0.5487092733383179, "learning_rate": 8.858934965147417e-05, "loss": 0.0628, "step": 51680 }, { "epoch": 1.7475235809188951, "grad_norm": 0.4718179702758789, "learning_rate": 8.858409207254018e-05, "loss": 0.0557, "step": 51690 }, { "epoch": 1.7478616586091484, "grad_norm": 0.5259551405906677, "learning_rate": 8.857883343873157e-05, "loss": 0.0506, "step": 51700 }, { "epoch": 1.7481997362994015, "grad_norm": 0.42357778549194336, "learning_rate": 8.857357375019218e-05, "loss": 0.0638, "step": 51710 }, { "epoch": 1.7485378139896548, "grad_norm": 0.39145755767822266, "learning_rate": 8.856831300706575e-05, "loss": 0.0615, "step": 51720 }, { "epoch": 1.748875891679908, "grad_norm": 0.6301740407943726, "learning_rate": 8.856305120949614e-05, "loss": 0.0627, "step": 51730 }, { "epoch": 1.7492139693701612, "grad_norm": 0.6510775685310364, "learning_rate": 8.85577883576272e-05, "loss": 0.0634, "step": 51740 }, { "epoch": 1.7495520470604145, "grad_norm": 0.8200224041938782, "learning_rate": 8.855252445160281e-05, "loss": 0.0606, "step": 51750 }, { "epoch": 1.7498901247506677, "grad_norm": 0.3797060251235962, "learning_rate": 8.854725949156688e-05, "loss": 0.0567, "step": 51760 }, { "epoch": 1.7502282024409208, "grad_norm": 0.715452253818512, "learning_rate": 8.854199347766337e-05, "loss": 0.0582, "step": 51770 }, { "epoch": 1.750566280131174, "grad_norm": 0.9123436808586121, "learning_rate": 8.853672641003623e-05, "loss": 0.062, "step": 51780 }, { "epoch": 1.7509043578214274, "grad_norm": 0.2897985577583313, "learning_rate": 8.853145828882948e-05, "loss": 0.0602, "step": 51790 }, { "epoch": 1.7512424355116805, "grad_norm": 0.2789511978626251, "learning_rate": 8.85261891141871e-05, "loss": 0.0553, "step": 51800 }, { "epoch": 1.7515805132019338, "grad_norm": 0.2961726784706116, "learning_rate": 8.852091888625322e-05, "loss": 0.056, "step": 51810 }, { "epoch": 1.751918590892187, "grad_norm": 0.26870861649513245, "learning_rate": 8.851564760517188e-05, "loss": 0.0592, "step": 51820 }, { "epoch": 1.7522566685824401, "grad_norm": 0.34393271803855896, "learning_rate": 8.85103752710872e-05, "loss": 0.0512, "step": 51830 }, { "epoch": 1.7525947462726936, "grad_norm": 0.2941077947616577, "learning_rate": 8.850510188414334e-05, "loss": 0.0615, "step": 51840 }, { "epoch": 1.7529328239629467, "grad_norm": 0.23492549359798431, "learning_rate": 8.849982744448444e-05, "loss": 0.0572, "step": 51850 }, { "epoch": 1.7532709016531998, "grad_norm": 0.31377536058425903, "learning_rate": 8.849455195225473e-05, "loss": 0.054, "step": 51860 }, { "epoch": 1.7536089793434533, "grad_norm": 0.35051146149635315, "learning_rate": 8.848927540759843e-05, "loss": 0.0555, "step": 51870 }, { "epoch": 1.7539470570337063, "grad_norm": 0.5207467675209045, "learning_rate": 8.848399781065981e-05, "loss": 0.0617, "step": 51880 }, { "epoch": 1.7542851347239594, "grad_norm": 0.4193461239337921, "learning_rate": 8.847871916158312e-05, "loss": 0.0543, "step": 51890 }, { "epoch": 1.754623212414213, "grad_norm": 0.3881557881832123, "learning_rate": 8.847343946051273e-05, "loss": 0.0597, "step": 51900 }, { "epoch": 1.754961290104466, "grad_norm": 0.4307311773300171, "learning_rate": 8.846815870759295e-05, "loss": 0.0574, "step": 51910 }, { "epoch": 1.755299367794719, "grad_norm": 0.4983651041984558, "learning_rate": 8.846287690296817e-05, "loss": 0.0535, "step": 51920 }, { "epoch": 1.7556374454849726, "grad_norm": 0.4195428490638733, "learning_rate": 8.845759404678278e-05, "loss": 0.0638, "step": 51930 }, { "epoch": 1.7559755231752256, "grad_norm": 0.37091729044914246, "learning_rate": 8.845231013918122e-05, "loss": 0.0541, "step": 51940 }, { "epoch": 1.756313600865479, "grad_norm": 0.4792543351650238, "learning_rate": 8.844702518030794e-05, "loss": 0.0607, "step": 51950 }, { "epoch": 1.7566516785557322, "grad_norm": 0.3205468952655792, "learning_rate": 8.844173917030744e-05, "loss": 0.0542, "step": 51960 }, { "epoch": 1.7569897562459853, "grad_norm": 0.32771968841552734, "learning_rate": 8.843645210932423e-05, "loss": 0.0659, "step": 51970 }, { "epoch": 1.7573278339362386, "grad_norm": 1.132272720336914, "learning_rate": 8.843116399750287e-05, "loss": 0.0633, "step": 51980 }, { "epoch": 1.7576659116264919, "grad_norm": 0.4049706757068634, "learning_rate": 8.842587483498791e-05, "loss": 0.0586, "step": 51990 }, { "epoch": 1.758003989316745, "grad_norm": 0.31178808212280273, "learning_rate": 8.842058462192397e-05, "loss": 0.0633, "step": 52000 }, { "epoch": 1.7583420670069982, "grad_norm": 0.385055810213089, "learning_rate": 8.841529335845569e-05, "loss": 0.0586, "step": 52010 }, { "epoch": 1.7586801446972515, "grad_norm": 0.8444207906723022, "learning_rate": 8.841000104472772e-05, "loss": 0.0587, "step": 52020 }, { "epoch": 1.7590182223875046, "grad_norm": 0.5409836769104004, "learning_rate": 8.840470768088473e-05, "loss": 0.0559, "step": 52030 }, { "epoch": 1.759356300077758, "grad_norm": 1.0734683275222778, "learning_rate": 8.839941326707149e-05, "loss": 0.0613, "step": 52040 }, { "epoch": 1.7596943777680112, "grad_norm": 0.29054778814315796, "learning_rate": 8.839411780343269e-05, "loss": 0.0637, "step": 52050 }, { "epoch": 1.7600324554582643, "grad_norm": 0.4514719843864441, "learning_rate": 8.838882129011317e-05, "loss": 0.0579, "step": 52060 }, { "epoch": 1.7603705331485175, "grad_norm": 0.44033095240592957, "learning_rate": 8.838352372725767e-05, "loss": 0.0581, "step": 52070 }, { "epoch": 1.7607086108387708, "grad_norm": 0.40646830201148987, "learning_rate": 8.837822511501105e-05, "loss": 0.0576, "step": 52080 }, { "epoch": 1.761046688529024, "grad_norm": 0.4503231346607208, "learning_rate": 8.837292545351818e-05, "loss": 0.0559, "step": 52090 }, { "epoch": 1.7613847662192772, "grad_norm": 0.6295009851455688, "learning_rate": 8.836762474292395e-05, "loss": 0.0575, "step": 52100 }, { "epoch": 1.7617228439095305, "grad_norm": 0.25405335426330566, "learning_rate": 8.836232298337326e-05, "loss": 0.0593, "step": 52110 }, { "epoch": 1.7620609215997836, "grad_norm": 1.211179256439209, "learning_rate": 8.835702017501108e-05, "loss": 0.0615, "step": 52120 }, { "epoch": 1.7623989992900368, "grad_norm": 0.4464603364467621, "learning_rate": 8.835171631798238e-05, "loss": 0.0544, "step": 52130 }, { "epoch": 1.7627370769802901, "grad_norm": 0.42731210589408875, "learning_rate": 8.834641141243217e-05, "loss": 0.0596, "step": 52140 }, { "epoch": 1.7630751546705432, "grad_norm": 0.5441881418228149, "learning_rate": 8.834110545850545e-05, "loss": 0.0606, "step": 52150 }, { "epoch": 1.7634132323607965, "grad_norm": 0.3652198612689972, "learning_rate": 8.833579845634735e-05, "loss": 0.0592, "step": 52160 }, { "epoch": 1.7637513100510498, "grad_norm": 0.4793745279312134, "learning_rate": 8.83304904061029e-05, "loss": 0.0626, "step": 52170 }, { "epoch": 1.7640893877413029, "grad_norm": 0.5895410776138306, "learning_rate": 8.832518130791724e-05, "loss": 0.0626, "step": 52180 }, { "epoch": 1.7644274654315562, "grad_norm": 0.5709414482116699, "learning_rate": 8.831987116193552e-05, "loss": 0.0608, "step": 52190 }, { "epoch": 1.7647655431218094, "grad_norm": 0.5832012295722961, "learning_rate": 8.831455996830294e-05, "loss": 0.0644, "step": 52200 }, { "epoch": 1.7651036208120625, "grad_norm": 0.5048625469207764, "learning_rate": 8.830924772716467e-05, "loss": 0.0661, "step": 52210 }, { "epoch": 1.7654416985023158, "grad_norm": 0.818747878074646, "learning_rate": 8.830393443866596e-05, "loss": 0.0659, "step": 52220 }, { "epoch": 1.765779776192569, "grad_norm": 0.7457579970359802, "learning_rate": 8.829862010295207e-05, "loss": 0.0704, "step": 52230 }, { "epoch": 1.7661178538828222, "grad_norm": 1.262976050376892, "learning_rate": 8.829330472016828e-05, "loss": 0.0667, "step": 52240 }, { "epoch": 1.7664559315730755, "grad_norm": 2.4139788150787354, "learning_rate": 8.828798829045994e-05, "loss": 0.065, "step": 52250 }, { "epoch": 1.7667940092633287, "grad_norm": 1.070237398147583, "learning_rate": 8.82826708139724e-05, "loss": 0.0735, "step": 52260 }, { "epoch": 1.7671320869535818, "grad_norm": 0.5928160548210144, "learning_rate": 8.827735229085101e-05, "loss": 0.0716, "step": 52270 }, { "epoch": 1.767470164643835, "grad_norm": 1.1204530000686646, "learning_rate": 8.827203272124119e-05, "loss": 0.0695, "step": 52280 }, { "epoch": 1.7678082423340884, "grad_norm": 0.3991992771625519, "learning_rate": 8.826671210528836e-05, "loss": 0.0671, "step": 52290 }, { "epoch": 1.7681463200243415, "grad_norm": 0.5517832040786743, "learning_rate": 8.826139044313801e-05, "loss": 0.0645, "step": 52300 }, { "epoch": 1.768484397714595, "grad_norm": 0.2836725413799286, "learning_rate": 8.825606773493561e-05, "loss": 0.0651, "step": 52310 }, { "epoch": 1.768822475404848, "grad_norm": 0.5483925342559814, "learning_rate": 8.82507439808267e-05, "loss": 0.0594, "step": 52320 }, { "epoch": 1.7691605530951011, "grad_norm": 0.3194272518157959, "learning_rate": 8.824541918095681e-05, "loss": 0.0639, "step": 52330 }, { "epoch": 1.7694986307853546, "grad_norm": 0.35510340332984924, "learning_rate": 8.824009333547154e-05, "loss": 0.0584, "step": 52340 }, { "epoch": 1.7698367084756077, "grad_norm": 0.8678958415985107, "learning_rate": 8.823476644451646e-05, "loss": 0.0602, "step": 52350 }, { "epoch": 1.7701747861658608, "grad_norm": 1.0873734951019287, "learning_rate": 8.822943850823726e-05, "loss": 0.0647, "step": 52360 }, { "epoch": 1.7705128638561143, "grad_norm": 1.074145793914795, "learning_rate": 8.822410952677956e-05, "loss": 0.0577, "step": 52370 }, { "epoch": 1.7708509415463674, "grad_norm": 0.43555933237075806, "learning_rate": 8.821877950028908e-05, "loss": 0.0638, "step": 52380 }, { "epoch": 1.7711890192366204, "grad_norm": 0.7554315328598022, "learning_rate": 8.821344842891151e-05, "loss": 0.0559, "step": 52390 }, { "epoch": 1.771527096926874, "grad_norm": 0.6439234018325806, "learning_rate": 8.820811631279262e-05, "loss": 0.067, "step": 52400 }, { "epoch": 1.771865174617127, "grad_norm": 0.3518700897693634, "learning_rate": 8.820278315207817e-05, "loss": 0.0596, "step": 52410 }, { "epoch": 1.77220325230738, "grad_norm": 0.4375792145729065, "learning_rate": 8.8197448946914e-05, "loss": 0.0738, "step": 52420 }, { "epoch": 1.7725413299976336, "grad_norm": 0.4937407076358795, "learning_rate": 8.81921136974459e-05, "loss": 0.0621, "step": 52430 }, { "epoch": 1.7728794076878867, "grad_norm": 0.5519753098487854, "learning_rate": 8.81867774038198e-05, "loss": 0.0621, "step": 52440 }, { "epoch": 1.77321748537814, "grad_norm": 0.55915766954422, "learning_rate": 8.818144006618154e-05, "loss": 0.0622, "step": 52450 }, { "epoch": 1.7735555630683932, "grad_norm": 0.3358355164527893, "learning_rate": 8.817610168467704e-05, "loss": 0.0555, "step": 52460 }, { "epoch": 1.7738936407586463, "grad_norm": 0.28630009293556213, "learning_rate": 8.817076225945227e-05, "loss": 0.0597, "step": 52470 }, { "epoch": 1.7742317184488996, "grad_norm": 1.462234377861023, "learning_rate": 8.816542179065319e-05, "loss": 0.0588, "step": 52480 }, { "epoch": 1.7745697961391529, "grad_norm": 0.29965174198150635, "learning_rate": 8.816008027842582e-05, "loss": 0.0624, "step": 52490 }, { "epoch": 1.774907873829406, "grad_norm": 0.9601502418518066, "learning_rate": 8.815473772291619e-05, "loss": 0.0578, "step": 52500 }, { "epoch": 1.7752459515196592, "grad_norm": 0.3625383973121643, "learning_rate": 8.814939412427037e-05, "loss": 0.061, "step": 52510 }, { "epoch": 1.7755840292099125, "grad_norm": 0.4621601998806, "learning_rate": 8.814404948263445e-05, "loss": 0.0645, "step": 52520 }, { "epoch": 1.7759221069001656, "grad_norm": 1.498390555381775, "learning_rate": 8.813870379815454e-05, "loss": 0.0589, "step": 52530 }, { "epoch": 1.776260184590419, "grad_norm": 0.6589806079864502, "learning_rate": 8.813335707097681e-05, "loss": 0.055, "step": 52540 }, { "epoch": 1.7765982622806722, "grad_norm": 0.36196616291999817, "learning_rate": 8.81280093012474e-05, "loss": 0.062, "step": 52550 }, { "epoch": 1.7769363399709253, "grad_norm": 0.48603248596191406, "learning_rate": 8.812266048911255e-05, "loss": 0.0612, "step": 52560 }, { "epoch": 1.7772744176611786, "grad_norm": 0.2750954031944275, "learning_rate": 8.811731063471849e-05, "loss": 0.0561, "step": 52570 }, { "epoch": 1.7776124953514318, "grad_norm": 0.3335220217704773, "learning_rate": 8.811195973821147e-05, "loss": 0.0613, "step": 52580 }, { "epoch": 1.777950573041685, "grad_norm": 0.45820242166519165, "learning_rate": 8.81066077997378e-05, "loss": 0.0601, "step": 52590 }, { "epoch": 1.7782886507319382, "grad_norm": 0.5716189742088318, "learning_rate": 8.810125481944377e-05, "loss": 0.0632, "step": 52600 }, { "epoch": 1.7786267284221915, "grad_norm": 0.42137444019317627, "learning_rate": 8.809590079747576e-05, "loss": 0.0599, "step": 52610 }, { "epoch": 1.7789648061124446, "grad_norm": 0.3688697814941406, "learning_rate": 8.809054573398015e-05, "loss": 0.0615, "step": 52620 }, { "epoch": 1.7793028838026979, "grad_norm": 0.2597053050994873, "learning_rate": 8.808518962910331e-05, "loss": 0.0635, "step": 52630 }, { "epoch": 1.7796409614929511, "grad_norm": 0.3116534948348999, "learning_rate": 8.807983248299169e-05, "loss": 0.0626, "step": 52640 }, { "epoch": 1.7799790391832042, "grad_norm": 0.5442503690719604, "learning_rate": 8.807447429579177e-05, "loss": 0.057, "step": 52650 }, { "epoch": 1.7803171168734575, "grad_norm": 0.5112340450286865, "learning_rate": 8.806911506765002e-05, "loss": 0.0612, "step": 52660 }, { "epoch": 1.7806551945637108, "grad_norm": 0.4924950897693634, "learning_rate": 8.806375479871295e-05, "loss": 0.0572, "step": 52670 }, { "epoch": 1.7809932722539639, "grad_norm": 0.38719528913497925, "learning_rate": 8.805839348912715e-05, "loss": 0.0589, "step": 52680 }, { "epoch": 1.7813313499442172, "grad_norm": 0.33649829030036926, "learning_rate": 8.805303113903915e-05, "loss": 0.0549, "step": 52690 }, { "epoch": 1.7816694276344704, "grad_norm": 0.468303918838501, "learning_rate": 8.804766774859557e-05, "loss": 0.0615, "step": 52700 }, { "epoch": 1.7820075053247235, "grad_norm": 0.4425952136516571, "learning_rate": 8.804230331794306e-05, "loss": 0.0592, "step": 52710 }, { "epoch": 1.7823455830149768, "grad_norm": 0.37022697925567627, "learning_rate": 8.803693784722827e-05, "loss": 0.0577, "step": 52720 }, { "epoch": 1.78268366070523, "grad_norm": 1.1086251735687256, "learning_rate": 8.803157133659789e-05, "loss": 0.0619, "step": 52730 }, { "epoch": 1.7830217383954832, "grad_norm": 0.5547161102294922, "learning_rate": 8.802620378619864e-05, "loss": 0.0633, "step": 52740 }, { "epoch": 1.7833598160857365, "grad_norm": 0.9783478379249573, "learning_rate": 8.802083519617725e-05, "loss": 0.0605, "step": 52750 }, { "epoch": 1.7836978937759898, "grad_norm": 0.6056403517723083, "learning_rate": 8.801546556668052e-05, "loss": 0.0619, "step": 52760 }, { "epoch": 1.7840359714662428, "grad_norm": 0.5754640698432922, "learning_rate": 8.801009489785526e-05, "loss": 0.0571, "step": 52770 }, { "epoch": 1.7843740491564961, "grad_norm": 0.38241833448410034, "learning_rate": 8.800472318984825e-05, "loss": 0.069, "step": 52780 }, { "epoch": 1.7847121268467494, "grad_norm": 0.42573997378349304, "learning_rate": 8.79993504428064e-05, "loss": 0.0579, "step": 52790 }, { "epoch": 1.7850502045370025, "grad_norm": 0.6490591764450073, "learning_rate": 8.79939766568766e-05, "loss": 0.0606, "step": 52800 }, { "epoch": 1.785388282227256, "grad_norm": 0.6560494303703308, "learning_rate": 8.798860183220575e-05, "loss": 0.0584, "step": 52810 }, { "epoch": 1.785726359917509, "grad_norm": 0.6355459690093994, "learning_rate": 8.798322596894077e-05, "loss": 0.066, "step": 52820 }, { "epoch": 1.7860644376077621, "grad_norm": 0.5713039040565491, "learning_rate": 8.797784906722868e-05, "loss": 0.0598, "step": 52830 }, { "epoch": 1.7864025152980156, "grad_norm": 0.4584882855415344, "learning_rate": 8.797247112721647e-05, "loss": 0.0621, "step": 52840 }, { "epoch": 1.7867405929882687, "grad_norm": 0.38700956106185913, "learning_rate": 8.796709214905116e-05, "loss": 0.056, "step": 52850 }, { "epoch": 1.7870786706785218, "grad_norm": 0.5397063493728638, "learning_rate": 8.79617121328798e-05, "loss": 0.059, "step": 52860 }, { "epoch": 1.7874167483687753, "grad_norm": 0.34136489033699036, "learning_rate": 8.795633107884951e-05, "loss": 0.055, "step": 52870 }, { "epoch": 1.7877548260590284, "grad_norm": 0.5285612940788269, "learning_rate": 8.795094898710739e-05, "loss": 0.0609, "step": 52880 }, { "epoch": 1.7880929037492814, "grad_norm": 0.38913312554359436, "learning_rate": 8.794556585780056e-05, "loss": 0.0591, "step": 52890 }, { "epoch": 1.788430981439535, "grad_norm": 0.3893885612487793, "learning_rate": 8.794018169107623e-05, "loss": 0.0601, "step": 52900 }, { "epoch": 1.788769059129788, "grad_norm": 0.5991348028182983, "learning_rate": 8.793479648708159e-05, "loss": 0.0666, "step": 52910 }, { "epoch": 1.7891071368200413, "grad_norm": 0.2857052981853485, "learning_rate": 8.792941024596386e-05, "loss": 0.0538, "step": 52920 }, { "epoch": 1.7894452145102946, "grad_norm": 0.561629593372345, "learning_rate": 8.79240229678703e-05, "loss": 0.0579, "step": 52930 }, { "epoch": 1.7897832922005477, "grad_norm": 0.3775210976600647, "learning_rate": 8.79186346529482e-05, "loss": 0.0595, "step": 52940 }, { "epoch": 1.790121369890801, "grad_norm": 0.35241371393203735, "learning_rate": 8.791324530134488e-05, "loss": 0.0616, "step": 52950 }, { "epoch": 1.7904594475810542, "grad_norm": 0.7336592674255371, "learning_rate": 8.790785491320767e-05, "loss": 0.0608, "step": 52960 }, { "epoch": 1.7907975252713073, "grad_norm": 0.35706037282943726, "learning_rate": 8.790246348868396e-05, "loss": 0.0547, "step": 52970 }, { "epoch": 1.7911356029615606, "grad_norm": 0.31577634811401367, "learning_rate": 8.789707102792112e-05, "loss": 0.0543, "step": 52980 }, { "epoch": 1.791473680651814, "grad_norm": 0.47462743520736694, "learning_rate": 8.789167753106661e-05, "loss": 0.0579, "step": 52990 }, { "epoch": 1.791811758342067, "grad_norm": 0.3986632823944092, "learning_rate": 8.788628299826786e-05, "loss": 0.0599, "step": 53000 }, { "epoch": 1.7921498360323203, "grad_norm": 0.7470908761024475, "learning_rate": 8.788088742967238e-05, "loss": 0.0595, "step": 53010 }, { "epoch": 1.7924879137225735, "grad_norm": 0.7141001224517822, "learning_rate": 8.787549082542766e-05, "loss": 0.062, "step": 53020 }, { "epoch": 1.7928259914128266, "grad_norm": 0.7154513597488403, "learning_rate": 8.787009318568126e-05, "loss": 0.054, "step": 53030 }, { "epoch": 1.79316406910308, "grad_norm": 0.6524508595466614, "learning_rate": 8.786469451058073e-05, "loss": 0.0585, "step": 53040 }, { "epoch": 1.7935021467933332, "grad_norm": 0.6620219349861145, "learning_rate": 8.785929480027367e-05, "loss": 0.0593, "step": 53050 }, { "epoch": 1.7938402244835863, "grad_norm": 1.2686902284622192, "learning_rate": 8.785389405490773e-05, "loss": 0.0584, "step": 53060 }, { "epoch": 1.7941783021738396, "grad_norm": 0.42393234372138977, "learning_rate": 8.784849227463054e-05, "loss": 0.0707, "step": 53070 }, { "epoch": 1.7945163798640928, "grad_norm": 0.4949718415737152, "learning_rate": 8.78430894595898e-05, "loss": 0.0621, "step": 53080 }, { "epoch": 1.794854457554346, "grad_norm": 0.5920323133468628, "learning_rate": 8.783768560993317e-05, "loss": 0.057, "step": 53090 }, { "epoch": 1.7951925352445992, "grad_norm": 0.4851129949092865, "learning_rate": 8.783228072580847e-05, "loss": 0.0571, "step": 53100 }, { "epoch": 1.7955306129348525, "grad_norm": 0.39900821447372437, "learning_rate": 8.782687480736343e-05, "loss": 0.0598, "step": 53110 }, { "epoch": 1.7958686906251056, "grad_norm": 0.3324642479419708, "learning_rate": 8.782146785474581e-05, "loss": 0.0596, "step": 53120 }, { "epoch": 1.7962067683153589, "grad_norm": 0.36370500922203064, "learning_rate": 8.78160598681035e-05, "loss": 0.0599, "step": 53130 }, { "epoch": 1.7965448460056122, "grad_norm": 0.3905181586742401, "learning_rate": 8.781065084758431e-05, "loss": 0.0625, "step": 53140 }, { "epoch": 1.7968829236958652, "grad_norm": 1.3814488649368286, "learning_rate": 8.780524079333613e-05, "loss": 0.0635, "step": 53150 }, { "epoch": 1.7972210013861185, "grad_norm": 0.48485442996025085, "learning_rate": 8.779982970550687e-05, "loss": 0.061, "step": 53160 }, { "epoch": 1.7975590790763718, "grad_norm": 0.4322613775730133, "learning_rate": 8.779441758424447e-05, "loss": 0.0651, "step": 53170 }, { "epoch": 1.7978971567666249, "grad_norm": 0.4008648097515106, "learning_rate": 8.778900442969688e-05, "loss": 0.0598, "step": 53180 }, { "epoch": 1.7982352344568782, "grad_norm": 0.4629569351673126, "learning_rate": 8.778359024201212e-05, "loss": 0.0564, "step": 53190 }, { "epoch": 1.7985733121471315, "grad_norm": 0.48193833231925964, "learning_rate": 8.777817502133819e-05, "loss": 0.0568, "step": 53200 }, { "epoch": 1.7989113898373845, "grad_norm": 0.4484761655330658, "learning_rate": 8.777275876782315e-05, "loss": 0.0575, "step": 53210 }, { "epoch": 1.7992494675276378, "grad_norm": 0.4811170697212219, "learning_rate": 8.776734148161507e-05, "loss": 0.0624, "step": 53220 }, { "epoch": 1.799587545217891, "grad_norm": 0.35236993432044983, "learning_rate": 8.776192316286207e-05, "loss": 0.0635, "step": 53230 }, { "epoch": 1.7999256229081442, "grad_norm": 1.023102045059204, "learning_rate": 8.775650381171227e-05, "loss": 0.0553, "step": 53240 }, { "epoch": 1.8002637005983975, "grad_norm": 3.924410104751587, "learning_rate": 8.775108342831385e-05, "loss": 0.0591, "step": 53250 }, { "epoch": 1.8006017782886508, "grad_norm": 0.3352293074131012, "learning_rate": 8.774566201281499e-05, "loss": 0.0573, "step": 53260 }, { "epoch": 1.8009398559789038, "grad_norm": 0.347663938999176, "learning_rate": 8.774023956536391e-05, "loss": 0.0556, "step": 53270 }, { "epoch": 1.8012779336691573, "grad_norm": 0.5537902116775513, "learning_rate": 8.773481608610886e-05, "loss": 0.0616, "step": 53280 }, { "epoch": 1.8016160113594104, "grad_norm": 0.48169228434562683, "learning_rate": 8.77293915751981e-05, "loss": 0.0654, "step": 53290 }, { "epoch": 1.8019540890496635, "grad_norm": 0.43665236234664917, "learning_rate": 8.772396603277997e-05, "loss": 0.0565, "step": 53300 }, { "epoch": 1.802292166739917, "grad_norm": 0.1954999715089798, "learning_rate": 8.771853945900277e-05, "loss": 0.0563, "step": 53310 }, { "epoch": 1.80263024443017, "grad_norm": 0.5289574861526489, "learning_rate": 8.771311185401487e-05, "loss": 0.0609, "step": 53320 }, { "epoch": 1.8029683221204231, "grad_norm": 0.43647342920303345, "learning_rate": 8.770768321796466e-05, "loss": 0.0615, "step": 53330 }, { "epoch": 1.8033063998106766, "grad_norm": 0.4662722945213318, "learning_rate": 8.770225355100054e-05, "loss": 0.0621, "step": 53340 }, { "epoch": 1.8036444775009297, "grad_norm": 0.48036426305770874, "learning_rate": 8.7696822853271e-05, "loss": 0.0577, "step": 53350 }, { "epoch": 1.8039825551911828, "grad_norm": 0.27043023705482483, "learning_rate": 8.769139112492447e-05, "loss": 0.0563, "step": 53360 }, { "epoch": 1.8043206328814363, "grad_norm": 0.36221951246261597, "learning_rate": 8.768595836610948e-05, "loss": 0.0587, "step": 53370 }, { "epoch": 1.8046587105716894, "grad_norm": 0.31739047169685364, "learning_rate": 8.768052457697454e-05, "loss": 0.0579, "step": 53380 }, { "epoch": 1.8049967882619424, "grad_norm": 0.40293923020362854, "learning_rate": 8.76750897576682e-05, "loss": 0.0623, "step": 53390 }, { "epoch": 1.805334865952196, "grad_norm": 0.22571060061454773, "learning_rate": 8.766965390833907e-05, "loss": 0.0571, "step": 53400 }, { "epoch": 1.805672943642449, "grad_norm": 0.34359651803970337, "learning_rate": 8.766421702913576e-05, "loss": 0.0594, "step": 53410 }, { "epoch": 1.8060110213327023, "grad_norm": 0.39075037837028503, "learning_rate": 8.76587791202069e-05, "loss": 0.0541, "step": 53420 }, { "epoch": 1.8063490990229556, "grad_norm": 0.36175301671028137, "learning_rate": 8.765334018170118e-05, "loss": 0.0547, "step": 53430 }, { "epoch": 1.8066871767132087, "grad_norm": 0.5352059602737427, "learning_rate": 8.764790021376728e-05, "loss": 0.0564, "step": 53440 }, { "epoch": 1.807025254403462, "grad_norm": 0.6494554281234741, "learning_rate": 8.764245921655391e-05, "loss": 0.0616, "step": 53450 }, { "epoch": 1.8073633320937152, "grad_norm": 0.3508084714412689, "learning_rate": 8.763701719020987e-05, "loss": 0.0561, "step": 53460 }, { "epoch": 1.8077014097839683, "grad_norm": 0.3213661313056946, "learning_rate": 8.763157413488392e-05, "loss": 0.0574, "step": 53470 }, { "epoch": 1.8080394874742216, "grad_norm": 0.38248535990715027, "learning_rate": 8.762613005072486e-05, "loss": 0.0615, "step": 53480 }, { "epoch": 1.808377565164475, "grad_norm": 0.6902522444725037, "learning_rate": 8.762068493788154e-05, "loss": 0.0622, "step": 53490 }, { "epoch": 1.808715642854728, "grad_norm": 0.5310094952583313, "learning_rate": 8.761523879650283e-05, "loss": 0.061, "step": 53500 }, { "epoch": 1.8090537205449813, "grad_norm": 0.43892449140548706, "learning_rate": 8.760979162673763e-05, "loss": 0.0583, "step": 53510 }, { "epoch": 1.8093917982352345, "grad_norm": 0.3713394105434418, "learning_rate": 8.760434342873484e-05, "loss": 0.0607, "step": 53520 }, { "epoch": 1.8097298759254876, "grad_norm": 0.2932042181491852, "learning_rate": 8.759889420264345e-05, "loss": 0.0594, "step": 53530 }, { "epoch": 1.810067953615741, "grad_norm": 0.38905420899391174, "learning_rate": 8.75934439486124e-05, "loss": 0.0604, "step": 53540 }, { "epoch": 1.8104060313059942, "grad_norm": 0.45598283410072327, "learning_rate": 8.758799266679071e-05, "loss": 0.058, "step": 53550 }, { "epoch": 1.8107441089962473, "grad_norm": 0.4287368655204773, "learning_rate": 8.758254035732744e-05, "loss": 0.0564, "step": 53560 }, { "epoch": 1.8110821866865006, "grad_norm": 0.6698809266090393, "learning_rate": 8.757708702037161e-05, "loss": 0.0569, "step": 53570 }, { "epoch": 1.8114202643767539, "grad_norm": 0.35883885622024536, "learning_rate": 8.757163265607237e-05, "loss": 0.0655, "step": 53580 }, { "epoch": 1.811758342067007, "grad_norm": 0.8727560043334961, "learning_rate": 8.756617726457878e-05, "loss": 0.0588, "step": 53590 }, { "epoch": 1.8120964197572602, "grad_norm": 0.5323042869567871, "learning_rate": 8.756072084604004e-05, "loss": 0.0543, "step": 53600 }, { "epoch": 1.8124344974475135, "grad_norm": 0.43429574370384216, "learning_rate": 8.75552634006053e-05, "loss": 0.0598, "step": 53610 }, { "epoch": 1.8127725751377666, "grad_norm": 0.35777872800827026, "learning_rate": 8.754980492842374e-05, "loss": 0.0582, "step": 53620 }, { "epoch": 1.8131106528280199, "grad_norm": 0.339161217212677, "learning_rate": 8.754434542964465e-05, "loss": 0.0559, "step": 53630 }, { "epoch": 1.8134487305182732, "grad_norm": 0.5936354994773865, "learning_rate": 8.753888490441725e-05, "loss": 0.0586, "step": 53640 }, { "epoch": 1.8137868082085262, "grad_norm": 0.3501465320587158, "learning_rate": 8.753342335289085e-05, "loss": 0.0607, "step": 53650 }, { "epoch": 1.8141248858987795, "grad_norm": 0.5364100337028503, "learning_rate": 8.752796077521475e-05, "loss": 0.0594, "step": 53660 }, { "epoch": 1.8144629635890328, "grad_norm": 0.6511762738227844, "learning_rate": 8.752249717153829e-05, "loss": 0.0604, "step": 53670 }, { "epoch": 1.8148010412792859, "grad_norm": 0.3459228277206421, "learning_rate": 8.751703254201087e-05, "loss": 0.0584, "step": 53680 }, { "epoch": 1.8151391189695392, "grad_norm": 0.3478519320487976, "learning_rate": 8.751156688678186e-05, "loss": 0.0525, "step": 53690 }, { "epoch": 1.8154771966597925, "grad_norm": 0.39522919058799744, "learning_rate": 8.750610020600072e-05, "loss": 0.0628, "step": 53700 }, { "epoch": 1.8158152743500455, "grad_norm": 0.5613885521888733, "learning_rate": 8.750063249981687e-05, "loss": 0.0564, "step": 53710 }, { "epoch": 1.8161533520402988, "grad_norm": 0.6069098711013794, "learning_rate": 8.749516376837983e-05, "loss": 0.06, "step": 53720 }, { "epoch": 1.816491429730552, "grad_norm": 0.5035673379898071, "learning_rate": 8.74896940118391e-05, "loss": 0.0556, "step": 53730 }, { "epoch": 1.8168295074208052, "grad_norm": 0.3445453345775604, "learning_rate": 8.748422323034421e-05, "loss": 0.061, "step": 53740 }, { "epoch": 1.8171675851110585, "grad_norm": 0.6572871208190918, "learning_rate": 8.747875142404475e-05, "loss": 0.0575, "step": 53750 }, { "epoch": 1.8175056628013118, "grad_norm": 0.36755111813545227, "learning_rate": 8.74732785930903e-05, "loss": 0.0578, "step": 53760 }, { "epoch": 1.8178437404915648, "grad_norm": 0.5889915823936462, "learning_rate": 8.746780473763049e-05, "loss": 0.0544, "step": 53770 }, { "epoch": 1.8181818181818183, "grad_norm": 0.4872758984565735, "learning_rate": 8.746232985781497e-05, "loss": 0.0577, "step": 53780 }, { "epoch": 1.8185198958720714, "grad_norm": 0.47576212882995605, "learning_rate": 8.745685395379345e-05, "loss": 0.0545, "step": 53790 }, { "epoch": 1.8188579735623245, "grad_norm": 0.41969743371009827, "learning_rate": 8.745137702571558e-05, "loss": 0.0634, "step": 53800 }, { "epoch": 1.819196051252578, "grad_norm": 0.42522522807121277, "learning_rate": 8.744589907373115e-05, "loss": 0.0609, "step": 53810 }, { "epoch": 1.819534128942831, "grad_norm": 0.7229530215263367, "learning_rate": 8.744042009798991e-05, "loss": 0.0651, "step": 53820 }, { "epoch": 1.8198722066330841, "grad_norm": 0.3418863117694855, "learning_rate": 8.743494009864165e-05, "loss": 0.0616, "step": 53830 }, { "epoch": 1.8202102843233376, "grad_norm": 0.6455805897712708, "learning_rate": 8.742945907583619e-05, "loss": 0.0647, "step": 53840 }, { "epoch": 1.8205483620135907, "grad_norm": 0.5331446528434753, "learning_rate": 8.742397702972338e-05, "loss": 0.0628, "step": 53850 }, { "epoch": 1.8208864397038438, "grad_norm": 0.32143333554267883, "learning_rate": 8.74184939604531e-05, "loss": 0.0574, "step": 53860 }, { "epoch": 1.8212245173940973, "grad_norm": 0.354842871427536, "learning_rate": 8.741300986817524e-05, "loss": 0.0635, "step": 53870 }, { "epoch": 1.8215625950843504, "grad_norm": 0.4847019612789154, "learning_rate": 8.740752475303975e-05, "loss": 0.0621, "step": 53880 }, { "epoch": 1.8219006727746037, "grad_norm": 0.5553790926933289, "learning_rate": 8.74020386151966e-05, "loss": 0.0534, "step": 53890 }, { "epoch": 1.822238750464857, "grad_norm": 0.3698694109916687, "learning_rate": 8.739655145479576e-05, "loss": 0.0636, "step": 53900 }, { "epoch": 1.82257682815511, "grad_norm": 0.5282677412033081, "learning_rate": 8.739106327198724e-05, "loss": 0.0602, "step": 53910 }, { "epoch": 1.8229149058453633, "grad_norm": 0.3111046850681305, "learning_rate": 8.73855740669211e-05, "loss": 0.0536, "step": 53920 }, { "epoch": 1.8232529835356166, "grad_norm": 0.49436378479003906, "learning_rate": 8.738008383974742e-05, "loss": 0.06, "step": 53930 }, { "epoch": 1.8235910612258697, "grad_norm": 0.4360782206058502, "learning_rate": 8.737459259061629e-05, "loss": 0.0607, "step": 53940 }, { "epoch": 1.823929138916123, "grad_norm": 0.34434810280799866, "learning_rate": 8.736910031967784e-05, "loss": 0.0591, "step": 53950 }, { "epoch": 1.8242672166063763, "grad_norm": 0.42647862434387207, "learning_rate": 8.736360702708222e-05, "loss": 0.0573, "step": 53960 }, { "epoch": 1.8246052942966293, "grad_norm": 0.33957144618034363, "learning_rate": 8.735811271297963e-05, "loss": 0.0631, "step": 53970 }, { "epoch": 1.8249433719868826, "grad_norm": 0.37700513005256653, "learning_rate": 8.735261737752025e-05, "loss": 0.0574, "step": 53980 }, { "epoch": 1.825281449677136, "grad_norm": 0.3583047688007355, "learning_rate": 8.734712102085436e-05, "loss": 0.0628, "step": 53990 }, { "epoch": 1.825619527367389, "grad_norm": 0.4205942749977112, "learning_rate": 8.734162364313222e-05, "loss": 0.0611, "step": 54000 }, { "epoch": 1.8259576050576423, "grad_norm": 0.3344241976737976, "learning_rate": 8.733612524450412e-05, "loss": 0.059, "step": 54010 }, { "epoch": 1.8262956827478956, "grad_norm": 0.8146212697029114, "learning_rate": 8.733062582512037e-05, "loss": 0.0614, "step": 54020 }, { "epoch": 1.8266337604381486, "grad_norm": 0.6001678109169006, "learning_rate": 8.732512538513134e-05, "loss": 0.0624, "step": 54030 }, { "epoch": 1.826971838128402, "grad_norm": 0.4552026391029358, "learning_rate": 8.73196239246874e-05, "loss": 0.0573, "step": 54040 }, { "epoch": 1.8273099158186552, "grad_norm": 0.40594181418418884, "learning_rate": 8.731412144393896e-05, "loss": 0.0589, "step": 54050 }, { "epoch": 1.8276479935089083, "grad_norm": 0.32372209429740906, "learning_rate": 8.730861794303646e-05, "loss": 0.0576, "step": 54060 }, { "epoch": 1.8279860711991616, "grad_norm": 0.2277791053056717, "learning_rate": 8.730311342213036e-05, "loss": 0.0552, "step": 54070 }, { "epoch": 1.8283241488894149, "grad_norm": 0.3369303047657013, "learning_rate": 8.729760788137116e-05, "loss": 0.0626, "step": 54080 }, { "epoch": 1.828662226579668, "grad_norm": 0.4721658527851105, "learning_rate": 8.729210132090938e-05, "loss": 0.0602, "step": 54090 }, { "epoch": 1.8290003042699212, "grad_norm": 0.28479740023612976, "learning_rate": 8.728659374089554e-05, "loss": 0.0576, "step": 54100 }, { "epoch": 1.8293383819601745, "grad_norm": 0.6782993674278259, "learning_rate": 8.728108514148023e-05, "loss": 0.0612, "step": 54110 }, { "epoch": 1.8296764596504276, "grad_norm": 0.2522174119949341, "learning_rate": 8.727557552281408e-05, "loss": 0.0503, "step": 54120 }, { "epoch": 1.8300145373406809, "grad_norm": 0.5055302381515503, "learning_rate": 8.72700648850477e-05, "loss": 0.0601, "step": 54130 }, { "epoch": 1.8303526150309342, "grad_norm": 1.0167561769485474, "learning_rate": 8.726455322833174e-05, "loss": 0.0532, "step": 54140 }, { "epoch": 1.8306906927211872, "grad_norm": 0.29456546902656555, "learning_rate": 8.725904055281688e-05, "loss": 0.0523, "step": 54150 }, { "epoch": 1.8310287704114405, "grad_norm": 0.3650056719779968, "learning_rate": 8.725352685865387e-05, "loss": 0.056, "step": 54160 }, { "epoch": 1.8313668481016938, "grad_norm": 0.4116411507129669, "learning_rate": 8.724801214599343e-05, "loss": 0.0632, "step": 54170 }, { "epoch": 1.8317049257919469, "grad_norm": 0.36849161982536316, "learning_rate": 8.724249641498632e-05, "loss": 0.0568, "step": 54180 }, { "epoch": 1.8320430034822002, "grad_norm": 0.44495970010757446, "learning_rate": 8.723697966578334e-05, "loss": 0.0606, "step": 54190 }, { "epoch": 1.8323810811724535, "grad_norm": 0.3134077489376068, "learning_rate": 8.723146189853535e-05, "loss": 0.0566, "step": 54200 }, { "epoch": 1.8327191588627065, "grad_norm": 0.616033136844635, "learning_rate": 8.722594311339317e-05, "loss": 0.0542, "step": 54210 }, { "epoch": 1.8330572365529598, "grad_norm": 0.3635402321815491, "learning_rate": 8.722042331050768e-05, "loss": 0.0616, "step": 54220 }, { "epoch": 1.8333953142432131, "grad_norm": 0.356032133102417, "learning_rate": 8.721490249002982e-05, "loss": 0.0592, "step": 54230 }, { "epoch": 1.8337333919334662, "grad_norm": 0.2544962167739868, "learning_rate": 8.72093806521105e-05, "loss": 0.0519, "step": 54240 }, { "epoch": 1.8340714696237197, "grad_norm": 0.41127637028694153, "learning_rate": 8.720385779690067e-05, "loss": 0.062, "step": 54250 }, { "epoch": 1.8344095473139728, "grad_norm": 0.2739834487438202, "learning_rate": 8.719833392455137e-05, "loss": 0.0604, "step": 54260 }, { "epoch": 1.8347476250042258, "grad_norm": 0.32791829109191895, "learning_rate": 8.719280903521358e-05, "loss": 0.0615, "step": 54270 }, { "epoch": 1.8350857026944793, "grad_norm": 0.3312573730945587, "learning_rate": 8.718728312903838e-05, "loss": 0.0559, "step": 54280 }, { "epoch": 1.8354237803847324, "grad_norm": 0.41440144181251526, "learning_rate": 8.718175620617681e-05, "loss": 0.0593, "step": 54290 }, { "epoch": 1.8357618580749855, "grad_norm": 0.2972918748855591, "learning_rate": 8.717622826678001e-05, "loss": 0.0532, "step": 54300 }, { "epoch": 1.836099935765239, "grad_norm": 0.21403174102306366, "learning_rate": 8.717069931099908e-05, "loss": 0.0608, "step": 54310 }, { "epoch": 1.836438013455492, "grad_norm": 0.22467546164989471, "learning_rate": 8.716516933898521e-05, "loss": 0.0618, "step": 54320 }, { "epoch": 1.8367760911457451, "grad_norm": 0.643121600151062, "learning_rate": 8.715963835088959e-05, "loss": 0.0572, "step": 54330 }, { "epoch": 1.8371141688359987, "grad_norm": 0.4421980082988739, "learning_rate": 8.715410634686339e-05, "loss": 0.0591, "step": 54340 }, { "epoch": 1.8374522465262517, "grad_norm": 0.34221869707107544, "learning_rate": 8.714857332705788e-05, "loss": 0.0585, "step": 54350 }, { "epoch": 1.8377903242165048, "grad_norm": 0.314264714717865, "learning_rate": 8.714303929162433e-05, "loss": 0.0545, "step": 54360 }, { "epoch": 1.8381284019067583, "grad_norm": 0.3687426447868347, "learning_rate": 8.713750424071405e-05, "loss": 0.0607, "step": 54370 }, { "epoch": 1.8384664795970114, "grad_norm": 0.35363873839378357, "learning_rate": 8.713196817447835e-05, "loss": 0.0561, "step": 54380 }, { "epoch": 1.8388045572872647, "grad_norm": 0.2378734052181244, "learning_rate": 8.71264310930686e-05, "loss": 0.0539, "step": 54390 }, { "epoch": 1.839142634977518, "grad_norm": 0.2926294803619385, "learning_rate": 8.712089299663617e-05, "loss": 0.0596, "step": 54400 }, { "epoch": 1.839480712667771, "grad_norm": 0.4974691569805145, "learning_rate": 8.711535388533246e-05, "loss": 0.0569, "step": 54410 }, { "epoch": 1.8398187903580243, "grad_norm": 0.2696552276611328, "learning_rate": 8.710981375930893e-05, "loss": 0.0536, "step": 54420 }, { "epoch": 1.8401568680482776, "grad_norm": 0.5110090970993042, "learning_rate": 8.710427261871703e-05, "loss": 0.0599, "step": 54430 }, { "epoch": 1.8404949457385307, "grad_norm": 0.37077796459198, "learning_rate": 8.709873046370825e-05, "loss": 0.0659, "step": 54440 }, { "epoch": 1.840833023428784, "grad_norm": 0.27176332473754883, "learning_rate": 8.709318729443411e-05, "loss": 0.0575, "step": 54450 }, { "epoch": 1.8411711011190373, "grad_norm": 0.331966757774353, "learning_rate": 8.708764311104618e-05, "loss": 0.0581, "step": 54460 }, { "epoch": 1.8415091788092903, "grad_norm": 0.29113781452178955, "learning_rate": 8.708209791369601e-05, "loss": 0.0589, "step": 54470 }, { "epoch": 1.8418472564995436, "grad_norm": 0.20615839958190918, "learning_rate": 8.707655170253524e-05, "loss": 0.0532, "step": 54480 }, { "epoch": 1.842185334189797, "grad_norm": 0.2971593737602234, "learning_rate": 8.707100447771545e-05, "loss": 0.0605, "step": 54490 }, { "epoch": 1.84252341188005, "grad_norm": 0.445136696100235, "learning_rate": 8.706545623938832e-05, "loss": 0.0553, "step": 54500 }, { "epoch": 1.8428614895703033, "grad_norm": 0.44568076729774475, "learning_rate": 8.705990698770556e-05, "loss": 0.0622, "step": 54510 }, { "epoch": 1.8431995672605566, "grad_norm": 1.8309743404388428, "learning_rate": 8.705435672281886e-05, "loss": 0.0614, "step": 54520 }, { "epoch": 1.8435376449508096, "grad_norm": 0.5214974284172058, "learning_rate": 8.704880544487995e-05, "loss": 0.0571, "step": 54530 }, { "epoch": 1.843875722641063, "grad_norm": 0.370595782995224, "learning_rate": 8.704325315404063e-05, "loss": 0.0625, "step": 54540 }, { "epoch": 1.8442138003313162, "grad_norm": 0.25815001130104065, "learning_rate": 8.70376998504527e-05, "loss": 0.0561, "step": 54550 }, { "epoch": 1.8445518780215693, "grad_norm": 0.6889346837997437, "learning_rate": 8.703214553426795e-05, "loss": 0.0589, "step": 54560 }, { "epoch": 1.8448899557118226, "grad_norm": 0.3874379098415375, "learning_rate": 8.702659020563826e-05, "loss": 0.0548, "step": 54570 }, { "epoch": 1.8452280334020759, "grad_norm": 0.23904535174369812, "learning_rate": 8.702103386471551e-05, "loss": 0.0577, "step": 54580 }, { "epoch": 1.845566111092329, "grad_norm": 0.37345823645591736, "learning_rate": 8.70154765116516e-05, "loss": 0.0546, "step": 54590 }, { "epoch": 1.8459041887825822, "grad_norm": 0.5641502737998962, "learning_rate": 8.700991814659847e-05, "loss": 0.0569, "step": 54600 }, { "epoch": 1.8462422664728355, "grad_norm": 0.39111486077308655, "learning_rate": 8.700435876970807e-05, "loss": 0.0566, "step": 54610 }, { "epoch": 1.8465803441630886, "grad_norm": 0.24781011044979095, "learning_rate": 8.69987983811324e-05, "loss": 0.0539, "step": 54620 }, { "epoch": 1.8469184218533419, "grad_norm": 0.43902587890625, "learning_rate": 8.699323698102351e-05, "loss": 0.0546, "step": 54630 }, { "epoch": 1.8472564995435952, "grad_norm": 0.39534908533096313, "learning_rate": 8.69876745695334e-05, "loss": 0.0557, "step": 54640 }, { "epoch": 1.8475945772338482, "grad_norm": 0.6932068467140198, "learning_rate": 8.698211114681417e-05, "loss": 0.0552, "step": 54650 }, { "epoch": 1.8479326549241015, "grad_norm": 0.34182319045066833, "learning_rate": 8.697654671301789e-05, "loss": 0.0554, "step": 54660 }, { "epoch": 1.8482707326143548, "grad_norm": 0.5439560413360596, "learning_rate": 8.697098126829675e-05, "loss": 0.0626, "step": 54670 }, { "epoch": 1.8486088103046079, "grad_norm": 0.31380271911621094, "learning_rate": 8.696541481280285e-05, "loss": 0.0554, "step": 54680 }, { "epoch": 1.8489468879948612, "grad_norm": 0.3275999128818512, "learning_rate": 8.69598473466884e-05, "loss": 0.057, "step": 54690 }, { "epoch": 1.8492849656851145, "grad_norm": 2.015157699584961, "learning_rate": 8.69542788701056e-05, "loss": 0.0551, "step": 54700 }, { "epoch": 1.8496230433753675, "grad_norm": 0.3256222903728485, "learning_rate": 8.694870938320671e-05, "loss": 0.0553, "step": 54710 }, { "epoch": 1.8499611210656208, "grad_norm": 0.8289569616317749, "learning_rate": 8.6943138886144e-05, "loss": 0.0541, "step": 54720 }, { "epoch": 1.8502991987558741, "grad_norm": 0.5885155200958252, "learning_rate": 8.693756737906972e-05, "loss": 0.0652, "step": 54730 }, { "epoch": 1.8506372764461272, "grad_norm": 0.4734085202217102, "learning_rate": 8.693199486213626e-05, "loss": 0.0558, "step": 54740 }, { "epoch": 1.8509753541363807, "grad_norm": 0.8158175349235535, "learning_rate": 8.692642133549591e-05, "loss": 0.0548, "step": 54750 }, { "epoch": 1.8513134318266338, "grad_norm": 0.6023409366607666, "learning_rate": 8.692084679930108e-05, "loss": 0.064, "step": 54760 }, { "epoch": 1.8516515095168868, "grad_norm": 0.3790176212787628, "learning_rate": 8.691527125370417e-05, "loss": 0.0618, "step": 54770 }, { "epoch": 1.8519895872071404, "grad_norm": 0.49755290150642395, "learning_rate": 8.690969469885761e-05, "loss": 0.0568, "step": 54780 }, { "epoch": 1.8523276648973934, "grad_norm": 0.2837907671928406, "learning_rate": 8.690411713491386e-05, "loss": 0.0623, "step": 54790 }, { "epoch": 1.8526657425876465, "grad_norm": 0.3112846612930298, "learning_rate": 8.689853856202541e-05, "loss": 0.0542, "step": 54800 }, { "epoch": 1.8530038202779, "grad_norm": 0.4099988043308258, "learning_rate": 8.68929589803448e-05, "loss": 0.0553, "step": 54810 }, { "epoch": 1.853341897968153, "grad_norm": 0.4920674264431, "learning_rate": 8.688737839002452e-05, "loss": 0.0565, "step": 54820 }, { "epoch": 1.8536799756584061, "grad_norm": 0.421640008687973, "learning_rate": 8.688179679121719e-05, "loss": 0.0673, "step": 54830 }, { "epoch": 1.8540180533486597, "grad_norm": 0.27963787317276, "learning_rate": 8.68762141840754e-05, "loss": 0.053, "step": 54840 }, { "epoch": 1.8543561310389127, "grad_norm": 1.1883231401443481, "learning_rate": 8.687063056875175e-05, "loss": 0.0592, "step": 54850 }, { "epoch": 1.854694208729166, "grad_norm": 0.6315132975578308, "learning_rate": 8.686504594539892e-05, "loss": 0.0521, "step": 54860 }, { "epoch": 1.8550322864194193, "grad_norm": 0.5337653756141663, "learning_rate": 8.68594603141696e-05, "loss": 0.059, "step": 54870 }, { "epoch": 1.8553703641096724, "grad_norm": 0.2931552231311798, "learning_rate": 8.685387367521646e-05, "loss": 0.0569, "step": 54880 }, { "epoch": 1.8557084417999257, "grad_norm": 0.39551350474357605, "learning_rate": 8.684828602869226e-05, "loss": 0.0623, "step": 54890 }, { "epoch": 1.856046519490179, "grad_norm": 0.3117876648902893, "learning_rate": 8.684269737474976e-05, "loss": 0.0572, "step": 54900 }, { "epoch": 1.856384597180432, "grad_norm": 0.3097190260887146, "learning_rate": 8.683710771354178e-05, "loss": 0.0551, "step": 54910 }, { "epoch": 1.8567226748706853, "grad_norm": 0.31240367889404297, "learning_rate": 8.683151704522109e-05, "loss": 0.0595, "step": 54920 }, { "epoch": 1.8570607525609386, "grad_norm": 0.36431553959846497, "learning_rate": 8.682592536994057e-05, "loss": 0.0541, "step": 54930 }, { "epoch": 1.8573988302511917, "grad_norm": 0.32013940811157227, "learning_rate": 8.682033268785308e-05, "loss": 0.0542, "step": 54940 }, { "epoch": 1.857736907941445, "grad_norm": 0.22759482264518738, "learning_rate": 8.681473899911154e-05, "loss": 0.0504, "step": 54950 }, { "epoch": 1.8580749856316983, "grad_norm": 0.3010881245136261, "learning_rate": 8.680914430386885e-05, "loss": 0.0552, "step": 54960 }, { "epoch": 1.8584130633219513, "grad_norm": 0.5570321083068848, "learning_rate": 8.6803548602278e-05, "loss": 0.0618, "step": 54970 }, { "epoch": 1.8587511410122046, "grad_norm": 0.36478734016418457, "learning_rate": 8.679795189449193e-05, "loss": 0.059, "step": 54980 }, { "epoch": 1.859089218702458, "grad_norm": 0.3603401780128479, "learning_rate": 8.679235418066371e-05, "loss": 0.0535, "step": 54990 }, { "epoch": 1.859427296392711, "grad_norm": 0.29230690002441406, "learning_rate": 8.678675546094633e-05, "loss": 0.0606, "step": 55000 }, { "epoch": 1.8597653740829643, "grad_norm": 0.3592151403427124, "learning_rate": 8.678115573549288e-05, "loss": 0.061, "step": 55010 }, { "epoch": 1.8601034517732176, "grad_norm": 0.6628940105438232, "learning_rate": 8.677555500445645e-05, "loss": 0.0595, "step": 55020 }, { "epoch": 1.8604415294634706, "grad_norm": 0.32866162061691284, "learning_rate": 8.676995326799017e-05, "loss": 0.0669, "step": 55030 }, { "epoch": 1.860779607153724, "grad_norm": 0.3469339907169342, "learning_rate": 8.676435052624719e-05, "loss": 0.0576, "step": 55040 }, { "epoch": 1.8611176848439772, "grad_norm": 0.418516606092453, "learning_rate": 8.675874677938066e-05, "loss": 0.0615, "step": 55050 }, { "epoch": 1.8614557625342303, "grad_norm": 0.38086196780204773, "learning_rate": 8.675314202754379e-05, "loss": 0.0565, "step": 55060 }, { "epoch": 1.8617938402244836, "grad_norm": 0.5033257007598877, "learning_rate": 8.674753627088985e-05, "loss": 0.0582, "step": 55070 }, { "epoch": 1.8621319179147369, "grad_norm": 0.3846341073513031, "learning_rate": 8.674192950957207e-05, "loss": 0.0557, "step": 55080 }, { "epoch": 1.86246999560499, "grad_norm": 0.44157078862190247, "learning_rate": 8.673632174374373e-05, "loss": 0.0606, "step": 55090 }, { "epoch": 1.8628080732952432, "grad_norm": 0.4563905894756317, "learning_rate": 8.673071297355817e-05, "loss": 0.0541, "step": 55100 }, { "epoch": 1.8631461509854965, "grad_norm": 0.33313432335853577, "learning_rate": 8.67251031991687e-05, "loss": 0.0503, "step": 55110 }, { "epoch": 1.8634842286757496, "grad_norm": 0.45702818036079407, "learning_rate": 8.671949242072873e-05, "loss": 0.0589, "step": 55120 }, { "epoch": 1.8638223063660029, "grad_norm": 0.3427799642086029, "learning_rate": 8.67138806383916e-05, "loss": 0.0563, "step": 55130 }, { "epoch": 1.8641603840562562, "grad_norm": 0.3165248930454254, "learning_rate": 8.670826785231078e-05, "loss": 0.06, "step": 55140 }, { "epoch": 1.8644984617465092, "grad_norm": 0.3667582869529724, "learning_rate": 8.670265406263971e-05, "loss": 0.0555, "step": 55150 }, { "epoch": 1.8648365394367625, "grad_norm": 0.31200161576271057, "learning_rate": 8.669703926953187e-05, "loss": 0.0594, "step": 55160 }, { "epoch": 1.8651746171270158, "grad_norm": 0.47233885526657104, "learning_rate": 8.669142347314074e-05, "loss": 0.0611, "step": 55170 }, { "epoch": 1.865512694817269, "grad_norm": 0.26554206013679504, "learning_rate": 8.668580667361991e-05, "loss": 0.0562, "step": 55180 }, { "epoch": 1.8658507725075222, "grad_norm": 0.5310185551643372, "learning_rate": 8.668018887112287e-05, "loss": 0.0628, "step": 55190 }, { "epoch": 1.8661888501977755, "grad_norm": 1.117676854133606, "learning_rate": 8.667457006580328e-05, "loss": 0.0589, "step": 55200 }, { "epoch": 1.8665269278880285, "grad_norm": 0.44501569867134094, "learning_rate": 8.666895025781471e-05, "loss": 0.0614, "step": 55210 }, { "epoch": 1.866865005578282, "grad_norm": 0.388232558965683, "learning_rate": 8.666332944731083e-05, "loss": 0.0616, "step": 55220 }, { "epoch": 1.8672030832685351, "grad_norm": 0.306672602891922, "learning_rate": 8.665770763444527e-05, "loss": 0.0619, "step": 55230 }, { "epoch": 1.8675411609587882, "grad_norm": 0.23165595531463623, "learning_rate": 8.665208481937176e-05, "loss": 0.0628, "step": 55240 }, { "epoch": 1.8678792386490417, "grad_norm": 0.33113792538642883, "learning_rate": 8.664646100224404e-05, "loss": 0.0607, "step": 55250 }, { "epoch": 1.8682173163392948, "grad_norm": 0.34110400080680847, "learning_rate": 8.664083618321583e-05, "loss": 0.0527, "step": 55260 }, { "epoch": 1.8685553940295478, "grad_norm": 0.21203766763210297, "learning_rate": 8.663521036244093e-05, "loss": 0.055, "step": 55270 }, { "epoch": 1.8688934717198014, "grad_norm": 0.35206425189971924, "learning_rate": 8.662958354007313e-05, "loss": 0.062, "step": 55280 }, { "epoch": 1.8692315494100544, "grad_norm": 0.26029902696609497, "learning_rate": 8.662395571626629e-05, "loss": 0.0589, "step": 55290 }, { "epoch": 1.8695696271003075, "grad_norm": 0.3169742226600647, "learning_rate": 8.661832689117426e-05, "loss": 0.0617, "step": 55300 }, { "epoch": 1.869907704790561, "grad_norm": 0.3944316804409027, "learning_rate": 8.661269706495092e-05, "loss": 0.0565, "step": 55310 }, { "epoch": 1.870245782480814, "grad_norm": 0.4801885485649109, "learning_rate": 8.660706623775021e-05, "loss": 0.0585, "step": 55320 }, { "epoch": 1.8705838601710671, "grad_norm": 0.32384222745895386, "learning_rate": 8.660143440972604e-05, "loss": 0.0604, "step": 55330 }, { "epoch": 1.8709219378613207, "grad_norm": 0.28702014684677124, "learning_rate": 8.659580158103242e-05, "loss": 0.0609, "step": 55340 }, { "epoch": 1.8712600155515737, "grad_norm": 0.44317036867141724, "learning_rate": 8.659016775182333e-05, "loss": 0.0557, "step": 55350 }, { "epoch": 1.871598093241827, "grad_norm": 0.35307449102401733, "learning_rate": 8.658453292225282e-05, "loss": 0.0541, "step": 55360 }, { "epoch": 1.8719361709320803, "grad_norm": 0.3110165297985077, "learning_rate": 8.657889709247489e-05, "loss": 0.0584, "step": 55370 }, { "epoch": 1.8722742486223334, "grad_norm": 0.31551891565322876, "learning_rate": 8.657326026264367e-05, "loss": 0.0544, "step": 55380 }, { "epoch": 1.8726123263125867, "grad_norm": 0.4439314007759094, "learning_rate": 8.656762243291325e-05, "loss": 0.06, "step": 55390 }, { "epoch": 1.87295040400284, "grad_norm": 0.31464749574661255, "learning_rate": 8.656198360343778e-05, "loss": 0.0582, "step": 55400 }, { "epoch": 1.873288481693093, "grad_norm": 0.4142618477344513, "learning_rate": 8.655634377437141e-05, "loss": 0.0554, "step": 55410 }, { "epoch": 1.8736265593833463, "grad_norm": 0.5455803871154785, "learning_rate": 8.655070294586833e-05, "loss": 0.0593, "step": 55420 }, { "epoch": 1.8739646370735996, "grad_norm": 0.27185067534446716, "learning_rate": 8.654506111808277e-05, "loss": 0.0566, "step": 55430 }, { "epoch": 1.8743027147638527, "grad_norm": 1.4835598468780518, "learning_rate": 8.653941829116896e-05, "loss": 0.0539, "step": 55440 }, { "epoch": 1.874640792454106, "grad_norm": 0.34207847714424133, "learning_rate": 8.653377446528119e-05, "loss": 0.0652, "step": 55450 }, { "epoch": 1.8749788701443593, "grad_norm": 0.2527855634689331, "learning_rate": 8.652812964057376e-05, "loss": 0.0571, "step": 55460 }, { "epoch": 1.8753169478346123, "grad_norm": 0.45473939180374146, "learning_rate": 8.652248381720097e-05, "loss": 0.058, "step": 55470 }, { "epoch": 1.8756550255248656, "grad_norm": 0.3986496031284332, "learning_rate": 8.65168369953172e-05, "loss": 0.0623, "step": 55480 }, { "epoch": 1.875993103215119, "grad_norm": 1.9175567626953125, "learning_rate": 8.651118917507681e-05, "loss": 0.0602, "step": 55490 }, { "epoch": 1.876331180905372, "grad_norm": 0.8268141746520996, "learning_rate": 8.650554035663425e-05, "loss": 0.055, "step": 55500 }, { "epoch": 1.8766692585956253, "grad_norm": 0.291044682264328, "learning_rate": 8.649989054014392e-05, "loss": 0.0598, "step": 55510 }, { "epoch": 1.8770073362858786, "grad_norm": 0.3186872601509094, "learning_rate": 8.64942397257603e-05, "loss": 0.0642, "step": 55520 }, { "epoch": 1.8773454139761316, "grad_norm": 0.35274738073349, "learning_rate": 8.648858791363786e-05, "loss": 0.0532, "step": 55530 }, { "epoch": 1.877683491666385, "grad_norm": 0.6953719258308411, "learning_rate": 8.648293510393115e-05, "loss": 0.0539, "step": 55540 }, { "epoch": 1.8780215693566382, "grad_norm": 0.28611916303634644, "learning_rate": 8.647728129679469e-05, "loss": 0.0565, "step": 55550 }, { "epoch": 1.8783596470468913, "grad_norm": 0.35885801911354065, "learning_rate": 8.647162649238307e-05, "loss": 0.058, "step": 55560 }, { "epoch": 1.8786977247371446, "grad_norm": 0.30378541350364685, "learning_rate": 8.64659706908509e-05, "loss": 0.0633, "step": 55570 }, { "epoch": 1.8790358024273979, "grad_norm": 0.39339494705200195, "learning_rate": 8.646031389235277e-05, "loss": 0.053, "step": 55580 }, { "epoch": 1.879373880117651, "grad_norm": 0.43804773688316345, "learning_rate": 8.645465609704335e-05, "loss": 0.0593, "step": 55590 }, { "epoch": 1.8797119578079042, "grad_norm": 0.3862711787223816, "learning_rate": 8.644899730507734e-05, "loss": 0.0613, "step": 55600 }, { "epoch": 1.8800500354981575, "grad_norm": 0.7840898633003235, "learning_rate": 8.644333751660944e-05, "loss": 0.0587, "step": 55610 }, { "epoch": 1.8803881131884106, "grad_norm": 0.2888069450855255, "learning_rate": 8.643767673179437e-05, "loss": 0.0522, "step": 55620 }, { "epoch": 1.8807261908786639, "grad_norm": 0.4762769639492035, "learning_rate": 8.643201495078692e-05, "loss": 0.0611, "step": 55630 }, { "epoch": 1.8810642685689172, "grad_norm": 0.39798614382743835, "learning_rate": 8.642635217374186e-05, "loss": 0.0669, "step": 55640 }, { "epoch": 1.8814023462591702, "grad_norm": 0.36497291922569275, "learning_rate": 8.6420688400814e-05, "loss": 0.0583, "step": 55650 }, { "epoch": 1.8817404239494235, "grad_norm": 0.2994627058506012, "learning_rate": 8.641502363215823e-05, "loss": 0.0622, "step": 55660 }, { "epoch": 1.8820785016396768, "grad_norm": 0.4874523878097534, "learning_rate": 8.64093578679294e-05, "loss": 0.0683, "step": 55670 }, { "epoch": 1.88241657932993, "grad_norm": 0.6189148426055908, "learning_rate": 8.640369110828238e-05, "loss": 0.0617, "step": 55680 }, { "epoch": 1.8827546570201832, "grad_norm": 0.6595147848129272, "learning_rate": 8.639802335337214e-05, "loss": 0.0554, "step": 55690 }, { "epoch": 1.8830927347104365, "grad_norm": 0.3708440959453583, "learning_rate": 8.63923546033536e-05, "loss": 0.0569, "step": 55700 }, { "epoch": 1.8834308124006895, "grad_norm": 0.6131412386894226, "learning_rate": 8.638668485838177e-05, "loss": 0.0601, "step": 55710 }, { "epoch": 1.883768890090943, "grad_norm": 0.5008652806282043, "learning_rate": 8.638101411861163e-05, "loss": 0.0553, "step": 55720 }, { "epoch": 1.8841069677811961, "grad_norm": 0.35773083567619324, "learning_rate": 8.637534238419825e-05, "loss": 0.0673, "step": 55730 }, { "epoch": 1.8844450454714492, "grad_norm": 0.4378318786621094, "learning_rate": 8.636966965529665e-05, "loss": 0.0659, "step": 55740 }, { "epoch": 1.8847831231617027, "grad_norm": 0.5352568626403809, "learning_rate": 8.636399593206197e-05, "loss": 0.0654, "step": 55750 }, { "epoch": 1.8851212008519558, "grad_norm": 1.099685549736023, "learning_rate": 8.63583212146493e-05, "loss": 0.059, "step": 55760 }, { "epoch": 1.8854592785422088, "grad_norm": 0.36791354417800903, "learning_rate": 8.635264550321378e-05, "loss": 0.0541, "step": 55770 }, { "epoch": 1.8857973562324624, "grad_norm": 0.35587283968925476, "learning_rate": 8.634696879791058e-05, "loss": 0.0552, "step": 55780 }, { "epoch": 1.8861354339227154, "grad_norm": 0.4054773151874542, "learning_rate": 8.634129109889493e-05, "loss": 0.0556, "step": 55790 }, { "epoch": 1.8864735116129685, "grad_norm": 0.4578629732131958, "learning_rate": 8.633561240632202e-05, "loss": 0.0587, "step": 55800 }, { "epoch": 1.886811589303222, "grad_norm": 0.5161023139953613, "learning_rate": 8.63299327203471e-05, "loss": 0.0635, "step": 55810 }, { "epoch": 1.887149666993475, "grad_norm": 0.47299522161483765, "learning_rate": 8.632425204112551e-05, "loss": 0.0564, "step": 55820 }, { "epoch": 1.8874877446837284, "grad_norm": 0.35877442359924316, "learning_rate": 8.631857036881248e-05, "loss": 0.0568, "step": 55830 }, { "epoch": 1.8878258223739817, "grad_norm": 0.4123840630054474, "learning_rate": 8.631288770356339e-05, "loss": 0.0577, "step": 55840 }, { "epoch": 1.8881639000642347, "grad_norm": 0.3537774682044983, "learning_rate": 8.63072040455336e-05, "loss": 0.0532, "step": 55850 }, { "epoch": 1.888501977754488, "grad_norm": 0.3659488260746002, "learning_rate": 8.630151939487849e-05, "loss": 0.0546, "step": 55860 }, { "epoch": 1.8888400554447413, "grad_norm": 0.4091181755065918, "learning_rate": 8.629583375175347e-05, "loss": 0.0553, "step": 55870 }, { "epoch": 1.8891781331349944, "grad_norm": 0.3601999580860138, "learning_rate": 8.6290147116314e-05, "loss": 0.056, "step": 55880 }, { "epoch": 1.8895162108252477, "grad_norm": 0.35396575927734375, "learning_rate": 8.628445948871553e-05, "loss": 0.0516, "step": 55890 }, { "epoch": 1.889854288515501, "grad_norm": 0.37033596634864807, "learning_rate": 8.627877086911358e-05, "loss": 0.0572, "step": 55900 }, { "epoch": 1.890192366205754, "grad_norm": 0.3911677598953247, "learning_rate": 8.627308125766366e-05, "loss": 0.0556, "step": 55910 }, { "epoch": 1.8905304438960073, "grad_norm": 0.313394159078598, "learning_rate": 8.626739065452132e-05, "loss": 0.0577, "step": 55920 }, { "epoch": 1.8908685215862606, "grad_norm": 0.2643034756183624, "learning_rate": 8.626169905984216e-05, "loss": 0.0513, "step": 55930 }, { "epoch": 1.8912065992765137, "grad_norm": 0.23444460332393646, "learning_rate": 8.625600647378175e-05, "loss": 0.0595, "step": 55940 }, { "epoch": 1.891544676966767, "grad_norm": 0.6218526363372803, "learning_rate": 8.625031289649576e-05, "loss": 0.0554, "step": 55950 }, { "epoch": 1.8918827546570203, "grad_norm": 0.3305155634880066, "learning_rate": 8.624461832813983e-05, "loss": 0.0608, "step": 55960 }, { "epoch": 1.8922208323472733, "grad_norm": 0.4630461037158966, "learning_rate": 8.623892276886964e-05, "loss": 0.0568, "step": 55970 }, { "epoch": 1.8925589100375266, "grad_norm": 0.4202554225921631, "learning_rate": 8.623322621884094e-05, "loss": 0.0586, "step": 55980 }, { "epoch": 1.89289698772778, "grad_norm": 0.606776773929596, "learning_rate": 8.622752867820942e-05, "loss": 0.0543, "step": 55990 }, { "epoch": 1.893235065418033, "grad_norm": 0.6030804514884949, "learning_rate": 8.622183014713089e-05, "loss": 0.0626, "step": 56000 }, { "epoch": 1.8935731431082863, "grad_norm": 0.46859806776046753, "learning_rate": 8.621613062576113e-05, "loss": 0.0556, "step": 56010 }, { "epoch": 1.8939112207985396, "grad_norm": 0.8954836130142212, "learning_rate": 8.621043011425597e-05, "loss": 0.0559, "step": 56020 }, { "epoch": 1.8942492984887926, "grad_norm": 0.374081015586853, "learning_rate": 8.620472861277125e-05, "loss": 0.0523, "step": 56030 }, { "epoch": 1.894587376179046, "grad_norm": 0.32226425409317017, "learning_rate": 8.619902612146284e-05, "loss": 0.0593, "step": 56040 }, { "epoch": 1.8949254538692992, "grad_norm": 0.4064125418663025, "learning_rate": 8.619332264048667e-05, "loss": 0.0642, "step": 56050 }, { "epoch": 1.8952635315595523, "grad_norm": 0.5090208053588867, "learning_rate": 8.618761816999865e-05, "loss": 0.0572, "step": 56060 }, { "epoch": 1.8956016092498056, "grad_norm": 0.35824471712112427, "learning_rate": 8.618191271015475e-05, "loss": 0.0552, "step": 56070 }, { "epoch": 1.8959396869400589, "grad_norm": 0.5264485478401184, "learning_rate": 8.617620626111094e-05, "loss": 0.0606, "step": 56080 }, { "epoch": 1.896277764630312, "grad_norm": 0.3909195065498352, "learning_rate": 8.617049882302326e-05, "loss": 0.0672, "step": 56090 }, { "epoch": 1.8966158423205652, "grad_norm": 0.2560864984989166, "learning_rate": 8.616479039604771e-05, "loss": 0.0643, "step": 56100 }, { "epoch": 1.8969539200108185, "grad_norm": 0.634775698184967, "learning_rate": 8.615908098034038e-05, "loss": 0.06, "step": 56110 }, { "epoch": 1.8972919977010716, "grad_norm": 1.8103867769241333, "learning_rate": 8.615337057605736e-05, "loss": 0.0607, "step": 56120 }, { "epoch": 1.8976300753913249, "grad_norm": 1.0950685739517212, "learning_rate": 8.614765918335479e-05, "loss": 0.0584, "step": 56130 }, { "epoch": 1.8979681530815782, "grad_norm": 0.7970686554908752, "learning_rate": 8.614194680238878e-05, "loss": 0.0627, "step": 56140 }, { "epoch": 1.8983062307718312, "grad_norm": 2.831230401992798, "learning_rate": 8.613623343331553e-05, "loss": 0.0642, "step": 56150 }, { "epoch": 1.8986443084620845, "grad_norm": 0.7503564953804016, "learning_rate": 8.613051907629123e-05, "loss": 0.0628, "step": 56160 }, { "epoch": 1.8989823861523378, "grad_norm": 0.5488095283508301, "learning_rate": 8.612480373147211e-05, "loss": 0.0713, "step": 56170 }, { "epoch": 1.899320463842591, "grad_norm": 0.48048049211502075, "learning_rate": 8.611908739901444e-05, "loss": 0.0677, "step": 56180 }, { "epoch": 1.8996585415328444, "grad_norm": 0.6830655336380005, "learning_rate": 8.611337007907448e-05, "loss": 0.0581, "step": 56190 }, { "epoch": 1.8999966192230975, "grad_norm": 1.7601631879806519, "learning_rate": 8.610765177180853e-05, "loss": 0.0662, "step": 56200 }, { "epoch": 1.9003346969133506, "grad_norm": 0.5331864356994629, "learning_rate": 8.610193247737297e-05, "loss": 0.066, "step": 56210 }, { "epoch": 1.900672774603604, "grad_norm": 0.5701102614402771, "learning_rate": 8.609621219592413e-05, "loss": 0.0606, "step": 56220 }, { "epoch": 1.9010108522938571, "grad_norm": 0.7447413802146912, "learning_rate": 8.60904909276184e-05, "loss": 0.0638, "step": 56230 }, { "epoch": 1.9013489299841102, "grad_norm": 0.4667551815509796, "learning_rate": 8.608476867261224e-05, "loss": 0.0621, "step": 56240 }, { "epoch": 1.9016870076743637, "grad_norm": 1.063193917274475, "learning_rate": 8.607904543106202e-05, "loss": 0.0583, "step": 56250 }, { "epoch": 1.9020250853646168, "grad_norm": 0.43474504351615906, "learning_rate": 8.607332120312428e-05, "loss": 0.0605, "step": 56260 }, { "epoch": 1.9023631630548699, "grad_norm": 0.3815651535987854, "learning_rate": 8.606759598895548e-05, "loss": 0.0614, "step": 56270 }, { "epoch": 1.9027012407451234, "grad_norm": 0.6089365482330322, "learning_rate": 8.606186978871215e-05, "loss": 0.0642, "step": 56280 }, { "epoch": 1.9030393184353764, "grad_norm": 0.45162075757980347, "learning_rate": 8.605614260255087e-05, "loss": 0.0605, "step": 56290 }, { "epoch": 1.9033773961256295, "grad_norm": 0.559087872505188, "learning_rate": 8.605041443062818e-05, "loss": 0.0607, "step": 56300 }, { "epoch": 1.903715473815883, "grad_norm": 0.42966964840888977, "learning_rate": 8.60446852731007e-05, "loss": 0.0622, "step": 56310 }, { "epoch": 1.904053551506136, "grad_norm": 0.4314900040626526, "learning_rate": 8.603895513012507e-05, "loss": 0.0618, "step": 56320 }, { "epoch": 1.9043916291963894, "grad_norm": 0.42849767208099365, "learning_rate": 8.603322400185796e-05, "loss": 0.0642, "step": 56330 }, { "epoch": 1.9047297068866427, "grad_norm": 0.3819114565849304, "learning_rate": 8.602749188845603e-05, "loss": 0.0595, "step": 56340 }, { "epoch": 1.9050677845768957, "grad_norm": 0.46132904291152954, "learning_rate": 8.6021758790076e-05, "loss": 0.0502, "step": 56350 }, { "epoch": 1.905405862267149, "grad_norm": 0.45236751437187195, "learning_rate": 8.601602470687462e-05, "loss": 0.0575, "step": 56360 }, { "epoch": 1.9057439399574023, "grad_norm": 0.3345819413661957, "learning_rate": 8.601028963900867e-05, "loss": 0.0591, "step": 56370 }, { "epoch": 1.9060820176476554, "grad_norm": 0.6215890645980835, "learning_rate": 8.60045535866349e-05, "loss": 0.0622, "step": 56380 }, { "epoch": 1.9064200953379087, "grad_norm": 0.7392917275428772, "learning_rate": 8.59988165499102e-05, "loss": 0.0612, "step": 56390 }, { "epoch": 1.906758173028162, "grad_norm": 0.64666748046875, "learning_rate": 8.599307852899135e-05, "loss": 0.063, "step": 56400 }, { "epoch": 1.907096250718415, "grad_norm": 0.3486059606075287, "learning_rate": 8.598733952403527e-05, "loss": 0.0571, "step": 56410 }, { "epoch": 1.9074343284086683, "grad_norm": 0.3749987781047821, "learning_rate": 8.598159953519885e-05, "loss": 0.066, "step": 56420 }, { "epoch": 1.9077724060989216, "grad_norm": 1.023800253868103, "learning_rate": 8.597585856263901e-05, "loss": 0.0629, "step": 56430 }, { "epoch": 1.9081104837891747, "grad_norm": 0.8552102446556091, "learning_rate": 8.597011660651271e-05, "loss": 0.06, "step": 56440 }, { "epoch": 1.908448561479428, "grad_norm": 2.194746971130371, "learning_rate": 8.596437366697692e-05, "loss": 0.0601, "step": 56450 }, { "epoch": 1.9087866391696813, "grad_norm": 0.49654874205589294, "learning_rate": 8.595862974418869e-05, "loss": 0.0568, "step": 56460 }, { "epoch": 1.9091247168599343, "grad_norm": 0.37688806653022766, "learning_rate": 8.595288483830502e-05, "loss": 0.0566, "step": 56470 }, { "epoch": 1.9094627945501876, "grad_norm": 0.4142686724662781, "learning_rate": 8.594713894948299e-05, "loss": 0.0567, "step": 56480 }, { "epoch": 1.909800872240441, "grad_norm": 0.3460366725921631, "learning_rate": 8.594139207787968e-05, "loss": 0.0577, "step": 56490 }, { "epoch": 1.910138949930694, "grad_norm": 0.38422638177871704, "learning_rate": 8.593564422365222e-05, "loss": 0.06, "step": 56500 }, { "epoch": 1.9104770276209473, "grad_norm": 0.41784825921058655, "learning_rate": 8.592989538695774e-05, "loss": 0.0626, "step": 56510 }, { "epoch": 1.9108151053112006, "grad_norm": 0.7300862073898315, "learning_rate": 8.592414556795343e-05, "loss": 0.0542, "step": 56520 }, { "epoch": 1.9111531830014536, "grad_norm": 0.4686973989009857, "learning_rate": 8.591839476679647e-05, "loss": 0.0614, "step": 56530 }, { "epoch": 1.911491260691707, "grad_norm": 0.3471318185329437, "learning_rate": 8.59126429836441e-05, "loss": 0.06, "step": 56540 }, { "epoch": 1.9118293383819602, "grad_norm": 0.33477333188056946, "learning_rate": 8.590689021865354e-05, "loss": 0.0627, "step": 56550 }, { "epoch": 1.9121674160722133, "grad_norm": 0.45990636944770813, "learning_rate": 8.590113647198211e-05, "loss": 0.0617, "step": 56560 }, { "epoch": 1.9125054937624666, "grad_norm": 0.49154502153396606, "learning_rate": 8.589538174378708e-05, "loss": 0.0615, "step": 56570 }, { "epoch": 1.9128435714527199, "grad_norm": 0.3901711404323578, "learning_rate": 8.58896260342258e-05, "loss": 0.0595, "step": 56580 }, { "epoch": 1.913181649142973, "grad_norm": 0.42222633957862854, "learning_rate": 8.588386934345565e-05, "loss": 0.0561, "step": 56590 }, { "epoch": 1.9135197268332262, "grad_norm": 0.8076237440109253, "learning_rate": 8.587811167163398e-05, "loss": 0.0586, "step": 56600 }, { "epoch": 1.9138578045234795, "grad_norm": 0.7310484051704407, "learning_rate": 8.587235301891822e-05, "loss": 0.0572, "step": 56610 }, { "epoch": 1.9141958822137326, "grad_norm": 0.3449760675430298, "learning_rate": 8.58665933854658e-05, "loss": 0.06, "step": 56620 }, { "epoch": 1.914533959903986, "grad_norm": 0.29825058579444885, "learning_rate": 8.586083277143419e-05, "loss": 0.0597, "step": 56630 }, { "epoch": 1.9148720375942392, "grad_norm": 0.6047582030296326, "learning_rate": 8.585507117698088e-05, "loss": 0.0569, "step": 56640 }, { "epoch": 1.9152101152844923, "grad_norm": 0.3341100513935089, "learning_rate": 8.584930860226341e-05, "loss": 0.0553, "step": 56650 }, { "epoch": 1.9155481929747455, "grad_norm": 0.3221413195133209, "learning_rate": 8.58435450474393e-05, "loss": 0.0624, "step": 56660 }, { "epoch": 1.9158862706649988, "grad_norm": 0.42894527316093445, "learning_rate": 8.583778051266613e-05, "loss": 0.0593, "step": 56670 }, { "epoch": 1.916224348355252, "grad_norm": 0.5540868043899536, "learning_rate": 8.583201499810152e-05, "loss": 0.0567, "step": 56680 }, { "epoch": 1.9165624260455054, "grad_norm": 0.39374521374702454, "learning_rate": 8.582624850390306e-05, "loss": 0.0584, "step": 56690 }, { "epoch": 1.9169005037357585, "grad_norm": 0.567483127117157, "learning_rate": 8.582048103022844e-05, "loss": 0.0582, "step": 56700 }, { "epoch": 1.9172385814260116, "grad_norm": 0.4403117597103119, "learning_rate": 8.581471257723532e-05, "loss": 0.0615, "step": 56710 }, { "epoch": 1.917576659116265, "grad_norm": 0.8013961315155029, "learning_rate": 8.58089431450814e-05, "loss": 0.0588, "step": 56720 }, { "epoch": 1.9179147368065181, "grad_norm": 0.6464091539382935, "learning_rate": 8.580317273392446e-05, "loss": 0.0646, "step": 56730 }, { "epoch": 1.9182528144967712, "grad_norm": 0.3580954968929291, "learning_rate": 8.57974013439222e-05, "loss": 0.0572, "step": 56740 }, { "epoch": 1.9185908921870247, "grad_norm": 0.4232107996940613, "learning_rate": 8.579162897523246e-05, "loss": 0.0543, "step": 56750 }, { "epoch": 1.9189289698772778, "grad_norm": 0.6078241467475891, "learning_rate": 8.5785855628013e-05, "loss": 0.0583, "step": 56760 }, { "epoch": 1.9192670475675309, "grad_norm": 0.6064618229866028, "learning_rate": 8.578008130242171e-05, "loss": 0.0601, "step": 56770 }, { "epoch": 1.9196051252577844, "grad_norm": 0.4876348376274109, "learning_rate": 8.577430599861644e-05, "loss": 0.06, "step": 56780 }, { "epoch": 1.9199432029480374, "grad_norm": 0.39178526401519775, "learning_rate": 8.576852971675509e-05, "loss": 0.061, "step": 56790 }, { "epoch": 1.9202812806382907, "grad_norm": 0.8477975726127625, "learning_rate": 8.576275245699555e-05, "loss": 0.0609, "step": 56800 }, { "epoch": 1.920619358328544, "grad_norm": 0.5811886787414551, "learning_rate": 8.575697421949582e-05, "loss": 0.0585, "step": 56810 }, { "epoch": 1.920957436018797, "grad_norm": 1.2006975412368774, "learning_rate": 8.575119500441383e-05, "loss": 0.0607, "step": 56820 }, { "epoch": 1.9212955137090504, "grad_norm": 0.7507680058479309, "learning_rate": 8.574541481190761e-05, "loss": 0.0605, "step": 56830 }, { "epoch": 1.9216335913993037, "grad_norm": 0.353771448135376, "learning_rate": 8.573963364213518e-05, "loss": 0.0631, "step": 56840 }, { "epoch": 1.9219716690895567, "grad_norm": 0.5201774835586548, "learning_rate": 8.573385149525458e-05, "loss": 0.0632, "step": 56850 }, { "epoch": 1.92230974677981, "grad_norm": 0.38994690775871277, "learning_rate": 8.572806837142391e-05, "loss": 0.0633, "step": 56860 }, { "epoch": 1.9226478244700633, "grad_norm": 0.3341738283634186, "learning_rate": 8.572228427080127e-05, "loss": 0.055, "step": 56870 }, { "epoch": 1.9229859021603164, "grad_norm": 0.5553536415100098, "learning_rate": 8.571649919354481e-05, "loss": 0.0603, "step": 56880 }, { "epoch": 1.9233239798505697, "grad_norm": 0.25376957654953003, "learning_rate": 8.571071313981267e-05, "loss": 0.0603, "step": 56890 }, { "epoch": 1.923662057540823, "grad_norm": 0.253079891204834, "learning_rate": 8.570492610976305e-05, "loss": 0.0582, "step": 56900 }, { "epoch": 1.924000135231076, "grad_norm": 0.44931381940841675, "learning_rate": 8.569913810355417e-05, "loss": 0.0611, "step": 56910 }, { "epoch": 1.9243382129213293, "grad_norm": 0.6303226351737976, "learning_rate": 8.569334912134426e-05, "loss": 0.0602, "step": 56920 }, { "epoch": 1.9246762906115826, "grad_norm": 0.30841121077537537, "learning_rate": 8.568755916329159e-05, "loss": 0.0562, "step": 56930 }, { "epoch": 1.9250143683018357, "grad_norm": 0.47119149565696716, "learning_rate": 8.568176822955447e-05, "loss": 0.0583, "step": 56940 }, { "epoch": 1.925352445992089, "grad_norm": 0.46570146083831787, "learning_rate": 8.567597632029121e-05, "loss": 0.0596, "step": 56950 }, { "epoch": 1.9256905236823423, "grad_norm": 0.40432825684547424, "learning_rate": 8.567018343566016e-05, "loss": 0.0525, "step": 56960 }, { "epoch": 1.9260286013725954, "grad_norm": 0.6030447483062744, "learning_rate": 8.566438957581968e-05, "loss": 0.0599, "step": 56970 }, { "epoch": 1.9263666790628486, "grad_norm": 0.31721270084381104, "learning_rate": 8.565859474092822e-05, "loss": 0.0541, "step": 56980 }, { "epoch": 1.926704756753102, "grad_norm": 0.515237033367157, "learning_rate": 8.565279893114415e-05, "loss": 0.0571, "step": 56990 }, { "epoch": 1.927042834443355, "grad_norm": 0.6808257699012756, "learning_rate": 8.564700214662595e-05, "loss": 0.0525, "step": 57000 }, { "epoch": 1.9273809121336083, "grad_norm": 0.39139652252197266, "learning_rate": 8.564120438753212e-05, "loss": 0.0545, "step": 57010 }, { "epoch": 1.9277189898238616, "grad_norm": 0.5909433960914612, "learning_rate": 8.563540565402115e-05, "loss": 0.0577, "step": 57020 }, { "epoch": 1.9280570675141147, "grad_norm": 0.948415994644165, "learning_rate": 8.562960594625155e-05, "loss": 0.0613, "step": 57030 }, { "epoch": 1.928395145204368, "grad_norm": 0.5436281561851501, "learning_rate": 8.562380526438194e-05, "loss": 0.0594, "step": 57040 }, { "epoch": 1.9287332228946212, "grad_norm": 0.5341663956642151, "learning_rate": 8.561800360857086e-05, "loss": 0.0549, "step": 57050 }, { "epoch": 1.9290713005848743, "grad_norm": 0.3994397521018982, "learning_rate": 8.561220097897696e-05, "loss": 0.0548, "step": 57060 }, { "epoch": 1.9294093782751276, "grad_norm": 0.4148274064064026, "learning_rate": 8.560639737575884e-05, "loss": 0.0634, "step": 57070 }, { "epoch": 1.9297474559653809, "grad_norm": 0.5882307291030884, "learning_rate": 8.56005927990752e-05, "loss": 0.0599, "step": 57080 }, { "epoch": 1.930085533655634, "grad_norm": 0.4320262372493744, "learning_rate": 8.559478724908474e-05, "loss": 0.0634, "step": 57090 }, { "epoch": 1.9304236113458872, "grad_norm": 0.4779087007045746, "learning_rate": 8.558898072594617e-05, "loss": 0.0613, "step": 57100 }, { "epoch": 1.9307616890361405, "grad_norm": 0.36193108558654785, "learning_rate": 8.558317322981821e-05, "loss": 0.0575, "step": 57110 }, { "epoch": 1.9310997667263936, "grad_norm": 0.31963059306144714, "learning_rate": 8.557736476085968e-05, "loss": 0.0635, "step": 57120 }, { "epoch": 1.931437844416647, "grad_norm": 0.5053792595863342, "learning_rate": 8.557155531922937e-05, "loss": 0.0615, "step": 57130 }, { "epoch": 1.9317759221069002, "grad_norm": 0.3851870894432068, "learning_rate": 8.556574490508611e-05, "loss": 0.056, "step": 57140 }, { "epoch": 1.9321139997971533, "grad_norm": 0.5058479905128479, "learning_rate": 8.555993351858874e-05, "loss": 0.0545, "step": 57150 }, { "epoch": 1.9324520774874068, "grad_norm": 0.3986932933330536, "learning_rate": 8.555412115989612e-05, "loss": 0.0615, "step": 57160 }, { "epoch": 1.9327901551776598, "grad_norm": 0.6105556488037109, "learning_rate": 8.55483078291672e-05, "loss": 0.065, "step": 57170 }, { "epoch": 1.933128232867913, "grad_norm": 0.43971216678619385, "learning_rate": 8.55424935265609e-05, "loss": 0.0605, "step": 57180 }, { "epoch": 1.9334663105581664, "grad_norm": 0.36433497071266174, "learning_rate": 8.55366782522362e-05, "loss": 0.0637, "step": 57190 }, { "epoch": 1.9338043882484195, "grad_norm": 1.255571961402893, "learning_rate": 8.553086200635203e-05, "loss": 0.0576, "step": 57200 }, { "epoch": 1.9341424659386726, "grad_norm": 0.3288270831108093, "learning_rate": 8.552504478906747e-05, "loss": 0.0617, "step": 57210 }, { "epoch": 1.934480543628926, "grad_norm": 0.4100351929664612, "learning_rate": 8.551922660054152e-05, "loss": 0.0569, "step": 57220 }, { "epoch": 1.9348186213191791, "grad_norm": 0.3547464907169342, "learning_rate": 8.551340744093325e-05, "loss": 0.0588, "step": 57230 }, { "epoch": 1.9351566990094322, "grad_norm": 0.29516860842704773, "learning_rate": 8.550758731040177e-05, "loss": 0.0565, "step": 57240 }, { "epoch": 1.9354947766996857, "grad_norm": 0.466441810131073, "learning_rate": 8.550176620910619e-05, "loss": 0.0614, "step": 57250 }, { "epoch": 1.9358328543899388, "grad_norm": 0.5313670635223389, "learning_rate": 8.549594413720568e-05, "loss": 0.0533, "step": 57260 }, { "epoch": 1.9361709320801919, "grad_norm": 0.30478763580322266, "learning_rate": 8.549012109485937e-05, "loss": 0.0577, "step": 57270 }, { "epoch": 1.9365090097704454, "grad_norm": 0.33947837352752686, "learning_rate": 8.548429708222648e-05, "loss": 0.0589, "step": 57280 }, { "epoch": 1.9368470874606984, "grad_norm": 0.3984563946723938, "learning_rate": 8.547847209946625e-05, "loss": 0.0553, "step": 57290 }, { "epoch": 1.9371851651509517, "grad_norm": 0.45888063311576843, "learning_rate": 8.547264614673792e-05, "loss": 0.0548, "step": 57300 }, { "epoch": 1.937523242841205, "grad_norm": 0.34037846326828003, "learning_rate": 8.546681922420076e-05, "loss": 0.0588, "step": 57310 }, { "epoch": 1.937861320531458, "grad_norm": 0.5609543919563293, "learning_rate": 8.54609913320141e-05, "loss": 0.0593, "step": 57320 }, { "epoch": 1.9381993982217114, "grad_norm": 0.5745946764945984, "learning_rate": 8.545516247033726e-05, "loss": 0.0579, "step": 57330 }, { "epoch": 1.9385374759119647, "grad_norm": 0.36386317014694214, "learning_rate": 8.544933263932959e-05, "loss": 0.0641, "step": 57340 }, { "epoch": 1.9388755536022177, "grad_norm": 0.5261662602424622, "learning_rate": 8.544350183915049e-05, "loss": 0.0647, "step": 57350 }, { "epoch": 1.939213631292471, "grad_norm": 0.4949020743370056, "learning_rate": 8.543767006995935e-05, "loss": 0.0625, "step": 57360 }, { "epoch": 1.9395517089827243, "grad_norm": 0.8312718868255615, "learning_rate": 8.543183733191564e-05, "loss": 0.0625, "step": 57370 }, { "epoch": 1.9398897866729774, "grad_norm": 0.38544800877571106, "learning_rate": 8.54260036251788e-05, "loss": 0.0613, "step": 57380 }, { "epoch": 1.9402278643632307, "grad_norm": 0.2915882468223572, "learning_rate": 8.542016894990834e-05, "loss": 0.0654, "step": 57390 }, { "epoch": 1.940565942053484, "grad_norm": 1.3573851585388184, "learning_rate": 8.541433330626377e-05, "loss": 0.0662, "step": 57400 }, { "epoch": 1.940904019743737, "grad_norm": 0.5113211274147034, "learning_rate": 8.54084966944046e-05, "loss": 0.0657, "step": 57410 }, { "epoch": 1.9412420974339903, "grad_norm": 0.4128546416759491, "learning_rate": 8.540265911449045e-05, "loss": 0.0623, "step": 57420 }, { "epoch": 1.9415801751242436, "grad_norm": 0.35151413083076477, "learning_rate": 8.539682056668092e-05, "loss": 0.0588, "step": 57430 }, { "epoch": 1.9419182528144967, "grad_norm": 0.5175586342811584, "learning_rate": 8.53909810511356e-05, "loss": 0.0638, "step": 57440 }, { "epoch": 1.94225633050475, "grad_norm": 0.2906821072101593, "learning_rate": 8.538514056801415e-05, "loss": 0.0619, "step": 57450 }, { "epoch": 1.9425944081950033, "grad_norm": 0.414999395608902, "learning_rate": 8.537929911747625e-05, "loss": 0.0517, "step": 57460 }, { "epoch": 1.9429324858852564, "grad_norm": 0.2950553297996521, "learning_rate": 8.537345669968161e-05, "loss": 0.0588, "step": 57470 }, { "epoch": 1.9432705635755096, "grad_norm": 0.45862337946891785, "learning_rate": 8.536761331478996e-05, "loss": 0.0595, "step": 57480 }, { "epoch": 1.943608641265763, "grad_norm": 0.28384530544281006, "learning_rate": 8.536176896296103e-05, "loss": 0.0584, "step": 57490 }, { "epoch": 1.943946718956016, "grad_norm": 0.3212166428565979, "learning_rate": 8.535592364435463e-05, "loss": 0.0559, "step": 57500 }, { "epoch": 1.9442847966462693, "grad_norm": 0.2888970673084259, "learning_rate": 8.535007735913056e-05, "loss": 0.0546, "step": 57510 }, { "epoch": 1.9446228743365226, "grad_norm": 0.4929901361465454, "learning_rate": 8.534423010744865e-05, "loss": 0.0571, "step": 57520 }, { "epoch": 1.9449609520267757, "grad_norm": 0.481574684381485, "learning_rate": 8.53383818894688e-05, "loss": 0.0555, "step": 57530 }, { "epoch": 1.945299029717029, "grad_norm": 0.6405019164085388, "learning_rate": 8.533253270535082e-05, "loss": 0.0554, "step": 57540 }, { "epoch": 1.9456371074072822, "grad_norm": 0.28562867641448975, "learning_rate": 8.532668255525468e-05, "loss": 0.0543, "step": 57550 }, { "epoch": 1.9459751850975353, "grad_norm": 0.33880138397216797, "learning_rate": 8.532083143934033e-05, "loss": 0.0543, "step": 57560 }, { "epoch": 1.9463132627877886, "grad_norm": 0.45123353600502014, "learning_rate": 8.53149793577677e-05, "loss": 0.0524, "step": 57570 }, { "epoch": 1.946651340478042, "grad_norm": 0.3346744775772095, "learning_rate": 8.530912631069681e-05, "loss": 0.0568, "step": 57580 }, { "epoch": 1.946989418168295, "grad_norm": 0.2970665395259857, "learning_rate": 8.530327229828767e-05, "loss": 0.0505, "step": 57590 }, { "epoch": 1.9473274958585483, "grad_norm": 0.46294569969177246, "learning_rate": 8.529741732070033e-05, "loss": 0.0618, "step": 57600 }, { "epoch": 1.9476655735488015, "grad_norm": 0.4584949314594269, "learning_rate": 8.529156137809484e-05, "loss": 0.0632, "step": 57610 }, { "epoch": 1.9480036512390546, "grad_norm": 0.345359206199646, "learning_rate": 8.528570447063136e-05, "loss": 0.0563, "step": 57620 }, { "epoch": 1.948341728929308, "grad_norm": 0.5609899759292603, "learning_rate": 8.527984659846995e-05, "loss": 0.061, "step": 57630 }, { "epoch": 1.9486798066195612, "grad_norm": 0.4345328211784363, "learning_rate": 8.527398776177078e-05, "loss": 0.057, "step": 57640 }, { "epoch": 1.9490178843098143, "grad_norm": 3.1995134353637695, "learning_rate": 8.526812796069405e-05, "loss": 0.0571, "step": 57650 }, { "epoch": 1.9493559620000678, "grad_norm": 0.4526677429676056, "learning_rate": 8.526226719539996e-05, "loss": 0.0597, "step": 57660 }, { "epoch": 1.9496940396903208, "grad_norm": 0.48079749941825867, "learning_rate": 8.525640546604871e-05, "loss": 0.0591, "step": 57670 }, { "epoch": 1.950032117380574, "grad_norm": 0.6251500844955444, "learning_rate": 8.525054277280061e-05, "loss": 0.0624, "step": 57680 }, { "epoch": 1.9503701950708274, "grad_norm": 0.5045092701911926, "learning_rate": 8.524467911581588e-05, "loss": 0.0506, "step": 57690 }, { "epoch": 1.9507082727610805, "grad_norm": 0.3929516673088074, "learning_rate": 8.523881449525488e-05, "loss": 0.0634, "step": 57700 }, { "epoch": 1.9510463504513336, "grad_norm": 0.4184246063232422, "learning_rate": 8.523294891127794e-05, "loss": 0.0609, "step": 57710 }, { "epoch": 1.951384428141587, "grad_norm": 0.2941223680973053, "learning_rate": 8.52270823640454e-05, "loss": 0.0521, "step": 57720 }, { "epoch": 1.9517225058318401, "grad_norm": 0.34841474890708923, "learning_rate": 8.522121485371767e-05, "loss": 0.0609, "step": 57730 }, { "epoch": 1.9520605835220932, "grad_norm": 0.42039796710014343, "learning_rate": 8.521534638045514e-05, "loss": 0.0571, "step": 57740 }, { "epoch": 1.9523986612123467, "grad_norm": 0.7518122792243958, "learning_rate": 8.52094769444183e-05, "loss": 0.0579, "step": 57750 }, { "epoch": 1.9527367389025998, "grad_norm": 0.531755268573761, "learning_rate": 8.520360654576757e-05, "loss": 0.0604, "step": 57760 }, { "epoch": 1.953074816592853, "grad_norm": 0.44358545541763306, "learning_rate": 8.519773518466346e-05, "loss": 0.0565, "step": 57770 }, { "epoch": 1.9534128942831064, "grad_norm": 0.709337055683136, "learning_rate": 8.51918628612665e-05, "loss": 0.0597, "step": 57780 }, { "epoch": 1.9537509719733595, "grad_norm": 0.7736034989356995, "learning_rate": 8.518598957573722e-05, "loss": 0.0606, "step": 57790 }, { "epoch": 1.9540890496636127, "grad_norm": 0.40160059928894043, "learning_rate": 8.518011532823621e-05, "loss": 0.0567, "step": 57800 }, { "epoch": 1.954427127353866, "grad_norm": 0.55115807056427, "learning_rate": 8.517424011892407e-05, "loss": 0.0513, "step": 57810 }, { "epoch": 1.954765205044119, "grad_norm": 0.3468240797519684, "learning_rate": 8.516836394796142e-05, "loss": 0.0564, "step": 57820 }, { "epoch": 1.9551032827343724, "grad_norm": 0.43067681789398193, "learning_rate": 8.51624868155089e-05, "loss": 0.0594, "step": 57830 }, { "epoch": 1.9554413604246257, "grad_norm": 0.4173279106616974, "learning_rate": 8.515660872172722e-05, "loss": 0.0583, "step": 57840 }, { "epoch": 1.9557794381148788, "grad_norm": 0.5337878465652466, "learning_rate": 8.515072966677704e-05, "loss": 0.0632, "step": 57850 }, { "epoch": 1.956117515805132, "grad_norm": 0.9476334452629089, "learning_rate": 8.514484965081913e-05, "loss": 0.0588, "step": 57860 }, { "epoch": 1.9564555934953853, "grad_norm": 0.3215279281139374, "learning_rate": 8.513896867401423e-05, "loss": 0.0537, "step": 57870 }, { "epoch": 1.9567936711856384, "grad_norm": 0.953158438205719, "learning_rate": 8.513308673652313e-05, "loss": 0.0599, "step": 57880 }, { "epoch": 1.9571317488758917, "grad_norm": 0.700183093547821, "learning_rate": 8.512720383850663e-05, "loss": 0.0605, "step": 57890 }, { "epoch": 1.957469826566145, "grad_norm": 0.2952534258365631, "learning_rate": 8.512131998012558e-05, "loss": 0.059, "step": 57900 }, { "epoch": 1.957807904256398, "grad_norm": 0.2980751693248749, "learning_rate": 8.511543516154084e-05, "loss": 0.0553, "step": 57910 }, { "epoch": 1.9581459819466513, "grad_norm": 0.3875002861022949, "learning_rate": 8.51095493829133e-05, "loss": 0.0586, "step": 57920 }, { "epoch": 1.9584840596369046, "grad_norm": 0.41277623176574707, "learning_rate": 8.510366264440384e-05, "loss": 0.0492, "step": 57930 }, { "epoch": 1.9588221373271577, "grad_norm": 0.386243999004364, "learning_rate": 8.509777494617346e-05, "loss": 0.0613, "step": 57940 }, { "epoch": 1.959160215017411, "grad_norm": 1.069344401359558, "learning_rate": 8.50918862883831e-05, "loss": 0.0548, "step": 57950 }, { "epoch": 1.9594982927076643, "grad_norm": 0.3751879930496216, "learning_rate": 8.508599667119374e-05, "loss": 0.0598, "step": 57960 }, { "epoch": 1.9598363703979174, "grad_norm": 0.48622867465019226, "learning_rate": 8.508010609476643e-05, "loss": 0.0578, "step": 57970 }, { "epoch": 1.9601744480881707, "grad_norm": 0.5586061477661133, "learning_rate": 8.507421455926217e-05, "loss": 0.0618, "step": 57980 }, { "epoch": 1.960512525778424, "grad_norm": 0.6377876996994019, "learning_rate": 8.506832206484207e-05, "loss": 0.0626, "step": 57990 }, { "epoch": 1.960850603468677, "grad_norm": 0.40280815958976746, "learning_rate": 8.506242861166723e-05, "loss": 0.0553, "step": 58000 }, { "epoch": 1.9611886811589303, "grad_norm": 0.5577664375305176, "learning_rate": 8.505653419989876e-05, "loss": 0.056, "step": 58010 }, { "epoch": 1.9615267588491836, "grad_norm": 0.4029608964920044, "learning_rate": 8.505063882969782e-05, "loss": 0.0622, "step": 58020 }, { "epoch": 1.9618648365394367, "grad_norm": 0.5403863787651062, "learning_rate": 8.504474250122558e-05, "loss": 0.0581, "step": 58030 }, { "epoch": 1.96220291422969, "grad_norm": 0.34695687890052795, "learning_rate": 8.503884521464322e-05, "loss": 0.0615, "step": 58040 }, { "epoch": 1.9625409919199432, "grad_norm": 0.39686787128448486, "learning_rate": 8.503294697011202e-05, "loss": 0.0629, "step": 58050 }, { "epoch": 1.9628790696101963, "grad_norm": 0.5429440140724182, "learning_rate": 8.50270477677932e-05, "loss": 0.0609, "step": 58060 }, { "epoch": 1.9632171473004496, "grad_norm": 0.40937286615371704, "learning_rate": 8.502114760784808e-05, "loss": 0.0573, "step": 58070 }, { "epoch": 1.963555224990703, "grad_norm": 0.32634592056274414, "learning_rate": 8.501524649043792e-05, "loss": 0.0556, "step": 58080 }, { "epoch": 1.963893302680956, "grad_norm": 0.5434418320655823, "learning_rate": 8.500934441572408e-05, "loss": 0.0561, "step": 58090 }, { "epoch": 1.9642313803712093, "grad_norm": 0.3875419795513153, "learning_rate": 8.50034413838679e-05, "loss": 0.0645, "step": 58100 }, { "epoch": 1.9645694580614625, "grad_norm": 0.4498380124568939, "learning_rate": 8.49975373950308e-05, "loss": 0.052, "step": 58110 }, { "epoch": 1.9649075357517156, "grad_norm": 0.6961074471473694, "learning_rate": 8.499163244937418e-05, "loss": 0.0555, "step": 58120 }, { "epoch": 1.9652456134419691, "grad_norm": 0.847576916217804, "learning_rate": 8.498572654705947e-05, "loss": 0.0616, "step": 58130 }, { "epoch": 1.9655836911322222, "grad_norm": 0.557930588722229, "learning_rate": 8.497981968824813e-05, "loss": 0.0555, "step": 58140 }, { "epoch": 1.9659217688224753, "grad_norm": 0.6273267865180969, "learning_rate": 8.497391187310166e-05, "loss": 0.0563, "step": 58150 }, { "epoch": 1.9662598465127288, "grad_norm": 0.2818664014339447, "learning_rate": 8.496800310178158e-05, "loss": 0.0607, "step": 58160 }, { "epoch": 1.9665979242029819, "grad_norm": 0.5787543058395386, "learning_rate": 8.496209337444945e-05, "loss": 0.0568, "step": 58170 }, { "epoch": 1.966936001893235, "grad_norm": 0.36085161566734314, "learning_rate": 8.495618269126681e-05, "loss": 0.0603, "step": 58180 }, { "epoch": 1.9672740795834884, "grad_norm": 0.5885505080223083, "learning_rate": 8.495027105239526e-05, "loss": 0.0596, "step": 58190 }, { "epoch": 1.9676121572737415, "grad_norm": 0.6703376173973083, "learning_rate": 8.494435845799645e-05, "loss": 0.065, "step": 58200 }, { "epoch": 1.9679502349639946, "grad_norm": 0.3643762767314911, "learning_rate": 8.493844490823199e-05, "loss": 0.0636, "step": 58210 }, { "epoch": 1.968288312654248, "grad_norm": 0.5027557611465454, "learning_rate": 8.493253040326358e-05, "loss": 0.0586, "step": 58220 }, { "epoch": 1.9686263903445012, "grad_norm": 0.30910757184028625, "learning_rate": 8.492661494325291e-05, "loss": 0.0604, "step": 58230 }, { "epoch": 1.9689644680347542, "grad_norm": 0.5208579301834106, "learning_rate": 8.49206985283617e-05, "loss": 0.0557, "step": 58240 }, { "epoch": 1.9693025457250077, "grad_norm": 0.6774007081985474, "learning_rate": 8.491478115875172e-05, "loss": 0.0625, "step": 58250 }, { "epoch": 1.9696406234152608, "grad_norm": 0.5704230070114136, "learning_rate": 8.490886283458474e-05, "loss": 0.0602, "step": 58260 }, { "epoch": 1.969978701105514, "grad_norm": 0.6393426060676575, "learning_rate": 8.490294355602258e-05, "loss": 0.0524, "step": 58270 }, { "epoch": 1.9703167787957674, "grad_norm": 0.5908071398735046, "learning_rate": 8.489702332322703e-05, "loss": 0.055, "step": 58280 }, { "epoch": 1.9706548564860205, "grad_norm": 0.42554599046707153, "learning_rate": 8.489110213635998e-05, "loss": 0.0551, "step": 58290 }, { "epoch": 1.9709929341762737, "grad_norm": 0.8355115652084351, "learning_rate": 8.488517999558331e-05, "loss": 0.0569, "step": 58300 }, { "epoch": 1.971331011866527, "grad_norm": 0.47727149724960327, "learning_rate": 8.487925690105891e-05, "loss": 0.06, "step": 58310 }, { "epoch": 1.97166908955678, "grad_norm": 0.8161004781723022, "learning_rate": 8.487333285294875e-05, "loss": 0.053, "step": 58320 }, { "epoch": 1.9720071672470334, "grad_norm": 0.3697737753391266, "learning_rate": 8.486740785141476e-05, "loss": 0.0578, "step": 58330 }, { "epoch": 1.9723452449372867, "grad_norm": 0.6359772682189941, "learning_rate": 8.486148189661895e-05, "loss": 0.057, "step": 58340 }, { "epoch": 1.9726833226275398, "grad_norm": 0.614509642124176, "learning_rate": 8.485555498872331e-05, "loss": 0.0564, "step": 58350 }, { "epoch": 1.973021400317793, "grad_norm": 0.3686063289642334, "learning_rate": 8.48496271278899e-05, "loss": 0.0584, "step": 58360 }, { "epoch": 1.9733594780080463, "grad_norm": 0.4726800322532654, "learning_rate": 8.484369831428076e-05, "loss": 0.0616, "step": 58370 }, { "epoch": 1.9736975556982994, "grad_norm": 0.7238774299621582, "learning_rate": 8.483776854805801e-05, "loss": 0.0562, "step": 58380 }, { "epoch": 1.9740356333885527, "grad_norm": 0.5619884133338928, "learning_rate": 8.483183782938375e-05, "loss": 0.064, "step": 58390 }, { "epoch": 1.974373711078806, "grad_norm": 0.46708860993385315, "learning_rate": 8.482590615842014e-05, "loss": 0.0655, "step": 58400 }, { "epoch": 1.974711788769059, "grad_norm": 0.4252980053424835, "learning_rate": 8.481997353532934e-05, "loss": 0.0564, "step": 58410 }, { "epoch": 1.9750498664593124, "grad_norm": 0.3483603000640869, "learning_rate": 8.481403996027354e-05, "loss": 0.0601, "step": 58420 }, { "epoch": 1.9753879441495656, "grad_norm": 0.2837541401386261, "learning_rate": 8.480810543341497e-05, "loss": 0.0574, "step": 58430 }, { "epoch": 1.9757260218398187, "grad_norm": 0.814701497554779, "learning_rate": 8.480216995491586e-05, "loss": 0.0593, "step": 58440 }, { "epoch": 1.976064099530072, "grad_norm": 0.9567320942878723, "learning_rate": 8.47962335249385e-05, "loss": 0.0529, "step": 58450 }, { "epoch": 1.9764021772203253, "grad_norm": 0.2801515460014343, "learning_rate": 8.47902961436452e-05, "loss": 0.0607, "step": 58460 }, { "epoch": 1.9767402549105784, "grad_norm": 0.24556374549865723, "learning_rate": 8.478435781119828e-05, "loss": 0.0636, "step": 58470 }, { "epoch": 1.9770783326008317, "grad_norm": 0.3016166687011719, "learning_rate": 8.477841852776008e-05, "loss": 0.0638, "step": 58480 }, { "epoch": 1.977416410291085, "grad_norm": 0.6890912055969238, "learning_rate": 8.477247829349297e-05, "loss": 0.0595, "step": 58490 }, { "epoch": 1.977754487981338, "grad_norm": 0.37097153067588806, "learning_rate": 8.476653710855938e-05, "loss": 0.0495, "step": 58500 }, { "epoch": 1.9780925656715913, "grad_norm": 0.3611009418964386, "learning_rate": 8.476059497312171e-05, "loss": 0.0586, "step": 58510 }, { "epoch": 1.9784306433618446, "grad_norm": 0.47010481357574463, "learning_rate": 8.475465188734244e-05, "loss": 0.0581, "step": 58520 }, { "epoch": 1.9787687210520977, "grad_norm": 0.5494914650917053, "learning_rate": 8.474870785138406e-05, "loss": 0.0613, "step": 58530 }, { "epoch": 1.979106798742351, "grad_norm": 0.511289656162262, "learning_rate": 8.474276286540905e-05, "loss": 0.061, "step": 58540 }, { "epoch": 1.9794448764326043, "grad_norm": 0.40631794929504395, "learning_rate": 8.473681692957996e-05, "loss": 0.0641, "step": 58550 }, { "epoch": 1.9797829541228573, "grad_norm": 0.44374266266822815, "learning_rate": 8.473087004405932e-05, "loss": 0.0616, "step": 58560 }, { "epoch": 1.9801210318131106, "grad_norm": 0.4465023875236511, "learning_rate": 8.472492220900976e-05, "loss": 0.0573, "step": 58570 }, { "epoch": 1.980459109503364, "grad_norm": 0.38961291313171387, "learning_rate": 8.471897342459389e-05, "loss": 0.0587, "step": 58580 }, { "epoch": 1.980797187193617, "grad_norm": 0.307024210691452, "learning_rate": 8.471302369097431e-05, "loss": 0.0542, "step": 58590 }, { "epoch": 1.9811352648838703, "grad_norm": 0.5418667197227478, "learning_rate": 8.470707300831371e-05, "loss": 0.0496, "step": 58600 }, { "epoch": 1.9814733425741236, "grad_norm": 0.5883227586746216, "learning_rate": 8.470112137677476e-05, "loss": 0.051, "step": 58610 }, { "epoch": 1.9818114202643766, "grad_norm": 0.43217670917510986, "learning_rate": 8.469516879652019e-05, "loss": 0.064, "step": 58620 }, { "epoch": 1.9821494979546301, "grad_norm": 0.4969958961009979, "learning_rate": 8.468921526771276e-05, "loss": 0.067, "step": 58630 }, { "epoch": 1.9824875756448832, "grad_norm": 0.6365874409675598, "learning_rate": 8.468326079051519e-05, "loss": 0.0632, "step": 58640 }, { "epoch": 1.9828256533351363, "grad_norm": 0.4849919080734253, "learning_rate": 8.467730536509033e-05, "loss": 0.0595, "step": 58650 }, { "epoch": 1.9831637310253898, "grad_norm": 0.597454845905304, "learning_rate": 8.467134899160093e-05, "loss": 0.0594, "step": 58660 }, { "epoch": 1.9835018087156429, "grad_norm": 0.416405588388443, "learning_rate": 8.46653916702099e-05, "loss": 0.0607, "step": 58670 }, { "epoch": 1.983839886405896, "grad_norm": 1.0711151361465454, "learning_rate": 8.46594334010801e-05, "loss": 0.0642, "step": 58680 }, { "epoch": 1.9841779640961494, "grad_norm": 0.38843053579330444, "learning_rate": 8.465347418437439e-05, "loss": 0.0637, "step": 58690 }, { "epoch": 1.9845160417864025, "grad_norm": 0.41843366622924805, "learning_rate": 8.46475140202557e-05, "loss": 0.054, "step": 58700 }, { "epoch": 1.9848541194766556, "grad_norm": 0.37617623805999756, "learning_rate": 8.464155290888703e-05, "loss": 0.0553, "step": 58710 }, { "epoch": 1.985192197166909, "grad_norm": 0.468329519033432, "learning_rate": 8.463559085043129e-05, "loss": 0.0596, "step": 58720 }, { "epoch": 1.9855302748571622, "grad_norm": 0.47366470098495483, "learning_rate": 8.462962784505152e-05, "loss": 0.0567, "step": 58730 }, { "epoch": 1.9858683525474154, "grad_norm": 0.8154898285865784, "learning_rate": 8.462366389291073e-05, "loss": 0.0601, "step": 58740 }, { "epoch": 1.9862064302376687, "grad_norm": 0.9256650805473328, "learning_rate": 8.461769899417198e-05, "loss": 0.053, "step": 58750 }, { "epoch": 1.9865445079279218, "grad_norm": 0.5936568379402161, "learning_rate": 8.461173314899833e-05, "loss": 0.0539, "step": 58760 }, { "epoch": 1.986882585618175, "grad_norm": 0.5536397099494934, "learning_rate": 8.460576635755292e-05, "loss": 0.0561, "step": 58770 }, { "epoch": 1.9872206633084284, "grad_norm": 0.38959378004074097, "learning_rate": 8.459979861999885e-05, "loss": 0.0562, "step": 58780 }, { "epoch": 1.9875587409986815, "grad_norm": 0.48771047592163086, "learning_rate": 8.45938299364993e-05, "loss": 0.0631, "step": 58790 }, { "epoch": 1.9878968186889348, "grad_norm": 0.9305754899978638, "learning_rate": 8.458786030721741e-05, "loss": 0.0621, "step": 58800 }, { "epoch": 1.988234896379188, "grad_norm": 0.5704796314239502, "learning_rate": 8.458188973231642e-05, "loss": 0.0555, "step": 58810 }, { "epoch": 1.9885729740694411, "grad_norm": 0.3281180262565613, "learning_rate": 8.457591821195956e-05, "loss": 0.0599, "step": 58820 }, { "epoch": 1.9889110517596944, "grad_norm": 0.4264548718929291, "learning_rate": 8.456994574631008e-05, "loss": 0.055, "step": 58830 }, { "epoch": 1.9892491294499477, "grad_norm": 0.4046817719936371, "learning_rate": 8.456397233553129e-05, "loss": 0.054, "step": 58840 }, { "epoch": 1.9895872071402008, "grad_norm": 0.336147665977478, "learning_rate": 8.455799797978645e-05, "loss": 0.0573, "step": 58850 }, { "epoch": 1.989925284830454, "grad_norm": 0.46994656324386597, "learning_rate": 8.455202267923895e-05, "loss": 0.0552, "step": 58860 }, { "epoch": 1.9902633625207073, "grad_norm": 0.49982795119285583, "learning_rate": 8.454604643405213e-05, "loss": 0.059, "step": 58870 }, { "epoch": 1.9906014402109604, "grad_norm": 0.469672292470932, "learning_rate": 8.454006924438937e-05, "loss": 0.0594, "step": 58880 }, { "epoch": 1.9909395179012137, "grad_norm": 0.5216494798660278, "learning_rate": 8.45340911104141e-05, "loss": 0.058, "step": 58890 }, { "epoch": 1.991277595591467, "grad_norm": 0.46288537979125977, "learning_rate": 8.452811203228974e-05, "loss": 0.0594, "step": 58900 }, { "epoch": 1.99161567328172, "grad_norm": 0.5838304162025452, "learning_rate": 8.45221320101798e-05, "loss": 0.0567, "step": 58910 }, { "epoch": 1.9919537509719734, "grad_norm": 1.0516436100006104, "learning_rate": 8.451615104424771e-05, "loss": 0.059, "step": 58920 }, { "epoch": 1.9922918286622266, "grad_norm": 0.42390599846839905, "learning_rate": 8.451016913465702e-05, "loss": 0.0619, "step": 58930 }, { "epoch": 1.9926299063524797, "grad_norm": 0.3950822651386261, "learning_rate": 8.450418628157128e-05, "loss": 0.0564, "step": 58940 }, { "epoch": 1.992967984042733, "grad_norm": 0.49198493361473083, "learning_rate": 8.449820248515404e-05, "loss": 0.0598, "step": 58950 }, { "epoch": 1.9933060617329863, "grad_norm": 0.418450266122818, "learning_rate": 8.449221774556892e-05, "loss": 0.056, "step": 58960 }, { "epoch": 1.9936441394232394, "grad_norm": 0.40427154302597046, "learning_rate": 8.448623206297951e-05, "loss": 0.0559, "step": 58970 }, { "epoch": 1.9939822171134927, "grad_norm": 0.5339241027832031, "learning_rate": 8.448024543754949e-05, "loss": 0.0602, "step": 58980 }, { "epoch": 1.994320294803746, "grad_norm": 0.43315473198890686, "learning_rate": 8.447425786944249e-05, "loss": 0.0553, "step": 58990 }, { "epoch": 1.994658372493999, "grad_norm": 0.2941559851169586, "learning_rate": 8.446826935882227e-05, "loss": 0.0503, "step": 59000 }, { "epoch": 1.9949964501842523, "grad_norm": 0.38738733530044556, "learning_rate": 8.446227990585248e-05, "loss": 0.0517, "step": 59010 }, { "epoch": 1.9953345278745056, "grad_norm": 0.4668222665786743, "learning_rate": 8.445628951069689e-05, "loss": 0.0588, "step": 59020 }, { "epoch": 1.9956726055647587, "grad_norm": 0.39163854718208313, "learning_rate": 8.445029817351933e-05, "loss": 0.0541, "step": 59030 }, { "epoch": 1.996010683255012, "grad_norm": 0.3871973752975464, "learning_rate": 8.444430589448354e-05, "loss": 0.0588, "step": 59040 }, { "epoch": 1.9963487609452653, "grad_norm": 0.3123009204864502, "learning_rate": 8.443831267375336e-05, "loss": 0.0582, "step": 59050 }, { "epoch": 1.9966868386355183, "grad_norm": 0.3560790717601776, "learning_rate": 8.443231851149266e-05, "loss": 0.0543, "step": 59060 }, { "epoch": 1.9970249163257716, "grad_norm": 0.43302762508392334, "learning_rate": 8.442632340786529e-05, "loss": 0.0534, "step": 59070 }, { "epoch": 1.997362994016025, "grad_norm": 0.2827969491481781, "learning_rate": 8.442032736303521e-05, "loss": 0.0578, "step": 59080 }, { "epoch": 1.997701071706278, "grad_norm": 0.4009947180747986, "learning_rate": 8.441433037716628e-05, "loss": 0.0521, "step": 59090 }, { "epoch": 1.9980391493965315, "grad_norm": 0.3136594295501709, "learning_rate": 8.440833245042249e-05, "loss": 0.0581, "step": 59100 }, { "epoch": 1.9983772270867846, "grad_norm": 0.6376173496246338, "learning_rate": 8.440233358296783e-05, "loss": 0.0604, "step": 59110 }, { "epoch": 1.9987153047770376, "grad_norm": 0.656425416469574, "learning_rate": 8.439633377496628e-05, "loss": 0.0564, "step": 59120 }, { "epoch": 1.9990533824672911, "grad_norm": 0.4069494605064392, "learning_rate": 8.43903330265819e-05, "loss": 0.056, "step": 59130 }, { "epoch": 1.9993914601575442, "grad_norm": 0.5222319960594177, "learning_rate": 8.438433133797871e-05, "loss": 0.0627, "step": 59140 }, { "epoch": 1.9997295378477973, "grad_norm": 0.7331944108009338, "learning_rate": 8.437832870932084e-05, "loss": 0.0498, "step": 59150 }, { "epoch": 2.000067615538051, "grad_norm": 0.723779022693634, "learning_rate": 8.437232514077238e-05, "loss": 0.0596, "step": 59160 }, { "epoch": 2.000405693228304, "grad_norm": 0.37290090322494507, "learning_rate": 8.436632063249745e-05, "loss": 0.0527, "step": 59170 }, { "epoch": 2.000743770918557, "grad_norm": 0.43820634484291077, "learning_rate": 8.436031518466024e-05, "loss": 0.0592, "step": 59180 }, { "epoch": 2.0010818486088104, "grad_norm": 0.4714776277542114, "learning_rate": 8.435430879742491e-05, "loss": 0.0464, "step": 59190 }, { "epoch": 2.0014199262990635, "grad_norm": 0.7208189964294434, "learning_rate": 8.434830147095569e-05, "loss": 0.068, "step": 59200 }, { "epoch": 2.0017580039893166, "grad_norm": 0.5872267484664917, "learning_rate": 8.434229320541681e-05, "loss": 0.0579, "step": 59210 }, { "epoch": 2.00209608167957, "grad_norm": 0.5266317129135132, "learning_rate": 8.433628400097253e-05, "loss": 0.0675, "step": 59220 }, { "epoch": 2.002434159369823, "grad_norm": 0.5801495313644409, "learning_rate": 8.433027385778716e-05, "loss": 0.0555, "step": 59230 }, { "epoch": 2.0027722370600762, "grad_norm": 0.9715628027915955, "learning_rate": 8.432426277602498e-05, "loss": 0.0572, "step": 59240 }, { "epoch": 2.0031103147503297, "grad_norm": 1.751460075378418, "learning_rate": 8.431825075585038e-05, "loss": 0.0559, "step": 59250 }, { "epoch": 2.003448392440583, "grad_norm": 0.9023072123527527, "learning_rate": 8.431223779742768e-05, "loss": 0.0644, "step": 59260 }, { "epoch": 2.003786470130836, "grad_norm": 1.085667371749878, "learning_rate": 8.430622390092128e-05, "loss": 0.0644, "step": 59270 }, { "epoch": 2.0041245478210894, "grad_norm": 0.47031766176223755, "learning_rate": 8.43002090664956e-05, "loss": 0.0612, "step": 59280 }, { "epoch": 2.0044626255113425, "grad_norm": 0.5183113813400269, "learning_rate": 8.429419329431511e-05, "loss": 0.0693, "step": 59290 }, { "epoch": 2.0048007032015955, "grad_norm": 0.2674892246723175, "learning_rate": 8.428817658454426e-05, "loss": 0.0637, "step": 59300 }, { "epoch": 2.005138780891849, "grad_norm": 0.3930831253528595, "learning_rate": 8.428215893734754e-05, "loss": 0.0544, "step": 59310 }, { "epoch": 2.005476858582102, "grad_norm": 0.32532939314842224, "learning_rate": 8.427614035288947e-05, "loss": 0.0616, "step": 59320 }, { "epoch": 2.005814936272355, "grad_norm": 0.4526248276233673, "learning_rate": 8.42701208313346e-05, "loss": 0.0533, "step": 59330 }, { "epoch": 2.0061530139626087, "grad_norm": 0.8865006566047668, "learning_rate": 8.42641003728475e-05, "loss": 0.0567, "step": 59340 }, { "epoch": 2.0064910916528618, "grad_norm": 0.34613272547721863, "learning_rate": 8.425807897759277e-05, "loss": 0.0585, "step": 59350 }, { "epoch": 2.006829169343115, "grad_norm": 0.5359576940536499, "learning_rate": 8.425205664573504e-05, "loss": 0.0593, "step": 59360 }, { "epoch": 2.0071672470333684, "grad_norm": 0.6939551830291748, "learning_rate": 8.424603337743892e-05, "loss": 0.065, "step": 59370 }, { "epoch": 2.0075053247236214, "grad_norm": 0.6180492043495178, "learning_rate": 8.424000917286914e-05, "loss": 0.0576, "step": 59380 }, { "epoch": 2.0078434024138745, "grad_norm": 0.6605892181396484, "learning_rate": 8.423398403219036e-05, "loss": 0.0604, "step": 59390 }, { "epoch": 2.008181480104128, "grad_norm": 0.3295442759990692, "learning_rate": 8.422795795556732e-05, "loss": 0.0585, "step": 59400 }, { "epoch": 2.008519557794381, "grad_norm": 0.37546712160110474, "learning_rate": 8.422193094316478e-05, "loss": 0.0594, "step": 59410 }, { "epoch": 2.0088576354846346, "grad_norm": 0.36842986941337585, "learning_rate": 8.42159029951475e-05, "loss": 0.0595, "step": 59420 }, { "epoch": 2.0091957131748877, "grad_norm": 0.9117990136146545, "learning_rate": 8.420987411168029e-05, "loss": 0.0639, "step": 59430 }, { "epoch": 2.0095337908651407, "grad_norm": 0.6185997724533081, "learning_rate": 8.420384429292796e-05, "loss": 0.0606, "step": 59440 }, { "epoch": 2.0098718685553942, "grad_norm": 0.4177935719490051, "learning_rate": 8.419781353905538e-05, "loss": 0.0559, "step": 59450 }, { "epoch": 2.0102099462456473, "grad_norm": 0.6357833743095398, "learning_rate": 8.419178185022743e-05, "loss": 0.0607, "step": 59460 }, { "epoch": 2.0105480239359004, "grad_norm": 0.4563556909561157, "learning_rate": 8.418574922660901e-05, "loss": 0.0638, "step": 59470 }, { "epoch": 2.010886101626154, "grad_norm": 0.43588584661483765, "learning_rate": 8.417971566836505e-05, "loss": 0.0587, "step": 59480 }, { "epoch": 2.011224179316407, "grad_norm": 0.42610880732536316, "learning_rate": 8.417368117566051e-05, "loss": 0.0617, "step": 59490 }, { "epoch": 2.01156225700666, "grad_norm": 0.572185754776001, "learning_rate": 8.416764574866035e-05, "loss": 0.0556, "step": 59500 }, { "epoch": 2.0119003346969135, "grad_norm": 0.368215948343277, "learning_rate": 8.416160938752961e-05, "loss": 0.0568, "step": 59510 }, { "epoch": 2.0122384123871666, "grad_norm": 0.5301235914230347, "learning_rate": 8.415557209243328e-05, "loss": 0.0522, "step": 59520 }, { "epoch": 2.0125764900774197, "grad_norm": 0.33902692794799805, "learning_rate": 8.414953386353645e-05, "loss": 0.0597, "step": 59530 }, { "epoch": 2.012914567767673, "grad_norm": 0.6266408562660217, "learning_rate": 8.41434947010042e-05, "loss": 0.0588, "step": 59540 }, { "epoch": 2.0132526454579263, "grad_norm": 1.1272170543670654, "learning_rate": 8.413745460500163e-05, "loss": 0.0616, "step": 59550 }, { "epoch": 2.0135907231481793, "grad_norm": 0.3929324746131897, "learning_rate": 8.413141357569387e-05, "loss": 0.0579, "step": 59560 }, { "epoch": 2.013928800838433, "grad_norm": 0.5152976512908936, "learning_rate": 8.41253716132461e-05, "loss": 0.0599, "step": 59570 }, { "epoch": 2.014266878528686, "grad_norm": 0.6931008696556091, "learning_rate": 8.411932871782347e-05, "loss": 0.055, "step": 59580 }, { "epoch": 2.014604956218939, "grad_norm": 0.5902965664863586, "learning_rate": 8.411328488959123e-05, "loss": 0.0608, "step": 59590 }, { "epoch": 2.0149430339091925, "grad_norm": 0.5205056667327881, "learning_rate": 8.410724012871459e-05, "loss": 0.063, "step": 59600 }, { "epoch": 2.0152811115994456, "grad_norm": 0.6533347964286804, "learning_rate": 8.41011944353588e-05, "loss": 0.0541, "step": 59610 }, { "epoch": 2.0156191892896986, "grad_norm": 0.41433775424957275, "learning_rate": 8.409514780968918e-05, "loss": 0.0559, "step": 59620 }, { "epoch": 2.015957266979952, "grad_norm": 0.4270528554916382, "learning_rate": 8.408910025187103e-05, "loss": 0.059, "step": 59630 }, { "epoch": 2.016295344670205, "grad_norm": 1.1750614643096924, "learning_rate": 8.40830517620697e-05, "loss": 0.0555, "step": 59640 }, { "epoch": 2.0166334223604583, "grad_norm": 0.5333185195922852, "learning_rate": 8.407700234045051e-05, "loss": 0.056, "step": 59650 }, { "epoch": 2.016971500050712, "grad_norm": 0.42014721035957336, "learning_rate": 8.407095198717889e-05, "loss": 0.0574, "step": 59660 }, { "epoch": 2.017309577740965, "grad_norm": 0.5800075531005859, "learning_rate": 8.406490070242024e-05, "loss": 0.0623, "step": 59670 }, { "epoch": 2.017647655431218, "grad_norm": 0.49792400002479553, "learning_rate": 8.405884848634003e-05, "loss": 0.0571, "step": 59680 }, { "epoch": 2.0179857331214714, "grad_norm": 0.31281042098999023, "learning_rate": 8.405279533910366e-05, "loss": 0.0577, "step": 59690 }, { "epoch": 2.0183238108117245, "grad_norm": 0.47095486521720886, "learning_rate": 8.404674126087668e-05, "loss": 0.0615, "step": 59700 }, { "epoch": 2.0186618885019776, "grad_norm": 0.3107616901397705, "learning_rate": 8.404068625182457e-05, "loss": 0.0572, "step": 59710 }, { "epoch": 2.018999966192231, "grad_norm": 0.8977746367454529, "learning_rate": 8.403463031211289e-05, "loss": 0.0534, "step": 59720 }, { "epoch": 2.019338043882484, "grad_norm": 0.444542795419693, "learning_rate": 8.402857344190722e-05, "loss": 0.0656, "step": 59730 }, { "epoch": 2.0196761215727372, "grad_norm": 0.7197684049606323, "learning_rate": 8.402251564137311e-05, "loss": 0.0552, "step": 59740 }, { "epoch": 2.0200141992629908, "grad_norm": 0.36467409133911133, "learning_rate": 8.401645691067624e-05, "loss": 0.0617, "step": 59750 }, { "epoch": 2.020352276953244, "grad_norm": 0.46656981110572815, "learning_rate": 8.401039724998218e-05, "loss": 0.0626, "step": 59760 }, { "epoch": 2.020690354643497, "grad_norm": 0.4943047761917114, "learning_rate": 8.400433665945665e-05, "loss": 0.0605, "step": 59770 }, { "epoch": 2.0210284323337504, "grad_norm": 0.520890474319458, "learning_rate": 8.399827513926534e-05, "loss": 0.063, "step": 59780 }, { "epoch": 2.0213665100240035, "grad_norm": 0.33224374055862427, "learning_rate": 8.399221268957395e-05, "loss": 0.0589, "step": 59790 }, { "epoch": 2.0217045877142565, "grad_norm": 0.477402001619339, "learning_rate": 8.398614931054825e-05, "loss": 0.0618, "step": 59800 }, { "epoch": 2.02204266540451, "grad_norm": 0.2717677056789398, "learning_rate": 8.398008500235397e-05, "loss": 0.0583, "step": 59810 }, { "epoch": 2.022380743094763, "grad_norm": 0.5594728589057922, "learning_rate": 8.397401976515695e-05, "loss": 0.0576, "step": 59820 }, { "epoch": 2.022718820785016, "grad_norm": 0.3322344124317169, "learning_rate": 8.396795359912298e-05, "loss": 0.0585, "step": 59830 }, { "epoch": 2.0230568984752697, "grad_norm": 0.5090760588645935, "learning_rate": 8.396188650441793e-05, "loss": 0.0553, "step": 59840 }, { "epoch": 2.0233949761655228, "grad_norm": 0.40230074524879456, "learning_rate": 8.395581848120766e-05, "loss": 0.0567, "step": 59850 }, { "epoch": 2.023733053855776, "grad_norm": 0.694970428943634, "learning_rate": 8.394974952965806e-05, "loss": 0.0593, "step": 59860 }, { "epoch": 2.0240711315460294, "grad_norm": 0.5620299577713013, "learning_rate": 8.394367964993508e-05, "loss": 0.06, "step": 59870 }, { "epoch": 2.0244092092362824, "grad_norm": 0.3792417049407959, "learning_rate": 8.393760884220465e-05, "loss": 0.053, "step": 59880 }, { "epoch": 2.0247472869265355, "grad_norm": 0.6927697062492371, "learning_rate": 8.393153710663274e-05, "loss": 0.0619, "step": 59890 }, { "epoch": 2.025085364616789, "grad_norm": 0.39967355132102966, "learning_rate": 8.392546444338534e-05, "loss": 0.0582, "step": 59900 }, { "epoch": 2.025423442307042, "grad_norm": 0.39070558547973633, "learning_rate": 8.39193908526285e-05, "loss": 0.0583, "step": 59910 }, { "epoch": 2.0257615199972956, "grad_norm": 0.3169401288032532, "learning_rate": 8.391331633452824e-05, "loss": 0.0545, "step": 59920 }, { "epoch": 2.0260995976875487, "grad_norm": 0.4901239275932312, "learning_rate": 8.390724088925068e-05, "loss": 0.059, "step": 59930 }, { "epoch": 2.0264376753778017, "grad_norm": 0.43912747502326965, "learning_rate": 8.390116451696188e-05, "loss": 0.0583, "step": 59940 }, { "epoch": 2.0267757530680552, "grad_norm": 0.7226338386535645, "learning_rate": 8.389508721782797e-05, "loss": 0.0611, "step": 59950 }, { "epoch": 2.0271138307583083, "grad_norm": 0.44491973519325256, "learning_rate": 8.388900899201511e-05, "loss": 0.0556, "step": 59960 }, { "epoch": 2.0274519084485614, "grad_norm": 0.40910565853118896, "learning_rate": 8.388292983968948e-05, "loss": 0.0598, "step": 59970 }, { "epoch": 2.027789986138815, "grad_norm": 0.501282811164856, "learning_rate": 8.387684976101728e-05, "loss": 0.0562, "step": 59980 }, { "epoch": 2.028128063829068, "grad_norm": 0.3702795207500458, "learning_rate": 8.387076875616473e-05, "loss": 0.0604, "step": 59990 }, { "epoch": 2.028466141519321, "grad_norm": 0.6454729437828064, "learning_rate": 8.38646868252981e-05, "loss": 0.0557, "step": 60000 }, { "epoch": 2.0288042192095745, "grad_norm": 0.42153671383857727, "learning_rate": 8.385860396858365e-05, "loss": 0.0577, "step": 60010 }, { "epoch": 2.0291422968998276, "grad_norm": 0.36053207516670227, "learning_rate": 8.385252018618767e-05, "loss": 0.0573, "step": 60020 }, { "epoch": 2.0294803745900807, "grad_norm": 0.5200681090354919, "learning_rate": 8.384643547827652e-05, "loss": 0.06, "step": 60030 }, { "epoch": 2.029818452280334, "grad_norm": 0.335883229970932, "learning_rate": 8.384034984501655e-05, "loss": 0.0559, "step": 60040 }, { "epoch": 2.0301565299705873, "grad_norm": 1.443895697593689, "learning_rate": 8.383426328657411e-05, "loss": 0.0616, "step": 60050 }, { "epoch": 2.0304946076608403, "grad_norm": 0.6276646852493286, "learning_rate": 8.382817580311562e-05, "loss": 0.0575, "step": 60060 }, { "epoch": 2.030832685351094, "grad_norm": 0.26016131043434143, "learning_rate": 8.382208739480752e-05, "loss": 0.0595, "step": 60070 }, { "epoch": 2.031170763041347, "grad_norm": 0.3363260328769684, "learning_rate": 8.381599806181625e-05, "loss": 0.0532, "step": 60080 }, { "epoch": 2.0315088407316, "grad_norm": 1.0781524181365967, "learning_rate": 8.380990780430831e-05, "loss": 0.0534, "step": 60090 }, { "epoch": 2.0318469184218535, "grad_norm": 0.3863368630409241, "learning_rate": 8.380381662245019e-05, "loss": 0.0617, "step": 60100 }, { "epoch": 2.0321849961121066, "grad_norm": 0.4945204555988312, "learning_rate": 8.379772451640841e-05, "loss": 0.0597, "step": 60110 }, { "epoch": 2.0325230738023596, "grad_norm": 0.41147708892822266, "learning_rate": 8.379163148634956e-05, "loss": 0.0555, "step": 60120 }, { "epoch": 2.032861151492613, "grad_norm": 0.41079598665237427, "learning_rate": 8.378553753244019e-05, "loss": 0.0619, "step": 60130 }, { "epoch": 2.033199229182866, "grad_norm": 0.371168315410614, "learning_rate": 8.377944265484691e-05, "loss": 0.0549, "step": 60140 }, { "epoch": 2.0335373068731193, "grad_norm": 0.27472612261772156, "learning_rate": 8.377334685373636e-05, "loss": 0.052, "step": 60150 }, { "epoch": 2.033875384563373, "grad_norm": 0.42795610427856445, "learning_rate": 8.376725012927522e-05, "loss": 0.0589, "step": 60160 }, { "epoch": 2.034213462253626, "grad_norm": 0.38332629203796387, "learning_rate": 8.376115248163012e-05, "loss": 0.0582, "step": 60170 }, { "epoch": 2.034551539943879, "grad_norm": 0.6489148139953613, "learning_rate": 8.37550539109678e-05, "loss": 0.0618, "step": 60180 }, { "epoch": 2.0348896176341325, "grad_norm": 0.27236083149909973, "learning_rate": 8.374895441745498e-05, "loss": 0.0535, "step": 60190 }, { "epoch": 2.0352276953243855, "grad_norm": 0.3726853132247925, "learning_rate": 8.374285400125844e-05, "loss": 0.0621, "step": 60200 }, { "epoch": 2.0355657730146386, "grad_norm": 0.3919707238674164, "learning_rate": 8.373675266254495e-05, "loss": 0.0557, "step": 60210 }, { "epoch": 2.035903850704892, "grad_norm": 1.2953476905822754, "learning_rate": 8.373065040148129e-05, "loss": 0.0532, "step": 60220 }, { "epoch": 2.036241928395145, "grad_norm": 0.4078902304172516, "learning_rate": 8.372454721823435e-05, "loss": 0.0612, "step": 60230 }, { "epoch": 2.0365800060853982, "grad_norm": 0.4725557863712311, "learning_rate": 8.371844311297094e-05, "loss": 0.0572, "step": 60240 }, { "epoch": 2.0369180837756518, "grad_norm": 0.3851125240325928, "learning_rate": 8.371233808585796e-05, "loss": 0.0558, "step": 60250 }, { "epoch": 2.037256161465905, "grad_norm": 0.5818977952003479, "learning_rate": 8.370623213706234e-05, "loss": 0.0631, "step": 60260 }, { "epoch": 2.037594239156158, "grad_norm": 0.4159165918827057, "learning_rate": 8.3700125266751e-05, "loss": 0.0622, "step": 60270 }, { "epoch": 2.0379323168464114, "grad_norm": 0.2312120944261551, "learning_rate": 8.369401747509088e-05, "loss": 0.0549, "step": 60280 }, { "epoch": 2.0382703945366645, "grad_norm": 0.34428051114082336, "learning_rate": 8.368790876224898e-05, "loss": 0.0589, "step": 60290 }, { "epoch": 2.0386084722269175, "grad_norm": 0.47857850790023804, "learning_rate": 8.368179912839231e-05, "loss": 0.0578, "step": 60300 }, { "epoch": 2.038946549917171, "grad_norm": 1.145686388015747, "learning_rate": 8.367568857368791e-05, "loss": 0.0548, "step": 60310 }, { "epoch": 2.039284627607424, "grad_norm": 0.5774300694465637, "learning_rate": 8.366957709830284e-05, "loss": 0.0603, "step": 60320 }, { "epoch": 2.039622705297677, "grad_norm": 1.0890377759933472, "learning_rate": 8.366346470240418e-05, "loss": 0.0551, "step": 60330 }, { "epoch": 2.0399607829879307, "grad_norm": 0.571401834487915, "learning_rate": 8.365735138615904e-05, "loss": 0.0612, "step": 60340 }, { "epoch": 2.040298860678184, "grad_norm": 0.3291206657886505, "learning_rate": 8.365123714973454e-05, "loss": 0.0505, "step": 60350 }, { "epoch": 2.0406369383684373, "grad_norm": 0.5950897336006165, "learning_rate": 8.364512199329789e-05, "loss": 0.0532, "step": 60360 }, { "epoch": 2.0409750160586904, "grad_norm": 0.5066255927085876, "learning_rate": 8.363900591701622e-05, "loss": 0.0624, "step": 60370 }, { "epoch": 2.0413130937489434, "grad_norm": 0.2974146008491516, "learning_rate": 8.363288892105679e-05, "loss": 0.0572, "step": 60380 }, { "epoch": 2.041651171439197, "grad_norm": 0.653049111366272, "learning_rate": 8.362677100558678e-05, "loss": 0.0592, "step": 60390 }, { "epoch": 2.04198924912945, "grad_norm": 0.21819955110549927, "learning_rate": 8.362065217077351e-05, "loss": 0.0523, "step": 60400 }, { "epoch": 2.042327326819703, "grad_norm": 0.4135172367095947, "learning_rate": 8.361453241678423e-05, "loss": 0.0567, "step": 60410 }, { "epoch": 2.0426654045099566, "grad_norm": 0.4253002107143402, "learning_rate": 8.360841174378626e-05, "loss": 0.0547, "step": 60420 }, { "epoch": 2.0430034822002097, "grad_norm": 0.43881818652153015, "learning_rate": 8.360229015194693e-05, "loss": 0.0547, "step": 60430 }, { "epoch": 2.0433415598904627, "grad_norm": 0.6732698082923889, "learning_rate": 8.359616764143363e-05, "loss": 0.0637, "step": 60440 }, { "epoch": 2.0436796375807162, "grad_norm": 0.7157981395721436, "learning_rate": 8.359004421241374e-05, "loss": 0.0585, "step": 60450 }, { "epoch": 2.0440177152709693, "grad_norm": 1.0947190523147583, "learning_rate": 8.358391986505463e-05, "loss": 0.0549, "step": 60460 }, { "epoch": 2.0443557929612224, "grad_norm": 0.4499533772468567, "learning_rate": 8.357779459952377e-05, "loss": 0.0596, "step": 60470 }, { "epoch": 2.044693870651476, "grad_norm": 0.3578614890575409, "learning_rate": 8.357166841598863e-05, "loss": 0.0608, "step": 60480 }, { "epoch": 2.045031948341729, "grad_norm": 0.5042595863342285, "learning_rate": 8.35655413146167e-05, "loss": 0.0558, "step": 60490 }, { "epoch": 2.045370026031982, "grad_norm": 0.5665378570556641, "learning_rate": 8.355941329557546e-05, "loss": 0.0602, "step": 60500 }, { "epoch": 2.0457081037222355, "grad_norm": 0.4965118169784546, "learning_rate": 8.355328435903247e-05, "loss": 0.0637, "step": 60510 }, { "epoch": 2.0460461814124886, "grad_norm": 0.3645302951335907, "learning_rate": 8.354715450515531e-05, "loss": 0.0545, "step": 60520 }, { "epoch": 2.0463842591027417, "grad_norm": 0.6034276485443115, "learning_rate": 8.354102373411155e-05, "loss": 0.0568, "step": 60530 }, { "epoch": 2.046722336792995, "grad_norm": 0.5741555690765381, "learning_rate": 8.353489204606877e-05, "loss": 0.057, "step": 60540 }, { "epoch": 2.0470604144832483, "grad_norm": 2.5081417560577393, "learning_rate": 8.352875944119465e-05, "loss": 0.0594, "step": 60550 }, { "epoch": 2.0473984921735013, "grad_norm": 2.0257275104522705, "learning_rate": 8.352262591965686e-05, "loss": 0.0657, "step": 60560 }, { "epoch": 2.047736569863755, "grad_norm": 4.157318592071533, "learning_rate": 8.351649148162306e-05, "loss": 0.0582, "step": 60570 }, { "epoch": 2.048074647554008, "grad_norm": 0.5899616479873657, "learning_rate": 8.351035612726099e-05, "loss": 0.0587, "step": 60580 }, { "epoch": 2.048412725244261, "grad_norm": 0.34095296263694763, "learning_rate": 8.350421985673836e-05, "loss": 0.0522, "step": 60590 }, { "epoch": 2.0487508029345145, "grad_norm": 1.2081818580627441, "learning_rate": 8.349808267022293e-05, "loss": 0.0545, "step": 60600 }, { "epoch": 2.0490888806247676, "grad_norm": 0.35433268547058105, "learning_rate": 8.349194456788252e-05, "loss": 0.0617, "step": 60610 }, { "epoch": 2.0494269583150206, "grad_norm": 0.9836782813072205, "learning_rate": 8.348580554988493e-05, "loss": 0.0583, "step": 60620 }, { "epoch": 2.049765036005274, "grad_norm": 0.40315011143684387, "learning_rate": 8.3479665616398e-05, "loss": 0.0596, "step": 60630 }, { "epoch": 2.0501031136955272, "grad_norm": 0.5723370909690857, "learning_rate": 8.347352476758959e-05, "loss": 0.0592, "step": 60640 }, { "epoch": 2.0504411913857803, "grad_norm": 0.7277238368988037, "learning_rate": 8.346738300362757e-05, "loss": 0.0643, "step": 60650 }, { "epoch": 2.050779269076034, "grad_norm": 0.3879607319831848, "learning_rate": 8.346124032467988e-05, "loss": 0.0542, "step": 60660 }, { "epoch": 2.051117346766287, "grad_norm": 0.524992048740387, "learning_rate": 8.345509673091446e-05, "loss": 0.0667, "step": 60670 }, { "epoch": 2.05145542445654, "grad_norm": 0.3220777213573456, "learning_rate": 8.344895222249926e-05, "loss": 0.0579, "step": 60680 }, { "epoch": 2.0517935021467935, "grad_norm": 0.5217911601066589, "learning_rate": 8.344280679960228e-05, "loss": 0.0618, "step": 60690 }, { "epoch": 2.0521315798370465, "grad_norm": 0.3594232201576233, "learning_rate": 8.343666046239152e-05, "loss": 0.0516, "step": 60700 }, { "epoch": 2.0524696575272996, "grad_norm": 0.8138737082481384, "learning_rate": 8.343051321103501e-05, "loss": 0.0578, "step": 60710 }, { "epoch": 2.052807735217553, "grad_norm": 0.5712116956710815, "learning_rate": 8.342436504570084e-05, "loss": 0.0619, "step": 60720 }, { "epoch": 2.053145812907806, "grad_norm": 1.317941665649414, "learning_rate": 8.341821596655708e-05, "loss": 0.0577, "step": 60730 }, { "epoch": 2.0534838905980592, "grad_norm": 1.5146563053131104, "learning_rate": 8.341206597377186e-05, "loss": 0.0592, "step": 60740 }, { "epoch": 2.0538219682883128, "grad_norm": 0.8440195918083191, "learning_rate": 8.34059150675133e-05, "loss": 0.0605, "step": 60750 }, { "epoch": 2.054160045978566, "grad_norm": 0.380001425743103, "learning_rate": 8.339976324794958e-05, "loss": 0.0601, "step": 60760 }, { "epoch": 2.054498123668819, "grad_norm": 0.40400949120521545, "learning_rate": 8.339361051524888e-05, "loss": 0.0524, "step": 60770 }, { "epoch": 2.0548362013590724, "grad_norm": 0.38152387738227844, "learning_rate": 8.338745686957942e-05, "loss": 0.057, "step": 60780 }, { "epoch": 2.0551742790493255, "grad_norm": 0.5077971816062927, "learning_rate": 8.338130231110941e-05, "loss": 0.0567, "step": 60790 }, { "epoch": 2.0555123567395785, "grad_norm": 0.6458737850189209, "learning_rate": 8.337514684000716e-05, "loss": 0.0568, "step": 60800 }, { "epoch": 2.055850434429832, "grad_norm": 0.6016498804092407, "learning_rate": 8.336899045644091e-05, "loss": 0.0595, "step": 60810 }, { "epoch": 2.056188512120085, "grad_norm": 0.38349980115890503, "learning_rate": 8.336283316057903e-05, "loss": 0.0552, "step": 60820 }, { "epoch": 2.056526589810338, "grad_norm": 0.45832911133766174, "learning_rate": 8.33566749525898e-05, "loss": 0.0596, "step": 60830 }, { "epoch": 2.0568646675005917, "grad_norm": 0.5783625841140747, "learning_rate": 8.335051583264161e-05, "loss": 0.061, "step": 60840 }, { "epoch": 2.057202745190845, "grad_norm": 0.42823120951652527, "learning_rate": 8.334435580090285e-05, "loss": 0.0551, "step": 60850 }, { "epoch": 2.057540822881098, "grad_norm": 0.48357105255126953, "learning_rate": 8.333819485754193e-05, "loss": 0.0637, "step": 60860 }, { "epoch": 2.0578789005713514, "grad_norm": 0.517127513885498, "learning_rate": 8.333203300272728e-05, "loss": 0.0623, "step": 60870 }, { "epoch": 2.0582169782616044, "grad_norm": 0.4053521156311035, "learning_rate": 8.332587023662738e-05, "loss": 0.0604, "step": 60880 }, { "epoch": 2.058555055951858, "grad_norm": 0.8821384310722351, "learning_rate": 8.33197065594107e-05, "loss": 0.0595, "step": 60890 }, { "epoch": 2.058893133642111, "grad_norm": 0.7868106365203857, "learning_rate": 8.331354197124576e-05, "loss": 0.0601, "step": 60900 }, { "epoch": 2.059231211332364, "grad_norm": 0.37029606103897095, "learning_rate": 8.330737647230111e-05, "loss": 0.0624, "step": 60910 }, { "epoch": 2.0595692890226176, "grad_norm": 0.3588319420814514, "learning_rate": 8.33012100627453e-05, "loss": 0.0551, "step": 60920 }, { "epoch": 2.0599073667128707, "grad_norm": 0.3848119378089905, "learning_rate": 8.32950427427469e-05, "loss": 0.0523, "step": 60930 }, { "epoch": 2.0602454444031237, "grad_norm": 1.2182915210723877, "learning_rate": 8.328887451247455e-05, "loss": 0.062, "step": 60940 }, { "epoch": 2.0605835220933773, "grad_norm": 0.40197688341140747, "learning_rate": 8.328270537209688e-05, "loss": 0.0588, "step": 60950 }, { "epoch": 2.0609215997836303, "grad_norm": 0.6172835230827332, "learning_rate": 8.327653532178254e-05, "loss": 0.0614, "step": 60960 }, { "epoch": 2.0612596774738834, "grad_norm": 2.0733776092529297, "learning_rate": 8.327036436170024e-05, "loss": 0.0558, "step": 60970 }, { "epoch": 2.061597755164137, "grad_norm": 2.661043643951416, "learning_rate": 8.326419249201867e-05, "loss": 0.0568, "step": 60980 }, { "epoch": 2.06193583285439, "grad_norm": 0.5798721313476562, "learning_rate": 8.325801971290657e-05, "loss": 0.0573, "step": 60990 }, { "epoch": 2.062273910544643, "grad_norm": 0.7200261950492859, "learning_rate": 8.32518460245327e-05, "loss": 0.0558, "step": 61000 }, { "epoch": 2.0626119882348966, "grad_norm": 0.8864979147911072, "learning_rate": 8.324567142706586e-05, "loss": 0.059, "step": 61010 }, { "epoch": 2.0629500659251496, "grad_norm": 0.3647385835647583, "learning_rate": 8.323949592067485e-05, "loss": 0.0616, "step": 61020 }, { "epoch": 2.0632881436154027, "grad_norm": 0.5912408828735352, "learning_rate": 8.323331950552851e-05, "loss": 0.0636, "step": 61030 }, { "epoch": 2.063626221305656, "grad_norm": 0.6645315885543823, "learning_rate": 8.322714218179571e-05, "loss": 0.0582, "step": 61040 }, { "epoch": 2.0639642989959093, "grad_norm": 0.5647448897361755, "learning_rate": 8.322096394964532e-05, "loss": 0.0636, "step": 61050 }, { "epoch": 2.0643023766861623, "grad_norm": 0.45025262236595154, "learning_rate": 8.321478480924624e-05, "loss": 0.0604, "step": 61060 }, { "epoch": 2.064640454376416, "grad_norm": 0.6042224764823914, "learning_rate": 8.320860476076745e-05, "loss": 0.0567, "step": 61070 }, { "epoch": 2.064978532066669, "grad_norm": 0.9461486339569092, "learning_rate": 8.320242380437786e-05, "loss": 0.0564, "step": 61080 }, { "epoch": 2.065316609756922, "grad_norm": 0.44811534881591797, "learning_rate": 8.319624194024649e-05, "loss": 0.0607, "step": 61090 }, { "epoch": 2.0656546874471755, "grad_norm": 2.1543023586273193, "learning_rate": 8.319005916854232e-05, "loss": 0.058, "step": 61100 }, { "epoch": 2.0659927651374286, "grad_norm": 0.38754400610923767, "learning_rate": 8.318387548943441e-05, "loss": 0.0609, "step": 61110 }, { "epoch": 2.0663308428276816, "grad_norm": 0.7717480063438416, "learning_rate": 8.317769090309182e-05, "loss": 0.0653, "step": 61120 }, { "epoch": 2.066668920517935, "grad_norm": 1.0184096097946167, "learning_rate": 8.317150540968362e-05, "loss": 0.0641, "step": 61130 }, { "epoch": 2.0670069982081882, "grad_norm": 0.518878698348999, "learning_rate": 8.316531900937891e-05, "loss": 0.0619, "step": 61140 }, { "epoch": 2.0673450758984413, "grad_norm": 1.3724215030670166, "learning_rate": 8.315913170234686e-05, "loss": 0.0652, "step": 61150 }, { "epoch": 2.067683153588695, "grad_norm": 0.5350247621536255, "learning_rate": 8.31529434887566e-05, "loss": 0.0632, "step": 61160 }, { "epoch": 2.068021231278948, "grad_norm": 0.4500349462032318, "learning_rate": 8.31467543687773e-05, "loss": 0.0586, "step": 61170 }, { "epoch": 2.068359308969201, "grad_norm": 0.5040259957313538, "learning_rate": 8.314056434257822e-05, "loss": 0.0626, "step": 61180 }, { "epoch": 2.0686973866594545, "grad_norm": 0.34946030378341675, "learning_rate": 8.313437341032853e-05, "loss": 0.061, "step": 61190 }, { "epoch": 2.0690354643497075, "grad_norm": 0.6447283029556274, "learning_rate": 8.312818157219755e-05, "loss": 0.0599, "step": 61200 }, { "epoch": 2.0693735420399606, "grad_norm": 0.7866971492767334, "learning_rate": 8.312198882835452e-05, "loss": 0.0631, "step": 61210 }, { "epoch": 2.069711619730214, "grad_norm": 1.2535263299942017, "learning_rate": 8.311579517896875e-05, "loss": 0.0599, "step": 61220 }, { "epoch": 2.070049697420467, "grad_norm": 0.6927381157875061, "learning_rate": 8.310960062420958e-05, "loss": 0.0623, "step": 61230 }, { "epoch": 2.0703877751107203, "grad_norm": 0.35701748728752136, "learning_rate": 8.310340516424637e-05, "loss": 0.0601, "step": 61240 }, { "epoch": 2.0707258528009738, "grad_norm": 0.3803139328956604, "learning_rate": 8.30972087992485e-05, "loss": 0.0572, "step": 61250 }, { "epoch": 2.071063930491227, "grad_norm": 0.9344675540924072, "learning_rate": 8.309101152938538e-05, "loss": 0.0599, "step": 61260 }, { "epoch": 2.07140200818148, "grad_norm": 0.5326681137084961, "learning_rate": 8.308481335482644e-05, "loss": 0.0563, "step": 61270 }, { "epoch": 2.0717400858717334, "grad_norm": 0.5087548494338989, "learning_rate": 8.307861427574113e-05, "loss": 0.0598, "step": 61280 }, { "epoch": 2.0720781635619865, "grad_norm": 0.44937631487846375, "learning_rate": 8.307241429229893e-05, "loss": 0.0598, "step": 61290 }, { "epoch": 2.0724162412522396, "grad_norm": 0.5376158952713013, "learning_rate": 8.306621340466934e-05, "loss": 0.0647, "step": 61300 }, { "epoch": 2.072754318942493, "grad_norm": 0.49855953454971313, "learning_rate": 8.30600116130219e-05, "loss": 0.06, "step": 61310 }, { "epoch": 2.073092396632746, "grad_norm": 0.46344494819641113, "learning_rate": 8.305380891752616e-05, "loss": 0.059, "step": 61320 }, { "epoch": 2.0734304743229997, "grad_norm": 1.376519799232483, "learning_rate": 8.304760531835172e-05, "loss": 0.0545, "step": 61330 }, { "epoch": 2.0737685520132527, "grad_norm": 0.4946962594985962, "learning_rate": 8.304140081566815e-05, "loss": 0.0568, "step": 61340 }, { "epoch": 2.074106629703506, "grad_norm": 0.4454230070114136, "learning_rate": 8.303519540964511e-05, "loss": 0.0601, "step": 61350 }, { "epoch": 2.0744447073937593, "grad_norm": 0.699123203754425, "learning_rate": 8.302898910045222e-05, "loss": 0.0598, "step": 61360 }, { "epoch": 2.0747827850840124, "grad_norm": 0.5872430801391602, "learning_rate": 8.302278188825918e-05, "loss": 0.0593, "step": 61370 }, { "epoch": 2.0751208627742654, "grad_norm": 0.8845314383506775, "learning_rate": 8.30165737732357e-05, "loss": 0.0604, "step": 61380 }, { "epoch": 2.075458940464519, "grad_norm": 0.3388826251029968, "learning_rate": 8.30103647555515e-05, "loss": 0.0593, "step": 61390 }, { "epoch": 2.075797018154772, "grad_norm": 0.6585696339607239, "learning_rate": 8.300415483537631e-05, "loss": 0.0626, "step": 61400 }, { "epoch": 2.076135095845025, "grad_norm": 0.29670748114585876, "learning_rate": 8.299794401287994e-05, "loss": 0.0594, "step": 61410 }, { "epoch": 2.0764731735352786, "grad_norm": 0.4654929041862488, "learning_rate": 8.299173228823219e-05, "loss": 0.0574, "step": 61420 }, { "epoch": 2.0768112512255317, "grad_norm": 0.5286101698875427, "learning_rate": 8.298551966160284e-05, "loss": 0.0583, "step": 61430 }, { "epoch": 2.0771493289157847, "grad_norm": 0.5034015774726868, "learning_rate": 8.297930613316178e-05, "loss": 0.055, "step": 61440 }, { "epoch": 2.0774874066060383, "grad_norm": 0.44004014134407043, "learning_rate": 8.29730917030789e-05, "loss": 0.0575, "step": 61450 }, { "epoch": 2.0778254842962913, "grad_norm": 1.0199344158172607, "learning_rate": 8.296687637152407e-05, "loss": 0.0581, "step": 61460 }, { "epoch": 2.0781635619865444, "grad_norm": 0.4883224070072174, "learning_rate": 8.296066013866723e-05, "loss": 0.0638, "step": 61470 }, { "epoch": 2.078501639676798, "grad_norm": 0.48828354477882385, "learning_rate": 8.295444300467832e-05, "loss": 0.0599, "step": 61480 }, { "epoch": 2.078839717367051, "grad_norm": 0.3134020268917084, "learning_rate": 8.294822496972732e-05, "loss": 0.0576, "step": 61490 }, { "epoch": 2.079177795057304, "grad_norm": 0.3369336426258087, "learning_rate": 8.294200603398423e-05, "loss": 0.0581, "step": 61500 }, { "epoch": 2.0795158727475576, "grad_norm": 0.8404374122619629, "learning_rate": 8.293578619761906e-05, "loss": 0.056, "step": 61510 }, { "epoch": 2.0798539504378106, "grad_norm": 0.6290745139122009, "learning_rate": 8.292956546080189e-05, "loss": 0.0589, "step": 61520 }, { "epoch": 2.0801920281280637, "grad_norm": 0.37898895144462585, "learning_rate": 8.292334382370275e-05, "loss": 0.0641, "step": 61530 }, { "epoch": 2.080530105818317, "grad_norm": 0.4579254388809204, "learning_rate": 8.291712128649177e-05, "loss": 0.0567, "step": 61540 }, { "epoch": 2.0808681835085703, "grad_norm": 0.35229840874671936, "learning_rate": 8.291089784933906e-05, "loss": 0.0551, "step": 61550 }, { "epoch": 2.0812062611988233, "grad_norm": 0.44579067826271057, "learning_rate": 8.290467351241475e-05, "loss": 0.0558, "step": 61560 }, { "epoch": 2.081544338889077, "grad_norm": 0.36301472783088684, "learning_rate": 8.289844827588906e-05, "loss": 0.0619, "step": 61570 }, { "epoch": 2.08188241657933, "grad_norm": 0.5107359290122986, "learning_rate": 8.289222213993213e-05, "loss": 0.0592, "step": 61580 }, { "epoch": 2.082220494269583, "grad_norm": 0.48165732622146606, "learning_rate": 8.288599510471421e-05, "loss": 0.0558, "step": 61590 }, { "epoch": 2.0825585719598365, "grad_norm": 0.28171950578689575, "learning_rate": 8.287976717040553e-05, "loss": 0.0564, "step": 61600 }, { "epoch": 2.0828966496500896, "grad_norm": 0.402536541223526, "learning_rate": 8.28735383371764e-05, "loss": 0.061, "step": 61610 }, { "epoch": 2.0832347273403427, "grad_norm": 0.39122119545936584, "learning_rate": 8.286730860519704e-05, "loss": 0.057, "step": 61620 }, { "epoch": 2.083572805030596, "grad_norm": 0.5118529796600342, "learning_rate": 8.286107797463782e-05, "loss": 0.06, "step": 61630 }, { "epoch": 2.0839108827208492, "grad_norm": 0.5928487777709961, "learning_rate": 8.285484644566909e-05, "loss": 0.0565, "step": 61640 }, { "epoch": 2.0842489604111023, "grad_norm": 0.5524837374687195, "learning_rate": 8.284861401846119e-05, "loss": 0.0558, "step": 61650 }, { "epoch": 2.084587038101356, "grad_norm": 0.5230762958526611, "learning_rate": 8.284238069318453e-05, "loss": 0.0611, "step": 61660 }, { "epoch": 2.084925115791609, "grad_norm": 0.3933207392692566, "learning_rate": 8.283614647000952e-05, "loss": 0.0595, "step": 61670 }, { "epoch": 2.085263193481862, "grad_norm": 0.3796652853488922, "learning_rate": 8.282991134910658e-05, "loss": 0.0647, "step": 61680 }, { "epoch": 2.0856012711721155, "grad_norm": 0.4311101734638214, "learning_rate": 8.282367533064624e-05, "loss": 0.058, "step": 61690 }, { "epoch": 2.0859393488623685, "grad_norm": 0.6720622777938843, "learning_rate": 8.281743841479891e-05, "loss": 0.0632, "step": 61700 }, { "epoch": 2.0862774265526216, "grad_norm": 0.6415063142776489, "learning_rate": 8.281120060173515e-05, "loss": 0.0607, "step": 61710 }, { "epoch": 2.086615504242875, "grad_norm": 0.5833033919334412, "learning_rate": 8.28049618916255e-05, "loss": 0.0576, "step": 61720 }, { "epoch": 2.086953581933128, "grad_norm": 0.7511762380599976, "learning_rate": 8.279872228464052e-05, "loss": 0.0575, "step": 61730 }, { "epoch": 2.0872916596233813, "grad_norm": 0.6966856122016907, "learning_rate": 8.279248178095078e-05, "loss": 0.0657, "step": 61740 }, { "epoch": 2.0876297373136348, "grad_norm": 0.4423602223396301, "learning_rate": 8.278624038072692e-05, "loss": 0.0654, "step": 61750 }, { "epoch": 2.087967815003888, "grad_norm": 0.614661693572998, "learning_rate": 8.277999808413956e-05, "loss": 0.0624, "step": 61760 }, { "epoch": 2.088305892694141, "grad_norm": 0.6823208332061768, "learning_rate": 8.277375489135937e-05, "loss": 0.0539, "step": 61770 }, { "epoch": 2.0886439703843944, "grad_norm": 0.5697227716445923, "learning_rate": 8.276751080255703e-05, "loss": 0.0613, "step": 61780 }, { "epoch": 2.0889820480746475, "grad_norm": 0.5093308091163635, "learning_rate": 8.276126581790327e-05, "loss": 0.0626, "step": 61790 }, { "epoch": 2.0893201257649006, "grad_norm": 0.5603193640708923, "learning_rate": 8.275501993756878e-05, "loss": 0.0602, "step": 61800 }, { "epoch": 2.089658203455154, "grad_norm": 0.4446425437927246, "learning_rate": 8.274877316172438e-05, "loss": 0.0632, "step": 61810 }, { "epoch": 2.089996281145407, "grad_norm": 0.6932706236839294, "learning_rate": 8.27425254905408e-05, "loss": 0.0604, "step": 61820 }, { "epoch": 2.09033435883566, "grad_norm": 2.8970847129821777, "learning_rate": 8.273627692418889e-05, "loss": 0.0589, "step": 61830 }, { "epoch": 2.0906724365259137, "grad_norm": 0.7534291744232178, "learning_rate": 8.273002746283946e-05, "loss": 0.0589, "step": 61840 }, { "epoch": 2.091010514216167, "grad_norm": 0.43859246373176575, "learning_rate": 8.272377710666339e-05, "loss": 0.0564, "step": 61850 }, { "epoch": 2.0913485919064203, "grad_norm": 0.44451722502708435, "learning_rate": 8.271752585583155e-05, "loss": 0.056, "step": 61860 }, { "epoch": 2.0916866695966734, "grad_norm": 0.8377527594566345, "learning_rate": 8.271127371051481e-05, "loss": 0.0609, "step": 61870 }, { "epoch": 2.0920247472869264, "grad_norm": 0.5018202662467957, "learning_rate": 8.270502067088417e-05, "loss": 0.0665, "step": 61880 }, { "epoch": 2.09236282497718, "grad_norm": 0.559216320514679, "learning_rate": 8.269876673711053e-05, "loss": 0.0587, "step": 61890 }, { "epoch": 2.092700902667433, "grad_norm": 0.7042052149772644, "learning_rate": 8.269251190936491e-05, "loss": 0.0645, "step": 61900 }, { "epoch": 2.093038980357686, "grad_norm": 0.9814149737358093, "learning_rate": 8.268625618781829e-05, "loss": 0.061, "step": 61910 }, { "epoch": 2.0933770580479396, "grad_norm": 0.6174700260162354, "learning_rate": 8.26799995726417e-05, "loss": 0.0547, "step": 61920 }, { "epoch": 2.0937151357381927, "grad_norm": 0.66315758228302, "learning_rate": 8.26737420640062e-05, "loss": 0.0664, "step": 61930 }, { "epoch": 2.0940532134284457, "grad_norm": 0.49772846698760986, "learning_rate": 8.266748366208285e-05, "loss": 0.0626, "step": 61940 }, { "epoch": 2.0943912911186993, "grad_norm": 1.2234407663345337, "learning_rate": 8.26612243670428e-05, "loss": 0.0616, "step": 61950 }, { "epoch": 2.0947293688089523, "grad_norm": 0.7830119729042053, "learning_rate": 8.265496417905714e-05, "loss": 0.0604, "step": 61960 }, { "epoch": 2.0950674464992054, "grad_norm": 0.7452471256256104, "learning_rate": 8.264870309829702e-05, "loss": 0.0547, "step": 61970 }, { "epoch": 2.095405524189459, "grad_norm": 0.7190441489219666, "learning_rate": 8.264244112493362e-05, "loss": 0.0561, "step": 61980 }, { "epoch": 2.095743601879712, "grad_norm": 5.034279823303223, "learning_rate": 8.263617825913814e-05, "loss": 0.0607, "step": 61990 }, { "epoch": 2.096081679569965, "grad_norm": 1.0178263187408447, "learning_rate": 8.262991450108182e-05, "loss": 0.0571, "step": 62000 }, { "epoch": 2.0964197572602186, "grad_norm": 0.3005647361278534, "learning_rate": 8.262364985093589e-05, "loss": 0.0599, "step": 62010 }, { "epoch": 2.0967578349504716, "grad_norm": 0.5618177652359009, "learning_rate": 8.261738430887163e-05, "loss": 0.0549, "step": 62020 }, { "epoch": 2.0970959126407247, "grad_norm": 0.6279397010803223, "learning_rate": 8.261111787506032e-05, "loss": 0.0631, "step": 62030 }, { "epoch": 2.097433990330978, "grad_norm": 0.6185021996498108, "learning_rate": 8.260485054967333e-05, "loss": 0.0591, "step": 62040 }, { "epoch": 2.0977720680212313, "grad_norm": 0.5925354361534119, "learning_rate": 8.259858233288196e-05, "loss": 0.0594, "step": 62050 }, { "epoch": 2.0981101457114844, "grad_norm": 0.59496009349823, "learning_rate": 8.259231322485759e-05, "loss": 0.061, "step": 62060 }, { "epoch": 2.098448223401738, "grad_norm": 0.4733593463897705, "learning_rate": 8.258604322577163e-05, "loss": 0.0567, "step": 62070 }, { "epoch": 2.098786301091991, "grad_norm": 0.9149970412254333, "learning_rate": 8.257977233579548e-05, "loss": 0.0689, "step": 62080 }, { "epoch": 2.099124378782244, "grad_norm": 0.5210924744606018, "learning_rate": 8.257350055510061e-05, "loss": 0.0621, "step": 62090 }, { "epoch": 2.0994624564724975, "grad_norm": 0.514714241027832, "learning_rate": 8.256722788385847e-05, "loss": 0.0611, "step": 62100 }, { "epoch": 2.0998005341627506, "grad_norm": 0.6196102499961853, "learning_rate": 8.256095432224053e-05, "loss": 0.0579, "step": 62110 }, { "epoch": 2.1001386118530037, "grad_norm": 1.3713406324386597, "learning_rate": 8.255467987041835e-05, "loss": 0.0596, "step": 62120 }, { "epoch": 2.100476689543257, "grad_norm": 1.0803518295288086, "learning_rate": 8.254840452856345e-05, "loss": 0.0565, "step": 62130 }, { "epoch": 2.1008147672335102, "grad_norm": 0.5822355151176453, "learning_rate": 8.25421282968474e-05, "loss": 0.0532, "step": 62140 }, { "epoch": 2.1011528449237633, "grad_norm": 0.6468656659126282, "learning_rate": 8.25358511754418e-05, "loss": 0.0584, "step": 62150 }, { "epoch": 2.101490922614017, "grad_norm": 0.574378252029419, "learning_rate": 8.252957316451824e-05, "loss": 0.0646, "step": 62160 }, { "epoch": 2.10182900030427, "grad_norm": 0.5632561445236206, "learning_rate": 8.252329426424837e-05, "loss": 0.0593, "step": 62170 }, { "epoch": 2.102167077994523, "grad_norm": 0.8038981556892395, "learning_rate": 8.251701447480384e-05, "loss": 0.072, "step": 62180 }, { "epoch": 2.1025051556847765, "grad_norm": 0.7042493224143982, "learning_rate": 8.251073379635637e-05, "loss": 0.0687, "step": 62190 }, { "epoch": 2.1028432333750295, "grad_norm": 0.5316424369812012, "learning_rate": 8.250445222907764e-05, "loss": 0.0601, "step": 62200 }, { "epoch": 2.1031813110652826, "grad_norm": 0.5165058970451355, "learning_rate": 8.24981697731394e-05, "loss": 0.0614, "step": 62210 }, { "epoch": 2.103519388755536, "grad_norm": 0.5778307318687439, "learning_rate": 8.249188642871341e-05, "loss": 0.0598, "step": 62220 }, { "epoch": 2.103857466445789, "grad_norm": 1.240118384361267, "learning_rate": 8.248560219597145e-05, "loss": 0.0621, "step": 62230 }, { "epoch": 2.1041955441360423, "grad_norm": 0.47698333859443665, "learning_rate": 8.247931707508534e-05, "loss": 0.0608, "step": 62240 }, { "epoch": 2.1045336218262958, "grad_norm": 0.9310121536254883, "learning_rate": 8.24730310662269e-05, "loss": 0.0609, "step": 62250 }, { "epoch": 2.104871699516549, "grad_norm": 1.7761589288711548, "learning_rate": 8.246674416956797e-05, "loss": 0.0581, "step": 62260 }, { "epoch": 2.105209777206802, "grad_norm": 0.7761666774749756, "learning_rate": 8.246045638528047e-05, "loss": 0.0593, "step": 62270 }, { "epoch": 2.1055478548970554, "grad_norm": 0.6618568301200867, "learning_rate": 8.245416771353629e-05, "loss": 0.0629, "step": 62280 }, { "epoch": 2.1058859325873085, "grad_norm": 3.0952258110046387, "learning_rate": 8.244787815450735e-05, "loss": 0.0709, "step": 62290 }, { "epoch": 2.106224010277562, "grad_norm": 1.141602635383606, "learning_rate": 8.244158770836561e-05, "loss": 0.0664, "step": 62300 }, { "epoch": 2.106562087967815, "grad_norm": 0.7716695666313171, "learning_rate": 8.243529637528306e-05, "loss": 0.0659, "step": 62310 }, { "epoch": 2.106900165658068, "grad_norm": 0.9844832420349121, "learning_rate": 8.242900415543167e-05, "loss": 0.067, "step": 62320 }, { "epoch": 2.1072382433483217, "grad_norm": 0.8641563057899475, "learning_rate": 8.242271104898352e-05, "loss": 0.0586, "step": 62330 }, { "epoch": 2.1075763210385747, "grad_norm": 0.8217832446098328, "learning_rate": 8.241641705611061e-05, "loss": 0.0631, "step": 62340 }, { "epoch": 2.107914398728828, "grad_norm": 0.6031192541122437, "learning_rate": 8.241012217698503e-05, "loss": 0.0616, "step": 62350 }, { "epoch": 2.1082524764190813, "grad_norm": 0.759389340877533, "learning_rate": 8.240382641177891e-05, "loss": 0.0657, "step": 62360 }, { "epoch": 2.1085905541093344, "grad_norm": 0.7521725296974182, "learning_rate": 8.239752976066432e-05, "loss": 0.0599, "step": 62370 }, { "epoch": 2.1089286317995874, "grad_norm": 0.7664703726768494, "learning_rate": 8.239123222381347e-05, "loss": 0.0631, "step": 62380 }, { "epoch": 2.109266709489841, "grad_norm": 0.4580742418766022, "learning_rate": 8.238493380139847e-05, "loss": 0.0654, "step": 62390 }, { "epoch": 2.109604787180094, "grad_norm": 0.7767530083656311, "learning_rate": 8.237863449359157e-05, "loss": 0.0584, "step": 62400 }, { "epoch": 2.109942864870347, "grad_norm": 0.5949085354804993, "learning_rate": 8.237233430056495e-05, "loss": 0.0576, "step": 62410 }, { "epoch": 2.1102809425606006, "grad_norm": 0.9031147360801697, "learning_rate": 8.236603322249088e-05, "loss": 0.0596, "step": 62420 }, { "epoch": 2.1106190202508537, "grad_norm": 0.6139145493507385, "learning_rate": 8.235973125954162e-05, "loss": 0.0535, "step": 62430 }, { "epoch": 2.1109570979411068, "grad_norm": 0.4002416729927063, "learning_rate": 8.235342841188946e-05, "loss": 0.063, "step": 62440 }, { "epoch": 2.1112951756313603, "grad_norm": 0.3478794991970062, "learning_rate": 8.234712467970673e-05, "loss": 0.0684, "step": 62450 }, { "epoch": 2.1116332533216133, "grad_norm": 0.6598034501075745, "learning_rate": 8.234082006316576e-05, "loss": 0.058, "step": 62460 }, { "epoch": 2.1119713310118664, "grad_norm": 0.8600452542304993, "learning_rate": 8.233451456243891e-05, "loss": 0.0592, "step": 62470 }, { "epoch": 2.11230940870212, "grad_norm": 0.47886037826538086, "learning_rate": 8.232820817769859e-05, "loss": 0.0656, "step": 62480 }, { "epoch": 2.112647486392373, "grad_norm": 0.40169546008110046, "learning_rate": 8.23219009091172e-05, "loss": 0.0643, "step": 62490 }, { "epoch": 2.112985564082626, "grad_norm": 0.4400400221347809, "learning_rate": 8.231559275686719e-05, "loss": 0.0572, "step": 62500 }, { "epoch": 2.1133236417728796, "grad_norm": 0.9855876564979553, "learning_rate": 8.2309283721121e-05, "loss": 0.0581, "step": 62510 }, { "epoch": 2.1136617194631326, "grad_norm": 0.6217101216316223, "learning_rate": 8.230297380205115e-05, "loss": 0.0548, "step": 62520 }, { "epoch": 2.1139997971533857, "grad_norm": 1.1427031755447388, "learning_rate": 8.229666299983011e-05, "loss": 0.0614, "step": 62530 }, { "epoch": 2.114337874843639, "grad_norm": 1.3132351636886597, "learning_rate": 8.229035131463045e-05, "loss": 0.0558, "step": 62540 }, { "epoch": 2.1146759525338923, "grad_norm": 0.5061676502227783, "learning_rate": 8.228403874662471e-05, "loss": 0.0567, "step": 62550 }, { "epoch": 2.1150140302241454, "grad_norm": 0.6206585764884949, "learning_rate": 8.22777252959855e-05, "loss": 0.0599, "step": 62560 }, { "epoch": 2.115352107914399, "grad_norm": 0.5848928689956665, "learning_rate": 8.227141096288537e-05, "loss": 0.0568, "step": 62570 }, { "epoch": 2.115690185604652, "grad_norm": 0.4230162501335144, "learning_rate": 8.226509574749701e-05, "loss": 0.0591, "step": 62580 }, { "epoch": 2.116028263294905, "grad_norm": 0.4813917577266693, "learning_rate": 8.225877964999304e-05, "loss": 0.0564, "step": 62590 }, { "epoch": 2.1163663409851585, "grad_norm": 0.5348475575447083, "learning_rate": 8.225246267054619e-05, "loss": 0.0626, "step": 62600 }, { "epoch": 2.1167044186754116, "grad_norm": 0.792969286441803, "learning_rate": 8.224614480932909e-05, "loss": 0.0562, "step": 62610 }, { "epoch": 2.1170424963656647, "grad_norm": 0.49249833822250366, "learning_rate": 8.223982606651452e-05, "loss": 0.0561, "step": 62620 }, { "epoch": 2.117380574055918, "grad_norm": 0.5152584314346313, "learning_rate": 8.223350644227521e-05, "loss": 0.0642, "step": 62630 }, { "epoch": 2.1177186517461712, "grad_norm": 0.39783090353012085, "learning_rate": 8.222718593678395e-05, "loss": 0.0584, "step": 62640 }, { "epoch": 2.1180567294364243, "grad_norm": 0.8957692384719849, "learning_rate": 8.222086455021353e-05, "loss": 0.0568, "step": 62650 }, { "epoch": 2.118394807126678, "grad_norm": 0.5035117268562317, "learning_rate": 8.221454228273678e-05, "loss": 0.0624, "step": 62660 }, { "epoch": 2.118732884816931, "grad_norm": 0.3604266941547394, "learning_rate": 8.220821913452655e-05, "loss": 0.0638, "step": 62670 }, { "epoch": 2.119070962507184, "grad_norm": 0.43296390771865845, "learning_rate": 8.22018951057557e-05, "loss": 0.0548, "step": 62680 }, { "epoch": 2.1194090401974375, "grad_norm": 0.4169316589832306, "learning_rate": 8.219557019659714e-05, "loss": 0.0588, "step": 62690 }, { "epoch": 2.1197471178876905, "grad_norm": 0.4682776927947998, "learning_rate": 8.218924440722378e-05, "loss": 0.0626, "step": 62700 }, { "epoch": 2.1200851955779436, "grad_norm": 0.34559404850006104, "learning_rate": 8.218291773780858e-05, "loss": 0.0603, "step": 62710 }, { "epoch": 2.120423273268197, "grad_norm": 0.7149161100387573, "learning_rate": 8.217659018852451e-05, "loss": 0.0584, "step": 62720 }, { "epoch": 2.12076135095845, "grad_norm": 0.5346312522888184, "learning_rate": 8.217026175954454e-05, "loss": 0.0619, "step": 62730 }, { "epoch": 2.1210994286487033, "grad_norm": 0.5566872954368591, "learning_rate": 8.216393245104169e-05, "loss": 0.0589, "step": 62740 }, { "epoch": 2.121437506338957, "grad_norm": 0.597074031829834, "learning_rate": 8.215760226318902e-05, "loss": 0.0644, "step": 62750 }, { "epoch": 2.12177558402921, "grad_norm": 0.45385241508483887, "learning_rate": 8.215127119615958e-05, "loss": 0.0541, "step": 62760 }, { "epoch": 2.122113661719463, "grad_norm": 0.3982880711555481, "learning_rate": 8.214493925012648e-05, "loss": 0.0611, "step": 62770 }, { "epoch": 2.1224517394097164, "grad_norm": 0.5343785285949707, "learning_rate": 8.213860642526278e-05, "loss": 0.0634, "step": 62780 }, { "epoch": 2.1227898170999695, "grad_norm": 0.5002942085266113, "learning_rate": 8.21322727217417e-05, "loss": 0.0647, "step": 62790 }, { "epoch": 2.1231278947902226, "grad_norm": 0.5556747317314148, "learning_rate": 8.212593813973632e-05, "loss": 0.0628, "step": 62800 }, { "epoch": 2.123465972480476, "grad_norm": 0.4922475218772888, "learning_rate": 8.21196026794199e-05, "loss": 0.0596, "step": 62810 }, { "epoch": 2.123804050170729, "grad_norm": 0.6181870102882385, "learning_rate": 8.211326634096556e-05, "loss": 0.0578, "step": 62820 }, { "epoch": 2.1241421278609827, "grad_norm": 0.8011096715927124, "learning_rate": 8.21069291245466e-05, "loss": 0.0619, "step": 62830 }, { "epoch": 2.1244802055512357, "grad_norm": 1.162994623184204, "learning_rate": 8.210059103033629e-05, "loss": 0.0612, "step": 62840 }, { "epoch": 2.124818283241489, "grad_norm": 1.270669937133789, "learning_rate": 8.209425205850785e-05, "loss": 0.0636, "step": 62850 }, { "epoch": 2.1251563609317423, "grad_norm": 0.664488673210144, "learning_rate": 8.208791220923462e-05, "loss": 0.0573, "step": 62860 }, { "epoch": 2.1254944386219954, "grad_norm": 0.45457348227500916, "learning_rate": 8.208157148268994e-05, "loss": 0.0565, "step": 62870 }, { "epoch": 2.1258325163122485, "grad_norm": 0.8275479674339294, "learning_rate": 8.207522987904714e-05, "loss": 0.0588, "step": 62880 }, { "epoch": 2.126170594002502, "grad_norm": 0.48872336745262146, "learning_rate": 8.20688873984796e-05, "loss": 0.0596, "step": 62890 }, { "epoch": 2.126508671692755, "grad_norm": 0.6623736023902893, "learning_rate": 8.206254404116072e-05, "loss": 0.058, "step": 62900 }, { "epoch": 2.126846749383008, "grad_norm": 1.1785037517547607, "learning_rate": 8.205619980726394e-05, "loss": 0.0566, "step": 62910 }, { "epoch": 2.1271848270732616, "grad_norm": 0.4514809846878052, "learning_rate": 8.204985469696271e-05, "loss": 0.0603, "step": 62920 }, { "epoch": 2.1275229047635147, "grad_norm": 0.46438702940940857, "learning_rate": 8.204350871043047e-05, "loss": 0.0611, "step": 62930 }, { "epoch": 2.1278609824537678, "grad_norm": 0.37784117460250854, "learning_rate": 8.203716184784075e-05, "loss": 0.0601, "step": 62940 }, { "epoch": 2.1281990601440213, "grad_norm": 0.4548548758029938, "learning_rate": 8.203081410936708e-05, "loss": 0.0586, "step": 62950 }, { "epoch": 2.1285371378342743, "grad_norm": 0.6663303971290588, "learning_rate": 8.202446549518297e-05, "loss": 0.0628, "step": 62960 }, { "epoch": 2.1288752155245274, "grad_norm": 0.48446112871170044, "learning_rate": 8.201811600546199e-05, "loss": 0.0558, "step": 62970 }, { "epoch": 2.129213293214781, "grad_norm": 14.398918151855469, "learning_rate": 8.201176564037777e-05, "loss": 0.0617, "step": 62980 }, { "epoch": 2.129551370905034, "grad_norm": 0.3771393597126007, "learning_rate": 8.20054144001039e-05, "loss": 0.0548, "step": 62990 }, { "epoch": 2.129889448595287, "grad_norm": 0.5514516234397888, "learning_rate": 8.199906228481402e-05, "loss": 0.0642, "step": 63000 }, { "epoch": 2.1302275262855406, "grad_norm": 0.36684083938598633, "learning_rate": 8.19927092946818e-05, "loss": 0.0575, "step": 63010 }, { "epoch": 2.1305656039757936, "grad_norm": 1.4335010051727295, "learning_rate": 8.198635542988092e-05, "loss": 0.0652, "step": 63020 }, { "epoch": 2.1309036816660467, "grad_norm": 0.45941162109375, "learning_rate": 8.198000069058509e-05, "loss": 0.0615, "step": 63030 }, { "epoch": 2.1312417593563002, "grad_norm": 0.3378903865814209, "learning_rate": 8.197364507696808e-05, "loss": 0.0585, "step": 63040 }, { "epoch": 2.1315798370465533, "grad_norm": 0.46674326062202454, "learning_rate": 8.196728858920361e-05, "loss": 0.0652, "step": 63050 }, { "epoch": 2.1319179147368064, "grad_norm": 0.6492018103599548, "learning_rate": 8.196093122746548e-05, "loss": 0.0555, "step": 63060 }, { "epoch": 2.13225599242706, "grad_norm": 0.3225058615207672, "learning_rate": 8.195457299192751e-05, "loss": 0.0568, "step": 63070 }, { "epoch": 2.132594070117313, "grad_norm": 0.7197138667106628, "learning_rate": 8.19482138827635e-05, "loss": 0.0678, "step": 63080 }, { "epoch": 2.132932147807566, "grad_norm": 0.35086897015571594, "learning_rate": 8.194185390014735e-05, "loss": 0.0567, "step": 63090 }, { "epoch": 2.1332702254978195, "grad_norm": 0.7245364189147949, "learning_rate": 8.19354930442529e-05, "loss": 0.0647, "step": 63100 }, { "epoch": 2.1336083031880726, "grad_norm": 0.761498749256134, "learning_rate": 8.192913131525406e-05, "loss": 0.0573, "step": 63110 }, { "epoch": 2.1339463808783257, "grad_norm": 0.5486121773719788, "learning_rate": 8.192276871332479e-05, "loss": 0.0631, "step": 63120 }, { "epoch": 2.134284458568579, "grad_norm": 0.5356922149658203, "learning_rate": 8.1916405238639e-05, "loss": 0.0609, "step": 63130 }, { "epoch": 2.1346225362588322, "grad_norm": 0.7142186760902405, "learning_rate": 8.191004089137066e-05, "loss": 0.0641, "step": 63140 }, { "epoch": 2.1349606139490853, "grad_norm": 0.8433762788772583, "learning_rate": 8.190367567169381e-05, "loss": 0.0661, "step": 63150 }, { "epoch": 2.135298691639339, "grad_norm": 0.7452905178070068, "learning_rate": 8.189730957978247e-05, "loss": 0.0637, "step": 63160 }, { "epoch": 2.135636769329592, "grad_norm": 1.0608266592025757, "learning_rate": 8.189094261581065e-05, "loss": 0.0579, "step": 63170 }, { "epoch": 2.135974847019845, "grad_norm": 0.7248814105987549, "learning_rate": 8.188457477995244e-05, "loss": 0.0585, "step": 63180 }, { "epoch": 2.1363129247100985, "grad_norm": 0.8686764240264893, "learning_rate": 8.187820607238194e-05, "loss": 0.0606, "step": 63190 }, { "epoch": 2.1366510024003516, "grad_norm": 0.44997331500053406, "learning_rate": 8.187183649327325e-05, "loss": 0.0582, "step": 63200 }, { "epoch": 2.1369890800906046, "grad_norm": 1.0253291130065918, "learning_rate": 8.186546604280053e-05, "loss": 0.0599, "step": 63210 }, { "epoch": 2.137327157780858, "grad_norm": 0.4880082607269287, "learning_rate": 8.185909472113795e-05, "loss": 0.0615, "step": 63220 }, { "epoch": 2.137665235471111, "grad_norm": 0.748543381690979, "learning_rate": 8.185272252845969e-05, "loss": 0.0654, "step": 63230 }, { "epoch": 2.1380033131613647, "grad_norm": 0.6050307750701904, "learning_rate": 8.184634946493994e-05, "loss": 0.0566, "step": 63240 }, { "epoch": 2.138341390851618, "grad_norm": 0.6697700619697571, "learning_rate": 8.183997553075298e-05, "loss": 0.0588, "step": 63250 }, { "epoch": 2.138679468541871, "grad_norm": 0.5675150752067566, "learning_rate": 8.183360072607304e-05, "loss": 0.0603, "step": 63260 }, { "epoch": 2.1390175462321244, "grad_norm": 0.5894567966461182, "learning_rate": 8.182722505107442e-05, "loss": 0.0608, "step": 63270 }, { "epoch": 2.1393556239223774, "grad_norm": 0.290075421333313, "learning_rate": 8.182084850593142e-05, "loss": 0.0603, "step": 63280 }, { "epoch": 2.1396937016126305, "grad_norm": 0.46384450793266296, "learning_rate": 8.181447109081839e-05, "loss": 0.0606, "step": 63290 }, { "epoch": 2.140031779302884, "grad_norm": 0.4527440071105957, "learning_rate": 8.180809280590966e-05, "loss": 0.0592, "step": 63300 }, { "epoch": 2.140369856993137, "grad_norm": 0.7303500175476074, "learning_rate": 8.180171365137963e-05, "loss": 0.0581, "step": 63310 }, { "epoch": 2.14070793468339, "grad_norm": 0.2820517420768738, "learning_rate": 8.17953336274027e-05, "loss": 0.0559, "step": 63320 }, { "epoch": 2.1410460123736437, "grad_norm": 0.7909488081932068, "learning_rate": 8.178895273415329e-05, "loss": 0.054, "step": 63330 }, { "epoch": 2.1413840900638967, "grad_norm": 0.6261566877365112, "learning_rate": 8.178257097180583e-05, "loss": 0.0621, "step": 63340 }, { "epoch": 2.14172216775415, "grad_norm": 0.4258913993835449, "learning_rate": 8.177618834053484e-05, "loss": 0.0539, "step": 63350 }, { "epoch": 2.1420602454444033, "grad_norm": 0.460395872592926, "learning_rate": 8.17698048405148e-05, "loss": 0.0598, "step": 63360 }, { "epoch": 2.1423983231346564, "grad_norm": 0.6445275545120239, "learning_rate": 8.176342047192023e-05, "loss": 0.0604, "step": 63370 }, { "epoch": 2.1427364008249095, "grad_norm": 0.6675060391426086, "learning_rate": 8.175703523492568e-05, "loss": 0.0594, "step": 63380 }, { "epoch": 2.143074478515163, "grad_norm": 0.34799206256866455, "learning_rate": 8.175064912970571e-05, "loss": 0.058, "step": 63390 }, { "epoch": 2.143412556205416, "grad_norm": 0.29565513134002686, "learning_rate": 8.174426215643493e-05, "loss": 0.0588, "step": 63400 }, { "epoch": 2.143750633895669, "grad_norm": 0.5024826526641846, "learning_rate": 8.173787431528794e-05, "loss": 0.0602, "step": 63410 }, { "epoch": 2.1440887115859226, "grad_norm": 0.47935235500335693, "learning_rate": 8.17314856064394e-05, "loss": 0.0607, "step": 63420 }, { "epoch": 2.1444267892761757, "grad_norm": 2.067432403564453, "learning_rate": 8.172509603006395e-05, "loss": 0.0574, "step": 63430 }, { "epoch": 2.1447648669664288, "grad_norm": 0.4666318893432617, "learning_rate": 8.17187055863363e-05, "loss": 0.0564, "step": 63440 }, { "epoch": 2.1451029446566823, "grad_norm": 0.9644191265106201, "learning_rate": 8.171231427543115e-05, "loss": 0.0543, "step": 63450 }, { "epoch": 2.1454410223469353, "grad_norm": 0.5044942498207092, "learning_rate": 8.170592209752326e-05, "loss": 0.059, "step": 63460 }, { "epoch": 2.1457791000371884, "grad_norm": 0.37301018834114075, "learning_rate": 8.169952905278734e-05, "loss": 0.0545, "step": 63470 }, { "epoch": 2.146117177727442, "grad_norm": 0.46573829650878906, "learning_rate": 8.169313514139823e-05, "loss": 0.0579, "step": 63480 }, { "epoch": 2.146455255417695, "grad_norm": 0.19285446405410767, "learning_rate": 8.16867403635307e-05, "loss": 0.0503, "step": 63490 }, { "epoch": 2.146793333107948, "grad_norm": 0.42432647943496704, "learning_rate": 8.16803447193596e-05, "loss": 0.0551, "step": 63500 }, { "epoch": 2.1471314107982016, "grad_norm": 0.5835754871368408, "learning_rate": 8.167394820905977e-05, "loss": 0.0552, "step": 63510 }, { "epoch": 2.1474694884884546, "grad_norm": 0.7156962752342224, "learning_rate": 8.16675508328061e-05, "loss": 0.0574, "step": 63520 }, { "epoch": 2.1478075661787077, "grad_norm": 0.385882169008255, "learning_rate": 8.166115259077349e-05, "loss": 0.0574, "step": 63530 }, { "epoch": 2.1481456438689612, "grad_norm": 0.4240606129169464, "learning_rate": 8.165475348313687e-05, "loss": 0.0619, "step": 63540 }, { "epoch": 2.1484837215592143, "grad_norm": 0.4728996455669403, "learning_rate": 8.164835351007117e-05, "loss": 0.0572, "step": 63550 }, { "epoch": 2.1488217992494674, "grad_norm": 0.37362048029899597, "learning_rate": 8.164195267175138e-05, "loss": 0.0523, "step": 63560 }, { "epoch": 2.149159876939721, "grad_norm": 0.3808407187461853, "learning_rate": 8.16355509683525e-05, "loss": 0.0651, "step": 63570 }, { "epoch": 2.149497954629974, "grad_norm": 0.5570014715194702, "learning_rate": 8.162914840004952e-05, "loss": 0.0591, "step": 63580 }, { "epoch": 2.149836032320227, "grad_norm": 0.41858628392219543, "learning_rate": 8.162274496701749e-05, "loss": 0.0554, "step": 63590 }, { "epoch": 2.1501741100104805, "grad_norm": 0.5598660111427307, "learning_rate": 8.161634066943153e-05, "loss": 0.0584, "step": 63600 }, { "epoch": 2.1505121877007336, "grad_norm": 0.46387937664985657, "learning_rate": 8.160993550746668e-05, "loss": 0.058, "step": 63610 }, { "epoch": 2.1508502653909867, "grad_norm": 0.4868294894695282, "learning_rate": 8.160352948129808e-05, "loss": 0.0635, "step": 63620 }, { "epoch": 2.15118834308124, "grad_norm": 0.36428889632225037, "learning_rate": 8.159712259110084e-05, "loss": 0.0617, "step": 63630 }, { "epoch": 2.1515264207714933, "grad_norm": 1.283949375152588, "learning_rate": 8.159071483705015e-05, "loss": 0.0617, "step": 63640 }, { "epoch": 2.1518644984617463, "grad_norm": 0.4900839924812317, "learning_rate": 8.158430621932118e-05, "loss": 0.0566, "step": 63650 }, { "epoch": 2.152202576152, "grad_norm": 0.4319070875644684, "learning_rate": 8.157789673808913e-05, "loss": 0.0558, "step": 63660 }, { "epoch": 2.152540653842253, "grad_norm": 0.5988159775733948, "learning_rate": 8.157148639352926e-05, "loss": 0.0593, "step": 63670 }, { "epoch": 2.152878731532506, "grad_norm": 0.6943286657333374, "learning_rate": 8.156507518581682e-05, "loss": 0.0624, "step": 63680 }, { "epoch": 2.1532168092227595, "grad_norm": 0.3318958282470703, "learning_rate": 8.155866311512708e-05, "loss": 0.0615, "step": 63690 }, { "epoch": 2.1535548869130126, "grad_norm": 0.2918451726436615, "learning_rate": 8.155225018163533e-05, "loss": 0.0557, "step": 63700 }, { "epoch": 2.1538929646032656, "grad_norm": 0.7393725514411926, "learning_rate": 8.154583638551694e-05, "loss": 0.0571, "step": 63710 }, { "epoch": 2.154231042293519, "grad_norm": 1.0177415609359741, "learning_rate": 8.153942172694721e-05, "loss": 0.0601, "step": 63720 }, { "epoch": 2.154569119983772, "grad_norm": 0.3954300880432129, "learning_rate": 8.153300620610154e-05, "loss": 0.0601, "step": 63730 }, { "epoch": 2.1549071976740253, "grad_norm": 0.3577790856361389, "learning_rate": 8.152658982315536e-05, "loss": 0.0541, "step": 63740 }, { "epoch": 2.155245275364279, "grad_norm": 0.39384180307388306, "learning_rate": 8.152017257828401e-05, "loss": 0.0607, "step": 63750 }, { "epoch": 2.155583353054532, "grad_norm": 0.5548979640007019, "learning_rate": 8.151375447166302e-05, "loss": 0.0591, "step": 63760 }, { "epoch": 2.155921430744785, "grad_norm": 0.38105687499046326, "learning_rate": 8.150733550346781e-05, "loss": 0.0596, "step": 63770 }, { "epoch": 2.1562595084350384, "grad_norm": 0.4511810541152954, "learning_rate": 8.150091567387389e-05, "loss": 0.0556, "step": 63780 }, { "epoch": 2.1565975861252915, "grad_norm": 0.511038601398468, "learning_rate": 8.149449498305674e-05, "loss": 0.0551, "step": 63790 }, { "epoch": 2.1569356638155446, "grad_norm": 0.2671245336532593, "learning_rate": 8.148807343119196e-05, "loss": 0.0558, "step": 63800 }, { "epoch": 2.157273741505798, "grad_norm": 0.5090627074241638, "learning_rate": 8.148165101845509e-05, "loss": 0.0555, "step": 63810 }, { "epoch": 2.157611819196051, "grad_norm": 0.3362866938114166, "learning_rate": 8.147522774502168e-05, "loss": 0.0547, "step": 63820 }, { "epoch": 2.1579498968863047, "grad_norm": 0.3426082134246826, "learning_rate": 8.146880361106739e-05, "loss": 0.0582, "step": 63830 }, { "epoch": 2.1582879745765577, "grad_norm": 0.40164273977279663, "learning_rate": 8.146237861676782e-05, "loss": 0.0544, "step": 63840 }, { "epoch": 2.158626052266811, "grad_norm": 0.48482832312583923, "learning_rate": 8.145595276229862e-05, "loss": 0.0592, "step": 63850 }, { "epoch": 2.1589641299570643, "grad_norm": 0.5188518166542053, "learning_rate": 8.144952604783553e-05, "loss": 0.0634, "step": 63860 }, { "epoch": 2.1593022076473174, "grad_norm": 0.4633658826351166, "learning_rate": 8.144309847355418e-05, "loss": 0.0628, "step": 63870 }, { "epoch": 2.1596402853375705, "grad_norm": 0.46490880846977234, "learning_rate": 8.143667003963034e-05, "loss": 0.0665, "step": 63880 }, { "epoch": 2.159978363027824, "grad_norm": 0.3005833923816681, "learning_rate": 8.143024074623975e-05, "loss": 0.0602, "step": 63890 }, { "epoch": 2.160316440718077, "grad_norm": 0.4216119647026062, "learning_rate": 8.14238105935582e-05, "loss": 0.0557, "step": 63900 }, { "epoch": 2.16065451840833, "grad_norm": 0.38180407881736755, "learning_rate": 8.141737958176146e-05, "loss": 0.0533, "step": 63910 }, { "epoch": 2.1609925960985836, "grad_norm": 0.39269816875457764, "learning_rate": 8.141094771102537e-05, "loss": 0.0595, "step": 63920 }, { "epoch": 2.1613306737888367, "grad_norm": 0.509120762348175, "learning_rate": 8.140451498152578e-05, "loss": 0.0599, "step": 63930 }, { "epoch": 2.1616687514790898, "grad_norm": 1.24948251247406, "learning_rate": 8.139808139343853e-05, "loss": 0.059, "step": 63940 }, { "epoch": 2.1620068291693433, "grad_norm": 0.6821429133415222, "learning_rate": 8.139164694693954e-05, "loss": 0.056, "step": 63950 }, { "epoch": 2.1623449068595963, "grad_norm": 0.4687446355819702, "learning_rate": 8.138521164220472e-05, "loss": 0.0561, "step": 63960 }, { "epoch": 2.1626829845498494, "grad_norm": 0.6151543259620667, "learning_rate": 8.137877547941e-05, "loss": 0.058, "step": 63970 }, { "epoch": 2.163021062240103, "grad_norm": 0.334204763174057, "learning_rate": 8.137233845873133e-05, "loss": 0.0564, "step": 63980 }, { "epoch": 2.163359139930356, "grad_norm": 0.3517671525478363, "learning_rate": 8.136590058034472e-05, "loss": 0.0524, "step": 63990 }, { "epoch": 2.163697217620609, "grad_norm": 1.5803542137145996, "learning_rate": 8.135946184442619e-05, "loss": 0.0582, "step": 64000 }, { "epoch": 2.1640352953108626, "grad_norm": 0.43227291107177734, "learning_rate": 8.135302225115175e-05, "loss": 0.0625, "step": 64010 }, { "epoch": 2.1643733730011157, "grad_norm": 0.3591621220111847, "learning_rate": 8.134658180069745e-05, "loss": 0.057, "step": 64020 }, { "epoch": 2.1647114506913687, "grad_norm": 0.4127628803253174, "learning_rate": 8.134014049323939e-05, "loss": 0.0591, "step": 64030 }, { "epoch": 2.1650495283816222, "grad_norm": 0.365093469619751, "learning_rate": 8.133369832895364e-05, "loss": 0.0559, "step": 64040 }, { "epoch": 2.1653876060718753, "grad_norm": 0.4810204803943634, "learning_rate": 8.132725530801638e-05, "loss": 0.0576, "step": 64050 }, { "epoch": 2.1657256837621284, "grad_norm": 0.5345925688743591, "learning_rate": 8.13208114306037e-05, "loss": 0.0608, "step": 64060 }, { "epoch": 2.166063761452382, "grad_norm": 0.5661726593971252, "learning_rate": 8.131436669689183e-05, "loss": 0.06, "step": 64070 }, { "epoch": 2.166401839142635, "grad_norm": 0.8506887555122375, "learning_rate": 8.130792110705692e-05, "loss": 0.0585, "step": 64080 }, { "epoch": 2.166739916832888, "grad_norm": 0.5717331171035767, "learning_rate": 8.130147466127522e-05, "loss": 0.0632, "step": 64090 }, { "epoch": 2.1670779945231415, "grad_norm": 0.8932076692581177, "learning_rate": 8.129502735972295e-05, "loss": 0.0634, "step": 64100 }, { "epoch": 2.1674160722133946, "grad_norm": 0.3118232190608978, "learning_rate": 8.12885792025764e-05, "loss": 0.0568, "step": 64110 }, { "epoch": 2.1677541499036477, "grad_norm": 0.3355270326137543, "learning_rate": 8.128213019001185e-05, "loss": 0.0596, "step": 64120 }, { "epoch": 2.168092227593901, "grad_norm": 0.47428783774375916, "learning_rate": 8.12756803222056e-05, "loss": 0.0621, "step": 64130 }, { "epoch": 2.1684303052841543, "grad_norm": 0.4410889446735382, "learning_rate": 8.126922959933401e-05, "loss": 0.0587, "step": 64140 }, { "epoch": 2.1687683829744073, "grad_norm": 1.0262387990951538, "learning_rate": 8.126277802157343e-05, "loss": 0.0623, "step": 64150 }, { "epoch": 2.169106460664661, "grad_norm": 0.9153083562850952, "learning_rate": 8.125632558910023e-05, "loss": 0.0668, "step": 64160 }, { "epoch": 2.169444538354914, "grad_norm": 0.637065052986145, "learning_rate": 8.124987230209085e-05, "loss": 0.0617, "step": 64170 }, { "epoch": 2.169782616045167, "grad_norm": 0.39642322063446045, "learning_rate": 8.12434181607217e-05, "loss": 0.0576, "step": 64180 }, { "epoch": 2.1701206937354205, "grad_norm": 0.6847009658813477, "learning_rate": 8.123696316516922e-05, "loss": 0.054, "step": 64190 }, { "epoch": 2.1704587714256736, "grad_norm": 0.4464573562145233, "learning_rate": 8.123050731560991e-05, "loss": 0.0595, "step": 64200 }, { "epoch": 2.170796849115927, "grad_norm": 1.8722981214523315, "learning_rate": 8.122405061222027e-05, "loss": 0.0562, "step": 64210 }, { "epoch": 2.17113492680618, "grad_norm": 0.5971588492393494, "learning_rate": 8.12175930551768e-05, "loss": 0.0529, "step": 64220 }, { "epoch": 2.171473004496433, "grad_norm": 13.858656883239746, "learning_rate": 8.121113464465608e-05, "loss": 0.0661, "step": 64230 }, { "epoch": 2.1718110821866867, "grad_norm": 1.254281759262085, "learning_rate": 8.120467538083465e-05, "loss": 0.0597, "step": 64240 }, { "epoch": 2.17214915987694, "grad_norm": 1.3013395071029663, "learning_rate": 8.119821526388913e-05, "loss": 0.0681, "step": 64250 }, { "epoch": 2.172487237567193, "grad_norm": 1.128735065460205, "learning_rate": 8.119175429399612e-05, "loss": 0.0626, "step": 64260 }, { "epoch": 2.1728253152574464, "grad_norm": 0.8089272379875183, "learning_rate": 8.118529247133227e-05, "loss": 0.0639, "step": 64270 }, { "epoch": 2.1731633929476994, "grad_norm": 5.365659236907959, "learning_rate": 8.117882979607425e-05, "loss": 0.0612, "step": 64280 }, { "epoch": 2.1735014706379525, "grad_norm": 0.7313252091407776, "learning_rate": 8.117236626839871e-05, "loss": 0.066, "step": 64290 }, { "epoch": 2.173839548328206, "grad_norm": 0.8294647336006165, "learning_rate": 8.116590188848241e-05, "loss": 0.0618, "step": 64300 }, { "epoch": 2.174177626018459, "grad_norm": 0.9020423293113708, "learning_rate": 8.115943665650205e-05, "loss": 0.0568, "step": 64310 }, { "epoch": 2.174515703708712, "grad_norm": 0.6046127676963806, "learning_rate": 8.11529705726344e-05, "loss": 0.0589, "step": 64320 }, { "epoch": 2.1748537813989657, "grad_norm": 0.5744115114212036, "learning_rate": 8.114650363705624e-05, "loss": 0.0663, "step": 64330 }, { "epoch": 2.1751918590892187, "grad_norm": 0.7166561484336853, "learning_rate": 8.114003584994436e-05, "loss": 0.0569, "step": 64340 }, { "epoch": 2.175529936779472, "grad_norm": 1.2905304431915283, "learning_rate": 8.113356721147561e-05, "loss": 0.0627, "step": 64350 }, { "epoch": 2.1758680144697253, "grad_norm": 0.4301573634147644, "learning_rate": 8.112709772182683e-05, "loss": 0.0712, "step": 64360 }, { "epoch": 2.1762060921599784, "grad_norm": 0.5348978042602539, "learning_rate": 8.11206273811749e-05, "loss": 0.0637, "step": 64370 }, { "epoch": 2.1765441698502315, "grad_norm": 1.7929153442382812, "learning_rate": 8.111415618969668e-05, "loss": 0.0636, "step": 64380 }, { "epoch": 2.176882247540485, "grad_norm": 0.7393127083778381, "learning_rate": 8.110768414756914e-05, "loss": 0.0599, "step": 64390 }, { "epoch": 2.177220325230738, "grad_norm": 0.7031688094139099, "learning_rate": 8.11012112549692e-05, "loss": 0.0621, "step": 64400 }, { "epoch": 2.177558402920991, "grad_norm": 0.482592910528183, "learning_rate": 8.109473751207384e-05, "loss": 0.0614, "step": 64410 }, { "epoch": 2.1778964806112446, "grad_norm": 0.3942035734653473, "learning_rate": 8.108826291906001e-05, "loss": 0.056, "step": 64420 }, { "epoch": 2.1782345583014977, "grad_norm": 0.520983874797821, "learning_rate": 8.108178747610477e-05, "loss": 0.0601, "step": 64430 }, { "epoch": 2.1785726359917508, "grad_norm": 1.0259965658187866, "learning_rate": 8.107531118338514e-05, "loss": 0.0603, "step": 64440 }, { "epoch": 2.1789107136820043, "grad_norm": 0.631351113319397, "learning_rate": 8.106883404107816e-05, "loss": 0.0623, "step": 64450 }, { "epoch": 2.1792487913722574, "grad_norm": 0.3310626447200775, "learning_rate": 8.106235604936094e-05, "loss": 0.0521, "step": 64460 }, { "epoch": 2.1795868690625104, "grad_norm": 0.5885518193244934, "learning_rate": 8.10558772084106e-05, "loss": 0.059, "step": 64470 }, { "epoch": 2.179924946752764, "grad_norm": 0.5463055372238159, "learning_rate": 8.10493975184042e-05, "loss": 0.056, "step": 64480 }, { "epoch": 2.180263024443017, "grad_norm": 1.6915135383605957, "learning_rate": 8.104291697951896e-05, "loss": 0.0651, "step": 64490 }, { "epoch": 2.18060110213327, "grad_norm": 0.4354300796985626, "learning_rate": 8.103643559193202e-05, "loss": 0.0637, "step": 64500 }, { "epoch": 2.1809391798235236, "grad_norm": 0.6079111099243164, "learning_rate": 8.10299533558206e-05, "loss": 0.0604, "step": 64510 }, { "epoch": 2.1812772575137767, "grad_norm": 0.4622866213321686, "learning_rate": 8.102347027136192e-05, "loss": 0.0597, "step": 64520 }, { "epoch": 2.1816153352040297, "grad_norm": 0.5835849642753601, "learning_rate": 8.101698633873321e-05, "loss": 0.0658, "step": 64530 }, { "epoch": 2.1819534128942832, "grad_norm": 0.6176680326461792, "learning_rate": 8.101050155811175e-05, "loss": 0.0614, "step": 64540 }, { "epoch": 2.1822914905845363, "grad_norm": 0.6097161769866943, "learning_rate": 8.100401592967481e-05, "loss": 0.0655, "step": 64550 }, { "epoch": 2.1826295682747894, "grad_norm": 0.9786903858184814, "learning_rate": 8.099752945359973e-05, "loss": 0.0544, "step": 64560 }, { "epoch": 2.182967645965043, "grad_norm": 0.747402012348175, "learning_rate": 8.099104213006385e-05, "loss": 0.06, "step": 64570 }, { "epoch": 2.183305723655296, "grad_norm": 0.5784555077552795, "learning_rate": 8.098455395924452e-05, "loss": 0.0664, "step": 64580 }, { "epoch": 2.183643801345549, "grad_norm": 0.4515858590602875, "learning_rate": 8.097806494131912e-05, "loss": 0.0612, "step": 64590 }, { "epoch": 2.1839818790358025, "grad_norm": 0.5879343748092651, "learning_rate": 8.097157507646507e-05, "loss": 0.0637, "step": 64600 }, { "epoch": 2.1843199567260556, "grad_norm": 0.7554988861083984, "learning_rate": 8.096508436485978e-05, "loss": 0.0612, "step": 64610 }, { "epoch": 2.1846580344163087, "grad_norm": 0.5162935256958008, "learning_rate": 8.095859280668072e-05, "loss": 0.062, "step": 64620 }, { "epoch": 2.184996112106562, "grad_norm": 0.6202875971794128, "learning_rate": 8.095210040210536e-05, "loss": 0.0648, "step": 64630 }, { "epoch": 2.1853341897968153, "grad_norm": 1.2519841194152832, "learning_rate": 8.09456071513112e-05, "loss": 0.0628, "step": 64640 }, { "epoch": 2.1856722674870683, "grad_norm": 0.4343077838420868, "learning_rate": 8.09391130544758e-05, "loss": 0.0618, "step": 64650 }, { "epoch": 2.186010345177322, "grad_norm": 0.6423685550689697, "learning_rate": 8.093261811177664e-05, "loss": 0.0627, "step": 64660 }, { "epoch": 2.186348422867575, "grad_norm": 0.5820552706718445, "learning_rate": 8.092612232339134e-05, "loss": 0.0665, "step": 64670 }, { "epoch": 2.186686500557828, "grad_norm": 0.5037172436714172, "learning_rate": 8.091962568949745e-05, "loss": 0.065, "step": 64680 }, { "epoch": 2.1870245782480815, "grad_norm": 0.7255784869194031, "learning_rate": 8.091312821027262e-05, "loss": 0.061, "step": 64690 }, { "epoch": 2.1873626559383346, "grad_norm": 0.6558296084403992, "learning_rate": 8.090662988589447e-05, "loss": 0.0635, "step": 64700 }, { "epoch": 2.1877007336285876, "grad_norm": 0.5234860181808472, "learning_rate": 8.09001307165407e-05, "loss": 0.0624, "step": 64710 }, { "epoch": 2.188038811318841, "grad_norm": 0.5680426955223083, "learning_rate": 8.089363070238894e-05, "loss": 0.0554, "step": 64720 }, { "epoch": 2.188376889009094, "grad_norm": 0.8699982166290283, "learning_rate": 8.088712984361694e-05, "loss": 0.0564, "step": 64730 }, { "epoch": 2.1887149666993473, "grad_norm": 0.44625839591026306, "learning_rate": 8.088062814040239e-05, "loss": 0.0645, "step": 64740 }, { "epoch": 2.189053044389601, "grad_norm": 0.5142084360122681, "learning_rate": 8.087412559292309e-05, "loss": 0.0617, "step": 64750 }, { "epoch": 2.189391122079854, "grad_norm": 0.6279720664024353, "learning_rate": 8.086762220135679e-05, "loss": 0.0633, "step": 64760 }, { "epoch": 2.189729199770107, "grad_norm": 0.8243870735168457, "learning_rate": 8.086111796588128e-05, "loss": 0.0591, "step": 64770 }, { "epoch": 2.1900672774603605, "grad_norm": 0.9192203283309937, "learning_rate": 8.085461288667442e-05, "loss": 0.0581, "step": 64780 }, { "epoch": 2.1904053551506135, "grad_norm": 0.5582347512245178, "learning_rate": 8.084810696391403e-05, "loss": 0.0592, "step": 64790 }, { "epoch": 2.190743432840867, "grad_norm": 0.7669507265090942, "learning_rate": 8.084160019777797e-05, "loss": 0.0636, "step": 64800 }, { "epoch": 2.19108151053112, "grad_norm": 1.3751407861709595, "learning_rate": 8.083509258844415e-05, "loss": 0.05, "step": 64810 }, { "epoch": 2.191419588221373, "grad_norm": 0.5579459071159363, "learning_rate": 8.08285841360905e-05, "loss": 0.0549, "step": 64820 }, { "epoch": 2.1917576659116267, "grad_norm": 0.8395970463752747, "learning_rate": 8.082207484089492e-05, "loss": 0.0597, "step": 64830 }, { "epoch": 2.1920957436018798, "grad_norm": 2.424353837966919, "learning_rate": 8.081556470303542e-05, "loss": 0.0626, "step": 64840 }, { "epoch": 2.192433821292133, "grad_norm": 0.6957279443740845, "learning_rate": 8.080905372268994e-05, "loss": 0.0565, "step": 64850 }, { "epoch": 2.1927718989823863, "grad_norm": 0.585409939289093, "learning_rate": 8.080254190003651e-05, "loss": 0.0654, "step": 64860 }, { "epoch": 2.1931099766726394, "grad_norm": 0.7580908536911011, "learning_rate": 8.079602923525315e-05, "loss": 0.0612, "step": 64870 }, { "epoch": 2.1934480543628925, "grad_norm": 0.47531458735466003, "learning_rate": 8.078951572851792e-05, "loss": 0.0557, "step": 64880 }, { "epoch": 2.193786132053146, "grad_norm": 0.49220672249794006, "learning_rate": 8.07830013800089e-05, "loss": 0.0555, "step": 64890 }, { "epoch": 2.194124209743399, "grad_norm": 0.9167153835296631, "learning_rate": 8.07764861899042e-05, "loss": 0.06, "step": 64900 }, { "epoch": 2.194462287433652, "grad_norm": 0.4390527307987213, "learning_rate": 8.076997015838192e-05, "loss": 0.0595, "step": 64910 }, { "epoch": 2.1948003651239056, "grad_norm": 0.629513680934906, "learning_rate": 8.07634532856202e-05, "loss": 0.0668, "step": 64920 }, { "epoch": 2.1951384428141587, "grad_norm": 0.7638389468193054, "learning_rate": 8.075693557179724e-05, "loss": 0.0592, "step": 64930 }, { "epoch": 2.1954765205044118, "grad_norm": 0.7544531226158142, "learning_rate": 8.075041701709123e-05, "loss": 0.0629, "step": 64940 }, { "epoch": 2.1958145981946653, "grad_norm": 0.5821909308433533, "learning_rate": 8.074389762168035e-05, "loss": 0.0539, "step": 64950 }, { "epoch": 2.1961526758849184, "grad_norm": 0.8610582947731018, "learning_rate": 8.073737738574288e-05, "loss": 0.059, "step": 64960 }, { "epoch": 2.1964907535751714, "grad_norm": 0.6325303316116333, "learning_rate": 8.073085630945703e-05, "loss": 0.0565, "step": 64970 }, { "epoch": 2.196828831265425, "grad_norm": 0.6663770079612732, "learning_rate": 8.072433439300113e-05, "loss": 0.0627, "step": 64980 }, { "epoch": 2.197166908955678, "grad_norm": 0.5140206217765808, "learning_rate": 8.071781163655346e-05, "loss": 0.0584, "step": 64990 }, { "epoch": 2.197504986645931, "grad_norm": 0.6283681988716125, "learning_rate": 8.071128804029238e-05, "loss": 0.0596, "step": 65000 }, { "epoch": 2.1978430643361846, "grad_norm": 0.5343438386917114, "learning_rate": 8.07047636043962e-05, "loss": 0.059, "step": 65010 }, { "epoch": 2.1981811420264377, "grad_norm": 0.8059038519859314, "learning_rate": 8.069823832904334e-05, "loss": 0.0644, "step": 65020 }, { "epoch": 2.1985192197166907, "grad_norm": 0.5422974824905396, "learning_rate": 8.069171221441217e-05, "loss": 0.056, "step": 65030 }, { "epoch": 2.1988572974069442, "grad_norm": 0.3215480446815491, "learning_rate": 8.068518526068112e-05, "loss": 0.0643, "step": 65040 }, { "epoch": 2.1991953750971973, "grad_norm": 0.605094850063324, "learning_rate": 8.067865746802862e-05, "loss": 0.0571, "step": 65050 }, { "epoch": 2.1995334527874504, "grad_norm": 0.3312607407569885, "learning_rate": 8.067212883663315e-05, "loss": 0.0534, "step": 65060 }, { "epoch": 2.199871530477704, "grad_norm": 0.7175371646881104, "learning_rate": 8.066559936667321e-05, "loss": 0.0645, "step": 65070 }, { "epoch": 2.200209608167957, "grad_norm": 0.41568490862846375, "learning_rate": 8.065906905832731e-05, "loss": 0.0554, "step": 65080 }, { "epoch": 2.20054768585821, "grad_norm": 0.512334406375885, "learning_rate": 8.065253791177396e-05, "loss": 0.0632, "step": 65090 }, { "epoch": 2.2008857635484635, "grad_norm": 0.43998441100120544, "learning_rate": 8.064600592719176e-05, "loss": 0.0551, "step": 65100 }, { "epoch": 2.2012238412387166, "grad_norm": 0.7629746198654175, "learning_rate": 8.063947310475927e-05, "loss": 0.056, "step": 65110 }, { "epoch": 2.2015619189289697, "grad_norm": 0.5511254668235779, "learning_rate": 8.063293944465508e-05, "loss": 0.0598, "step": 65120 }, { "epoch": 2.201899996619223, "grad_norm": 0.7663003206253052, "learning_rate": 8.062640494705784e-05, "loss": 0.0593, "step": 65130 }, { "epoch": 2.2022380743094763, "grad_norm": 0.5774022936820984, "learning_rate": 8.06198696121462e-05, "loss": 0.0611, "step": 65140 }, { "epoch": 2.2025761519997293, "grad_norm": 2.292177200317383, "learning_rate": 8.061333344009882e-05, "loss": 0.0595, "step": 65150 }, { "epoch": 2.202914229689983, "grad_norm": 0.6591470241546631, "learning_rate": 8.060679643109442e-05, "loss": 0.0591, "step": 65160 }, { "epoch": 2.203252307380236, "grad_norm": 2.191585063934326, "learning_rate": 8.06002585853117e-05, "loss": 0.0601, "step": 65170 }, { "epoch": 2.2035903850704894, "grad_norm": 0.4746268093585968, "learning_rate": 8.059371990292939e-05, "loss": 0.0536, "step": 65180 }, { "epoch": 2.2039284627607425, "grad_norm": 0.9536274671554565, "learning_rate": 8.058718038412629e-05, "loss": 0.0556, "step": 65190 }, { "epoch": 2.2042665404509956, "grad_norm": 0.4812895357608795, "learning_rate": 8.058064002908116e-05, "loss": 0.0561, "step": 65200 }, { "epoch": 2.204604618141249, "grad_norm": 0.7460231184959412, "learning_rate": 8.057409883797282e-05, "loss": 0.0634, "step": 65210 }, { "epoch": 2.204942695831502, "grad_norm": 0.47286176681518555, "learning_rate": 8.05675568109801e-05, "loss": 0.0523, "step": 65220 }, { "epoch": 2.2052807735217552, "grad_norm": 0.9653750658035278, "learning_rate": 8.056101394828188e-05, "loss": 0.0635, "step": 65230 }, { "epoch": 2.2056188512120087, "grad_norm": 0.8595234751701355, "learning_rate": 8.0554470250057e-05, "loss": 0.0482, "step": 65240 }, { "epoch": 2.205956928902262, "grad_norm": 0.4884827733039856, "learning_rate": 8.054792571648438e-05, "loss": 0.0639, "step": 65250 }, { "epoch": 2.206295006592515, "grad_norm": 0.6535950303077698, "learning_rate": 8.054138034774297e-05, "loss": 0.0595, "step": 65260 }, { "epoch": 2.2066330842827684, "grad_norm": 0.6294102072715759, "learning_rate": 8.053483414401167e-05, "loss": 0.05, "step": 65270 }, { "epoch": 2.2069711619730215, "grad_norm": 0.6083894371986389, "learning_rate": 8.05282871054695e-05, "loss": 0.0612, "step": 65280 }, { "epoch": 2.2073092396632745, "grad_norm": 0.9878699779510498, "learning_rate": 8.05217392322954e-05, "loss": 0.062, "step": 65290 }, { "epoch": 2.207647317353528, "grad_norm": 0.4584379494190216, "learning_rate": 8.051519052466843e-05, "loss": 0.0611, "step": 65300 }, { "epoch": 2.207985395043781, "grad_norm": 0.6978819966316223, "learning_rate": 8.050864098276762e-05, "loss": 0.0599, "step": 65310 }, { "epoch": 2.208323472734034, "grad_norm": 1.3055419921875, "learning_rate": 8.050209060677203e-05, "loss": 0.0646, "step": 65320 }, { "epoch": 2.2086615504242877, "grad_norm": 0.3595762252807617, "learning_rate": 8.049553939686074e-05, "loss": 0.0592, "step": 65330 }, { "epoch": 2.2089996281145408, "grad_norm": 0.40771788358688354, "learning_rate": 8.048898735321284e-05, "loss": 0.0586, "step": 65340 }, { "epoch": 2.209337705804794, "grad_norm": 0.5315342545509338, "learning_rate": 8.048243447600749e-05, "loss": 0.0554, "step": 65350 }, { "epoch": 2.2096757834950473, "grad_norm": 0.811765730381012, "learning_rate": 8.047588076542386e-05, "loss": 0.0581, "step": 65360 }, { "epoch": 2.2100138611853004, "grad_norm": 0.49085354804992676, "learning_rate": 8.046932622164107e-05, "loss": 0.0556, "step": 65370 }, { "epoch": 2.2103519388755535, "grad_norm": 0.5102158188819885, "learning_rate": 8.046277084483835e-05, "loss": 0.0577, "step": 65380 }, { "epoch": 2.210690016565807, "grad_norm": 0.39882567524909973, "learning_rate": 8.045621463519493e-05, "loss": 0.0573, "step": 65390 }, { "epoch": 2.21102809425606, "grad_norm": 0.9457538723945618, "learning_rate": 8.044965759289003e-05, "loss": 0.0601, "step": 65400 }, { "epoch": 2.211366171946313, "grad_norm": 0.5349503755569458, "learning_rate": 8.044309971810294e-05, "loss": 0.0619, "step": 65410 }, { "epoch": 2.2117042496365666, "grad_norm": 0.48738691210746765, "learning_rate": 8.043654101101294e-05, "loss": 0.0635, "step": 65420 }, { "epoch": 2.2120423273268197, "grad_norm": 0.558101236820221, "learning_rate": 8.042998147179934e-05, "loss": 0.059, "step": 65430 }, { "epoch": 2.212380405017073, "grad_norm": 1.343916416168213, "learning_rate": 8.042342110064148e-05, "loss": 0.0566, "step": 65440 }, { "epoch": 2.2127184827073263, "grad_norm": 1.384135127067566, "learning_rate": 8.041685989771873e-05, "loss": 0.0679, "step": 65450 }, { "epoch": 2.2130565603975794, "grad_norm": 0.560066282749176, "learning_rate": 8.041029786321044e-05, "loss": 0.065, "step": 65460 }, { "epoch": 2.2133946380878324, "grad_norm": 0.6734886169433594, "learning_rate": 8.040373499729604e-05, "loss": 0.0612, "step": 65470 }, { "epoch": 2.213732715778086, "grad_norm": 2.2940237522125244, "learning_rate": 8.039717130015494e-05, "loss": 0.0687, "step": 65480 }, { "epoch": 2.214070793468339, "grad_norm": 0.4495968520641327, "learning_rate": 8.039060677196661e-05, "loss": 0.0639, "step": 65490 }, { "epoch": 2.214408871158592, "grad_norm": 0.589106023311615, "learning_rate": 8.038404141291049e-05, "loss": 0.0591, "step": 65500 }, { "epoch": 2.2147469488488456, "grad_norm": 2.150047540664673, "learning_rate": 8.03774752231661e-05, "loss": 0.0637, "step": 65510 }, { "epoch": 2.2150850265390987, "grad_norm": 0.7425894141197205, "learning_rate": 8.037090820291296e-05, "loss": 0.0663, "step": 65520 }, { "epoch": 2.2154231042293517, "grad_norm": 0.6094037294387817, "learning_rate": 8.036434035233058e-05, "loss": 0.0619, "step": 65530 }, { "epoch": 2.2157611819196052, "grad_norm": 0.6473750472068787, "learning_rate": 8.035777167159856e-05, "loss": 0.0657, "step": 65540 }, { "epoch": 2.2160992596098583, "grad_norm": 0.4404483437538147, "learning_rate": 8.035120216089647e-05, "loss": 0.0543, "step": 65550 }, { "epoch": 2.2164373373001114, "grad_norm": 0.7593993544578552, "learning_rate": 8.034463182040389e-05, "loss": 0.0585, "step": 65560 }, { "epoch": 2.216775414990365, "grad_norm": 0.4593430459499359, "learning_rate": 8.03380606503005e-05, "loss": 0.0604, "step": 65570 }, { "epoch": 2.217113492680618, "grad_norm": 0.7454952001571655, "learning_rate": 8.033148865076592e-05, "loss": 0.0561, "step": 65580 }, { "epoch": 2.217451570370871, "grad_norm": 0.7401496171951294, "learning_rate": 8.032491582197983e-05, "loss": 0.059, "step": 65590 }, { "epoch": 2.2177896480611246, "grad_norm": 0.6411807537078857, "learning_rate": 8.031834216412195e-05, "loss": 0.062, "step": 65600 }, { "epoch": 2.2181277257513776, "grad_norm": 0.8644804358482361, "learning_rate": 8.031176767737196e-05, "loss": 0.0598, "step": 65610 }, { "epoch": 2.2184658034416307, "grad_norm": 0.4527592957019806, "learning_rate": 8.030519236190966e-05, "loss": 0.0607, "step": 65620 }, { "epoch": 2.218803881131884, "grad_norm": 0.4266522228717804, "learning_rate": 8.029861621791476e-05, "loss": 0.0603, "step": 65630 }, { "epoch": 2.2191419588221373, "grad_norm": 0.3406639099121094, "learning_rate": 8.029203924556708e-05, "loss": 0.0598, "step": 65640 }, { "epoch": 2.2194800365123903, "grad_norm": 1.0790168046951294, "learning_rate": 8.028546144504643e-05, "loss": 0.0558, "step": 65650 }, { "epoch": 2.219818114202644, "grad_norm": 1.2417786121368408, "learning_rate": 8.027888281653265e-05, "loss": 0.0577, "step": 65660 }, { "epoch": 2.220156191892897, "grad_norm": 0.46601563692092896, "learning_rate": 8.027230336020559e-05, "loss": 0.0552, "step": 65670 }, { "epoch": 2.22049426958315, "grad_norm": 0.37138795852661133, "learning_rate": 8.026572307624512e-05, "loss": 0.0573, "step": 65680 }, { "epoch": 2.2208323472734035, "grad_norm": 0.5320441722869873, "learning_rate": 8.025914196483116e-05, "loss": 0.0663, "step": 65690 }, { "epoch": 2.2211704249636566, "grad_norm": 0.5926468372344971, "learning_rate": 8.025256002614361e-05, "loss": 0.0629, "step": 65700 }, { "epoch": 2.2215085026539096, "grad_norm": 0.5988909602165222, "learning_rate": 8.024597726036245e-05, "loss": 0.0614, "step": 65710 }, { "epoch": 2.221846580344163, "grad_norm": 0.7175000309944153, "learning_rate": 8.023939366766763e-05, "loss": 0.0622, "step": 65720 }, { "epoch": 2.2221846580344162, "grad_norm": 0.7763135433197021, "learning_rate": 8.023280924823915e-05, "loss": 0.0658, "step": 65730 }, { "epoch": 2.2225227357246693, "grad_norm": 0.7753355503082275, "learning_rate": 8.022622400225701e-05, "loss": 0.0634, "step": 65740 }, { "epoch": 2.222860813414923, "grad_norm": 0.44986405968666077, "learning_rate": 8.021963792990129e-05, "loss": 0.0588, "step": 65750 }, { "epoch": 2.223198891105176, "grad_norm": 0.5321260094642639, "learning_rate": 8.0213051031352e-05, "loss": 0.0599, "step": 65760 }, { "epoch": 2.2235369687954294, "grad_norm": 0.5952684879302979, "learning_rate": 8.020646330678924e-05, "loss": 0.0531, "step": 65770 }, { "epoch": 2.2238750464856825, "grad_norm": 0.8829144239425659, "learning_rate": 8.019987475639312e-05, "loss": 0.0631, "step": 65780 }, { "epoch": 2.2242131241759355, "grad_norm": 2.0442113876342773, "learning_rate": 8.019328538034378e-05, "loss": 0.0599, "step": 65790 }, { "epoch": 2.224551201866189, "grad_norm": 0.42009237408638, "learning_rate": 8.018669517882136e-05, "loss": 0.056, "step": 65800 }, { "epoch": 2.224889279556442, "grad_norm": 0.826644241809845, "learning_rate": 8.018010415200604e-05, "loss": 0.0563, "step": 65810 }, { "epoch": 2.225227357246695, "grad_norm": 0.899606466293335, "learning_rate": 8.017351230007802e-05, "loss": 0.0627, "step": 65820 }, { "epoch": 2.2255654349369487, "grad_norm": 0.8915049433708191, "learning_rate": 8.016691962321748e-05, "loss": 0.0648, "step": 65830 }, { "epoch": 2.2259035126272018, "grad_norm": 0.8274115324020386, "learning_rate": 8.016032612160472e-05, "loss": 0.0642, "step": 65840 }, { "epoch": 2.226241590317455, "grad_norm": 0.549790620803833, "learning_rate": 8.015373179541995e-05, "loss": 0.0605, "step": 65850 }, { "epoch": 2.2265796680077083, "grad_norm": 0.7557091116905212, "learning_rate": 8.014713664484349e-05, "loss": 0.0573, "step": 65860 }, { "epoch": 2.2269177456979614, "grad_norm": 0.9283783435821533, "learning_rate": 8.014054067005564e-05, "loss": 0.0549, "step": 65870 }, { "epoch": 2.2272558233882145, "grad_norm": 1.1044508218765259, "learning_rate": 8.013394387123674e-05, "loss": 0.0624, "step": 65880 }, { "epoch": 2.227593901078468, "grad_norm": 0.4908212125301361, "learning_rate": 8.012734624856712e-05, "loss": 0.0604, "step": 65890 }, { "epoch": 2.227931978768721, "grad_norm": 0.5346165299415588, "learning_rate": 8.012074780222718e-05, "loss": 0.0576, "step": 65900 }, { "epoch": 2.228270056458974, "grad_norm": 0.35300570726394653, "learning_rate": 8.011414853239732e-05, "loss": 0.0535, "step": 65910 }, { "epoch": 2.2286081341492276, "grad_norm": 0.8687924742698669, "learning_rate": 8.010754843925794e-05, "loss": 0.0563, "step": 65920 }, { "epoch": 2.2289462118394807, "grad_norm": 0.5345227718353271, "learning_rate": 8.01009475229895e-05, "loss": 0.0604, "step": 65930 }, { "epoch": 2.229284289529734, "grad_norm": 0.47522345185279846, "learning_rate": 8.009434578377246e-05, "loss": 0.0563, "step": 65940 }, { "epoch": 2.2296223672199873, "grad_norm": 0.8121903538703918, "learning_rate": 8.00877432217873e-05, "loss": 0.0583, "step": 65950 }, { "epoch": 2.2299604449102404, "grad_norm": 0.5737133026123047, "learning_rate": 8.008113983721457e-05, "loss": 0.058, "step": 65960 }, { "epoch": 2.2302985226004934, "grad_norm": 0.6180956959724426, "learning_rate": 8.007453563023477e-05, "loss": 0.064, "step": 65970 }, { "epoch": 2.230636600290747, "grad_norm": 0.417612761259079, "learning_rate": 8.006793060102845e-05, "loss": 0.0581, "step": 65980 }, { "epoch": 2.230974677981, "grad_norm": 0.6766437888145447, "learning_rate": 8.006132474977622e-05, "loss": 0.0589, "step": 65990 }, { "epoch": 2.231312755671253, "grad_norm": 0.5331613421440125, "learning_rate": 8.005471807665866e-05, "loss": 0.0642, "step": 66000 }, { "epoch": 2.2316508333615066, "grad_norm": 0.38946303725242615, "learning_rate": 8.004811058185639e-05, "loss": 0.0561, "step": 66010 }, { "epoch": 2.2319889110517597, "grad_norm": 1.2141838073730469, "learning_rate": 8.004150226555008e-05, "loss": 0.0591, "step": 66020 }, { "epoch": 2.2323269887420127, "grad_norm": 1.5534725189208984, "learning_rate": 8.003489312792038e-05, "loss": 0.0587, "step": 66030 }, { "epoch": 2.2326650664322663, "grad_norm": 0.558381974697113, "learning_rate": 8.002828316914797e-05, "loss": 0.0649, "step": 66040 }, { "epoch": 2.2330031441225193, "grad_norm": 0.650886595249176, "learning_rate": 8.00216723894136e-05, "loss": 0.0615, "step": 66050 }, { "epoch": 2.2333412218127724, "grad_norm": 0.6230465173721313, "learning_rate": 8.001506078889798e-05, "loss": 0.0611, "step": 66060 }, { "epoch": 2.233679299503026, "grad_norm": 0.953870952129364, "learning_rate": 8.000844836778188e-05, "loss": 0.0626, "step": 66070 }, { "epoch": 2.234017377193279, "grad_norm": 0.41388872265815735, "learning_rate": 8.000183512624604e-05, "loss": 0.0625, "step": 66080 }, { "epoch": 2.234355454883532, "grad_norm": 0.6151312589645386, "learning_rate": 7.999522106447133e-05, "loss": 0.0608, "step": 66090 }, { "epoch": 2.2346935325737856, "grad_norm": 0.8795665502548218, "learning_rate": 7.998860618263854e-05, "loss": 0.0582, "step": 66100 }, { "epoch": 2.2350316102640386, "grad_norm": 0.9132431149482727, "learning_rate": 7.998199048092854e-05, "loss": 0.0626, "step": 66110 }, { "epoch": 2.2353696879542917, "grad_norm": 0.4674036502838135, "learning_rate": 7.997537395952215e-05, "loss": 0.0653, "step": 66120 }, { "epoch": 2.235707765644545, "grad_norm": 0.49872463941574097, "learning_rate": 7.996875661860032e-05, "loss": 0.0596, "step": 66130 }, { "epoch": 2.2360458433347983, "grad_norm": 1.3281978368759155, "learning_rate": 7.996213845834393e-05, "loss": 0.0544, "step": 66140 }, { "epoch": 2.236383921025052, "grad_norm": 6.521726608276367, "learning_rate": 7.995551947893394e-05, "loss": 0.0625, "step": 66150 }, { "epoch": 2.236721998715305, "grad_norm": 0.3835180103778839, "learning_rate": 7.99488996805513e-05, "loss": 0.0593, "step": 66160 }, { "epoch": 2.237060076405558, "grad_norm": 0.5620945692062378, "learning_rate": 7.9942279063377e-05, "loss": 0.0598, "step": 66170 }, { "epoch": 2.2373981540958114, "grad_norm": 0.5151807069778442, "learning_rate": 7.993565762759203e-05, "loss": 0.0546, "step": 66180 }, { "epoch": 2.2377362317860645, "grad_norm": 0.5717644095420837, "learning_rate": 7.992903537337743e-05, "loss": 0.0567, "step": 66190 }, { "epoch": 2.2380743094763176, "grad_norm": 0.7924454212188721, "learning_rate": 7.992241230091425e-05, "loss": 0.0567, "step": 66200 }, { "epoch": 2.238412387166571, "grad_norm": 0.3432428538799286, "learning_rate": 7.991578841038355e-05, "loss": 0.0608, "step": 66210 }, { "epoch": 2.238750464856824, "grad_norm": 0.6749627590179443, "learning_rate": 7.990916370196644e-05, "loss": 0.0546, "step": 66220 }, { "epoch": 2.2390885425470772, "grad_norm": 0.6980394124984741, "learning_rate": 7.990253817584403e-05, "loss": 0.062, "step": 66230 }, { "epoch": 2.2394266202373307, "grad_norm": 1.0009148120880127, "learning_rate": 7.989591183219747e-05, "loss": 0.0545, "step": 66240 }, { "epoch": 2.239764697927584, "grad_norm": 0.40764620900154114, "learning_rate": 7.988928467120789e-05, "loss": 0.0614, "step": 66250 }, { "epoch": 2.240102775617837, "grad_norm": 0.6826432347297668, "learning_rate": 7.98826566930565e-05, "loss": 0.0633, "step": 66260 }, { "epoch": 2.2404408533080904, "grad_norm": 0.3795216977596283, "learning_rate": 7.987602789792452e-05, "loss": 0.0599, "step": 66270 }, { "epoch": 2.2407789309983435, "grad_norm": 0.40840527415275574, "learning_rate": 7.986939828599315e-05, "loss": 0.0597, "step": 66280 }, { "epoch": 2.2411170086885965, "grad_norm": 0.5967975854873657, "learning_rate": 7.986276785744365e-05, "loss": 0.0611, "step": 66290 }, { "epoch": 2.24145508637885, "grad_norm": 0.36213257908821106, "learning_rate": 7.98561366124573e-05, "loss": 0.0515, "step": 66300 }, { "epoch": 2.241793164069103, "grad_norm": 0.6120102405548096, "learning_rate": 7.98495045512154e-05, "loss": 0.055, "step": 66310 }, { "epoch": 2.242131241759356, "grad_norm": 0.6779550909996033, "learning_rate": 7.984287167389924e-05, "loss": 0.062, "step": 66320 }, { "epoch": 2.2424693194496097, "grad_norm": 0.7736868858337402, "learning_rate": 7.98362379806902e-05, "loss": 0.0541, "step": 66330 }, { "epoch": 2.2428073971398628, "grad_norm": 0.5561549663543701, "learning_rate": 7.98296034717696e-05, "loss": 0.0574, "step": 66340 }, { "epoch": 2.243145474830116, "grad_norm": 1.3246511220932007, "learning_rate": 7.982296814731886e-05, "loss": 0.067, "step": 66350 }, { "epoch": 2.2434835525203694, "grad_norm": 0.5752294659614563, "learning_rate": 7.981633200751937e-05, "loss": 0.0586, "step": 66360 }, { "epoch": 2.2438216302106224, "grad_norm": 0.41245466470718384, "learning_rate": 7.980969505255258e-05, "loss": 0.0585, "step": 66370 }, { "epoch": 2.2441597079008755, "grad_norm": 0.7288932800292969, "learning_rate": 7.980305728259991e-05, "loss": 0.0569, "step": 66380 }, { "epoch": 2.244497785591129, "grad_norm": 0.5895681381225586, "learning_rate": 7.979641869784286e-05, "loss": 0.0542, "step": 66390 }, { "epoch": 2.244835863281382, "grad_norm": 0.7516427040100098, "learning_rate": 7.978977929846291e-05, "loss": 0.059, "step": 66400 }, { "epoch": 2.245173940971635, "grad_norm": 0.8551175594329834, "learning_rate": 7.97831390846416e-05, "loss": 0.0621, "step": 66410 }, { "epoch": 2.2455120186618887, "grad_norm": 1.439774751663208, "learning_rate": 7.977649805656043e-05, "loss": 0.0576, "step": 66420 }, { "epoch": 2.2458500963521417, "grad_norm": 0.7662843465805054, "learning_rate": 7.9769856214401e-05, "loss": 0.0615, "step": 66430 }, { "epoch": 2.246188174042395, "grad_norm": 0.6383765339851379, "learning_rate": 7.97632135583449e-05, "loss": 0.0629, "step": 66440 }, { "epoch": 2.2465262517326483, "grad_norm": 0.4489523768424988, "learning_rate": 7.97565700885737e-05, "loss": 0.0521, "step": 66450 }, { "epoch": 2.2468643294229014, "grad_norm": 1.1108856201171875, "learning_rate": 7.974992580526907e-05, "loss": 0.0614, "step": 66460 }, { "epoch": 2.2472024071131544, "grad_norm": 1.1636981964111328, "learning_rate": 7.974328070861264e-05, "loss": 0.0662, "step": 66470 }, { "epoch": 2.247540484803408, "grad_norm": 0.5036073923110962, "learning_rate": 7.97366347987861e-05, "loss": 0.0608, "step": 66480 }, { "epoch": 2.247878562493661, "grad_norm": 1.4786334037780762, "learning_rate": 7.972998807597112e-05, "loss": 0.0592, "step": 66490 }, { "epoch": 2.248216640183914, "grad_norm": 0.42036452889442444, "learning_rate": 7.972334054034944e-05, "loss": 0.0531, "step": 66500 }, { "epoch": 2.2485547178741676, "grad_norm": 0.7258294820785522, "learning_rate": 7.971669219210281e-05, "loss": 0.057, "step": 66510 }, { "epoch": 2.2488927955644207, "grad_norm": 0.5092190504074097, "learning_rate": 7.971004303141295e-05, "loss": 0.0548, "step": 66520 }, { "epoch": 2.2492308732546737, "grad_norm": 0.5926225185394287, "learning_rate": 7.970339305846169e-05, "loss": 0.0614, "step": 66530 }, { "epoch": 2.2495689509449273, "grad_norm": 0.40596216917037964, "learning_rate": 7.969674227343082e-05, "loss": 0.0564, "step": 66540 }, { "epoch": 2.2499070286351803, "grad_norm": 0.28936800360679626, "learning_rate": 7.969009067650219e-05, "loss": 0.0524, "step": 66550 }, { "epoch": 2.2502451063254334, "grad_norm": 0.6052042245864868, "learning_rate": 7.968343826785762e-05, "loss": 0.0599, "step": 66560 }, { "epoch": 2.250583184015687, "grad_norm": 0.6350416541099548, "learning_rate": 7.967678504767899e-05, "loss": 0.0593, "step": 66570 }, { "epoch": 2.25092126170594, "grad_norm": 1.1925681829452515, "learning_rate": 7.967013101614822e-05, "loss": 0.0565, "step": 66580 }, { "epoch": 2.251259339396193, "grad_norm": 0.5072146654129028, "learning_rate": 7.96634761734472e-05, "loss": 0.0599, "step": 66590 }, { "epoch": 2.2515974170864466, "grad_norm": 0.5818628072738647, "learning_rate": 7.965682051975792e-05, "loss": 0.062, "step": 66600 }, { "epoch": 2.2519354947766996, "grad_norm": 2.3185839653015137, "learning_rate": 7.965016405526228e-05, "loss": 0.0591, "step": 66610 }, { "epoch": 2.2522735724669527, "grad_norm": 1.692866563796997, "learning_rate": 7.96435067801423e-05, "loss": 0.0554, "step": 66620 }, { "epoch": 2.252611650157206, "grad_norm": 0.5065956711769104, "learning_rate": 7.963684869457998e-05, "loss": 0.0527, "step": 66630 }, { "epoch": 2.2529497278474593, "grad_norm": 0.3599620759487152, "learning_rate": 7.963018979875735e-05, "loss": 0.0647, "step": 66640 }, { "epoch": 2.2532878055377124, "grad_norm": 0.6941595077514648, "learning_rate": 7.962353009285646e-05, "loss": 0.0621, "step": 66650 }, { "epoch": 2.253625883227966, "grad_norm": 0.6973366737365723, "learning_rate": 7.961686957705938e-05, "loss": 0.0576, "step": 66660 }, { "epoch": 2.253963960918219, "grad_norm": 0.37665194272994995, "learning_rate": 7.961020825154824e-05, "loss": 0.0578, "step": 66670 }, { "epoch": 2.254302038608472, "grad_norm": 0.45644956827163696, "learning_rate": 7.960354611650512e-05, "loss": 0.0552, "step": 66680 }, { "epoch": 2.2546401162987255, "grad_norm": 1.1045234203338623, "learning_rate": 7.959688317211216e-05, "loss": 0.0579, "step": 66690 }, { "epoch": 2.2549781939889786, "grad_norm": 0.8624269366264343, "learning_rate": 7.959021941855154e-05, "loss": 0.0611, "step": 66700 }, { "epoch": 2.2553162716792317, "grad_norm": 0.6139153838157654, "learning_rate": 7.958355485600542e-05, "loss": 0.059, "step": 66710 }, { "epoch": 2.255654349369485, "grad_norm": 0.6794095635414124, "learning_rate": 7.957688948465604e-05, "loss": 0.0577, "step": 66720 }, { "epoch": 2.2559924270597382, "grad_norm": 0.44207942485809326, "learning_rate": 7.957022330468561e-05, "loss": 0.0607, "step": 66730 }, { "epoch": 2.2563305047499913, "grad_norm": 0.476116418838501, "learning_rate": 7.95635563162764e-05, "loss": 0.0576, "step": 66740 }, { "epoch": 2.256668582440245, "grad_norm": 0.37923020124435425, "learning_rate": 7.955688851961064e-05, "loss": 0.0577, "step": 66750 }, { "epoch": 2.257006660130498, "grad_norm": 2.3284921646118164, "learning_rate": 7.955021991487067e-05, "loss": 0.0585, "step": 66760 }, { "epoch": 2.2573447378207514, "grad_norm": 0.30813226103782654, "learning_rate": 7.954355050223879e-05, "loss": 0.0551, "step": 66770 }, { "epoch": 2.2576828155110045, "grad_norm": 0.3553866147994995, "learning_rate": 7.953688028189732e-05, "loss": 0.0575, "step": 66780 }, { "epoch": 2.2580208932012575, "grad_norm": 0.26840418577194214, "learning_rate": 7.953020925402865e-05, "loss": 0.0594, "step": 66790 }, { "epoch": 2.258358970891511, "grad_norm": 0.48593607544898987, "learning_rate": 7.952353741881516e-05, "loss": 0.0541, "step": 66800 }, { "epoch": 2.258697048581764, "grad_norm": 0.5530974268913269, "learning_rate": 7.951686477643922e-05, "loss": 0.0532, "step": 66810 }, { "epoch": 2.259035126272017, "grad_norm": 0.558120608329773, "learning_rate": 7.951019132708333e-05, "loss": 0.0599, "step": 66820 }, { "epoch": 2.2593732039622707, "grad_norm": 0.5850324630737305, "learning_rate": 7.950351707092987e-05, "loss": 0.0574, "step": 66830 }, { "epoch": 2.2597112816525238, "grad_norm": 0.45022791624069214, "learning_rate": 7.949684200816133e-05, "loss": 0.0581, "step": 66840 }, { "epoch": 2.260049359342777, "grad_norm": 0.8069096207618713, "learning_rate": 7.949016613896023e-05, "loss": 0.0497, "step": 66850 }, { "epoch": 2.2603874370330304, "grad_norm": 0.5292481184005737, "learning_rate": 7.948348946350906e-05, "loss": 0.0515, "step": 66860 }, { "epoch": 2.2607255147232834, "grad_norm": 0.30156898498535156, "learning_rate": 7.947681198199036e-05, "loss": 0.0585, "step": 66870 }, { "epoch": 2.2610635924135365, "grad_norm": 0.5663498044013977, "learning_rate": 7.94701336945867e-05, "loss": 0.0637, "step": 66880 }, { "epoch": 2.26140167010379, "grad_norm": 0.6542547345161438, "learning_rate": 7.946345460148065e-05, "loss": 0.0607, "step": 66890 }, { "epoch": 2.261739747794043, "grad_norm": 0.5347739458084106, "learning_rate": 7.945677470285482e-05, "loss": 0.0551, "step": 66900 }, { "epoch": 2.262077825484296, "grad_norm": 0.7820920348167419, "learning_rate": 7.945009399889185e-05, "loss": 0.0671, "step": 66910 }, { "epoch": 2.2624159031745497, "grad_norm": 0.5369885563850403, "learning_rate": 7.944341248977436e-05, "loss": 0.0623, "step": 66920 }, { "epoch": 2.2627539808648027, "grad_norm": 0.360678493976593, "learning_rate": 7.943673017568504e-05, "loss": 0.0607, "step": 66930 }, { "epoch": 2.263092058555056, "grad_norm": 0.4847082197666168, "learning_rate": 7.943004705680656e-05, "loss": 0.0563, "step": 66940 }, { "epoch": 2.2634301362453093, "grad_norm": 1.5283987522125244, "learning_rate": 7.942336313332169e-05, "loss": 0.0579, "step": 66950 }, { "epoch": 2.2637682139355624, "grad_norm": 0.36819690465927124, "learning_rate": 7.941667840541309e-05, "loss": 0.0596, "step": 66960 }, { "epoch": 2.2641062916258154, "grad_norm": 0.3531498610973358, "learning_rate": 7.940999287326356e-05, "loss": 0.0532, "step": 66970 }, { "epoch": 2.264444369316069, "grad_norm": 0.44751447439193726, "learning_rate": 7.940330653705588e-05, "loss": 0.0535, "step": 66980 }, { "epoch": 2.264782447006322, "grad_norm": 0.31763240694999695, "learning_rate": 7.939661939697284e-05, "loss": 0.0608, "step": 66990 }, { "epoch": 2.265120524696575, "grad_norm": 1.307051658630371, "learning_rate": 7.938993145319727e-05, "loss": 0.0589, "step": 67000 }, { "epoch": 2.2654586023868286, "grad_norm": 0.37096327543258667, "learning_rate": 7.938324270591202e-05, "loss": 0.0559, "step": 67010 }, { "epoch": 2.2657966800770817, "grad_norm": 0.5414878129959106, "learning_rate": 7.937655315529995e-05, "loss": 0.0599, "step": 67020 }, { "epoch": 2.2661347577673348, "grad_norm": 0.32287272810935974, "learning_rate": 7.936986280154395e-05, "loss": 0.0565, "step": 67030 }, { "epoch": 2.2664728354575883, "grad_norm": 0.38924479484558105, "learning_rate": 7.936317164482694e-05, "loss": 0.0578, "step": 67040 }, { "epoch": 2.2668109131478413, "grad_norm": 0.5361484289169312, "learning_rate": 7.935647968533184e-05, "loss": 0.0571, "step": 67050 }, { "epoch": 2.2671489908380944, "grad_norm": 0.41635990142822266, "learning_rate": 7.934978692324162e-05, "loss": 0.0532, "step": 67060 }, { "epoch": 2.267487068528348, "grad_norm": 0.971569836139679, "learning_rate": 7.934309335873927e-05, "loss": 0.0546, "step": 67070 }, { "epoch": 2.267825146218601, "grad_norm": 0.46734336018562317, "learning_rate": 7.933639899200774e-05, "loss": 0.0576, "step": 67080 }, { "epoch": 2.2681632239088545, "grad_norm": 0.369846910238266, "learning_rate": 7.932970382323011e-05, "loss": 0.053, "step": 67090 }, { "epoch": 2.2685013015991076, "grad_norm": 0.4296899139881134, "learning_rate": 7.932300785258937e-05, "loss": 0.0565, "step": 67100 }, { "epoch": 2.2688393792893606, "grad_norm": 0.45194652676582336, "learning_rate": 7.931631108026862e-05, "loss": 0.0563, "step": 67110 }, { "epoch": 2.269177456979614, "grad_norm": 0.38032686710357666, "learning_rate": 7.930961350645095e-05, "loss": 0.0546, "step": 67120 }, { "epoch": 2.269515534669867, "grad_norm": 1.1052640676498413, "learning_rate": 7.930291513131944e-05, "loss": 0.064, "step": 67130 }, { "epoch": 2.2698536123601203, "grad_norm": 0.49983522295951843, "learning_rate": 7.929621595505726e-05, "loss": 0.0611, "step": 67140 }, { "epoch": 2.270191690050374, "grad_norm": 0.3010047674179077, "learning_rate": 7.928951597784753e-05, "loss": 0.0555, "step": 67150 }, { "epoch": 2.270529767740627, "grad_norm": 0.9573617577552795, "learning_rate": 7.928281519987343e-05, "loss": 0.0576, "step": 67160 }, { "epoch": 2.27086784543088, "grad_norm": 0.40125134587287903, "learning_rate": 7.927611362131817e-05, "loss": 0.0587, "step": 67170 }, { "epoch": 2.2712059231211335, "grad_norm": 0.3695377707481384, "learning_rate": 7.926941124236497e-05, "loss": 0.0558, "step": 67180 }, { "epoch": 2.2715440008113865, "grad_norm": 0.43816351890563965, "learning_rate": 7.926270806319705e-05, "loss": 0.0545, "step": 67190 }, { "epoch": 2.2718820785016396, "grad_norm": 0.36903560161590576, "learning_rate": 7.92560040839977e-05, "loss": 0.0617, "step": 67200 }, { "epoch": 2.272220156191893, "grad_norm": 0.34263676404953003, "learning_rate": 7.924929930495018e-05, "loss": 0.0501, "step": 67210 }, { "epoch": 2.272558233882146, "grad_norm": 0.5137849450111389, "learning_rate": 7.92425937262378e-05, "loss": 0.0597, "step": 67220 }, { "epoch": 2.2728963115723992, "grad_norm": 0.5550275444984436, "learning_rate": 7.92358873480439e-05, "loss": 0.0568, "step": 67230 }, { "epoch": 2.2732343892626528, "grad_norm": 0.28067582845687866, "learning_rate": 7.922918017055183e-05, "loss": 0.057, "step": 67240 }, { "epoch": 2.273572466952906, "grad_norm": 0.49505001306533813, "learning_rate": 7.922247219394494e-05, "loss": 0.0575, "step": 67250 }, { "epoch": 2.273910544643159, "grad_norm": 3.121304512023926, "learning_rate": 7.921576341840664e-05, "loss": 0.0643, "step": 67260 }, { "epoch": 2.2742486223334124, "grad_norm": 0.4507353901863098, "learning_rate": 7.920905384412036e-05, "loss": 0.0463, "step": 67270 }, { "epoch": 2.2745867000236655, "grad_norm": 0.734964907169342, "learning_rate": 7.920234347126951e-05, "loss": 0.0589, "step": 67280 }, { "epoch": 2.2749247777139185, "grad_norm": 0.4693196415901184, "learning_rate": 7.919563230003755e-05, "loss": 0.0631, "step": 67290 }, { "epoch": 2.275262855404172, "grad_norm": 0.3275657296180725, "learning_rate": 7.918892033060799e-05, "loss": 0.0617, "step": 67300 }, { "epoch": 2.275600933094425, "grad_norm": 0.38775312900543213, "learning_rate": 7.918220756316432e-05, "loss": 0.0509, "step": 67310 }, { "epoch": 2.275939010784678, "grad_norm": 0.5360870361328125, "learning_rate": 7.917549399789004e-05, "loss": 0.0546, "step": 67320 }, { "epoch": 2.2762770884749317, "grad_norm": 0.6115170121192932, "learning_rate": 7.916877963496871e-05, "loss": 0.0593, "step": 67330 }, { "epoch": 2.276615166165185, "grad_norm": 1.71281099319458, "learning_rate": 7.916206447458391e-05, "loss": 0.0534, "step": 67340 }, { "epoch": 2.276953243855438, "grad_norm": 0.3638629615306854, "learning_rate": 7.915534851691923e-05, "loss": 0.0605, "step": 67350 }, { "epoch": 2.2772913215456914, "grad_norm": 0.4488765001296997, "learning_rate": 7.914863176215828e-05, "loss": 0.0535, "step": 67360 }, { "epoch": 2.2776293992359444, "grad_norm": 0.6448969841003418, "learning_rate": 7.914191421048467e-05, "loss": 0.0506, "step": 67370 }, { "epoch": 2.2779674769261975, "grad_norm": 0.5304858088493347, "learning_rate": 7.91351958620821e-05, "loss": 0.0599, "step": 67380 }, { "epoch": 2.278305554616451, "grad_norm": 0.7725871205329895, "learning_rate": 7.91284767171342e-05, "loss": 0.0624, "step": 67390 }, { "epoch": 2.278643632306704, "grad_norm": 0.8832029700279236, "learning_rate": 7.91217567758247e-05, "loss": 0.0606, "step": 67400 }, { "epoch": 2.278981709996957, "grad_norm": 0.5723350644111633, "learning_rate": 7.911503603833729e-05, "loss": 0.0608, "step": 67410 }, { "epoch": 2.2793197876872107, "grad_norm": 0.48617079854011536, "learning_rate": 7.910831450485574e-05, "loss": 0.0578, "step": 67420 }, { "epoch": 2.2796578653774637, "grad_norm": 0.6069732904434204, "learning_rate": 7.910159217556382e-05, "loss": 0.0566, "step": 67430 }, { "epoch": 2.279995943067717, "grad_norm": 0.34987470507621765, "learning_rate": 7.909486905064529e-05, "loss": 0.059, "step": 67440 }, { "epoch": 2.2803340207579703, "grad_norm": 0.5238209962844849, "learning_rate": 7.908814513028398e-05, "loss": 0.0602, "step": 67450 }, { "epoch": 2.2806720984482234, "grad_norm": 0.3294472098350525, "learning_rate": 7.90814204146637e-05, "loss": 0.0546, "step": 67460 }, { "epoch": 2.2810101761384765, "grad_norm": 0.48586344718933105, "learning_rate": 7.907469490396832e-05, "loss": 0.0554, "step": 67470 }, { "epoch": 2.28134825382873, "grad_norm": 0.37151628732681274, "learning_rate": 7.90679685983817e-05, "loss": 0.0522, "step": 67480 }, { "epoch": 2.281686331518983, "grad_norm": 0.48200586438179016, "learning_rate": 7.906124149808776e-05, "loss": 0.0558, "step": 67490 }, { "epoch": 2.282024409209236, "grad_norm": 0.5102343559265137, "learning_rate": 7.905451360327037e-05, "loss": 0.06, "step": 67500 }, { "epoch": 2.2823624868994896, "grad_norm": 0.522857129573822, "learning_rate": 7.904778491411349e-05, "loss": 0.0619, "step": 67510 }, { "epoch": 2.2827005645897427, "grad_norm": 0.531410276889801, "learning_rate": 7.90410554308011e-05, "loss": 0.0563, "step": 67520 }, { "epoch": 2.2830386422799958, "grad_norm": 0.45622518658638, "learning_rate": 7.903432515351716e-05, "loss": 0.0535, "step": 67530 }, { "epoch": 2.2833767199702493, "grad_norm": 0.6207053065299988, "learning_rate": 7.902759408244566e-05, "loss": 0.0593, "step": 67540 }, { "epoch": 2.2837147976605023, "grad_norm": 0.38479849696159363, "learning_rate": 7.902086221777066e-05, "loss": 0.0596, "step": 67550 }, { "epoch": 2.2840528753507554, "grad_norm": 0.5687040090560913, "learning_rate": 7.901412955967619e-05, "loss": 0.0604, "step": 67560 }, { "epoch": 2.284390953041009, "grad_norm": 0.48557502031326294, "learning_rate": 7.900739610834632e-05, "loss": 0.0597, "step": 67570 }, { "epoch": 2.284729030731262, "grad_norm": 0.4470102787017822, "learning_rate": 7.900066186396513e-05, "loss": 0.0588, "step": 67580 }, { "epoch": 2.285067108421515, "grad_norm": 0.8984741568565369, "learning_rate": 7.899392682671675e-05, "loss": 0.0566, "step": 67590 }, { "epoch": 2.2854051861117686, "grad_norm": 0.5980446338653564, "learning_rate": 7.898719099678529e-05, "loss": 0.0598, "step": 67600 }, { "epoch": 2.2857432638020216, "grad_norm": 0.3465314209461212, "learning_rate": 7.898045437435493e-05, "loss": 0.0613, "step": 67610 }, { "epoch": 2.2860813414922747, "grad_norm": 0.3532034456729889, "learning_rate": 7.897371695960982e-05, "loss": 0.0573, "step": 67620 }, { "epoch": 2.2864194191825282, "grad_norm": 0.38513919711112976, "learning_rate": 7.89669787527342e-05, "loss": 0.058, "step": 67630 }, { "epoch": 2.2867574968727813, "grad_norm": 0.44665324687957764, "learning_rate": 7.896023975391224e-05, "loss": 0.057, "step": 67640 }, { "epoch": 2.2870955745630344, "grad_norm": 0.5253501534461975, "learning_rate": 7.895349996332821e-05, "loss": 0.059, "step": 67650 }, { "epoch": 2.287433652253288, "grad_norm": 0.2609068751335144, "learning_rate": 7.894675938116637e-05, "loss": 0.0553, "step": 67660 }, { "epoch": 2.287771729943541, "grad_norm": 0.30880481004714966, "learning_rate": 7.8940018007611e-05, "loss": 0.0537, "step": 67670 }, { "epoch": 2.288109807633794, "grad_norm": 0.44981279969215393, "learning_rate": 7.893327584284642e-05, "loss": 0.0637, "step": 67680 }, { "epoch": 2.2884478853240475, "grad_norm": 0.5096108317375183, "learning_rate": 7.892653288705693e-05, "loss": 0.0542, "step": 67690 }, { "epoch": 2.2887859630143006, "grad_norm": 0.5516252517700195, "learning_rate": 7.891978914042691e-05, "loss": 0.0582, "step": 67700 }, { "epoch": 2.2891240407045537, "grad_norm": 0.42458605766296387, "learning_rate": 7.891304460314073e-05, "loss": 0.0598, "step": 67710 }, { "epoch": 2.289462118394807, "grad_norm": 0.3005657196044922, "learning_rate": 7.890629927538276e-05, "loss": 0.0574, "step": 67720 }, { "epoch": 2.2898001960850602, "grad_norm": 0.5613632202148438, "learning_rate": 7.889955315733742e-05, "loss": 0.0566, "step": 67730 }, { "epoch": 2.2901382737753138, "grad_norm": 0.40040475130081177, "learning_rate": 7.889280624918917e-05, "loss": 0.0563, "step": 67740 }, { "epoch": 2.290476351465567, "grad_norm": 0.6070114374160767, "learning_rate": 7.888605855112244e-05, "loss": 0.0527, "step": 67750 }, { "epoch": 2.29081442915582, "grad_norm": 0.47576531767845154, "learning_rate": 7.887931006332173e-05, "loss": 0.0586, "step": 67760 }, { "epoch": 2.2911525068460734, "grad_norm": 0.36896392703056335, "learning_rate": 7.887256078597152e-05, "loss": 0.052, "step": 67770 }, { "epoch": 2.2914905845363265, "grad_norm": 0.2954486906528473, "learning_rate": 7.886581071925636e-05, "loss": 0.0556, "step": 67780 }, { "epoch": 2.2918286622265795, "grad_norm": 0.43400344252586365, "learning_rate": 7.885905986336077e-05, "loss": 0.0568, "step": 67790 }, { "epoch": 2.292166739916833, "grad_norm": 0.4238629639148712, "learning_rate": 7.885230821846931e-05, "loss": 0.0618, "step": 67800 }, { "epoch": 2.292504817607086, "grad_norm": 0.4552905857563019, "learning_rate": 7.88455557847666e-05, "loss": 0.0575, "step": 67810 }, { "epoch": 2.292842895297339, "grad_norm": 1.189861536026001, "learning_rate": 7.883880256243724e-05, "loss": 0.0534, "step": 67820 }, { "epoch": 2.2931809729875927, "grad_norm": 0.5563831329345703, "learning_rate": 7.883204855166584e-05, "loss": 0.0625, "step": 67830 }, { "epoch": 2.293519050677846, "grad_norm": 0.38153019547462463, "learning_rate": 7.882529375263707e-05, "loss": 0.0572, "step": 67840 }, { "epoch": 2.293857128368099, "grad_norm": 0.2782488167285919, "learning_rate": 7.88185381655356e-05, "loss": 0.0593, "step": 67850 }, { "epoch": 2.2941952060583524, "grad_norm": 0.4983495771884918, "learning_rate": 7.881178179054611e-05, "loss": 0.0577, "step": 67860 }, { "epoch": 2.2945332837486054, "grad_norm": 0.5712295770645142, "learning_rate": 7.880502462785335e-05, "loss": 0.0618, "step": 67870 }, { "epoch": 2.2948713614388585, "grad_norm": 0.6195652484893799, "learning_rate": 7.879826667764203e-05, "loss": 0.0635, "step": 67880 }, { "epoch": 2.295209439129112, "grad_norm": 0.3440183699131012, "learning_rate": 7.879150794009691e-05, "loss": 0.0667, "step": 67890 }, { "epoch": 2.295547516819365, "grad_norm": 0.2633994519710541, "learning_rate": 7.878474841540278e-05, "loss": 0.0602, "step": 67900 }, { "epoch": 2.295885594509618, "grad_norm": 0.4944263696670532, "learning_rate": 7.877798810374446e-05, "loss": 0.0528, "step": 67910 }, { "epoch": 2.2962236721998717, "grad_norm": 0.752147912979126, "learning_rate": 7.877122700530672e-05, "loss": 0.0555, "step": 67920 }, { "epoch": 2.2965617498901247, "grad_norm": 1.3208098411560059, "learning_rate": 7.876446512027448e-05, "loss": 0.0576, "step": 67930 }, { "epoch": 2.296899827580378, "grad_norm": 0.6635867357254028, "learning_rate": 7.875770244883256e-05, "loss": 0.0549, "step": 67940 }, { "epoch": 2.2972379052706313, "grad_norm": 0.5590217113494873, "learning_rate": 7.875093899116586e-05, "loss": 0.0564, "step": 67950 }, { "epoch": 2.2975759829608844, "grad_norm": 0.49745017290115356, "learning_rate": 7.874417474745928e-05, "loss": 0.0535, "step": 67960 }, { "epoch": 2.2979140606511375, "grad_norm": 0.3174096345901489, "learning_rate": 7.873740971789775e-05, "loss": 0.0555, "step": 67970 }, { "epoch": 2.298252138341391, "grad_norm": 0.5994954705238342, "learning_rate": 7.873064390266625e-05, "loss": 0.0611, "step": 67980 }, { "epoch": 2.298590216031644, "grad_norm": 0.47625336050987244, "learning_rate": 7.872387730194974e-05, "loss": 0.064, "step": 67990 }, { "epoch": 2.298928293721897, "grad_norm": 0.5612407326698303, "learning_rate": 7.871710991593322e-05, "loss": 0.0634, "step": 68000 }, { "epoch": 2.2992663714121506, "grad_norm": 0.3432050049304962, "learning_rate": 7.87103417448017e-05, "loss": 0.0673, "step": 68010 }, { "epoch": 2.2996044491024037, "grad_norm": 0.4966222643852234, "learning_rate": 7.870357278874021e-05, "loss": 0.0584, "step": 68020 }, { "epoch": 2.2999425267926568, "grad_norm": 0.5097094178199768, "learning_rate": 7.869680304793385e-05, "loss": 0.0546, "step": 68030 }, { "epoch": 2.3002806044829103, "grad_norm": 1.537979006767273, "learning_rate": 7.869003252256765e-05, "loss": 0.053, "step": 68040 }, { "epoch": 2.3006186821731633, "grad_norm": 0.9031146764755249, "learning_rate": 7.868326121282674e-05, "loss": 0.0639, "step": 68050 }, { "epoch": 2.300956759863417, "grad_norm": 0.5023351311683655, "learning_rate": 7.867648911889625e-05, "loss": 0.0588, "step": 68060 }, { "epoch": 2.30129483755367, "grad_norm": 0.4911010265350342, "learning_rate": 7.866971624096133e-05, "loss": 0.0535, "step": 68070 }, { "epoch": 2.301632915243923, "grad_norm": 0.6872576475143433, "learning_rate": 7.866294257920715e-05, "loss": 0.0592, "step": 68080 }, { "epoch": 2.3019709929341765, "grad_norm": 0.688103199005127, "learning_rate": 7.865616813381887e-05, "loss": 0.0562, "step": 68090 }, { "epoch": 2.3023090706244296, "grad_norm": 0.6605334281921387, "learning_rate": 7.864939290498171e-05, "loss": 0.074, "step": 68100 }, { "epoch": 2.3026471483146826, "grad_norm": 0.5469846725463867, "learning_rate": 7.86426168928809e-05, "loss": 0.0626, "step": 68110 }, { "epoch": 2.302985226004936, "grad_norm": 0.4422467350959778, "learning_rate": 7.863584009770173e-05, "loss": 0.0604, "step": 68120 }, { "epoch": 2.3033233036951892, "grad_norm": 0.9615902900695801, "learning_rate": 7.862906251962944e-05, "loss": 0.0537, "step": 68130 }, { "epoch": 2.3036613813854423, "grad_norm": 0.4837368130683899, "learning_rate": 7.862228415884936e-05, "loss": 0.061, "step": 68140 }, { "epoch": 2.303999459075696, "grad_norm": 0.4811903238296509, "learning_rate": 7.861550501554676e-05, "loss": 0.0641, "step": 68150 }, { "epoch": 2.304337536765949, "grad_norm": 0.8557948470115662, "learning_rate": 7.860872508990701e-05, "loss": 0.0614, "step": 68160 }, { "epoch": 2.304675614456202, "grad_norm": 0.5874652862548828, "learning_rate": 7.860194438211545e-05, "loss": 0.0673, "step": 68170 }, { "epoch": 2.3050136921464555, "grad_norm": 0.6844581365585327, "learning_rate": 7.859516289235748e-05, "loss": 0.0549, "step": 68180 }, { "epoch": 2.3053517698367085, "grad_norm": 0.5204153656959534, "learning_rate": 7.858838062081853e-05, "loss": 0.0518, "step": 68190 }, { "epoch": 2.3056898475269616, "grad_norm": 0.5269920229911804, "learning_rate": 7.858159756768396e-05, "loss": 0.0589, "step": 68200 }, { "epoch": 2.306027925217215, "grad_norm": 1.1785558462142944, "learning_rate": 7.857481373313926e-05, "loss": 0.0595, "step": 68210 }, { "epoch": 2.306366002907468, "grad_norm": 0.3732657730579376, "learning_rate": 7.856802911736989e-05, "loss": 0.0608, "step": 68220 }, { "epoch": 2.3067040805977213, "grad_norm": 0.5767748951911926, "learning_rate": 7.856124372056133e-05, "loss": 0.058, "step": 68230 }, { "epoch": 2.3070421582879748, "grad_norm": 0.5547166466712952, "learning_rate": 7.855445754289911e-05, "loss": 0.0557, "step": 68240 }, { "epoch": 2.307380235978228, "grad_norm": 0.8563647866249084, "learning_rate": 7.854767058456873e-05, "loss": 0.063, "step": 68250 }, { "epoch": 2.307718313668481, "grad_norm": 0.619892954826355, "learning_rate": 7.854088284575579e-05, "loss": 0.0576, "step": 68260 }, { "epoch": 2.3080563913587344, "grad_norm": 0.36380714178085327, "learning_rate": 7.85340943266458e-05, "loss": 0.0558, "step": 68270 }, { "epoch": 2.3083944690489875, "grad_norm": 0.5440890192985535, "learning_rate": 7.852730502742441e-05, "loss": 0.0531, "step": 68280 }, { "epoch": 2.3087325467392406, "grad_norm": 1.0857629776000977, "learning_rate": 7.852051494827722e-05, "loss": 0.0597, "step": 68290 }, { "epoch": 2.309070624429494, "grad_norm": 0.5843140482902527, "learning_rate": 7.851372408938985e-05, "loss": 0.0579, "step": 68300 }, { "epoch": 2.309408702119747, "grad_norm": 0.44978970289230347, "learning_rate": 7.8506932450948e-05, "loss": 0.059, "step": 68310 }, { "epoch": 2.30974677981, "grad_norm": 0.30619531869888306, "learning_rate": 7.85001400331373e-05, "loss": 0.0576, "step": 68320 }, { "epoch": 2.3100848575002537, "grad_norm": 0.44520848989486694, "learning_rate": 7.849334683614349e-05, "loss": 0.0543, "step": 68330 }, { "epoch": 2.310422935190507, "grad_norm": 0.6039217710494995, "learning_rate": 7.848655286015229e-05, "loss": 0.0585, "step": 68340 }, { "epoch": 2.31076101288076, "grad_norm": 0.5791872143745422, "learning_rate": 7.847975810534943e-05, "loss": 0.0599, "step": 68350 }, { "epoch": 2.3110990905710134, "grad_norm": 0.3546658158302307, "learning_rate": 7.847296257192069e-05, "loss": 0.0597, "step": 68360 }, { "epoch": 2.3114371682612664, "grad_norm": 0.5524914860725403, "learning_rate": 7.846616626005183e-05, "loss": 0.0544, "step": 68370 }, { "epoch": 2.3117752459515195, "grad_norm": 0.837976336479187, "learning_rate": 7.845936916992869e-05, "loss": 0.0605, "step": 68380 }, { "epoch": 2.312113323641773, "grad_norm": 0.7536798715591431, "learning_rate": 7.845257130173709e-05, "loss": 0.0561, "step": 68390 }, { "epoch": 2.312451401332026, "grad_norm": 0.45288804173469543, "learning_rate": 7.844577265566289e-05, "loss": 0.0547, "step": 68400 }, { "epoch": 2.312789479022279, "grad_norm": 0.9394175410270691, "learning_rate": 7.843897323189194e-05, "loss": 0.0501, "step": 68410 }, { "epoch": 2.3131275567125327, "grad_norm": 0.5232017040252686, "learning_rate": 7.843217303061014e-05, "loss": 0.0599, "step": 68420 }, { "epoch": 2.3134656344027857, "grad_norm": 0.5795779228210449, "learning_rate": 7.842537205200342e-05, "loss": 0.0552, "step": 68430 }, { "epoch": 2.313803712093039, "grad_norm": 0.3857165575027466, "learning_rate": 7.841857029625771e-05, "loss": 0.0563, "step": 68440 }, { "epoch": 2.3141417897832923, "grad_norm": 0.6607354879379272, "learning_rate": 7.841176776355897e-05, "loss": 0.0581, "step": 68450 }, { "epoch": 2.3144798674735454, "grad_norm": 0.5665524005889893, "learning_rate": 7.840496445409316e-05, "loss": 0.0554, "step": 68460 }, { "epoch": 2.3148179451637985, "grad_norm": 0.8548925518989563, "learning_rate": 7.839816036804628e-05, "loss": 0.0591, "step": 68470 }, { "epoch": 2.315156022854052, "grad_norm": 0.6604849100112915, "learning_rate": 7.83913555056044e-05, "loss": 0.0549, "step": 68480 }, { "epoch": 2.315494100544305, "grad_norm": 0.5374889969825745, "learning_rate": 7.83845498669535e-05, "loss": 0.0498, "step": 68490 }, { "epoch": 2.315832178234558, "grad_norm": 0.3147391378879547, "learning_rate": 7.837774345227966e-05, "loss": 0.0571, "step": 68500 }, { "epoch": 2.3161702559248116, "grad_norm": 0.5657776594161987, "learning_rate": 7.8370936261769e-05, "loss": 0.0574, "step": 68510 }, { "epoch": 2.3165083336150647, "grad_norm": 0.7565991878509521, "learning_rate": 7.83641282956076e-05, "loss": 0.0609, "step": 68520 }, { "epoch": 2.3168464113053178, "grad_norm": 0.370758980512619, "learning_rate": 7.835731955398158e-05, "loss": 0.0628, "step": 68530 }, { "epoch": 2.3171844889955713, "grad_norm": 0.47566020488739014, "learning_rate": 7.83505100370771e-05, "loss": 0.0582, "step": 68540 }, { "epoch": 2.3175225666858243, "grad_norm": 0.5885264873504639, "learning_rate": 7.834369974508033e-05, "loss": 0.0552, "step": 68550 }, { "epoch": 2.3178606443760774, "grad_norm": 0.5433377027511597, "learning_rate": 7.833688867817744e-05, "loss": 0.0557, "step": 68560 }, { "epoch": 2.318198722066331, "grad_norm": 1.340983271598816, "learning_rate": 7.833007683655466e-05, "loss": 0.0585, "step": 68570 }, { "epoch": 2.318536799756584, "grad_norm": 0.4307805895805359, "learning_rate": 7.832326422039823e-05, "loss": 0.0571, "step": 68580 }, { "epoch": 2.318874877446837, "grad_norm": 0.4872375726699829, "learning_rate": 7.831645082989441e-05, "loss": 0.0643, "step": 68590 }, { "epoch": 2.3192129551370906, "grad_norm": 0.5578970313072205, "learning_rate": 7.830963666522946e-05, "loss": 0.0579, "step": 68600 }, { "epoch": 2.3195510328273437, "grad_norm": 0.7818518877029419, "learning_rate": 7.830282172658967e-05, "loss": 0.0596, "step": 68610 }, { "epoch": 2.3198891105175967, "grad_norm": 0.6472999453544617, "learning_rate": 7.829600601416136e-05, "loss": 0.0572, "step": 68620 }, { "epoch": 2.3202271882078502, "grad_norm": 0.6738526225090027, "learning_rate": 7.82891895281309e-05, "loss": 0.0601, "step": 68630 }, { "epoch": 2.3205652658981033, "grad_norm": 0.6897426247596741, "learning_rate": 7.82823722686846e-05, "loss": 0.0603, "step": 68640 }, { "epoch": 2.3209033435883564, "grad_norm": 0.4921148419380188, "learning_rate": 7.827555423600889e-05, "loss": 0.0597, "step": 68650 }, { "epoch": 2.32124142127861, "grad_norm": 1.522993803024292, "learning_rate": 7.826873543029013e-05, "loss": 0.0582, "step": 68660 }, { "epoch": 2.321579498968863, "grad_norm": 0.6625575423240662, "learning_rate": 7.826191585171477e-05, "loss": 0.0555, "step": 68670 }, { "epoch": 2.321917576659116, "grad_norm": 0.8186103105545044, "learning_rate": 7.825509550046926e-05, "loss": 0.0618, "step": 68680 }, { "epoch": 2.3222556543493695, "grad_norm": 1.0623893737792969, "learning_rate": 7.824827437674002e-05, "loss": 0.063, "step": 68690 }, { "epoch": 2.3225937320396226, "grad_norm": 0.8539764881134033, "learning_rate": 7.824145248071361e-05, "loss": 0.0606, "step": 68700 }, { "epoch": 2.322931809729876, "grad_norm": 0.9251818060874939, "learning_rate": 7.823462981257648e-05, "loss": 0.0645, "step": 68710 }, { "epoch": 2.323269887420129, "grad_norm": 0.9054690599441528, "learning_rate": 7.822780637251518e-05, "loss": 0.0559, "step": 68720 }, { "epoch": 2.3236079651103823, "grad_norm": 0.7889193892478943, "learning_rate": 7.822098216071626e-05, "loss": 0.0599, "step": 68730 }, { "epoch": 2.3239460428006358, "grad_norm": 0.8251268267631531, "learning_rate": 7.821415717736629e-05, "loss": 0.0602, "step": 68740 }, { "epoch": 2.324284120490889, "grad_norm": 0.46852797269821167, "learning_rate": 7.820733142265186e-05, "loss": 0.0611, "step": 68750 }, { "epoch": 2.324622198181142, "grad_norm": 0.4649711847305298, "learning_rate": 7.820050489675959e-05, "loss": 0.0578, "step": 68760 }, { "epoch": 2.3249602758713954, "grad_norm": 0.5148841142654419, "learning_rate": 7.81936775998761e-05, "loss": 0.0523, "step": 68770 }, { "epoch": 2.3252983535616485, "grad_norm": 1.0202136039733887, "learning_rate": 7.818684953218807e-05, "loss": 0.0642, "step": 68780 }, { "epoch": 2.3256364312519016, "grad_norm": 0.5615414977073669, "learning_rate": 7.818002069388215e-05, "loss": 0.0615, "step": 68790 }, { "epoch": 2.325974508942155, "grad_norm": 0.5269373655319214, "learning_rate": 7.817319108514507e-05, "loss": 0.0635, "step": 68800 }, { "epoch": 2.326312586632408, "grad_norm": 0.594315230846405, "learning_rate": 7.816636070616352e-05, "loss": 0.0652, "step": 68810 }, { "epoch": 2.326650664322661, "grad_norm": 0.5600699186325073, "learning_rate": 7.815952955712425e-05, "loss": 0.0543, "step": 68820 }, { "epoch": 2.3269887420129147, "grad_norm": 0.5877501964569092, "learning_rate": 7.815269763821401e-05, "loss": 0.0621, "step": 68830 }, { "epoch": 2.327326819703168, "grad_norm": 0.6025657057762146, "learning_rate": 7.814586494961961e-05, "loss": 0.0629, "step": 68840 }, { "epoch": 2.327664897393421, "grad_norm": 0.4014146029949188, "learning_rate": 7.813903149152783e-05, "loss": 0.0559, "step": 68850 }, { "epoch": 2.3280029750836744, "grad_norm": 0.7804610133171082, "learning_rate": 7.813219726412549e-05, "loss": 0.0587, "step": 68860 }, { "epoch": 2.3283410527739274, "grad_norm": 0.6891326904296875, "learning_rate": 7.812536226759945e-05, "loss": 0.0564, "step": 68870 }, { "epoch": 2.3286791304641805, "grad_norm": 0.44596022367477417, "learning_rate": 7.811852650213658e-05, "loss": 0.0628, "step": 68880 }, { "epoch": 2.329017208154434, "grad_norm": 0.6730589270591736, "learning_rate": 7.811168996792376e-05, "loss": 0.0602, "step": 68890 }, { "epoch": 2.329355285844687, "grad_norm": 0.6338079571723938, "learning_rate": 7.810485266514789e-05, "loss": 0.0612, "step": 68900 }, { "epoch": 2.32969336353494, "grad_norm": 0.4840381145477295, "learning_rate": 7.809801459399592e-05, "loss": 0.0555, "step": 68910 }, { "epoch": 2.3300314412251937, "grad_norm": 0.5350661277770996, "learning_rate": 7.809117575465476e-05, "loss": 0.0604, "step": 68920 }, { "epoch": 2.3303695189154467, "grad_norm": 0.7815043330192566, "learning_rate": 7.808433614731144e-05, "loss": 0.0573, "step": 68930 }, { "epoch": 2.3307075966057, "grad_norm": 0.7333648800849915, "learning_rate": 7.80774957721529e-05, "loss": 0.0621, "step": 68940 }, { "epoch": 2.3310456742959533, "grad_norm": 0.5363223552703857, "learning_rate": 7.807065462936618e-05, "loss": 0.0582, "step": 68950 }, { "epoch": 2.3313837519862064, "grad_norm": 0.5706125497817993, "learning_rate": 7.80638127191383e-05, "loss": 0.0687, "step": 68960 }, { "epoch": 2.3317218296764595, "grad_norm": 0.42012539505958557, "learning_rate": 7.805697004165633e-05, "loss": 0.0546, "step": 68970 }, { "epoch": 2.332059907366713, "grad_norm": 0.8520643711090088, "learning_rate": 7.805012659710734e-05, "loss": 0.0573, "step": 68980 }, { "epoch": 2.332397985056966, "grad_norm": 0.8064644932746887, "learning_rate": 7.804328238567844e-05, "loss": 0.0584, "step": 68990 }, { "epoch": 2.332736062747219, "grad_norm": 0.6713204383850098, "learning_rate": 7.803643740755672e-05, "loss": 0.0609, "step": 69000 }, { "epoch": 2.3330741404374726, "grad_norm": 2.81687068939209, "learning_rate": 7.802959166292932e-05, "loss": 0.0646, "step": 69010 }, { "epoch": 2.3334122181277257, "grad_norm": 0.6308516263961792, "learning_rate": 7.802274515198343e-05, "loss": 0.0651, "step": 69020 }, { "epoch": 2.333750295817979, "grad_norm": 0.5208409428596497, "learning_rate": 7.801589787490622e-05, "loss": 0.0606, "step": 69030 }, { "epoch": 2.3340883735082323, "grad_norm": 5.369491100311279, "learning_rate": 7.800904983188488e-05, "loss": 0.0559, "step": 69040 }, { "epoch": 2.3344264511984854, "grad_norm": 0.5461375117301941, "learning_rate": 7.800220102310665e-05, "loss": 0.0606, "step": 69050 }, { "epoch": 2.334764528888739, "grad_norm": 0.5543222427368164, "learning_rate": 7.799535144875875e-05, "loss": 0.0607, "step": 69060 }, { "epoch": 2.335102606578992, "grad_norm": 0.6996274590492249, "learning_rate": 7.798850110902847e-05, "loss": 0.06, "step": 69070 }, { "epoch": 2.335440684269245, "grad_norm": 0.4534449279308319, "learning_rate": 7.798165000410308e-05, "loss": 0.0579, "step": 69080 }, { "epoch": 2.3357787619594985, "grad_norm": 0.8309826850891113, "learning_rate": 7.797479813416989e-05, "loss": 0.0575, "step": 69090 }, { "epoch": 2.3361168396497516, "grad_norm": 0.4234558045864105, "learning_rate": 7.796794549941622e-05, "loss": 0.0602, "step": 69100 }, { "epoch": 2.3364549173400047, "grad_norm": 0.5529680848121643, "learning_rate": 7.796109210002945e-05, "loss": 0.0566, "step": 69110 }, { "epoch": 2.336792995030258, "grad_norm": 0.5437840819358826, "learning_rate": 7.795423793619692e-05, "loss": 0.0571, "step": 69120 }, { "epoch": 2.3371310727205112, "grad_norm": 0.8358700275421143, "learning_rate": 7.7947383008106e-05, "loss": 0.0612, "step": 69130 }, { "epoch": 2.3374691504107643, "grad_norm": 0.48025596141815186, "learning_rate": 7.794052731594415e-05, "loss": 0.0597, "step": 69140 }, { "epoch": 2.337807228101018, "grad_norm": 1.6934219598770142, "learning_rate": 7.793367085989877e-05, "loss": 0.0582, "step": 69150 }, { "epoch": 2.338145305791271, "grad_norm": 0.5028706789016724, "learning_rate": 7.792681364015735e-05, "loss": 0.0662, "step": 69160 }, { "epoch": 2.338483383481524, "grad_norm": 0.4916921854019165, "learning_rate": 7.791995565690731e-05, "loss": 0.0575, "step": 69170 }, { "epoch": 2.3388214611717775, "grad_norm": 0.6862167716026306, "learning_rate": 7.791309691033619e-05, "loss": 0.0583, "step": 69180 }, { "epoch": 2.3391595388620305, "grad_norm": 0.5416519045829773, "learning_rate": 7.790623740063147e-05, "loss": 0.0551, "step": 69190 }, { "epoch": 2.3394976165522836, "grad_norm": 0.5135542750358582, "learning_rate": 7.78993771279807e-05, "loss": 0.0612, "step": 69200 }, { "epoch": 2.339835694242537, "grad_norm": 3.1114745140075684, "learning_rate": 7.789251609257145e-05, "loss": 0.0593, "step": 69210 }, { "epoch": 2.34017377193279, "grad_norm": 0.9324019551277161, "learning_rate": 7.788565429459129e-05, "loss": 0.0639, "step": 69220 }, { "epoch": 2.3405118496230433, "grad_norm": 0.6297482848167419, "learning_rate": 7.787879173422782e-05, "loss": 0.0621, "step": 69230 }, { "epoch": 2.3408499273132968, "grad_norm": 37.89619445800781, "learning_rate": 7.787192841166866e-05, "loss": 0.0586, "step": 69240 }, { "epoch": 2.34118800500355, "grad_norm": 0.35560786724090576, "learning_rate": 7.786506432710144e-05, "loss": 0.0592, "step": 69250 }, { "epoch": 2.341526082693803, "grad_norm": 0.7722256183624268, "learning_rate": 7.785819948071382e-05, "loss": 0.0653, "step": 69260 }, { "epoch": 2.3418641603840564, "grad_norm": 0.40586188435554504, "learning_rate": 7.78513338726935e-05, "loss": 0.0631, "step": 69270 }, { "epoch": 2.3422022380743095, "grad_norm": 1.1661673784255981, "learning_rate": 7.784446750322818e-05, "loss": 0.0593, "step": 69280 }, { "epoch": 2.3425403157645626, "grad_norm": 0.8237074613571167, "learning_rate": 7.783760037250558e-05, "loss": 0.0596, "step": 69290 }, { "epoch": 2.342878393454816, "grad_norm": 0.4148918092250824, "learning_rate": 7.783073248071344e-05, "loss": 0.0577, "step": 69300 }, { "epoch": 2.343216471145069, "grad_norm": 0.6588377952575684, "learning_rate": 7.782386382803954e-05, "loss": 0.0548, "step": 69310 }, { "epoch": 2.343554548835322, "grad_norm": 0.42385706305503845, "learning_rate": 7.781699441467163e-05, "loss": 0.0625, "step": 69320 }, { "epoch": 2.3438926265255757, "grad_norm": 0.3676016628742218, "learning_rate": 7.781012424079757e-05, "loss": 0.0511, "step": 69330 }, { "epoch": 2.344230704215829, "grad_norm": 0.3398737907409668, "learning_rate": 7.780325330660516e-05, "loss": 0.0628, "step": 69340 }, { "epoch": 2.344568781906082, "grad_norm": 0.5133020877838135, "learning_rate": 7.779638161228225e-05, "loss": 0.0568, "step": 69350 }, { "epoch": 2.3449068595963354, "grad_norm": 0.6136677265167236, "learning_rate": 7.778950915801669e-05, "loss": 0.0631, "step": 69360 }, { "epoch": 2.3452449372865884, "grad_norm": 0.4824298322200775, "learning_rate": 7.778263594399639e-05, "loss": 0.0565, "step": 69370 }, { "epoch": 2.3455830149768415, "grad_norm": 0.5163981914520264, "learning_rate": 7.777576197040927e-05, "loss": 0.0528, "step": 69380 }, { "epoch": 2.345921092667095, "grad_norm": 0.4402904212474823, "learning_rate": 7.776888723744327e-05, "loss": 0.0541, "step": 69390 }, { "epoch": 2.346259170357348, "grad_norm": 0.9322385191917419, "learning_rate": 7.77620117452863e-05, "loss": 0.0667, "step": 69400 }, { "epoch": 2.346597248047601, "grad_norm": 0.48141342401504517, "learning_rate": 7.775513549412637e-05, "loss": 0.0572, "step": 69410 }, { "epoch": 2.3469353257378547, "grad_norm": 0.4422679543495178, "learning_rate": 7.774825848415146e-05, "loss": 0.0585, "step": 69420 }, { "epoch": 2.3472734034281078, "grad_norm": 0.8271393775939941, "learning_rate": 7.774138071554958e-05, "loss": 0.066, "step": 69430 }, { "epoch": 2.347611481118361, "grad_norm": 0.44242700934410095, "learning_rate": 7.773450218850878e-05, "loss": 0.0622, "step": 69440 }, { "epoch": 2.3479495588086143, "grad_norm": 0.5265247225761414, "learning_rate": 7.772762290321711e-05, "loss": 0.0657, "step": 69450 }, { "epoch": 2.3482876364988674, "grad_norm": 0.36212044954299927, "learning_rate": 7.772074285986263e-05, "loss": 0.0655, "step": 69460 }, { "epoch": 2.3486257141891205, "grad_norm": 0.4063819348812103, "learning_rate": 7.77138620586335e-05, "loss": 0.0554, "step": 69470 }, { "epoch": 2.348963791879374, "grad_norm": 0.9832977056503296, "learning_rate": 7.770698049971776e-05, "loss": 0.0534, "step": 69480 }, { "epoch": 2.349301869569627, "grad_norm": 0.621142566204071, "learning_rate": 7.770009818330359e-05, "loss": 0.0574, "step": 69490 }, { "epoch": 2.34963994725988, "grad_norm": 0.5598986744880676, "learning_rate": 7.769321510957913e-05, "loss": 0.0573, "step": 69500 }, { "epoch": 2.3499780249501336, "grad_norm": 0.529160737991333, "learning_rate": 7.768633127873257e-05, "loss": 0.0632, "step": 69510 }, { "epoch": 2.3503161026403867, "grad_norm": 0.4005160629749298, "learning_rate": 7.767944669095213e-05, "loss": 0.0592, "step": 69520 }, { "epoch": 2.3506541803306398, "grad_norm": 0.4662475287914276, "learning_rate": 7.767256134642602e-05, "loss": 0.0559, "step": 69530 }, { "epoch": 2.3509922580208933, "grad_norm": 0.5619776844978333, "learning_rate": 7.766567524534248e-05, "loss": 0.0567, "step": 69540 }, { "epoch": 2.3513303357111464, "grad_norm": 0.5599259734153748, "learning_rate": 7.765878838788976e-05, "loss": 0.0623, "step": 69550 }, { "epoch": 2.3516684134013994, "grad_norm": 0.6290231943130493, "learning_rate": 7.765190077425615e-05, "loss": 0.0658, "step": 69560 }, { "epoch": 2.352006491091653, "grad_norm": 0.494303435087204, "learning_rate": 7.764501240462998e-05, "loss": 0.0569, "step": 69570 }, { "epoch": 2.352344568781906, "grad_norm": 0.54974764585495, "learning_rate": 7.763812327919953e-05, "loss": 0.0583, "step": 69580 }, { "epoch": 2.352682646472159, "grad_norm": 0.5965532064437866, "learning_rate": 7.76312333981532e-05, "loss": 0.0554, "step": 69590 }, { "epoch": 2.3530207241624126, "grad_norm": 0.7030461430549622, "learning_rate": 7.762434276167931e-05, "loss": 0.0634, "step": 69600 }, { "epoch": 2.3533588018526657, "grad_norm": 0.5887495875358582, "learning_rate": 7.761745136996628e-05, "loss": 0.0534, "step": 69610 }, { "epoch": 2.3536968795429187, "grad_norm": 0.5023970007896423, "learning_rate": 7.76105592232025e-05, "loss": 0.0582, "step": 69620 }, { "epoch": 2.3540349572331722, "grad_norm": 0.4718407988548279, "learning_rate": 7.76036663215764e-05, "loss": 0.0542, "step": 69630 }, { "epoch": 2.3543730349234253, "grad_norm": 1.007577896118164, "learning_rate": 7.759677266527644e-05, "loss": 0.0578, "step": 69640 }, { "epoch": 2.3547111126136784, "grad_norm": 0.6420173645019531, "learning_rate": 7.758987825449108e-05, "loss": 0.0581, "step": 69650 }, { "epoch": 2.355049190303932, "grad_norm": 0.8814259171485901, "learning_rate": 7.75829830894088e-05, "loss": 0.0509, "step": 69660 }, { "epoch": 2.355387267994185, "grad_norm": 0.42029398679733276, "learning_rate": 7.757608717021816e-05, "loss": 0.0622, "step": 69670 }, { "epoch": 2.3557253456844385, "grad_norm": 0.7249559164047241, "learning_rate": 7.756919049710762e-05, "loss": 0.0526, "step": 69680 }, { "epoch": 2.3560634233746915, "grad_norm": 0.42978635430336, "learning_rate": 7.75622930702658e-05, "loss": 0.0518, "step": 69690 }, { "epoch": 2.3564015010649446, "grad_norm": 0.510968804359436, "learning_rate": 7.755539488988122e-05, "loss": 0.0606, "step": 69700 }, { "epoch": 2.356739578755198, "grad_norm": 0.6775065660476685, "learning_rate": 7.75484959561425e-05, "loss": 0.0536, "step": 69710 }, { "epoch": 2.357077656445451, "grad_norm": 0.5315871834754944, "learning_rate": 7.754159626923824e-05, "loss": 0.0629, "step": 69720 }, { "epoch": 2.3574157341357043, "grad_norm": 0.41866424679756165, "learning_rate": 7.753469582935711e-05, "loss": 0.0579, "step": 69730 }, { "epoch": 2.357753811825958, "grad_norm": 0.5572869181632996, "learning_rate": 7.752779463668773e-05, "loss": 0.0604, "step": 69740 }, { "epoch": 2.358091889516211, "grad_norm": 0.9782142639160156, "learning_rate": 7.752089269141878e-05, "loss": 0.0645, "step": 69750 }, { "epoch": 2.358429967206464, "grad_norm": 0.4667436480522156, "learning_rate": 7.751398999373898e-05, "loss": 0.0586, "step": 69760 }, { "epoch": 2.3587680448967174, "grad_norm": 0.39126500487327576, "learning_rate": 7.7507086543837e-05, "loss": 0.0599, "step": 69770 }, { "epoch": 2.3591061225869705, "grad_norm": 0.4212413430213928, "learning_rate": 7.750018234190163e-05, "loss": 0.0513, "step": 69780 }, { "epoch": 2.3594442002772236, "grad_norm": 0.5297970175743103, "learning_rate": 7.74932773881216e-05, "loss": 0.0579, "step": 69790 }, { "epoch": 2.359782277967477, "grad_norm": 0.7468355894088745, "learning_rate": 7.74863716826857e-05, "loss": 0.0553, "step": 69800 }, { "epoch": 2.36012035565773, "grad_norm": 0.6562986373901367, "learning_rate": 7.747946522578272e-05, "loss": 0.0625, "step": 69810 }, { "epoch": 2.360458433347983, "grad_norm": 0.8331825137138367, "learning_rate": 7.747255801760149e-05, "loss": 0.0596, "step": 69820 }, { "epoch": 2.3607965110382367, "grad_norm": 2.2448010444641113, "learning_rate": 7.746565005833083e-05, "loss": 0.057, "step": 69830 }, { "epoch": 2.36113458872849, "grad_norm": 0.41000574827194214, "learning_rate": 7.745874134815962e-05, "loss": 0.0598, "step": 69840 }, { "epoch": 2.361472666418743, "grad_norm": 0.9806844592094421, "learning_rate": 7.745183188727674e-05, "loss": 0.0553, "step": 69850 }, { "epoch": 2.3618107441089964, "grad_norm": 0.8564830422401428, "learning_rate": 7.74449216758711e-05, "loss": 0.0641, "step": 69860 }, { "epoch": 2.3621488217992495, "grad_norm": 1.0831493139266968, "learning_rate": 7.743801071413161e-05, "loss": 0.0563, "step": 69870 }, { "epoch": 2.3624868994895025, "grad_norm": 0.45047998428344727, "learning_rate": 7.743109900224721e-05, "loss": 0.0572, "step": 69880 }, { "epoch": 2.362824977179756, "grad_norm": 0.4811283349990845, "learning_rate": 7.742418654040686e-05, "loss": 0.0573, "step": 69890 }, { "epoch": 2.363163054870009, "grad_norm": 5.693653106689453, "learning_rate": 7.741727332879956e-05, "loss": 0.0629, "step": 69900 }, { "epoch": 2.363501132560262, "grad_norm": 0.7372121810913086, "learning_rate": 7.74103593676143e-05, "loss": 0.0567, "step": 69910 }, { "epoch": 2.3638392102505157, "grad_norm": 0.6221588850021362, "learning_rate": 7.740344465704013e-05, "loss": 0.068, "step": 69920 }, { "epoch": 2.3641772879407688, "grad_norm": 0.8250868320465088, "learning_rate": 7.739652919726606e-05, "loss": 0.0563, "step": 69930 }, { "epoch": 2.364515365631022, "grad_norm": 1.1311684846878052, "learning_rate": 7.73896129884812e-05, "loss": 0.0612, "step": 69940 }, { "epoch": 2.3648534433212753, "grad_norm": 1.0182749032974243, "learning_rate": 7.738269603087458e-05, "loss": 0.0648, "step": 69950 }, { "epoch": 2.3651915210115284, "grad_norm": 1.1477560997009277, "learning_rate": 7.737577832463534e-05, "loss": 0.0607, "step": 69960 }, { "epoch": 2.3655295987017815, "grad_norm": 0.6201339960098267, "learning_rate": 7.736885986995262e-05, "loss": 0.0606, "step": 69970 }, { "epoch": 2.365867676392035, "grad_norm": 0.5433847904205322, "learning_rate": 7.736194066701555e-05, "loss": 0.0582, "step": 69980 }, { "epoch": 2.366205754082288, "grad_norm": 0.6389403343200684, "learning_rate": 7.73550207160133e-05, "loss": 0.0645, "step": 69990 }, { "epoch": 2.3665438317725416, "grad_norm": 0.3588375449180603, "learning_rate": 7.734810001713507e-05, "loss": 0.0676, "step": 70000 }, { "epoch": 2.3668819094627946, "grad_norm": 0.5894716382026672, "learning_rate": 7.734117857057005e-05, "loss": 0.0569, "step": 70010 }, { "epoch": 2.3672199871530477, "grad_norm": 0.375410795211792, "learning_rate": 7.733425637650747e-05, "loss": 0.0581, "step": 70020 }, { "epoch": 2.3675580648433012, "grad_norm": 0.5817031264305115, "learning_rate": 7.732733343513661e-05, "loss": 0.0628, "step": 70030 }, { "epoch": 2.3678961425335543, "grad_norm": 0.5136577486991882, "learning_rate": 7.732040974664671e-05, "loss": 0.0567, "step": 70040 }, { "epoch": 2.3682342202238074, "grad_norm": 0.40058591961860657, "learning_rate": 7.731348531122706e-05, "loss": 0.0547, "step": 70050 }, { "epoch": 2.368572297914061, "grad_norm": 0.37116575241088867, "learning_rate": 7.730656012906701e-05, "loss": 0.0575, "step": 70060 }, { "epoch": 2.368910375604314, "grad_norm": 0.6368156671524048, "learning_rate": 7.729963420035584e-05, "loss": 0.0573, "step": 70070 }, { "epoch": 2.369248453294567, "grad_norm": 0.6960996985435486, "learning_rate": 7.729270752528295e-05, "loss": 0.0581, "step": 70080 }, { "epoch": 2.3695865309848205, "grad_norm": 0.6384339332580566, "learning_rate": 7.728578010403766e-05, "loss": 0.0607, "step": 70090 }, { "epoch": 2.3699246086750736, "grad_norm": 0.7226912975311279, "learning_rate": 7.727885193680942e-05, "loss": 0.0529, "step": 70100 }, { "epoch": 2.3702626863653267, "grad_norm": 0.5301994681358337, "learning_rate": 7.727192302378761e-05, "loss": 0.0637, "step": 70110 }, { "epoch": 2.37060076405558, "grad_norm": 0.5751892924308777, "learning_rate": 7.726499336516165e-05, "loss": 0.0635, "step": 70120 }, { "epoch": 2.3709388417458332, "grad_norm": 1.2426520586013794, "learning_rate": 7.725806296112104e-05, "loss": 0.0616, "step": 70130 }, { "epoch": 2.3712769194360863, "grad_norm": 1.4395211935043335, "learning_rate": 7.725113181185521e-05, "loss": 0.0571, "step": 70140 }, { "epoch": 2.37161499712634, "grad_norm": 0.35084035992622375, "learning_rate": 7.724419991755367e-05, "loss": 0.0602, "step": 70150 }, { "epoch": 2.371953074816593, "grad_norm": 0.6456379890441895, "learning_rate": 7.723726727840596e-05, "loss": 0.0655, "step": 70160 }, { "epoch": 2.372291152506846, "grad_norm": 0.4993957281112671, "learning_rate": 7.723033389460159e-05, "loss": 0.0574, "step": 70170 }, { "epoch": 2.3726292301970995, "grad_norm": 0.5104308128356934, "learning_rate": 7.722339976633012e-05, "loss": 0.063, "step": 70180 }, { "epoch": 2.3729673078873526, "grad_norm": 1.5258249044418335, "learning_rate": 7.721646489378112e-05, "loss": 0.0519, "step": 70190 }, { "epoch": 2.3733053855776056, "grad_norm": 0.6491238474845886, "learning_rate": 7.72095292771442e-05, "loss": 0.0579, "step": 70200 }, { "epoch": 2.373643463267859, "grad_norm": 0.6840458512306213, "learning_rate": 7.720259291660898e-05, "loss": 0.0609, "step": 70210 }, { "epoch": 2.373981540958112, "grad_norm": 0.3636995255947113, "learning_rate": 7.719565581236508e-05, "loss": 0.0581, "step": 70220 }, { "epoch": 2.3743196186483653, "grad_norm": 0.3425467014312744, "learning_rate": 7.718871796460217e-05, "loss": 0.0639, "step": 70230 }, { "epoch": 2.374657696338619, "grad_norm": 0.6284688711166382, "learning_rate": 7.718177937350992e-05, "loss": 0.0603, "step": 70240 }, { "epoch": 2.374995774028872, "grad_norm": 0.5020453333854675, "learning_rate": 7.717484003927804e-05, "loss": 0.0597, "step": 70250 }, { "epoch": 2.375333851719125, "grad_norm": 0.42189934849739075, "learning_rate": 7.716789996209623e-05, "loss": 0.0609, "step": 70260 }, { "epoch": 2.3756719294093784, "grad_norm": 0.693902850151062, "learning_rate": 7.716095914215426e-05, "loss": 0.0545, "step": 70270 }, { "epoch": 2.3760100070996315, "grad_norm": 0.9096542596817017, "learning_rate": 7.715401757964184e-05, "loss": 0.0618, "step": 70280 }, { "epoch": 2.3763480847898846, "grad_norm": 0.36981743574142456, "learning_rate": 7.714707527474882e-05, "loss": 0.0612, "step": 70290 }, { "epoch": 2.376686162480138, "grad_norm": 0.7179379463195801, "learning_rate": 7.714013222766494e-05, "loss": 0.0535, "step": 70300 }, { "epoch": 2.377024240170391, "grad_norm": 0.4759906828403473, "learning_rate": 7.713318843858006e-05, "loss": 0.0649, "step": 70310 }, { "epoch": 2.3773623178606442, "grad_norm": 0.48463520407676697, "learning_rate": 7.712624390768397e-05, "loss": 0.0576, "step": 70320 }, { "epoch": 2.3777003955508977, "grad_norm": 0.9949231743812561, "learning_rate": 7.711929863516658e-05, "loss": 0.0675, "step": 70330 }, { "epoch": 2.378038473241151, "grad_norm": 0.5801993608474731, "learning_rate": 7.711235262121774e-05, "loss": 0.0596, "step": 70340 }, { "epoch": 2.378376550931404, "grad_norm": 0.9339672327041626, "learning_rate": 7.710540586602739e-05, "loss": 0.0596, "step": 70350 }, { "epoch": 2.3787146286216574, "grad_norm": 0.8642628192901611, "learning_rate": 7.709845836978542e-05, "loss": 0.0611, "step": 70360 }, { "epoch": 2.3790527063119105, "grad_norm": 0.8735095262527466, "learning_rate": 7.709151013268177e-05, "loss": 0.0537, "step": 70370 }, { "epoch": 2.3793907840021635, "grad_norm": 0.5548751354217529, "learning_rate": 7.70845611549064e-05, "loss": 0.0652, "step": 70380 }, { "epoch": 2.379728861692417, "grad_norm": 0.9597299098968506, "learning_rate": 7.707761143664933e-05, "loss": 0.0597, "step": 70390 }, { "epoch": 2.38006693938267, "grad_norm": 0.5381062030792236, "learning_rate": 7.707066097810052e-05, "loss": 0.0617, "step": 70400 }, { "epoch": 2.380405017072923, "grad_norm": 0.5657317638397217, "learning_rate": 7.706370977945e-05, "loss": 0.0644, "step": 70410 }, { "epoch": 2.3807430947631767, "grad_norm": 0.5387363433837891, "learning_rate": 7.705675784088784e-05, "loss": 0.0633, "step": 70420 }, { "epoch": 2.3810811724534298, "grad_norm": 3.7187647819519043, "learning_rate": 7.704980516260407e-05, "loss": 0.0634, "step": 70430 }, { "epoch": 2.381419250143683, "grad_norm": 0.47831690311431885, "learning_rate": 7.70428517447888e-05, "loss": 0.0617, "step": 70440 }, { "epoch": 2.3817573278339363, "grad_norm": 0.5083218216896057, "learning_rate": 7.703589758763212e-05, "loss": 0.0491, "step": 70450 }, { "epoch": 2.3820954055241894, "grad_norm": 0.7438653707504272, "learning_rate": 7.702894269132417e-05, "loss": 0.0547, "step": 70460 }, { "epoch": 2.3824334832144425, "grad_norm": 0.532710611820221, "learning_rate": 7.702198705605506e-05, "loss": 0.0653, "step": 70470 }, { "epoch": 2.382771560904696, "grad_norm": 0.9352258443832397, "learning_rate": 7.701503068201499e-05, "loss": 0.0683, "step": 70480 }, { "epoch": 2.383109638594949, "grad_norm": 1.752407431602478, "learning_rate": 7.700807356939411e-05, "loss": 0.0578, "step": 70490 }, { "epoch": 2.383447716285202, "grad_norm": 0.9425315856933594, "learning_rate": 7.700111571838266e-05, "loss": 0.06, "step": 70500 }, { "epoch": 2.3837857939754556, "grad_norm": 0.37636542320251465, "learning_rate": 7.699415712917087e-05, "loss": 0.0589, "step": 70510 }, { "epoch": 2.3841238716657087, "grad_norm": 0.3936779797077179, "learning_rate": 7.698719780194892e-05, "loss": 0.0552, "step": 70520 }, { "epoch": 2.384461949355962, "grad_norm": 0.32140955328941345, "learning_rate": 7.698023773690714e-05, "loss": 0.0614, "step": 70530 }, { "epoch": 2.3848000270462153, "grad_norm": 0.8341812491416931, "learning_rate": 7.697327693423581e-05, "loss": 0.0587, "step": 70540 }, { "epoch": 2.3851381047364684, "grad_norm": 0.37331515550613403, "learning_rate": 7.696631539412523e-05, "loss": 0.055, "step": 70550 }, { "epoch": 2.3854761824267214, "grad_norm": 0.5268539786338806, "learning_rate": 7.695935311676568e-05, "loss": 0.0597, "step": 70560 }, { "epoch": 2.385814260116975, "grad_norm": 0.6112158894538879, "learning_rate": 7.695239010234758e-05, "loss": 0.0573, "step": 70570 }, { "epoch": 2.386152337807228, "grad_norm": 0.3634938597679138, "learning_rate": 7.694542635106123e-05, "loss": 0.0607, "step": 70580 }, { "epoch": 2.386490415497481, "grad_norm": 0.6995195150375366, "learning_rate": 7.693846186309707e-05, "loss": 0.0649, "step": 70590 }, { "epoch": 2.3868284931877346, "grad_norm": 1.255185842514038, "learning_rate": 7.693149663864548e-05, "loss": 0.0568, "step": 70600 }, { "epoch": 2.3871665708779877, "grad_norm": 0.4375886023044586, "learning_rate": 7.692453067789687e-05, "loss": 0.0613, "step": 70610 }, { "epoch": 2.3875046485682407, "grad_norm": 1.7050892114639282, "learning_rate": 7.691756398104173e-05, "loss": 0.0569, "step": 70620 }, { "epoch": 2.3878427262584943, "grad_norm": 0.46516603231430054, "learning_rate": 7.691059654827049e-05, "loss": 0.0603, "step": 70630 }, { "epoch": 2.3881808039487473, "grad_norm": 1.1796588897705078, "learning_rate": 7.690362837977365e-05, "loss": 0.0603, "step": 70640 }, { "epoch": 2.388518881639001, "grad_norm": 0.8948869705200195, "learning_rate": 7.689665947574173e-05, "loss": 0.0642, "step": 70650 }, { "epoch": 2.388856959329254, "grad_norm": 2.616274356842041, "learning_rate": 7.688968983636522e-05, "loss": 0.0615, "step": 70660 }, { "epoch": 2.389195037019507, "grad_norm": 0.43442583084106445, "learning_rate": 7.688271946183471e-05, "loss": 0.0654, "step": 70670 }, { "epoch": 2.3895331147097605, "grad_norm": 1.2512867450714111, "learning_rate": 7.687574835234075e-05, "loss": 0.0616, "step": 70680 }, { "epoch": 2.3898711924000136, "grad_norm": 0.49663054943084717, "learning_rate": 7.686877650807394e-05, "loss": 0.0636, "step": 70690 }, { "epoch": 2.3902092700902666, "grad_norm": 0.5412122011184692, "learning_rate": 7.686180392922485e-05, "loss": 0.062, "step": 70700 }, { "epoch": 2.39054734778052, "grad_norm": 0.5323091745376587, "learning_rate": 7.685483061598413e-05, "loss": 0.0648, "step": 70710 }, { "epoch": 2.390885425470773, "grad_norm": 0.7421901226043701, "learning_rate": 7.684785656854246e-05, "loss": 0.0629, "step": 70720 }, { "epoch": 2.3912235031610263, "grad_norm": 0.7093856334686279, "learning_rate": 7.684088178709045e-05, "loss": 0.0556, "step": 70730 }, { "epoch": 2.39156158085128, "grad_norm": 0.5615379214286804, "learning_rate": 7.683390627181881e-05, "loss": 0.0686, "step": 70740 }, { "epoch": 2.391899658541533, "grad_norm": 0.6090462803840637, "learning_rate": 7.682693002291827e-05, "loss": 0.0587, "step": 70750 }, { "epoch": 2.392237736231786, "grad_norm": 0.31339648365974426, "learning_rate": 7.681995304057956e-05, "loss": 0.0582, "step": 70760 }, { "epoch": 2.3925758139220394, "grad_norm": 1.0683895349502563, "learning_rate": 7.681297532499337e-05, "loss": 0.0585, "step": 70770 }, { "epoch": 2.3929138916122925, "grad_norm": 1.0211560726165771, "learning_rate": 7.680599687635055e-05, "loss": 0.061, "step": 70780 }, { "epoch": 2.3932519693025456, "grad_norm": 0.4219883680343628, "learning_rate": 7.679901769484179e-05, "loss": 0.0623, "step": 70790 }, { "epoch": 2.393590046992799, "grad_norm": 0.42624378204345703, "learning_rate": 7.6792037780658e-05, "loss": 0.0608, "step": 70800 }, { "epoch": 2.393928124683052, "grad_norm": 0.6905614137649536, "learning_rate": 7.678505713398994e-05, "loss": 0.0625, "step": 70810 }, { "epoch": 2.3942662023733052, "grad_norm": 1.7854679822921753, "learning_rate": 7.67780757550285e-05, "loss": 0.0546, "step": 70820 }, { "epoch": 2.3946042800635587, "grad_norm": 0.5969488620758057, "learning_rate": 7.67710936439645e-05, "loss": 0.0554, "step": 70830 }, { "epoch": 2.394942357753812, "grad_norm": 0.5557528734207153, "learning_rate": 7.67641108009889e-05, "loss": 0.0569, "step": 70840 }, { "epoch": 2.395280435444065, "grad_norm": 0.5748619437217712, "learning_rate": 7.675712722629253e-05, "loss": 0.0615, "step": 70850 }, { "epoch": 2.3956185131343184, "grad_norm": 0.41095104813575745, "learning_rate": 7.675014292006637e-05, "loss": 0.0595, "step": 70860 }, { "epoch": 2.3959565908245715, "grad_norm": 0.781376302242279, "learning_rate": 7.674315788250134e-05, "loss": 0.0577, "step": 70870 }, { "epoch": 2.3962946685148245, "grad_norm": 0.6057633757591248, "learning_rate": 7.673617211378842e-05, "loss": 0.0589, "step": 70880 }, { "epoch": 2.396632746205078, "grad_norm": 0.6997587084770203, "learning_rate": 7.672918561411862e-05, "loss": 0.0575, "step": 70890 }, { "epoch": 2.396970823895331, "grad_norm": 0.4821854531764984, "learning_rate": 7.672219838368291e-05, "loss": 0.0612, "step": 70900 }, { "epoch": 2.397308901585584, "grad_norm": 0.4960592985153198, "learning_rate": 7.671521042267233e-05, "loss": 0.0611, "step": 70910 }, { "epoch": 2.3976469792758377, "grad_norm": 0.46221452951431274, "learning_rate": 7.670822173127794e-05, "loss": 0.0612, "step": 70920 }, { "epoch": 2.3979850569660908, "grad_norm": 0.43757563829421997, "learning_rate": 7.670123230969081e-05, "loss": 0.0576, "step": 70930 }, { "epoch": 2.398323134656344, "grad_norm": 0.39241230487823486, "learning_rate": 7.669424215810201e-05, "loss": 0.0605, "step": 70940 }, { "epoch": 2.3986612123465973, "grad_norm": 0.45869091153144836, "learning_rate": 7.668725127670267e-05, "loss": 0.0602, "step": 70950 }, { "epoch": 2.3989992900368504, "grad_norm": 0.7234908938407898, "learning_rate": 7.668025966568388e-05, "loss": 0.0594, "step": 70960 }, { "epoch": 2.399337367727104, "grad_norm": 1.0143812894821167, "learning_rate": 7.667326732523685e-05, "loss": 0.0552, "step": 70970 }, { "epoch": 2.399675445417357, "grad_norm": 0.4309951961040497, "learning_rate": 7.666627425555269e-05, "loss": 0.0577, "step": 70980 }, { "epoch": 2.40001352310761, "grad_norm": 0.49414628744125366, "learning_rate": 7.66592804568226e-05, "loss": 0.0518, "step": 70990 }, { "epoch": 2.4003516007978636, "grad_norm": 1.5363879203796387, "learning_rate": 7.665228592923782e-05, "loss": 0.0569, "step": 71000 }, { "epoch": 2.4006896784881167, "grad_norm": 0.7280935049057007, "learning_rate": 7.664529067298954e-05, "loss": 0.0579, "step": 71010 }, { "epoch": 2.4010277561783697, "grad_norm": 0.5101234912872314, "learning_rate": 7.663829468826905e-05, "loss": 0.0605, "step": 71020 }, { "epoch": 2.4013658338686232, "grad_norm": 0.7792870998382568, "learning_rate": 7.663129797526755e-05, "loss": 0.0584, "step": 71030 }, { "epoch": 2.4017039115588763, "grad_norm": 0.5619886517524719, "learning_rate": 7.662430053417638e-05, "loss": 0.0613, "step": 71040 }, { "epoch": 2.4020419892491294, "grad_norm": 0.6947982311248779, "learning_rate": 7.661730236518685e-05, "loss": 0.0559, "step": 71050 }, { "epoch": 2.402380066939383, "grad_norm": 0.5385177731513977, "learning_rate": 7.661030346849027e-05, "loss": 0.0668, "step": 71060 }, { "epoch": 2.402718144629636, "grad_norm": 0.9882221817970276, "learning_rate": 7.6603303844278e-05, "loss": 0.0659, "step": 71070 }, { "epoch": 2.403056222319889, "grad_norm": 0.5716262459754944, "learning_rate": 7.659630349274138e-05, "loss": 0.0595, "step": 71080 }, { "epoch": 2.4033943000101425, "grad_norm": 0.3501792252063751, "learning_rate": 7.658930241407183e-05, "loss": 0.0606, "step": 71090 }, { "epoch": 2.4037323777003956, "grad_norm": 0.3905206620693207, "learning_rate": 7.658230060846072e-05, "loss": 0.0625, "step": 71100 }, { "epoch": 2.4040704553906487, "grad_norm": 0.45863839983940125, "learning_rate": 7.65752980760995e-05, "loss": 0.0616, "step": 71110 }, { "epoch": 2.404408533080902, "grad_norm": 0.4152370095252991, "learning_rate": 7.656829481717962e-05, "loss": 0.0581, "step": 71120 }, { "epoch": 2.4047466107711553, "grad_norm": 0.5184576511383057, "learning_rate": 7.656129083189255e-05, "loss": 0.0548, "step": 71130 }, { "epoch": 2.4050846884614083, "grad_norm": 0.6199729442596436, "learning_rate": 7.655428612042976e-05, "loss": 0.0574, "step": 71140 }, { "epoch": 2.405422766151662, "grad_norm": 0.9452703595161438, "learning_rate": 7.654728068298277e-05, "loss": 0.0569, "step": 71150 }, { "epoch": 2.405760843841915, "grad_norm": 0.269502729177475, "learning_rate": 7.654027451974309e-05, "loss": 0.0543, "step": 71160 }, { "epoch": 2.406098921532168, "grad_norm": 0.6066786646842957, "learning_rate": 7.653326763090227e-05, "loss": 0.0615, "step": 71170 }, { "epoch": 2.4064369992224215, "grad_norm": 0.8076184988021851, "learning_rate": 7.652626001665189e-05, "loss": 0.0635, "step": 71180 }, { "epoch": 2.4067750769126746, "grad_norm": 0.3675038516521454, "learning_rate": 7.651925167718353e-05, "loss": 0.0578, "step": 71190 }, { "epoch": 2.4071131546029276, "grad_norm": 0.3735675811767578, "learning_rate": 7.651224261268878e-05, "loss": 0.056, "step": 71200 }, { "epoch": 2.407451232293181, "grad_norm": 0.5275079011917114, "learning_rate": 7.650523282335929e-05, "loss": 0.0597, "step": 71210 }, { "epoch": 2.407789309983434, "grad_norm": 0.4646707773208618, "learning_rate": 7.64982223093867e-05, "loss": 0.0531, "step": 71220 }, { "epoch": 2.4081273876736873, "grad_norm": 0.5918118357658386, "learning_rate": 7.649121107096264e-05, "loss": 0.0597, "step": 71230 }, { "epoch": 2.408465465363941, "grad_norm": 0.4308556616306305, "learning_rate": 7.648419910827884e-05, "loss": 0.0582, "step": 71240 }, { "epoch": 2.408803543054194, "grad_norm": 0.8444586396217346, "learning_rate": 7.647718642152698e-05, "loss": 0.0616, "step": 71250 }, { "epoch": 2.409141620744447, "grad_norm": 0.4950779974460602, "learning_rate": 7.64701730108988e-05, "loss": 0.0498, "step": 71260 }, { "epoch": 2.4094796984347004, "grad_norm": 0.8978794813156128, "learning_rate": 7.646315887658603e-05, "loss": 0.055, "step": 71270 }, { "epoch": 2.4098177761249535, "grad_norm": 0.26807960867881775, "learning_rate": 7.645614401878044e-05, "loss": 0.0544, "step": 71280 }, { "epoch": 2.4101558538152066, "grad_norm": 0.37775805592536926, "learning_rate": 7.64491284376738e-05, "loss": 0.0622, "step": 71290 }, { "epoch": 2.41049393150546, "grad_norm": 0.3981759548187256, "learning_rate": 7.644211213345794e-05, "loss": 0.0561, "step": 71300 }, { "epoch": 2.410832009195713, "grad_norm": 0.5744835734367371, "learning_rate": 7.643509510632465e-05, "loss": 0.0604, "step": 71310 }, { "epoch": 2.4111700868859662, "grad_norm": 0.5279887914657593, "learning_rate": 7.64280773564658e-05, "loss": 0.0554, "step": 71320 }, { "epoch": 2.4115081645762197, "grad_norm": 0.9337408542633057, "learning_rate": 7.642105888407326e-05, "loss": 0.0574, "step": 71330 }, { "epoch": 2.411846242266473, "grad_norm": 0.5475181937217712, "learning_rate": 7.641403968933887e-05, "loss": 0.058, "step": 71340 }, { "epoch": 2.412184319956726, "grad_norm": 0.6537285447120667, "learning_rate": 7.640701977245456e-05, "loss": 0.0606, "step": 71350 }, { "epoch": 2.4125223976469794, "grad_norm": 0.5162078738212585, "learning_rate": 7.639999913361227e-05, "loss": 0.0583, "step": 71360 }, { "epoch": 2.4128604753372325, "grad_norm": 0.5159935355186462, "learning_rate": 7.639297777300389e-05, "loss": 0.0637, "step": 71370 }, { "epoch": 2.4131985530274855, "grad_norm": 0.4714280068874359, "learning_rate": 7.638595569082145e-05, "loss": 0.065, "step": 71380 }, { "epoch": 2.413536630717739, "grad_norm": 0.35619503259658813, "learning_rate": 7.637893288725688e-05, "loss": 0.0566, "step": 71390 }, { "epoch": 2.413874708407992, "grad_norm": 0.5512820482254028, "learning_rate": 7.637190936250219e-05, "loss": 0.0596, "step": 71400 }, { "epoch": 2.414212786098245, "grad_norm": 0.48216351866722107, "learning_rate": 7.63648851167494e-05, "loss": 0.0583, "step": 71410 }, { "epoch": 2.4145508637884987, "grad_norm": 0.46295103430747986, "learning_rate": 7.635786015019058e-05, "loss": 0.0543, "step": 71420 }, { "epoch": 2.4148889414787518, "grad_norm": 0.5034775733947754, "learning_rate": 7.635083446301774e-05, "loss": 0.054, "step": 71430 }, { "epoch": 2.415227019169005, "grad_norm": 0.5768901705741882, "learning_rate": 7.6343808055423e-05, "loss": 0.0544, "step": 71440 }, { "epoch": 2.4155650968592584, "grad_norm": 0.4461383521556854, "learning_rate": 7.633678092759843e-05, "loss": 0.0597, "step": 71450 }, { "epoch": 2.4159031745495114, "grad_norm": 0.3477180004119873, "learning_rate": 7.632975307973618e-05, "loss": 0.0588, "step": 71460 }, { "epoch": 2.4162412522397645, "grad_norm": 0.34820884466171265, "learning_rate": 7.632272451202837e-05, "loss": 0.0623, "step": 71470 }, { "epoch": 2.416579329930018, "grad_norm": 0.5604353547096252, "learning_rate": 7.631569522466715e-05, "loss": 0.0591, "step": 71480 }, { "epoch": 2.416917407620271, "grad_norm": 0.713837742805481, "learning_rate": 7.63086652178447e-05, "loss": 0.0563, "step": 71490 }, { "epoch": 2.417255485310524, "grad_norm": 0.5489259958267212, "learning_rate": 7.630163449175325e-05, "loss": 0.058, "step": 71500 }, { "epoch": 2.4175935630007777, "grad_norm": 0.5400086641311646, "learning_rate": 7.6294603046585e-05, "loss": 0.059, "step": 71510 }, { "epoch": 2.4179316406910307, "grad_norm": 0.6083783507347107, "learning_rate": 7.628757088253216e-05, "loss": 0.0578, "step": 71520 }, { "epoch": 2.418269718381284, "grad_norm": 0.6083508133888245, "learning_rate": 7.628053799978702e-05, "loss": 0.0579, "step": 71530 }, { "epoch": 2.4186077960715373, "grad_norm": 0.3693058490753174, "learning_rate": 7.627350439854183e-05, "loss": 0.0548, "step": 71540 }, { "epoch": 2.4189458737617904, "grad_norm": 0.48211199045181274, "learning_rate": 7.626647007898891e-05, "loss": 0.0551, "step": 71550 }, { "epoch": 2.4192839514520434, "grad_norm": 0.43872663378715515, "learning_rate": 7.625943504132056e-05, "loss": 0.0573, "step": 71560 }, { "epoch": 2.419622029142297, "grad_norm": 0.3238127529621124, "learning_rate": 7.625239928572914e-05, "loss": 0.0557, "step": 71570 }, { "epoch": 2.41996010683255, "grad_norm": 0.9124661684036255, "learning_rate": 7.624536281240698e-05, "loss": 0.0639, "step": 71580 }, { "epoch": 2.420298184522803, "grad_norm": 0.41233325004577637, "learning_rate": 7.623832562154645e-05, "loss": 0.0614, "step": 71590 }, { "epoch": 2.4206362622130566, "grad_norm": 0.5252602696418762, "learning_rate": 7.623128771333995e-05, "loss": 0.0653, "step": 71600 }, { "epoch": 2.4209743399033097, "grad_norm": 1.0342284440994263, "learning_rate": 7.622424908797992e-05, "loss": 0.0567, "step": 71610 }, { "epoch": 2.421312417593563, "grad_norm": 0.45643284916877747, "learning_rate": 7.621720974565874e-05, "loss": 0.0579, "step": 71620 }, { "epoch": 2.4216504952838163, "grad_norm": 0.3772644102573395, "learning_rate": 7.621016968656892e-05, "loss": 0.0588, "step": 71630 }, { "epoch": 2.4219885729740693, "grad_norm": 2.287966251373291, "learning_rate": 7.62031289109029e-05, "loss": 0.0592, "step": 71640 }, { "epoch": 2.422326650664323, "grad_norm": 0.7436572909355164, "learning_rate": 7.619608741885318e-05, "loss": 0.0571, "step": 71650 }, { "epoch": 2.422664728354576, "grad_norm": 1.451687216758728, "learning_rate": 7.618904521061227e-05, "loss": 0.0633, "step": 71660 }, { "epoch": 2.423002806044829, "grad_norm": 0.5372000336647034, "learning_rate": 7.618200228637271e-05, "loss": 0.0543, "step": 71670 }, { "epoch": 2.4233408837350825, "grad_norm": 0.8186752200126648, "learning_rate": 7.617495864632703e-05, "loss": 0.0595, "step": 71680 }, { "epoch": 2.4236789614253356, "grad_norm": 0.470565527677536, "learning_rate": 7.616791429066782e-05, "loss": 0.0561, "step": 71690 }, { "epoch": 2.4240170391155886, "grad_norm": 0.6950828433036804, "learning_rate": 7.616086921958767e-05, "loss": 0.0564, "step": 71700 }, { "epoch": 2.424355116805842, "grad_norm": 0.871040940284729, "learning_rate": 7.615382343327918e-05, "loss": 0.0562, "step": 71710 }, { "epoch": 2.424693194496095, "grad_norm": 0.7926183342933655, "learning_rate": 7.614677693193499e-05, "loss": 0.059, "step": 71720 }, { "epoch": 2.4250312721863483, "grad_norm": 0.9177138209342957, "learning_rate": 7.613972971574773e-05, "loss": 0.0551, "step": 71730 }, { "epoch": 2.425369349876602, "grad_norm": 2.2889535427093506, "learning_rate": 7.613268178491009e-05, "loss": 0.0552, "step": 71740 }, { "epoch": 2.425707427566855, "grad_norm": 0.38445866107940674, "learning_rate": 7.612563313961473e-05, "loss": 0.0567, "step": 71750 }, { "epoch": 2.426045505257108, "grad_norm": 0.6153709292411804, "learning_rate": 7.61185837800544e-05, "loss": 0.0563, "step": 71760 }, { "epoch": 2.4263835829473615, "grad_norm": 0.7080597281455994, "learning_rate": 7.611153370642178e-05, "loss": 0.0584, "step": 71770 }, { "epoch": 2.4267216606376145, "grad_norm": 0.40304499864578247, "learning_rate": 7.610448291890966e-05, "loss": 0.0655, "step": 71780 }, { "epoch": 2.4270597383278676, "grad_norm": 0.5868215560913086, "learning_rate": 7.609743141771076e-05, "loss": 0.0573, "step": 71790 }, { "epoch": 2.427397816018121, "grad_norm": 0.8968347311019897, "learning_rate": 7.609037920301792e-05, "loss": 0.058, "step": 71800 }, { "epoch": 2.427735893708374, "grad_norm": 0.5402724146842957, "learning_rate": 7.608332627502387e-05, "loss": 0.0555, "step": 71810 }, { "epoch": 2.4280739713986272, "grad_norm": 0.43820443749427795, "learning_rate": 7.607627263392151e-05, "loss": 0.0606, "step": 71820 }, { "epoch": 2.4284120490888808, "grad_norm": 0.8554500937461853, "learning_rate": 7.606921827990364e-05, "loss": 0.0615, "step": 71830 }, { "epoch": 2.428750126779134, "grad_norm": 0.6340339183807373, "learning_rate": 7.606216321316315e-05, "loss": 0.0606, "step": 71840 }, { "epoch": 2.429088204469387, "grad_norm": 0.47463926672935486, "learning_rate": 7.605510743389291e-05, "loss": 0.0671, "step": 71850 }, { "epoch": 2.4294262821596404, "grad_norm": 0.35705944895744324, "learning_rate": 7.60480509422858e-05, "loss": 0.0603, "step": 71860 }, { "epoch": 2.4297643598498935, "grad_norm": 0.5447651743888855, "learning_rate": 7.604099373853478e-05, "loss": 0.0538, "step": 71870 }, { "epoch": 2.4301024375401465, "grad_norm": 0.6102834939956665, "learning_rate": 7.603393582283275e-05, "loss": 0.0575, "step": 71880 }, { "epoch": 2.4304405152304, "grad_norm": 0.40114012360572815, "learning_rate": 7.602687719537273e-05, "loss": 0.0587, "step": 71890 }, { "epoch": 2.430778592920653, "grad_norm": 0.6645365357398987, "learning_rate": 7.601981785634765e-05, "loss": 0.0594, "step": 71900 }, { "epoch": 2.431116670610906, "grad_norm": 1.0562338829040527, "learning_rate": 7.601275780595052e-05, "loss": 0.0563, "step": 71910 }, { "epoch": 2.4314547483011597, "grad_norm": 0.34487301111221313, "learning_rate": 7.600569704437437e-05, "loss": 0.0601, "step": 71920 }, { "epoch": 2.4317928259914128, "grad_norm": 0.755765438079834, "learning_rate": 7.599863557181222e-05, "loss": 0.0569, "step": 71930 }, { "epoch": 2.4321309036816663, "grad_norm": 0.49912548065185547, "learning_rate": 7.599157338845717e-05, "loss": 0.0573, "step": 71940 }, { "epoch": 2.4324689813719194, "grad_norm": 0.49698150157928467, "learning_rate": 7.598451049450223e-05, "loss": 0.062, "step": 71950 }, { "epoch": 2.4328070590621724, "grad_norm": 0.45522356033325195, "learning_rate": 7.597744689014057e-05, "loss": 0.0557, "step": 71960 }, { "epoch": 2.433145136752426, "grad_norm": 0.4932296574115753, "learning_rate": 7.597038257556525e-05, "loss": 0.0607, "step": 71970 }, { "epoch": 2.433483214442679, "grad_norm": 0.8914423584938049, "learning_rate": 7.596331755096944e-05, "loss": 0.0627, "step": 71980 }, { "epoch": 2.433821292132932, "grad_norm": 0.6506635546684265, "learning_rate": 7.595625181654628e-05, "loss": 0.0608, "step": 71990 }, { "epoch": 2.4341593698231856, "grad_norm": 0.5168605446815491, "learning_rate": 7.594918537248893e-05, "loss": 0.0655, "step": 72000 }, { "epoch": 2.4344974475134387, "grad_norm": 0.5796452760696411, "learning_rate": 7.594211821899063e-05, "loss": 0.0569, "step": 72010 }, { "epoch": 2.4348355252036917, "grad_norm": 1.7296059131622314, "learning_rate": 7.593505035624453e-05, "loss": 0.0603, "step": 72020 }, { "epoch": 2.4351736028939452, "grad_norm": 0.47264227271080017, "learning_rate": 7.592798178444392e-05, "loss": 0.0574, "step": 72030 }, { "epoch": 2.4355116805841983, "grad_norm": 0.8723598122596741, "learning_rate": 7.592091250378201e-05, "loss": 0.0547, "step": 72040 }, { "epoch": 2.4358497582744514, "grad_norm": 0.6170424818992615, "learning_rate": 7.591384251445213e-05, "loss": 0.064, "step": 72050 }, { "epoch": 2.436187835964705, "grad_norm": 0.4693480432033539, "learning_rate": 7.590677181664749e-05, "loss": 0.0594, "step": 72060 }, { "epoch": 2.436525913654958, "grad_norm": 0.9486886262893677, "learning_rate": 7.589970041056144e-05, "loss": 0.0532, "step": 72070 }, { "epoch": 2.436863991345211, "grad_norm": 0.5590530037879944, "learning_rate": 7.589262829638734e-05, "loss": 0.0666, "step": 72080 }, { "epoch": 2.4372020690354645, "grad_norm": 0.8994885683059692, "learning_rate": 7.58855554743185e-05, "loss": 0.0561, "step": 72090 }, { "epoch": 2.4375401467257176, "grad_norm": 0.3744420111179352, "learning_rate": 7.587848194454828e-05, "loss": 0.0581, "step": 72100 }, { "epoch": 2.4378782244159707, "grad_norm": 0.3749316334724426, "learning_rate": 7.587140770727012e-05, "loss": 0.0678, "step": 72110 }, { "epoch": 2.438216302106224, "grad_norm": 0.7799550294876099, "learning_rate": 7.586433276267738e-05, "loss": 0.064, "step": 72120 }, { "epoch": 2.4385543797964773, "grad_norm": 1.1332528591156006, "learning_rate": 7.585725711096348e-05, "loss": 0.0637, "step": 72130 }, { "epoch": 2.4388924574867303, "grad_norm": 0.5123032927513123, "learning_rate": 7.585018075232188e-05, "loss": 0.0528, "step": 72140 }, { "epoch": 2.439230535176984, "grad_norm": 0.6158207654953003, "learning_rate": 7.584310368694607e-05, "loss": 0.0607, "step": 72150 }, { "epoch": 2.439568612867237, "grad_norm": 0.3983602523803711, "learning_rate": 7.583602591502952e-05, "loss": 0.058, "step": 72160 }, { "epoch": 2.43990669055749, "grad_norm": 0.3852970004081726, "learning_rate": 7.58289474367657e-05, "loss": 0.0607, "step": 72170 }, { "epoch": 2.4402447682477435, "grad_norm": 1.1239631175994873, "learning_rate": 7.582186825234818e-05, "loss": 0.0581, "step": 72180 }, { "epoch": 2.4405828459379966, "grad_norm": 1.0095055103302002, "learning_rate": 7.581478836197049e-05, "loss": 0.0628, "step": 72190 }, { "epoch": 2.4409209236282496, "grad_norm": 0.46011579036712646, "learning_rate": 7.580770776582616e-05, "loss": 0.064, "step": 72200 }, { "epoch": 2.441259001318503, "grad_norm": 0.4702374041080475, "learning_rate": 7.580062646410882e-05, "loss": 0.0573, "step": 72210 }, { "epoch": 2.441597079008756, "grad_norm": 0.8712277412414551, "learning_rate": 7.579354445701204e-05, "loss": 0.0567, "step": 72220 }, { "epoch": 2.4419351566990093, "grad_norm": 0.5536755323410034, "learning_rate": 7.578646174472943e-05, "loss": 0.0575, "step": 72230 }, { "epoch": 2.442273234389263, "grad_norm": 0.3485574424266815, "learning_rate": 7.577937832745466e-05, "loss": 0.0607, "step": 72240 }, { "epoch": 2.442611312079516, "grad_norm": 0.826938807964325, "learning_rate": 7.577229420538135e-05, "loss": 0.0617, "step": 72250 }, { "epoch": 2.442949389769769, "grad_norm": 0.5497678518295288, "learning_rate": 7.576520937870322e-05, "loss": 0.0602, "step": 72260 }, { "epoch": 2.4432874674600225, "grad_norm": 0.6475471258163452, "learning_rate": 7.575812384761394e-05, "loss": 0.0589, "step": 72270 }, { "epoch": 2.4436255451502755, "grad_norm": 0.5375613570213318, "learning_rate": 7.575103761230724e-05, "loss": 0.0556, "step": 72280 }, { "epoch": 2.4439636228405286, "grad_norm": 0.6424856185913086, "learning_rate": 7.574395067297683e-05, "loss": 0.0552, "step": 72290 }, { "epoch": 2.444301700530782, "grad_norm": 0.39724719524383545, "learning_rate": 7.573686302981648e-05, "loss": 0.0603, "step": 72300 }, { "epoch": 2.444639778221035, "grad_norm": 0.5612545013427734, "learning_rate": 7.572977468301997e-05, "loss": 0.0574, "step": 72310 }, { "epoch": 2.4449778559112882, "grad_norm": 0.34143874049186707, "learning_rate": 7.572268563278109e-05, "loss": 0.0622, "step": 72320 }, { "epoch": 2.4453159336015418, "grad_norm": 0.48234736919403076, "learning_rate": 7.571559587929364e-05, "loss": 0.0597, "step": 72330 }, { "epoch": 2.445654011291795, "grad_norm": 0.3131656348705292, "learning_rate": 7.570850542275146e-05, "loss": 0.0566, "step": 72340 }, { "epoch": 2.445992088982048, "grad_norm": 0.8536360859870911, "learning_rate": 7.57014142633484e-05, "loss": 0.0537, "step": 72350 }, { "epoch": 2.4463301666723014, "grad_norm": 0.7747331261634827, "learning_rate": 7.569432240127833e-05, "loss": 0.0594, "step": 72360 }, { "epoch": 2.4466682443625545, "grad_norm": 0.6245951056480408, "learning_rate": 7.568722983673516e-05, "loss": 0.0578, "step": 72370 }, { "epoch": 2.4470063220528075, "grad_norm": 0.8325949311256409, "learning_rate": 7.568013656991275e-05, "loss": 0.0654, "step": 72380 }, { "epoch": 2.447344399743061, "grad_norm": 0.3695763051509857, "learning_rate": 7.567304260100505e-05, "loss": 0.0578, "step": 72390 }, { "epoch": 2.447682477433314, "grad_norm": 0.5702596306800842, "learning_rate": 7.566594793020602e-05, "loss": 0.0544, "step": 72400 }, { "epoch": 2.448020555123567, "grad_norm": 0.45583003759384155, "learning_rate": 7.565885255770964e-05, "loss": 0.0574, "step": 72410 }, { "epoch": 2.4483586328138207, "grad_norm": 0.8815855979919434, "learning_rate": 7.565175648370983e-05, "loss": 0.0628, "step": 72420 }, { "epoch": 2.448696710504074, "grad_norm": 0.9296838641166687, "learning_rate": 7.564465970840067e-05, "loss": 0.0578, "step": 72430 }, { "epoch": 2.449034788194327, "grad_norm": 1.498517632484436, "learning_rate": 7.563756223197613e-05, "loss": 0.0592, "step": 72440 }, { "epoch": 2.4493728658845804, "grad_norm": 0.5068128705024719, "learning_rate": 7.563046405463026e-05, "loss": 0.0575, "step": 72450 }, { "epoch": 2.4497109435748334, "grad_norm": 0.547368049621582, "learning_rate": 7.562336517655714e-05, "loss": 0.0633, "step": 72460 }, { "epoch": 2.4500490212650865, "grad_norm": 0.5029283165931702, "learning_rate": 7.561626559795085e-05, "loss": 0.0575, "step": 72470 }, { "epoch": 2.45038709895534, "grad_norm": 0.962081253528595, "learning_rate": 7.560916531900549e-05, "loss": 0.0552, "step": 72480 }, { "epoch": 2.450725176645593, "grad_norm": 0.6498042941093445, "learning_rate": 7.560206433991515e-05, "loss": 0.0577, "step": 72490 }, { "epoch": 2.451063254335846, "grad_norm": 0.5773380994796753, "learning_rate": 7.5594962660874e-05, "loss": 0.0556, "step": 72500 }, { "epoch": 2.4514013320260997, "grad_norm": 0.4557045102119446, "learning_rate": 7.558786028207617e-05, "loss": 0.0588, "step": 72510 }, { "epoch": 2.4517394097163527, "grad_norm": 1.0327483415603638, "learning_rate": 7.558075720371587e-05, "loss": 0.0583, "step": 72520 }, { "epoch": 2.452077487406606, "grad_norm": 0.5117234587669373, "learning_rate": 7.557365342598728e-05, "loss": 0.0597, "step": 72530 }, { "epoch": 2.4524155650968593, "grad_norm": 0.5206335186958313, "learning_rate": 7.556654894908461e-05, "loss": 0.0557, "step": 72540 }, { "epoch": 2.4527536427871124, "grad_norm": 1.060741901397705, "learning_rate": 7.55594437732021e-05, "loss": 0.0573, "step": 72550 }, { "epoch": 2.4530917204773655, "grad_norm": 0.5739669799804688, "learning_rate": 7.5552337898534e-05, "loss": 0.0579, "step": 72560 }, { "epoch": 2.453429798167619, "grad_norm": 0.6944803595542908, "learning_rate": 7.554523132527455e-05, "loss": 0.059, "step": 72570 }, { "epoch": 2.453767875857872, "grad_norm": 0.4910019338130951, "learning_rate": 7.553812405361811e-05, "loss": 0.0531, "step": 72580 }, { "epoch": 2.4541059535481256, "grad_norm": 0.357685923576355, "learning_rate": 7.553101608375894e-05, "loss": 0.0611, "step": 72590 }, { "epoch": 2.4544440312383786, "grad_norm": 0.615389347076416, "learning_rate": 7.552390741589137e-05, "loss": 0.0604, "step": 72600 }, { "epoch": 2.4547821089286317, "grad_norm": 1.0298727750778198, "learning_rate": 7.551679805020979e-05, "loss": 0.0583, "step": 72610 }, { "epoch": 2.455120186618885, "grad_norm": 0.6257211565971375, "learning_rate": 7.550968798690852e-05, "loss": 0.0615, "step": 72620 }, { "epoch": 2.4554582643091383, "grad_norm": 0.6694914102554321, "learning_rate": 7.550257722618194e-05, "loss": 0.0569, "step": 72630 }, { "epoch": 2.4557963419993913, "grad_norm": 0.4098210036754608, "learning_rate": 7.549546576822452e-05, "loss": 0.0605, "step": 72640 }, { "epoch": 2.456134419689645, "grad_norm": 0.6725311279296875, "learning_rate": 7.548835361323064e-05, "loss": 0.0566, "step": 72650 }, { "epoch": 2.456472497379898, "grad_norm": 0.9937531352043152, "learning_rate": 7.548124076139471e-05, "loss": 0.0562, "step": 72660 }, { "epoch": 2.456810575070151, "grad_norm": 0.527283251285553, "learning_rate": 7.547412721291126e-05, "loss": 0.0597, "step": 72670 }, { "epoch": 2.4571486527604045, "grad_norm": 0.566506028175354, "learning_rate": 7.546701296797474e-05, "loss": 0.0566, "step": 72680 }, { "epoch": 2.4574867304506576, "grad_norm": 0.9973638653755188, "learning_rate": 7.545989802677966e-05, "loss": 0.0623, "step": 72690 }, { "epoch": 2.4578248081409106, "grad_norm": 1.7703616619110107, "learning_rate": 7.545278238952052e-05, "loss": 0.063, "step": 72700 }, { "epoch": 2.458162885831164, "grad_norm": 0.5304450392723083, "learning_rate": 7.544566605639188e-05, "loss": 0.0519, "step": 72710 }, { "epoch": 2.4585009635214172, "grad_norm": 0.42501300573349, "learning_rate": 7.543854902758828e-05, "loss": 0.0611, "step": 72720 }, { "epoch": 2.4588390412116703, "grad_norm": 0.44812771677970886, "learning_rate": 7.54314313033043e-05, "loss": 0.0547, "step": 72730 }, { "epoch": 2.459177118901924, "grad_norm": 0.3135029375553131, "learning_rate": 7.542431288373457e-05, "loss": 0.0578, "step": 72740 }, { "epoch": 2.459515196592177, "grad_norm": 0.5304620265960693, "learning_rate": 7.541719376907366e-05, "loss": 0.0578, "step": 72750 }, { "epoch": 2.45985327428243, "grad_norm": 0.4433593153953552, "learning_rate": 7.541007395951622e-05, "loss": 0.0608, "step": 72760 }, { "epoch": 2.4601913519726835, "grad_norm": 1.0726990699768066, "learning_rate": 7.540295345525689e-05, "loss": 0.0494, "step": 72770 }, { "epoch": 2.4605294296629365, "grad_norm": 0.2999347150325775, "learning_rate": 7.539583225649039e-05, "loss": 0.0625, "step": 72780 }, { "epoch": 2.4608675073531896, "grad_norm": 0.49649953842163086, "learning_rate": 7.538871036341135e-05, "loss": 0.0568, "step": 72790 }, { "epoch": 2.461205585043443, "grad_norm": 0.5228790044784546, "learning_rate": 7.538158777621452e-05, "loss": 0.0595, "step": 72800 }, { "epoch": 2.461543662733696, "grad_norm": 0.6448682546615601, "learning_rate": 7.537446449509461e-05, "loss": 0.0547, "step": 72810 }, { "epoch": 2.4618817404239492, "grad_norm": 0.8104101419448853, "learning_rate": 7.536734052024638e-05, "loss": 0.0524, "step": 72820 }, { "epoch": 2.4622198181142028, "grad_norm": 1.0367763042449951, "learning_rate": 7.536021585186456e-05, "loss": 0.0546, "step": 72830 }, { "epoch": 2.462557895804456, "grad_norm": 0.3250342905521393, "learning_rate": 7.5353090490144e-05, "loss": 0.0613, "step": 72840 }, { "epoch": 2.462895973494709, "grad_norm": 0.7430943250656128, "learning_rate": 7.534596443527946e-05, "loss": 0.0559, "step": 72850 }, { "epoch": 2.4632340511849624, "grad_norm": 1.3881617784500122, "learning_rate": 7.53388376874658e-05, "loss": 0.0619, "step": 72860 }, { "epoch": 2.4635721288752155, "grad_norm": 0.5867689251899719, "learning_rate": 7.533171024689781e-05, "loss": 0.055, "step": 72870 }, { "epoch": 2.4639102065654686, "grad_norm": 0.5866740345954895, "learning_rate": 7.532458211377037e-05, "loss": 0.059, "step": 72880 }, { "epoch": 2.464248284255722, "grad_norm": 0.24000340700149536, "learning_rate": 7.531745328827837e-05, "loss": 0.0487, "step": 72890 }, { "epoch": 2.464586361945975, "grad_norm": 0.41611137986183167, "learning_rate": 7.531032377061673e-05, "loss": 0.0559, "step": 72900 }, { "epoch": 2.4649244396362286, "grad_norm": 0.8049718737602234, "learning_rate": 7.530319356098033e-05, "loss": 0.0596, "step": 72910 }, { "epoch": 2.4652625173264817, "grad_norm": 0.8288143873214722, "learning_rate": 7.529606265956412e-05, "loss": 0.0531, "step": 72920 }, { "epoch": 2.465600595016735, "grad_norm": 0.50758957862854, "learning_rate": 7.528893106656308e-05, "loss": 0.0588, "step": 72930 }, { "epoch": 2.4659386727069883, "grad_norm": 0.7700445652008057, "learning_rate": 7.528179878217216e-05, "loss": 0.0593, "step": 72940 }, { "epoch": 2.4662767503972414, "grad_norm": 0.5067596435546875, "learning_rate": 7.527466580658635e-05, "loss": 0.0579, "step": 72950 }, { "epoch": 2.4666148280874944, "grad_norm": 0.4752533733844757, "learning_rate": 7.526753214000068e-05, "loss": 0.0569, "step": 72960 }, { "epoch": 2.466952905777748, "grad_norm": 0.5518107414245605, "learning_rate": 7.526039778261018e-05, "loss": 0.0644, "step": 72970 }, { "epoch": 2.467290983468001, "grad_norm": 0.6177627444267273, "learning_rate": 7.525326273460986e-05, "loss": 0.0621, "step": 72980 }, { "epoch": 2.467629061158254, "grad_norm": 0.4379371404647827, "learning_rate": 7.524612699619487e-05, "loss": 0.0583, "step": 72990 }, { "epoch": 2.4679671388485076, "grad_norm": 0.5830535888671875, "learning_rate": 7.523899056756023e-05, "loss": 0.0562, "step": 73000 }, { "epoch": 2.4683052165387607, "grad_norm": 0.8729007244110107, "learning_rate": 7.523185344890107e-05, "loss": 0.0589, "step": 73010 }, { "epoch": 2.4686432942290137, "grad_norm": 0.441535621881485, "learning_rate": 7.522471564041251e-05, "loss": 0.0572, "step": 73020 }, { "epoch": 2.4689813719192673, "grad_norm": 0.4102623164653778, "learning_rate": 7.521757714228971e-05, "loss": 0.0658, "step": 73030 }, { "epoch": 2.4693194496095203, "grad_norm": 0.9018625617027283, "learning_rate": 7.521043795472781e-05, "loss": 0.0592, "step": 73040 }, { "epoch": 2.4696575272997734, "grad_norm": 0.41025128960609436, "learning_rate": 7.520329807792204e-05, "loss": 0.0557, "step": 73050 }, { "epoch": 2.469995604990027, "grad_norm": 0.4789799749851227, "learning_rate": 7.519615751206755e-05, "loss": 0.0557, "step": 73060 }, { "epoch": 2.47033368268028, "grad_norm": 0.44244804978370667, "learning_rate": 7.518901625735957e-05, "loss": 0.0569, "step": 73070 }, { "epoch": 2.470671760370533, "grad_norm": 1.1279162168502808, "learning_rate": 7.518187431399335e-05, "loss": 0.0538, "step": 73080 }, { "epoch": 2.4710098380607866, "grad_norm": 0.4725927710533142, "learning_rate": 7.517473168216415e-05, "loss": 0.0584, "step": 73090 }, { "epoch": 2.4713479157510396, "grad_norm": 0.5373665690422058, "learning_rate": 7.516758836206726e-05, "loss": 0.06, "step": 73100 }, { "epoch": 2.4716859934412927, "grad_norm": 0.39337703585624695, "learning_rate": 7.516044435389795e-05, "loss": 0.057, "step": 73110 }, { "epoch": 2.472024071131546, "grad_norm": 0.8741632103919983, "learning_rate": 7.515329965785154e-05, "loss": 0.0577, "step": 73120 }, { "epoch": 2.4723621488217993, "grad_norm": 0.6700582504272461, "learning_rate": 7.514615427412336e-05, "loss": 0.0581, "step": 73130 }, { "epoch": 2.4727002265120523, "grad_norm": 0.3568941652774811, "learning_rate": 7.513900820290877e-05, "loss": 0.0578, "step": 73140 }, { "epoch": 2.473038304202306, "grad_norm": 0.549213707447052, "learning_rate": 7.513186144440315e-05, "loss": 0.0563, "step": 73150 }, { "epoch": 2.473376381892559, "grad_norm": 0.4920866787433624, "learning_rate": 7.512471399880188e-05, "loss": 0.0592, "step": 73160 }, { "epoch": 2.473714459582812, "grad_norm": 0.37631285190582275, "learning_rate": 7.511756586630036e-05, "loss": 0.0586, "step": 73170 }, { "epoch": 2.4740525372730655, "grad_norm": 0.5299208164215088, "learning_rate": 7.511041704709403e-05, "loss": 0.0624, "step": 73180 }, { "epoch": 2.4743906149633186, "grad_norm": 0.7342093586921692, "learning_rate": 7.510326754137834e-05, "loss": 0.0591, "step": 73190 }, { "epoch": 2.4747286926535716, "grad_norm": 0.44718989729881287, "learning_rate": 7.509611734934873e-05, "loss": 0.0624, "step": 73200 }, { "epoch": 2.475066770343825, "grad_norm": 0.5476096868515015, "learning_rate": 7.50889664712007e-05, "loss": 0.058, "step": 73210 }, { "epoch": 2.4754048480340782, "grad_norm": 0.5104821920394897, "learning_rate": 7.508181490712975e-05, "loss": 0.0627, "step": 73220 }, { "epoch": 2.4757429257243313, "grad_norm": 0.5796894431114197, "learning_rate": 7.507466265733143e-05, "loss": 0.0598, "step": 73230 }, { "epoch": 2.476081003414585, "grad_norm": 0.570681631565094, "learning_rate": 7.506750972200123e-05, "loss": 0.0525, "step": 73240 }, { "epoch": 2.476419081104838, "grad_norm": 0.4211031496524811, "learning_rate": 7.506035610133474e-05, "loss": 0.0541, "step": 73250 }, { "epoch": 2.476757158795091, "grad_norm": 0.6596526503562927, "learning_rate": 7.505320179552753e-05, "loss": 0.0565, "step": 73260 }, { "epoch": 2.4770952364853445, "grad_norm": 0.7660709619522095, "learning_rate": 7.504604680477518e-05, "loss": 0.0644, "step": 73270 }, { "epoch": 2.4774333141755975, "grad_norm": 0.7578754425048828, "learning_rate": 7.503889112927334e-05, "loss": 0.0639, "step": 73280 }, { "epoch": 2.4777713918658506, "grad_norm": 0.5238327383995056, "learning_rate": 7.50317347692176e-05, "loss": 0.0626, "step": 73290 }, { "epoch": 2.478109469556104, "grad_norm": 0.48214879631996155, "learning_rate": 7.502457772480367e-05, "loss": 0.0549, "step": 73300 }, { "epoch": 2.478447547246357, "grad_norm": 0.5628460645675659, "learning_rate": 7.501741999622716e-05, "loss": 0.0517, "step": 73310 }, { "epoch": 2.4787856249366103, "grad_norm": 0.7672939896583557, "learning_rate": 7.501026158368379e-05, "loss": 0.0706, "step": 73320 }, { "epoch": 2.4791237026268638, "grad_norm": 0.28748077154159546, "learning_rate": 7.500310248736926e-05, "loss": 0.0546, "step": 73330 }, { "epoch": 2.479461780317117, "grad_norm": 1.395861029624939, "learning_rate": 7.499594270747932e-05, "loss": 0.0608, "step": 73340 }, { "epoch": 2.47979985800737, "grad_norm": 0.6604917049407959, "learning_rate": 7.498878224420967e-05, "loss": 0.0613, "step": 73350 }, { "epoch": 2.4801379356976234, "grad_norm": 0.49955475330352783, "learning_rate": 7.498162109775613e-05, "loss": 0.0616, "step": 73360 }, { "epoch": 2.4804760133878765, "grad_norm": 0.5532594919204712, "learning_rate": 7.497445926831444e-05, "loss": 0.0581, "step": 73370 }, { "epoch": 2.4808140910781296, "grad_norm": 0.39290711283683777, "learning_rate": 7.49672967560804e-05, "loss": 0.0549, "step": 73380 }, { "epoch": 2.481152168768383, "grad_norm": 0.3797379732131958, "learning_rate": 7.496013356124987e-05, "loss": 0.0574, "step": 73390 }, { "epoch": 2.481490246458636, "grad_norm": 0.5443952679634094, "learning_rate": 7.495296968401865e-05, "loss": 0.0594, "step": 73400 }, { "epoch": 2.481828324148889, "grad_norm": 2.2946534156799316, "learning_rate": 7.494580512458263e-05, "loss": 0.059, "step": 73410 }, { "epoch": 2.4821664018391427, "grad_norm": 1.0532208681106567, "learning_rate": 7.493863988313765e-05, "loss": 0.0574, "step": 73420 }, { "epoch": 2.482504479529396, "grad_norm": 0.4483238458633423, "learning_rate": 7.493147395987964e-05, "loss": 0.06, "step": 73430 }, { "epoch": 2.482842557219649, "grad_norm": 0.6950351595878601, "learning_rate": 7.492430735500449e-05, "loss": 0.0555, "step": 73440 }, { "epoch": 2.4831806349099024, "grad_norm": 0.3697393536567688, "learning_rate": 7.491714006870815e-05, "loss": 0.0577, "step": 73450 }, { "epoch": 2.4835187126001554, "grad_norm": 1.0280555486679077, "learning_rate": 7.490997210118655e-05, "loss": 0.0551, "step": 73460 }, { "epoch": 2.4838567902904085, "grad_norm": 0.8477169275283813, "learning_rate": 7.490280345263568e-05, "loss": 0.0603, "step": 73470 }, { "epoch": 2.484194867980662, "grad_norm": 0.3539755046367645, "learning_rate": 7.489563412325151e-05, "loss": 0.0573, "step": 73480 }, { "epoch": 2.484532945670915, "grad_norm": 0.5268971920013428, "learning_rate": 7.488846411323007e-05, "loss": 0.0542, "step": 73490 }, { "epoch": 2.484871023361168, "grad_norm": 0.5684064626693726, "learning_rate": 7.488129342276736e-05, "loss": 0.0605, "step": 73500 }, { "epoch": 2.4852091010514217, "grad_norm": 0.4009830057621002, "learning_rate": 7.487412205205943e-05, "loss": 0.0538, "step": 73510 }, { "epoch": 2.4855471787416747, "grad_norm": 0.4412812292575836, "learning_rate": 7.486695000130237e-05, "loss": 0.0617, "step": 73520 }, { "epoch": 2.485885256431928, "grad_norm": 0.48574599623680115, "learning_rate": 7.485977727069221e-05, "loss": 0.0628, "step": 73530 }, { "epoch": 2.4862233341221813, "grad_norm": 0.9055690765380859, "learning_rate": 7.485260386042509e-05, "loss": 0.0593, "step": 73540 }, { "epoch": 2.4865614118124344, "grad_norm": 0.7586053013801575, "learning_rate": 7.484542977069713e-05, "loss": 0.0597, "step": 73550 }, { "epoch": 2.486899489502688, "grad_norm": 1.0267404317855835, "learning_rate": 7.483825500170445e-05, "loss": 0.0588, "step": 73560 }, { "epoch": 2.487237567192941, "grad_norm": 0.5748680830001831, "learning_rate": 7.48310795536432e-05, "loss": 0.0518, "step": 73570 }, { "epoch": 2.487575644883194, "grad_norm": 0.5566640496253967, "learning_rate": 7.482390342670958e-05, "loss": 0.0533, "step": 73580 }, { "epoch": 2.4879137225734476, "grad_norm": 0.46989259123802185, "learning_rate": 7.481672662109974e-05, "loss": 0.0554, "step": 73590 }, { "epoch": 2.4882518002637006, "grad_norm": 0.42716675996780396, "learning_rate": 7.480954913700992e-05, "loss": 0.059, "step": 73600 }, { "epoch": 2.4885898779539537, "grad_norm": 0.9775055646896362, "learning_rate": 7.480237097463637e-05, "loss": 0.0601, "step": 73610 }, { "epoch": 2.488927955644207, "grad_norm": 1.5423239469528198, "learning_rate": 7.47951921341753e-05, "loss": 0.0567, "step": 73620 }, { "epoch": 2.4892660333344603, "grad_norm": 1.651768445968628, "learning_rate": 7.4788012615823e-05, "loss": 0.0624, "step": 73630 }, { "epoch": 2.4896041110247134, "grad_norm": 1.0910837650299072, "learning_rate": 7.478083241977575e-05, "loss": 0.0685, "step": 73640 }, { "epoch": 2.489942188714967, "grad_norm": 2.074331760406494, "learning_rate": 7.477365154622983e-05, "loss": 0.0717, "step": 73650 }, { "epoch": 2.49028026640522, "grad_norm": 1.347781777381897, "learning_rate": 7.476646999538161e-05, "loss": 0.0647, "step": 73660 }, { "epoch": 2.490618344095473, "grad_norm": 1.0833771228790283, "learning_rate": 7.47592877674274e-05, "loss": 0.0629, "step": 73670 }, { "epoch": 2.4909564217857265, "grad_norm": 0.5404566526412964, "learning_rate": 7.475210486256356e-05, "loss": 0.0696, "step": 73680 }, { "epoch": 2.4912944994759796, "grad_norm": 0.6509020328521729, "learning_rate": 7.474492128098646e-05, "loss": 0.0576, "step": 73690 }, { "epoch": 2.4916325771662327, "grad_norm": 0.7778757810592651, "learning_rate": 7.473773702289253e-05, "loss": 0.0559, "step": 73700 }, { "epoch": 2.491970654856486, "grad_norm": 0.49022120237350464, "learning_rate": 7.473055208847815e-05, "loss": 0.0642, "step": 73710 }, { "epoch": 2.4923087325467392, "grad_norm": 0.5917701721191406, "learning_rate": 7.472336647793976e-05, "loss": 0.06, "step": 73720 }, { "epoch": 2.4926468102369923, "grad_norm": 0.40537476539611816, "learning_rate": 7.471618019147381e-05, "loss": 0.0603, "step": 73730 }, { "epoch": 2.492984887927246, "grad_norm": 0.28353917598724365, "learning_rate": 7.470899322927681e-05, "loss": 0.0559, "step": 73740 }, { "epoch": 2.493322965617499, "grad_norm": 0.597540557384491, "learning_rate": 7.47018055915452e-05, "loss": 0.061, "step": 73750 }, { "epoch": 2.493661043307752, "grad_norm": 0.5011661648750305, "learning_rate": 7.469461727847553e-05, "loss": 0.0595, "step": 73760 }, { "epoch": 2.4939991209980055, "grad_norm": 1.4028156995773315, "learning_rate": 7.468742829026425e-05, "loss": 0.0594, "step": 73770 }, { "epoch": 2.4943371986882585, "grad_norm": 1.3327820301055908, "learning_rate": 7.4680238627108e-05, "loss": 0.0618, "step": 73780 }, { "epoch": 2.4946752763785116, "grad_norm": 1.149761438369751, "learning_rate": 7.467304828920328e-05, "loss": 0.0642, "step": 73790 }, { "epoch": 2.495013354068765, "grad_norm": 1.0573060512542725, "learning_rate": 7.466585727674669e-05, "loss": 0.0619, "step": 73800 }, { "epoch": 2.495351431759018, "grad_norm": 0.6210426092147827, "learning_rate": 7.465866558993483e-05, "loss": 0.063, "step": 73810 }, { "epoch": 2.4956895094492713, "grad_norm": 0.5400168299674988, "learning_rate": 7.46514732289643e-05, "loss": 0.0592, "step": 73820 }, { "epoch": 2.4960275871395248, "grad_norm": 0.7076559662818909, "learning_rate": 7.46442801940318e-05, "loss": 0.0617, "step": 73830 }, { "epoch": 2.496365664829778, "grad_norm": 0.8186512589454651, "learning_rate": 7.463708648533389e-05, "loss": 0.0539, "step": 73840 }, { "epoch": 2.496703742520031, "grad_norm": 0.5357182025909424, "learning_rate": 7.462989210306731e-05, "loss": 0.062, "step": 73850 }, { "epoch": 2.4970418202102844, "grad_norm": 1.2646684646606445, "learning_rate": 7.462269704742872e-05, "loss": 0.0643, "step": 73860 }, { "epoch": 2.4973798979005375, "grad_norm": 0.8732843399047852, "learning_rate": 7.461550131861488e-05, "loss": 0.0752, "step": 73870 }, { "epoch": 2.497717975590791, "grad_norm": 1.2702689170837402, "learning_rate": 7.460830491682245e-05, "loss": 0.0801, "step": 73880 }, { "epoch": 2.498056053281044, "grad_norm": 1.4605907201766968, "learning_rate": 7.460110784224822e-05, "loss": 0.0815, "step": 73890 }, { "epoch": 2.498394130971297, "grad_norm": 1.7768795490264893, "learning_rate": 7.459391009508894e-05, "loss": 0.0681, "step": 73900 }, { "epoch": 2.4987322086615507, "grad_norm": 1.1626189947128296, "learning_rate": 7.458671167554138e-05, "loss": 0.0686, "step": 73910 }, { "epoch": 2.4990702863518037, "grad_norm": 1.0882290601730347, "learning_rate": 7.457951258380238e-05, "loss": 0.074, "step": 73920 }, { "epoch": 2.499408364042057, "grad_norm": 1.2823684215545654, "learning_rate": 7.457231282006875e-05, "loss": 0.0657, "step": 73930 }, { "epoch": 2.4997464417323103, "grad_norm": 1.0923858880996704, "learning_rate": 7.456511238453729e-05, "loss": 0.063, "step": 73940 }, { "epoch": 2.5000845194225634, "grad_norm": 1.9678354263305664, "learning_rate": 7.45579112774049e-05, "loss": 0.0641, "step": 73950 }, { "epoch": 2.5004225971128164, "grad_norm": 1.1971722841262817, "learning_rate": 7.455070949886845e-05, "loss": 0.0588, "step": 73960 }, { "epoch": 2.50076067480307, "grad_norm": 1.5513026714324951, "learning_rate": 7.454350704912482e-05, "loss": 0.0656, "step": 73970 }, { "epoch": 2.501098752493323, "grad_norm": 0.962022602558136, "learning_rate": 7.453630392837092e-05, "loss": 0.0668, "step": 73980 }, { "epoch": 2.501436830183576, "grad_norm": 2.3643946647644043, "learning_rate": 7.452910013680369e-05, "loss": 0.0659, "step": 73990 }, { "epoch": 2.5017749078738296, "grad_norm": 0.8415169715881348, "learning_rate": 7.452189567462008e-05, "loss": 0.0673, "step": 74000 }, { "epoch": 2.5021129855640827, "grad_norm": 1.6555083990097046, "learning_rate": 7.451469054201708e-05, "loss": 0.0662, "step": 74010 }, { "epoch": 2.5024510632543358, "grad_norm": 3.0832417011260986, "learning_rate": 7.450748473919161e-05, "loss": 0.0683, "step": 74020 }, { "epoch": 2.5027891409445893, "grad_norm": 1.950346827507019, "learning_rate": 7.450027826634074e-05, "loss": 0.0751, "step": 74030 }, { "epoch": 2.5031272186348423, "grad_norm": 2.034736156463623, "learning_rate": 7.449307112366145e-05, "loss": 0.0694, "step": 74040 }, { "epoch": 2.5034652963250954, "grad_norm": 2.144394636154175, "learning_rate": 7.448586331135081e-05, "loss": 0.0651, "step": 74050 }, { "epoch": 2.503803374015349, "grad_norm": 5.136597633361816, "learning_rate": 7.447865482960585e-05, "loss": 0.0688, "step": 74060 }, { "epoch": 2.504141451705602, "grad_norm": 0.9959430694580078, "learning_rate": 7.447144567862367e-05, "loss": 0.0704, "step": 74070 }, { "epoch": 2.504479529395855, "grad_norm": 1.6375471353530884, "learning_rate": 7.446423585860138e-05, "loss": 0.0689, "step": 74080 }, { "epoch": 2.5048176070861086, "grad_norm": 2.184041738510132, "learning_rate": 7.445702536973604e-05, "loss": 0.069, "step": 74090 }, { "epoch": 2.5051556847763616, "grad_norm": 8.214812278747559, "learning_rate": 7.444981421222483e-05, "loss": 0.0752, "step": 74100 }, { "epoch": 2.5054937624666147, "grad_norm": 1.4884878396987915, "learning_rate": 7.44426023862649e-05, "loss": 0.0699, "step": 74110 }, { "epoch": 2.505831840156868, "grad_norm": 1.764352798461914, "learning_rate": 7.443538989205339e-05, "loss": 0.071, "step": 74120 }, { "epoch": 2.5061699178471213, "grad_norm": 1.5295737981796265, "learning_rate": 7.442817672978749e-05, "loss": 0.0671, "step": 74130 }, { "epoch": 2.5065079955373744, "grad_norm": 1.7500265836715698, "learning_rate": 7.442096289966443e-05, "loss": 0.0634, "step": 74140 }, { "epoch": 2.506846073227628, "grad_norm": 1.4442976713180542, "learning_rate": 7.441374840188141e-05, "loss": 0.0731, "step": 74150 }, { "epoch": 2.507184150917881, "grad_norm": 4.1318583488464355, "learning_rate": 7.440653323663568e-05, "loss": 0.0699, "step": 74160 }, { "epoch": 2.507522228608134, "grad_norm": 0.6636614203453064, "learning_rate": 7.439931740412449e-05, "loss": 0.06, "step": 74170 }, { "epoch": 2.5078603062983875, "grad_norm": 0.6285256147384644, "learning_rate": 7.439210090454517e-05, "loss": 0.0688, "step": 74180 }, { "epoch": 2.5081983839886406, "grad_norm": 0.6817486882209778, "learning_rate": 7.438488373809494e-05, "loss": 0.0623, "step": 74190 }, { "epoch": 2.5085364616788937, "grad_norm": 0.9488841891288757, "learning_rate": 7.437766590497115e-05, "loss": 0.0629, "step": 74200 }, { "epoch": 2.508874539369147, "grad_norm": 1.27613365650177, "learning_rate": 7.437044740537114e-05, "loss": 0.057, "step": 74210 }, { "epoch": 2.5092126170594002, "grad_norm": 0.7293279767036438, "learning_rate": 7.436322823949225e-05, "loss": 0.0624, "step": 74220 }, { "epoch": 2.5095506947496533, "grad_norm": 1.629823088645935, "learning_rate": 7.435600840753185e-05, "loss": 0.0594, "step": 74230 }, { "epoch": 2.509888772439907, "grad_norm": 0.5570058822631836, "learning_rate": 7.434878790968735e-05, "loss": 0.0618, "step": 74240 }, { "epoch": 2.51022685013016, "grad_norm": 0.9904621839523315, "learning_rate": 7.434156674615611e-05, "loss": 0.0649, "step": 74250 }, { "epoch": 2.510564927820413, "grad_norm": 1.1299989223480225, "learning_rate": 7.433434491713561e-05, "loss": 0.0595, "step": 74260 }, { "epoch": 2.5109030055106665, "grad_norm": 1.5607342720031738, "learning_rate": 7.432712242282323e-05, "loss": 0.0709, "step": 74270 }, { "epoch": 2.5112410832009195, "grad_norm": 2.0162978172302246, "learning_rate": 7.431989926341648e-05, "loss": 0.063, "step": 74280 }, { "epoch": 2.5115791608911726, "grad_norm": 1.568433403968811, "learning_rate": 7.43126754391128e-05, "loss": 0.0714, "step": 74290 }, { "epoch": 2.511917238581426, "grad_norm": 1.989469289779663, "learning_rate": 7.430545095010973e-05, "loss": 0.0608, "step": 74300 }, { "epoch": 2.512255316271679, "grad_norm": 1.7771862745285034, "learning_rate": 7.429822579660476e-05, "loss": 0.0621, "step": 74310 }, { "epoch": 2.5125933939619323, "grad_norm": 1.4893709421157837, "learning_rate": 7.42909999787954e-05, "loss": 0.0641, "step": 74320 }, { "epoch": 2.5129314716521858, "grad_norm": 1.9013149738311768, "learning_rate": 7.428377349687927e-05, "loss": 0.0608, "step": 74330 }, { "epoch": 2.513269549342439, "grad_norm": 2.5331199169158936, "learning_rate": 7.427654635105387e-05, "loss": 0.0622, "step": 74340 }, { "epoch": 2.513607627032692, "grad_norm": 1.051651954650879, "learning_rate": 7.42693185415168e-05, "loss": 0.0601, "step": 74350 }, { "epoch": 2.5139457047229454, "grad_norm": 0.822921872138977, "learning_rate": 7.42620900684657e-05, "loss": 0.058, "step": 74360 }, { "epoch": 2.5142837824131985, "grad_norm": 2.158716917037964, "learning_rate": 7.425486093209817e-05, "loss": 0.062, "step": 74370 }, { "epoch": 2.5146218601034516, "grad_norm": 6.463254928588867, "learning_rate": 7.424763113261185e-05, "loss": 0.0611, "step": 74380 }, { "epoch": 2.514959937793705, "grad_norm": 2.0407536029815674, "learning_rate": 7.424040067020442e-05, "loss": 0.0703, "step": 74390 }, { "epoch": 2.515298015483958, "grad_norm": 0.9195684194564819, "learning_rate": 7.423316954507352e-05, "loss": 0.0618, "step": 74400 }, { "epoch": 2.515636093174211, "grad_norm": 1.6853877305984497, "learning_rate": 7.422593775741689e-05, "loss": 0.0762, "step": 74410 }, { "epoch": 2.5159741708644647, "grad_norm": 1.8672981262207031, "learning_rate": 7.421870530743218e-05, "loss": 0.0694, "step": 74420 }, { "epoch": 2.516312248554718, "grad_norm": 1.0266737937927246, "learning_rate": 7.421147219531721e-05, "loss": 0.064, "step": 74430 }, { "epoch": 2.516650326244971, "grad_norm": 1.2198679447174072, "learning_rate": 7.420423842126966e-05, "loss": 0.0674, "step": 74440 }, { "epoch": 2.5169884039352244, "grad_norm": 0.7035315036773682, "learning_rate": 7.419700398548735e-05, "loss": 0.0673, "step": 74450 }, { "epoch": 2.5173264816254775, "grad_norm": 1.1670678853988647, "learning_rate": 7.418976888816802e-05, "loss": 0.0662, "step": 74460 }, { "epoch": 2.5176645593157305, "grad_norm": 2.7899980545043945, "learning_rate": 7.418253312950951e-05, "loss": 0.0634, "step": 74470 }, { "epoch": 2.518002637005984, "grad_norm": 0.9658267498016357, "learning_rate": 7.41752967097096e-05, "loss": 0.064, "step": 74480 }, { "epoch": 2.518340714696237, "grad_norm": 0.8110589385032654, "learning_rate": 7.41680596289662e-05, "loss": 0.0616, "step": 74490 }, { "epoch": 2.51867879238649, "grad_norm": 1.6473616361618042, "learning_rate": 7.416082188747711e-05, "loss": 0.0567, "step": 74500 }, { "epoch": 2.5190168700767437, "grad_norm": 0.8889740109443665, "learning_rate": 7.415358348544023e-05, "loss": 0.0622, "step": 74510 }, { "epoch": 2.5193549477669968, "grad_norm": 1.1681740283966064, "learning_rate": 7.414634442305345e-05, "loss": 0.0609, "step": 74520 }, { "epoch": 2.51969302545725, "grad_norm": 1.1201727390289307, "learning_rate": 7.413910470051469e-05, "loss": 0.058, "step": 74530 }, { "epoch": 2.5200311031475033, "grad_norm": 2.031996488571167, "learning_rate": 7.413186431802186e-05, "loss": 0.0635, "step": 74540 }, { "epoch": 2.5203691808377564, "grad_norm": 0.9991660118103027, "learning_rate": 7.412462327577294e-05, "loss": 0.0669, "step": 74550 }, { "epoch": 2.5207072585280095, "grad_norm": 0.5568290948867798, "learning_rate": 7.411738157396587e-05, "loss": 0.0613, "step": 74560 }, { "epoch": 2.521045336218263, "grad_norm": 0.5915117859840393, "learning_rate": 7.411013921279866e-05, "loss": 0.0619, "step": 74570 }, { "epoch": 2.521383413908516, "grad_norm": 0.6153187155723572, "learning_rate": 7.41028961924693e-05, "loss": 0.0631, "step": 74580 }, { "epoch": 2.521721491598769, "grad_norm": 2.4938762187957764, "learning_rate": 7.40956525131758e-05, "loss": 0.067, "step": 74590 }, { "epoch": 2.5220595692890226, "grad_norm": 0.7132799029350281, "learning_rate": 7.408840817511624e-05, "loss": 0.0573, "step": 74600 }, { "epoch": 2.5223976469792757, "grad_norm": 0.3911531865596771, "learning_rate": 7.408116317848862e-05, "loss": 0.062, "step": 74610 }, { "epoch": 2.5227357246695292, "grad_norm": 0.6362413763999939, "learning_rate": 7.407391752349107e-05, "loss": 0.057, "step": 74620 }, { "epoch": 2.5230738023597823, "grad_norm": 0.5498132109642029, "learning_rate": 7.406667121032165e-05, "loss": 0.0606, "step": 74630 }, { "epoch": 2.5234118800500354, "grad_norm": 0.4344049096107483, "learning_rate": 7.405942423917849e-05, "loss": 0.058, "step": 74640 }, { "epoch": 2.523749957740289, "grad_norm": 0.4941018223762512, "learning_rate": 7.405217661025969e-05, "loss": 0.0604, "step": 74650 }, { "epoch": 2.524088035430542, "grad_norm": 0.7168843746185303, "learning_rate": 7.404492832376343e-05, "loss": 0.0617, "step": 74660 }, { "epoch": 2.524426113120795, "grad_norm": 0.6571817994117737, "learning_rate": 7.403767937988785e-05, "loss": 0.0602, "step": 74670 }, { "epoch": 2.5247641908110485, "grad_norm": 0.6047806739807129, "learning_rate": 7.403042977883117e-05, "loss": 0.0634, "step": 74680 }, { "epoch": 2.5251022685013016, "grad_norm": 1.421478271484375, "learning_rate": 7.402317952079154e-05, "loss": 0.0632, "step": 74690 }, { "epoch": 2.5254403461915547, "grad_norm": 0.5349745750427246, "learning_rate": 7.401592860596723e-05, "loss": 0.0568, "step": 74700 }, { "epoch": 2.525778423881808, "grad_norm": 0.8604224324226379, "learning_rate": 7.400867703455644e-05, "loss": 0.0521, "step": 74710 }, { "epoch": 2.5261165015720612, "grad_norm": 0.5501282215118408, "learning_rate": 7.400142480675746e-05, "loss": 0.0619, "step": 74720 }, { "epoch": 2.5264545792623143, "grad_norm": 0.43877559900283813, "learning_rate": 7.399417192276851e-05, "loss": 0.0593, "step": 74730 }, { "epoch": 2.526792656952568, "grad_norm": 0.5552314519882202, "learning_rate": 7.398691838278794e-05, "loss": 0.0597, "step": 74740 }, { "epoch": 2.527130734642821, "grad_norm": 0.378418505191803, "learning_rate": 7.397966418701401e-05, "loss": 0.059, "step": 74750 }, { "epoch": 2.5274688123330744, "grad_norm": 0.5774987936019897, "learning_rate": 7.39724093356451e-05, "loss": 0.0615, "step": 74760 }, { "epoch": 2.5278068900233275, "grad_norm": 0.4788055121898651, "learning_rate": 7.396515382887951e-05, "loss": 0.0584, "step": 74770 }, { "epoch": 2.5281449677135805, "grad_norm": 0.3635983467102051, "learning_rate": 7.395789766691561e-05, "loss": 0.0644, "step": 74780 }, { "epoch": 2.528483045403834, "grad_norm": 0.37707632780075073, "learning_rate": 7.395064084995181e-05, "loss": 0.0604, "step": 74790 }, { "epoch": 2.528821123094087, "grad_norm": 0.3362261652946472, "learning_rate": 7.394338337818646e-05, "loss": 0.0593, "step": 74800 }, { "epoch": 2.52915920078434, "grad_norm": 0.6794439554214478, "learning_rate": 7.393612525181801e-05, "loss": 0.0634, "step": 74810 }, { "epoch": 2.5294972784745937, "grad_norm": 0.5340309739112854, "learning_rate": 7.392886647104488e-05, "loss": 0.051, "step": 74820 }, { "epoch": 2.529835356164847, "grad_norm": 0.7767943739891052, "learning_rate": 7.392160703606556e-05, "loss": 0.0547, "step": 74830 }, { "epoch": 2.5301734338551, "grad_norm": 0.2537946403026581, "learning_rate": 7.391434694707847e-05, "loss": 0.0558, "step": 74840 }, { "epoch": 2.5305115115453534, "grad_norm": 0.5062519907951355, "learning_rate": 7.390708620428213e-05, "loss": 0.0542, "step": 74850 }, { "epoch": 2.5308495892356064, "grad_norm": 0.4771672785282135, "learning_rate": 7.389982480787501e-05, "loss": 0.061, "step": 74860 }, { "epoch": 2.5311876669258595, "grad_norm": 0.44159382581710815, "learning_rate": 7.38925627580557e-05, "loss": 0.0542, "step": 74870 }, { "epoch": 2.531525744616113, "grad_norm": 0.7100340723991394, "learning_rate": 7.388530005502266e-05, "loss": 0.0593, "step": 74880 }, { "epoch": 2.531863822306366, "grad_norm": 0.5906128883361816, "learning_rate": 7.387803669897452e-05, "loss": 0.0571, "step": 74890 }, { "epoch": 2.532201899996619, "grad_norm": 0.5406297445297241, "learning_rate": 7.387077269010981e-05, "loss": 0.0604, "step": 74900 }, { "epoch": 2.5325399776868727, "grad_norm": 0.5563213229179382, "learning_rate": 7.386350802862715e-05, "loss": 0.0564, "step": 74910 }, { "epoch": 2.5328780553771257, "grad_norm": 0.7352524399757385, "learning_rate": 7.385624271472513e-05, "loss": 0.06, "step": 74920 }, { "epoch": 2.533216133067379, "grad_norm": 0.462523877620697, "learning_rate": 7.384897674860241e-05, "loss": 0.0575, "step": 74930 }, { "epoch": 2.5335542107576323, "grad_norm": 0.5422052145004272, "learning_rate": 7.384171013045762e-05, "loss": 0.0557, "step": 74940 }, { "epoch": 2.5338922884478854, "grad_norm": 0.8715444803237915, "learning_rate": 7.383444286048944e-05, "loss": 0.0611, "step": 74950 }, { "epoch": 2.5342303661381385, "grad_norm": 0.9780318737030029, "learning_rate": 7.382717493889654e-05, "loss": 0.0563, "step": 74960 }, { "epoch": 2.534568443828392, "grad_norm": 0.3398556411266327, "learning_rate": 7.381990636587763e-05, "loss": 0.0499, "step": 74970 }, { "epoch": 2.534906521518645, "grad_norm": 0.7279512286186218, "learning_rate": 7.381263714163144e-05, "loss": 0.0555, "step": 74980 }, { "epoch": 2.535244599208898, "grad_norm": 0.4699952006340027, "learning_rate": 7.380536726635668e-05, "loss": 0.0603, "step": 74990 }, { "epoch": 2.5355826768991516, "grad_norm": 1.2641282081604004, "learning_rate": 7.379809674025212e-05, "loss": 0.0574, "step": 75000 }, { "epoch": 2.5359207545894047, "grad_norm": 1.5195231437683105, "learning_rate": 7.379082556351655e-05, "loss": 0.0488, "step": 75010 }, { "epoch": 2.5362588322796578, "grad_norm": 1.110005259513855, "learning_rate": 7.378355373634873e-05, "loss": 0.0644, "step": 75020 }, { "epoch": 2.5365969099699113, "grad_norm": 0.5150746703147888, "learning_rate": 7.37762812589475e-05, "loss": 0.0615, "step": 75030 }, { "epoch": 2.5369349876601643, "grad_norm": 0.6256832480430603, "learning_rate": 7.376900813151168e-05, "loss": 0.057, "step": 75040 }, { "epoch": 2.5372730653504174, "grad_norm": 0.7252966165542603, "learning_rate": 7.376173435424008e-05, "loss": 0.0569, "step": 75050 }, { "epoch": 2.537611143040671, "grad_norm": 0.7759541869163513, "learning_rate": 7.375445992733161e-05, "loss": 0.0623, "step": 75060 }, { "epoch": 2.537949220730924, "grad_norm": 0.4613436460494995, "learning_rate": 7.374718485098513e-05, "loss": 0.0529, "step": 75070 }, { "epoch": 2.538287298421177, "grad_norm": 1.0532565116882324, "learning_rate": 7.373990912539953e-05, "loss": 0.0551, "step": 75080 }, { "epoch": 2.5386253761114306, "grad_norm": 0.9621742963790894, "learning_rate": 7.373263275077374e-05, "loss": 0.0563, "step": 75090 }, { "epoch": 2.5389634538016836, "grad_norm": 0.4680982232093811, "learning_rate": 7.372535572730669e-05, "loss": 0.0668, "step": 75100 }, { "epoch": 2.5393015314919367, "grad_norm": 0.4614895284175873, "learning_rate": 7.371807805519732e-05, "loss": 0.0538, "step": 75110 }, { "epoch": 2.5396396091821902, "grad_norm": 0.3960627317428589, "learning_rate": 7.37107997346446e-05, "loss": 0.0575, "step": 75120 }, { "epoch": 2.5399776868724433, "grad_norm": 0.5501511693000793, "learning_rate": 7.370352076584753e-05, "loss": 0.0585, "step": 75130 }, { "epoch": 2.5403157645626964, "grad_norm": 1.075110912322998, "learning_rate": 7.369624114900511e-05, "loss": 0.0575, "step": 75140 }, { "epoch": 2.54065384225295, "grad_norm": 0.28099995851516724, "learning_rate": 7.368896088431635e-05, "loss": 0.0543, "step": 75150 }, { "epoch": 2.540991919943203, "grad_norm": 0.3772404193878174, "learning_rate": 7.368167997198031e-05, "loss": 0.0495, "step": 75160 }, { "epoch": 2.541329997633456, "grad_norm": 0.6056807041168213, "learning_rate": 7.367439841219603e-05, "loss": 0.0615, "step": 75170 }, { "epoch": 2.5416680753237095, "grad_norm": 0.46893060207366943, "learning_rate": 7.366711620516258e-05, "loss": 0.0598, "step": 75180 }, { "epoch": 2.5420061530139626, "grad_norm": 0.7697925567626953, "learning_rate": 7.365983335107908e-05, "loss": 0.0555, "step": 75190 }, { "epoch": 2.5423442307042157, "grad_norm": 0.6037213802337646, "learning_rate": 7.365254985014464e-05, "loss": 0.0583, "step": 75200 }, { "epoch": 2.542682308394469, "grad_norm": 1.7696352005004883, "learning_rate": 7.364526570255835e-05, "loss": 0.0566, "step": 75210 }, { "epoch": 2.5430203860847223, "grad_norm": 0.8807423710823059, "learning_rate": 7.36379809085194e-05, "loss": 0.0604, "step": 75220 }, { "epoch": 2.5433584637749753, "grad_norm": 0.4850938022136688, "learning_rate": 7.36306954682269e-05, "loss": 0.0564, "step": 75230 }, { "epoch": 2.543696541465229, "grad_norm": 0.5195220708847046, "learning_rate": 7.362340938188008e-05, "loss": 0.0568, "step": 75240 }, { "epoch": 2.544034619155482, "grad_norm": 0.8308311104774475, "learning_rate": 7.361612264967811e-05, "loss": 0.0629, "step": 75250 }, { "epoch": 2.544372696845735, "grad_norm": 0.8122499585151672, "learning_rate": 7.360883527182025e-05, "loss": 0.0602, "step": 75260 }, { "epoch": 2.5447107745359885, "grad_norm": 0.6154493093490601, "learning_rate": 7.360154724850568e-05, "loss": 0.0606, "step": 75270 }, { "epoch": 2.5450488522262416, "grad_norm": 1.2936290502548218, "learning_rate": 7.359425857993367e-05, "loss": 0.0591, "step": 75280 }, { "epoch": 2.5453869299164946, "grad_norm": 0.5146687030792236, "learning_rate": 7.358696926630351e-05, "loss": 0.0634, "step": 75290 }, { "epoch": 2.545725007606748, "grad_norm": 0.9612672328948975, "learning_rate": 7.357967930781446e-05, "loss": 0.0578, "step": 75300 }, { "epoch": 2.546063085297001, "grad_norm": 0.5207695364952087, "learning_rate": 7.357238870466585e-05, "loss": 0.063, "step": 75310 }, { "epoch": 2.5464011629872543, "grad_norm": 0.483169287443161, "learning_rate": 7.356509745705697e-05, "loss": 0.0666, "step": 75320 }, { "epoch": 2.546739240677508, "grad_norm": 0.49384069442749023, "learning_rate": 7.355780556518719e-05, "loss": 0.0617, "step": 75330 }, { "epoch": 2.547077318367761, "grad_norm": 2.8948683738708496, "learning_rate": 7.355051302925585e-05, "loss": 0.061, "step": 75340 }, { "epoch": 2.547415396058014, "grad_norm": 0.969551682472229, "learning_rate": 7.354321984946234e-05, "loss": 0.0684, "step": 75350 }, { "epoch": 2.5477534737482674, "grad_norm": 0.942054808139801, "learning_rate": 7.353592602600605e-05, "loss": 0.0612, "step": 75360 }, { "epoch": 2.5480915514385205, "grad_norm": 0.7615219950675964, "learning_rate": 7.352863155908636e-05, "loss": 0.0587, "step": 75370 }, { "epoch": 2.5484296291287736, "grad_norm": 0.7552902698516846, "learning_rate": 7.352133644890273e-05, "loss": 0.0647, "step": 75380 }, { "epoch": 2.548767706819027, "grad_norm": 0.6253749132156372, "learning_rate": 7.351404069565462e-05, "loss": 0.064, "step": 75390 }, { "epoch": 2.54910578450928, "grad_norm": 0.9284663796424866, "learning_rate": 7.350674429954146e-05, "loss": 0.0602, "step": 75400 }, { "epoch": 2.5494438621995332, "grad_norm": 1.2981423139572144, "learning_rate": 7.349944726076274e-05, "loss": 0.0675, "step": 75410 }, { "epoch": 2.5497819398897867, "grad_norm": 0.9621978998184204, "learning_rate": 7.349214957951795e-05, "loss": 0.0566, "step": 75420 }, { "epoch": 2.55012001758004, "grad_norm": 0.950251042842865, "learning_rate": 7.348485125600663e-05, "loss": 0.0541, "step": 75430 }, { "epoch": 2.550458095270293, "grad_norm": 0.8455981612205505, "learning_rate": 7.347755229042828e-05, "loss": 0.0629, "step": 75440 }, { "epoch": 2.5507961729605464, "grad_norm": 0.5898106694221497, "learning_rate": 7.347025268298249e-05, "loss": 0.0639, "step": 75450 }, { "epoch": 2.5511342506507995, "grad_norm": 0.4142121374607086, "learning_rate": 7.34629524338688e-05, "loss": 0.0532, "step": 75460 }, { "epoch": 2.5514723283410525, "grad_norm": 0.4141894578933716, "learning_rate": 7.345565154328682e-05, "loss": 0.0559, "step": 75470 }, { "epoch": 2.551810406031306, "grad_norm": 0.5458618998527527, "learning_rate": 7.344835001143611e-05, "loss": 0.0577, "step": 75480 }, { "epoch": 2.552148483721559, "grad_norm": 0.5194174647331238, "learning_rate": 7.344104783851634e-05, "loss": 0.0615, "step": 75490 }, { "epoch": 2.552486561411812, "grad_norm": 1.1008485555648804, "learning_rate": 7.343374502472713e-05, "loss": 0.0597, "step": 75500 }, { "epoch": 2.5528246391020657, "grad_norm": 1.640953779220581, "learning_rate": 7.342644157026813e-05, "loss": 0.06, "step": 75510 }, { "epoch": 2.5531627167923188, "grad_norm": 0.6904051899909973, "learning_rate": 7.341913747533901e-05, "loss": 0.0648, "step": 75520 }, { "epoch": 2.553500794482572, "grad_norm": 0.6032142043113708, "learning_rate": 7.341183274013948e-05, "loss": 0.0603, "step": 75530 }, { "epoch": 2.5538388721728253, "grad_norm": 0.6102293729782104, "learning_rate": 7.340452736486925e-05, "loss": 0.058, "step": 75540 }, { "epoch": 2.5541769498630784, "grad_norm": 0.8214420080184937, "learning_rate": 7.3397221349728e-05, "loss": 0.0643, "step": 75550 }, { "epoch": 2.5545150275533315, "grad_norm": 0.5173282623291016, "learning_rate": 7.338991469491554e-05, "loss": 0.0608, "step": 75560 }, { "epoch": 2.554853105243585, "grad_norm": 0.6781930923461914, "learning_rate": 7.338260740063159e-05, "loss": 0.0579, "step": 75570 }, { "epoch": 2.555191182933838, "grad_norm": 0.4383901059627533, "learning_rate": 7.337529946707593e-05, "loss": 0.0567, "step": 75580 }, { "epoch": 2.5555292606240916, "grad_norm": 0.5540868639945984, "learning_rate": 7.336799089444837e-05, "loss": 0.0605, "step": 75590 }, { "epoch": 2.5558673383143447, "grad_norm": 0.8496267795562744, "learning_rate": 7.336068168294873e-05, "loss": 0.0547, "step": 75600 }, { "epoch": 2.5562054160045977, "grad_norm": 0.4505930244922638, "learning_rate": 7.335337183277682e-05, "loss": 0.0602, "step": 75610 }, { "epoch": 2.5565434936948512, "grad_norm": 0.4628024697303772, "learning_rate": 7.334606134413249e-05, "loss": 0.0567, "step": 75620 }, { "epoch": 2.5568815713851043, "grad_norm": 0.38296806812286377, "learning_rate": 7.333875021721562e-05, "loss": 0.0621, "step": 75630 }, { "epoch": 2.5572196490753574, "grad_norm": 0.4181617796421051, "learning_rate": 7.333143845222607e-05, "loss": 0.0529, "step": 75640 }, { "epoch": 2.557557726765611, "grad_norm": 0.6921057105064392, "learning_rate": 7.332412604936378e-05, "loss": 0.058, "step": 75650 }, { "epoch": 2.557895804455864, "grad_norm": 0.5148507356643677, "learning_rate": 7.331681300882865e-05, "loss": 0.0593, "step": 75660 }, { "epoch": 2.558233882146117, "grad_norm": 0.7014822363853455, "learning_rate": 7.33094993308206e-05, "loss": 0.0568, "step": 75670 }, { "epoch": 2.5585719598363705, "grad_norm": 2.6628565788269043, "learning_rate": 7.33021850155396e-05, "loss": 0.0575, "step": 75680 }, { "epoch": 2.5589100375266236, "grad_norm": 0.7361537218093872, "learning_rate": 7.329487006318561e-05, "loss": 0.062, "step": 75690 }, { "epoch": 2.5592481152168767, "grad_norm": 1.1764682531356812, "learning_rate": 7.328755447395863e-05, "loss": 0.0624, "step": 75700 }, { "epoch": 2.55958619290713, "grad_norm": 0.3880343437194824, "learning_rate": 7.328023824805865e-05, "loss": 0.0587, "step": 75710 }, { "epoch": 2.5599242705973833, "grad_norm": 0.8073248267173767, "learning_rate": 7.327292138568571e-05, "loss": 0.0639, "step": 75720 }, { "epoch": 2.5602623482876368, "grad_norm": 0.5664547681808472, "learning_rate": 7.326560388703985e-05, "loss": 0.0599, "step": 75730 }, { "epoch": 2.56060042597789, "grad_norm": 0.5104001760482788, "learning_rate": 7.325828575232109e-05, "loss": 0.0588, "step": 75740 }, { "epoch": 2.560938503668143, "grad_norm": 0.33268481492996216, "learning_rate": 7.325096698172956e-05, "loss": 0.0603, "step": 75750 }, { "epoch": 2.5612765813583964, "grad_norm": 0.5167682766914368, "learning_rate": 7.324364757546534e-05, "loss": 0.06, "step": 75760 }, { "epoch": 2.5616146590486495, "grad_norm": 0.6728656888008118, "learning_rate": 7.32363275337285e-05, "loss": 0.0608, "step": 75770 }, { "epoch": 2.5619527367389026, "grad_norm": 0.8520151972770691, "learning_rate": 7.322900685671921e-05, "loss": 0.0567, "step": 75780 }, { "epoch": 2.562290814429156, "grad_norm": 0.5271980166435242, "learning_rate": 7.32216855446376e-05, "loss": 0.0547, "step": 75790 }, { "epoch": 2.562628892119409, "grad_norm": 0.6224406361579895, "learning_rate": 7.321436359768382e-05, "loss": 0.056, "step": 75800 }, { "epoch": 2.562966969809662, "grad_norm": 0.7110934257507324, "learning_rate": 7.320704101605808e-05, "loss": 0.0577, "step": 75810 }, { "epoch": 2.5633050474999157, "grad_norm": 0.6162413358688354, "learning_rate": 7.319971779996053e-05, "loss": 0.0592, "step": 75820 }, { "epoch": 2.563643125190169, "grad_norm": 1.2864724397659302, "learning_rate": 7.319239394959143e-05, "loss": 0.0574, "step": 75830 }, { "epoch": 2.563981202880422, "grad_norm": 1.0173994302749634, "learning_rate": 7.3185069465151e-05, "loss": 0.0544, "step": 75840 }, { "epoch": 2.5643192805706754, "grad_norm": 0.9861511588096619, "learning_rate": 7.317774434683946e-05, "loss": 0.0626, "step": 75850 }, { "epoch": 2.5646573582609284, "grad_norm": 0.47978708148002625, "learning_rate": 7.31704185948571e-05, "loss": 0.051, "step": 75860 }, { "epoch": 2.5649954359511815, "grad_norm": 1.2561794519424438, "learning_rate": 7.316309220940422e-05, "loss": 0.0564, "step": 75870 }, { "epoch": 2.565333513641435, "grad_norm": 0.44772395491600037, "learning_rate": 7.315576519068108e-05, "loss": 0.0575, "step": 75880 }, { "epoch": 2.565671591331688, "grad_norm": 0.594319760799408, "learning_rate": 7.314843753888803e-05, "loss": 0.0489, "step": 75890 }, { "epoch": 2.566009669021941, "grad_norm": 0.8282900452613831, "learning_rate": 7.314110925422539e-05, "loss": 0.0562, "step": 75900 }, { "epoch": 2.5663477467121947, "grad_norm": 0.5900702476501465, "learning_rate": 7.313378033689352e-05, "loss": 0.0597, "step": 75910 }, { "epoch": 2.5666858244024477, "grad_norm": 0.98430335521698, "learning_rate": 7.312645078709278e-05, "loss": 0.0515, "step": 75920 }, { "epoch": 2.567023902092701, "grad_norm": 0.7182077169418335, "learning_rate": 7.311912060502357e-05, "loss": 0.0599, "step": 75930 }, { "epoch": 2.5673619797829543, "grad_norm": 0.6501374244689941, "learning_rate": 7.311178979088628e-05, "loss": 0.0525, "step": 75940 }, { "epoch": 2.5677000574732074, "grad_norm": 1.7665609121322632, "learning_rate": 7.310445834488136e-05, "loss": 0.0661, "step": 75950 }, { "epoch": 2.5680381351634605, "grad_norm": 0.5355004072189331, "learning_rate": 7.30971262672092e-05, "loss": 0.05, "step": 75960 }, { "epoch": 2.568376212853714, "grad_norm": 0.6921966671943665, "learning_rate": 7.308979355807031e-05, "loss": 0.0582, "step": 75970 }, { "epoch": 2.568714290543967, "grad_norm": 0.6398014426231384, "learning_rate": 7.308246021766511e-05, "loss": 0.062, "step": 75980 }, { "epoch": 2.56905236823422, "grad_norm": 0.6107906699180603, "learning_rate": 7.307512624619415e-05, "loss": 0.0561, "step": 75990 }, { "epoch": 2.5693904459244736, "grad_norm": 0.701644241809845, "learning_rate": 7.30677916438579e-05, "loss": 0.0626, "step": 76000 }, { "epoch": 2.5697285236147267, "grad_norm": 1.1299786567687988, "learning_rate": 7.306045641085688e-05, "loss": 0.0587, "step": 76010 }, { "epoch": 2.5700666013049798, "grad_norm": 0.8130368590354919, "learning_rate": 7.305312054739166e-05, "loss": 0.057, "step": 76020 }, { "epoch": 2.5704046789952333, "grad_norm": 0.5192604064941406, "learning_rate": 7.304578405366276e-05, "loss": 0.0597, "step": 76030 }, { "epoch": 2.5707427566854864, "grad_norm": 0.9254018068313599, "learning_rate": 7.303844692987081e-05, "loss": 0.0584, "step": 76040 }, { "epoch": 2.5710808343757394, "grad_norm": 0.5740879774093628, "learning_rate": 7.303110917621637e-05, "loss": 0.0576, "step": 76050 }, { "epoch": 2.571418912065993, "grad_norm": 0.7145863175392151, "learning_rate": 7.302377079290004e-05, "loss": 0.0572, "step": 76060 }, { "epoch": 2.571756989756246, "grad_norm": 0.5635077357292175, "learning_rate": 7.301643178012247e-05, "loss": 0.0592, "step": 76070 }, { "epoch": 2.572095067446499, "grad_norm": 0.9348610043525696, "learning_rate": 7.300909213808432e-05, "loss": 0.0586, "step": 76080 }, { "epoch": 2.5724331451367526, "grad_norm": 0.5611786246299744, "learning_rate": 7.300175186698622e-05, "loss": 0.0543, "step": 76090 }, { "epoch": 2.5727712228270057, "grad_norm": 0.8444593548774719, "learning_rate": 7.299441096702888e-05, "loss": 0.0552, "step": 76100 }, { "epoch": 2.5731093005172587, "grad_norm": 0.44027289748191833, "learning_rate": 7.298706943841298e-05, "loss": 0.0573, "step": 76110 }, { "epoch": 2.5734473782075122, "grad_norm": 0.7410919070243835, "learning_rate": 7.297972728133924e-05, "loss": 0.0618, "step": 76120 }, { "epoch": 2.5737854558977653, "grad_norm": 0.5046793222427368, "learning_rate": 7.297238449600837e-05, "loss": 0.0587, "step": 76130 }, { "epoch": 2.5741235335880184, "grad_norm": 0.6635203957557678, "learning_rate": 7.296504108262114e-05, "loss": 0.0506, "step": 76140 }, { "epoch": 2.574461611278272, "grad_norm": 0.8162645697593689, "learning_rate": 7.295769704137832e-05, "loss": 0.0549, "step": 76150 }, { "epoch": 2.574799688968525, "grad_norm": 0.7842698693275452, "learning_rate": 7.29503523724807e-05, "loss": 0.0581, "step": 76160 }, { "epoch": 2.575137766658778, "grad_norm": 0.5693742632865906, "learning_rate": 7.294300707612905e-05, "loss": 0.0523, "step": 76170 }, { "epoch": 2.5754758443490315, "grad_norm": 0.6069871783256531, "learning_rate": 7.293566115252423e-05, "loss": 0.0576, "step": 76180 }, { "epoch": 2.5758139220392846, "grad_norm": 0.6614569425582886, "learning_rate": 7.292831460186704e-05, "loss": 0.0533, "step": 76190 }, { "epoch": 2.5761519997295377, "grad_norm": 0.644379198551178, "learning_rate": 7.292096742435833e-05, "loss": 0.0563, "step": 76200 }, { "epoch": 2.576490077419791, "grad_norm": 0.664341390132904, "learning_rate": 7.291361962019898e-05, "loss": 0.0619, "step": 76210 }, { "epoch": 2.5768281551100443, "grad_norm": 0.8529562950134277, "learning_rate": 7.29062711895899e-05, "loss": 0.0555, "step": 76220 }, { "epoch": 2.5771662328002973, "grad_norm": 0.5139126777648926, "learning_rate": 7.289892213273195e-05, "loss": 0.0597, "step": 76230 }, { "epoch": 2.577504310490551, "grad_norm": 0.865452229976654, "learning_rate": 7.289157244982607e-05, "loss": 0.0596, "step": 76240 }, { "epoch": 2.577842388180804, "grad_norm": 0.6830989122390747, "learning_rate": 7.288422214107322e-05, "loss": 0.0564, "step": 76250 }, { "epoch": 2.578180465871057, "grad_norm": 0.5900964140892029, "learning_rate": 7.287687120667432e-05, "loss": 0.0647, "step": 76260 }, { "epoch": 2.5785185435613105, "grad_norm": 1.0556007623672485, "learning_rate": 7.286951964683036e-05, "loss": 0.0633, "step": 76270 }, { "epoch": 2.5788566212515636, "grad_norm": 0.961440920829773, "learning_rate": 7.286216746174233e-05, "loss": 0.0561, "step": 76280 }, { "epoch": 2.5791946989418166, "grad_norm": 0.6846992373466492, "learning_rate": 7.285481465161123e-05, "loss": 0.0572, "step": 76290 }, { "epoch": 2.57953277663207, "grad_norm": 0.4142138659954071, "learning_rate": 7.284746121663807e-05, "loss": 0.0571, "step": 76300 }, { "epoch": 2.579870854322323, "grad_norm": 0.5598763227462769, "learning_rate": 7.284010715702392e-05, "loss": 0.0556, "step": 76310 }, { "epoch": 2.5802089320125763, "grad_norm": 0.5529046058654785, "learning_rate": 7.28327524729698e-05, "loss": 0.0659, "step": 76320 }, { "epoch": 2.58054700970283, "grad_norm": 1.0132213830947876, "learning_rate": 7.282539716467682e-05, "loss": 0.06, "step": 76330 }, { "epoch": 2.580885087393083, "grad_norm": 0.5799549221992493, "learning_rate": 7.281804123234605e-05, "loss": 0.0578, "step": 76340 }, { "epoch": 2.581223165083336, "grad_norm": 0.505401611328125, "learning_rate": 7.281068467617862e-05, "loss": 0.0612, "step": 76350 }, { "epoch": 2.5815612427735894, "grad_norm": 0.7501521110534668, "learning_rate": 7.280332749637561e-05, "loss": 0.0603, "step": 76360 }, { "epoch": 2.5818993204638425, "grad_norm": 0.6262615919113159, "learning_rate": 7.279596969313823e-05, "loss": 0.0661, "step": 76370 }, { "epoch": 2.5822373981540956, "grad_norm": 0.4972258508205414, "learning_rate": 7.278861126666758e-05, "loss": 0.0543, "step": 76380 }, { "epoch": 2.582575475844349, "grad_norm": 2.055833339691162, "learning_rate": 7.278125221716486e-05, "loss": 0.0605, "step": 76390 }, { "epoch": 2.582913553534602, "grad_norm": 0.6380904316902161, "learning_rate": 7.277389254483127e-05, "loss": 0.0636, "step": 76400 }, { "epoch": 2.5832516312248552, "grad_norm": 0.626889169216156, "learning_rate": 7.276653224986801e-05, "loss": 0.0584, "step": 76410 }, { "epoch": 2.5835897089151088, "grad_norm": 0.8466479778289795, "learning_rate": 7.275917133247632e-05, "loss": 0.0572, "step": 76420 }, { "epoch": 2.583927786605362, "grad_norm": 0.7438868284225464, "learning_rate": 7.275180979285743e-05, "loss": 0.0622, "step": 76430 }, { "epoch": 2.584265864295615, "grad_norm": 0.35839390754699707, "learning_rate": 7.27444476312126e-05, "loss": 0.0621, "step": 76440 }, { "epoch": 2.5846039419858684, "grad_norm": 0.7013611793518066, "learning_rate": 7.273708484774312e-05, "loss": 0.0594, "step": 76450 }, { "epoch": 2.5849420196761215, "grad_norm": 0.6763083338737488, "learning_rate": 7.272972144265029e-05, "loss": 0.0533, "step": 76460 }, { "epoch": 2.5852800973663745, "grad_norm": 0.5343705415725708, "learning_rate": 7.272235741613541e-05, "loss": 0.0567, "step": 76470 }, { "epoch": 2.585618175056628, "grad_norm": 0.8431388139724731, "learning_rate": 7.271499276839982e-05, "loss": 0.0607, "step": 76480 }, { "epoch": 2.585956252746881, "grad_norm": 0.44362887740135193, "learning_rate": 7.270762749964485e-05, "loss": 0.0581, "step": 76490 }, { "epoch": 2.586294330437134, "grad_norm": 0.9354041218757629, "learning_rate": 7.270026161007188e-05, "loss": 0.0623, "step": 76500 }, { "epoch": 2.5866324081273877, "grad_norm": 0.5869567394256592, "learning_rate": 7.269289509988229e-05, "loss": 0.0588, "step": 76510 }, { "epoch": 2.5869704858176408, "grad_norm": 1.1328191757202148, "learning_rate": 7.268552796927745e-05, "loss": 0.0611, "step": 76520 }, { "epoch": 2.587308563507894, "grad_norm": 0.471796452999115, "learning_rate": 7.267816021845882e-05, "loss": 0.0571, "step": 76530 }, { "epoch": 2.5876466411981474, "grad_norm": 0.5946875810623169, "learning_rate": 7.267079184762781e-05, "loss": 0.06, "step": 76540 }, { "epoch": 2.5879847188884004, "grad_norm": 0.7413159012794495, "learning_rate": 7.266342285698587e-05, "loss": 0.0563, "step": 76550 }, { "epoch": 2.588322796578654, "grad_norm": 0.7733597755432129, "learning_rate": 7.265605324673447e-05, "loss": 0.0588, "step": 76560 }, { "epoch": 2.588660874268907, "grad_norm": 0.6632053852081299, "learning_rate": 7.264868301707507e-05, "loss": 0.0616, "step": 76570 }, { "epoch": 2.58899895195916, "grad_norm": 1.2950150966644287, "learning_rate": 7.264131216820918e-05, "loss": 0.0585, "step": 76580 }, { "epoch": 2.5893370296494136, "grad_norm": 0.9104880690574646, "learning_rate": 7.263394070033834e-05, "loss": 0.0624, "step": 76590 }, { "epoch": 2.5896751073396667, "grad_norm": 13.547212600708008, "learning_rate": 7.262656861366405e-05, "loss": 0.059, "step": 76600 }, { "epoch": 2.5900131850299197, "grad_norm": 0.5880537033081055, "learning_rate": 7.261919590838788e-05, "loss": 0.0541, "step": 76610 }, { "epoch": 2.5903512627201732, "grad_norm": 0.4375222325325012, "learning_rate": 7.261182258471139e-05, "loss": 0.0535, "step": 76620 }, { "epoch": 2.5906893404104263, "grad_norm": 0.7344573736190796, "learning_rate": 7.260444864283616e-05, "loss": 0.0564, "step": 76630 }, { "epoch": 2.5910274181006794, "grad_norm": 0.4813941419124603, "learning_rate": 7.259707408296381e-05, "loss": 0.0593, "step": 76640 }, { "epoch": 2.591365495790933, "grad_norm": 0.5386002659797668, "learning_rate": 7.258969890529593e-05, "loss": 0.0565, "step": 76650 }, { "epoch": 2.591703573481186, "grad_norm": 0.6512331962585449, "learning_rate": 7.258232311003418e-05, "loss": 0.0598, "step": 76660 }, { "epoch": 2.592041651171439, "grad_norm": 1.0975894927978516, "learning_rate": 7.257494669738018e-05, "loss": 0.0571, "step": 76670 }, { "epoch": 2.5923797288616925, "grad_norm": 1.174404501914978, "learning_rate": 7.256756966753565e-05, "loss": 0.0571, "step": 76680 }, { "epoch": 2.5927178065519456, "grad_norm": 0.6546758413314819, "learning_rate": 7.256019202070222e-05, "loss": 0.0606, "step": 76690 }, { "epoch": 2.593055884242199, "grad_norm": 2.42187762260437, "learning_rate": 7.255281375708162e-05, "loss": 0.0625, "step": 76700 }, { "epoch": 2.593393961932452, "grad_norm": 0.45185431838035583, "learning_rate": 7.254543487687558e-05, "loss": 0.0576, "step": 76710 }, { "epoch": 2.5937320396227053, "grad_norm": 1.0025631189346313, "learning_rate": 7.25380553802858e-05, "loss": 0.0584, "step": 76720 }, { "epoch": 2.594070117312959, "grad_norm": 0.7078099846839905, "learning_rate": 7.253067526751406e-05, "loss": 0.0615, "step": 76730 }, { "epoch": 2.594408195003212, "grad_norm": 0.818277895450592, "learning_rate": 7.252329453876212e-05, "loss": 0.0653, "step": 76740 }, { "epoch": 2.594746272693465, "grad_norm": 0.6993176937103271, "learning_rate": 7.251591319423177e-05, "loss": 0.054, "step": 76750 }, { "epoch": 2.5950843503837184, "grad_norm": 0.6321616768836975, "learning_rate": 7.250853123412482e-05, "loss": 0.0548, "step": 76760 }, { "epoch": 2.5954224280739715, "grad_norm": 3.3899648189544678, "learning_rate": 7.250114865864307e-05, "loss": 0.0563, "step": 76770 }, { "epoch": 2.5957605057642246, "grad_norm": 0.6721659898757935, "learning_rate": 7.249376546798838e-05, "loss": 0.0573, "step": 76780 }, { "epoch": 2.596098583454478, "grad_norm": 0.5842357277870178, "learning_rate": 7.248638166236259e-05, "loss": 0.0575, "step": 76790 }, { "epoch": 2.596436661144731, "grad_norm": 0.7003534436225891, "learning_rate": 7.247899724196756e-05, "loss": 0.0576, "step": 76800 }, { "epoch": 2.596774738834984, "grad_norm": 0.5920403599739075, "learning_rate": 7.247161220700522e-05, "loss": 0.0577, "step": 76810 }, { "epoch": 2.5971128165252377, "grad_norm": 0.7620457410812378, "learning_rate": 7.246422655767742e-05, "loss": 0.0597, "step": 76820 }, { "epoch": 2.597450894215491, "grad_norm": 0.6984841227531433, "learning_rate": 7.24568402941861e-05, "loss": 0.0595, "step": 76830 }, { "epoch": 2.597788971905744, "grad_norm": 0.4183480441570282, "learning_rate": 7.244945341673322e-05, "loss": 0.0547, "step": 76840 }, { "epoch": 2.5981270495959974, "grad_norm": 0.7193061709403992, "learning_rate": 7.244206592552071e-05, "loss": 0.0643, "step": 76850 }, { "epoch": 2.5984651272862505, "grad_norm": 0.7948480248451233, "learning_rate": 7.243467782075054e-05, "loss": 0.0625, "step": 76860 }, { "epoch": 2.5988032049765035, "grad_norm": 0.7800285220146179, "learning_rate": 7.242728910262472e-05, "loss": 0.0619, "step": 76870 }, { "epoch": 2.599141282666757, "grad_norm": 0.6562572717666626, "learning_rate": 7.241989977134524e-05, "loss": 0.0591, "step": 76880 }, { "epoch": 2.59947936035701, "grad_norm": 0.8098133206367493, "learning_rate": 7.241250982711413e-05, "loss": 0.06, "step": 76890 }, { "epoch": 2.599817438047263, "grad_norm": 0.45641985535621643, "learning_rate": 7.24051192701334e-05, "loss": 0.0574, "step": 76900 }, { "epoch": 2.6001555157375167, "grad_norm": 0.5173595547676086, "learning_rate": 7.239772810060515e-05, "loss": 0.0599, "step": 76910 }, { "epoch": 2.6004935934277698, "grad_norm": 0.6201762557029724, "learning_rate": 7.23903363187314e-05, "loss": 0.0644, "step": 76920 }, { "epoch": 2.600831671118023, "grad_norm": 0.30713948607444763, "learning_rate": 7.23829439247143e-05, "loss": 0.0557, "step": 76930 }, { "epoch": 2.6011697488082763, "grad_norm": 1.008413553237915, "learning_rate": 7.23755509187559e-05, "loss": 0.0607, "step": 76940 }, { "epoch": 2.6015078264985294, "grad_norm": 0.5072164535522461, "learning_rate": 7.236815730105835e-05, "loss": 0.0594, "step": 76950 }, { "epoch": 2.6018459041887825, "grad_norm": 0.688162088394165, "learning_rate": 7.236076307182378e-05, "loss": 0.0613, "step": 76960 }, { "epoch": 2.602183981879036, "grad_norm": 0.3847099244594574, "learning_rate": 7.235336823125434e-05, "loss": 0.0615, "step": 76970 }, { "epoch": 2.602522059569289, "grad_norm": 0.5836260318756104, "learning_rate": 7.234597277955222e-05, "loss": 0.0603, "step": 76980 }, { "epoch": 2.602860137259542, "grad_norm": 1.4877774715423584, "learning_rate": 7.23385767169196e-05, "loss": 0.0609, "step": 76990 }, { "epoch": 2.6031982149497956, "grad_norm": 0.5882909893989563, "learning_rate": 7.23311800435587e-05, "loss": 0.0565, "step": 77000 }, { "epoch": 2.6035362926400487, "grad_norm": 0.5626007914543152, "learning_rate": 7.232378275967172e-05, "loss": 0.0612, "step": 77010 }, { "epoch": 2.603874370330302, "grad_norm": 0.6758400797843933, "learning_rate": 7.231638486546089e-05, "loss": 0.0568, "step": 77020 }, { "epoch": 2.6042124480205553, "grad_norm": 0.4064030647277832, "learning_rate": 7.230898636112848e-05, "loss": 0.0509, "step": 77030 }, { "epoch": 2.6045505257108084, "grad_norm": 0.5070512890815735, "learning_rate": 7.230158724687678e-05, "loss": 0.0594, "step": 77040 }, { "epoch": 2.6048886034010614, "grad_norm": 1.1009764671325684, "learning_rate": 7.229418752290805e-05, "loss": 0.0667, "step": 77050 }, { "epoch": 2.605226681091315, "grad_norm": 0.7722114324569702, "learning_rate": 7.228678718942463e-05, "loss": 0.0607, "step": 77060 }, { "epoch": 2.605564758781568, "grad_norm": 0.545076310634613, "learning_rate": 7.22793862466288e-05, "loss": 0.0572, "step": 77070 }, { "epoch": 2.605902836471821, "grad_norm": 0.8491846919059753, "learning_rate": 7.227198469472292e-05, "loss": 0.058, "step": 77080 }, { "epoch": 2.6062409141620746, "grad_norm": 0.5392397046089172, "learning_rate": 7.226458253390935e-05, "loss": 0.0538, "step": 77090 }, { "epoch": 2.6065789918523277, "grad_norm": 0.799324095249176, "learning_rate": 7.225717976439046e-05, "loss": 0.0637, "step": 77100 }, { "epoch": 2.6069170695425807, "grad_norm": 0.6506450772285461, "learning_rate": 7.224977638636863e-05, "loss": 0.0541, "step": 77110 }, { "epoch": 2.6072551472328342, "grad_norm": 2.4652178287506104, "learning_rate": 7.224237240004628e-05, "loss": 0.0633, "step": 77120 }, { "epoch": 2.6075932249230873, "grad_norm": 1.0199090242385864, "learning_rate": 7.223496780562583e-05, "loss": 0.0568, "step": 77130 }, { "epoch": 2.6079313026133404, "grad_norm": 0.6958945393562317, "learning_rate": 7.222756260330972e-05, "loss": 0.0575, "step": 77140 }, { "epoch": 2.608269380303594, "grad_norm": 0.8010122776031494, "learning_rate": 7.222015679330037e-05, "loss": 0.0589, "step": 77150 }, { "epoch": 2.608607457993847, "grad_norm": 0.6925544738769531, "learning_rate": 7.221275037580029e-05, "loss": 0.0563, "step": 77160 }, { "epoch": 2.6089455356841, "grad_norm": 0.8046688437461853, "learning_rate": 7.220534335101197e-05, "loss": 0.059, "step": 77170 }, { "epoch": 2.6092836133743535, "grad_norm": 2.26094913482666, "learning_rate": 7.219793571913792e-05, "loss": 0.0563, "step": 77180 }, { "epoch": 2.6096216910646066, "grad_norm": 0.4867696166038513, "learning_rate": 7.219052748038062e-05, "loss": 0.0597, "step": 77190 }, { "epoch": 2.6099597687548597, "grad_norm": 0.4396221339702606, "learning_rate": 7.218311863494263e-05, "loss": 0.0585, "step": 77200 }, { "epoch": 2.610297846445113, "grad_norm": 0.8310909271240234, "learning_rate": 7.217570918302655e-05, "loss": 0.0573, "step": 77210 }, { "epoch": 2.6106359241353663, "grad_norm": 0.3715589642524719, "learning_rate": 7.216829912483486e-05, "loss": 0.0639, "step": 77220 }, { "epoch": 2.6109740018256193, "grad_norm": 0.8006253838539124, "learning_rate": 7.216088846057022e-05, "loss": 0.062, "step": 77230 }, { "epoch": 2.611312079515873, "grad_norm": 0.6625715494155884, "learning_rate": 7.215347719043524e-05, "loss": 0.0681, "step": 77240 }, { "epoch": 2.611650157206126, "grad_norm": 0.5948194265365601, "learning_rate": 7.21460653146325e-05, "loss": 0.0603, "step": 77250 }, { "epoch": 2.611988234896379, "grad_norm": 0.6141760945320129, "learning_rate": 7.213865283336464e-05, "loss": 0.0592, "step": 77260 }, { "epoch": 2.6123263125866325, "grad_norm": 0.46584081649780273, "learning_rate": 7.213123974683436e-05, "loss": 0.0584, "step": 77270 }, { "epoch": 2.6126643902768856, "grad_norm": 0.4748814105987549, "learning_rate": 7.212382605524425e-05, "loss": 0.0599, "step": 77280 }, { "epoch": 2.6130024679671386, "grad_norm": 0.5618825554847717, "learning_rate": 7.211641175879709e-05, "loss": 0.054, "step": 77290 }, { "epoch": 2.613340545657392, "grad_norm": 0.7182773947715759, "learning_rate": 7.210899685769552e-05, "loss": 0.0584, "step": 77300 }, { "epoch": 2.6136786233476452, "grad_norm": 0.43279340863227844, "learning_rate": 7.21015813521423e-05, "loss": 0.0584, "step": 77310 }, { "epoch": 2.6140167010378983, "grad_norm": 0.9438582062721252, "learning_rate": 7.209416524234013e-05, "loss": 0.0586, "step": 77320 }, { "epoch": 2.614354778728152, "grad_norm": 0.4353700280189514, "learning_rate": 7.208674852849181e-05, "loss": 0.0587, "step": 77330 }, { "epoch": 2.614692856418405, "grad_norm": 0.6952443718910217, "learning_rate": 7.207933121080006e-05, "loss": 0.0566, "step": 77340 }, { "epoch": 2.615030934108658, "grad_norm": 0.5159597396850586, "learning_rate": 7.20719132894677e-05, "loss": 0.0592, "step": 77350 }, { "epoch": 2.6153690117989115, "grad_norm": 0.5461724400520325, "learning_rate": 7.20644947646975e-05, "loss": 0.0637, "step": 77360 }, { "epoch": 2.6157070894891645, "grad_norm": 0.4663631021976471, "learning_rate": 7.205707563669232e-05, "loss": 0.0573, "step": 77370 }, { "epoch": 2.6160451671794176, "grad_norm": 0.7724102139472961, "learning_rate": 7.204965590565497e-05, "loss": 0.0635, "step": 77380 }, { "epoch": 2.616383244869671, "grad_norm": 0.7869552373886108, "learning_rate": 7.204223557178834e-05, "loss": 0.06, "step": 77390 }, { "epoch": 2.616721322559924, "grad_norm": 0.4796316921710968, "learning_rate": 7.203481463529525e-05, "loss": 0.0579, "step": 77400 }, { "epoch": 2.6170594002501772, "grad_norm": 0.5807232856750488, "learning_rate": 7.20273930963786e-05, "loss": 0.0532, "step": 77410 }, { "epoch": 2.6173974779404308, "grad_norm": 0.6664503812789917, "learning_rate": 7.20199709552413e-05, "loss": 0.0541, "step": 77420 }, { "epoch": 2.617735555630684, "grad_norm": 0.6082444190979004, "learning_rate": 7.201254821208627e-05, "loss": 0.0557, "step": 77430 }, { "epoch": 2.618073633320937, "grad_norm": 0.5402559041976929, "learning_rate": 7.200512486711644e-05, "loss": 0.0596, "step": 77440 }, { "epoch": 2.6184117110111904, "grad_norm": 0.30733123421669006, "learning_rate": 7.199770092053476e-05, "loss": 0.057, "step": 77450 }, { "epoch": 2.6187497887014435, "grad_norm": 1.0502880811691284, "learning_rate": 7.199027637254421e-05, "loss": 0.0562, "step": 77460 }, { "epoch": 2.6190878663916966, "grad_norm": 0.45494893193244934, "learning_rate": 7.198285122334775e-05, "loss": 0.057, "step": 77470 }, { "epoch": 2.61942594408195, "grad_norm": 0.5643818378448486, "learning_rate": 7.197542547314841e-05, "loss": 0.0543, "step": 77480 }, { "epoch": 2.619764021772203, "grad_norm": 0.9567685723304749, "learning_rate": 7.19679991221492e-05, "loss": 0.0614, "step": 77490 }, { "epoch": 2.620102099462456, "grad_norm": 1.1146467924118042, "learning_rate": 7.196057217055315e-05, "loss": 0.0677, "step": 77500 }, { "epoch": 2.6204401771527097, "grad_norm": 1.0057953596115112, "learning_rate": 7.195314461856328e-05, "loss": 0.0613, "step": 77510 }, { "epoch": 2.620778254842963, "grad_norm": 0.46457359194755554, "learning_rate": 7.194571646638271e-05, "loss": 0.0673, "step": 77520 }, { "epoch": 2.6211163325332163, "grad_norm": 0.4767102003097534, "learning_rate": 7.193828771421447e-05, "loss": 0.0537, "step": 77530 }, { "epoch": 2.6214544102234694, "grad_norm": 0.9036865234375, "learning_rate": 7.19308583622617e-05, "loss": 0.0582, "step": 77540 }, { "epoch": 2.6217924879137224, "grad_norm": 1.5970308780670166, "learning_rate": 7.19234284107275e-05, "loss": 0.0597, "step": 77550 }, { "epoch": 2.622130565603976, "grad_norm": 2.045900583267212, "learning_rate": 7.191599785981502e-05, "loss": 0.0612, "step": 77560 }, { "epoch": 2.622468643294229, "grad_norm": 0.8974219560623169, "learning_rate": 7.190856670972737e-05, "loss": 0.0614, "step": 77570 }, { "epoch": 2.622806720984482, "grad_norm": 0.7830896377563477, "learning_rate": 7.190113496066775e-05, "loss": 0.0654, "step": 77580 }, { "epoch": 2.6231447986747356, "grad_norm": 0.6355934739112854, "learning_rate": 7.189370261283933e-05, "loss": 0.0558, "step": 77590 }, { "epoch": 2.6234828763649887, "grad_norm": 1.9107297658920288, "learning_rate": 7.188626966644528e-05, "loss": 0.0559, "step": 77600 }, { "epoch": 2.6238209540552417, "grad_norm": 0.543176531791687, "learning_rate": 7.187883612168886e-05, "loss": 0.0596, "step": 77610 }, { "epoch": 2.6241590317454953, "grad_norm": 1.1089924573898315, "learning_rate": 7.187140197877328e-05, "loss": 0.0655, "step": 77620 }, { "epoch": 2.6244971094357483, "grad_norm": 0.5821645855903625, "learning_rate": 7.186396723790179e-05, "loss": 0.0613, "step": 77630 }, { "epoch": 2.6248351871260014, "grad_norm": 0.6205069422721863, "learning_rate": 7.185653189927763e-05, "loss": 0.0617, "step": 77640 }, { "epoch": 2.625173264816255, "grad_norm": 0.41729480028152466, "learning_rate": 7.184909596310412e-05, "loss": 0.0639, "step": 77650 }, { "epoch": 2.625511342506508, "grad_norm": 0.7335444092750549, "learning_rate": 7.184165942958452e-05, "loss": 0.0617, "step": 77660 }, { "epoch": 2.6258494201967615, "grad_norm": 0.5084216594696045, "learning_rate": 7.183422229892217e-05, "loss": 0.0638, "step": 77670 }, { "epoch": 2.6261874978870146, "grad_norm": 0.45816782116889954, "learning_rate": 7.182678457132039e-05, "loss": 0.0575, "step": 77680 }, { "epoch": 2.6265255755772676, "grad_norm": 1.3454418182373047, "learning_rate": 7.18193462469825e-05, "loss": 0.0649, "step": 77690 }, { "epoch": 2.626863653267521, "grad_norm": 0.9099161624908447, "learning_rate": 7.181190732611189e-05, "loss": 0.0654, "step": 77700 }, { "epoch": 2.627201730957774, "grad_norm": 0.737383246421814, "learning_rate": 7.180446780891192e-05, "loss": 0.0633, "step": 77710 }, { "epoch": 2.6275398086480273, "grad_norm": 1.163262963294983, "learning_rate": 7.179702769558601e-05, "loss": 0.0569, "step": 77720 }, { "epoch": 2.627877886338281, "grad_norm": 0.7317689061164856, "learning_rate": 7.178958698633753e-05, "loss": 0.0634, "step": 77730 }, { "epoch": 2.628215964028534, "grad_norm": 1.3020834922790527, "learning_rate": 7.178214568136993e-05, "loss": 0.06, "step": 77740 }, { "epoch": 2.628554041718787, "grad_norm": 1.1101152896881104, "learning_rate": 7.177470378088667e-05, "loss": 0.0607, "step": 77750 }, { "epoch": 2.6288921194090404, "grad_norm": 0.9561934471130371, "learning_rate": 7.176726128509117e-05, "loss": 0.0589, "step": 77760 }, { "epoch": 2.6292301970992935, "grad_norm": 1.0171805620193481, "learning_rate": 7.175981819418695e-05, "loss": 0.0587, "step": 77770 }, { "epoch": 2.6295682747895466, "grad_norm": 1.5550087690353394, "learning_rate": 7.175237450837743e-05, "loss": 0.0584, "step": 77780 }, { "epoch": 2.6299063524798, "grad_norm": 1.5739784240722656, "learning_rate": 7.174493022786621e-05, "loss": 0.0712, "step": 77790 }, { "epoch": 2.630244430170053, "grad_norm": 1.700255036354065, "learning_rate": 7.173748535285674e-05, "loss": 0.0591, "step": 77800 }, { "epoch": 2.6305825078603062, "grad_norm": 0.8625608086585999, "learning_rate": 7.17300398835526e-05, "loss": 0.0639, "step": 77810 }, { "epoch": 2.6309205855505597, "grad_norm": 0.7958372235298157, "learning_rate": 7.172259382015733e-05, "loss": 0.0554, "step": 77820 }, { "epoch": 2.631258663240813, "grad_norm": 0.6954524517059326, "learning_rate": 7.17151471628745e-05, "loss": 0.0651, "step": 77830 }, { "epoch": 2.631596740931066, "grad_norm": 0.608246922492981, "learning_rate": 7.170769991190771e-05, "loss": 0.0607, "step": 77840 }, { "epoch": 2.6319348186213194, "grad_norm": 0.5973051190376282, "learning_rate": 7.170025206746055e-05, "loss": 0.0559, "step": 77850 }, { "epoch": 2.6322728963115725, "grad_norm": 0.6032540202140808, "learning_rate": 7.169280362973665e-05, "loss": 0.0568, "step": 77860 }, { "epoch": 2.6326109740018255, "grad_norm": 0.8577111959457397, "learning_rate": 7.168535459893966e-05, "loss": 0.062, "step": 77870 }, { "epoch": 2.632949051692079, "grad_norm": 0.6399845480918884, "learning_rate": 7.167790497527323e-05, "loss": 0.0581, "step": 77880 }, { "epoch": 2.633287129382332, "grad_norm": 0.572027325630188, "learning_rate": 7.1670454758941e-05, "loss": 0.0579, "step": 77890 }, { "epoch": 2.633625207072585, "grad_norm": 0.48943856358528137, "learning_rate": 7.16630039501467e-05, "loss": 0.0606, "step": 77900 }, { "epoch": 2.6339632847628387, "grad_norm": 0.9505097270011902, "learning_rate": 7.165555254909398e-05, "loss": 0.0572, "step": 77910 }, { "epoch": 2.6343013624530918, "grad_norm": 0.6708853840827942, "learning_rate": 7.164810055598661e-05, "loss": 0.058, "step": 77920 }, { "epoch": 2.634639440143345, "grad_norm": 0.6021277904510498, "learning_rate": 7.164064797102831e-05, "loss": 0.0547, "step": 77930 }, { "epoch": 2.6349775178335983, "grad_norm": 0.37955015897750854, "learning_rate": 7.16331947944228e-05, "loss": 0.056, "step": 77940 }, { "epoch": 2.6353155955238514, "grad_norm": 0.40871959924697876, "learning_rate": 7.16257410263739e-05, "loss": 0.0601, "step": 77950 }, { "epoch": 2.6356536732141045, "grad_norm": 0.6007097959518433, "learning_rate": 7.161828666708535e-05, "loss": 0.0555, "step": 77960 }, { "epoch": 2.635991750904358, "grad_norm": 0.41629695892333984, "learning_rate": 7.161083171676097e-05, "loss": 0.0599, "step": 77970 }, { "epoch": 2.636329828594611, "grad_norm": 0.4565883278846741, "learning_rate": 7.160337617560455e-05, "loss": 0.0626, "step": 77980 }, { "epoch": 2.636667906284864, "grad_norm": 0.6208024621009827, "learning_rate": 7.159592004381996e-05, "loss": 0.0606, "step": 77990 }, { "epoch": 2.6370059839751177, "grad_norm": 0.4862643778324127, "learning_rate": 7.158846332161104e-05, "loss": 0.063, "step": 78000 }, { "epoch": 2.6373440616653707, "grad_norm": 0.3716149628162384, "learning_rate": 7.158100600918163e-05, "loss": 0.0579, "step": 78010 }, { "epoch": 2.637682139355624, "grad_norm": 0.4128357470035553, "learning_rate": 7.157354810673562e-05, "loss": 0.0607, "step": 78020 }, { "epoch": 2.6380202170458773, "grad_norm": 0.4268608093261719, "learning_rate": 7.15660896144769e-05, "loss": 0.0538, "step": 78030 }, { "epoch": 2.6383582947361304, "grad_norm": 0.4366587698459625, "learning_rate": 7.155863053260942e-05, "loss": 0.0605, "step": 78040 }, { "epoch": 2.6386963724263834, "grad_norm": 0.6351027488708496, "learning_rate": 7.155117086133705e-05, "loss": 0.0612, "step": 78050 }, { "epoch": 2.639034450116637, "grad_norm": 0.3592607378959656, "learning_rate": 7.154371060086378e-05, "loss": 0.0625, "step": 78060 }, { "epoch": 2.63937252780689, "grad_norm": 0.5482779741287231, "learning_rate": 7.153624975139356e-05, "loss": 0.0621, "step": 78070 }, { "epoch": 2.639710605497143, "grad_norm": 0.44816938042640686, "learning_rate": 7.152878831313037e-05, "loss": 0.0477, "step": 78080 }, { "epoch": 2.6400486831873966, "grad_norm": 0.4467684030532837, "learning_rate": 7.152132628627818e-05, "loss": 0.0574, "step": 78090 }, { "epoch": 2.6403867608776497, "grad_norm": 0.665565550327301, "learning_rate": 7.151386367104102e-05, "loss": 0.0574, "step": 78100 }, { "epoch": 2.6407248385679027, "grad_norm": 0.5179789066314697, "learning_rate": 7.150640046762289e-05, "loss": 0.0561, "step": 78110 }, { "epoch": 2.6410629162581563, "grad_norm": 0.7135960459709167, "learning_rate": 7.149893667622789e-05, "loss": 0.06, "step": 78120 }, { "epoch": 2.6414009939484093, "grad_norm": 0.5655475854873657, "learning_rate": 7.149147229706e-05, "loss": 0.0575, "step": 78130 }, { "epoch": 2.6417390716386624, "grad_norm": 0.5050358176231384, "learning_rate": 7.148400733032337e-05, "loss": 0.0587, "step": 78140 }, { "epoch": 2.642077149328916, "grad_norm": 0.4911496937274933, "learning_rate": 7.147654177622202e-05, "loss": 0.0591, "step": 78150 }, { "epoch": 2.642415227019169, "grad_norm": 0.49924522638320923, "learning_rate": 7.146907563496008e-05, "loss": 0.0603, "step": 78160 }, { "epoch": 2.642753304709422, "grad_norm": 0.584469735622406, "learning_rate": 7.146160890674169e-05, "loss": 0.0581, "step": 78170 }, { "epoch": 2.6430913823996756, "grad_norm": 0.4163244962692261, "learning_rate": 7.145414159177099e-05, "loss": 0.0587, "step": 78180 }, { "epoch": 2.6434294600899286, "grad_norm": 0.45937785506248474, "learning_rate": 7.144667369025211e-05, "loss": 0.0621, "step": 78190 }, { "epoch": 2.6437675377801817, "grad_norm": 0.9585686326026917, "learning_rate": 7.143920520238921e-05, "loss": 0.0584, "step": 78200 }, { "epoch": 2.644105615470435, "grad_norm": 0.5447681546211243, "learning_rate": 7.143173612838652e-05, "loss": 0.055, "step": 78210 }, { "epoch": 2.6444436931606883, "grad_norm": 0.44194918870925903, "learning_rate": 7.14242664684482e-05, "loss": 0.0574, "step": 78220 }, { "epoch": 2.6447817708509413, "grad_norm": 0.4946926534175873, "learning_rate": 7.14167962227785e-05, "loss": 0.0593, "step": 78230 }, { "epoch": 2.645119848541195, "grad_norm": 0.38895076513290405, "learning_rate": 7.140932539158162e-05, "loss": 0.0658, "step": 78240 }, { "epoch": 2.645457926231448, "grad_norm": 0.4714232385158539, "learning_rate": 7.140185397506182e-05, "loss": 0.0539, "step": 78250 }, { "epoch": 2.645796003921701, "grad_norm": 1.187643051147461, "learning_rate": 7.139438197342339e-05, "loss": 0.0563, "step": 78260 }, { "epoch": 2.6461340816119545, "grad_norm": 0.4570578932762146, "learning_rate": 7.13869093868706e-05, "loss": 0.0528, "step": 78270 }, { "epoch": 2.6464721593022076, "grad_norm": 0.8322599530220032, "learning_rate": 7.137943621560773e-05, "loss": 0.0578, "step": 78280 }, { "epoch": 2.6468102369924607, "grad_norm": 0.844781756401062, "learning_rate": 7.137196245983911e-05, "loss": 0.059, "step": 78290 }, { "epoch": 2.647148314682714, "grad_norm": 0.5383767485618591, "learning_rate": 7.136448811976905e-05, "loss": 0.0593, "step": 78300 }, { "epoch": 2.6474863923729672, "grad_norm": 1.598142385482788, "learning_rate": 7.135701319560193e-05, "loss": 0.0548, "step": 78310 }, { "epoch": 2.6478244700632203, "grad_norm": 0.6799598932266235, "learning_rate": 7.134953768754208e-05, "loss": 0.057, "step": 78320 }, { "epoch": 2.648162547753474, "grad_norm": 0.5548301935195923, "learning_rate": 7.13420615957939e-05, "loss": 0.0562, "step": 78330 }, { "epoch": 2.648500625443727, "grad_norm": 1.1547648906707764, "learning_rate": 7.133458492056178e-05, "loss": 0.0645, "step": 78340 }, { "epoch": 2.64883870313398, "grad_norm": 0.564031183719635, "learning_rate": 7.132710766205012e-05, "loss": 0.0529, "step": 78350 }, { "epoch": 2.6491767808242335, "grad_norm": 0.8727381229400635, "learning_rate": 7.131962982046335e-05, "loss": 0.0609, "step": 78360 }, { "epoch": 2.6495148585144865, "grad_norm": 0.39232754707336426, "learning_rate": 7.131215139600589e-05, "loss": 0.0584, "step": 78370 }, { "epoch": 2.6498529362047396, "grad_norm": 0.7807379961013794, "learning_rate": 7.130467238888222e-05, "loss": 0.0517, "step": 78380 }, { "epoch": 2.650191013894993, "grad_norm": 0.8295179009437561, "learning_rate": 7.129719279929682e-05, "loss": 0.0555, "step": 78390 }, { "epoch": 2.650529091585246, "grad_norm": 0.6025496125221252, "learning_rate": 7.128971262745419e-05, "loss": 0.0619, "step": 78400 }, { "epoch": 2.6508671692754993, "grad_norm": 0.7219594120979309, "learning_rate": 7.128223187355878e-05, "loss": 0.0554, "step": 78410 }, { "epoch": 2.6512052469657528, "grad_norm": 0.9676351547241211, "learning_rate": 7.127475053781517e-05, "loss": 0.0714, "step": 78420 }, { "epoch": 2.651543324656006, "grad_norm": 0.6558446288108826, "learning_rate": 7.126726862042786e-05, "loss": 0.0548, "step": 78430 }, { "epoch": 2.651881402346259, "grad_norm": 0.601250946521759, "learning_rate": 7.125978612160143e-05, "loss": 0.0571, "step": 78440 }, { "epoch": 2.6522194800365124, "grad_norm": 0.8354202508926392, "learning_rate": 7.125230304154042e-05, "loss": 0.0559, "step": 78450 }, { "epoch": 2.6525575577267655, "grad_norm": 1.1422638893127441, "learning_rate": 7.124481938044942e-05, "loss": 0.0551, "step": 78460 }, { "epoch": 2.6528956354170186, "grad_norm": 0.8713324666023254, "learning_rate": 7.123733513853305e-05, "loss": 0.0549, "step": 78470 }, { "epoch": 2.653233713107272, "grad_norm": 0.8089393973350525, "learning_rate": 7.122985031599594e-05, "loss": 0.064, "step": 78480 }, { "epoch": 2.653571790797525, "grad_norm": 0.6512552499771118, "learning_rate": 7.122236491304266e-05, "loss": 0.0552, "step": 78490 }, { "epoch": 2.6539098684877787, "grad_norm": 0.43348678946495056, "learning_rate": 7.121487892987792e-05, "loss": 0.0534, "step": 78500 }, { "epoch": 2.6542479461780317, "grad_norm": 0.6205800175666809, "learning_rate": 7.120739236670634e-05, "loss": 0.0619, "step": 78510 }, { "epoch": 2.654586023868285, "grad_norm": 0.5970582365989685, "learning_rate": 7.119990522373265e-05, "loss": 0.0599, "step": 78520 }, { "epoch": 2.6549241015585383, "grad_norm": 0.5563194751739502, "learning_rate": 7.11924175011615e-05, "loss": 0.0574, "step": 78530 }, { "epoch": 2.6552621792487914, "grad_norm": 0.7246320247650146, "learning_rate": 7.118492919919761e-05, "loss": 0.0605, "step": 78540 }, { "epoch": 2.6556002569390444, "grad_norm": 0.5691744685173035, "learning_rate": 7.117744031804572e-05, "loss": 0.0548, "step": 78550 }, { "epoch": 2.655938334629298, "grad_norm": 0.6624094843864441, "learning_rate": 7.116995085791057e-05, "loss": 0.059, "step": 78560 }, { "epoch": 2.656276412319551, "grad_norm": 0.7093308568000793, "learning_rate": 7.116246081899691e-05, "loss": 0.0539, "step": 78570 }, { "epoch": 2.656614490009804, "grad_norm": 0.7300847768783569, "learning_rate": 7.115497020150953e-05, "loss": 0.059, "step": 78580 }, { "epoch": 2.6569525677000576, "grad_norm": 0.5730242133140564, "learning_rate": 7.11474790056532e-05, "loss": 0.0644, "step": 78590 }, { "epoch": 2.6572906453903107, "grad_norm": 0.8098510503768921, "learning_rate": 7.113998723163277e-05, "loss": 0.0593, "step": 78600 }, { "epoch": 2.6576287230805637, "grad_norm": 1.239587664604187, "learning_rate": 7.1132494879653e-05, "loss": 0.0613, "step": 78610 }, { "epoch": 2.6579668007708173, "grad_norm": 0.7078257203102112, "learning_rate": 7.112500194991876e-05, "loss": 0.06, "step": 78620 }, { "epoch": 2.6583048784610703, "grad_norm": 1.7940709590911865, "learning_rate": 7.111750844263491e-05, "loss": 0.0559, "step": 78630 }, { "epoch": 2.658642956151324, "grad_norm": 0.5176776647567749, "learning_rate": 7.111001435800632e-05, "loss": 0.0516, "step": 78640 }, { "epoch": 2.658981033841577, "grad_norm": 0.45831140875816345, "learning_rate": 7.110251969623785e-05, "loss": 0.0628, "step": 78650 }, { "epoch": 2.65931911153183, "grad_norm": 0.4302363991737366, "learning_rate": 7.109502445753444e-05, "loss": 0.0588, "step": 78660 }, { "epoch": 2.6596571892220835, "grad_norm": 0.583135724067688, "learning_rate": 7.108752864210098e-05, "loss": 0.053, "step": 78670 }, { "epoch": 2.6599952669123366, "grad_norm": 0.7412582039833069, "learning_rate": 7.10800322501424e-05, "loss": 0.061, "step": 78680 }, { "epoch": 2.6603333446025896, "grad_norm": 0.5380671620368958, "learning_rate": 7.107253528186366e-05, "loss": 0.0623, "step": 78690 }, { "epoch": 2.660671422292843, "grad_norm": 1.6223622560501099, "learning_rate": 7.106503773746972e-05, "loss": 0.058, "step": 78700 }, { "epoch": 2.661009499983096, "grad_norm": 0.6452276706695557, "learning_rate": 7.105753961716557e-05, "loss": 0.0654, "step": 78710 }, { "epoch": 2.6613475776733493, "grad_norm": 0.533847987651825, "learning_rate": 7.105004092115619e-05, "loss": 0.0608, "step": 78720 }, { "epoch": 2.661685655363603, "grad_norm": 0.6577204465866089, "learning_rate": 7.104254164964662e-05, "loss": 0.0634, "step": 78730 }, { "epoch": 2.662023733053856, "grad_norm": 0.7506848573684692, "learning_rate": 7.103504180284185e-05, "loss": 0.0613, "step": 78740 }, { "epoch": 2.662361810744109, "grad_norm": 0.47102028131484985, "learning_rate": 7.102754138094693e-05, "loss": 0.0615, "step": 78750 }, { "epoch": 2.6626998884343624, "grad_norm": 0.9150038957595825, "learning_rate": 7.102004038416694e-05, "loss": 0.0618, "step": 78760 }, { "epoch": 2.6630379661246155, "grad_norm": 0.4243102967739105, "learning_rate": 7.101253881270695e-05, "loss": 0.0632, "step": 78770 }, { "epoch": 2.6633760438148686, "grad_norm": 1.0738258361816406, "learning_rate": 7.100503666677202e-05, "loss": 0.0579, "step": 78780 }, { "epoch": 2.663714121505122, "grad_norm": 0.4375089108943939, "learning_rate": 7.099753394656732e-05, "loss": 0.0563, "step": 78790 }, { "epoch": 2.664052199195375, "grad_norm": 0.49004098773002625, "learning_rate": 7.099003065229789e-05, "loss": 0.0548, "step": 78800 }, { "epoch": 2.6643902768856282, "grad_norm": 0.8648068904876709, "learning_rate": 7.098252678416892e-05, "loss": 0.0601, "step": 78810 }, { "epoch": 2.6647283545758818, "grad_norm": 0.9739906787872314, "learning_rate": 7.097502234238555e-05, "loss": 0.0661, "step": 78820 }, { "epoch": 2.665066432266135, "grad_norm": 0.5055481195449829, "learning_rate": 7.096751732715297e-05, "loss": 0.0594, "step": 78830 }, { "epoch": 2.665404509956388, "grad_norm": 0.576543927192688, "learning_rate": 7.096001173867631e-05, "loss": 0.0721, "step": 78840 }, { "epoch": 2.6657425876466414, "grad_norm": 0.9011796116828918, "learning_rate": 7.095250557716082e-05, "loss": 0.0595, "step": 78850 }, { "epoch": 2.6660806653368945, "grad_norm": 0.4757072627544403, "learning_rate": 7.094499884281169e-05, "loss": 0.0611, "step": 78860 }, { "epoch": 2.6664187430271475, "grad_norm": 0.617011308670044, "learning_rate": 7.093749153583417e-05, "loss": 0.0599, "step": 78870 }, { "epoch": 2.666756820717401, "grad_norm": 1.5907131433486938, "learning_rate": 7.09299836564335e-05, "loss": 0.0591, "step": 78880 }, { "epoch": 2.667094898407654, "grad_norm": 0.5555822253227234, "learning_rate": 7.092247520481495e-05, "loss": 0.0619, "step": 78890 }, { "epoch": 2.667432976097907, "grad_norm": 3.8262746334075928, "learning_rate": 7.091496618118376e-05, "loss": 0.0585, "step": 78900 }, { "epoch": 2.6677710537881607, "grad_norm": 0.7404365539550781, "learning_rate": 7.090745658574526e-05, "loss": 0.0693, "step": 78910 }, { "epoch": 2.6681091314784138, "grad_norm": 0.9843640923500061, "learning_rate": 7.089994641870477e-05, "loss": 0.0698, "step": 78920 }, { "epoch": 2.668447209168667, "grad_norm": 0.5346428155899048, "learning_rate": 7.089243568026756e-05, "loss": 0.057, "step": 78930 }, { "epoch": 2.6687852868589204, "grad_norm": 1.4355642795562744, "learning_rate": 7.088492437063904e-05, "loss": 0.0579, "step": 78940 }, { "epoch": 2.6691233645491734, "grad_norm": 0.46763142943382263, "learning_rate": 7.087741249002452e-05, "loss": 0.0615, "step": 78950 }, { "epoch": 2.6694614422394265, "grad_norm": 1.0763955116271973, "learning_rate": 7.086990003862939e-05, "loss": 0.0596, "step": 78960 }, { "epoch": 2.66979951992968, "grad_norm": 0.5727894306182861, "learning_rate": 7.086238701665904e-05, "loss": 0.0552, "step": 78970 }, { "epoch": 2.670137597619933, "grad_norm": 0.9754548072814941, "learning_rate": 7.085487342431886e-05, "loss": 0.0666, "step": 78980 }, { "epoch": 2.670475675310186, "grad_norm": 1.6695460081100464, "learning_rate": 7.084735926181428e-05, "loss": 0.0607, "step": 78990 }, { "epoch": 2.6708137530004397, "grad_norm": 0.5639792084693909, "learning_rate": 7.083984452935071e-05, "loss": 0.0623, "step": 79000 }, { "epoch": 2.6711518306906927, "grad_norm": 0.6066418886184692, "learning_rate": 7.083232922713364e-05, "loss": 0.0614, "step": 79010 }, { "epoch": 2.671489908380946, "grad_norm": 0.40791141986846924, "learning_rate": 7.082481335536851e-05, "loss": 0.0638, "step": 79020 }, { "epoch": 2.6718279860711993, "grad_norm": 0.5552513003349304, "learning_rate": 7.08172969142608e-05, "loss": 0.062, "step": 79030 }, { "epoch": 2.6721660637614524, "grad_norm": 0.3424718976020813, "learning_rate": 7.080977990401602e-05, "loss": 0.059, "step": 79040 }, { "epoch": 2.6725041414517055, "grad_norm": 0.6481580138206482, "learning_rate": 7.080226232483968e-05, "loss": 0.0582, "step": 79050 }, { "epoch": 2.672842219141959, "grad_norm": 0.6378081440925598, "learning_rate": 7.07947441769373e-05, "loss": 0.0602, "step": 79060 }, { "epoch": 2.673180296832212, "grad_norm": 1.4256447553634644, "learning_rate": 7.078722546051442e-05, "loss": 0.0608, "step": 79070 }, { "epoch": 2.673518374522465, "grad_norm": 0.7082626819610596, "learning_rate": 7.077970617577662e-05, "loss": 0.0639, "step": 79080 }, { "epoch": 2.6738564522127186, "grad_norm": 0.9340940117835999, "learning_rate": 7.077218632292944e-05, "loss": 0.0557, "step": 79090 }, { "epoch": 2.6741945299029717, "grad_norm": 0.6136195659637451, "learning_rate": 7.07646659021785e-05, "loss": 0.0614, "step": 79100 }, { "epoch": 2.6745326075932248, "grad_norm": 0.6028650999069214, "learning_rate": 7.07571449137294e-05, "loss": 0.0626, "step": 79110 }, { "epoch": 2.6748706852834783, "grad_norm": 0.6076703667640686, "learning_rate": 7.074962335778775e-05, "loss": 0.0608, "step": 79120 }, { "epoch": 2.6752087629737313, "grad_norm": 0.49759748578071594, "learning_rate": 7.074210123455918e-05, "loss": 0.0608, "step": 79130 }, { "epoch": 2.6755468406639844, "grad_norm": 0.6895293593406677, "learning_rate": 7.073457854424938e-05, "loss": 0.0646, "step": 79140 }, { "epoch": 2.675884918354238, "grad_norm": 0.9005990624427795, "learning_rate": 7.072705528706396e-05, "loss": 0.0707, "step": 79150 }, { "epoch": 2.676222996044491, "grad_norm": 0.8556886315345764, "learning_rate": 7.071953146320867e-05, "loss": 0.0639, "step": 79160 }, { "epoch": 2.676561073734744, "grad_norm": 0.9086946249008179, "learning_rate": 7.071200707288916e-05, "loss": 0.0613, "step": 79170 }, { "epoch": 2.6768991514249976, "grad_norm": 1.1638215780258179, "learning_rate": 7.070448211631115e-05, "loss": 0.0598, "step": 79180 }, { "epoch": 2.6772372291152506, "grad_norm": 0.8366913795471191, "learning_rate": 7.069695659368038e-05, "loss": 0.0594, "step": 79190 }, { "epoch": 2.6775753068055037, "grad_norm": 0.9206525087356567, "learning_rate": 7.06894305052026e-05, "loss": 0.0676, "step": 79200 }, { "epoch": 2.677913384495757, "grad_norm": 0.7075870633125305, "learning_rate": 7.068190385108356e-05, "loss": 0.0569, "step": 79210 }, { "epoch": 2.6782514621860103, "grad_norm": 1.3648746013641357, "learning_rate": 7.067437663152903e-05, "loss": 0.0546, "step": 79220 }, { "epoch": 2.6785895398762634, "grad_norm": 0.6698693037033081, "learning_rate": 7.066684884674482e-05, "loss": 0.0598, "step": 79230 }, { "epoch": 2.678927617566517, "grad_norm": 0.8757779002189636, "learning_rate": 7.065932049693673e-05, "loss": 0.0567, "step": 79240 }, { "epoch": 2.67926569525677, "grad_norm": 0.6413248777389526, "learning_rate": 7.065179158231058e-05, "loss": 0.0629, "step": 79250 }, { "epoch": 2.679603772947023, "grad_norm": 0.671105146408081, "learning_rate": 7.06442621030722e-05, "loss": 0.059, "step": 79260 }, { "epoch": 2.6799418506372765, "grad_norm": 1.6420975923538208, "learning_rate": 7.063673205942745e-05, "loss": 0.0584, "step": 79270 }, { "epoch": 2.6802799283275296, "grad_norm": 0.8346603512763977, "learning_rate": 7.062920145158219e-05, "loss": 0.0579, "step": 79280 }, { "epoch": 2.6806180060177827, "grad_norm": 0.6447412967681885, "learning_rate": 7.062167027974233e-05, "loss": 0.0583, "step": 79290 }, { "epoch": 2.680956083708036, "grad_norm": 0.691703200340271, "learning_rate": 7.061413854411373e-05, "loss": 0.0571, "step": 79300 }, { "epoch": 2.6812941613982892, "grad_norm": 4.144936561584473, "learning_rate": 7.060660624490234e-05, "loss": 0.055, "step": 79310 }, { "epoch": 2.6816322390885423, "grad_norm": 0.39897432923316956, "learning_rate": 7.059907338231407e-05, "loss": 0.0577, "step": 79320 }, { "epoch": 2.681970316778796, "grad_norm": 1.4391131401062012, "learning_rate": 7.059153995655488e-05, "loss": 0.0647, "step": 79330 }, { "epoch": 2.682308394469049, "grad_norm": 0.4382613003253937, "learning_rate": 7.058400596783072e-05, "loss": 0.0556, "step": 79340 }, { "epoch": 2.682646472159302, "grad_norm": 0.8327407836914062, "learning_rate": 7.057647141634758e-05, "loss": 0.0643, "step": 79350 }, { "epoch": 2.6829845498495555, "grad_norm": 0.6359192728996277, "learning_rate": 7.056893630231143e-05, "loss": 0.0561, "step": 79360 }, { "epoch": 2.6833226275398085, "grad_norm": 1.00132155418396, "learning_rate": 7.056140062592828e-05, "loss": 0.0559, "step": 79370 }, { "epoch": 2.6836607052300616, "grad_norm": 0.8028372526168823, "learning_rate": 7.055386438740418e-05, "loss": 0.0596, "step": 79380 }, { "epoch": 2.683998782920315, "grad_norm": 0.6235703825950623, "learning_rate": 7.054632758694514e-05, "loss": 0.0562, "step": 79390 }, { "epoch": 2.684336860610568, "grad_norm": 0.8623664975166321, "learning_rate": 7.053879022475722e-05, "loss": 0.0584, "step": 79400 }, { "epoch": 2.6846749383008213, "grad_norm": 0.49271655082702637, "learning_rate": 7.05312523010465e-05, "loss": 0.0562, "step": 79410 }, { "epoch": 2.685013015991075, "grad_norm": 0.7432698607444763, "learning_rate": 7.052371381601906e-05, "loss": 0.0527, "step": 79420 }, { "epoch": 2.685351093681328, "grad_norm": 1.5663477182388306, "learning_rate": 7.051617476988098e-05, "loss": 0.058, "step": 79430 }, { "epoch": 2.685689171371581, "grad_norm": 1.132853627204895, "learning_rate": 7.05086351628384e-05, "loss": 0.057, "step": 79440 }, { "epoch": 2.6860272490618344, "grad_norm": 1.4501043558120728, "learning_rate": 7.050109499509742e-05, "loss": 0.0641, "step": 79450 }, { "epoch": 2.6863653267520875, "grad_norm": 1.3158934116363525, "learning_rate": 7.049355426686423e-05, "loss": 0.0613, "step": 79460 }, { "epoch": 2.686703404442341, "grad_norm": 0.5508637428283691, "learning_rate": 7.048601297834495e-05, "loss": 0.0591, "step": 79470 }, { "epoch": 2.687041482132594, "grad_norm": 0.762679934501648, "learning_rate": 7.047847112974578e-05, "loss": 0.0582, "step": 79480 }, { "epoch": 2.687379559822847, "grad_norm": 0.9154617786407471, "learning_rate": 7.04709287212729e-05, "loss": 0.0596, "step": 79490 }, { "epoch": 2.6877176375131007, "grad_norm": 1.2568888664245605, "learning_rate": 7.046338575313253e-05, "loss": 0.0681, "step": 79500 }, { "epoch": 2.6880557152033537, "grad_norm": 0.7662172913551331, "learning_rate": 7.045584222553086e-05, "loss": 0.0633, "step": 79510 }, { "epoch": 2.688393792893607, "grad_norm": 0.4126603901386261, "learning_rate": 7.044829813867416e-05, "loss": 0.0645, "step": 79520 }, { "epoch": 2.6887318705838603, "grad_norm": 0.6413425803184509, "learning_rate": 7.044075349276866e-05, "loss": 0.0672, "step": 79530 }, { "epoch": 2.6890699482741134, "grad_norm": 1.1948161125183105, "learning_rate": 7.043320828802066e-05, "loss": 0.0626, "step": 79540 }, { "epoch": 2.6894080259643665, "grad_norm": 0.5425118803977966, "learning_rate": 7.042566252463641e-05, "loss": 0.0614, "step": 79550 }, { "epoch": 2.68974610365462, "grad_norm": 0.6661133766174316, "learning_rate": 7.041811620282223e-05, "loss": 0.0618, "step": 79560 }, { "epoch": 2.690084181344873, "grad_norm": 1.0035345554351807, "learning_rate": 7.04105693227844e-05, "loss": 0.0631, "step": 79570 }, { "epoch": 2.690422259035126, "grad_norm": 0.6516170501708984, "learning_rate": 7.04030218847293e-05, "loss": 0.0592, "step": 79580 }, { "epoch": 2.6907603367253796, "grad_norm": 0.8929372429847717, "learning_rate": 7.039547388886322e-05, "loss": 0.0614, "step": 79590 }, { "epoch": 2.6910984144156327, "grad_norm": 0.710709810256958, "learning_rate": 7.038792533539255e-05, "loss": 0.0661, "step": 79600 }, { "epoch": 2.691436492105886, "grad_norm": 0.512667715549469, "learning_rate": 7.038037622452368e-05, "loss": 0.0669, "step": 79610 }, { "epoch": 2.6917745697961393, "grad_norm": 0.8116409182548523, "learning_rate": 7.037282655646297e-05, "loss": 0.0631, "step": 79620 }, { "epoch": 2.6921126474863923, "grad_norm": 0.5636340975761414, "learning_rate": 7.036527633141684e-05, "loss": 0.0598, "step": 79630 }, { "epoch": 2.692450725176646, "grad_norm": 1.1811292171478271, "learning_rate": 7.03577255495917e-05, "loss": 0.0691, "step": 79640 }, { "epoch": 2.692788802866899, "grad_norm": 0.5083993673324585, "learning_rate": 7.035017421119399e-05, "loss": 0.062, "step": 79650 }, { "epoch": 2.693126880557152, "grad_norm": 2.06071400642395, "learning_rate": 7.034262231643016e-05, "loss": 0.0565, "step": 79660 }, { "epoch": 2.6934649582474055, "grad_norm": 0.6968011260032654, "learning_rate": 7.033506986550669e-05, "loss": 0.0678, "step": 79670 }, { "epoch": 2.6938030359376586, "grad_norm": 0.5290502905845642, "learning_rate": 7.032751685863004e-05, "loss": 0.0612, "step": 79680 }, { "epoch": 2.6941411136279116, "grad_norm": 0.7382467985153198, "learning_rate": 7.031996329600673e-05, "loss": 0.0662, "step": 79690 }, { "epoch": 2.694479191318165, "grad_norm": 0.5455809831619263, "learning_rate": 7.031240917784323e-05, "loss": 0.0586, "step": 79700 }, { "epoch": 2.6948172690084182, "grad_norm": 0.5176076292991638, "learning_rate": 7.030485450434612e-05, "loss": 0.0609, "step": 79710 }, { "epoch": 2.6951553466986713, "grad_norm": 0.7857068777084351, "learning_rate": 7.02972992757219e-05, "loss": 0.0624, "step": 79720 }, { "epoch": 2.695493424388925, "grad_norm": 0.6516279578208923, "learning_rate": 7.028974349217716e-05, "loss": 0.0601, "step": 79730 }, { "epoch": 2.695831502079178, "grad_norm": 1.060985803604126, "learning_rate": 7.028218715391845e-05, "loss": 0.0602, "step": 79740 }, { "epoch": 2.696169579769431, "grad_norm": 0.6624082922935486, "learning_rate": 7.027463026115235e-05, "loss": 0.0622, "step": 79750 }, { "epoch": 2.6965076574596845, "grad_norm": 0.6251511573791504, "learning_rate": 7.026707281408549e-05, "loss": 0.0635, "step": 79760 }, { "epoch": 2.6968457351499375, "grad_norm": 0.5934754014015198, "learning_rate": 7.025951481292446e-05, "loss": 0.066, "step": 79770 }, { "epoch": 2.6971838128401906, "grad_norm": 1.1162360906600952, "learning_rate": 7.02519562578759e-05, "loss": 0.0632, "step": 79780 }, { "epoch": 2.697521890530444, "grad_norm": 0.6504145860671997, "learning_rate": 7.02443971491465e-05, "loss": 0.0586, "step": 79790 }, { "epoch": 2.697859968220697, "grad_norm": 0.7673811316490173, "learning_rate": 7.023683748694286e-05, "loss": 0.0571, "step": 79800 }, { "epoch": 2.6981980459109502, "grad_norm": 0.9900097846984863, "learning_rate": 7.022927727147169e-05, "loss": 0.0678, "step": 79810 }, { "epoch": 2.6985361236012038, "grad_norm": 1.246661901473999, "learning_rate": 7.022171650293968e-05, "loss": 0.0579, "step": 79820 }, { "epoch": 2.698874201291457, "grad_norm": 0.4865904748439789, "learning_rate": 7.021415518155354e-05, "loss": 0.0542, "step": 79830 }, { "epoch": 2.69921227898171, "grad_norm": 0.9826365113258362, "learning_rate": 7.020659330751997e-05, "loss": 0.0606, "step": 79840 }, { "epoch": 2.6995503566719634, "grad_norm": 0.8451858162879944, "learning_rate": 7.019903088104576e-05, "loss": 0.0626, "step": 79850 }, { "epoch": 2.6998884343622165, "grad_norm": 0.6808516979217529, "learning_rate": 7.019146790233762e-05, "loss": 0.073, "step": 79860 }, { "epoch": 2.7002265120524696, "grad_norm": 0.8369019031524658, "learning_rate": 7.018390437160233e-05, "loss": 0.0622, "step": 79870 }, { "epoch": 2.700564589742723, "grad_norm": 0.7882893085479736, "learning_rate": 7.01763402890467e-05, "loss": 0.0611, "step": 79880 }, { "epoch": 2.700902667432976, "grad_norm": 0.6854357123374939, "learning_rate": 7.016877565487747e-05, "loss": 0.0625, "step": 79890 }, { "epoch": 2.701240745123229, "grad_norm": 0.8051100373268127, "learning_rate": 7.016121046930151e-05, "loss": 0.0589, "step": 79900 }, { "epoch": 2.7015788228134827, "grad_norm": 0.7544891834259033, "learning_rate": 7.015364473252563e-05, "loss": 0.0579, "step": 79910 }, { "epoch": 2.701916900503736, "grad_norm": 1.0344922542572021, "learning_rate": 7.014607844475669e-05, "loss": 0.0663, "step": 79920 }, { "epoch": 2.702254978193989, "grad_norm": 0.5124989748001099, "learning_rate": 7.013851160620152e-05, "loss": 0.0591, "step": 79930 }, { "epoch": 2.7025930558842424, "grad_norm": 0.5236947536468506, "learning_rate": 7.013094421706701e-05, "loss": 0.0611, "step": 79940 }, { "epoch": 2.7029311335744954, "grad_norm": 0.42497506737709045, "learning_rate": 7.012337627756005e-05, "loss": 0.061, "step": 79950 }, { "epoch": 2.7032692112647485, "grad_norm": 0.7216018438339233, "learning_rate": 7.011580778788754e-05, "loss": 0.0624, "step": 79960 }, { "epoch": 2.703607288955002, "grad_norm": 1.0984938144683838, "learning_rate": 7.010823874825638e-05, "loss": 0.0581, "step": 79970 }, { "epoch": 2.703945366645255, "grad_norm": 0.36157944798469543, "learning_rate": 7.010066915887356e-05, "loss": 0.0609, "step": 79980 }, { "epoch": 2.704283444335508, "grad_norm": 0.6449263095855713, "learning_rate": 7.009309901994598e-05, "loss": 0.0648, "step": 79990 }, { "epoch": 2.7046215220257617, "grad_norm": 0.47833314538002014, "learning_rate": 7.008552833168063e-05, "loss": 0.0629, "step": 80000 }, { "epoch": 2.7049595997160147, "grad_norm": 0.5852586030960083, "learning_rate": 7.007795709428447e-05, "loss": 0.0622, "step": 80010 }, { "epoch": 2.705297677406268, "grad_norm": 0.5416476130485535, "learning_rate": 7.007038530796451e-05, "loss": 0.0693, "step": 80020 }, { "epoch": 2.7056357550965213, "grad_norm": 0.9951139688491821, "learning_rate": 7.006281297292775e-05, "loss": 0.0614, "step": 80030 }, { "epoch": 2.7059738327867744, "grad_norm": 0.7022889852523804, "learning_rate": 7.005524008938124e-05, "loss": 0.063, "step": 80040 }, { "epoch": 2.7063119104770275, "grad_norm": 0.7503043413162231, "learning_rate": 7.004766665753198e-05, "loss": 0.058, "step": 80050 }, { "epoch": 2.706649988167281, "grad_norm": 1.1439125537872314, "learning_rate": 7.004009267758705e-05, "loss": 0.0621, "step": 80060 }, { "epoch": 2.706988065857534, "grad_norm": 0.6653104424476624, "learning_rate": 7.003251814975353e-05, "loss": 0.0601, "step": 80070 }, { "epoch": 2.707326143547787, "grad_norm": 0.5915131568908691, "learning_rate": 7.002494307423849e-05, "loss": 0.0583, "step": 80080 }, { "epoch": 2.7076642212380406, "grad_norm": 1.184659481048584, "learning_rate": 7.001736745124901e-05, "loss": 0.0637, "step": 80090 }, { "epoch": 2.7080022989282937, "grad_norm": 0.9579296112060547, "learning_rate": 7.000979128099223e-05, "loss": 0.0595, "step": 80100 }, { "epoch": 2.7083403766185468, "grad_norm": 0.5676302909851074, "learning_rate": 7.000221456367527e-05, "loss": 0.0644, "step": 80110 }, { "epoch": 2.7086784543088003, "grad_norm": 2.2646148204803467, "learning_rate": 6.999463729950528e-05, "loss": 0.0602, "step": 80120 }, { "epoch": 2.7090165319990533, "grad_norm": 0.5006230473518372, "learning_rate": 6.998705948868944e-05, "loss": 0.0666, "step": 80130 }, { "epoch": 2.7093546096893064, "grad_norm": 1.21571946144104, "learning_rate": 6.997948113143488e-05, "loss": 0.0592, "step": 80140 }, { "epoch": 2.70969268737956, "grad_norm": 1.0203486680984497, "learning_rate": 6.99719022279488e-05, "loss": 0.0702, "step": 80150 }, { "epoch": 2.710030765069813, "grad_norm": 0.5282381772994995, "learning_rate": 6.996432277843843e-05, "loss": 0.0619, "step": 80160 }, { "epoch": 2.710368842760066, "grad_norm": 1.7548543214797974, "learning_rate": 6.9956742783111e-05, "loss": 0.0596, "step": 80170 }, { "epoch": 2.7107069204503196, "grad_norm": 0.4999825358390808, "learning_rate": 6.994916224217368e-05, "loss": 0.0533, "step": 80180 }, { "epoch": 2.7110449981405726, "grad_norm": 0.7544301748275757, "learning_rate": 6.99415811558338e-05, "loss": 0.0593, "step": 80190 }, { "epoch": 2.7113830758308257, "grad_norm": 0.46052587032318115, "learning_rate": 6.993399952429854e-05, "loss": 0.0655, "step": 80200 }, { "epoch": 2.7117211535210792, "grad_norm": 0.9881262183189392, "learning_rate": 6.992641734777524e-05, "loss": 0.0604, "step": 80210 }, { "epoch": 2.7120592312113323, "grad_norm": 0.48546990752220154, "learning_rate": 6.991883462647117e-05, "loss": 0.0559, "step": 80220 }, { "epoch": 2.7123973089015854, "grad_norm": 0.773314893245697, "learning_rate": 6.991125136059366e-05, "loss": 0.058, "step": 80230 }, { "epoch": 2.712735386591839, "grad_norm": 1.3672001361846924, "learning_rate": 6.990366755035001e-05, "loss": 0.0605, "step": 80240 }, { "epoch": 2.713073464282092, "grad_norm": 0.9003579616546631, "learning_rate": 6.989608319594756e-05, "loss": 0.0539, "step": 80250 }, { "epoch": 2.713411541972345, "grad_norm": 1.0981851816177368, "learning_rate": 6.988849829759369e-05, "loss": 0.0615, "step": 80260 }, { "epoch": 2.7137496196625985, "grad_norm": 1.0380593538284302, "learning_rate": 6.988091285549573e-05, "loss": 0.0549, "step": 80270 }, { "epoch": 2.7140876973528516, "grad_norm": 0.45102620124816895, "learning_rate": 6.987332686986108e-05, "loss": 0.0567, "step": 80280 }, { "epoch": 2.7144257750431047, "grad_norm": 0.7066352367401123, "learning_rate": 6.986574034089714e-05, "loss": 0.0563, "step": 80290 }, { "epoch": 2.714763852733358, "grad_norm": 0.602484941482544, "learning_rate": 6.985815326881133e-05, "loss": 0.0606, "step": 80300 }, { "epoch": 2.7151019304236113, "grad_norm": 0.6678237318992615, "learning_rate": 6.985056565381107e-05, "loss": 0.059, "step": 80310 }, { "epoch": 2.7154400081138643, "grad_norm": 0.9295937418937683, "learning_rate": 6.98429774961038e-05, "loss": 0.0603, "step": 80320 }, { "epoch": 2.715778085804118, "grad_norm": 0.5972285866737366, "learning_rate": 6.983538879589698e-05, "loss": 0.0546, "step": 80330 }, { "epoch": 2.716116163494371, "grad_norm": 0.4953761696815491, "learning_rate": 6.982779955339808e-05, "loss": 0.0621, "step": 80340 }, { "epoch": 2.716454241184624, "grad_norm": 0.5117977261543274, "learning_rate": 6.982020976881459e-05, "loss": 0.059, "step": 80350 }, { "epoch": 2.7167923188748775, "grad_norm": 1.0383667945861816, "learning_rate": 6.981261944235402e-05, "loss": 0.061, "step": 80360 }, { "epoch": 2.7171303965651306, "grad_norm": 0.7556383013725281, "learning_rate": 6.980502857422387e-05, "loss": 0.0548, "step": 80370 }, { "epoch": 2.7174684742553836, "grad_norm": 0.4152728021144867, "learning_rate": 6.979743716463168e-05, "loss": 0.0599, "step": 80380 }, { "epoch": 2.717806551945637, "grad_norm": 0.6331555843353271, "learning_rate": 6.9789845213785e-05, "loss": 0.0547, "step": 80390 }, { "epoch": 2.71814462963589, "grad_norm": 0.6726093292236328, "learning_rate": 6.978225272189138e-05, "loss": 0.0614, "step": 80400 }, { "epoch": 2.7184827073261433, "grad_norm": 0.662964403629303, "learning_rate": 6.977465968915839e-05, "loss": 0.0598, "step": 80410 }, { "epoch": 2.718820785016397, "grad_norm": 0.8984435200691223, "learning_rate": 6.976706611579365e-05, "loss": 0.0619, "step": 80420 }, { "epoch": 2.71915886270665, "grad_norm": 1.129873514175415, "learning_rate": 6.975947200200475e-05, "loss": 0.0591, "step": 80430 }, { "epoch": 2.7194969403969034, "grad_norm": 0.6217281222343445, "learning_rate": 6.975187734799931e-05, "loss": 0.0601, "step": 80440 }, { "epoch": 2.7198350180871564, "grad_norm": 0.7562757730484009, "learning_rate": 6.974428215398495e-05, "loss": 0.0616, "step": 80450 }, { "epoch": 2.7201730957774095, "grad_norm": 0.977419912815094, "learning_rate": 6.973668642016935e-05, "loss": 0.0534, "step": 80460 }, { "epoch": 2.720511173467663, "grad_norm": 0.8659193515777588, "learning_rate": 6.972909014676013e-05, "loss": 0.0577, "step": 80470 }, { "epoch": 2.720849251157916, "grad_norm": 1.0492831468582153, "learning_rate": 6.972149333396502e-05, "loss": 0.0583, "step": 80480 }, { "epoch": 2.721187328848169, "grad_norm": 1.2280728816986084, "learning_rate": 6.97138959819917e-05, "loss": 0.062, "step": 80490 }, { "epoch": 2.7215254065384227, "grad_norm": 0.4182039499282837, "learning_rate": 6.970629809104784e-05, "loss": 0.0542, "step": 80500 }, { "epoch": 2.7218634842286757, "grad_norm": 1.0299478769302368, "learning_rate": 6.969869966134123e-05, "loss": 0.0595, "step": 80510 }, { "epoch": 2.722201561918929, "grad_norm": 1.006932258605957, "learning_rate": 6.969110069307955e-05, "loss": 0.0648, "step": 80520 }, { "epoch": 2.7225396396091823, "grad_norm": 0.6501926183700562, "learning_rate": 6.968350118647057e-05, "loss": 0.0568, "step": 80530 }, { "epoch": 2.7228777172994354, "grad_norm": 1.807374119758606, "learning_rate": 6.967590114172207e-05, "loss": 0.059, "step": 80540 }, { "epoch": 2.7232157949896885, "grad_norm": 0.7325424551963806, "learning_rate": 6.966830055904183e-05, "loss": 0.059, "step": 80550 }, { "epoch": 2.723553872679942, "grad_norm": 1.1819576025009155, "learning_rate": 6.966069943863763e-05, "loss": 0.0602, "step": 80560 }, { "epoch": 2.723891950370195, "grad_norm": 0.9425497651100159, "learning_rate": 6.965309778071731e-05, "loss": 0.0547, "step": 80570 }, { "epoch": 2.7242300280604486, "grad_norm": 0.7498512864112854, "learning_rate": 6.964549558548866e-05, "loss": 0.0639, "step": 80580 }, { "epoch": 2.7245681057507016, "grad_norm": 1.2411788702011108, "learning_rate": 6.963789285315954e-05, "loss": 0.0614, "step": 80590 }, { "epoch": 2.7249061834409547, "grad_norm": 1.1999918222427368, "learning_rate": 6.963028958393783e-05, "loss": 0.0564, "step": 80600 }, { "epoch": 2.725244261131208, "grad_norm": 0.572547197341919, "learning_rate": 6.962268577803137e-05, "loss": 0.0537, "step": 80610 }, { "epoch": 2.7255823388214613, "grad_norm": 1.6985365152359009, "learning_rate": 6.961508143564805e-05, "loss": 0.0646, "step": 80620 }, { "epoch": 2.7259204165117144, "grad_norm": 0.49763426184654236, "learning_rate": 6.960747655699578e-05, "loss": 0.059, "step": 80630 }, { "epoch": 2.726258494201968, "grad_norm": 1.3538233041763306, "learning_rate": 6.959987114228245e-05, "loss": 0.0579, "step": 80640 }, { "epoch": 2.726596571892221, "grad_norm": 0.9793647527694702, "learning_rate": 6.959226519171602e-05, "loss": 0.0591, "step": 80650 }, { "epoch": 2.726934649582474, "grad_norm": 0.749593198299408, "learning_rate": 6.95846587055044e-05, "loss": 0.0594, "step": 80660 }, { "epoch": 2.7272727272727275, "grad_norm": 1.0364105701446533, "learning_rate": 6.957705168385559e-05, "loss": 0.0697, "step": 80670 }, { "epoch": 2.7276108049629806, "grad_norm": 0.6102093458175659, "learning_rate": 6.956944412697753e-05, "loss": 0.0549, "step": 80680 }, { "epoch": 2.7279488826532337, "grad_norm": 0.5618035793304443, "learning_rate": 6.956183603507824e-05, "loss": 0.0642, "step": 80690 }, { "epoch": 2.728286960343487, "grad_norm": 0.54581218957901, "learning_rate": 6.955422740836568e-05, "loss": 0.0565, "step": 80700 }, { "epoch": 2.7286250380337402, "grad_norm": 0.48231810331344604, "learning_rate": 6.954661824704789e-05, "loss": 0.0622, "step": 80710 }, { "epoch": 2.7289631157239933, "grad_norm": 1.9159965515136719, "learning_rate": 6.95390085513329e-05, "loss": 0.0563, "step": 80720 }, { "epoch": 2.729301193414247, "grad_norm": 0.850035548210144, "learning_rate": 6.953139832142876e-05, "loss": 0.0591, "step": 80730 }, { "epoch": 2.7296392711045, "grad_norm": 1.3553121089935303, "learning_rate": 6.952378755754353e-05, "loss": 0.0579, "step": 80740 }, { "epoch": 2.729977348794753, "grad_norm": 0.48505493998527527, "learning_rate": 6.951617625988528e-05, "loss": 0.059, "step": 80750 }, { "epoch": 2.7303154264850065, "grad_norm": 1.3788280487060547, "learning_rate": 6.950856442866212e-05, "loss": 0.0589, "step": 80760 }, { "epoch": 2.7306535041752595, "grad_norm": 0.7125118374824524, "learning_rate": 6.950095206408211e-05, "loss": 0.0509, "step": 80770 }, { "epoch": 2.7309915818655126, "grad_norm": 0.8499259948730469, "learning_rate": 6.949333916635342e-05, "loss": 0.0562, "step": 80780 }, { "epoch": 2.731329659555766, "grad_norm": 2.153278112411499, "learning_rate": 6.948572573568414e-05, "loss": 0.0539, "step": 80790 }, { "epoch": 2.731667737246019, "grad_norm": 0.688831627368927, "learning_rate": 6.947811177228244e-05, "loss": 0.0631, "step": 80800 }, { "epoch": 2.7320058149362723, "grad_norm": 0.9451429843902588, "learning_rate": 6.947049727635651e-05, "loss": 0.0623, "step": 80810 }, { "epoch": 2.7323438926265258, "grad_norm": 0.7317196130752563, "learning_rate": 6.946288224811448e-05, "loss": 0.0563, "step": 80820 }, { "epoch": 2.732681970316779, "grad_norm": 0.5739426612854004, "learning_rate": 6.945526668776456e-05, "loss": 0.0565, "step": 80830 }, { "epoch": 2.733020048007032, "grad_norm": 0.677989661693573, "learning_rate": 6.944765059551496e-05, "loss": 0.0625, "step": 80840 }, { "epoch": 2.7333581256972854, "grad_norm": 1.046613097190857, "learning_rate": 6.944003397157388e-05, "loss": 0.0647, "step": 80850 }, { "epoch": 2.7336962033875385, "grad_norm": 0.7924309372901917, "learning_rate": 6.94324168161496e-05, "loss": 0.062, "step": 80860 }, { "epoch": 2.7340342810777916, "grad_norm": 1.3072165250778198, "learning_rate": 6.942479912945033e-05, "loss": 0.0597, "step": 80870 }, { "epoch": 2.734372358768045, "grad_norm": 0.7518268823623657, "learning_rate": 6.941718091168436e-05, "loss": 0.0614, "step": 80880 }, { "epoch": 2.734710436458298, "grad_norm": 1.5996581315994263, "learning_rate": 6.940956216305996e-05, "loss": 0.0636, "step": 80890 }, { "epoch": 2.735048514148551, "grad_norm": 0.8771799802780151, "learning_rate": 6.940194288378543e-05, "loss": 0.0611, "step": 80900 }, { "epoch": 2.7353865918388047, "grad_norm": 0.6693007349967957, "learning_rate": 6.939432307406904e-05, "loss": 0.0625, "step": 80910 }, { "epoch": 2.735724669529058, "grad_norm": 0.749671459197998, "learning_rate": 6.938670273411916e-05, "loss": 0.0672, "step": 80920 }, { "epoch": 2.736062747219311, "grad_norm": 1.0531120300292969, "learning_rate": 6.937908186414411e-05, "loss": 0.06, "step": 80930 }, { "epoch": 2.7364008249095644, "grad_norm": 1.2953543663024902, "learning_rate": 6.937146046435226e-05, "loss": 0.0682, "step": 80940 }, { "epoch": 2.7367389025998174, "grad_norm": 0.9084087610244751, "learning_rate": 6.936383853495194e-05, "loss": 0.0609, "step": 80950 }, { "epoch": 2.7370769802900705, "grad_norm": 0.8689128160476685, "learning_rate": 6.935621607615155e-05, "loss": 0.0606, "step": 80960 }, { "epoch": 2.737415057980324, "grad_norm": 1.5214695930480957, "learning_rate": 6.934859308815949e-05, "loss": 0.0624, "step": 80970 }, { "epoch": 2.737753135670577, "grad_norm": 1.3142311573028564, "learning_rate": 6.934096957118416e-05, "loss": 0.0603, "step": 80980 }, { "epoch": 2.73809121336083, "grad_norm": 0.8187546133995056, "learning_rate": 6.933334552543398e-05, "loss": 0.0608, "step": 80990 }, { "epoch": 2.7384292910510837, "grad_norm": 0.436417818069458, "learning_rate": 6.93257209511174e-05, "loss": 0.0637, "step": 81000 }, { "epoch": 2.7387673687413367, "grad_norm": 0.431591659784317, "learning_rate": 6.931809584844287e-05, "loss": 0.0604, "step": 81010 }, { "epoch": 2.73910544643159, "grad_norm": 2.767024040222168, "learning_rate": 6.931047021761886e-05, "loss": 0.0694, "step": 81020 }, { "epoch": 2.7394435241218433, "grad_norm": 0.6818819642066956, "learning_rate": 6.930284405885385e-05, "loss": 0.0581, "step": 81030 }, { "epoch": 2.7397816018120964, "grad_norm": 1.854690670967102, "learning_rate": 6.929521737235634e-05, "loss": 0.0582, "step": 81040 }, { "epoch": 2.7401196795023495, "grad_norm": 1.1116411685943604, "learning_rate": 6.928759015833483e-05, "loss": 0.061, "step": 81050 }, { "epoch": 2.740457757192603, "grad_norm": 0.8272482752799988, "learning_rate": 6.927996241699785e-05, "loss": 0.0651, "step": 81060 }, { "epoch": 2.740795834882856, "grad_norm": 1.357516884803772, "learning_rate": 6.927233414855394e-05, "loss": 0.0664, "step": 81070 }, { "epoch": 2.741133912573109, "grad_norm": 1.2547111511230469, "learning_rate": 6.926470535321166e-05, "loss": 0.0512, "step": 81080 }, { "epoch": 2.7414719902633626, "grad_norm": 1.107461929321289, "learning_rate": 6.925707603117958e-05, "loss": 0.0598, "step": 81090 }, { "epoch": 2.7418100679536157, "grad_norm": 0.8268128037452698, "learning_rate": 6.924944618266624e-05, "loss": 0.0605, "step": 81100 }, { "epoch": 2.7421481456438688, "grad_norm": 0.9973679184913635, "learning_rate": 6.924181580788031e-05, "loss": 0.0558, "step": 81110 }, { "epoch": 2.7424862233341223, "grad_norm": 0.7862439751625061, "learning_rate": 6.923418490703036e-05, "loss": 0.0554, "step": 81120 }, { "epoch": 2.7428243010243754, "grad_norm": 0.652726411819458, "learning_rate": 6.922655348032501e-05, "loss": 0.0581, "step": 81130 }, { "epoch": 2.7431623787146284, "grad_norm": 0.5270068645477295, "learning_rate": 6.921892152797291e-05, "loss": 0.0617, "step": 81140 }, { "epoch": 2.743500456404882, "grad_norm": 0.9018769860267639, "learning_rate": 6.921128905018274e-05, "loss": 0.0591, "step": 81150 }, { "epoch": 2.743838534095135, "grad_norm": 0.6161333322525024, "learning_rate": 6.920365604716311e-05, "loss": 0.064, "step": 81160 }, { "epoch": 2.744176611785388, "grad_norm": 1.086464762687683, "learning_rate": 6.919602251912276e-05, "loss": 0.0594, "step": 81170 }, { "epoch": 2.7445146894756416, "grad_norm": 0.8030464053153992, "learning_rate": 6.918838846627034e-05, "loss": 0.0577, "step": 81180 }, { "epoch": 2.7448527671658947, "grad_norm": 0.724375307559967, "learning_rate": 6.918075388881461e-05, "loss": 0.0624, "step": 81190 }, { "epoch": 2.7451908448561477, "grad_norm": 0.8112912178039551, "learning_rate": 6.917311878696428e-05, "loss": 0.0596, "step": 81200 }, { "epoch": 2.7455289225464012, "grad_norm": 0.6449520587921143, "learning_rate": 6.916548316092809e-05, "loss": 0.0633, "step": 81210 }, { "epoch": 2.7458670002366543, "grad_norm": 3.845759391784668, "learning_rate": 6.915784701091476e-05, "loss": 0.0568, "step": 81220 }, { "epoch": 2.7462050779269074, "grad_norm": 0.8393079042434692, "learning_rate": 6.915021033713311e-05, "loss": 0.0667, "step": 81230 }, { "epoch": 2.746543155617161, "grad_norm": 0.5403466820716858, "learning_rate": 6.914257313979192e-05, "loss": 0.0549, "step": 81240 }, { "epoch": 2.746881233307414, "grad_norm": 1.172938585281372, "learning_rate": 6.913493541909995e-05, "loss": 0.062, "step": 81250 }, { "epoch": 2.747219310997667, "grad_norm": 0.6316571235656738, "learning_rate": 6.912729717526602e-05, "loss": 0.0593, "step": 81260 }, { "epoch": 2.7475573886879205, "grad_norm": 1.9744583368301392, "learning_rate": 6.9119658408499e-05, "loss": 0.0629, "step": 81270 }, { "epoch": 2.7478954663781736, "grad_norm": 0.686037003993988, "learning_rate": 6.91120191190077e-05, "loss": 0.0559, "step": 81280 }, { "epoch": 2.7482335440684267, "grad_norm": 0.919161319732666, "learning_rate": 6.910437930700095e-05, "loss": 0.0648, "step": 81290 }, { "epoch": 2.74857162175868, "grad_norm": 0.7498604655265808, "learning_rate": 6.909673897268767e-05, "loss": 0.0533, "step": 81300 }, { "epoch": 2.7489096994489333, "grad_norm": 1.0430623292922974, "learning_rate": 6.90890981162767e-05, "loss": 0.0641, "step": 81310 }, { "epoch": 2.7492477771391863, "grad_norm": 0.9892024397850037, "learning_rate": 6.908145673797699e-05, "loss": 0.057, "step": 81320 }, { "epoch": 2.74958585482944, "grad_norm": 1.580163598060608, "learning_rate": 6.90738148379974e-05, "loss": 0.0575, "step": 81330 }, { "epoch": 2.749923932519693, "grad_norm": 1.410193920135498, "learning_rate": 6.906617241654689e-05, "loss": 0.0603, "step": 81340 }, { "epoch": 2.750262010209946, "grad_norm": 0.6997145414352417, "learning_rate": 6.90585294738344e-05, "loss": 0.0642, "step": 81350 }, { "epoch": 2.7506000879001995, "grad_norm": 1.6019985675811768, "learning_rate": 6.905088601006884e-05, "loss": 0.0624, "step": 81360 }, { "epoch": 2.7509381655904526, "grad_norm": 1.3160450458526611, "learning_rate": 6.904324202545923e-05, "loss": 0.069, "step": 81370 }, { "epoch": 2.7512762432807056, "grad_norm": 1.2408159971237183, "learning_rate": 6.903559752021455e-05, "loss": 0.0587, "step": 81380 }, { "epoch": 2.751614320970959, "grad_norm": 1.0345045328140259, "learning_rate": 6.902795249454377e-05, "loss": 0.0622, "step": 81390 }, { "epoch": 2.751952398661212, "grad_norm": 1.8202039003372192, "learning_rate": 6.902030694865592e-05, "loss": 0.0631, "step": 81400 }, { "epoch": 2.7522904763514657, "grad_norm": 1.1963838338851929, "learning_rate": 6.901266088276003e-05, "loss": 0.0623, "step": 81410 }, { "epoch": 2.752628554041719, "grad_norm": 1.4800474643707275, "learning_rate": 6.900501429706511e-05, "loss": 0.0587, "step": 81420 }, { "epoch": 2.752966631731972, "grad_norm": 3.0571999549865723, "learning_rate": 6.899736719178025e-05, "loss": 0.055, "step": 81430 }, { "epoch": 2.7533047094222254, "grad_norm": 2.0829389095306396, "learning_rate": 6.898971956711452e-05, "loss": 0.0641, "step": 81440 }, { "epoch": 2.7536427871124785, "grad_norm": 0.6128553748130798, "learning_rate": 6.8982071423277e-05, "loss": 0.0632, "step": 81450 }, { "epoch": 2.7539808648027315, "grad_norm": 1.0242670774459839, "learning_rate": 6.897442276047677e-05, "loss": 0.0601, "step": 81460 }, { "epoch": 2.754318942492985, "grad_norm": 0.732150673866272, "learning_rate": 6.896677357892294e-05, "loss": 0.0588, "step": 81470 }, { "epoch": 2.754657020183238, "grad_norm": 1.0460941791534424, "learning_rate": 6.895912387882465e-05, "loss": 0.065, "step": 81480 }, { "epoch": 2.754995097873491, "grad_norm": 0.5532605051994324, "learning_rate": 6.895147366039103e-05, "loss": 0.0526, "step": 81490 }, { "epoch": 2.7553331755637447, "grad_norm": 0.8528335690498352, "learning_rate": 6.894382292383126e-05, "loss": 0.0612, "step": 81500 }, { "epoch": 2.7556712532539978, "grad_norm": 1.2458081245422363, "learning_rate": 6.893617166935446e-05, "loss": 0.0585, "step": 81510 }, { "epoch": 2.756009330944251, "grad_norm": 0.40799087285995483, "learning_rate": 6.892851989716986e-05, "loss": 0.058, "step": 81520 }, { "epoch": 2.7563474086345043, "grad_norm": 0.8966576457023621, "learning_rate": 6.892086760748664e-05, "loss": 0.056, "step": 81530 }, { "epoch": 2.7566854863247574, "grad_norm": 0.4319189786911011, "learning_rate": 6.891321480051399e-05, "loss": 0.0577, "step": 81540 }, { "epoch": 2.757023564015011, "grad_norm": 0.621278703212738, "learning_rate": 6.890556147646117e-05, "loss": 0.0591, "step": 81550 }, { "epoch": 2.757361641705264, "grad_norm": 0.6217780709266663, "learning_rate": 6.889790763553738e-05, "loss": 0.0535, "step": 81560 }, { "epoch": 2.757699719395517, "grad_norm": 0.3617999851703644, "learning_rate": 6.889025327795191e-05, "loss": 0.0584, "step": 81570 }, { "epoch": 2.7580377970857706, "grad_norm": 0.6141430139541626, "learning_rate": 6.888259840391402e-05, "loss": 0.0625, "step": 81580 }, { "epoch": 2.7583758747760236, "grad_norm": 3.3065619468688965, "learning_rate": 6.887494301363299e-05, "loss": 0.0558, "step": 81590 }, { "epoch": 2.7587139524662767, "grad_norm": 0.5049740076065063, "learning_rate": 6.88672871073181e-05, "loss": 0.0575, "step": 81600 }, { "epoch": 2.75905203015653, "grad_norm": 0.5148761868476868, "learning_rate": 6.885963068517864e-05, "loss": 0.0581, "step": 81610 }, { "epoch": 2.7593901078467833, "grad_norm": 9.137247085571289, "learning_rate": 6.885197374742398e-05, "loss": 0.0599, "step": 81620 }, { "epoch": 2.7597281855370364, "grad_norm": 0.6813557744026184, "learning_rate": 6.884431629426345e-05, "loss": 0.0605, "step": 81630 }, { "epoch": 2.76006626322729, "grad_norm": 2.199441909790039, "learning_rate": 6.883665832590639e-05, "loss": 0.0649, "step": 81640 }, { "epoch": 2.760404340917543, "grad_norm": 0.927905797958374, "learning_rate": 6.882899984256216e-05, "loss": 0.0561, "step": 81650 }, { "epoch": 2.760742418607796, "grad_norm": 1.0097405910491943, "learning_rate": 6.882134084444017e-05, "loss": 0.0542, "step": 81660 }, { "epoch": 2.7610804962980495, "grad_norm": 0.5623427033424377, "learning_rate": 6.881368133174977e-05, "loss": 0.0552, "step": 81670 }, { "epoch": 2.7614185739883026, "grad_norm": 0.4466843605041504, "learning_rate": 6.880602130470039e-05, "loss": 0.0575, "step": 81680 }, { "epoch": 2.7617566516785557, "grad_norm": 1.4035272598266602, "learning_rate": 6.879836076350146e-05, "loss": 0.056, "step": 81690 }, { "epoch": 2.762094729368809, "grad_norm": 1.3678277730941772, "learning_rate": 6.87906997083624e-05, "loss": 0.0625, "step": 81700 }, { "epoch": 2.7624328070590622, "grad_norm": 0.8282526135444641, "learning_rate": 6.87830381394927e-05, "loss": 0.0623, "step": 81710 }, { "epoch": 2.7627708847493153, "grad_norm": 0.6692232489585876, "learning_rate": 6.877537605710176e-05, "loss": 0.0581, "step": 81720 }, { "epoch": 2.763108962439569, "grad_norm": 9.084258079528809, "learning_rate": 6.876771346139909e-05, "loss": 0.0645, "step": 81730 }, { "epoch": 2.763447040129822, "grad_norm": 1.2499375343322754, "learning_rate": 6.876005035259419e-05, "loss": 0.0578, "step": 81740 }, { "epoch": 2.763785117820075, "grad_norm": 0.7370465993881226, "learning_rate": 6.87523867308966e-05, "loss": 0.0572, "step": 81750 }, { "epoch": 2.7641231955103285, "grad_norm": 1.961175560951233, "learning_rate": 6.874472259651575e-05, "loss": 0.0571, "step": 81760 }, { "epoch": 2.7644612732005815, "grad_norm": 1.012095332145691, "learning_rate": 6.873705794966124e-05, "loss": 0.0604, "step": 81770 }, { "epoch": 2.7647993508908346, "grad_norm": 2.509416103363037, "learning_rate": 6.872939279054263e-05, "loss": 0.0612, "step": 81780 }, { "epoch": 2.765137428581088, "grad_norm": 0.8691924214363098, "learning_rate": 6.872172711936943e-05, "loss": 0.0619, "step": 81790 }, { "epoch": 2.765475506271341, "grad_norm": 0.9206948280334473, "learning_rate": 6.871406093635124e-05, "loss": 0.0604, "step": 81800 }, { "epoch": 2.7658135839615943, "grad_norm": 0.9230184555053711, "learning_rate": 6.870639424169767e-05, "loss": 0.0557, "step": 81810 }, { "epoch": 2.766151661651848, "grad_norm": 2.072953224182129, "learning_rate": 6.869872703561831e-05, "loss": 0.0519, "step": 81820 }, { "epoch": 2.766489739342101, "grad_norm": 1.251133918762207, "learning_rate": 6.869105931832277e-05, "loss": 0.0584, "step": 81830 }, { "epoch": 2.766827817032354, "grad_norm": 0.744563639163971, "learning_rate": 6.86833910900207e-05, "loss": 0.0641, "step": 81840 }, { "epoch": 2.7671658947226074, "grad_norm": 0.6336174011230469, "learning_rate": 6.867572235092173e-05, "loss": 0.0592, "step": 81850 }, { "epoch": 2.7675039724128605, "grad_norm": 1.175147294998169, "learning_rate": 6.866805310123554e-05, "loss": 0.0605, "step": 81860 }, { "epoch": 2.7678420501031136, "grad_norm": 0.4946523606777191, "learning_rate": 6.866038334117176e-05, "loss": 0.0664, "step": 81870 }, { "epoch": 2.768180127793367, "grad_norm": 1.851828694343567, "learning_rate": 6.865271307094014e-05, "loss": 0.0608, "step": 81880 }, { "epoch": 2.76851820548362, "grad_norm": 0.6494681239128113, "learning_rate": 6.864504229075033e-05, "loss": 0.0575, "step": 81890 }, { "epoch": 2.7688562831738732, "grad_norm": 0.7945733070373535, "learning_rate": 6.863737100081209e-05, "loss": 0.0531, "step": 81900 }, { "epoch": 2.7691943608641267, "grad_norm": 1.915814757347107, "learning_rate": 6.862969920133512e-05, "loss": 0.0646, "step": 81910 }, { "epoch": 2.76953243855438, "grad_norm": 3.511406660079956, "learning_rate": 6.862202689252917e-05, "loss": 0.0639, "step": 81920 }, { "epoch": 2.769870516244633, "grad_norm": 0.8587055206298828, "learning_rate": 6.861435407460399e-05, "loss": 0.054, "step": 81930 }, { "epoch": 2.7702085939348864, "grad_norm": 1.9971123933792114, "learning_rate": 6.860668074776938e-05, "loss": 0.0655, "step": 81940 }, { "epoch": 2.7705466716251395, "grad_norm": 1.845643401145935, "learning_rate": 6.859900691223509e-05, "loss": 0.0626, "step": 81950 }, { "epoch": 2.7708847493153925, "grad_norm": 0.9476697444915771, "learning_rate": 6.859133256821096e-05, "loss": 0.0566, "step": 81960 }, { "epoch": 2.771222827005646, "grad_norm": 1.0338807106018066, "learning_rate": 6.858365771590678e-05, "loss": 0.0583, "step": 81970 }, { "epoch": 2.771560904695899, "grad_norm": 1.019657015800476, "learning_rate": 6.857598235553237e-05, "loss": 0.0532, "step": 81980 }, { "epoch": 2.771898982386152, "grad_norm": 0.9296405911445618, "learning_rate": 6.856830648729758e-05, "loss": 0.062, "step": 81990 }, { "epoch": 2.7722370600764057, "grad_norm": 1.910800814628601, "learning_rate": 6.856063011141227e-05, "loss": 0.0623, "step": 82000 }, { "epoch": 2.7725751377666588, "grad_norm": 1.3671610355377197, "learning_rate": 6.85529532280863e-05, "loss": 0.0608, "step": 82010 }, { "epoch": 2.772913215456912, "grad_norm": 0.6913853883743286, "learning_rate": 6.854527583752956e-05, "loss": 0.071, "step": 82020 }, { "epoch": 2.7732512931471653, "grad_norm": 0.6695452928543091, "learning_rate": 6.853759793995196e-05, "loss": 0.0573, "step": 82030 }, { "epoch": 2.7735893708374184, "grad_norm": 1.3358683586120605, "learning_rate": 6.85299195355634e-05, "loss": 0.0607, "step": 82040 }, { "epoch": 2.7739274485276715, "grad_norm": 1.4235748052597046, "learning_rate": 6.852224062457378e-05, "loss": 0.0639, "step": 82050 }, { "epoch": 2.774265526217925, "grad_norm": 0.4121151268482208, "learning_rate": 6.851456120719306e-05, "loss": 0.0584, "step": 82060 }, { "epoch": 2.774603603908178, "grad_norm": 0.5607067942619324, "learning_rate": 6.85068812836312e-05, "loss": 0.0604, "step": 82070 }, { "epoch": 2.774941681598431, "grad_norm": 0.6109222769737244, "learning_rate": 6.849920085409816e-05, "loss": 0.0639, "step": 82080 }, { "epoch": 2.7752797592886846, "grad_norm": 0.6913357973098755, "learning_rate": 6.849151991880392e-05, "loss": 0.0591, "step": 82090 }, { "epoch": 2.7756178369789377, "grad_norm": 0.5263227820396423, "learning_rate": 6.848383847795847e-05, "loss": 0.0569, "step": 82100 }, { "epoch": 2.775955914669191, "grad_norm": 0.5468730330467224, "learning_rate": 6.847615653177182e-05, "loss": 0.0612, "step": 82110 }, { "epoch": 2.7762939923594443, "grad_norm": 0.45917797088623047, "learning_rate": 6.846847408045398e-05, "loss": 0.057, "step": 82120 }, { "epoch": 2.7766320700496974, "grad_norm": 1.167466163635254, "learning_rate": 6.846079112421501e-05, "loss": 0.0644, "step": 82130 }, { "epoch": 2.7769701477399504, "grad_norm": 0.5334057807922363, "learning_rate": 6.845310766326494e-05, "loss": 0.0653, "step": 82140 }, { "epoch": 2.777308225430204, "grad_norm": 1.3539131879806519, "learning_rate": 6.844542369781385e-05, "loss": 0.0608, "step": 82150 }, { "epoch": 2.777646303120457, "grad_norm": 1.173792839050293, "learning_rate": 6.84377392280718e-05, "loss": 0.0574, "step": 82160 }, { "epoch": 2.77798438081071, "grad_norm": 0.9178383946418762, "learning_rate": 6.84300542542489e-05, "loss": 0.0599, "step": 82170 }, { "epoch": 2.7783224585009636, "grad_norm": 0.601858377456665, "learning_rate": 6.842236877655521e-05, "loss": 0.0648, "step": 82180 }, { "epoch": 2.7786605361912167, "grad_norm": 1.4901516437530518, "learning_rate": 6.84146827952009e-05, "loss": 0.0656, "step": 82190 }, { "epoch": 2.7789986138814697, "grad_norm": 0.7385057210922241, "learning_rate": 6.840699631039606e-05, "loss": 0.057, "step": 82200 }, { "epoch": 2.7793366915717233, "grad_norm": 0.5238521695137024, "learning_rate": 6.839930932235089e-05, "loss": 0.0608, "step": 82210 }, { "epoch": 2.7796747692619763, "grad_norm": 0.9183527231216431, "learning_rate": 6.83916218312755e-05, "loss": 0.0677, "step": 82220 }, { "epoch": 2.7800128469522294, "grad_norm": 0.6478486657142639, "learning_rate": 6.838393383738006e-05, "loss": 0.0603, "step": 82230 }, { "epoch": 2.780350924642483, "grad_norm": 0.7385034561157227, "learning_rate": 6.83762453408748e-05, "loss": 0.0708, "step": 82240 }, { "epoch": 2.780689002332736, "grad_norm": 0.8981940150260925, "learning_rate": 6.836855634196988e-05, "loss": 0.0686, "step": 82250 }, { "epoch": 2.781027080022989, "grad_norm": 0.9783680438995361, "learning_rate": 6.836086684087553e-05, "loss": 0.066, "step": 82260 }, { "epoch": 2.7813651577132426, "grad_norm": 2.3041677474975586, "learning_rate": 6.835317683780198e-05, "loss": 0.0628, "step": 82270 }, { "epoch": 2.7817032354034956, "grad_norm": 0.6703003644943237, "learning_rate": 6.834548633295948e-05, "loss": 0.0587, "step": 82280 }, { "epoch": 2.7820413130937487, "grad_norm": 0.7556859254837036, "learning_rate": 6.833779532655828e-05, "loss": 0.0544, "step": 82290 }, { "epoch": 2.782379390784002, "grad_norm": 0.6976072788238525, "learning_rate": 6.833010381880864e-05, "loss": 0.0598, "step": 82300 }, { "epoch": 2.7827174684742553, "grad_norm": 3.10611891746521, "learning_rate": 6.832241180992083e-05, "loss": 0.0644, "step": 82310 }, { "epoch": 2.7830555461645083, "grad_norm": 0.796281099319458, "learning_rate": 6.831471930010518e-05, "loss": 0.0681, "step": 82320 }, { "epoch": 2.783393623854762, "grad_norm": 0.6364321708679199, "learning_rate": 6.830702628957198e-05, "loss": 0.0599, "step": 82330 }, { "epoch": 2.783731701545015, "grad_norm": 0.999504566192627, "learning_rate": 6.829933277853158e-05, "loss": 0.0665, "step": 82340 }, { "epoch": 2.784069779235268, "grad_norm": 1.8636318445205688, "learning_rate": 6.829163876719426e-05, "loss": 0.0566, "step": 82350 }, { "epoch": 2.7844078569255215, "grad_norm": 0.7692185044288635, "learning_rate": 6.828394425577044e-05, "loss": 0.0606, "step": 82360 }, { "epoch": 2.7847459346157746, "grad_norm": 0.8647109270095825, "learning_rate": 6.827624924447044e-05, "loss": 0.068, "step": 82370 }, { "epoch": 2.785084012306028, "grad_norm": 0.865967333316803, "learning_rate": 6.826855373350465e-05, "loss": 0.0605, "step": 82380 }, { "epoch": 2.785422089996281, "grad_norm": 0.9541834592819214, "learning_rate": 6.826085772308346e-05, "loss": 0.0682, "step": 82390 }, { "epoch": 2.7857601676865342, "grad_norm": 3.7471234798431396, "learning_rate": 6.825316121341729e-05, "loss": 0.0576, "step": 82400 }, { "epoch": 2.7860982453767877, "grad_norm": 0.5174198746681213, "learning_rate": 6.824546420471653e-05, "loss": 0.0641, "step": 82410 }, { "epoch": 2.786436323067041, "grad_norm": 1.2604907751083374, "learning_rate": 6.823776669719166e-05, "loss": 0.0592, "step": 82420 }, { "epoch": 2.786774400757294, "grad_norm": 0.719809889793396, "learning_rate": 6.823006869105308e-05, "loss": 0.0606, "step": 82430 }, { "epoch": 2.7871124784475474, "grad_norm": 0.9160181879997253, "learning_rate": 6.822237018651129e-05, "loss": 0.056, "step": 82440 }, { "epoch": 2.7874505561378005, "grad_norm": 0.8946492075920105, "learning_rate": 6.821467118377672e-05, "loss": 0.0636, "step": 82450 }, { "epoch": 2.7877886338280535, "grad_norm": 1.1184176206588745, "learning_rate": 6.820697168305992e-05, "loss": 0.0634, "step": 82460 }, { "epoch": 2.788126711518307, "grad_norm": 0.9036959409713745, "learning_rate": 6.819927168457131e-05, "loss": 0.0589, "step": 82470 }, { "epoch": 2.78846478920856, "grad_norm": 0.8338983058929443, "learning_rate": 6.819157118852147e-05, "loss": 0.0621, "step": 82480 }, { "epoch": 2.788802866898813, "grad_norm": 0.6729753613471985, "learning_rate": 6.818387019512092e-05, "loss": 0.0694, "step": 82490 }, { "epoch": 2.7891409445890667, "grad_norm": 1.009842038154602, "learning_rate": 6.817616870458018e-05, "loss": 0.0669, "step": 82500 }, { "epoch": 2.7894790222793198, "grad_norm": 0.7534738183021545, "learning_rate": 6.816846671710981e-05, "loss": 0.0577, "step": 82510 }, { "epoch": 2.7898170999695733, "grad_norm": 0.5664520859718323, "learning_rate": 6.816076423292038e-05, "loss": 0.0642, "step": 82520 }, { "epoch": 2.7901551776598263, "grad_norm": 1.0040102005004883, "learning_rate": 6.815306125222249e-05, "loss": 0.0627, "step": 82530 }, { "epoch": 2.7904932553500794, "grad_norm": 2.1253228187561035, "learning_rate": 6.814535777522673e-05, "loss": 0.0611, "step": 82540 }, { "epoch": 2.790831333040333, "grad_norm": 0.6622933149337769, "learning_rate": 6.81376538021437e-05, "loss": 0.0562, "step": 82550 }, { "epoch": 2.791169410730586, "grad_norm": 0.760624885559082, "learning_rate": 6.812994933318403e-05, "loss": 0.0558, "step": 82560 }, { "epoch": 2.791507488420839, "grad_norm": 3.3356363773345947, "learning_rate": 6.812224436855836e-05, "loss": 0.0634, "step": 82570 }, { "epoch": 2.7918455661110926, "grad_norm": 0.6291145086288452, "learning_rate": 6.811453890847732e-05, "loss": 0.0631, "step": 82580 }, { "epoch": 2.7921836438013456, "grad_norm": 0.7068415880203247, "learning_rate": 6.810683295315162e-05, "loss": 0.0561, "step": 82590 }, { "epoch": 2.7925217214915987, "grad_norm": 1.202404499053955, "learning_rate": 6.809912650279188e-05, "loss": 0.0576, "step": 82600 }, { "epoch": 2.7928597991818522, "grad_norm": 1.3407200574874878, "learning_rate": 6.809141955760885e-05, "loss": 0.0604, "step": 82610 }, { "epoch": 2.7931978768721053, "grad_norm": 1.0720733404159546, "learning_rate": 6.808371211781318e-05, "loss": 0.0659, "step": 82620 }, { "epoch": 2.7935359545623584, "grad_norm": 0.6329993605613708, "learning_rate": 6.807600418361562e-05, "loss": 0.064, "step": 82630 }, { "epoch": 2.793874032252612, "grad_norm": 1.125667691230774, "learning_rate": 6.80682957552269e-05, "loss": 0.0632, "step": 82640 }, { "epoch": 2.794212109942865, "grad_norm": 0.358125239610672, "learning_rate": 6.806058683285776e-05, "loss": 0.0618, "step": 82650 }, { "epoch": 2.794550187633118, "grad_norm": 0.3739187717437744, "learning_rate": 6.805287741671896e-05, "loss": 0.0572, "step": 82660 }, { "epoch": 2.7948882653233715, "grad_norm": 0.45633795857429504, "learning_rate": 6.804516750702129e-05, "loss": 0.0569, "step": 82670 }, { "epoch": 2.7952263430136246, "grad_norm": 1.0577001571655273, "learning_rate": 6.803745710397549e-05, "loss": 0.0596, "step": 82680 }, { "epoch": 2.7955644207038777, "grad_norm": 1.294227957725525, "learning_rate": 6.802974620779241e-05, "loss": 0.0728, "step": 82690 }, { "epoch": 2.795902498394131, "grad_norm": 0.7805055975914001, "learning_rate": 6.802203481868283e-05, "loss": 0.0594, "step": 82700 }, { "epoch": 2.7962405760843843, "grad_norm": 0.5931764245033264, "learning_rate": 6.801432293685761e-05, "loss": 0.0634, "step": 82710 }, { "epoch": 2.7965786537746373, "grad_norm": 0.8209718465805054, "learning_rate": 6.800661056252755e-05, "loss": 0.0588, "step": 82720 }, { "epoch": 2.796916731464891, "grad_norm": 1.300818681716919, "learning_rate": 6.799889769590353e-05, "loss": 0.0565, "step": 82730 }, { "epoch": 2.797254809155144, "grad_norm": 0.6855236291885376, "learning_rate": 6.79911843371964e-05, "loss": 0.0588, "step": 82740 }, { "epoch": 2.797592886845397, "grad_norm": 0.7630497813224792, "learning_rate": 6.798347048661706e-05, "loss": 0.0558, "step": 82750 }, { "epoch": 2.7979309645356505, "grad_norm": 0.7556053400039673, "learning_rate": 6.797575614437639e-05, "loss": 0.06, "step": 82760 }, { "epoch": 2.7982690422259036, "grad_norm": 1.9687591791152954, "learning_rate": 6.79680413106853e-05, "loss": 0.0603, "step": 82770 }, { "epoch": 2.7986071199161566, "grad_norm": 1.5800268650054932, "learning_rate": 6.796032598575471e-05, "loss": 0.0585, "step": 82780 }, { "epoch": 2.79894519760641, "grad_norm": 0.692314624786377, "learning_rate": 6.795261016979555e-05, "loss": 0.06, "step": 82790 }, { "epoch": 2.799283275296663, "grad_norm": 0.771334707736969, "learning_rate": 6.794489386301878e-05, "loss": 0.0575, "step": 82800 }, { "epoch": 2.7996213529869163, "grad_norm": 0.7825608253479004, "learning_rate": 6.793717706563534e-05, "loss": 0.0631, "step": 82810 }, { "epoch": 2.79995943067717, "grad_norm": 1.522841215133667, "learning_rate": 6.792945977785623e-05, "loss": 0.0519, "step": 82820 }, { "epoch": 2.800297508367423, "grad_norm": 1.6590421199798584, "learning_rate": 6.792174199989243e-05, "loss": 0.0595, "step": 82830 }, { "epoch": 2.800635586057676, "grad_norm": 0.8083429932594299, "learning_rate": 6.791402373195493e-05, "loss": 0.0614, "step": 82840 }, { "epoch": 2.8009736637479294, "grad_norm": 0.7311067581176758, "learning_rate": 6.790630497425475e-05, "loss": 0.0568, "step": 82850 }, { "epoch": 2.8013117414381825, "grad_norm": 1.2689844369888306, "learning_rate": 6.789858572700293e-05, "loss": 0.0578, "step": 82860 }, { "epoch": 2.8016498191284356, "grad_norm": 0.5705965161323547, "learning_rate": 6.789086599041047e-05, "loss": 0.0609, "step": 82870 }, { "epoch": 2.801987896818689, "grad_norm": 2.800250768661499, "learning_rate": 6.788314576468849e-05, "loss": 0.0631, "step": 82880 }, { "epoch": 2.802325974508942, "grad_norm": 1.0094163417816162, "learning_rate": 6.7875425050048e-05, "loss": 0.0642, "step": 82890 }, { "epoch": 2.8026640521991952, "grad_norm": 0.9748022556304932, "learning_rate": 6.786770384670013e-05, "loss": 0.0586, "step": 82900 }, { "epoch": 2.8030021298894487, "grad_norm": 0.7118015885353088, "learning_rate": 6.785998215485591e-05, "loss": 0.0619, "step": 82910 }, { "epoch": 2.803340207579702, "grad_norm": 0.8905333876609802, "learning_rate": 6.785225997472653e-05, "loss": 0.0581, "step": 82920 }, { "epoch": 2.803678285269955, "grad_norm": 1.4882711172103882, "learning_rate": 6.784453730652305e-05, "loss": 0.0611, "step": 82930 }, { "epoch": 2.8040163629602084, "grad_norm": 0.7416388392448425, "learning_rate": 6.783681415045661e-05, "loss": 0.0564, "step": 82940 }, { "epoch": 2.8043544406504615, "grad_norm": 0.9018976092338562, "learning_rate": 6.782909050673839e-05, "loss": 0.0665, "step": 82950 }, { "epoch": 2.8046925183407145, "grad_norm": 2.7344958782196045, "learning_rate": 6.782136637557953e-05, "loss": 0.0593, "step": 82960 }, { "epoch": 2.805030596030968, "grad_norm": 0.47908079624176025, "learning_rate": 6.781364175719121e-05, "loss": 0.0513, "step": 82970 }, { "epoch": 2.805368673721221, "grad_norm": 1.2442147731781006, "learning_rate": 6.780591665178461e-05, "loss": 0.0569, "step": 82980 }, { "epoch": 2.805706751411474, "grad_norm": 0.8772178292274475, "learning_rate": 6.779819105957095e-05, "loss": 0.0656, "step": 82990 }, { "epoch": 2.8060448291017277, "grad_norm": 0.6527968049049377, "learning_rate": 6.779046498076143e-05, "loss": 0.0664, "step": 83000 }, { "epoch": 2.8063829067919808, "grad_norm": 0.5605630874633789, "learning_rate": 6.778273841556725e-05, "loss": 0.0604, "step": 83010 }, { "epoch": 2.806720984482234, "grad_norm": 0.7141987681388855, "learning_rate": 6.777501136419972e-05, "loss": 0.0617, "step": 83020 }, { "epoch": 2.8070590621724874, "grad_norm": 0.9201030731201172, "learning_rate": 6.776728382687005e-05, "loss": 0.0661, "step": 83030 }, { "epoch": 2.8073971398627404, "grad_norm": 1.2799545526504517, "learning_rate": 6.775955580378951e-05, "loss": 0.061, "step": 83040 }, { "epoch": 2.8077352175529935, "grad_norm": 0.842852771282196, "learning_rate": 6.775182729516939e-05, "loss": 0.061, "step": 83050 }, { "epoch": 2.808073295243247, "grad_norm": 0.8589309453964233, "learning_rate": 6.774409830122097e-05, "loss": 0.0552, "step": 83060 }, { "epoch": 2.8084113729335, "grad_norm": 0.5288635492324829, "learning_rate": 6.77363688221556e-05, "loss": 0.0608, "step": 83070 }, { "epoch": 2.808749450623753, "grad_norm": 2.239069700241089, "learning_rate": 6.772863885818452e-05, "loss": 0.0601, "step": 83080 }, { "epoch": 2.8090875283140067, "grad_norm": 0.9853573441505432, "learning_rate": 6.772090840951915e-05, "loss": 0.0617, "step": 83090 }, { "epoch": 2.8094256060042597, "grad_norm": 1.0394556522369385, "learning_rate": 6.771317747637078e-05, "loss": 0.0552, "step": 83100 }, { "epoch": 2.809763683694513, "grad_norm": 0.5742775797843933, "learning_rate": 6.770544605895082e-05, "loss": 0.0556, "step": 83110 }, { "epoch": 2.8101017613847663, "grad_norm": 0.8255060911178589, "learning_rate": 6.769771415747059e-05, "loss": 0.0625, "step": 83120 }, { "epoch": 2.8104398390750194, "grad_norm": 0.5632453560829163, "learning_rate": 6.768998177214153e-05, "loss": 0.0586, "step": 83130 }, { "epoch": 2.8107779167652724, "grad_norm": 2.6542704105377197, "learning_rate": 6.768224890317499e-05, "loss": 0.0599, "step": 83140 }, { "epoch": 2.811115994455526, "grad_norm": 0.6540679931640625, "learning_rate": 6.767451555078242e-05, "loss": 0.0647, "step": 83150 }, { "epoch": 2.811454072145779, "grad_norm": 0.7592980265617371, "learning_rate": 6.766678171517522e-05, "loss": 0.0627, "step": 83160 }, { "epoch": 2.811792149836032, "grad_norm": 0.9108603000640869, "learning_rate": 6.765904739656486e-05, "loss": 0.0563, "step": 83170 }, { "epoch": 2.8121302275262856, "grad_norm": 0.6299149394035339, "learning_rate": 6.765131259516277e-05, "loss": 0.0631, "step": 83180 }, { "epoch": 2.8124683052165387, "grad_norm": 0.9543861150741577, "learning_rate": 6.764357731118043e-05, "loss": 0.0647, "step": 83190 }, { "epoch": 2.8128063829067917, "grad_norm": 0.7135117650032043, "learning_rate": 6.763584154482932e-05, "loss": 0.0602, "step": 83200 }, { "epoch": 2.8131444605970453, "grad_norm": 0.751322865486145, "learning_rate": 6.762810529632092e-05, "loss": 0.0608, "step": 83210 }, { "epoch": 2.8134825382872983, "grad_norm": 0.8724144697189331, "learning_rate": 6.762036856586674e-05, "loss": 0.0651, "step": 83220 }, { "epoch": 2.8138206159775514, "grad_norm": 0.8098955154418945, "learning_rate": 6.76126313536783e-05, "loss": 0.0587, "step": 83230 }, { "epoch": 2.814158693667805, "grad_norm": 0.7534717917442322, "learning_rate": 6.760489365996717e-05, "loss": 0.0607, "step": 83240 }, { "epoch": 2.814496771358058, "grad_norm": 0.8955134749412537, "learning_rate": 6.759715548494482e-05, "loss": 0.0593, "step": 83250 }, { "epoch": 2.814834849048311, "grad_norm": 0.847656786441803, "learning_rate": 6.758941682882287e-05, "loss": 0.0635, "step": 83260 }, { "epoch": 2.8151729267385646, "grad_norm": 1.2791534662246704, "learning_rate": 6.758167769181285e-05, "loss": 0.0639, "step": 83270 }, { "epoch": 2.8155110044288176, "grad_norm": 0.5814121961593628, "learning_rate": 6.757393807412639e-05, "loss": 0.0578, "step": 83280 }, { "epoch": 2.8158490821190707, "grad_norm": 0.7214401364326477, "learning_rate": 6.756619797597506e-05, "loss": 0.0568, "step": 83290 }, { "epoch": 2.816187159809324, "grad_norm": 0.8421252965927124, "learning_rate": 6.755845739757048e-05, "loss": 0.0595, "step": 83300 }, { "epoch": 2.8165252374995773, "grad_norm": 1.0849982500076294, "learning_rate": 6.755071633912426e-05, "loss": 0.0655, "step": 83310 }, { "epoch": 2.8168633151898304, "grad_norm": 0.5964885950088501, "learning_rate": 6.754297480084807e-05, "loss": 0.0555, "step": 83320 }, { "epoch": 2.817201392880084, "grad_norm": 0.6351515650749207, "learning_rate": 6.75352327829535e-05, "loss": 0.0586, "step": 83330 }, { "epoch": 2.817539470570337, "grad_norm": 1.40559983253479, "learning_rate": 6.752749028565229e-05, "loss": 0.0602, "step": 83340 }, { "epoch": 2.8178775482605904, "grad_norm": 0.5665598511695862, "learning_rate": 6.751974730915607e-05, "loss": 0.0568, "step": 83350 }, { "epoch": 2.8182156259508435, "grad_norm": 0.8552013039588928, "learning_rate": 6.751200385367655e-05, "loss": 0.0602, "step": 83360 }, { "epoch": 2.8185537036410966, "grad_norm": 0.8188120126724243, "learning_rate": 6.75042599194254e-05, "loss": 0.0643, "step": 83370 }, { "epoch": 2.81889178133135, "grad_norm": 0.5663717985153198, "learning_rate": 6.749651550661438e-05, "loss": 0.0614, "step": 83380 }, { "epoch": 2.819229859021603, "grad_norm": 0.4039260447025299, "learning_rate": 6.74887706154552e-05, "loss": 0.0562, "step": 83390 }, { "epoch": 2.8195679367118562, "grad_norm": 1.0834972858428955, "learning_rate": 6.748102524615959e-05, "loss": 0.054, "step": 83400 }, { "epoch": 2.8199060144021098, "grad_norm": 0.6283162832260132, "learning_rate": 6.747327939893931e-05, "loss": 0.0597, "step": 83410 }, { "epoch": 2.820244092092363, "grad_norm": 5.9336838722229, "learning_rate": 6.746553307400617e-05, "loss": 0.062, "step": 83420 }, { "epoch": 2.820582169782616, "grad_norm": 0.7434399724006653, "learning_rate": 6.745778627157187e-05, "loss": 0.0588, "step": 83430 }, { "epoch": 2.8209202474728694, "grad_norm": 0.8763619065284729, "learning_rate": 6.745003899184827e-05, "loss": 0.0615, "step": 83440 }, { "epoch": 2.8212583251631225, "grad_norm": 1.1243139505386353, "learning_rate": 6.744229123504718e-05, "loss": 0.0564, "step": 83450 }, { "epoch": 2.8215964028533755, "grad_norm": 0.5223084688186646, "learning_rate": 6.74345430013804e-05, "loss": 0.0585, "step": 83460 }, { "epoch": 2.821934480543629, "grad_norm": 0.6550795435905457, "learning_rate": 6.742679429105974e-05, "loss": 0.0607, "step": 83470 }, { "epoch": 2.822272558233882, "grad_norm": 0.8656832575798035, "learning_rate": 6.741904510429706e-05, "loss": 0.0587, "step": 83480 }, { "epoch": 2.8226106359241356, "grad_norm": 2.3504703044891357, "learning_rate": 6.741129544130427e-05, "loss": 0.0595, "step": 83490 }, { "epoch": 2.8229487136143887, "grad_norm": 0.8467559218406677, "learning_rate": 6.740354530229318e-05, "loss": 0.0634, "step": 83500 }, { "epoch": 2.8232867913046418, "grad_norm": 0.4832301437854767, "learning_rate": 6.739579468747571e-05, "loss": 0.0631, "step": 83510 }, { "epoch": 2.8236248689948953, "grad_norm": 0.35696715116500854, "learning_rate": 6.738804359706373e-05, "loss": 0.0553, "step": 83520 }, { "epoch": 2.8239629466851484, "grad_norm": 0.7130741477012634, "learning_rate": 6.738029203126918e-05, "loss": 0.0583, "step": 83530 }, { "epoch": 2.8243010243754014, "grad_norm": 0.47027742862701416, "learning_rate": 6.7372539990304e-05, "loss": 0.0536, "step": 83540 }, { "epoch": 2.824639102065655, "grad_norm": 1.389323353767395, "learning_rate": 6.736478747438007e-05, "loss": 0.0537, "step": 83550 }, { "epoch": 2.824977179755908, "grad_norm": 1.0212727785110474, "learning_rate": 6.735703448370938e-05, "loss": 0.062, "step": 83560 }, { "epoch": 2.825315257446161, "grad_norm": 1.0818573236465454, "learning_rate": 6.734928101850391e-05, "loss": 0.0601, "step": 83570 }, { "epoch": 2.8256533351364146, "grad_norm": 1.094029188156128, "learning_rate": 6.73415270789756e-05, "loss": 0.0582, "step": 83580 }, { "epoch": 2.8259914128266677, "grad_norm": 1.0455249547958374, "learning_rate": 6.733377266533644e-05, "loss": 0.0637, "step": 83590 }, { "epoch": 2.8263294905169207, "grad_norm": 0.5681179761886597, "learning_rate": 6.732601777779847e-05, "loss": 0.0549, "step": 83600 }, { "epoch": 2.8266675682071742, "grad_norm": 1.3134247064590454, "learning_rate": 6.731826241657368e-05, "loss": 0.0617, "step": 83610 }, { "epoch": 2.8270056458974273, "grad_norm": 0.8835235834121704, "learning_rate": 6.73105065818741e-05, "loss": 0.0592, "step": 83620 }, { "epoch": 2.8273437235876804, "grad_norm": 0.6141197085380554, "learning_rate": 6.730275027391178e-05, "loss": 0.0522, "step": 83630 }, { "epoch": 2.827681801277934, "grad_norm": 0.9713426232337952, "learning_rate": 6.729499349289874e-05, "loss": 0.0574, "step": 83640 }, { "epoch": 2.828019878968187, "grad_norm": 0.6011172533035278, "learning_rate": 6.728723623904709e-05, "loss": 0.0628, "step": 83650 }, { "epoch": 2.82835795665844, "grad_norm": 0.9906256198883057, "learning_rate": 6.72794785125689e-05, "loss": 0.0556, "step": 83660 }, { "epoch": 2.8286960343486935, "grad_norm": 0.7538473606109619, "learning_rate": 6.727172031367627e-05, "loss": 0.0596, "step": 83670 }, { "epoch": 2.8290341120389466, "grad_norm": 0.3642056882381439, "learning_rate": 6.726396164258128e-05, "loss": 0.0548, "step": 83680 }, { "epoch": 2.8293721897291997, "grad_norm": 0.849642276763916, "learning_rate": 6.725620249949606e-05, "loss": 0.0654, "step": 83690 }, { "epoch": 2.829710267419453, "grad_norm": 1.4165033102035522, "learning_rate": 6.724844288463275e-05, "loss": 0.0603, "step": 83700 }, { "epoch": 2.8300483451097063, "grad_norm": 0.7215090394020081, "learning_rate": 6.72406827982035e-05, "loss": 0.0606, "step": 83710 }, { "epoch": 2.8303864227999593, "grad_norm": 0.8253834247589111, "learning_rate": 6.723292224042045e-05, "loss": 0.067, "step": 83720 }, { "epoch": 2.830724500490213, "grad_norm": 0.956598699092865, "learning_rate": 6.722516121149578e-05, "loss": 0.0653, "step": 83730 }, { "epoch": 2.831062578180466, "grad_norm": 0.9132595062255859, "learning_rate": 6.721739971164169e-05, "loss": 0.0589, "step": 83740 }, { "epoch": 2.831400655870719, "grad_norm": 0.5129356980323792, "learning_rate": 6.720963774107035e-05, "loss": 0.0593, "step": 83750 }, { "epoch": 2.8317387335609725, "grad_norm": 0.6550308465957642, "learning_rate": 6.720187529999398e-05, "loss": 0.0538, "step": 83760 }, { "epoch": 2.8320768112512256, "grad_norm": 0.9005833268165588, "learning_rate": 6.71941123886248e-05, "loss": 0.0589, "step": 83770 }, { "epoch": 2.8324148889414786, "grad_norm": 1.7887530326843262, "learning_rate": 6.718634900717504e-05, "loss": 0.0666, "step": 83780 }, { "epoch": 2.832752966631732, "grad_norm": 0.8957923650741577, "learning_rate": 6.717858515585696e-05, "loss": 0.0575, "step": 83790 }, { "epoch": 2.833091044321985, "grad_norm": 1.101004958152771, "learning_rate": 6.717082083488282e-05, "loss": 0.058, "step": 83800 }, { "epoch": 2.8334291220122383, "grad_norm": 0.436016321182251, "learning_rate": 6.716305604446488e-05, "loss": 0.0601, "step": 83810 }, { "epoch": 2.833767199702492, "grad_norm": 0.48413822054862976, "learning_rate": 6.715529078481545e-05, "loss": 0.0671, "step": 83820 }, { "epoch": 2.834105277392745, "grad_norm": 1.250483751296997, "learning_rate": 6.714752505614681e-05, "loss": 0.0627, "step": 83830 }, { "epoch": 2.834443355082998, "grad_norm": 2.5121026039123535, "learning_rate": 6.713975885867126e-05, "loss": 0.0628, "step": 83840 }, { "epoch": 2.8347814327732515, "grad_norm": 0.8302661776542664, "learning_rate": 6.713199219260116e-05, "loss": 0.0572, "step": 83850 }, { "epoch": 2.8351195104635045, "grad_norm": 2.1187071800231934, "learning_rate": 6.712422505814882e-05, "loss": 0.0589, "step": 83860 }, { "epoch": 2.8354575881537576, "grad_norm": 2.0974714756011963, "learning_rate": 6.711645745552661e-05, "loss": 0.0635, "step": 83870 }, { "epoch": 2.835795665844011, "grad_norm": 0.937352180480957, "learning_rate": 6.710868938494688e-05, "loss": 0.0583, "step": 83880 }, { "epoch": 2.836133743534264, "grad_norm": 1.4175485372543335, "learning_rate": 6.710092084662201e-05, "loss": 0.0731, "step": 83890 }, { "epoch": 2.8364718212245172, "grad_norm": 0.7814223766326904, "learning_rate": 6.709315184076438e-05, "loss": 0.0585, "step": 83900 }, { "epoch": 2.8368098989147708, "grad_norm": 1.3899048566818237, "learning_rate": 6.708538236758642e-05, "loss": 0.0684, "step": 83910 }, { "epoch": 2.837147976605024, "grad_norm": 0.5813559889793396, "learning_rate": 6.70776124273005e-05, "loss": 0.0605, "step": 83920 }, { "epoch": 2.837486054295277, "grad_norm": 1.9075816869735718, "learning_rate": 6.70698420201191e-05, "loss": 0.0654, "step": 83930 }, { "epoch": 2.8378241319855304, "grad_norm": 0.9137285947799683, "learning_rate": 6.706207114625461e-05, "loss": 0.0619, "step": 83940 }, { "epoch": 2.8381622096757835, "grad_norm": 0.5232746601104736, "learning_rate": 6.705429980591951e-05, "loss": 0.0576, "step": 83950 }, { "epoch": 2.8385002873660365, "grad_norm": 1.1701868772506714, "learning_rate": 6.704652799932625e-05, "loss": 0.0549, "step": 83960 }, { "epoch": 2.83883836505629, "grad_norm": 0.5751649737358093, "learning_rate": 6.703875572668733e-05, "loss": 0.0651, "step": 83970 }, { "epoch": 2.839176442746543, "grad_norm": 0.7853454351425171, "learning_rate": 6.703098298821523e-05, "loss": 0.0574, "step": 83980 }, { "epoch": 2.839514520436796, "grad_norm": 1.8193401098251343, "learning_rate": 6.702320978412246e-05, "loss": 0.0571, "step": 83990 }, { "epoch": 2.8398525981270497, "grad_norm": 0.5793631076812744, "learning_rate": 6.701543611462152e-05, "loss": 0.0606, "step": 84000 }, { "epoch": 2.840190675817303, "grad_norm": 0.5170610547065735, "learning_rate": 6.700766197992495e-05, "loss": 0.0597, "step": 84010 }, { "epoch": 2.840528753507556, "grad_norm": 0.5108944773674011, "learning_rate": 6.69998873802453e-05, "loss": 0.0593, "step": 84020 }, { "epoch": 2.8408668311978094, "grad_norm": 0.767723798751831, "learning_rate": 6.699211231579509e-05, "loss": 0.0558, "step": 84030 }, { "epoch": 2.8412049088880624, "grad_norm": 6.276208400726318, "learning_rate": 6.698433678678693e-05, "loss": 0.0598, "step": 84040 }, { "epoch": 2.8415429865783155, "grad_norm": 0.5968562960624695, "learning_rate": 6.69765607934334e-05, "loss": 0.0658, "step": 84050 }, { "epoch": 2.841881064268569, "grad_norm": 0.8498018383979797, "learning_rate": 6.696878433594706e-05, "loss": 0.0613, "step": 84060 }, { "epoch": 2.842219141958822, "grad_norm": 0.9642738103866577, "learning_rate": 6.696100741454054e-05, "loss": 0.0573, "step": 84070 }, { "epoch": 2.842557219649075, "grad_norm": 0.6097720861434937, "learning_rate": 6.695323002942642e-05, "loss": 0.0516, "step": 84080 }, { "epoch": 2.8428952973393287, "grad_norm": 0.7087414264678955, "learning_rate": 6.69454521808174e-05, "loss": 0.0586, "step": 84090 }, { "epoch": 2.8432333750295817, "grad_norm": 0.44878363609313965, "learning_rate": 6.693767386892605e-05, "loss": 0.0603, "step": 84100 }, { "epoch": 2.843571452719835, "grad_norm": 1.1159528493881226, "learning_rate": 6.692989509396508e-05, "loss": 0.0617, "step": 84110 }, { "epoch": 2.8439095304100883, "grad_norm": 0.6764934659004211, "learning_rate": 6.692211585614711e-05, "loss": 0.0565, "step": 84120 }, { "epoch": 2.8442476081003414, "grad_norm": 1.8929864168167114, "learning_rate": 6.691433615568489e-05, "loss": 0.0631, "step": 84130 }, { "epoch": 2.8445856857905945, "grad_norm": 0.4744660556316376, "learning_rate": 6.690655599279106e-05, "loss": 0.0597, "step": 84140 }, { "epoch": 2.844923763480848, "grad_norm": 0.6506770253181458, "learning_rate": 6.689877536767833e-05, "loss": 0.0598, "step": 84150 }, { "epoch": 2.845261841171101, "grad_norm": 1.0567325353622437, "learning_rate": 6.689099428055942e-05, "loss": 0.0559, "step": 84160 }, { "epoch": 2.845599918861354, "grad_norm": 0.7949625849723816, "learning_rate": 6.688321273164711e-05, "loss": 0.0663, "step": 84170 }, { "epoch": 2.8459379965516076, "grad_norm": 0.9727532267570496, "learning_rate": 6.687543072115408e-05, "loss": 0.0615, "step": 84180 }, { "epoch": 2.8462760742418607, "grad_norm": 1.1653404235839844, "learning_rate": 6.686764824929311e-05, "loss": 0.0595, "step": 84190 }, { "epoch": 2.8466141519321138, "grad_norm": 0.47267743945121765, "learning_rate": 6.6859865316277e-05, "loss": 0.0545, "step": 84200 }, { "epoch": 2.8469522296223673, "grad_norm": 0.7825753688812256, "learning_rate": 6.685208192231846e-05, "loss": 0.0592, "step": 84210 }, { "epoch": 2.8472903073126203, "grad_norm": 0.5425163507461548, "learning_rate": 6.684429806763038e-05, "loss": 0.0629, "step": 84220 }, { "epoch": 2.8476283850028734, "grad_norm": 1.0071446895599365, "learning_rate": 6.683651375242548e-05, "loss": 0.0609, "step": 84230 }, { "epoch": 2.847966462693127, "grad_norm": 1.1966875791549683, "learning_rate": 6.682872897691665e-05, "loss": 0.057, "step": 84240 }, { "epoch": 2.84830454038338, "grad_norm": 0.3691728711128235, "learning_rate": 6.682094374131668e-05, "loss": 0.0583, "step": 84250 }, { "epoch": 2.848642618073633, "grad_norm": 0.619797945022583, "learning_rate": 6.681315804583843e-05, "loss": 0.0562, "step": 84260 }, { "epoch": 2.8489806957638866, "grad_norm": 0.6364789009094238, "learning_rate": 6.680537189069475e-05, "loss": 0.0554, "step": 84270 }, { "epoch": 2.8493187734541396, "grad_norm": 1.1636322736740112, "learning_rate": 6.679758527609852e-05, "loss": 0.0595, "step": 84280 }, { "epoch": 2.8496568511443927, "grad_norm": 0.6951487064361572, "learning_rate": 6.678979820226261e-05, "loss": 0.0571, "step": 84290 }, { "epoch": 2.8499949288346462, "grad_norm": 0.9733210206031799, "learning_rate": 6.678201066939993e-05, "loss": 0.0563, "step": 84300 }, { "epoch": 2.8503330065248993, "grad_norm": 0.75555020570755, "learning_rate": 6.677422267772338e-05, "loss": 0.06, "step": 84310 }, { "epoch": 2.850671084215153, "grad_norm": 0.49963998794555664, "learning_rate": 6.67664342274459e-05, "loss": 0.0597, "step": 84320 }, { "epoch": 2.851009161905406, "grad_norm": 0.7412177920341492, "learning_rate": 6.67586453187804e-05, "loss": 0.0593, "step": 84330 }, { "epoch": 2.851347239595659, "grad_norm": 0.6100654602050781, "learning_rate": 6.675085595193984e-05, "loss": 0.0553, "step": 84340 }, { "epoch": 2.8516853172859125, "grad_norm": 1.2915992736816406, "learning_rate": 6.674306612713714e-05, "loss": 0.0554, "step": 84350 }, { "epoch": 2.8520233949761655, "grad_norm": 0.8679530024528503, "learning_rate": 6.673527584458534e-05, "loss": 0.0612, "step": 84360 }, { "epoch": 2.8523614726664186, "grad_norm": 0.6095782518386841, "learning_rate": 6.672748510449736e-05, "loss": 0.0558, "step": 84370 }, { "epoch": 2.852699550356672, "grad_norm": 1.6700561046600342, "learning_rate": 6.671969390708623e-05, "loss": 0.0613, "step": 84380 }, { "epoch": 2.853037628046925, "grad_norm": 0.5949579477310181, "learning_rate": 6.671190225256496e-05, "loss": 0.0582, "step": 84390 }, { "epoch": 2.8533757057371782, "grad_norm": 1.2294570207595825, "learning_rate": 6.670411014114653e-05, "loss": 0.055, "step": 84400 }, { "epoch": 2.8537137834274318, "grad_norm": 0.7410945296287537, "learning_rate": 6.669631757304403e-05, "loss": 0.0551, "step": 84410 }, { "epoch": 2.854051861117685, "grad_norm": 2.5377607345581055, "learning_rate": 6.668852454847048e-05, "loss": 0.0577, "step": 84420 }, { "epoch": 2.854389938807938, "grad_norm": 0.9031684994697571, "learning_rate": 6.668073106763893e-05, "loss": 0.0561, "step": 84430 }, { "epoch": 2.8547280164981914, "grad_norm": 0.6450600028038025, "learning_rate": 6.667293713076246e-05, "loss": 0.0614, "step": 84440 }, { "epoch": 2.8550660941884445, "grad_norm": 0.5081115961074829, "learning_rate": 6.666514273805416e-05, "loss": 0.0547, "step": 84450 }, { "epoch": 2.855404171878698, "grad_norm": 1.1727477312088013, "learning_rate": 6.665734788972711e-05, "loss": 0.0619, "step": 84460 }, { "epoch": 2.855742249568951, "grad_norm": 1.6707873344421387, "learning_rate": 6.664955258599444e-05, "loss": 0.0606, "step": 84470 }, { "epoch": 2.856080327259204, "grad_norm": 1.0030187368392944, "learning_rate": 6.664175682706924e-05, "loss": 0.0596, "step": 84480 }, { "epoch": 2.8564184049494576, "grad_norm": 0.9083448648452759, "learning_rate": 6.663396061316467e-05, "loss": 0.06, "step": 84490 }, { "epoch": 2.8567564826397107, "grad_norm": 0.7520106434822083, "learning_rate": 6.662616394449385e-05, "loss": 0.0571, "step": 84500 }, { "epoch": 2.857094560329964, "grad_norm": 1.3325164318084717, "learning_rate": 6.661836682126998e-05, "loss": 0.0572, "step": 84510 }, { "epoch": 2.8574326380202173, "grad_norm": 1.2932692766189575, "learning_rate": 6.661056924370618e-05, "loss": 0.0538, "step": 84520 }, { "epoch": 2.8577707157104704, "grad_norm": 0.8961428999900818, "learning_rate": 6.660277121201567e-05, "loss": 0.0572, "step": 84530 }, { "epoch": 2.8581087934007234, "grad_norm": 1.541192650794983, "learning_rate": 6.659497272641163e-05, "loss": 0.0559, "step": 84540 }, { "epoch": 2.858446871090977, "grad_norm": 1.0644314289093018, "learning_rate": 6.658717378710728e-05, "loss": 0.0572, "step": 84550 }, { "epoch": 2.85878494878123, "grad_norm": 2.7165234088897705, "learning_rate": 6.657937439431582e-05, "loss": 0.0543, "step": 84560 }, { "epoch": 2.859123026471483, "grad_norm": 1.8388407230377197, "learning_rate": 6.65715745482505e-05, "loss": 0.066, "step": 84570 }, { "epoch": 2.8594611041617366, "grad_norm": 1.3376319408416748, "learning_rate": 6.656377424912455e-05, "loss": 0.0568, "step": 84580 }, { "epoch": 2.8597991818519897, "grad_norm": 0.9755160808563232, "learning_rate": 6.655597349715125e-05, "loss": 0.0605, "step": 84590 }, { "epoch": 2.8601372595422427, "grad_norm": 2.3466312885284424, "learning_rate": 6.654817229254383e-05, "loss": 0.0602, "step": 84600 }, { "epoch": 2.8604753372324963, "grad_norm": 0.775360107421875, "learning_rate": 6.65403706355156e-05, "loss": 0.0548, "step": 84610 }, { "epoch": 2.8608134149227493, "grad_norm": 1.4625493288040161, "learning_rate": 6.653256852627987e-05, "loss": 0.0603, "step": 84620 }, { "epoch": 2.8611514926130024, "grad_norm": 1.2749338150024414, "learning_rate": 6.652476596504991e-05, "loss": 0.0658, "step": 84630 }, { "epoch": 2.861489570303256, "grad_norm": 1.4050973653793335, "learning_rate": 6.651696295203907e-05, "loss": 0.0609, "step": 84640 }, { "epoch": 2.861827647993509, "grad_norm": 3.902038335800171, "learning_rate": 6.650915948746064e-05, "loss": 0.073, "step": 84650 }, { "epoch": 2.862165725683762, "grad_norm": 0.8534451127052307, "learning_rate": 6.650135557152802e-05, "loss": 0.0632, "step": 84660 }, { "epoch": 2.8625038033740156, "grad_norm": 1.1969236135482788, "learning_rate": 6.649355120445453e-05, "loss": 0.0664, "step": 84670 }, { "epoch": 2.8628418810642686, "grad_norm": 1.6901767253875732, "learning_rate": 6.648574638645355e-05, "loss": 0.0695, "step": 84680 }, { "epoch": 2.8631799587545217, "grad_norm": 1.2408740520477295, "learning_rate": 6.647794111773843e-05, "loss": 0.0566, "step": 84690 }, { "epoch": 2.863518036444775, "grad_norm": 0.8740754127502441, "learning_rate": 6.647013539852264e-05, "loss": 0.0567, "step": 84700 }, { "epoch": 2.8638561141350283, "grad_norm": 1.6095713376998901, "learning_rate": 6.646232922901951e-05, "loss": 0.0627, "step": 84710 }, { "epoch": 2.8641941918252813, "grad_norm": 1.9838767051696777, "learning_rate": 6.645452260944248e-05, "loss": 0.0601, "step": 84720 }, { "epoch": 2.864532269515535, "grad_norm": 0.7435847520828247, "learning_rate": 6.644671554000498e-05, "loss": 0.0632, "step": 84730 }, { "epoch": 2.864870347205788, "grad_norm": 1.2179399728775024, "learning_rate": 6.643890802092046e-05, "loss": 0.0619, "step": 84740 }, { "epoch": 2.865208424896041, "grad_norm": 1.7280313968658447, "learning_rate": 6.643110005240237e-05, "loss": 0.0597, "step": 84750 }, { "epoch": 2.8655465025862945, "grad_norm": 3.4371020793914795, "learning_rate": 6.642329163466418e-05, "loss": 0.0609, "step": 84760 }, { "epoch": 2.8658845802765476, "grad_norm": 1.3287975788116455, "learning_rate": 6.641548276791938e-05, "loss": 0.0577, "step": 84770 }, { "epoch": 2.8662226579668006, "grad_norm": 0.9371561408042908, "learning_rate": 6.640767345238143e-05, "loss": 0.0643, "step": 84780 }, { "epoch": 2.866560735657054, "grad_norm": 2.194324493408203, "learning_rate": 6.639986368826386e-05, "loss": 0.0652, "step": 84790 }, { "epoch": 2.8668988133473072, "grad_norm": 0.9277855157852173, "learning_rate": 6.639205347578017e-05, "loss": 0.0674, "step": 84800 }, { "epoch": 2.8672368910375603, "grad_norm": 0.9682068228721619, "learning_rate": 6.638424281514391e-05, "loss": 0.0638, "step": 84810 }, { "epoch": 2.867574968727814, "grad_norm": 0.7209556698799133, "learning_rate": 6.63764317065686e-05, "loss": 0.0649, "step": 84820 }, { "epoch": 2.867913046418067, "grad_norm": 1.004522442817688, "learning_rate": 6.63686201502678e-05, "loss": 0.0652, "step": 84830 }, { "epoch": 2.86825112410832, "grad_norm": 0.6761596202850342, "learning_rate": 6.636080814645506e-05, "loss": 0.0592, "step": 84840 }, { "epoch": 2.8685892017985735, "grad_norm": 1.852986216545105, "learning_rate": 6.635299569534398e-05, "loss": 0.0595, "step": 84850 }, { "epoch": 2.8689272794888265, "grad_norm": 0.6166626214981079, "learning_rate": 6.634518279714815e-05, "loss": 0.0605, "step": 84860 }, { "epoch": 2.8692653571790796, "grad_norm": 0.95627760887146, "learning_rate": 6.633736945208115e-05, "loss": 0.0705, "step": 84870 }, { "epoch": 2.869603434869333, "grad_norm": 1.5584867000579834, "learning_rate": 6.632955566035663e-05, "loss": 0.0568, "step": 84880 }, { "epoch": 2.869941512559586, "grad_norm": 1.0500582456588745, "learning_rate": 6.632174142218816e-05, "loss": 0.0643, "step": 84890 }, { "epoch": 2.8702795902498393, "grad_norm": 0.9104843735694885, "learning_rate": 6.631392673778943e-05, "loss": 0.0634, "step": 84900 }, { "epoch": 2.8706176679400928, "grad_norm": 0.6736263036727905, "learning_rate": 6.630611160737408e-05, "loss": 0.0555, "step": 84910 }, { "epoch": 2.870955745630346, "grad_norm": 1.5216081142425537, "learning_rate": 6.629829603115576e-05, "loss": 0.0591, "step": 84920 }, { "epoch": 2.871293823320599, "grad_norm": 0.706312358379364, "learning_rate": 6.629048000934814e-05, "loss": 0.0564, "step": 84930 }, { "epoch": 2.8716319010108524, "grad_norm": 0.5249950885772705, "learning_rate": 6.628266354216491e-05, "loss": 0.0572, "step": 84940 }, { "epoch": 2.8719699787011055, "grad_norm": 0.9540312886238098, "learning_rate": 6.62748466298198e-05, "loss": 0.0576, "step": 84950 }, { "epoch": 2.8723080563913586, "grad_norm": 0.8051903247833252, "learning_rate": 6.626702927252648e-05, "loss": 0.0629, "step": 84960 }, { "epoch": 2.872646134081612, "grad_norm": 0.7665297389030457, "learning_rate": 6.625921147049871e-05, "loss": 0.0577, "step": 84970 }, { "epoch": 2.872984211771865, "grad_norm": 0.9119502305984497, "learning_rate": 6.625139322395019e-05, "loss": 0.0646, "step": 84980 }, { "epoch": 2.873322289462118, "grad_norm": 0.6811774969100952, "learning_rate": 6.624357453309468e-05, "loss": 0.0621, "step": 84990 }, { "epoch": 2.8736603671523717, "grad_norm": 0.6440602540969849, "learning_rate": 6.623575539814595e-05, "loss": 0.0609, "step": 85000 }, { "epoch": 2.873998444842625, "grad_norm": 1.053259253501892, "learning_rate": 6.622793581931776e-05, "loss": 0.0597, "step": 85010 }, { "epoch": 2.874336522532878, "grad_norm": 1.8088607788085938, "learning_rate": 6.622011579682392e-05, "loss": 0.0616, "step": 85020 }, { "epoch": 2.8746746002231314, "grad_norm": 0.4661022126674652, "learning_rate": 6.621229533087821e-05, "loss": 0.0633, "step": 85030 }, { "epoch": 2.8750126779133844, "grad_norm": 0.677878737449646, "learning_rate": 6.620447442169441e-05, "loss": 0.0594, "step": 85040 }, { "epoch": 2.8753507556036375, "grad_norm": 0.6819843053817749, "learning_rate": 6.619665306948638e-05, "loss": 0.0583, "step": 85050 }, { "epoch": 2.875688833293891, "grad_norm": 0.47314557433128357, "learning_rate": 6.618883127446793e-05, "loss": 0.0546, "step": 85060 }, { "epoch": 2.876026910984144, "grad_norm": 1.1055539846420288, "learning_rate": 6.618100903685294e-05, "loss": 0.0597, "step": 85070 }, { "epoch": 2.876364988674397, "grad_norm": 0.3413068950176239, "learning_rate": 6.617318635685522e-05, "loss": 0.0601, "step": 85080 }, { "epoch": 2.8767030663646507, "grad_norm": 0.8601422309875488, "learning_rate": 6.616536323468868e-05, "loss": 0.06, "step": 85090 }, { "epoch": 2.8770411440549037, "grad_norm": 0.6375895142555237, "learning_rate": 6.615753967056717e-05, "loss": 0.0541, "step": 85100 }, { "epoch": 2.877379221745157, "grad_norm": 0.4075665771961212, "learning_rate": 6.614971566470459e-05, "loss": 0.0586, "step": 85110 }, { "epoch": 2.8777172994354103, "grad_norm": 1.4064900875091553, "learning_rate": 6.614189121731486e-05, "loss": 0.0531, "step": 85120 }, { "epoch": 2.8780553771256634, "grad_norm": 0.7114781141281128, "learning_rate": 6.613406632861189e-05, "loss": 0.0649, "step": 85130 }, { "epoch": 2.8783934548159165, "grad_norm": 0.4879671633243561, "learning_rate": 6.61262409988096e-05, "loss": 0.0607, "step": 85140 }, { "epoch": 2.87873153250617, "grad_norm": 4.380626201629639, "learning_rate": 6.611841522812194e-05, "loss": 0.0565, "step": 85150 }, { "epoch": 2.879069610196423, "grad_norm": 0.5212727785110474, "learning_rate": 6.611058901676287e-05, "loss": 0.0586, "step": 85160 }, { "epoch": 2.879407687886676, "grad_norm": 0.6295210123062134, "learning_rate": 6.610276236494633e-05, "loss": 0.0515, "step": 85170 }, { "epoch": 2.8797457655769296, "grad_norm": 1.1291757822036743, "learning_rate": 6.609493527288634e-05, "loss": 0.0619, "step": 85180 }, { "epoch": 2.8800838432671827, "grad_norm": 1.8478500843048096, "learning_rate": 6.608710774079685e-05, "loss": 0.0569, "step": 85190 }, { "epoch": 2.8804219209574358, "grad_norm": 0.5118518471717834, "learning_rate": 6.607927976889189e-05, "loss": 0.0617, "step": 85200 }, { "epoch": 2.8807599986476893, "grad_norm": 1.5557969808578491, "learning_rate": 6.607145135738547e-05, "loss": 0.0566, "step": 85210 }, { "epoch": 2.8810980763379423, "grad_norm": 0.3507290184497833, "learning_rate": 6.606362250649158e-05, "loss": 0.0627, "step": 85220 }, { "epoch": 2.8814361540281954, "grad_norm": 0.5728627443313599, "learning_rate": 6.60557932164243e-05, "loss": 0.0541, "step": 85230 }, { "epoch": 2.881774231718449, "grad_norm": 0.5253920555114746, "learning_rate": 6.604796348739767e-05, "loss": 0.062, "step": 85240 }, { "epoch": 2.882112309408702, "grad_norm": 1.228994369506836, "learning_rate": 6.604013331962574e-05, "loss": 0.0603, "step": 85250 }, { "epoch": 2.882450387098955, "grad_norm": 0.7142907977104187, "learning_rate": 6.60323027133226e-05, "loss": 0.0543, "step": 85260 }, { "epoch": 2.8827884647892086, "grad_norm": 0.4451817274093628, "learning_rate": 6.602447166870231e-05, "loss": 0.059, "step": 85270 }, { "epoch": 2.8831265424794617, "grad_norm": 1.1129214763641357, "learning_rate": 6.6016640185979e-05, "loss": 0.0587, "step": 85280 }, { "epoch": 2.883464620169715, "grad_norm": 0.4882674217224121, "learning_rate": 6.600880826536677e-05, "loss": 0.0637, "step": 85290 }, { "epoch": 2.8838026978599682, "grad_norm": 0.518145740032196, "learning_rate": 6.600097590707972e-05, "loss": 0.0565, "step": 85300 }, { "epoch": 2.8841407755502213, "grad_norm": 0.5823004245758057, "learning_rate": 6.5993143111332e-05, "loss": 0.0565, "step": 85310 }, { "epoch": 2.884478853240475, "grad_norm": 0.5162341594696045, "learning_rate": 6.598530987833776e-05, "loss": 0.0647, "step": 85320 }, { "epoch": 2.884816930930728, "grad_norm": 0.5645075440406799, "learning_rate": 6.597747620831116e-05, "loss": 0.0539, "step": 85330 }, { "epoch": 2.885155008620981, "grad_norm": 1.1018123626708984, "learning_rate": 6.596964210146636e-05, "loss": 0.0632, "step": 85340 }, { "epoch": 2.8854930863112345, "grad_norm": 0.8245125412940979, "learning_rate": 6.596180755801754e-05, "loss": 0.0593, "step": 85350 }, { "epoch": 2.8858311640014875, "grad_norm": 0.8069275617599487, "learning_rate": 6.59539725781789e-05, "loss": 0.0572, "step": 85360 }, { "epoch": 2.8861692416917406, "grad_norm": 0.8413758873939514, "learning_rate": 6.594613716216464e-05, "loss": 0.0625, "step": 85370 }, { "epoch": 2.886507319381994, "grad_norm": 1.5890295505523682, "learning_rate": 6.593830131018898e-05, "loss": 0.0626, "step": 85380 }, { "epoch": 2.886845397072247, "grad_norm": 0.5954417586326599, "learning_rate": 6.593046502246616e-05, "loss": 0.0561, "step": 85390 }, { "epoch": 2.8871834747625003, "grad_norm": 0.563956081867218, "learning_rate": 6.59226282992104e-05, "loss": 0.0586, "step": 85400 }, { "epoch": 2.8875215524527538, "grad_norm": 0.748414933681488, "learning_rate": 6.591479114063597e-05, "loss": 0.0563, "step": 85410 }, { "epoch": 2.887859630143007, "grad_norm": 0.9937997460365295, "learning_rate": 6.590695354695714e-05, "loss": 0.0628, "step": 85420 }, { "epoch": 2.8881977078332604, "grad_norm": 0.42362791299819946, "learning_rate": 6.589911551838815e-05, "loss": 0.0591, "step": 85430 }, { "epoch": 2.8885357855235134, "grad_norm": 0.48464882373809814, "learning_rate": 6.589127705514334e-05, "loss": 0.0511, "step": 85440 }, { "epoch": 2.8888738632137665, "grad_norm": 0.5476806163787842, "learning_rate": 6.588343815743697e-05, "loss": 0.0591, "step": 85450 }, { "epoch": 2.88921194090402, "grad_norm": 0.9691774845123291, "learning_rate": 6.587559882548337e-05, "loss": 0.0662, "step": 85460 }, { "epoch": 2.889550018594273, "grad_norm": 0.5731031894683838, "learning_rate": 6.586775905949687e-05, "loss": 0.0571, "step": 85470 }, { "epoch": 2.889888096284526, "grad_norm": 0.48343124985694885, "learning_rate": 6.585991885969179e-05, "loss": 0.0588, "step": 85480 }, { "epoch": 2.8902261739747797, "grad_norm": 0.5666456818580627, "learning_rate": 6.58520782262825e-05, "loss": 0.0642, "step": 85490 }, { "epoch": 2.8905642516650327, "grad_norm": 0.5891483426094055, "learning_rate": 6.584423715948332e-05, "loss": 0.0643, "step": 85500 }, { "epoch": 2.890902329355286, "grad_norm": 0.5252359509468079, "learning_rate": 6.583639565950868e-05, "loss": 0.0636, "step": 85510 }, { "epoch": 2.8912404070455393, "grad_norm": 0.6600485444068909, "learning_rate": 6.58285537265729e-05, "loss": 0.0624, "step": 85520 }, { "epoch": 2.8915784847357924, "grad_norm": 0.7324530482292175, "learning_rate": 6.582071136089044e-05, "loss": 0.0591, "step": 85530 }, { "epoch": 2.8919165624260454, "grad_norm": 0.48724010586738586, "learning_rate": 6.581286856267565e-05, "loss": 0.0582, "step": 85540 }, { "epoch": 2.892254640116299, "grad_norm": 0.6949857473373413, "learning_rate": 6.5805025332143e-05, "loss": 0.0623, "step": 85550 }, { "epoch": 2.892592717806552, "grad_norm": 0.6228746771812439, "learning_rate": 6.579718166950686e-05, "loss": 0.065, "step": 85560 }, { "epoch": 2.892930795496805, "grad_norm": 0.781182050704956, "learning_rate": 6.578933757498174e-05, "loss": 0.0612, "step": 85570 }, { "epoch": 2.8932688731870586, "grad_norm": 0.5780925154685974, "learning_rate": 6.578149304878204e-05, "loss": 0.0641, "step": 85580 }, { "epoch": 2.8936069508773117, "grad_norm": 0.6700589656829834, "learning_rate": 6.577364809112227e-05, "loss": 0.0574, "step": 85590 }, { "epoch": 2.8939450285675647, "grad_norm": 2.178802013397217, "learning_rate": 6.576580270221687e-05, "loss": 0.0588, "step": 85600 }, { "epoch": 2.8942831062578183, "grad_norm": 2.1143743991851807, "learning_rate": 6.575795688228035e-05, "loss": 0.0592, "step": 85610 }, { "epoch": 2.8946211839480713, "grad_norm": 1.858174443244934, "learning_rate": 6.575011063152721e-05, "loss": 0.0592, "step": 85620 }, { "epoch": 2.8949592616383244, "grad_norm": 1.407701015472412, "learning_rate": 6.574226395017196e-05, "loss": 0.0632, "step": 85630 }, { "epoch": 2.895297339328578, "grad_norm": 0.4885462522506714, "learning_rate": 6.573441683842913e-05, "loss": 0.0632, "step": 85640 }, { "epoch": 2.895635417018831, "grad_norm": 0.8399273753166199, "learning_rate": 6.572656929651327e-05, "loss": 0.0655, "step": 85650 }, { "epoch": 2.895973494709084, "grad_norm": 0.4933319091796875, "learning_rate": 6.57187213246389e-05, "loss": 0.0548, "step": 85660 }, { "epoch": 2.8963115723993376, "grad_norm": 0.8443335890769958, "learning_rate": 6.571087292302059e-05, "loss": 0.0545, "step": 85670 }, { "epoch": 2.8966496500895906, "grad_norm": 0.4389195442199707, "learning_rate": 6.570302409187292e-05, "loss": 0.0614, "step": 85680 }, { "epoch": 2.8969877277798437, "grad_norm": 0.6800221800804138, "learning_rate": 6.569517483141047e-05, "loss": 0.0589, "step": 85690 }, { "epoch": 2.897325805470097, "grad_norm": 0.6513642072677612, "learning_rate": 6.568732514184784e-05, "loss": 0.0623, "step": 85700 }, { "epoch": 2.8976638831603503, "grad_norm": 0.7380992770195007, "learning_rate": 6.567947502339962e-05, "loss": 0.0592, "step": 85710 }, { "epoch": 2.8980019608506034, "grad_norm": 1.5570374727249146, "learning_rate": 6.567162447628048e-05, "loss": 0.0568, "step": 85720 }, { "epoch": 2.898340038540857, "grad_norm": 0.6237925887107849, "learning_rate": 6.566377350070499e-05, "loss": 0.0596, "step": 85730 }, { "epoch": 2.89867811623111, "grad_norm": 0.7145414352416992, "learning_rate": 6.565592209688781e-05, "loss": 0.0608, "step": 85740 }, { "epoch": 2.899016193921363, "grad_norm": 0.442921906709671, "learning_rate": 6.56480702650436e-05, "loss": 0.0612, "step": 85750 }, { "epoch": 2.8993542716116165, "grad_norm": 0.6098846197128296, "learning_rate": 6.564021800538705e-05, "loss": 0.065, "step": 85760 }, { "epoch": 2.8996923493018696, "grad_norm": 0.49945250153541565, "learning_rate": 6.563236531813278e-05, "loss": 0.0572, "step": 85770 }, { "epoch": 2.9000304269921227, "grad_norm": 1.639212727546692, "learning_rate": 6.562451220349555e-05, "loss": 0.0574, "step": 85780 }, { "epoch": 2.900368504682376, "grad_norm": 1.0426716804504395, "learning_rate": 6.561665866169001e-05, "loss": 0.0594, "step": 85790 }, { "epoch": 2.9007065823726292, "grad_norm": 0.5704402327537537, "learning_rate": 6.560880469293089e-05, "loss": 0.0616, "step": 85800 }, { "epoch": 2.9010446600628823, "grad_norm": 0.9686047434806824, "learning_rate": 6.56009502974329e-05, "loss": 0.056, "step": 85810 }, { "epoch": 2.901382737753136, "grad_norm": 1.3674817085266113, "learning_rate": 6.559309547541083e-05, "loss": 0.0575, "step": 85820 }, { "epoch": 2.901720815443389, "grad_norm": 2.7352473735809326, "learning_rate": 6.558524022707935e-05, "loss": 0.0559, "step": 85830 }, { "epoch": 2.902058893133642, "grad_norm": 1.5116277933120728, "learning_rate": 6.557738455265327e-05, "loss": 0.0637, "step": 85840 }, { "epoch": 2.9023969708238955, "grad_norm": 2.191676616668701, "learning_rate": 6.556952845234735e-05, "loss": 0.0563, "step": 85850 }, { "epoch": 2.9027350485141485, "grad_norm": 2.5676181316375732, "learning_rate": 6.556167192637637e-05, "loss": 0.0611, "step": 85860 }, { "epoch": 2.9030731262044016, "grad_norm": 1.4767247438430786, "learning_rate": 6.555381497495514e-05, "loss": 0.0639, "step": 85870 }, { "epoch": 2.903411203894655, "grad_norm": 3.1507742404937744, "learning_rate": 6.554595759829842e-05, "loss": 0.0656, "step": 85880 }, { "epoch": 2.903749281584908, "grad_norm": 1.9351544380187988, "learning_rate": 6.553809979662109e-05, "loss": 0.0576, "step": 85890 }, { "epoch": 2.9040873592751613, "grad_norm": 2.520773410797119, "learning_rate": 6.553024157013793e-05, "loss": 0.0594, "step": 85900 }, { "epoch": 2.9044254369654148, "grad_norm": 3.2412338256835938, "learning_rate": 6.552238291906381e-05, "loss": 0.0572, "step": 85910 }, { "epoch": 2.904763514655668, "grad_norm": 1.0581729412078857, "learning_rate": 6.551452384361356e-05, "loss": 0.0591, "step": 85920 }, { "epoch": 2.905101592345921, "grad_norm": 2.9283041954040527, "learning_rate": 6.550666434400207e-05, "loss": 0.0622, "step": 85930 }, { "epoch": 2.9054396700361744, "grad_norm": 4.095746040344238, "learning_rate": 6.54988044204442e-05, "loss": 0.0605, "step": 85940 }, { "epoch": 2.9057777477264275, "grad_norm": 4.739996433258057, "learning_rate": 6.549094407315484e-05, "loss": 0.0608, "step": 85950 }, { "epoch": 2.9061158254166806, "grad_norm": 1.2219213247299194, "learning_rate": 6.548308330234887e-05, "loss": 0.0641, "step": 85960 }, { "epoch": 2.906453903106934, "grad_norm": 1.0251166820526123, "learning_rate": 6.547522210824126e-05, "loss": 0.0605, "step": 85970 }, { "epoch": 2.906791980797187, "grad_norm": 1.2184727191925049, "learning_rate": 6.546736049104686e-05, "loss": 0.0675, "step": 85980 }, { "epoch": 2.90713005848744, "grad_norm": 2.2918994426727295, "learning_rate": 6.545949845098064e-05, "loss": 0.0591, "step": 85990 }, { "epoch": 2.9074681361776937, "grad_norm": 1.487625241279602, "learning_rate": 6.545163598825755e-05, "loss": 0.0593, "step": 86000 }, { "epoch": 2.907806213867947, "grad_norm": 0.8814974427223206, "learning_rate": 6.544377310309252e-05, "loss": 0.0563, "step": 86010 }, { "epoch": 2.9081442915582, "grad_norm": 2.3765645027160645, "learning_rate": 6.543590979570053e-05, "loss": 0.0629, "step": 86020 }, { "epoch": 2.9084823692484534, "grad_norm": 4.521368503570557, "learning_rate": 6.54280460662966e-05, "loss": 0.0567, "step": 86030 }, { "epoch": 2.9088204469387064, "grad_norm": 1.9833890199661255, "learning_rate": 6.542018191509565e-05, "loss": 0.0675, "step": 86040 }, { "epoch": 2.9091585246289595, "grad_norm": 3.5602355003356934, "learning_rate": 6.541231734231275e-05, "loss": 0.0625, "step": 86050 }, { "epoch": 2.909496602319213, "grad_norm": 1.5445406436920166, "learning_rate": 6.540445234816287e-05, "loss": 0.0688, "step": 86060 }, { "epoch": 2.909834680009466, "grad_norm": 2.0839407444000244, "learning_rate": 6.539658693286105e-05, "loss": 0.0607, "step": 86070 }, { "epoch": 2.910172757699719, "grad_norm": 2.760575771331787, "learning_rate": 6.538872109662232e-05, "loss": 0.0654, "step": 86080 }, { "epoch": 2.9105108353899727, "grad_norm": 1.1096545457839966, "learning_rate": 6.538085483966174e-05, "loss": 0.0649, "step": 86090 }, { "epoch": 2.9108489130802258, "grad_norm": 1.8139592409133911, "learning_rate": 6.537298816219437e-05, "loss": 0.0664, "step": 86100 }, { "epoch": 2.911186990770479, "grad_norm": 2.4818360805511475, "learning_rate": 6.536512106443529e-05, "loss": 0.064, "step": 86110 }, { "epoch": 2.9115250684607323, "grad_norm": 0.8702841997146606, "learning_rate": 6.535725354659956e-05, "loss": 0.0582, "step": 86120 }, { "epoch": 2.9118631461509854, "grad_norm": 1.9091565608978271, "learning_rate": 6.534938560890229e-05, "loss": 0.0703, "step": 86130 }, { "epoch": 2.9122012238412385, "grad_norm": 1.6850658655166626, "learning_rate": 6.534151725155858e-05, "loss": 0.0635, "step": 86140 }, { "epoch": 2.912539301531492, "grad_norm": 0.9032342433929443, "learning_rate": 6.533364847478355e-05, "loss": 0.0606, "step": 86150 }, { "epoch": 2.912877379221745, "grad_norm": 0.9306918978691101, "learning_rate": 6.532577927879235e-05, "loss": 0.0617, "step": 86160 }, { "epoch": 2.913215456911998, "grad_norm": 0.8061515092849731, "learning_rate": 6.531790966380008e-05, "loss": 0.0657, "step": 86170 }, { "epoch": 2.9135535346022516, "grad_norm": 0.8578261137008667, "learning_rate": 6.531003963002193e-05, "loss": 0.068, "step": 86180 }, { "epoch": 2.9138916122925047, "grad_norm": 0.9487699270248413, "learning_rate": 6.530216917767304e-05, "loss": 0.0611, "step": 86190 }, { "epoch": 2.9142296899827578, "grad_norm": 2.045192241668701, "learning_rate": 6.529429830696859e-05, "loss": 0.0647, "step": 86200 }, { "epoch": 2.9145677676730113, "grad_norm": 11.97752857208252, "learning_rate": 6.528642701812378e-05, "loss": 0.0648, "step": 86210 }, { "epoch": 2.9149058453632644, "grad_norm": 2.7013258934020996, "learning_rate": 6.527855531135379e-05, "loss": 0.0678, "step": 86220 }, { "epoch": 2.9152439230535174, "grad_norm": 1.3697054386138916, "learning_rate": 6.527068318687384e-05, "loss": 0.0633, "step": 86230 }, { "epoch": 2.915582000743771, "grad_norm": 3.274371862411499, "learning_rate": 6.526281064489916e-05, "loss": 0.0688, "step": 86240 }, { "epoch": 2.915920078434024, "grad_norm": 7.838155746459961, "learning_rate": 6.525493768564496e-05, "loss": 0.0682, "step": 86250 }, { "epoch": 2.9162581561242775, "grad_norm": 1.5610289573669434, "learning_rate": 6.52470643093265e-05, "loss": 0.0716, "step": 86260 }, { "epoch": 2.9165962338145306, "grad_norm": 2.3103020191192627, "learning_rate": 6.5239190516159e-05, "loss": 0.0691, "step": 86270 }, { "epoch": 2.9169343115047837, "grad_norm": 3.8122634887695312, "learning_rate": 6.523131630635779e-05, "loss": 0.0708, "step": 86280 }, { "epoch": 2.917272389195037, "grad_norm": 1.517092227935791, "learning_rate": 6.522344168013811e-05, "loss": 0.067, "step": 86290 }, { "epoch": 2.9176104668852902, "grad_norm": 1.4438518285751343, "learning_rate": 6.521556663771524e-05, "loss": 0.0625, "step": 86300 }, { "epoch": 2.9179485445755433, "grad_norm": 1.240980625152588, "learning_rate": 6.520769117930452e-05, "loss": 0.0665, "step": 86310 }, { "epoch": 2.918286622265797, "grad_norm": 1.2878367900848389, "learning_rate": 6.519981530512121e-05, "loss": 0.064, "step": 86320 }, { "epoch": 2.91862469995605, "grad_norm": 1.0579822063446045, "learning_rate": 6.519193901538066e-05, "loss": 0.0614, "step": 86330 }, { "epoch": 2.918962777646303, "grad_norm": 2.6409456729888916, "learning_rate": 6.518406231029822e-05, "loss": 0.0693, "step": 86340 }, { "epoch": 2.9193008553365565, "grad_norm": 0.8855243921279907, "learning_rate": 6.517618519008922e-05, "loss": 0.0602, "step": 86350 }, { "epoch": 2.9196389330268095, "grad_norm": 1.4155683517456055, "learning_rate": 6.516830765496902e-05, "loss": 0.0588, "step": 86360 }, { "epoch": 2.9199770107170626, "grad_norm": 0.6957307457923889, "learning_rate": 6.516042970515298e-05, "loss": 0.065, "step": 86370 }, { "epoch": 2.920315088407316, "grad_norm": 0.7027822136878967, "learning_rate": 6.515255134085649e-05, "loss": 0.0613, "step": 86380 }, { "epoch": 2.920653166097569, "grad_norm": 1.2178852558135986, "learning_rate": 6.514467256229493e-05, "loss": 0.0604, "step": 86390 }, { "epoch": 2.9209912437878227, "grad_norm": 3.6960341930389404, "learning_rate": 6.513679336968372e-05, "loss": 0.0676, "step": 86400 }, { "epoch": 2.921329321478076, "grad_norm": 1.4712083339691162, "learning_rate": 6.512891376323827e-05, "loss": 0.0724, "step": 86410 }, { "epoch": 2.921667399168329, "grad_norm": 3.181610107421875, "learning_rate": 6.5121033743174e-05, "loss": 0.0671, "step": 86420 }, { "epoch": 2.9220054768585824, "grad_norm": 0.9390160441398621, "learning_rate": 6.511315330970635e-05, "loss": 0.0703, "step": 86430 }, { "epoch": 2.9223435545488354, "grad_norm": 1.2577552795410156, "learning_rate": 6.510527246305075e-05, "loss": 0.0675, "step": 86440 }, { "epoch": 2.9226816322390885, "grad_norm": 8.819046974182129, "learning_rate": 6.509739120342269e-05, "loss": 0.0645, "step": 86450 }, { "epoch": 2.923019709929342, "grad_norm": 1.0179754495620728, "learning_rate": 6.50895095310376e-05, "loss": 0.0601, "step": 86460 }, { "epoch": 2.923357787619595, "grad_norm": 2.2934868335723877, "learning_rate": 6.508162744611104e-05, "loss": 0.0637, "step": 86470 }, { "epoch": 2.923695865309848, "grad_norm": 1.6002724170684814, "learning_rate": 6.507374494885841e-05, "loss": 0.0628, "step": 86480 }, { "epoch": 2.9240339430001017, "grad_norm": 0.8336268663406372, "learning_rate": 6.506586203949528e-05, "loss": 0.0597, "step": 86490 }, { "epoch": 2.9243720206903547, "grad_norm": 1.3089543581008911, "learning_rate": 6.505797871823712e-05, "loss": 0.0614, "step": 86500 }, { "epoch": 2.924710098380608, "grad_norm": 0.6787895560264587, "learning_rate": 6.505009498529949e-05, "loss": 0.061, "step": 86510 }, { "epoch": 2.9250481760708613, "grad_norm": 0.7643451690673828, "learning_rate": 6.504221084089791e-05, "loss": 0.0619, "step": 86520 }, { "epoch": 2.9253862537611144, "grad_norm": 0.873477578163147, "learning_rate": 6.503432628524795e-05, "loss": 0.062, "step": 86530 }, { "epoch": 2.9257243314513675, "grad_norm": 2.5005409717559814, "learning_rate": 6.502644131856514e-05, "loss": 0.0703, "step": 86540 }, { "epoch": 2.926062409141621, "grad_norm": 0.789378821849823, "learning_rate": 6.501855594106508e-05, "loss": 0.0587, "step": 86550 }, { "epoch": 2.926400486831874, "grad_norm": 0.8335222601890564, "learning_rate": 6.501067015296336e-05, "loss": 0.0604, "step": 86560 }, { "epoch": 2.926738564522127, "grad_norm": 0.815604031085968, "learning_rate": 6.500278395447553e-05, "loss": 0.0601, "step": 86570 }, { "epoch": 2.9270766422123806, "grad_norm": 1.1438829898834229, "learning_rate": 6.499489734581724e-05, "loss": 0.0613, "step": 86580 }, { "epoch": 2.9274147199026337, "grad_norm": 0.9266244173049927, "learning_rate": 6.498701032720406e-05, "loss": 0.0675, "step": 86590 }, { "epoch": 2.9277527975928868, "grad_norm": 1.2283917665481567, "learning_rate": 6.497912289885169e-05, "loss": 0.0592, "step": 86600 }, { "epoch": 2.9280908752831403, "grad_norm": 0.7487992644309998, "learning_rate": 6.497123506097571e-05, "loss": 0.0688, "step": 86610 }, { "epoch": 2.9284289529733933, "grad_norm": 0.5620054602622986, "learning_rate": 6.496334681379179e-05, "loss": 0.0574, "step": 86620 }, { "epoch": 2.9287670306636464, "grad_norm": 1.2432996034622192, "learning_rate": 6.495545815751558e-05, "loss": 0.0642, "step": 86630 }, { "epoch": 2.9291051083539, "grad_norm": 0.9540862441062927, "learning_rate": 6.494756909236277e-05, "loss": 0.062, "step": 86640 }, { "epoch": 2.929443186044153, "grad_norm": 1.285940408706665, "learning_rate": 6.493967961854903e-05, "loss": 0.0593, "step": 86650 }, { "epoch": 2.929781263734406, "grad_norm": 1.1421352624893188, "learning_rate": 6.493178973629007e-05, "loss": 0.0626, "step": 86660 }, { "epoch": 2.9301193414246596, "grad_norm": 0.8641250133514404, "learning_rate": 6.492389944580159e-05, "loss": 0.0545, "step": 86670 }, { "epoch": 2.9304574191149126, "grad_norm": 0.6111859083175659, "learning_rate": 6.491600874729931e-05, "loss": 0.0628, "step": 86680 }, { "epoch": 2.9307954968051657, "grad_norm": 1.186697006225586, "learning_rate": 6.490811764099893e-05, "loss": 0.0647, "step": 86690 }, { "epoch": 2.9311335744954192, "grad_norm": 0.4764697253704071, "learning_rate": 6.490022612711624e-05, "loss": 0.0672, "step": 86700 }, { "epoch": 2.9314716521856723, "grad_norm": 0.9523037672042847, "learning_rate": 6.489233420586695e-05, "loss": 0.0632, "step": 86710 }, { "epoch": 2.9318097298759254, "grad_norm": 0.7658494710922241, "learning_rate": 6.488444187746684e-05, "loss": 0.0605, "step": 86720 }, { "epoch": 2.932147807566179, "grad_norm": 0.6721420884132385, "learning_rate": 6.487654914213168e-05, "loss": 0.0617, "step": 86730 }, { "epoch": 2.932485885256432, "grad_norm": 0.9961069226264954, "learning_rate": 6.486865600007726e-05, "loss": 0.0568, "step": 86740 }, { "epoch": 2.932823962946685, "grad_norm": 0.5525308847427368, "learning_rate": 6.486076245151937e-05, "loss": 0.0638, "step": 86750 }, { "epoch": 2.9331620406369385, "grad_norm": 0.4872891902923584, "learning_rate": 6.485286849667382e-05, "loss": 0.0572, "step": 86760 }, { "epoch": 2.9335001183271916, "grad_norm": 3.2989413738250732, "learning_rate": 6.484497413575643e-05, "loss": 0.0565, "step": 86770 }, { "epoch": 2.9338381960174447, "grad_norm": 1.1075736284255981, "learning_rate": 6.483707936898301e-05, "loss": 0.0574, "step": 86780 }, { "epoch": 2.934176273707698, "grad_norm": 0.9186370968818665, "learning_rate": 6.482918419656941e-05, "loss": 0.0577, "step": 86790 }, { "epoch": 2.9345143513979512, "grad_norm": 0.6979706883430481, "learning_rate": 6.48212886187315e-05, "loss": 0.0571, "step": 86800 }, { "epoch": 2.9348524290882043, "grad_norm": 0.9489243030548096, "learning_rate": 6.481339263568513e-05, "loss": 0.0608, "step": 86810 }, { "epoch": 2.935190506778458, "grad_norm": 0.745445728302002, "learning_rate": 6.480549624764615e-05, "loss": 0.0604, "step": 86820 }, { "epoch": 2.935528584468711, "grad_norm": 4.012001037597656, "learning_rate": 6.479759945483048e-05, "loss": 0.0668, "step": 86830 }, { "epoch": 2.935866662158964, "grad_norm": 0.6843268275260925, "learning_rate": 6.4789702257454e-05, "loss": 0.0625, "step": 86840 }, { "epoch": 2.9362047398492175, "grad_norm": 1.4916576147079468, "learning_rate": 6.478180465573263e-05, "loss": 0.0578, "step": 86850 }, { "epoch": 2.9365428175394706, "grad_norm": 1.286845088005066, "learning_rate": 6.477390664988226e-05, "loss": 0.0672, "step": 86860 }, { "epoch": 2.9368808952297236, "grad_norm": 0.7256174683570862, "learning_rate": 6.476600824011883e-05, "loss": 0.0563, "step": 86870 }, { "epoch": 2.937218972919977, "grad_norm": 1.2423653602600098, "learning_rate": 6.475810942665828e-05, "loss": 0.0601, "step": 86880 }, { "epoch": 2.93755705061023, "grad_norm": 1.5106431245803833, "learning_rate": 6.475021020971657e-05, "loss": 0.0596, "step": 86890 }, { "epoch": 2.9378951283004833, "grad_norm": 1.8419502973556519, "learning_rate": 6.474231058950963e-05, "loss": 0.0664, "step": 86900 }, { "epoch": 2.938233205990737, "grad_norm": 1.175952434539795, "learning_rate": 6.473441056625349e-05, "loss": 0.0631, "step": 86910 }, { "epoch": 2.93857128368099, "grad_norm": 3.0465891361236572, "learning_rate": 6.472651014016408e-05, "loss": 0.0588, "step": 86920 }, { "epoch": 2.938909361371243, "grad_norm": 2.455813407897949, "learning_rate": 6.471860931145744e-05, "loss": 0.0587, "step": 86930 }, { "epoch": 2.9392474390614964, "grad_norm": 1.8346624374389648, "learning_rate": 6.471070808034953e-05, "loss": 0.0559, "step": 86940 }, { "epoch": 2.9395855167517495, "grad_norm": 1.871725082397461, "learning_rate": 6.47028064470564e-05, "loss": 0.059, "step": 86950 }, { "epoch": 2.9399235944420026, "grad_norm": 2.8554885387420654, "learning_rate": 6.469490441179405e-05, "loss": 0.0663, "step": 86960 }, { "epoch": 2.940261672132256, "grad_norm": 1.0504149198532104, "learning_rate": 6.468700197477853e-05, "loss": 0.0644, "step": 86970 }, { "epoch": 2.940599749822509, "grad_norm": 3.5199058055877686, "learning_rate": 6.46790991362259e-05, "loss": 0.0673, "step": 86980 }, { "epoch": 2.9409378275127622, "grad_norm": 1.3192323446273804, "learning_rate": 6.467119589635223e-05, "loss": 0.0666, "step": 86990 }, { "epoch": 2.9412759052030157, "grad_norm": 2.3249454498291016, "learning_rate": 6.466329225537356e-05, "loss": 0.0705, "step": 87000 }, { "epoch": 2.941613982893269, "grad_norm": 1.7016487121582031, "learning_rate": 6.4655388213506e-05, "loss": 0.0694, "step": 87010 }, { "epoch": 2.941952060583522, "grad_norm": 3.8157479763031006, "learning_rate": 6.464748377096562e-05, "loss": 0.0634, "step": 87020 }, { "epoch": 2.9422901382737754, "grad_norm": 4.959635257720947, "learning_rate": 6.463957892796854e-05, "loss": 0.0667, "step": 87030 }, { "epoch": 2.9426282159640285, "grad_norm": 2.36569881439209, "learning_rate": 6.463167368473087e-05, "loss": 0.0613, "step": 87040 }, { "epoch": 2.9429662936542815, "grad_norm": 3.8356199264526367, "learning_rate": 6.462376804146874e-05, "loss": 0.0599, "step": 87050 }, { "epoch": 2.943304371344535, "grad_norm": 3.031334638595581, "learning_rate": 6.461586199839831e-05, "loss": 0.0641, "step": 87060 }, { "epoch": 2.943642449034788, "grad_norm": 6.006106853485107, "learning_rate": 6.460795555573567e-05, "loss": 0.0615, "step": 87070 }, { "epoch": 2.943980526725041, "grad_norm": 13.971837997436523, "learning_rate": 6.460004871369703e-05, "loss": 0.063, "step": 87080 }, { "epoch": 2.9443186044152947, "grad_norm": 1.629205584526062, "learning_rate": 6.459214147249852e-05, "loss": 0.0586, "step": 87090 }, { "epoch": 2.9446566821055478, "grad_norm": 3.983452081680298, "learning_rate": 6.458423383235638e-05, "loss": 0.0661, "step": 87100 }, { "epoch": 2.944994759795801, "grad_norm": 0.9936608672142029, "learning_rate": 6.457632579348674e-05, "loss": 0.0603, "step": 87110 }, { "epoch": 2.9453328374860543, "grad_norm": 4.060879707336426, "learning_rate": 6.456841735610585e-05, "loss": 0.0622, "step": 87120 }, { "epoch": 2.9456709151763074, "grad_norm": 1.0409421920776367, "learning_rate": 6.456050852042987e-05, "loss": 0.0624, "step": 87130 }, { "epoch": 2.9460089928665605, "grad_norm": 1.065726637840271, "learning_rate": 6.45525992866751e-05, "loss": 0.0607, "step": 87140 }, { "epoch": 2.946347070556814, "grad_norm": 0.7404862642288208, "learning_rate": 6.454468965505768e-05, "loss": 0.0542, "step": 87150 }, { "epoch": 2.946685148247067, "grad_norm": 1.0934393405914307, "learning_rate": 6.453677962579396e-05, "loss": 0.0585, "step": 87160 }, { "epoch": 2.94702322593732, "grad_norm": 1.680971622467041, "learning_rate": 6.452886919910012e-05, "loss": 0.0618, "step": 87170 }, { "epoch": 2.9473613036275736, "grad_norm": 18.31496810913086, "learning_rate": 6.452095837519246e-05, "loss": 0.0654, "step": 87180 }, { "epoch": 2.9476993813178267, "grad_norm": 0.9528282284736633, "learning_rate": 6.451304715428725e-05, "loss": 0.0638, "step": 87190 }, { "epoch": 2.94803745900808, "grad_norm": 0.9437657594680786, "learning_rate": 6.450513553660079e-05, "loss": 0.0594, "step": 87200 }, { "epoch": 2.9483755366983333, "grad_norm": 2.907097101211548, "learning_rate": 6.449722352234937e-05, "loss": 0.0619, "step": 87210 }, { "epoch": 2.9487136143885864, "grad_norm": 1.6173959970474243, "learning_rate": 6.448931111174929e-05, "loss": 0.0634, "step": 87220 }, { "epoch": 2.94905169207884, "grad_norm": 0.8642231225967407, "learning_rate": 6.448139830501692e-05, "loss": 0.0623, "step": 87230 }, { "epoch": 2.949389769769093, "grad_norm": 0.723974883556366, "learning_rate": 6.447348510236853e-05, "loss": 0.0616, "step": 87240 }, { "epoch": 2.949727847459346, "grad_norm": 1.3063790798187256, "learning_rate": 6.446557150402051e-05, "loss": 0.0634, "step": 87250 }, { "epoch": 2.9500659251495995, "grad_norm": 0.8520880341529846, "learning_rate": 6.44576575101892e-05, "loss": 0.0667, "step": 87260 }, { "epoch": 2.9504040028398526, "grad_norm": 8.917579650878906, "learning_rate": 6.444974312109097e-05, "loss": 0.0623, "step": 87270 }, { "epoch": 2.9507420805301057, "grad_norm": 0.6199980974197388, "learning_rate": 6.444182833694218e-05, "loss": 0.0652, "step": 87280 }, { "epoch": 2.951080158220359, "grad_norm": 0.8363950252532959, "learning_rate": 6.443391315795923e-05, "loss": 0.06, "step": 87290 }, { "epoch": 2.9514182359106123, "grad_norm": 2.7172605991363525, "learning_rate": 6.442599758435851e-05, "loss": 0.0649, "step": 87300 }, { "epoch": 2.9517563136008653, "grad_norm": 1.8534964323043823, "learning_rate": 6.441808161635645e-05, "loss": 0.0611, "step": 87310 }, { "epoch": 2.952094391291119, "grad_norm": 0.8282179832458496, "learning_rate": 6.441016525416945e-05, "loss": 0.0591, "step": 87320 }, { "epoch": 2.952432468981372, "grad_norm": 0.8678667545318604, "learning_rate": 6.440224849801395e-05, "loss": 0.0636, "step": 87330 }, { "epoch": 2.952770546671625, "grad_norm": 0.9479806423187256, "learning_rate": 6.439433134810638e-05, "loss": 0.062, "step": 87340 }, { "epoch": 2.9531086243618785, "grad_norm": 10.481294631958008, "learning_rate": 6.43864138046632e-05, "loss": 0.0638, "step": 87350 }, { "epoch": 2.9534467020521316, "grad_norm": 0.803175687789917, "learning_rate": 6.437849586790087e-05, "loss": 0.054, "step": 87360 }, { "epoch": 2.953784779742385, "grad_norm": 1.533729910850525, "learning_rate": 6.437057753803588e-05, "loss": 0.0688, "step": 87370 }, { "epoch": 2.954122857432638, "grad_norm": 0.6572540998458862, "learning_rate": 6.436265881528468e-05, "loss": 0.0638, "step": 87380 }, { "epoch": 2.954460935122891, "grad_norm": 2.338513135910034, "learning_rate": 6.43547396998638e-05, "loss": 0.0649, "step": 87390 }, { "epoch": 2.9547990128131447, "grad_norm": 1.1900670528411865, "learning_rate": 6.434682019198971e-05, "loss": 0.0605, "step": 87400 }, { "epoch": 2.955137090503398, "grad_norm": 1.4396075010299683, "learning_rate": 6.433890029187896e-05, "loss": 0.0631, "step": 87410 }, { "epoch": 2.955475168193651, "grad_norm": 5.762599468231201, "learning_rate": 6.433097999974805e-05, "loss": 0.0575, "step": 87420 }, { "epoch": 2.9558132458839044, "grad_norm": 1.111777663230896, "learning_rate": 6.432305931581355e-05, "loss": 0.0577, "step": 87430 }, { "epoch": 2.9561513235741574, "grad_norm": 0.8839204907417297, "learning_rate": 6.431513824029197e-05, "loss": 0.0597, "step": 87440 }, { "epoch": 2.9564894012644105, "grad_norm": 1.0724443197250366, "learning_rate": 6.430721677339991e-05, "loss": 0.0674, "step": 87450 }, { "epoch": 2.956827478954664, "grad_norm": 1.6368156671524048, "learning_rate": 6.429929491535389e-05, "loss": 0.0623, "step": 87460 }, { "epoch": 2.957165556644917, "grad_norm": 1.105153203010559, "learning_rate": 6.429137266637053e-05, "loss": 0.0624, "step": 87470 }, { "epoch": 2.95750363433517, "grad_norm": 1.4689586162567139, "learning_rate": 6.428345002666641e-05, "loss": 0.0672, "step": 87480 }, { "epoch": 2.9578417120254237, "grad_norm": 0.6863720417022705, "learning_rate": 6.427552699645814e-05, "loss": 0.067, "step": 87490 }, { "epoch": 2.9581797897156767, "grad_norm": 0.7404821515083313, "learning_rate": 6.426760357596234e-05, "loss": 0.0662, "step": 87500 }, { "epoch": 2.95851786740593, "grad_norm": 1.5002449750900269, "learning_rate": 6.425967976539558e-05, "loss": 0.0626, "step": 87510 }, { "epoch": 2.9588559450961833, "grad_norm": 2.05070424079895, "learning_rate": 6.425175556497456e-05, "loss": 0.0675, "step": 87520 }, { "epoch": 2.9591940227864364, "grad_norm": 1.4952788352966309, "learning_rate": 6.424383097491589e-05, "loss": 0.0625, "step": 87530 }, { "epoch": 2.9595321004766895, "grad_norm": 3.206022024154663, "learning_rate": 6.423590599543624e-05, "loss": 0.0633, "step": 87540 }, { "epoch": 2.959870178166943, "grad_norm": 1.817947268486023, "learning_rate": 6.422798062675224e-05, "loss": 0.0676, "step": 87550 }, { "epoch": 2.960208255857196, "grad_norm": 0.8641737103462219, "learning_rate": 6.422005486908062e-05, "loss": 0.0626, "step": 87560 }, { "epoch": 2.960546333547449, "grad_norm": 0.6591285467147827, "learning_rate": 6.421212872263805e-05, "loss": 0.0675, "step": 87570 }, { "epoch": 2.9608844112377026, "grad_norm": 2.3929660320281982, "learning_rate": 6.42042021876412e-05, "loss": 0.0624, "step": 87580 }, { "epoch": 2.9612224889279557, "grad_norm": 7.108625888824463, "learning_rate": 6.41962752643068e-05, "loss": 0.055, "step": 87590 }, { "epoch": 2.9615605666182088, "grad_norm": 0.7927741408348083, "learning_rate": 6.418834795285158e-05, "loss": 0.0616, "step": 87600 }, { "epoch": 2.9618986443084623, "grad_norm": 1.7907575368881226, "learning_rate": 6.418042025349224e-05, "loss": 0.0658, "step": 87610 }, { "epoch": 2.9622367219987153, "grad_norm": 1.6853692531585693, "learning_rate": 6.417249216644555e-05, "loss": 0.0664, "step": 87620 }, { "epoch": 2.9625747996889684, "grad_norm": 2.7922468185424805, "learning_rate": 6.416456369192825e-05, "loss": 0.057, "step": 87630 }, { "epoch": 2.962912877379222, "grad_norm": 1.8010717630386353, "learning_rate": 6.41566348301571e-05, "loss": 0.0625, "step": 87640 }, { "epoch": 2.963250955069475, "grad_norm": 3.0820465087890625, "learning_rate": 6.414870558134886e-05, "loss": 0.0589, "step": 87650 }, { "epoch": 2.963589032759728, "grad_norm": 1.1488854885101318, "learning_rate": 6.414077594572033e-05, "loss": 0.0658, "step": 87660 }, { "epoch": 2.9639271104499816, "grad_norm": 3.2298905849456787, "learning_rate": 6.413284592348831e-05, "loss": 0.0629, "step": 87670 }, { "epoch": 2.9642651881402347, "grad_norm": 1.3292783498764038, "learning_rate": 6.412491551486959e-05, "loss": 0.0622, "step": 87680 }, { "epoch": 2.9646032658304877, "grad_norm": 8.467966079711914, "learning_rate": 6.411698472008097e-05, "loss": 0.0605, "step": 87690 }, { "epoch": 2.9649413435207412, "grad_norm": 0.6624161005020142, "learning_rate": 6.410905353933931e-05, "loss": 0.0584, "step": 87700 }, { "epoch": 2.9652794212109943, "grad_norm": 1.3858070373535156, "learning_rate": 6.41011219728614e-05, "loss": 0.0624, "step": 87710 }, { "epoch": 2.9656174989012474, "grad_norm": 0.9599841833114624, "learning_rate": 6.409319002086414e-05, "loss": 0.059, "step": 87720 }, { "epoch": 2.965955576591501, "grad_norm": 1.5098611116409302, "learning_rate": 6.408525768356435e-05, "loss": 0.0659, "step": 87730 }, { "epoch": 2.966293654281754, "grad_norm": 3.876798391342163, "learning_rate": 6.407732496117891e-05, "loss": 0.0656, "step": 87740 }, { "epoch": 2.966631731972007, "grad_norm": 6.684793472290039, "learning_rate": 6.406939185392469e-05, "loss": 0.0638, "step": 87750 }, { "epoch": 2.9669698096622605, "grad_norm": 15.943984985351562, "learning_rate": 6.406145836201858e-05, "loss": 0.061, "step": 87760 }, { "epoch": 2.9673078873525136, "grad_norm": 1.038718819618225, "learning_rate": 6.405352448567748e-05, "loss": 0.0605, "step": 87770 }, { "epoch": 2.9676459650427667, "grad_norm": 1.2055538892745972, "learning_rate": 6.40455902251183e-05, "loss": 0.0605, "step": 87780 }, { "epoch": 2.96798404273302, "grad_norm": 1.4965596199035645, "learning_rate": 6.403765558055796e-05, "loss": 0.0564, "step": 87790 }, { "epoch": 2.9683221204232733, "grad_norm": 1.6449559926986694, "learning_rate": 6.40297205522134e-05, "loss": 0.0636, "step": 87800 }, { "epoch": 2.9686601981135263, "grad_norm": 1.7805227041244507, "learning_rate": 6.402178514030154e-05, "loss": 0.0641, "step": 87810 }, { "epoch": 2.96899827580378, "grad_norm": 1.4944205284118652, "learning_rate": 6.401384934503935e-05, "loss": 0.0585, "step": 87820 }, { "epoch": 2.969336353494033, "grad_norm": 3.2361032962799072, "learning_rate": 6.400591316664379e-05, "loss": 0.0576, "step": 87830 }, { "epoch": 2.969674431184286, "grad_norm": 1.8009928464889526, "learning_rate": 6.39979766053318e-05, "loss": 0.0625, "step": 87840 }, { "epoch": 2.9700125088745395, "grad_norm": 9.140613555908203, "learning_rate": 6.399003966132043e-05, "loss": 0.0584, "step": 87850 }, { "epoch": 2.9703505865647926, "grad_norm": 291.0454406738281, "learning_rate": 6.398210233482659e-05, "loss": 0.065, "step": 87860 }, { "epoch": 2.9706886642550456, "grad_norm": 1.4150588512420654, "learning_rate": 6.397416462606734e-05, "loss": 0.0585, "step": 87870 }, { "epoch": 2.971026741945299, "grad_norm": 2.5157458782196045, "learning_rate": 6.396622653525967e-05, "loss": 0.0583, "step": 87880 }, { "epoch": 2.971364819635552, "grad_norm": 1.9672075510025024, "learning_rate": 6.395828806262065e-05, "loss": 0.0652, "step": 87890 }, { "epoch": 2.9717028973258053, "grad_norm": 1.16034996509552, "learning_rate": 6.395034920836724e-05, "loss": 0.0687, "step": 87900 }, { "epoch": 2.972040975016059, "grad_norm": 1.5532255172729492, "learning_rate": 6.394240997271655e-05, "loss": 0.0566, "step": 87910 }, { "epoch": 2.972379052706312, "grad_norm": 0.8351938128471375, "learning_rate": 6.393447035588558e-05, "loss": 0.0674, "step": 87920 }, { "epoch": 2.972717130396565, "grad_norm": 3.5048024654388428, "learning_rate": 6.392653035809144e-05, "loss": 0.0662, "step": 87930 }, { "epoch": 2.9730552080868184, "grad_norm": 1.880731225013733, "learning_rate": 6.39185899795512e-05, "loss": 0.06, "step": 87940 }, { "epoch": 2.9733932857770715, "grad_norm": 1.1833854913711548, "learning_rate": 6.391064922048193e-05, "loss": 0.0689, "step": 87950 }, { "epoch": 2.9737313634673246, "grad_norm": 1.327363133430481, "learning_rate": 6.390270808110074e-05, "loss": 0.0626, "step": 87960 }, { "epoch": 2.974069441157578, "grad_norm": 1.267584204673767, "learning_rate": 6.389476656162472e-05, "loss": 0.0605, "step": 87970 }, { "epoch": 2.974407518847831, "grad_norm": 1.3732190132141113, "learning_rate": 6.388682466227101e-05, "loss": 0.0696, "step": 87980 }, { "epoch": 2.9747455965380842, "grad_norm": 1.1971559524536133, "learning_rate": 6.387888238325674e-05, "loss": 0.064, "step": 87990 }, { "epoch": 2.9750836742283377, "grad_norm": 3.2788846492767334, "learning_rate": 6.387093972479903e-05, "loss": 0.065, "step": 88000 }, { "epoch": 2.975421751918591, "grad_norm": 2.1616129875183105, "learning_rate": 6.386299668711504e-05, "loss": 0.0588, "step": 88010 }, { "epoch": 2.975759829608844, "grad_norm": 1.9133747816085815, "learning_rate": 6.385505327042193e-05, "loss": 0.067, "step": 88020 }, { "epoch": 2.9760979072990974, "grad_norm": 7.60457706451416, "learning_rate": 6.384710947493689e-05, "loss": 0.0627, "step": 88030 }, { "epoch": 2.9764359849893505, "grad_norm": 1.6138534545898438, "learning_rate": 6.383916530087704e-05, "loss": 0.0591, "step": 88040 }, { "epoch": 2.9767740626796035, "grad_norm": 1.5399742126464844, "learning_rate": 6.383122074845962e-05, "loss": 0.0606, "step": 88050 }, { "epoch": 2.977112140369857, "grad_norm": 1.6698685884475708, "learning_rate": 6.382327581790185e-05, "loss": 0.0674, "step": 88060 }, { "epoch": 2.97745021806011, "grad_norm": 1.7602202892303467, "learning_rate": 6.38153305094209e-05, "loss": 0.0671, "step": 88070 }, { "epoch": 2.977788295750363, "grad_norm": 3.3659253120422363, "learning_rate": 6.380738482323398e-05, "loss": 0.0642, "step": 88080 }, { "epoch": 2.9781263734406167, "grad_norm": 1.9337074756622314, "learning_rate": 6.379943875955837e-05, "loss": 0.0674, "step": 88090 }, { "epoch": 2.9784644511308698, "grad_norm": 1.9563374519348145, "learning_rate": 6.379149231861129e-05, "loss": 0.0692, "step": 88100 }, { "epoch": 2.978802528821123, "grad_norm": 1.8595764636993408, "learning_rate": 6.378354550060997e-05, "loss": 0.072, "step": 88110 }, { "epoch": 2.9791406065113764, "grad_norm": 2.0136427879333496, "learning_rate": 6.377559830577171e-05, "loss": 0.0672, "step": 88120 }, { "epoch": 2.9794786842016294, "grad_norm": 1.8812185525894165, "learning_rate": 6.376765073431377e-05, "loss": 0.0711, "step": 88130 }, { "epoch": 2.9798167618918825, "grad_norm": 1.55075204372406, "learning_rate": 6.375970278645342e-05, "loss": 0.0682, "step": 88140 }, { "epoch": 2.980154839582136, "grad_norm": 1.2865760326385498, "learning_rate": 6.375175446240797e-05, "loss": 0.0584, "step": 88150 }, { "epoch": 2.980492917272389, "grad_norm": 1.4237077236175537, "learning_rate": 6.374380576239473e-05, "loss": 0.0637, "step": 88160 }, { "epoch": 2.980830994962642, "grad_norm": 1.336493968963623, "learning_rate": 6.373585668663098e-05, "loss": 0.0629, "step": 88170 }, { "epoch": 2.9811690726528957, "grad_norm": 7.288848400115967, "learning_rate": 6.372790723533408e-05, "loss": 0.069, "step": 88180 }, { "epoch": 2.9815071503431487, "grad_norm": 1.2334541082382202, "learning_rate": 6.371995740872136e-05, "loss": 0.0644, "step": 88190 }, { "epoch": 2.9818452280334022, "grad_norm": 5.010420799255371, "learning_rate": 6.371200720701016e-05, "loss": 0.0634, "step": 88200 }, { "epoch": 2.9821833057236553, "grad_norm": 17.92904281616211, "learning_rate": 6.370405663041783e-05, "loss": 0.0661, "step": 88210 }, { "epoch": 2.9825213834139084, "grad_norm": 9.568615913391113, "learning_rate": 6.369610567916173e-05, "loss": 0.0666, "step": 88220 }, { "epoch": 2.982859461104162, "grad_norm": 1.904308557510376, "learning_rate": 6.368815435345924e-05, "loss": 0.0697, "step": 88230 }, { "epoch": 2.983197538794415, "grad_norm": 1.9413845539093018, "learning_rate": 6.368020265352779e-05, "loss": 0.066, "step": 88240 }, { "epoch": 2.983535616484668, "grad_norm": 2.6301896572113037, "learning_rate": 6.36722505795847e-05, "loss": 0.0726, "step": 88250 }, { "epoch": 2.9838736941749215, "grad_norm": 5.841256141662598, "learning_rate": 6.366429813184744e-05, "loss": 0.0669, "step": 88260 }, { "epoch": 2.9842117718651746, "grad_norm": 2.387972116470337, "learning_rate": 6.36563453105334e-05, "loss": 0.0653, "step": 88270 }, { "epoch": 2.9845498495554277, "grad_norm": 8.355180740356445, "learning_rate": 6.364839211586e-05, "loss": 0.0619, "step": 88280 }, { "epoch": 2.984887927245681, "grad_norm": 2.3002498149871826, "learning_rate": 6.36404385480447e-05, "loss": 0.0619, "step": 88290 }, { "epoch": 2.9852260049359343, "grad_norm": 2.1792683601379395, "learning_rate": 6.36324846073049e-05, "loss": 0.0629, "step": 88300 }, { "epoch": 2.9855640826261873, "grad_norm": 3.3201944828033447, "learning_rate": 6.362453029385813e-05, "loss": 0.0588, "step": 88310 }, { "epoch": 2.985902160316441, "grad_norm": 2.6194405555725098, "learning_rate": 6.36165756079218e-05, "loss": 0.0608, "step": 88320 }, { "epoch": 2.986240238006694, "grad_norm": 2.0119590759277344, "learning_rate": 6.360862054971344e-05, "loss": 0.0647, "step": 88330 }, { "epoch": 2.9865783156969474, "grad_norm": 4.051635265350342, "learning_rate": 6.360066511945047e-05, "loss": 0.0639, "step": 88340 }, { "epoch": 2.9869163933872005, "grad_norm": 3.3445076942443848, "learning_rate": 6.359270931735044e-05, "loss": 0.0665, "step": 88350 }, { "epoch": 2.9872544710774536, "grad_norm": 4.965936183929443, "learning_rate": 6.358475314363084e-05, "loss": 0.0651, "step": 88360 }, { "epoch": 2.987592548767707, "grad_norm": 3.075456142425537, "learning_rate": 6.357679659850919e-05, "loss": 0.0682, "step": 88370 }, { "epoch": 2.98793062645796, "grad_norm": 36.25916290283203, "learning_rate": 6.356883968220303e-05, "loss": 0.065, "step": 88380 }, { "epoch": 2.988268704148213, "grad_norm": 29.817161560058594, "learning_rate": 6.35608823949299e-05, "loss": 0.0669, "step": 88390 }, { "epoch": 2.9886067818384667, "grad_norm": 5.639437675476074, "learning_rate": 6.355292473690732e-05, "loss": 0.0746, "step": 88400 }, { "epoch": 2.98894485952872, "grad_norm": 8.32291316986084, "learning_rate": 6.354496670835288e-05, "loss": 0.0776, "step": 88410 }, { "epoch": 2.989282937218973, "grad_norm": 4.6644182205200195, "learning_rate": 6.353700830948412e-05, "loss": 0.0722, "step": 88420 }, { "epoch": 2.9896210149092264, "grad_norm": 7.924507141113281, "learning_rate": 6.352904954051865e-05, "loss": 0.0765, "step": 88430 }, { "epoch": 2.9899590925994795, "grad_norm": 4.398316383361816, "learning_rate": 6.352109040167407e-05, "loss": 0.0743, "step": 88440 }, { "epoch": 2.9902971702897325, "grad_norm": 3.0222129821777344, "learning_rate": 6.351313089316794e-05, "loss": 0.0719, "step": 88450 }, { "epoch": 2.990635247979986, "grad_norm": 3.1128463745117188, "learning_rate": 6.350517101521788e-05, "loss": 0.0696, "step": 88460 }, { "epoch": 2.990973325670239, "grad_norm": 2.780407667160034, "learning_rate": 6.349721076804152e-05, "loss": 0.0703, "step": 88470 }, { "epoch": 2.991311403360492, "grad_norm": 3.632399320602417, "learning_rate": 6.34892501518565e-05, "loss": 0.0727, "step": 88480 }, { "epoch": 2.9916494810507457, "grad_norm": 11.355003356933594, "learning_rate": 6.348128916688045e-05, "loss": 0.0737, "step": 88490 }, { "epoch": 2.9919875587409988, "grad_norm": 8.47912883758545, "learning_rate": 6.3473327813331e-05, "loss": 0.0638, "step": 88500 }, { "epoch": 2.992325636431252, "grad_norm": 3.402588367462158, "learning_rate": 6.346536609142584e-05, "loss": 0.0662, "step": 88510 }, { "epoch": 2.9926637141215053, "grad_norm": 6.1389617919921875, "learning_rate": 6.345740400138264e-05, "loss": 0.0655, "step": 88520 }, { "epoch": 2.9930017918117584, "grad_norm": 3.714315414428711, "learning_rate": 6.344944154341905e-05, "loss": 0.0661, "step": 88530 }, { "epoch": 2.9933398695020115, "grad_norm": 7.766320705413818, "learning_rate": 6.34414787177528e-05, "loss": 0.0647, "step": 88540 }, { "epoch": 2.993677947192265, "grad_norm": 9.186469078063965, "learning_rate": 6.343351552460155e-05, "loss": 0.064, "step": 88550 }, { "epoch": 2.994016024882518, "grad_norm": 2.4414267539978027, "learning_rate": 6.342555196418306e-05, "loss": 0.0649, "step": 88560 }, { "epoch": 2.994354102572771, "grad_norm": 19.505699157714844, "learning_rate": 6.3417588036715e-05, "loss": 0.0631, "step": 88570 }, { "epoch": 2.9946921802630246, "grad_norm": 1.6717098951339722, "learning_rate": 6.340962374241514e-05, "loss": 0.072, "step": 88580 }, { "epoch": 2.9950302579532777, "grad_norm": 4.444525241851807, "learning_rate": 6.340165908150119e-05, "loss": 0.062, "step": 88590 }, { "epoch": 2.9953683356435308, "grad_norm": 1.239223599433899, "learning_rate": 6.339369405419093e-05, "loss": 0.069, "step": 88600 }, { "epoch": 2.9957064133337843, "grad_norm": 1.7681583166122437, "learning_rate": 6.338572866070209e-05, "loss": 0.0687, "step": 88610 }, { "epoch": 2.9960444910240374, "grad_norm": 1.8486260175704956, "learning_rate": 6.337776290125247e-05, "loss": 0.0727, "step": 88620 }, { "epoch": 2.9963825687142904, "grad_norm": 3.002162218093872, "learning_rate": 6.336979677605983e-05, "loss": 0.0656, "step": 88630 }, { "epoch": 2.996720646404544, "grad_norm": 12.192002296447754, "learning_rate": 6.336183028534198e-05, "loss": 0.0719, "step": 88640 }, { "epoch": 2.997058724094797, "grad_norm": 4.119073390960693, "learning_rate": 6.33538634293167e-05, "loss": 0.0695, "step": 88650 }, { "epoch": 2.99739680178505, "grad_norm": 2.41217303276062, "learning_rate": 6.334589620820182e-05, "loss": 0.0642, "step": 88660 }, { "epoch": 2.9977348794753036, "grad_norm": 2.0731489658355713, "learning_rate": 6.333792862221514e-05, "loss": 0.0701, "step": 88670 }, { "epoch": 2.9980729571655567, "grad_norm": 3.425382137298584, "learning_rate": 6.33299606715745e-05, "loss": 0.0681, "step": 88680 }, { "epoch": 2.9984110348558097, "grad_norm": 3.7888758182525635, "learning_rate": 6.332199235649776e-05, "loss": 0.0669, "step": 88690 }, { "epoch": 2.9987491125460632, "grad_norm": 1.2278037071228027, "learning_rate": 6.331402367720275e-05, "loss": 0.0699, "step": 88700 }, { "epoch": 2.9990871902363163, "grad_norm": 2.7767481803894043, "learning_rate": 6.330605463390732e-05, "loss": 0.0706, "step": 88710 }, { "epoch": 2.9994252679265694, "grad_norm": 2.322593927383423, "learning_rate": 6.329808522682937e-05, "loss": 0.0705, "step": 88720 }, { "epoch": 2.999763345616823, "grad_norm": 3.464599847793579, "learning_rate": 6.329011545618676e-05, "loss": 0.0654, "step": 88730 }, { "epoch": 3.000101423307076, "grad_norm": 4.991198539733887, "learning_rate": 6.328214532219739e-05, "loss": 0.0609, "step": 88740 }, { "epoch": 3.000439500997329, "grad_norm": 1.7113763093948364, "learning_rate": 6.327417482507914e-05, "loss": 0.0636, "step": 88750 }, { "epoch": 3.0007775786875825, "grad_norm": 4.971979141235352, "learning_rate": 6.326620396504994e-05, "loss": 0.0687, "step": 88760 }, { "epoch": 3.0011156563778356, "grad_norm": 1.5828213691711426, "learning_rate": 6.325823274232773e-05, "loss": 0.0647, "step": 88770 }, { "epoch": 3.0014537340680887, "grad_norm": 3.150599718093872, "learning_rate": 6.325026115713039e-05, "loss": 0.0754, "step": 88780 }, { "epoch": 3.001791811758342, "grad_norm": 5.455735206604004, "learning_rate": 6.324228920967592e-05, "loss": 0.0705, "step": 88790 }, { "epoch": 3.0021298894485953, "grad_norm": 3.112517833709717, "learning_rate": 6.323431690018223e-05, "loss": 0.0647, "step": 88800 }, { "epoch": 3.0024679671388483, "grad_norm": 2.175280809402466, "learning_rate": 6.322634422886726e-05, "loss": 0.068, "step": 88810 }, { "epoch": 3.002806044829102, "grad_norm": 6.6352458000183105, "learning_rate": 6.321837119594902e-05, "loss": 0.0671, "step": 88820 }, { "epoch": 3.003144122519355, "grad_norm": 4.433596134185791, "learning_rate": 6.32103978016455e-05, "loss": 0.0629, "step": 88830 }, { "epoch": 3.003482200209608, "grad_norm": 3.8824195861816406, "learning_rate": 6.320242404617465e-05, "loss": 0.0663, "step": 88840 }, { "epoch": 3.0038202778998615, "grad_norm": 2.542445182800293, "learning_rate": 6.31944499297545e-05, "loss": 0.0627, "step": 88850 }, { "epoch": 3.0041583555901146, "grad_norm": 12.051742553710938, "learning_rate": 6.318647545260302e-05, "loss": 0.0716, "step": 88860 }, { "epoch": 3.0044964332803676, "grad_norm": 9.661172866821289, "learning_rate": 6.317850061493827e-05, "loss": 0.0663, "step": 88870 }, { "epoch": 3.004834510970621, "grad_norm": 2.294205665588379, "learning_rate": 6.317052541697826e-05, "loss": 0.0687, "step": 88880 }, { "epoch": 3.0051725886608742, "grad_norm": 5.248610973358154, "learning_rate": 6.316254985894103e-05, "loss": 0.0681, "step": 88890 }, { "epoch": 3.0055106663511273, "grad_norm": 5.297034740447998, "learning_rate": 6.315457394104463e-05, "loss": 0.0628, "step": 88900 }, { "epoch": 3.005848744041381, "grad_norm": 6.763940334320068, "learning_rate": 6.314659766350714e-05, "loss": 0.067, "step": 88910 }, { "epoch": 3.006186821731634, "grad_norm": 3.2229456901550293, "learning_rate": 6.31386210265466e-05, "loss": 0.0698, "step": 88920 }, { "epoch": 3.006524899421887, "grad_norm": 1.6825886964797974, "learning_rate": 6.31306440303811e-05, "loss": 0.0735, "step": 88930 }, { "epoch": 3.0068629771121405, "grad_norm": 3.596261978149414, "learning_rate": 6.31226666752287e-05, "loss": 0.0643, "step": 88940 }, { "epoch": 3.0072010548023935, "grad_norm": 4.092008590698242, "learning_rate": 6.311468896130755e-05, "loss": 0.0654, "step": 88950 }, { "epoch": 3.0075391324926466, "grad_norm": 2.993201494216919, "learning_rate": 6.310671088883571e-05, "loss": 0.064, "step": 88960 }, { "epoch": 3.0078772101829, "grad_norm": 2.409148931503296, "learning_rate": 6.309873245803135e-05, "loss": 0.0647, "step": 88970 }, { "epoch": 3.008215287873153, "grad_norm": 1.8777291774749756, "learning_rate": 6.309075366911255e-05, "loss": 0.0649, "step": 88980 }, { "epoch": 3.0085533655634067, "grad_norm": 6.244673252105713, "learning_rate": 6.308277452229744e-05, "loss": 0.0564, "step": 88990 }, { "epoch": 3.0088914432536598, "grad_norm": 4.157722473144531, "learning_rate": 6.30747950178042e-05, "loss": 0.0676, "step": 89000 }, { "epoch": 3.009229520943913, "grad_norm": 9.798083305358887, "learning_rate": 6.3066815155851e-05, "loss": 0.0754, "step": 89010 }, { "epoch": 3.0095675986341663, "grad_norm": 1.2644832134246826, "learning_rate": 6.305883493665596e-05, "loss": 0.0673, "step": 89020 }, { "epoch": 3.0099056763244194, "grad_norm": 5.170543193817139, "learning_rate": 6.30508543604373e-05, "loss": 0.0724, "step": 89030 }, { "epoch": 3.0102437540146725, "grad_norm": 3.069108486175537, "learning_rate": 6.304287342741317e-05, "loss": 0.0722, "step": 89040 }, { "epoch": 3.010581831704926, "grad_norm": 2.2823374271392822, "learning_rate": 6.303489213780178e-05, "loss": 0.0666, "step": 89050 }, { "epoch": 3.010919909395179, "grad_norm": 9.38670825958252, "learning_rate": 6.302691049182133e-05, "loss": 0.0739, "step": 89060 }, { "epoch": 3.011257987085432, "grad_norm": 2.5917062759399414, "learning_rate": 6.301892848969004e-05, "loss": 0.0627, "step": 89070 }, { "epoch": 3.0115960647756856, "grad_norm": 2.075301170349121, "learning_rate": 6.301094613162616e-05, "loss": 0.0677, "step": 89080 }, { "epoch": 3.0119341424659387, "grad_norm": 1.662882685661316, "learning_rate": 6.300296341784788e-05, "loss": 0.0734, "step": 89090 }, { "epoch": 3.012272220156192, "grad_norm": 11.655284881591797, "learning_rate": 6.299498034857348e-05, "loss": 0.0654, "step": 89100 }, { "epoch": 3.0126102978464453, "grad_norm": 2.0130677223205566, "learning_rate": 6.298699692402118e-05, "loss": 0.0736, "step": 89110 }, { "epoch": 3.0129483755366984, "grad_norm": 1.323124885559082, "learning_rate": 6.297901314440929e-05, "loss": 0.0687, "step": 89120 }, { "epoch": 3.0132864532269514, "grad_norm": 2.415867328643799, "learning_rate": 6.297102900995602e-05, "loss": 0.0697, "step": 89130 }, { "epoch": 3.013624530917205, "grad_norm": 2.3509414196014404, "learning_rate": 6.296304452087971e-05, "loss": 0.0719, "step": 89140 }, { "epoch": 3.013962608607458, "grad_norm": 4.2878098487854, "learning_rate": 6.295505967739863e-05, "loss": 0.0588, "step": 89150 }, { "epoch": 3.014300686297711, "grad_norm": 4.405576229095459, "learning_rate": 6.294707447973109e-05, "loss": 0.0672, "step": 89160 }, { "epoch": 3.0146387639879646, "grad_norm": 1.1096245050430298, "learning_rate": 6.29390889280954e-05, "loss": 0.0689, "step": 89170 }, { "epoch": 3.0149768416782177, "grad_norm": 1.6156481504440308, "learning_rate": 6.293110302270987e-05, "loss": 0.0684, "step": 89180 }, { "epoch": 3.0153149193684707, "grad_norm": 5.139451026916504, "learning_rate": 6.292311676379284e-05, "loss": 0.0714, "step": 89190 }, { "epoch": 3.0156529970587242, "grad_norm": 1.749884843826294, "learning_rate": 6.291513015156267e-05, "loss": 0.0672, "step": 89200 }, { "epoch": 3.0159910747489773, "grad_norm": 3.216710329055786, "learning_rate": 6.290714318623767e-05, "loss": 0.062, "step": 89210 }, { "epoch": 3.0163291524392304, "grad_norm": 8.855215072631836, "learning_rate": 6.289915586803625e-05, "loss": 0.0639, "step": 89220 }, { "epoch": 3.016667230129484, "grad_norm": 3.23626708984375, "learning_rate": 6.289116819717675e-05, "loss": 0.065, "step": 89230 }, { "epoch": 3.017005307819737, "grad_norm": 2.5335779190063477, "learning_rate": 6.288318017387753e-05, "loss": 0.0676, "step": 89240 }, { "epoch": 3.01734338550999, "grad_norm": 4.5034356117248535, "learning_rate": 6.287519179835702e-05, "loss": 0.0652, "step": 89250 }, { "epoch": 3.0176814632002436, "grad_norm": 9.893404960632324, "learning_rate": 6.28672030708336e-05, "loss": 0.0701, "step": 89260 }, { "epoch": 3.0180195408904966, "grad_norm": 2.3050425052642822, "learning_rate": 6.285921399152569e-05, "loss": 0.0696, "step": 89270 }, { "epoch": 3.0183576185807497, "grad_norm": 5.563252925872803, "learning_rate": 6.285122456065169e-05, "loss": 0.0698, "step": 89280 }, { "epoch": 3.018695696271003, "grad_norm": 1.6764119863510132, "learning_rate": 6.284323477843005e-05, "loss": 0.0642, "step": 89290 }, { "epoch": 3.0190337739612563, "grad_norm": 1.3632583618164062, "learning_rate": 6.28352446450792e-05, "loss": 0.0705, "step": 89300 }, { "epoch": 3.0193718516515093, "grad_norm": 1.4408493041992188, "learning_rate": 6.282725416081757e-05, "loss": 0.0671, "step": 89310 }, { "epoch": 3.019709929341763, "grad_norm": 7.691853046417236, "learning_rate": 6.281926332586362e-05, "loss": 0.0678, "step": 89320 }, { "epoch": 3.020048007032016, "grad_norm": 1.2218331098556519, "learning_rate": 6.281127214043586e-05, "loss": 0.0641, "step": 89330 }, { "epoch": 3.020386084722269, "grad_norm": 1.5817033052444458, "learning_rate": 6.280328060475269e-05, "loss": 0.0625, "step": 89340 }, { "epoch": 3.0207241624125225, "grad_norm": 1.6800349950790405, "learning_rate": 6.279528871903267e-05, "loss": 0.0703, "step": 89350 }, { "epoch": 3.0210622401027756, "grad_norm": 9.443280220031738, "learning_rate": 6.278729648349426e-05, "loss": 0.0632, "step": 89360 }, { "epoch": 3.0214003177930286, "grad_norm": 4.047760009765625, "learning_rate": 6.277930389835597e-05, "loss": 0.0701, "step": 89370 }, { "epoch": 3.021738395483282, "grad_norm": 1.4053634405136108, "learning_rate": 6.277131096383628e-05, "loss": 0.0664, "step": 89380 }, { "epoch": 3.0220764731735352, "grad_norm": 1.3339282274246216, "learning_rate": 6.276331768015378e-05, "loss": 0.0772, "step": 89390 }, { "epoch": 3.0224145508637883, "grad_norm": 2.7550504207611084, "learning_rate": 6.275532404752696e-05, "loss": 0.0636, "step": 89400 }, { "epoch": 3.022752628554042, "grad_norm": 2.2657408714294434, "learning_rate": 6.274733006617438e-05, "loss": 0.061, "step": 89410 }, { "epoch": 3.023090706244295, "grad_norm": 3.9905788898468018, "learning_rate": 6.273933573631457e-05, "loss": 0.0652, "step": 89420 }, { "epoch": 3.023428783934548, "grad_norm": 5.215058326721191, "learning_rate": 6.273134105816612e-05, "loss": 0.0676, "step": 89430 }, { "epoch": 3.0237668616248015, "grad_norm": 2.7689428329467773, "learning_rate": 6.272334603194758e-05, "loss": 0.0794, "step": 89440 }, { "epoch": 3.0241049393150545, "grad_norm": 7.005691051483154, "learning_rate": 6.271535065787754e-05, "loss": 0.0693, "step": 89450 }, { "epoch": 3.024443017005308, "grad_norm": 2.8810997009277344, "learning_rate": 6.27073549361746e-05, "loss": 0.0685, "step": 89460 }, { "epoch": 3.024781094695561, "grad_norm": 2.1948318481445312, "learning_rate": 6.269935886705733e-05, "loss": 0.0653, "step": 89470 }, { "epoch": 3.025119172385814, "grad_norm": 1.8235995769500732, "learning_rate": 6.269136245074439e-05, "loss": 0.0623, "step": 89480 }, { "epoch": 3.0254572500760677, "grad_norm": 2.9674761295318604, "learning_rate": 6.268336568745434e-05, "loss": 0.0692, "step": 89490 }, { "epoch": 3.0257953277663208, "grad_norm": 2.544743776321411, "learning_rate": 6.267536857740585e-05, "loss": 0.0674, "step": 89500 }, { "epoch": 3.026133405456574, "grad_norm": 2.848917245864868, "learning_rate": 6.266737112081754e-05, "loss": 0.0596, "step": 89510 }, { "epoch": 3.0264714831468273, "grad_norm": 1.4739614725112915, "learning_rate": 6.265937331790808e-05, "loss": 0.061, "step": 89520 }, { "epoch": 3.0268095608370804, "grad_norm": 1.6697790622711182, "learning_rate": 6.265137516889609e-05, "loss": 0.0664, "step": 89530 }, { "epoch": 3.0271476385273335, "grad_norm": 1.7851272821426392, "learning_rate": 6.264337667400026e-05, "loss": 0.0648, "step": 89540 }, { "epoch": 3.027485716217587, "grad_norm": 3.525214433670044, "learning_rate": 6.263537783343927e-05, "loss": 0.0665, "step": 89550 }, { "epoch": 3.02782379390784, "grad_norm": 2.1576035022735596, "learning_rate": 6.262737864743179e-05, "loss": 0.0627, "step": 89560 }, { "epoch": 3.028161871598093, "grad_norm": 5.474564552307129, "learning_rate": 6.26193791161965e-05, "loss": 0.0758, "step": 89570 }, { "epoch": 3.0284999492883466, "grad_norm": 1.5978902578353882, "learning_rate": 6.261137923995216e-05, "loss": 0.0669, "step": 89580 }, { "epoch": 3.0288380269785997, "grad_norm": 1.164754867553711, "learning_rate": 6.260337901891743e-05, "loss": 0.0752, "step": 89590 }, { "epoch": 3.029176104668853, "grad_norm": 1.1324338912963867, "learning_rate": 6.259537845331107e-05, "loss": 0.0743, "step": 89600 }, { "epoch": 3.0295141823591063, "grad_norm": 1.9957975149154663, "learning_rate": 6.258737754335178e-05, "loss": 0.0684, "step": 89610 }, { "epoch": 3.0298522600493594, "grad_norm": 3.223153591156006, "learning_rate": 6.257937628925832e-05, "loss": 0.0748, "step": 89620 }, { "epoch": 3.0301903377396124, "grad_norm": 2.5309929847717285, "learning_rate": 6.257137469124944e-05, "loss": 0.0747, "step": 89630 }, { "epoch": 3.030528415429866, "grad_norm": 1.8131259679794312, "learning_rate": 6.25633727495439e-05, "loss": 0.0707, "step": 89640 }, { "epoch": 3.030866493120119, "grad_norm": 2.3067140579223633, "learning_rate": 6.255537046436047e-05, "loss": 0.0812, "step": 89650 }, { "epoch": 3.031204570810372, "grad_norm": 1.7394371032714844, "learning_rate": 6.254736783591793e-05, "loss": 0.0789, "step": 89660 }, { "epoch": 3.0315426485006256, "grad_norm": 2.095268964767456, "learning_rate": 6.253936486443504e-05, "loss": 0.0726, "step": 89670 }, { "epoch": 3.0318807261908787, "grad_norm": 3.811133623123169, "learning_rate": 6.253136155013066e-05, "loss": 0.0667, "step": 89680 }, { "epoch": 3.0322188038811317, "grad_norm": 1.1261180639266968, "learning_rate": 6.252335789322355e-05, "loss": 0.0756, "step": 89690 }, { "epoch": 3.0325568815713853, "grad_norm": 7.362478733062744, "learning_rate": 6.251535389393254e-05, "loss": 0.0749, "step": 89700 }, { "epoch": 3.0328949592616383, "grad_norm": 1.9310994148254395, "learning_rate": 6.250734955247644e-05, "loss": 0.0678, "step": 89710 }, { "epoch": 3.0332330369518914, "grad_norm": 1.5351420640945435, "learning_rate": 6.249934486907413e-05, "loss": 0.0715, "step": 89720 }, { "epoch": 3.033571114642145, "grad_norm": 2.3738882541656494, "learning_rate": 6.249133984394442e-05, "loss": 0.07, "step": 89730 }, { "epoch": 3.033909192332398, "grad_norm": 2.591438055038452, "learning_rate": 6.248333447730617e-05, "loss": 0.0685, "step": 89740 }, { "epoch": 3.034247270022651, "grad_norm": 2.6845035552978516, "learning_rate": 6.247532876937825e-05, "loss": 0.0715, "step": 89750 }, { "epoch": 3.0345853477129046, "grad_norm": 2.0384411811828613, "learning_rate": 6.24673227203795e-05, "loss": 0.0713, "step": 89760 }, { "epoch": 3.0349234254031576, "grad_norm": 3.0031330585479736, "learning_rate": 6.245931633052885e-05, "loss": 0.0685, "step": 89770 }, { "epoch": 3.0352615030934107, "grad_norm": 2.06917405128479, "learning_rate": 6.245130960004517e-05, "loss": 0.0764, "step": 89780 }, { "epoch": 3.035599580783664, "grad_norm": 2.011648416519165, "learning_rate": 6.244330252914738e-05, "loss": 0.0715, "step": 89790 }, { "epoch": 3.0359376584739173, "grad_norm": 5.676229476928711, "learning_rate": 6.243529511805435e-05, "loss": 0.0702, "step": 89800 }, { "epoch": 3.0362757361641703, "grad_norm": 1.977105975151062, "learning_rate": 6.242728736698504e-05, "loss": 0.0731, "step": 89810 }, { "epoch": 3.036613813854424, "grad_norm": 1.4202309846878052, "learning_rate": 6.241927927615836e-05, "loss": 0.0743, "step": 89820 }, { "epoch": 3.036951891544677, "grad_norm": 1.6664222478866577, "learning_rate": 6.241127084579323e-05, "loss": 0.0636, "step": 89830 }, { "epoch": 3.03728996923493, "grad_norm": 1.2874341011047363, "learning_rate": 6.240326207610864e-05, "loss": 0.0702, "step": 89840 }, { "epoch": 3.0376280469251835, "grad_norm": 2.86848783493042, "learning_rate": 6.239525296732352e-05, "loss": 0.0691, "step": 89850 }, { "epoch": 3.0379661246154366, "grad_norm": 1.7254929542541504, "learning_rate": 6.238724351965683e-05, "loss": 0.0663, "step": 89860 }, { "epoch": 3.0383042023056896, "grad_norm": 1.4797166585922241, "learning_rate": 6.237923373332758e-05, "loss": 0.0697, "step": 89870 }, { "epoch": 3.038642279995943, "grad_norm": 3.245588541030884, "learning_rate": 6.237122360855472e-05, "loss": 0.0778, "step": 89880 }, { "epoch": 3.0389803576861962, "grad_norm": 1.8144629001617432, "learning_rate": 6.236321314555725e-05, "loss": 0.0698, "step": 89890 }, { "epoch": 3.0393184353764493, "grad_norm": 1.054405689239502, "learning_rate": 6.235520234455417e-05, "loss": 0.072, "step": 89900 }, { "epoch": 3.039656513066703, "grad_norm": 2.929729700088501, "learning_rate": 6.234719120576451e-05, "loss": 0.0742, "step": 89910 }, { "epoch": 3.039994590756956, "grad_norm": 1.5282498598098755, "learning_rate": 6.233917972940727e-05, "loss": 0.0765, "step": 89920 }, { "epoch": 3.040332668447209, "grad_norm": 3.4519622325897217, "learning_rate": 6.23311679157015e-05, "loss": 0.0718, "step": 89930 }, { "epoch": 3.0406707461374625, "grad_norm": 3.3180668354034424, "learning_rate": 6.232315576486622e-05, "loss": 0.0623, "step": 89940 }, { "epoch": 3.0410088238277155, "grad_norm": 1.6896440982818604, "learning_rate": 6.231514327712052e-05, "loss": 0.0635, "step": 89950 }, { "epoch": 3.0413469015179686, "grad_norm": 1.885602593421936, "learning_rate": 6.230713045268339e-05, "loss": 0.0684, "step": 89960 }, { "epoch": 3.041684979208222, "grad_norm": 2.0632431507110596, "learning_rate": 6.229911729177396e-05, "loss": 0.0719, "step": 89970 }, { "epoch": 3.042023056898475, "grad_norm": 1.4282571077346802, "learning_rate": 6.22911037946113e-05, "loss": 0.0675, "step": 89980 }, { "epoch": 3.0423611345887287, "grad_norm": 1.327248215675354, "learning_rate": 6.228308996141446e-05, "loss": 0.0764, "step": 89990 }, { "epoch": 3.0426992122789818, "grad_norm": 3.630736827850342, "learning_rate": 6.227507579240254e-05, "loss": 0.0697, "step": 90000 }, { "epoch": 3.043037289969235, "grad_norm": 1.9584159851074219, "learning_rate": 6.226706128779468e-05, "loss": 0.0685, "step": 90010 }, { "epoch": 3.0433753676594884, "grad_norm": 1.414780616760254, "learning_rate": 6.225904644780997e-05, "loss": 0.077, "step": 90020 }, { "epoch": 3.0437134453497414, "grad_norm": 7.947182655334473, "learning_rate": 6.225103127266753e-05, "loss": 0.0684, "step": 90030 }, { "epoch": 3.0440515230399945, "grad_norm": 1.3879529237747192, "learning_rate": 6.224301576258651e-05, "loss": 0.0737, "step": 90040 }, { "epoch": 3.044389600730248, "grad_norm": 1.7315391302108765, "learning_rate": 6.223499991778604e-05, "loss": 0.071, "step": 90050 }, { "epoch": 3.044727678420501, "grad_norm": 4.828361988067627, "learning_rate": 6.222698373848525e-05, "loss": 0.0772, "step": 90060 }, { "epoch": 3.045065756110754, "grad_norm": 1.648937463760376, "learning_rate": 6.221896722490333e-05, "loss": 0.0703, "step": 90070 }, { "epoch": 3.0454038338010077, "grad_norm": 3.158200740814209, "learning_rate": 6.221095037725942e-05, "loss": 0.0656, "step": 90080 }, { "epoch": 3.0457419114912607, "grad_norm": 1.8656270503997803, "learning_rate": 6.220293319577273e-05, "loss": 0.0622, "step": 90090 }, { "epoch": 3.046079989181514, "grad_norm": 1.7993193864822388, "learning_rate": 6.219491568066243e-05, "loss": 0.0679, "step": 90100 }, { "epoch": 3.0464180668717673, "grad_norm": 1.6334941387176514, "learning_rate": 6.218689783214772e-05, "loss": 0.0664, "step": 90110 }, { "epoch": 3.0467561445620204, "grad_norm": 1.838052749633789, "learning_rate": 6.21788796504478e-05, "loss": 0.0638, "step": 90120 }, { "epoch": 3.0470942222522734, "grad_norm": 1.828321099281311, "learning_rate": 6.217086113578187e-05, "loss": 0.0622, "step": 90130 }, { "epoch": 3.047432299942527, "grad_norm": 1.2434165477752686, "learning_rate": 6.21628422883692e-05, "loss": 0.0697, "step": 90140 }, { "epoch": 3.04777037763278, "grad_norm": 1.7307946681976318, "learning_rate": 6.215482310842896e-05, "loss": 0.0628, "step": 90150 }, { "epoch": 3.048108455323033, "grad_norm": 3.3551340103149414, "learning_rate": 6.214680359618044e-05, "loss": 0.068, "step": 90160 }, { "epoch": 3.0484465330132866, "grad_norm": 1.4022715091705322, "learning_rate": 6.213878375184287e-05, "loss": 0.0693, "step": 90170 }, { "epoch": 3.0487846107035397, "grad_norm": 2.3703761100769043, "learning_rate": 6.21307635756355e-05, "loss": 0.0654, "step": 90180 }, { "epoch": 3.0491226883937927, "grad_norm": 1.6064516305923462, "learning_rate": 6.212274306777764e-05, "loss": 0.0717, "step": 90190 }, { "epoch": 3.0494607660840463, "grad_norm": 1.379130244255066, "learning_rate": 6.21147222284885e-05, "loss": 0.0675, "step": 90200 }, { "epoch": 3.0497988437742993, "grad_norm": 6.689403057098389, "learning_rate": 6.210670105798744e-05, "loss": 0.0647, "step": 90210 }, { "epoch": 3.0501369214645524, "grad_norm": 1.616525650024414, "learning_rate": 6.20986795564937e-05, "loss": 0.0618, "step": 90220 }, { "epoch": 3.050474999154806, "grad_norm": 1.3647453784942627, "learning_rate": 6.209065772422661e-05, "loss": 0.0673, "step": 90230 }, { "epoch": 3.050813076845059, "grad_norm": 1.2375963926315308, "learning_rate": 6.208263556140549e-05, "loss": 0.0766, "step": 90240 }, { "epoch": 3.051151154535312, "grad_norm": 1.5514194965362549, "learning_rate": 6.207461306824965e-05, "loss": 0.0751, "step": 90250 }, { "epoch": 3.0514892322255656, "grad_norm": 0.9154356122016907, "learning_rate": 6.206659024497843e-05, "loss": 0.0662, "step": 90260 }, { "epoch": 3.0518273099158186, "grad_norm": 0.6026983857154846, "learning_rate": 6.205856709181115e-05, "loss": 0.062, "step": 90270 }, { "epoch": 3.0521653876060717, "grad_norm": 0.6796011328697205, "learning_rate": 6.205054360896718e-05, "loss": 0.0657, "step": 90280 }, { "epoch": 3.052503465296325, "grad_norm": 2.140939950942993, "learning_rate": 6.204251979666589e-05, "loss": 0.0663, "step": 90290 }, { "epoch": 3.0528415429865783, "grad_norm": 1.7180238962173462, "learning_rate": 6.203449565512661e-05, "loss": 0.065, "step": 90300 }, { "epoch": 3.0531796206768314, "grad_norm": 0.7280929684638977, "learning_rate": 6.202647118456877e-05, "loss": 0.0678, "step": 90310 }, { "epoch": 3.053517698367085, "grad_norm": 0.978592574596405, "learning_rate": 6.20184463852117e-05, "loss": 0.067, "step": 90320 }, { "epoch": 3.053855776057338, "grad_norm": 1.3194653987884521, "learning_rate": 6.201042125727483e-05, "loss": 0.0726, "step": 90330 }, { "epoch": 3.054193853747591, "grad_norm": 1.8087348937988281, "learning_rate": 6.200239580097754e-05, "loss": 0.0685, "step": 90340 }, { "epoch": 3.0545319314378445, "grad_norm": 1.6852961778640747, "learning_rate": 6.199437001653929e-05, "loss": 0.067, "step": 90350 }, { "epoch": 3.0548700091280976, "grad_norm": 1.8426740169525146, "learning_rate": 6.198634390417944e-05, "loss": 0.0672, "step": 90360 }, { "epoch": 3.0552080868183507, "grad_norm": 0.6097111701965332, "learning_rate": 6.197831746411748e-05, "loss": 0.0689, "step": 90370 }, { "epoch": 3.055546164508604, "grad_norm": 1.0139747858047485, "learning_rate": 6.197029069657279e-05, "loss": 0.064, "step": 90380 }, { "epoch": 3.0558842421988572, "grad_norm": 3.831913709640503, "learning_rate": 6.196226360176486e-05, "loss": 0.0714, "step": 90390 }, { "epoch": 3.0562223198891103, "grad_norm": 1.306066870689392, "learning_rate": 6.195423617991313e-05, "loss": 0.0629, "step": 90400 }, { "epoch": 3.056560397579364, "grad_norm": 1.4489161968231201, "learning_rate": 6.194620843123709e-05, "loss": 0.0667, "step": 90410 }, { "epoch": 3.056898475269617, "grad_norm": 0.9795233011245728, "learning_rate": 6.19381803559562e-05, "loss": 0.064, "step": 90420 }, { "epoch": 3.0572365529598704, "grad_norm": 1.0440833568572998, "learning_rate": 6.193015195428993e-05, "loss": 0.067, "step": 90430 }, { "epoch": 3.0575746306501235, "grad_norm": 0.6983064413070679, "learning_rate": 6.19221232264578e-05, "loss": 0.0662, "step": 90440 }, { "epoch": 3.0579127083403765, "grad_norm": 0.8755689859390259, "learning_rate": 6.191409417267929e-05, "loss": 0.0664, "step": 90450 }, { "epoch": 3.05825078603063, "grad_norm": 1.0870691537857056, "learning_rate": 6.190606479317392e-05, "loss": 0.0682, "step": 90460 }, { "epoch": 3.058588863720883, "grad_norm": 0.9884268045425415, "learning_rate": 6.189803508816121e-05, "loss": 0.0658, "step": 90470 }, { "epoch": 3.058926941411136, "grad_norm": 0.9527239203453064, "learning_rate": 6.189000505786071e-05, "loss": 0.0699, "step": 90480 }, { "epoch": 3.0592650191013897, "grad_norm": 1.1329272985458374, "learning_rate": 6.188197470249192e-05, "loss": 0.0581, "step": 90490 }, { "epoch": 3.0596030967916428, "grad_norm": 0.7404751777648926, "learning_rate": 6.187394402227442e-05, "loss": 0.0699, "step": 90500 }, { "epoch": 3.059941174481896, "grad_norm": 0.8157228827476501, "learning_rate": 6.186591301742774e-05, "loss": 0.0637, "step": 90510 }, { "epoch": 3.0602792521721494, "grad_norm": 1.2056305408477783, "learning_rate": 6.185788168817146e-05, "loss": 0.0604, "step": 90520 }, { "epoch": 3.0606173298624024, "grad_norm": 0.9072842001914978, "learning_rate": 6.184985003472513e-05, "loss": 0.0676, "step": 90530 }, { "epoch": 3.0609554075526555, "grad_norm": 0.6615332961082458, "learning_rate": 6.184181805730838e-05, "loss": 0.0607, "step": 90540 }, { "epoch": 3.061293485242909, "grad_norm": 0.8484323620796204, "learning_rate": 6.183378575614077e-05, "loss": 0.0709, "step": 90550 }, { "epoch": 3.061631562933162, "grad_norm": 0.7185263633728027, "learning_rate": 6.18257531314419e-05, "loss": 0.062, "step": 90560 }, { "epoch": 3.061969640623415, "grad_norm": 0.7599573135375977, "learning_rate": 6.181772018343137e-05, "loss": 0.0608, "step": 90570 }, { "epoch": 3.0623077183136687, "grad_norm": 1.959012746810913, "learning_rate": 6.180968691232882e-05, "loss": 0.0725, "step": 90580 }, { "epoch": 3.0626457960039217, "grad_norm": 0.8256846070289612, "learning_rate": 6.180165331835386e-05, "loss": 0.0658, "step": 90590 }, { "epoch": 3.062983873694175, "grad_norm": 1.2900583744049072, "learning_rate": 6.179361940172614e-05, "loss": 0.0601, "step": 90600 }, { "epoch": 3.0633219513844283, "grad_norm": 0.9991148114204407, "learning_rate": 6.178558516266529e-05, "loss": 0.0571, "step": 90610 }, { "epoch": 3.0636600290746814, "grad_norm": 0.9491769671440125, "learning_rate": 6.177755060139098e-05, "loss": 0.0626, "step": 90620 }, { "epoch": 3.0639981067649344, "grad_norm": 0.9594551920890808, "learning_rate": 6.176951571812285e-05, "loss": 0.0586, "step": 90630 }, { "epoch": 3.064336184455188, "grad_norm": 0.7379465103149414, "learning_rate": 6.176148051308059e-05, "loss": 0.0693, "step": 90640 }, { "epoch": 3.064674262145441, "grad_norm": 1.9668853282928467, "learning_rate": 6.175344498648385e-05, "loss": 0.0564, "step": 90650 }, { "epoch": 3.065012339835694, "grad_norm": 1.224648356437683, "learning_rate": 6.174540913855237e-05, "loss": 0.0616, "step": 90660 }, { "epoch": 3.0653504175259476, "grad_norm": 1.1348624229431152, "learning_rate": 6.173737296950579e-05, "loss": 0.0653, "step": 90670 }, { "epoch": 3.0656884952162007, "grad_norm": 0.9582610726356506, "learning_rate": 6.172933647956386e-05, "loss": 0.0584, "step": 90680 }, { "epoch": 3.0660265729064538, "grad_norm": 1.5602519512176514, "learning_rate": 6.172129966894627e-05, "loss": 0.0633, "step": 90690 }, { "epoch": 3.0663646505967073, "grad_norm": 1.6957226991653442, "learning_rate": 6.171326253787275e-05, "loss": 0.0604, "step": 90700 }, { "epoch": 3.0667027282869603, "grad_norm": 2.9886717796325684, "learning_rate": 6.170522508656304e-05, "loss": 0.0599, "step": 90710 }, { "epoch": 3.0670408059772134, "grad_norm": 0.6646662950515747, "learning_rate": 6.169718731523686e-05, "loss": 0.0598, "step": 90720 }, { "epoch": 3.067378883667467, "grad_norm": 2.402547836303711, "learning_rate": 6.1689149224114e-05, "loss": 0.0659, "step": 90730 }, { "epoch": 3.06771696135772, "grad_norm": 0.6926572322845459, "learning_rate": 6.168111081341417e-05, "loss": 0.0626, "step": 90740 }, { "epoch": 3.068055039047973, "grad_norm": 0.6744198203086853, "learning_rate": 6.167307208335716e-05, "loss": 0.0618, "step": 90750 }, { "epoch": 3.0683931167382266, "grad_norm": 1.6826128959655762, "learning_rate": 6.166503303416274e-05, "loss": 0.0653, "step": 90760 }, { "epoch": 3.0687311944284796, "grad_norm": 0.9204127192497253, "learning_rate": 6.165699366605072e-05, "loss": 0.0576, "step": 90770 }, { "epoch": 3.0690692721187327, "grad_norm": 1.003116488456726, "learning_rate": 6.164895397924085e-05, "loss": 0.0664, "step": 90780 }, { "epoch": 3.069407349808986, "grad_norm": 0.7195785045623779, "learning_rate": 6.164091397395296e-05, "loss": 0.0662, "step": 90790 }, { "epoch": 3.0697454274992393, "grad_norm": 1.9905855655670166, "learning_rate": 6.163287365040687e-05, "loss": 0.0689, "step": 90800 }, { "epoch": 3.0700835051894924, "grad_norm": 16.22673797607422, "learning_rate": 6.162483300882237e-05, "loss": 0.0696, "step": 90810 }, { "epoch": 3.070421582879746, "grad_norm": 1.2366873025894165, "learning_rate": 6.16167920494193e-05, "loss": 0.0602, "step": 90820 }, { "epoch": 3.070759660569999, "grad_norm": 1.4053677320480347, "learning_rate": 6.160875077241753e-05, "loss": 0.0553, "step": 90830 }, { "epoch": 3.071097738260252, "grad_norm": 0.7188838720321655, "learning_rate": 6.160070917803684e-05, "loss": 0.0697, "step": 90840 }, { "epoch": 3.0714358159505055, "grad_norm": 0.7547112703323364, "learning_rate": 6.159266726649713e-05, "loss": 0.0607, "step": 90850 }, { "epoch": 3.0717738936407586, "grad_norm": 0.8185462355613708, "learning_rate": 6.158462503801824e-05, "loss": 0.066, "step": 90860 }, { "epoch": 3.0721119713310117, "grad_norm": 0.969115674495697, "learning_rate": 6.157658249282007e-05, "loss": 0.0626, "step": 90870 }, { "epoch": 3.072450049021265, "grad_norm": 1.2035791873931885, "learning_rate": 6.156853963112248e-05, "loss": 0.0604, "step": 90880 }, { "epoch": 3.0727881267115182, "grad_norm": 1.8255081176757812, "learning_rate": 6.156049645314537e-05, "loss": 0.059, "step": 90890 }, { "epoch": 3.0731262044017713, "grad_norm": 2.0278592109680176, "learning_rate": 6.155245295910863e-05, "loss": 0.0619, "step": 90900 }, { "epoch": 3.073464282092025, "grad_norm": 1.204256296157837, "learning_rate": 6.154440914923215e-05, "loss": 0.0648, "step": 90910 }, { "epoch": 3.073802359782278, "grad_norm": 0.9444422125816345, "learning_rate": 6.153636502373586e-05, "loss": 0.0627, "step": 90920 }, { "epoch": 3.074140437472531, "grad_norm": 1.6729531288146973, "learning_rate": 6.152832058283969e-05, "loss": 0.0627, "step": 90930 }, { "epoch": 3.0744785151627845, "grad_norm": 0.7428798079490662, "learning_rate": 6.152027582676359e-05, "loss": 0.0621, "step": 90940 }, { "epoch": 3.0748165928530375, "grad_norm": 1.7031170129776, "learning_rate": 6.151223075572744e-05, "loss": 0.0702, "step": 90950 }, { "epoch": 3.075154670543291, "grad_norm": 1.6803895235061646, "learning_rate": 6.150418536995125e-05, "loss": 0.0652, "step": 90960 }, { "epoch": 3.075492748233544, "grad_norm": 0.8199710845947266, "learning_rate": 6.149613966965493e-05, "loss": 0.0601, "step": 90970 }, { "epoch": 3.075830825923797, "grad_norm": 0.6088709831237793, "learning_rate": 6.14880936550585e-05, "loss": 0.063, "step": 90980 }, { "epoch": 3.0761689036140507, "grad_norm": 0.9129000306129456, "learning_rate": 6.148004732638188e-05, "loss": 0.0611, "step": 90990 }, { "epoch": 3.076506981304304, "grad_norm": 1.7514575719833374, "learning_rate": 6.147200068384511e-05, "loss": 0.0656, "step": 91000 }, { "epoch": 3.076845058994557, "grad_norm": 2.1306777000427246, "learning_rate": 6.146395372766812e-05, "loss": 0.0661, "step": 91010 }, { "epoch": 3.0771831366848104, "grad_norm": 2.19429874420166, "learning_rate": 6.145590645807097e-05, "loss": 0.0592, "step": 91020 }, { "epoch": 3.0775212143750634, "grad_norm": 0.8582308292388916, "learning_rate": 6.144785887527362e-05, "loss": 0.07, "step": 91030 }, { "epoch": 3.0778592920653165, "grad_norm": 0.7861725687980652, "learning_rate": 6.143981097949611e-05, "loss": 0.0589, "step": 91040 }, { "epoch": 3.07819736975557, "grad_norm": 1.1573951244354248, "learning_rate": 6.143176277095847e-05, "loss": 0.0622, "step": 91050 }, { "epoch": 3.078535447445823, "grad_norm": 0.9539462327957153, "learning_rate": 6.142371424988074e-05, "loss": 0.0628, "step": 91060 }, { "epoch": 3.078873525136076, "grad_norm": 0.9393266439437866, "learning_rate": 6.141566541648294e-05, "loss": 0.0568, "step": 91070 }, { "epoch": 3.0792116028263297, "grad_norm": 0.8303041458129883, "learning_rate": 6.140761627098515e-05, "loss": 0.0643, "step": 91080 }, { "epoch": 3.0795496805165827, "grad_norm": 1.5563619136810303, "learning_rate": 6.13995668136074e-05, "loss": 0.0678, "step": 91090 }, { "epoch": 3.079887758206836, "grad_norm": 0.8002598285675049, "learning_rate": 6.139151704456978e-05, "loss": 0.062, "step": 91100 }, { "epoch": 3.0802258358970893, "grad_norm": 0.8548060059547424, "learning_rate": 6.138346696409236e-05, "loss": 0.0577, "step": 91110 }, { "epoch": 3.0805639135873424, "grad_norm": 0.7012787461280823, "learning_rate": 6.137541657239525e-05, "loss": 0.0564, "step": 91120 }, { "epoch": 3.0809019912775955, "grad_norm": 0.9511988759040833, "learning_rate": 6.136736586969851e-05, "loss": 0.0592, "step": 91130 }, { "epoch": 3.081240068967849, "grad_norm": 0.8986364603042603, "learning_rate": 6.135931485622225e-05, "loss": 0.0572, "step": 91140 }, { "epoch": 3.081578146658102, "grad_norm": 8.155345916748047, "learning_rate": 6.135126353218661e-05, "loss": 0.0591, "step": 91150 }, { "epoch": 3.081916224348355, "grad_norm": 1.2236850261688232, "learning_rate": 6.134321189781166e-05, "loss": 0.0644, "step": 91160 }, { "epoch": 3.0822543020386086, "grad_norm": 0.9033488631248474, "learning_rate": 6.13351599533176e-05, "loss": 0.0578, "step": 91170 }, { "epoch": 3.0825923797288617, "grad_norm": 1.132009744644165, "learning_rate": 6.132710769892449e-05, "loss": 0.0653, "step": 91180 }, { "epoch": 3.0829304574191148, "grad_norm": 10.05895709991455, "learning_rate": 6.131905513485252e-05, "loss": 0.0634, "step": 91190 }, { "epoch": 3.0832685351093683, "grad_norm": 1.0354355573654175, "learning_rate": 6.131100226132184e-05, "loss": 0.0613, "step": 91200 }, { "epoch": 3.0836066127996213, "grad_norm": 2.503243923187256, "learning_rate": 6.130294907855261e-05, "loss": 0.0648, "step": 91210 }, { "epoch": 3.0839446904898744, "grad_norm": 2.0148279666900635, "learning_rate": 6.129489558676499e-05, "loss": 0.0615, "step": 91220 }, { "epoch": 3.084282768180128, "grad_norm": 0.9038591980934143, "learning_rate": 6.128684178617917e-05, "loss": 0.0648, "step": 91230 }, { "epoch": 3.084620845870381, "grad_norm": 0.6354416012763977, "learning_rate": 6.127878767701533e-05, "loss": 0.0619, "step": 91240 }, { "epoch": 3.084958923560634, "grad_norm": 0.9259825348854065, "learning_rate": 6.127073325949368e-05, "loss": 0.0611, "step": 91250 }, { "epoch": 3.0852970012508876, "grad_norm": 1.1668726205825806, "learning_rate": 6.126267853383442e-05, "loss": 0.0639, "step": 91260 }, { "epoch": 3.0856350789411406, "grad_norm": 0.9041481018066406, "learning_rate": 6.125462350025776e-05, "loss": 0.0636, "step": 91270 }, { "epoch": 3.0859731566313937, "grad_norm": 2.088789224624634, "learning_rate": 6.124656815898392e-05, "loss": 0.0563, "step": 91280 }, { "epoch": 3.0863112343216472, "grad_norm": 0.8914148211479187, "learning_rate": 6.123851251023313e-05, "loss": 0.0645, "step": 91290 }, { "epoch": 3.0866493120119003, "grad_norm": 0.8195071220397949, "learning_rate": 6.123045655422563e-05, "loss": 0.0597, "step": 91300 }, { "epoch": 3.0869873897021534, "grad_norm": 1.7317265272140503, "learning_rate": 6.122240029118168e-05, "loss": 0.0569, "step": 91310 }, { "epoch": 3.087325467392407, "grad_norm": 1.3503016233444214, "learning_rate": 6.121434372132149e-05, "loss": 0.0618, "step": 91320 }, { "epoch": 3.08766354508266, "grad_norm": 1.1984888315200806, "learning_rate": 6.120628684486538e-05, "loss": 0.0614, "step": 91330 }, { "epoch": 3.088001622772913, "grad_norm": 1.0541155338287354, "learning_rate": 6.119822966203359e-05, "loss": 0.0658, "step": 91340 }, { "epoch": 3.0883397004631665, "grad_norm": 0.7461380362510681, "learning_rate": 6.119017217304641e-05, "loss": 0.0629, "step": 91350 }, { "epoch": 3.0886777781534196, "grad_norm": 0.6616825461387634, "learning_rate": 6.118211437812412e-05, "loss": 0.0627, "step": 91360 }, { "epoch": 3.0890158558436727, "grad_norm": 1.9157798290252686, "learning_rate": 6.117405627748703e-05, "loss": 0.0559, "step": 91370 }, { "epoch": 3.089353933533926, "grad_norm": 0.9167060256004333, "learning_rate": 6.116599787135545e-05, "loss": 0.0648, "step": 91380 }, { "epoch": 3.0896920112241792, "grad_norm": 2.4139668941497803, "learning_rate": 6.115793915994968e-05, "loss": 0.061, "step": 91390 }, { "epoch": 3.0900300889144328, "grad_norm": 0.8236344456672668, "learning_rate": 6.114988014349004e-05, "loss": 0.0608, "step": 91400 }, { "epoch": 3.090368166604686, "grad_norm": 1.3238352537155151, "learning_rate": 6.114182082219686e-05, "loss": 0.0592, "step": 91410 }, { "epoch": 3.090706244294939, "grad_norm": 0.6072728037834167, "learning_rate": 6.113376119629049e-05, "loss": 0.0607, "step": 91420 }, { "epoch": 3.0910443219851924, "grad_norm": 0.7949411273002625, "learning_rate": 6.112570126599125e-05, "loss": 0.0639, "step": 91430 }, { "epoch": 3.0913823996754455, "grad_norm": 3.3982815742492676, "learning_rate": 6.111764103151955e-05, "loss": 0.0599, "step": 91440 }, { "epoch": 3.0917204773656985, "grad_norm": 1.5822783708572388, "learning_rate": 6.11095804930957e-05, "loss": 0.0657, "step": 91450 }, { "epoch": 3.092058555055952, "grad_norm": 1.3140963315963745, "learning_rate": 6.110151965094011e-05, "loss": 0.0611, "step": 91460 }, { "epoch": 3.092396632746205, "grad_norm": 0.7496457695960999, "learning_rate": 6.109345850527314e-05, "loss": 0.0579, "step": 91470 }, { "epoch": 3.092734710436458, "grad_norm": 1.0216070413589478, "learning_rate": 6.108539705631517e-05, "loss": 0.0618, "step": 91480 }, { "epoch": 3.0930727881267117, "grad_norm": 0.6817342042922974, "learning_rate": 6.107733530428661e-05, "loss": 0.0574, "step": 91490 }, { "epoch": 3.093410865816965, "grad_norm": 0.7573391199111938, "learning_rate": 6.106927324940788e-05, "loss": 0.0609, "step": 91500 }, { "epoch": 3.093748943507218, "grad_norm": 0.8610051870346069, "learning_rate": 6.106121089189937e-05, "loss": 0.0658, "step": 91510 }, { "epoch": 3.0940870211974714, "grad_norm": 1.6370657682418823, "learning_rate": 6.10531482319815e-05, "loss": 0.0652, "step": 91520 }, { "epoch": 3.0944250988877244, "grad_norm": 0.7429580688476562, "learning_rate": 6.104508526987472e-05, "loss": 0.0661, "step": 91530 }, { "epoch": 3.0947631765779775, "grad_norm": 1.094343900680542, "learning_rate": 6.103702200579946e-05, "loss": 0.0621, "step": 91540 }, { "epoch": 3.095101254268231, "grad_norm": 1.4426119327545166, "learning_rate": 6.1028958439976146e-05, "loss": 0.0596, "step": 91550 }, { "epoch": 3.095439331958484, "grad_norm": 1.8163537979125977, "learning_rate": 6.1020894572625275e-05, "loss": 0.0641, "step": 91560 }, { "epoch": 3.095777409648737, "grad_norm": 0.8192230463027954, "learning_rate": 6.101283040396728e-05, "loss": 0.0672, "step": 91570 }, { "epoch": 3.0961154873389907, "grad_norm": 1.2492506504058838, "learning_rate": 6.100476593422263e-05, "loss": 0.0578, "step": 91580 }, { "epoch": 3.0964535650292437, "grad_norm": 0.7532088756561279, "learning_rate": 6.099670116361182e-05, "loss": 0.059, "step": 91590 }, { "epoch": 3.096791642719497, "grad_norm": 0.6067982912063599, "learning_rate": 6.0988636092355335e-05, "loss": 0.0591, "step": 91600 }, { "epoch": 3.0971297204097503, "grad_norm": 0.7114478945732117, "learning_rate": 6.098057072067367e-05, "loss": 0.0552, "step": 91610 }, { "epoch": 3.0974677981000034, "grad_norm": 1.0750370025634766, "learning_rate": 6.097250504878731e-05, "loss": 0.0543, "step": 91620 }, { "epoch": 3.0978058757902565, "grad_norm": 0.7078461647033691, "learning_rate": 6.09644390769168e-05, "loss": 0.0587, "step": 91630 }, { "epoch": 3.09814395348051, "grad_norm": 0.8717377781867981, "learning_rate": 6.0956372805282634e-05, "loss": 0.0611, "step": 91640 }, { "epoch": 3.098482031170763, "grad_norm": 0.9010372757911682, "learning_rate": 6.094830623410537e-05, "loss": 0.0692, "step": 91650 }, { "epoch": 3.098820108861016, "grad_norm": 0.8321270942687988, "learning_rate": 6.094023936360551e-05, "loss": 0.0597, "step": 91660 }, { "epoch": 3.0991581865512696, "grad_norm": 0.5002503395080566, "learning_rate": 6.093217219400362e-05, "loss": 0.0608, "step": 91670 }, { "epoch": 3.0994962642415227, "grad_norm": 1.7655870914459229, "learning_rate": 6.092410472552025e-05, "loss": 0.0615, "step": 91680 }, { "epoch": 3.0998343419317758, "grad_norm": 0.9957497715950012, "learning_rate": 6.0916036958375966e-05, "loss": 0.0587, "step": 91690 }, { "epoch": 3.1001724196220293, "grad_norm": 0.607448935508728, "learning_rate": 6.090796889279132e-05, "loss": 0.059, "step": 91700 }, { "epoch": 3.1005104973122823, "grad_norm": 0.8430808782577515, "learning_rate": 6.089990052898692e-05, "loss": 0.0634, "step": 91710 }, { "epoch": 3.1008485750025354, "grad_norm": 0.830194890499115, "learning_rate": 6.0891831867183316e-05, "loss": 0.0615, "step": 91720 }, { "epoch": 3.101186652692789, "grad_norm": 0.5604565739631653, "learning_rate": 6.088376290760114e-05, "loss": 0.0574, "step": 91730 }, { "epoch": 3.101524730383042, "grad_norm": 0.8839113116264343, "learning_rate": 6.087569365046095e-05, "loss": 0.0587, "step": 91740 }, { "epoch": 3.101862808073295, "grad_norm": 0.9341572523117065, "learning_rate": 6.08676240959834e-05, "loss": 0.0618, "step": 91750 }, { "epoch": 3.1022008857635486, "grad_norm": 0.8115862011909485, "learning_rate": 6.085955424438908e-05, "loss": 0.0683, "step": 91760 }, { "epoch": 3.1025389634538016, "grad_norm": 0.7010939717292786, "learning_rate": 6.085148409589864e-05, "loss": 0.06, "step": 91770 }, { "epoch": 3.1028770411440547, "grad_norm": 0.7713167667388916, "learning_rate": 6.0843413650732685e-05, "loss": 0.0629, "step": 91780 }, { "epoch": 3.1032151188343082, "grad_norm": 0.4684099853038788, "learning_rate": 6.083534290911188e-05, "loss": 0.0584, "step": 91790 }, { "epoch": 3.1035531965245613, "grad_norm": 1.1363756656646729, "learning_rate": 6.082727187125688e-05, "loss": 0.0567, "step": 91800 }, { "epoch": 3.1038912742148144, "grad_norm": 2.0178983211517334, "learning_rate": 6.081920053738832e-05, "loss": 0.0519, "step": 91810 }, { "epoch": 3.104229351905068, "grad_norm": 0.47991567850112915, "learning_rate": 6.081112890772689e-05, "loss": 0.0603, "step": 91820 }, { "epoch": 3.104567429595321, "grad_norm": 1.1689878702163696, "learning_rate": 6.0803056982493255e-05, "loss": 0.0641, "step": 91830 }, { "epoch": 3.104905507285574, "grad_norm": 0.8570847511291504, "learning_rate": 6.0794984761908114e-05, "loss": 0.0575, "step": 91840 }, { "epoch": 3.1052435849758275, "grad_norm": 1.5688971281051636, "learning_rate": 6.078691224619213e-05, "loss": 0.065, "step": 91850 }, { "epoch": 3.1055816626660806, "grad_norm": 0.6181672215461731, "learning_rate": 6.0778839435566014e-05, "loss": 0.0611, "step": 91860 }, { "epoch": 3.1059197403563337, "grad_norm": 0.5466929078102112, "learning_rate": 6.077076633025049e-05, "loss": 0.0622, "step": 91870 }, { "epoch": 3.106257818046587, "grad_norm": 0.5793380737304688, "learning_rate": 6.076269293046626e-05, "loss": 0.0591, "step": 91880 }, { "epoch": 3.1065958957368403, "grad_norm": 0.49509167671203613, "learning_rate": 6.075461923643403e-05, "loss": 0.0597, "step": 91890 }, { "epoch": 3.1069339734270933, "grad_norm": 2.5036299228668213, "learning_rate": 6.074654524837458e-05, "loss": 0.0536, "step": 91900 }, { "epoch": 3.107272051117347, "grad_norm": 1.3688374757766724, "learning_rate": 6.07384709665086e-05, "loss": 0.0588, "step": 91910 }, { "epoch": 3.1076101288076, "grad_norm": 2.1891088485717773, "learning_rate": 6.073039639105686e-05, "loss": 0.0658, "step": 91920 }, { "epoch": 3.1079482064978534, "grad_norm": 0.5810253620147705, "learning_rate": 6.072232152224011e-05, "loss": 0.0621, "step": 91930 }, { "epoch": 3.1082862841881065, "grad_norm": 0.9255905747413635, "learning_rate": 6.071424636027912e-05, "loss": 0.0663, "step": 91940 }, { "epoch": 3.1086243618783596, "grad_norm": 1.2286971807479858, "learning_rate": 6.0706170905394665e-05, "loss": 0.0586, "step": 91950 }, { "epoch": 3.108962439568613, "grad_norm": 2.359813928604126, "learning_rate": 6.069809515780753e-05, "loss": 0.065, "step": 91960 }, { "epoch": 3.109300517258866, "grad_norm": 0.751932680606842, "learning_rate": 6.069001911773847e-05, "loss": 0.0669, "step": 91970 }, { "epoch": 3.109638594949119, "grad_norm": 1.201062560081482, "learning_rate": 6.068194278540833e-05, "loss": 0.0674, "step": 91980 }, { "epoch": 3.1099766726393727, "grad_norm": 1.6797387599945068, "learning_rate": 6.067386616103785e-05, "loss": 0.0638, "step": 91990 }, { "epoch": 3.110314750329626, "grad_norm": 3.2656424045562744, "learning_rate": 6.066578924484791e-05, "loss": 0.0584, "step": 92000 }, { "epoch": 3.110652828019879, "grad_norm": 1.6763815879821777, "learning_rate": 6.065771203705929e-05, "loss": 0.0625, "step": 92010 }, { "epoch": 3.1109909057101324, "grad_norm": 2.1852385997772217, "learning_rate": 6.064963453789283e-05, "loss": 0.0597, "step": 92020 }, { "epoch": 3.1113289834003854, "grad_norm": 1.7241278886795044, "learning_rate": 6.064155674756936e-05, "loss": 0.0623, "step": 92030 }, { "epoch": 3.1116670610906385, "grad_norm": 0.8901191353797913, "learning_rate": 6.0633478666309714e-05, "loss": 0.0665, "step": 92040 }, { "epoch": 3.112005138780892, "grad_norm": 1.1272348165512085, "learning_rate": 6.0625400294334765e-05, "loss": 0.0677, "step": 92050 }, { "epoch": 3.112343216471145, "grad_norm": 0.3758770227432251, "learning_rate": 6.061732163186537e-05, "loss": 0.0581, "step": 92060 }, { "epoch": 3.112681294161398, "grad_norm": 0.7187371253967285, "learning_rate": 6.060924267912239e-05, "loss": 0.0567, "step": 92070 }, { "epoch": 3.1130193718516517, "grad_norm": 0.8833014965057373, "learning_rate": 6.060116343632669e-05, "loss": 0.0629, "step": 92080 }, { "epoch": 3.1133574495419047, "grad_norm": 0.8087466955184937, "learning_rate": 6.059308390369919e-05, "loss": 0.0596, "step": 92090 }, { "epoch": 3.113695527232158, "grad_norm": 1.0654582977294922, "learning_rate": 6.0585004081460725e-05, "loss": 0.0587, "step": 92100 }, { "epoch": 3.1140336049224113, "grad_norm": 0.5920616388320923, "learning_rate": 6.057692396983224e-05, "loss": 0.067, "step": 92110 }, { "epoch": 3.1143716826126644, "grad_norm": 0.7867711186408997, "learning_rate": 6.056884356903462e-05, "loss": 0.061, "step": 92120 }, { "epoch": 3.1147097603029175, "grad_norm": 0.7626368403434753, "learning_rate": 6.05607628792888e-05, "loss": 0.0606, "step": 92130 }, { "epoch": 3.115047837993171, "grad_norm": 1.0688847303390503, "learning_rate": 6.055268190081568e-05, "loss": 0.0641, "step": 92140 }, { "epoch": 3.115385915683424, "grad_norm": 0.7585014700889587, "learning_rate": 6.0544600633836214e-05, "loss": 0.0608, "step": 92150 }, { "epoch": 3.115723993373677, "grad_norm": 0.5534887313842773, "learning_rate": 6.053651907857131e-05, "loss": 0.0595, "step": 92160 }, { "epoch": 3.1160620710639306, "grad_norm": 0.9721804857254028, "learning_rate": 6.052843723524195e-05, "loss": 0.0677, "step": 92170 }, { "epoch": 3.1164001487541837, "grad_norm": 0.584949254989624, "learning_rate": 6.052035510406905e-05, "loss": 0.0656, "step": 92180 }, { "epoch": 3.1167382264444368, "grad_norm": 0.3959891200065613, "learning_rate": 6.0512272685273605e-05, "loss": 0.0573, "step": 92190 }, { "epoch": 3.1170763041346903, "grad_norm": 0.7210841178894043, "learning_rate": 6.0504189979076574e-05, "loss": 0.0669, "step": 92200 }, { "epoch": 3.1174143818249433, "grad_norm": 0.6333999633789062, "learning_rate": 6.0496106985698944e-05, "loss": 0.0546, "step": 92210 }, { "epoch": 3.1177524595151964, "grad_norm": 0.5394607186317444, "learning_rate": 6.048802370536168e-05, "loss": 0.0617, "step": 92220 }, { "epoch": 3.11809053720545, "grad_norm": 0.6441255807876587, "learning_rate": 6.04799401382858e-05, "loss": 0.0565, "step": 92230 }, { "epoch": 3.118428614895703, "grad_norm": 0.5575159192085266, "learning_rate": 6.0471856284692286e-05, "loss": 0.0595, "step": 92240 }, { "epoch": 3.118766692585956, "grad_norm": 0.46912863850593567, "learning_rate": 6.0463772144802145e-05, "loss": 0.0623, "step": 92250 }, { "epoch": 3.1191047702762096, "grad_norm": 1.0048768520355225, "learning_rate": 6.045568771883642e-05, "loss": 0.0586, "step": 92260 }, { "epoch": 3.1194428479664627, "grad_norm": 1.1341632604599, "learning_rate": 6.044760300701612e-05, "loss": 0.0612, "step": 92270 }, { "epoch": 3.1197809256567157, "grad_norm": 0.7819059491157532, "learning_rate": 6.043951800956227e-05, "loss": 0.0528, "step": 92280 }, { "epoch": 3.1201190033469692, "grad_norm": 0.9839978218078613, "learning_rate": 6.043143272669591e-05, "loss": 0.064, "step": 92290 }, { "epoch": 3.1204570810372223, "grad_norm": 0.8735873103141785, "learning_rate": 6.0423347158638134e-05, "loss": 0.0642, "step": 92300 }, { "epoch": 3.1207951587274754, "grad_norm": 1.375313639640808, "learning_rate": 6.0415261305609925e-05, "loss": 0.0603, "step": 92310 }, { "epoch": 3.121133236417729, "grad_norm": 0.8189809322357178, "learning_rate": 6.040717516783241e-05, "loss": 0.0505, "step": 92320 }, { "epoch": 3.121471314107982, "grad_norm": 3.0665719509124756, "learning_rate": 6.039908874552661e-05, "loss": 0.0632, "step": 92330 }, { "epoch": 3.121809391798235, "grad_norm": 1.117332100868225, "learning_rate": 6.039100203891366e-05, "loss": 0.0651, "step": 92340 }, { "epoch": 3.1221474694884885, "grad_norm": 0.7241660952568054, "learning_rate": 6.038291504821461e-05, "loss": 0.0569, "step": 92350 }, { "epoch": 3.1224855471787416, "grad_norm": 0.9639180898666382, "learning_rate": 6.037482777365056e-05, "loss": 0.0615, "step": 92360 }, { "epoch": 3.122823624868995, "grad_norm": 0.6386992335319519, "learning_rate": 6.0366740215442595e-05, "loss": 0.0566, "step": 92370 }, { "epoch": 3.123161702559248, "grad_norm": 0.6479867696762085, "learning_rate": 6.035865237381187e-05, "loss": 0.0606, "step": 92380 }, { "epoch": 3.1234997802495013, "grad_norm": 0.7179564833641052, "learning_rate": 6.035056424897948e-05, "loss": 0.0625, "step": 92390 }, { "epoch": 3.1238378579397548, "grad_norm": 0.7610864043235779, "learning_rate": 6.034247584116656e-05, "loss": 0.0596, "step": 92400 }, { "epoch": 3.124175935630008, "grad_norm": 0.7499779462814331, "learning_rate": 6.033438715059422e-05, "loss": 0.0627, "step": 92410 }, { "epoch": 3.124514013320261, "grad_norm": 0.6538314819335938, "learning_rate": 6.032629817748363e-05, "loss": 0.0595, "step": 92420 }, { "epoch": 3.1248520910105144, "grad_norm": 0.5030601620674133, "learning_rate": 6.0318208922055905e-05, "loss": 0.0568, "step": 92430 }, { "epoch": 3.1251901687007675, "grad_norm": 0.6265441179275513, "learning_rate": 6.031011938453224e-05, "loss": 0.0606, "step": 92440 }, { "epoch": 3.1255282463910206, "grad_norm": 0.5628311038017273, "learning_rate": 6.0302029565133775e-05, "loss": 0.06, "step": 92450 }, { "epoch": 3.125866324081274, "grad_norm": 0.951859176158905, "learning_rate": 6.029393946408172e-05, "loss": 0.0614, "step": 92460 }, { "epoch": 3.126204401771527, "grad_norm": 0.6669337153434753, "learning_rate": 6.02858490815972e-05, "loss": 0.0668, "step": 92470 }, { "epoch": 3.12654247946178, "grad_norm": 0.42423373460769653, "learning_rate": 6.0277758417901444e-05, "loss": 0.0538, "step": 92480 }, { "epoch": 3.1268805571520337, "grad_norm": 0.6447725296020508, "learning_rate": 6.026966747321563e-05, "loss": 0.0603, "step": 92490 }, { "epoch": 3.127218634842287, "grad_norm": 1.010148286819458, "learning_rate": 6.026157624776095e-05, "loss": 0.0546, "step": 92500 }, { "epoch": 3.12755671253254, "grad_norm": 0.7049450874328613, "learning_rate": 6.025348474175865e-05, "loss": 0.0612, "step": 92510 }, { "epoch": 3.1278947902227934, "grad_norm": 0.924684464931488, "learning_rate": 6.0245392955429924e-05, "loss": 0.0656, "step": 92520 }, { "epoch": 3.1282328679130464, "grad_norm": 0.47478732466697693, "learning_rate": 6.023730088899601e-05, "loss": 0.0618, "step": 92530 }, { "epoch": 3.1285709456032995, "grad_norm": 0.6606351733207703, "learning_rate": 6.022920854267812e-05, "loss": 0.0544, "step": 92540 }, { "epoch": 3.128909023293553, "grad_norm": 0.35601791739463806, "learning_rate": 6.0221115916697524e-05, "loss": 0.0591, "step": 92550 }, { "epoch": 3.129247100983806, "grad_norm": 0.4318622946739197, "learning_rate": 6.021302301127545e-05, "loss": 0.0576, "step": 92560 }, { "epoch": 3.129585178674059, "grad_norm": 0.8421592116355896, "learning_rate": 6.020492982663317e-05, "loss": 0.0564, "step": 92570 }, { "epoch": 3.1299232563643127, "grad_norm": 0.7800697088241577, "learning_rate": 6.0196836362991935e-05, "loss": 0.065, "step": 92580 }, { "epoch": 3.1302613340545657, "grad_norm": 1.0432003736495972, "learning_rate": 6.0188742620573034e-05, "loss": 0.0673, "step": 92590 }, { "epoch": 3.130599411744819, "grad_norm": 1.0819529294967651, "learning_rate": 6.018064859959773e-05, "loss": 0.0527, "step": 92600 }, { "epoch": 3.1309374894350723, "grad_norm": 0.9151217341423035, "learning_rate": 6.017255430028733e-05, "loss": 0.0596, "step": 92610 }, { "epoch": 3.1312755671253254, "grad_norm": 1.1105695962905884, "learning_rate": 6.016445972286311e-05, "loss": 0.0625, "step": 92620 }, { "epoch": 3.1316136448155785, "grad_norm": 0.7352160215377808, "learning_rate": 6.015636486754638e-05, "loss": 0.0555, "step": 92630 }, { "epoch": 3.131951722505832, "grad_norm": 0.8454814553260803, "learning_rate": 6.014826973455844e-05, "loss": 0.0555, "step": 92640 }, { "epoch": 3.132289800196085, "grad_norm": 0.6544723510742188, "learning_rate": 6.014017432412063e-05, "loss": 0.0599, "step": 92650 }, { "epoch": 3.132627877886338, "grad_norm": 1.1222028732299805, "learning_rate": 6.013207863645426e-05, "loss": 0.0649, "step": 92660 }, { "epoch": 3.1329659555765916, "grad_norm": 1.3178898096084595, "learning_rate": 6.012398267178068e-05, "loss": 0.0676, "step": 92670 }, { "epoch": 3.1333040332668447, "grad_norm": 1.4415950775146484, "learning_rate": 6.01158864303212e-05, "loss": 0.0582, "step": 92680 }, { "epoch": 3.1336421109570978, "grad_norm": 0.9318828582763672, "learning_rate": 6.0107789912297185e-05, "loss": 0.0603, "step": 92690 }, { "epoch": 3.1339801886473513, "grad_norm": 0.8361856937408447, "learning_rate": 6.009969311792999e-05, "loss": 0.0581, "step": 92700 }, { "epoch": 3.1343182663376044, "grad_norm": 1.8940318822860718, "learning_rate": 6.009159604744099e-05, "loss": 0.0587, "step": 92710 }, { "epoch": 3.1346563440278574, "grad_norm": 1.2084805965423584, "learning_rate": 6.0083498701051545e-05, "loss": 0.0599, "step": 92720 }, { "epoch": 3.134994421718111, "grad_norm": 1.1745096445083618, "learning_rate": 6.007540107898303e-05, "loss": 0.0657, "step": 92730 }, { "epoch": 3.135332499408364, "grad_norm": 1.3821160793304443, "learning_rate": 6.0067303181456834e-05, "loss": 0.0676, "step": 92740 }, { "epoch": 3.135670577098617, "grad_norm": 1.0709731578826904, "learning_rate": 6.005920500869434e-05, "loss": 0.063, "step": 92750 }, { "epoch": 3.1360086547888706, "grad_norm": 3.5783679485321045, "learning_rate": 6.005110656091697e-05, "loss": 0.0609, "step": 92760 }, { "epoch": 3.1363467324791237, "grad_norm": 1.7142874002456665, "learning_rate": 6.004300783834613e-05, "loss": 0.0696, "step": 92770 }, { "epoch": 3.1366848101693767, "grad_norm": 1.8504365682601929, "learning_rate": 6.003490884120322e-05, "loss": 0.063, "step": 92780 }, { "epoch": 3.1370228878596302, "grad_norm": 1.4646263122558594, "learning_rate": 6.0026809569709675e-05, "loss": 0.0619, "step": 92790 }, { "epoch": 3.1373609655498833, "grad_norm": 0.7176147103309631, "learning_rate": 6.001871002408692e-05, "loss": 0.0613, "step": 92800 }, { "epoch": 3.1376990432401364, "grad_norm": 2.1151785850524902, "learning_rate": 6.001061020455641e-05, "loss": 0.0605, "step": 92810 }, { "epoch": 3.13803712093039, "grad_norm": 1.4558829069137573, "learning_rate": 6.000251011133956e-05, "loss": 0.0594, "step": 92820 }, { "epoch": 3.138375198620643, "grad_norm": 0.9202450513839722, "learning_rate": 5.999440974465784e-05, "loss": 0.062, "step": 92830 }, { "epoch": 3.138713276310896, "grad_norm": 1.5222876071929932, "learning_rate": 5.998630910473273e-05, "loss": 0.0569, "step": 92840 }, { "epoch": 3.1390513540011495, "grad_norm": 2.333970546722412, "learning_rate": 5.997820819178567e-05, "loss": 0.0567, "step": 92850 }, { "epoch": 3.1393894316914026, "grad_norm": 3.0178709030151367, "learning_rate": 5.9970107006038155e-05, "loss": 0.0645, "step": 92860 }, { "epoch": 3.1397275093816557, "grad_norm": 1.1308540105819702, "learning_rate": 5.996200554771165e-05, "loss": 0.0683, "step": 92870 }, { "epoch": 3.140065587071909, "grad_norm": 1.1479380130767822, "learning_rate": 5.9953903817027656e-05, "loss": 0.0658, "step": 92880 }, { "epoch": 3.1404036647621623, "grad_norm": 3.723206043243408, "learning_rate": 5.994580181420767e-05, "loss": 0.059, "step": 92890 }, { "epoch": 3.1407417424524153, "grad_norm": 3.960710287094116, "learning_rate": 5.9937699539473215e-05, "loss": 0.0609, "step": 92900 }, { "epoch": 3.141079820142669, "grad_norm": 1.228244423866272, "learning_rate": 5.9929596993045766e-05, "loss": 0.0661, "step": 92910 }, { "epoch": 3.141417897832922, "grad_norm": 0.9687169194221497, "learning_rate": 5.992149417514689e-05, "loss": 0.0614, "step": 92920 }, { "epoch": 3.1417559755231754, "grad_norm": 2.314814567565918, "learning_rate": 5.991339108599807e-05, "loss": 0.0662, "step": 92930 }, { "epoch": 3.1420940532134285, "grad_norm": 0.8952468633651733, "learning_rate": 5.990528772582087e-05, "loss": 0.0593, "step": 92940 }, { "epoch": 3.1424321309036816, "grad_norm": 2.0235095024108887, "learning_rate": 5.989718409483681e-05, "loss": 0.058, "step": 92950 }, { "epoch": 3.142770208593935, "grad_norm": 1.063615322113037, "learning_rate": 5.9889080193267475e-05, "loss": 0.061, "step": 92960 }, { "epoch": 3.143108286284188, "grad_norm": 3.758364200592041, "learning_rate": 5.98809760213344e-05, "loss": 0.0625, "step": 92970 }, { "epoch": 3.143446363974441, "grad_norm": 1.4684723615646362, "learning_rate": 5.987287157925915e-05, "loss": 0.0583, "step": 92980 }, { "epoch": 3.1437844416646947, "grad_norm": 0.9780966639518738, "learning_rate": 5.98647668672633e-05, "loss": 0.0597, "step": 92990 }, { "epoch": 3.144122519354948, "grad_norm": 1.2694460153579712, "learning_rate": 5.985666188556842e-05, "loss": 0.0615, "step": 93000 }, { "epoch": 3.144460597045201, "grad_norm": 2.845475196838379, "learning_rate": 5.9848556634396115e-05, "loss": 0.06, "step": 93010 }, { "epoch": 3.1447986747354544, "grad_norm": 0.9781643152236938, "learning_rate": 5.9840451113967975e-05, "loss": 0.0587, "step": 93020 }, { "epoch": 3.1451367524257074, "grad_norm": 3.8030707836151123, "learning_rate": 5.98323453245056e-05, "loss": 0.0672, "step": 93030 }, { "epoch": 3.1454748301159605, "grad_norm": 2.270996332168579, "learning_rate": 5.982423926623059e-05, "loss": 0.0681, "step": 93040 }, { "epoch": 3.145812907806214, "grad_norm": 3.6120238304138184, "learning_rate": 5.981613293936458e-05, "loss": 0.0586, "step": 93050 }, { "epoch": 3.146150985496467, "grad_norm": 1.4016342163085938, "learning_rate": 5.980802634412916e-05, "loss": 0.0673, "step": 93060 }, { "epoch": 3.14648906318672, "grad_norm": 1.0355018377304077, "learning_rate": 5.9799919480746006e-05, "loss": 0.061, "step": 93070 }, { "epoch": 3.1468271408769737, "grad_norm": 1.3790591955184937, "learning_rate": 5.9791812349436726e-05, "loss": 0.0684, "step": 93080 }, { "epoch": 3.1471652185672268, "grad_norm": 5.63170051574707, "learning_rate": 5.978370495042298e-05, "loss": 0.0694, "step": 93090 }, { "epoch": 3.14750329625748, "grad_norm": 1.6209466457366943, "learning_rate": 5.977559728392641e-05, "loss": 0.0635, "step": 93100 }, { "epoch": 3.1478413739477333, "grad_norm": 1.1026724576950073, "learning_rate": 5.97674893501687e-05, "loss": 0.0619, "step": 93110 }, { "epoch": 3.1481794516379864, "grad_norm": 1.868091344833374, "learning_rate": 5.9759381149371486e-05, "loss": 0.0653, "step": 93120 }, { "epoch": 3.1485175293282395, "grad_norm": 1.5836467742919922, "learning_rate": 5.9751272681756465e-05, "loss": 0.0707, "step": 93130 }, { "epoch": 3.148855607018493, "grad_norm": 2.8232855796813965, "learning_rate": 5.974316394754529e-05, "loss": 0.07, "step": 93140 }, { "epoch": 3.149193684708746, "grad_norm": 1.8452484607696533, "learning_rate": 5.97350549469597e-05, "loss": 0.0706, "step": 93150 }, { "epoch": 3.149531762398999, "grad_norm": 0.9380190372467041, "learning_rate": 5.9726945680221344e-05, "loss": 0.0657, "step": 93160 }, { "epoch": 3.1498698400892526, "grad_norm": 6.491121768951416, "learning_rate": 5.9718836147551974e-05, "loss": 0.0594, "step": 93170 }, { "epoch": 3.1502079177795057, "grad_norm": 3.488801956176758, "learning_rate": 5.9710726349173254e-05, "loss": 0.0569, "step": 93180 }, { "epoch": 3.1505459954697588, "grad_norm": 1.837222695350647, "learning_rate": 5.970261628530693e-05, "loss": 0.056, "step": 93190 }, { "epoch": 3.1508840731600123, "grad_norm": 1.471008062362671, "learning_rate": 5.969450595617472e-05, "loss": 0.0705, "step": 93200 }, { "epoch": 3.1512221508502654, "grad_norm": 1.5594265460968018, "learning_rate": 5.968639536199836e-05, "loss": 0.0576, "step": 93210 }, { "epoch": 3.1515602285405184, "grad_norm": 1.266489028930664, "learning_rate": 5.9678284502999596e-05, "loss": 0.0593, "step": 93220 }, { "epoch": 3.151898306230772, "grad_norm": 3.1093270778656006, "learning_rate": 5.967017337940016e-05, "loss": 0.0672, "step": 93230 }, { "epoch": 3.152236383921025, "grad_norm": 2.8688313961029053, "learning_rate": 5.966206199142182e-05, "loss": 0.0625, "step": 93240 }, { "epoch": 3.152574461611278, "grad_norm": 0.8257875442504883, "learning_rate": 5.9653950339286337e-05, "loss": 0.058, "step": 93250 }, { "epoch": 3.1529125393015316, "grad_norm": 2.2925708293914795, "learning_rate": 5.964583842321547e-05, "loss": 0.0662, "step": 93260 }, { "epoch": 3.1532506169917847, "grad_norm": 1.8909025192260742, "learning_rate": 5.963772624343101e-05, "loss": 0.0637, "step": 93270 }, { "epoch": 3.1535886946820377, "grad_norm": 2.1575326919555664, "learning_rate": 5.962961380015474e-05, "loss": 0.0618, "step": 93280 }, { "epoch": 3.1539267723722912, "grad_norm": 2.3730969429016113, "learning_rate": 5.9621501093608445e-05, "loss": 0.061, "step": 93290 }, { "epoch": 3.1542648500625443, "grad_norm": 2.5905532836914062, "learning_rate": 5.9613388124013926e-05, "loss": 0.0607, "step": 93300 }, { "epoch": 3.154602927752798, "grad_norm": 1.716593623161316, "learning_rate": 5.960527489159298e-05, "loss": 0.0634, "step": 93310 }, { "epoch": 3.154941005443051, "grad_norm": 1.275941252708435, "learning_rate": 5.9597161396567435e-05, "loss": 0.069, "step": 93320 }, { "epoch": 3.155279083133304, "grad_norm": 2.472234010696411, "learning_rate": 5.958904763915909e-05, "loss": 0.0608, "step": 93330 }, { "epoch": 3.1556171608235575, "grad_norm": 3.3450870513916016, "learning_rate": 5.9580933619589796e-05, "loss": 0.0633, "step": 93340 }, { "epoch": 3.1559552385138105, "grad_norm": 1.014845609664917, "learning_rate": 5.9572819338081375e-05, "loss": 0.063, "step": 93350 }, { "epoch": 3.1562933162040636, "grad_norm": 2.8776724338531494, "learning_rate": 5.956470479485568e-05, "loss": 0.0609, "step": 93360 }, { "epoch": 3.156631393894317, "grad_norm": 2.153233289718628, "learning_rate": 5.9556589990134524e-05, "loss": 0.0604, "step": 93370 }, { "epoch": 3.15696947158457, "grad_norm": 1.374650478363037, "learning_rate": 5.954847492413981e-05, "loss": 0.0663, "step": 93380 }, { "epoch": 3.1573075492748233, "grad_norm": 1.3559566736221313, "learning_rate": 5.9540359597093356e-05, "loss": 0.0593, "step": 93390 }, { "epoch": 3.157645626965077, "grad_norm": 0.7127299308776855, "learning_rate": 5.953224400921707e-05, "loss": 0.0645, "step": 93400 }, { "epoch": 3.15798370465533, "grad_norm": 2.1048898696899414, "learning_rate": 5.9524128160732814e-05, "loss": 0.063, "step": 93410 }, { "epoch": 3.158321782345583, "grad_norm": 1.9506291151046753, "learning_rate": 5.951601205186248e-05, "loss": 0.0658, "step": 93420 }, { "epoch": 3.1586598600358364, "grad_norm": 0.8655610680580139, "learning_rate": 5.950789568282794e-05, "loss": 0.0655, "step": 93430 }, { "epoch": 3.1589979377260895, "grad_norm": 1.5136620998382568, "learning_rate": 5.9499779053851114e-05, "loss": 0.064, "step": 93440 }, { "epoch": 3.1593360154163426, "grad_norm": 1.5472244024276733, "learning_rate": 5.9491662165153896e-05, "loss": 0.057, "step": 93450 }, { "epoch": 3.159674093106596, "grad_norm": 1.9806089401245117, "learning_rate": 5.9483545016958186e-05, "loss": 0.0638, "step": 93460 }, { "epoch": 3.160012170796849, "grad_norm": 2.1240298748016357, "learning_rate": 5.947542760948592e-05, "loss": 0.0576, "step": 93470 }, { "epoch": 3.160350248487102, "grad_norm": 3.2128732204437256, "learning_rate": 5.946730994295904e-05, "loss": 0.0627, "step": 93480 }, { "epoch": 3.1606883261773557, "grad_norm": 4.467906475067139, "learning_rate": 5.945919201759944e-05, "loss": 0.071, "step": 93490 }, { "epoch": 3.161026403867609, "grad_norm": 3.4157769680023193, "learning_rate": 5.94510738336291e-05, "loss": 0.0631, "step": 93500 }, { "epoch": 3.161364481557862, "grad_norm": 3.1800999641418457, "learning_rate": 5.944295539126995e-05, "loss": 0.0592, "step": 93510 }, { "epoch": 3.1617025592481154, "grad_norm": 2.3858442306518555, "learning_rate": 5.9434836690743944e-05, "loss": 0.061, "step": 93520 }, { "epoch": 3.1620406369383685, "grad_norm": 0.9423102140426636, "learning_rate": 5.942671773227305e-05, "loss": 0.0624, "step": 93530 }, { "epoch": 3.1623787146286215, "grad_norm": 3.7347607612609863, "learning_rate": 5.9418598516079226e-05, "loss": 0.0622, "step": 93540 }, { "epoch": 3.162716792318875, "grad_norm": 1.001185417175293, "learning_rate": 5.941047904238447e-05, "loss": 0.069, "step": 93550 }, { "epoch": 3.163054870009128, "grad_norm": 2.632382392883301, "learning_rate": 5.940235931141075e-05, "loss": 0.0665, "step": 93560 }, { "epoch": 3.163392947699381, "grad_norm": 3.1893727779388428, "learning_rate": 5.9394239323380064e-05, "loss": 0.0686, "step": 93570 }, { "epoch": 3.1637310253896347, "grad_norm": 1.257602334022522, "learning_rate": 5.938611907851438e-05, "loss": 0.0655, "step": 93580 }, { "epoch": 3.1640691030798878, "grad_norm": 1.3061171770095825, "learning_rate": 5.937799857703574e-05, "loss": 0.0632, "step": 93590 }, { "epoch": 3.164407180770141, "grad_norm": 3.2687366008758545, "learning_rate": 5.9369877819166145e-05, "loss": 0.0626, "step": 93600 }, { "epoch": 3.1647452584603943, "grad_norm": 1.0296778678894043, "learning_rate": 5.936175680512762e-05, "loss": 0.0623, "step": 93610 }, { "epoch": 3.1650833361506474, "grad_norm": 2.404789447784424, "learning_rate": 5.935363553514217e-05, "loss": 0.0679, "step": 93620 }, { "epoch": 3.1654214138409005, "grad_norm": 1.6063036918640137, "learning_rate": 5.934551400943185e-05, "loss": 0.0687, "step": 93630 }, { "epoch": 3.165759491531154, "grad_norm": 3.5702855587005615, "learning_rate": 5.9337392228218666e-05, "loss": 0.0665, "step": 93640 }, { "epoch": 3.166097569221407, "grad_norm": 1.672536849975586, "learning_rate": 5.9329270191724694e-05, "loss": 0.0577, "step": 93650 }, { "epoch": 3.16643564691166, "grad_norm": 3.7278811931610107, "learning_rate": 5.932114790017198e-05, "loss": 0.0658, "step": 93660 }, { "epoch": 3.1667737246019136, "grad_norm": 1.9586803913116455, "learning_rate": 5.931302535378259e-05, "loss": 0.0671, "step": 93670 }, { "epoch": 3.1671118022921667, "grad_norm": 2.024223804473877, "learning_rate": 5.930490255277859e-05, "loss": 0.0679, "step": 93680 }, { "epoch": 3.16744987998242, "grad_norm": 0.6667715907096863, "learning_rate": 5.929677949738205e-05, "loss": 0.0636, "step": 93690 }, { "epoch": 3.1677879576726733, "grad_norm": 1.5739926099777222, "learning_rate": 5.928865618781504e-05, "loss": 0.0603, "step": 93700 }, { "epoch": 3.1681260353629264, "grad_norm": 1.3190919160842896, "learning_rate": 5.9280532624299654e-05, "loss": 0.0529, "step": 93710 }, { "epoch": 3.1684641130531794, "grad_norm": 1.494966745376587, "learning_rate": 5.9272408807058e-05, "loss": 0.0614, "step": 93720 }, { "epoch": 3.168802190743433, "grad_norm": 1.4298864603042603, "learning_rate": 5.926428473631218e-05, "loss": 0.0614, "step": 93730 }, { "epoch": 3.169140268433686, "grad_norm": 0.9272691011428833, "learning_rate": 5.925616041228429e-05, "loss": 0.0672, "step": 93740 }, { "epoch": 3.169478346123939, "grad_norm": 1.9913023710250854, "learning_rate": 5.924803583519645e-05, "loss": 0.0687, "step": 93750 }, { "epoch": 3.1698164238141926, "grad_norm": 2.2373437881469727, "learning_rate": 5.92399110052708e-05, "loss": 0.0671, "step": 93760 }, { "epoch": 3.1701545015044457, "grad_norm": 1.3407988548278809, "learning_rate": 5.923178592272943e-05, "loss": 0.0691, "step": 93770 }, { "epoch": 3.1704925791946987, "grad_norm": 1.1880486011505127, "learning_rate": 5.9223660587794506e-05, "loss": 0.0617, "step": 93780 }, { "epoch": 3.1708306568849522, "grad_norm": 1.1340734958648682, "learning_rate": 5.921553500068817e-05, "loss": 0.0625, "step": 93790 }, { "epoch": 3.1711687345752053, "grad_norm": 1.7487928867340088, "learning_rate": 5.9207409161632585e-05, "loss": 0.0631, "step": 93800 }, { "epoch": 3.1715068122654584, "grad_norm": 2.6364660263061523, "learning_rate": 5.9199283070849875e-05, "loss": 0.0651, "step": 93810 }, { "epoch": 3.171844889955712, "grad_norm": 1.144649624824524, "learning_rate": 5.919115672856223e-05, "loss": 0.0634, "step": 93820 }, { "epoch": 3.172182967645965, "grad_norm": 2.7209112644195557, "learning_rate": 5.918303013499181e-05, "loss": 0.0684, "step": 93830 }, { "epoch": 3.172521045336218, "grad_norm": 1.1247268915176392, "learning_rate": 5.91749032903608e-05, "loss": 0.0616, "step": 93840 }, { "epoch": 3.1728591230264716, "grad_norm": 0.925990879535675, "learning_rate": 5.916677619489136e-05, "loss": 0.0593, "step": 93850 }, { "epoch": 3.1731972007167246, "grad_norm": 1.6088887453079224, "learning_rate": 5.915864884880573e-05, "loss": 0.0594, "step": 93860 }, { "epoch": 3.1735352784069777, "grad_norm": 2.4800219535827637, "learning_rate": 5.915052125232607e-05, "loss": 0.0661, "step": 93870 }, { "epoch": 3.173873356097231, "grad_norm": 0.9418466687202454, "learning_rate": 5.91423934056746e-05, "loss": 0.0606, "step": 93880 }, { "epoch": 3.1742114337874843, "grad_norm": 2.8997979164123535, "learning_rate": 5.9134265309073525e-05, "loss": 0.0609, "step": 93890 }, { "epoch": 3.174549511477738, "grad_norm": 1.2106956243515015, "learning_rate": 5.912613696274507e-05, "loss": 0.0699, "step": 93900 }, { "epoch": 3.174887589167991, "grad_norm": 1.1483947038650513, "learning_rate": 5.911800836691145e-05, "loss": 0.071, "step": 93910 }, { "epoch": 3.175225666858244, "grad_norm": 2.9290945529937744, "learning_rate": 5.910987952179492e-05, "loss": 0.055, "step": 93920 }, { "epoch": 3.1755637445484974, "grad_norm": 1.337178111076355, "learning_rate": 5.91017504276177e-05, "loss": 0.0664, "step": 93930 }, { "epoch": 3.1759018222387505, "grad_norm": 0.977120041847229, "learning_rate": 5.909362108460205e-05, "loss": 0.0622, "step": 93940 }, { "epoch": 3.1762398999290036, "grad_norm": 1.3348742723464966, "learning_rate": 5.9085491492970205e-05, "loss": 0.0612, "step": 93950 }, { "epoch": 3.176577977619257, "grad_norm": 1.2841421365737915, "learning_rate": 5.907736165294444e-05, "loss": 0.0675, "step": 93960 }, { "epoch": 3.17691605530951, "grad_norm": 1.0515419244766235, "learning_rate": 5.906923156474702e-05, "loss": 0.0664, "step": 93970 }, { "epoch": 3.1772541329997632, "grad_norm": 1.5340385437011719, "learning_rate": 5.906110122860021e-05, "loss": 0.0677, "step": 93980 }, { "epoch": 3.1775922106900167, "grad_norm": 0.9538862705230713, "learning_rate": 5.90529706447263e-05, "loss": 0.0605, "step": 93990 }, { "epoch": 3.17793028838027, "grad_norm": 1.8865079879760742, "learning_rate": 5.904483981334757e-05, "loss": 0.062, "step": 94000 }, { "epoch": 3.178268366070523, "grad_norm": 1.2048721313476562, "learning_rate": 5.903670873468633e-05, "loss": 0.0568, "step": 94010 }, { "epoch": 3.1786064437607764, "grad_norm": 1.7154247760772705, "learning_rate": 5.902857740896486e-05, "loss": 0.0594, "step": 94020 }, { "epoch": 3.1789445214510295, "grad_norm": 0.9344229698181152, "learning_rate": 5.902044583640547e-05, "loss": 0.0658, "step": 94030 }, { "epoch": 3.1792825991412825, "grad_norm": 0.744956374168396, "learning_rate": 5.901231401723049e-05, "loss": 0.0647, "step": 94040 }, { "epoch": 3.179620676831536, "grad_norm": 0.7583875060081482, "learning_rate": 5.900418195166223e-05, "loss": 0.0665, "step": 94050 }, { "epoch": 3.179958754521789, "grad_norm": 1.86846923828125, "learning_rate": 5.8996049639923e-05, "loss": 0.0662, "step": 94060 }, { "epoch": 3.180296832212042, "grad_norm": 0.8462681174278259, "learning_rate": 5.898791708223518e-05, "loss": 0.0576, "step": 94070 }, { "epoch": 3.1806349099022957, "grad_norm": 2.7303149700164795, "learning_rate": 5.897978427882107e-05, "loss": 0.0636, "step": 94080 }, { "epoch": 3.1809729875925488, "grad_norm": 0.9053163528442383, "learning_rate": 5.897165122990303e-05, "loss": 0.0603, "step": 94090 }, { "epoch": 3.181311065282802, "grad_norm": 1.015401005744934, "learning_rate": 5.8963517935703404e-05, "loss": 0.0616, "step": 94100 }, { "epoch": 3.1816491429730553, "grad_norm": 1.1327446699142456, "learning_rate": 5.8955384396444593e-05, "loss": 0.0647, "step": 94110 }, { "epoch": 3.1819872206633084, "grad_norm": 0.9080742001533508, "learning_rate": 5.894725061234892e-05, "loss": 0.0567, "step": 94120 }, { "epoch": 3.1823252983535615, "grad_norm": 0.8964744210243225, "learning_rate": 5.893911658363878e-05, "loss": 0.0626, "step": 94130 }, { "epoch": 3.182663376043815, "grad_norm": 0.7677438259124756, "learning_rate": 5.8930982310536556e-05, "loss": 0.0561, "step": 94140 }, { "epoch": 3.183001453734068, "grad_norm": 1.5139895677566528, "learning_rate": 5.8922847793264624e-05, "loss": 0.0606, "step": 94150 }, { "epoch": 3.183339531424321, "grad_norm": 1.3915860652923584, "learning_rate": 5.8914713032045366e-05, "loss": 0.06, "step": 94160 }, { "epoch": 3.1836776091145746, "grad_norm": 1.9839222431182861, "learning_rate": 5.8906578027101235e-05, "loss": 0.0649, "step": 94170 }, { "epoch": 3.1840156868048277, "grad_norm": 0.9015912413597107, "learning_rate": 5.889844277865459e-05, "loss": 0.0653, "step": 94180 }, { "epoch": 3.184353764495081, "grad_norm": 1.436361312866211, "learning_rate": 5.8890307286927884e-05, "loss": 0.0674, "step": 94190 }, { "epoch": 3.1846918421853343, "grad_norm": 1.2792786359786987, "learning_rate": 5.8882171552143486e-05, "loss": 0.067, "step": 94200 }, { "epoch": 3.1850299198755874, "grad_norm": 1.1595135927200317, "learning_rate": 5.887403557452387e-05, "loss": 0.0665, "step": 94210 }, { "epoch": 3.1853679975658404, "grad_norm": 0.9749572277069092, "learning_rate": 5.8865899354291445e-05, "loss": 0.0601, "step": 94220 }, { "epoch": 3.185706075256094, "grad_norm": 0.9978528618812561, "learning_rate": 5.885776289166868e-05, "loss": 0.0578, "step": 94230 }, { "epoch": 3.186044152946347, "grad_norm": 1.0127952098846436, "learning_rate": 5.884962618687799e-05, "loss": 0.0591, "step": 94240 }, { "epoch": 3.1863822306366, "grad_norm": 1.410111904144287, "learning_rate": 5.8841489240141856e-05, "loss": 0.0683, "step": 94250 }, { "epoch": 3.1867203083268536, "grad_norm": 1.0082179307937622, "learning_rate": 5.8833352051682725e-05, "loss": 0.0638, "step": 94260 }, { "epoch": 3.1870583860171067, "grad_norm": 0.8240951299667358, "learning_rate": 5.8825214621723056e-05, "loss": 0.0641, "step": 94270 }, { "epoch": 3.18739646370736, "grad_norm": 0.5445860028266907, "learning_rate": 5.881707695048535e-05, "loss": 0.0649, "step": 94280 }, { "epoch": 3.1877345413976133, "grad_norm": 3.7330071926116943, "learning_rate": 5.880893903819206e-05, "loss": 0.0641, "step": 94290 }, { "epoch": 3.1880726190878663, "grad_norm": 0.5367873907089233, "learning_rate": 5.88008008850657e-05, "loss": 0.06, "step": 94300 }, { "epoch": 3.18841069677812, "grad_norm": 1.4367643594741821, "learning_rate": 5.879266249132873e-05, "loss": 0.0671, "step": 94310 }, { "epoch": 3.188748774468373, "grad_norm": 0.5473456382751465, "learning_rate": 5.87845238572037e-05, "loss": 0.0674, "step": 94320 }, { "epoch": 3.189086852158626, "grad_norm": 1.0223698616027832, "learning_rate": 5.877638498291307e-05, "loss": 0.0622, "step": 94330 }, { "epoch": 3.1894249298488795, "grad_norm": 0.6411276459693909, "learning_rate": 5.876824586867937e-05, "loss": 0.0624, "step": 94340 }, { "epoch": 3.1897630075391326, "grad_norm": 0.9655399322509766, "learning_rate": 5.876010651472512e-05, "loss": 0.0636, "step": 94350 }, { "epoch": 3.1901010852293856, "grad_norm": 0.9365026950836182, "learning_rate": 5.875196692127286e-05, "loss": 0.0601, "step": 94360 }, { "epoch": 3.190439162919639, "grad_norm": 0.7289236783981323, "learning_rate": 5.87438270885451e-05, "loss": 0.0535, "step": 94370 }, { "epoch": 3.190777240609892, "grad_norm": 0.7191102504730225, "learning_rate": 5.873568701676441e-05, "loss": 0.0632, "step": 94380 }, { "epoch": 3.1911153183001453, "grad_norm": 1.2197450399398804, "learning_rate": 5.8727546706153305e-05, "loss": 0.0641, "step": 94390 }, { "epoch": 3.191453395990399, "grad_norm": 0.6999905109405518, "learning_rate": 5.8719406156934357e-05, "loss": 0.0622, "step": 94400 }, { "epoch": 3.191791473680652, "grad_norm": 1.066542387008667, "learning_rate": 5.87112653693301e-05, "loss": 0.0593, "step": 94410 }, { "epoch": 3.192129551370905, "grad_norm": 0.9739336371421814, "learning_rate": 5.870312434356315e-05, "loss": 0.0651, "step": 94420 }, { "epoch": 3.1924676290611584, "grad_norm": 0.7497197389602661, "learning_rate": 5.869498307985604e-05, "loss": 0.0605, "step": 94430 }, { "epoch": 3.1928057067514115, "grad_norm": 0.7849961519241333, "learning_rate": 5.868684157843136e-05, "loss": 0.0598, "step": 94440 }, { "epoch": 3.1931437844416646, "grad_norm": 2.0883548259735107, "learning_rate": 5.867869983951168e-05, "loss": 0.0656, "step": 94450 }, { "epoch": 3.193481862131918, "grad_norm": 2.042407274246216, "learning_rate": 5.867055786331962e-05, "loss": 0.0609, "step": 94460 }, { "epoch": 3.193819939822171, "grad_norm": 0.9951142072677612, "learning_rate": 5.866241565007776e-05, "loss": 0.0578, "step": 94470 }, { "epoch": 3.1941580175124242, "grad_norm": 1.3957239389419556, "learning_rate": 5.865427320000872e-05, "loss": 0.0632, "step": 94480 }, { "epoch": 3.1944960952026777, "grad_norm": 1.5591126680374146, "learning_rate": 5.86461305133351e-05, "loss": 0.0613, "step": 94490 }, { "epoch": 3.194834172892931, "grad_norm": 1.105820894241333, "learning_rate": 5.863798759027951e-05, "loss": 0.0588, "step": 94500 }, { "epoch": 3.195172250583184, "grad_norm": 1.4381208419799805, "learning_rate": 5.862984443106461e-05, "loss": 0.0619, "step": 94510 }, { "epoch": 3.1955103282734374, "grad_norm": 1.0608052015304565, "learning_rate": 5.862170103591298e-05, "loss": 0.0614, "step": 94520 }, { "epoch": 3.1958484059636905, "grad_norm": 0.9315150380134583, "learning_rate": 5.861355740504729e-05, "loss": 0.0608, "step": 94530 }, { "epoch": 3.1961864836539435, "grad_norm": 1.3264479637145996, "learning_rate": 5.8605413538690154e-05, "loss": 0.0629, "step": 94540 }, { "epoch": 3.196524561344197, "grad_norm": 0.8459938168525696, "learning_rate": 5.859726943706428e-05, "loss": 0.0603, "step": 94550 }, { "epoch": 3.19686263903445, "grad_norm": 0.7627761960029602, "learning_rate": 5.8589125100392276e-05, "loss": 0.0645, "step": 94560 }, { "epoch": 3.197200716724703, "grad_norm": 1.6947267055511475, "learning_rate": 5.8580980528896824e-05, "loss": 0.0612, "step": 94570 }, { "epoch": 3.1975387944149567, "grad_norm": 0.6976479291915894, "learning_rate": 5.857283572280058e-05, "loss": 0.0591, "step": 94580 }, { "epoch": 3.1978768721052098, "grad_norm": 0.90023273229599, "learning_rate": 5.8564690682326237e-05, "loss": 0.06, "step": 94590 }, { "epoch": 3.198214949795463, "grad_norm": 1.5500378608703613, "learning_rate": 5.855654540769645e-05, "loss": 0.0636, "step": 94600 }, { "epoch": 3.1985530274857163, "grad_norm": 2.386444330215454, "learning_rate": 5.8548399899133944e-05, "loss": 0.0637, "step": 94610 }, { "epoch": 3.1988911051759694, "grad_norm": 1.3770451545715332, "learning_rate": 5.854025415686139e-05, "loss": 0.0623, "step": 94620 }, { "epoch": 3.1992291828662225, "grad_norm": 0.8898617625236511, "learning_rate": 5.8532108181101496e-05, "loss": 0.0597, "step": 94630 }, { "epoch": 3.199567260556476, "grad_norm": 0.6751539707183838, "learning_rate": 5.852396197207697e-05, "loss": 0.0574, "step": 94640 }, { "epoch": 3.199905338246729, "grad_norm": 1.1487255096435547, "learning_rate": 5.851581553001053e-05, "loss": 0.0584, "step": 94650 }, { "epoch": 3.200243415936982, "grad_norm": 1.3689132928848267, "learning_rate": 5.8507668855124885e-05, "loss": 0.0587, "step": 94660 }, { "epoch": 3.2005814936272357, "grad_norm": 0.7055611610412598, "learning_rate": 5.8499521947642764e-05, "loss": 0.0579, "step": 94670 }, { "epoch": 3.2009195713174887, "grad_norm": 0.671643078327179, "learning_rate": 5.849137480778691e-05, "loss": 0.0587, "step": 94680 }, { "epoch": 3.201257649007742, "grad_norm": 0.6248870491981506, "learning_rate": 5.848322743578008e-05, "loss": 0.0669, "step": 94690 }, { "epoch": 3.2015957266979953, "grad_norm": 1.045222282409668, "learning_rate": 5.8475079831844984e-05, "loss": 0.0702, "step": 94700 }, { "epoch": 3.2019338043882484, "grad_norm": 0.8842941522598267, "learning_rate": 5.846693199620438e-05, "loss": 0.0623, "step": 94710 }, { "epoch": 3.2022718820785014, "grad_norm": 0.5924085974693298, "learning_rate": 5.845878392908105e-05, "loss": 0.0566, "step": 94720 }, { "epoch": 3.202609959768755, "grad_norm": 0.9051300883293152, "learning_rate": 5.845063563069774e-05, "loss": 0.0613, "step": 94730 }, { "epoch": 3.202948037459008, "grad_norm": 0.8751972913742065, "learning_rate": 5.844248710127722e-05, "loss": 0.0639, "step": 94740 }, { "epoch": 3.203286115149261, "grad_norm": 0.8393830060958862, "learning_rate": 5.843433834104227e-05, "loss": 0.0598, "step": 94750 }, { "epoch": 3.2036241928395146, "grad_norm": 0.6561406850814819, "learning_rate": 5.8426189350215696e-05, "loss": 0.0597, "step": 94760 }, { "epoch": 3.2039622705297677, "grad_norm": 6.2687458992004395, "learning_rate": 5.841804012902027e-05, "loss": 0.0557, "step": 94770 }, { "epoch": 3.2043003482200207, "grad_norm": 1.542360544204712, "learning_rate": 5.840989067767878e-05, "loss": 0.0563, "step": 94780 }, { "epoch": 3.2046384259102743, "grad_norm": 2.003201484680176, "learning_rate": 5.8401740996414034e-05, "loss": 0.0582, "step": 94790 }, { "epoch": 3.2049765036005273, "grad_norm": 0.731541097164154, "learning_rate": 5.8393591085448865e-05, "loss": 0.06, "step": 94800 }, { "epoch": 3.2053145812907804, "grad_norm": 7.277970314025879, "learning_rate": 5.8385440945006044e-05, "loss": 0.072, "step": 94810 }, { "epoch": 3.205652658981034, "grad_norm": 0.8958252668380737, "learning_rate": 5.8377290575308444e-05, "loss": 0.0653, "step": 94820 }, { "epoch": 3.205990736671287, "grad_norm": 0.6585374474525452, "learning_rate": 5.836913997657885e-05, "loss": 0.063, "step": 94830 }, { "epoch": 3.20632881436154, "grad_norm": 3.606546401977539, "learning_rate": 5.8360989149040125e-05, "loss": 0.0647, "step": 94840 }, { "epoch": 3.2066668920517936, "grad_norm": 0.5907716155052185, "learning_rate": 5.835283809291509e-05, "loss": 0.0577, "step": 94850 }, { "epoch": 3.2070049697420466, "grad_norm": 0.8647741675376892, "learning_rate": 5.834468680842659e-05, "loss": 0.056, "step": 94860 }, { "epoch": 3.2073430474323, "grad_norm": 0.8461915254592896, "learning_rate": 5.833653529579749e-05, "loss": 0.0628, "step": 94870 }, { "epoch": 3.207681125122553, "grad_norm": 1.1757439374923706, "learning_rate": 5.832838355525065e-05, "loss": 0.0664, "step": 94880 }, { "epoch": 3.2080192028128063, "grad_norm": 1.147413969039917, "learning_rate": 5.8320231587008946e-05, "loss": 0.0572, "step": 94890 }, { "epoch": 3.20835728050306, "grad_norm": 1.3831064701080322, "learning_rate": 5.831207939129523e-05, "loss": 0.0659, "step": 94900 }, { "epoch": 3.208695358193313, "grad_norm": 0.8523349761962891, "learning_rate": 5.8303926968332375e-05, "loss": 0.0593, "step": 94910 }, { "epoch": 3.209033435883566, "grad_norm": 1.0363779067993164, "learning_rate": 5.829577431834328e-05, "loss": 0.0662, "step": 94920 }, { "epoch": 3.2093715135738194, "grad_norm": 1.0462267398834229, "learning_rate": 5.828762144155083e-05, "loss": 0.0603, "step": 94930 }, { "epoch": 3.2097095912640725, "grad_norm": 1.2316064834594727, "learning_rate": 5.827946833817795e-05, "loss": 0.0627, "step": 94940 }, { "epoch": 3.2100476689543256, "grad_norm": 1.17412269115448, "learning_rate": 5.827131500844748e-05, "loss": 0.0626, "step": 94950 }, { "epoch": 3.210385746644579, "grad_norm": 1.59730863571167, "learning_rate": 5.8263161452582384e-05, "loss": 0.0612, "step": 94960 }, { "epoch": 3.210723824334832, "grad_norm": 0.5992162823677063, "learning_rate": 5.8255007670805564e-05, "loss": 0.0573, "step": 94970 }, { "epoch": 3.2110619020250852, "grad_norm": 2.822506904602051, "learning_rate": 5.824685366333992e-05, "loss": 0.056, "step": 94980 }, { "epoch": 3.2113999797153387, "grad_norm": 1.091239094734192, "learning_rate": 5.8238699430408394e-05, "loss": 0.058, "step": 94990 }, { "epoch": 3.211738057405592, "grad_norm": 0.646369457244873, "learning_rate": 5.823054497223392e-05, "loss": 0.0636, "step": 95000 }, { "epoch": 3.212076135095845, "grad_norm": 1.1382791996002197, "learning_rate": 5.8222390289039456e-05, "loss": 0.0594, "step": 95010 }, { "epoch": 3.2124142127860984, "grad_norm": 2.209652900695801, "learning_rate": 5.8214235381047914e-05, "loss": 0.0596, "step": 95020 }, { "epoch": 3.2127522904763515, "grad_norm": 0.601723313331604, "learning_rate": 5.8206080248482275e-05, "loss": 0.0606, "step": 95030 }, { "epoch": 3.2130903681666045, "grad_norm": 1.3746254444122314, "learning_rate": 5.819792489156548e-05, "loss": 0.0645, "step": 95040 }, { "epoch": 3.213428445856858, "grad_norm": 0.8649618029594421, "learning_rate": 5.8189769310520495e-05, "loss": 0.0571, "step": 95050 }, { "epoch": 3.213766523547111, "grad_norm": 0.9560039043426514, "learning_rate": 5.8181613505570284e-05, "loss": 0.0602, "step": 95060 }, { "epoch": 3.214104601237364, "grad_norm": 2.286437511444092, "learning_rate": 5.8173457476937856e-05, "loss": 0.0593, "step": 95070 }, { "epoch": 3.2144426789276177, "grad_norm": 0.7502277493476868, "learning_rate": 5.8165301224846156e-05, "loss": 0.0576, "step": 95080 }, { "epoch": 3.2147807566178708, "grad_norm": 0.8084967136383057, "learning_rate": 5.81571447495182e-05, "loss": 0.0594, "step": 95090 }, { "epoch": 3.215118834308124, "grad_norm": 2.949042797088623, "learning_rate": 5.8148988051176954e-05, "loss": 0.058, "step": 95100 }, { "epoch": 3.2154569119983774, "grad_norm": 1.641026496887207, "learning_rate": 5.8140831130045445e-05, "loss": 0.0622, "step": 95110 }, { "epoch": 3.2157949896886304, "grad_norm": 0.959877610206604, "learning_rate": 5.813267398634668e-05, "loss": 0.062, "step": 95120 }, { "epoch": 3.2161330673788835, "grad_norm": 0.9769473075866699, "learning_rate": 5.812451662030366e-05, "loss": 0.0623, "step": 95130 }, { "epoch": 3.216471145069137, "grad_norm": 0.8202447891235352, "learning_rate": 5.811635903213939e-05, "loss": 0.0616, "step": 95140 }, { "epoch": 3.21680922275939, "grad_norm": 1.3093233108520508, "learning_rate": 5.8108201222076944e-05, "loss": 0.0659, "step": 95150 }, { "epoch": 3.217147300449643, "grad_norm": 1.208803415298462, "learning_rate": 5.8100043190339295e-05, "loss": 0.0716, "step": 95160 }, { "epoch": 3.2174853781398967, "grad_norm": 1.0276062488555908, "learning_rate": 5.809188493714952e-05, "loss": 0.0651, "step": 95170 }, { "epoch": 3.2178234558301497, "grad_norm": 1.4289870262145996, "learning_rate": 5.8083726462730635e-05, "loss": 0.0571, "step": 95180 }, { "epoch": 3.218161533520403, "grad_norm": 0.9298065900802612, "learning_rate": 5.807556776730574e-05, "loss": 0.0595, "step": 95190 }, { "epoch": 3.2184996112106563, "grad_norm": 0.9373744130134583, "learning_rate": 5.8067408851097825e-05, "loss": 0.0659, "step": 95200 }, { "epoch": 3.2188376889009094, "grad_norm": 1.3416533470153809, "learning_rate": 5.805924971432999e-05, "loss": 0.0638, "step": 95210 }, { "epoch": 3.2191757665911624, "grad_norm": 3.2797772884368896, "learning_rate": 5.8051090357225314e-05, "loss": 0.0612, "step": 95220 }, { "epoch": 3.219513844281416, "grad_norm": 1.0838981866836548, "learning_rate": 5.804293078000683e-05, "loss": 0.0621, "step": 95230 }, { "epoch": 3.219851921971669, "grad_norm": 0.996858537197113, "learning_rate": 5.803477098289765e-05, "loss": 0.0694, "step": 95240 }, { "epoch": 3.2201899996619225, "grad_norm": 1.0502049922943115, "learning_rate": 5.802661096612083e-05, "loss": 0.0639, "step": 95250 }, { "epoch": 3.2205280773521756, "grad_norm": 0.8515275120735168, "learning_rate": 5.801845072989951e-05, "loss": 0.0629, "step": 95260 }, { "epoch": 3.2208661550424287, "grad_norm": 2.1720786094665527, "learning_rate": 5.801029027445675e-05, "loss": 0.0633, "step": 95270 }, { "epoch": 3.221204232732682, "grad_norm": 1.2150758504867554, "learning_rate": 5.800212960001567e-05, "loss": 0.0611, "step": 95280 }, { "epoch": 3.2215423104229353, "grad_norm": 0.8866048455238342, "learning_rate": 5.7993968706799365e-05, "loss": 0.0643, "step": 95290 }, { "epoch": 3.2218803881131883, "grad_norm": 1.6261838674545288, "learning_rate": 5.798580759503095e-05, "loss": 0.0683, "step": 95300 }, { "epoch": 3.222218465803442, "grad_norm": 3.387711524963379, "learning_rate": 5.797764626493356e-05, "loss": 0.0588, "step": 95310 }, { "epoch": 3.222556543493695, "grad_norm": 1.1433390378952026, "learning_rate": 5.796948471673033e-05, "loss": 0.0665, "step": 95320 }, { "epoch": 3.222894621183948, "grad_norm": 1.4126802682876587, "learning_rate": 5.796132295064437e-05, "loss": 0.0607, "step": 95330 }, { "epoch": 3.2232326988742015, "grad_norm": 3.761636972427368, "learning_rate": 5.7953160966898846e-05, "loss": 0.0628, "step": 95340 }, { "epoch": 3.2235707765644546, "grad_norm": 1.5146042108535767, "learning_rate": 5.7944998765716866e-05, "loss": 0.0601, "step": 95350 }, { "epoch": 3.2239088542547076, "grad_norm": 7.4082536697387695, "learning_rate": 5.7936836347321634e-05, "loss": 0.0639, "step": 95360 }, { "epoch": 3.224246931944961, "grad_norm": 1.148462176322937, "learning_rate": 5.7928673711936247e-05, "loss": 0.0625, "step": 95370 }, { "epoch": 3.224585009635214, "grad_norm": 1.3229161500930786, "learning_rate": 5.792051085978391e-05, "loss": 0.0624, "step": 95380 }, { "epoch": 3.2249230873254673, "grad_norm": 0.7383999824523926, "learning_rate": 5.791234779108778e-05, "loss": 0.0629, "step": 95390 }, { "epoch": 3.225261165015721, "grad_norm": 0.7862010598182678, "learning_rate": 5.7904184506071046e-05, "loss": 0.0582, "step": 95400 }, { "epoch": 3.225599242705974, "grad_norm": 1.3529688119888306, "learning_rate": 5.789602100495687e-05, "loss": 0.0679, "step": 95410 }, { "epoch": 3.225937320396227, "grad_norm": 0.7031984329223633, "learning_rate": 5.788785728796844e-05, "loss": 0.0542, "step": 95420 }, { "epoch": 3.2262753980864805, "grad_norm": 2.3250770568847656, "learning_rate": 5.7879693355328954e-05, "loss": 0.0581, "step": 95430 }, { "epoch": 3.2266134757767335, "grad_norm": 2.1777405738830566, "learning_rate": 5.787152920726163e-05, "loss": 0.0649, "step": 95440 }, { "epoch": 3.2269515534669866, "grad_norm": 1.5892245769500732, "learning_rate": 5.786336484398964e-05, "loss": 0.0622, "step": 95450 }, { "epoch": 3.22728963115724, "grad_norm": 1.1664111614227295, "learning_rate": 5.785520026573621e-05, "loss": 0.06, "step": 95460 }, { "epoch": 3.227627708847493, "grad_norm": 3.220127582550049, "learning_rate": 5.784703547272457e-05, "loss": 0.0635, "step": 95470 }, { "epoch": 3.2279657865377462, "grad_norm": 1.1125226020812988, "learning_rate": 5.783887046517791e-05, "loss": 0.0607, "step": 95480 }, { "epoch": 3.2283038642279998, "grad_norm": 2.658057689666748, "learning_rate": 5.7830705243319486e-05, "loss": 0.0608, "step": 95490 }, { "epoch": 3.228641941918253, "grad_norm": 0.9344154000282288, "learning_rate": 5.782253980737251e-05, "loss": 0.0561, "step": 95500 }, { "epoch": 3.228980019608506, "grad_norm": 1.7024537324905396, "learning_rate": 5.781437415756026e-05, "loss": 0.0647, "step": 95510 }, { "epoch": 3.2293180972987594, "grad_norm": 2.9492034912109375, "learning_rate": 5.780620829410594e-05, "loss": 0.0607, "step": 95520 }, { "epoch": 3.2296561749890125, "grad_norm": 1.5346399545669556, "learning_rate": 5.779804221723282e-05, "loss": 0.0688, "step": 95530 }, { "epoch": 3.2299942526792655, "grad_norm": 8.171263694763184, "learning_rate": 5.7789875927164147e-05, "loss": 0.0665, "step": 95540 }, { "epoch": 3.230332330369519, "grad_norm": 1.3605899810791016, "learning_rate": 5.778170942412321e-05, "loss": 0.0624, "step": 95550 }, { "epoch": 3.230670408059772, "grad_norm": 2.027716636657715, "learning_rate": 5.777354270833324e-05, "loss": 0.0592, "step": 95560 }, { "epoch": 3.231008485750025, "grad_norm": 4.109302997589111, "learning_rate": 5.776537578001755e-05, "loss": 0.0602, "step": 95570 }, { "epoch": 3.2313465634402787, "grad_norm": 1.8005956411361694, "learning_rate": 5.775720863939939e-05, "loss": 0.0662, "step": 95580 }, { "epoch": 3.2316846411305318, "grad_norm": 3.783994436264038, "learning_rate": 5.774904128670206e-05, "loss": 0.0559, "step": 95590 }, { "epoch": 3.232022718820785, "grad_norm": 24.68793296813965, "learning_rate": 5.774087372214887e-05, "loss": 0.0575, "step": 95600 }, { "epoch": 3.2323607965110384, "grad_norm": 3.1456546783447266, "learning_rate": 5.773270594596309e-05, "loss": 0.0599, "step": 95610 }, { "epoch": 3.2326988742012914, "grad_norm": 1.8132065534591675, "learning_rate": 5.772453795836803e-05, "loss": 0.0666, "step": 95620 }, { "epoch": 3.2330369518915445, "grad_norm": 1.630898356437683, "learning_rate": 5.771636975958701e-05, "loss": 0.0645, "step": 95630 }, { "epoch": 3.233375029581798, "grad_norm": 3.787104368209839, "learning_rate": 5.770820134984334e-05, "loss": 0.0645, "step": 95640 }, { "epoch": 3.233713107272051, "grad_norm": 6.4312896728515625, "learning_rate": 5.7700032729360346e-05, "loss": 0.07, "step": 95650 }, { "epoch": 3.234051184962304, "grad_norm": 4.745448112487793, "learning_rate": 5.769186389836133e-05, "loss": 0.0645, "step": 95660 }, { "epoch": 3.2343892626525577, "grad_norm": 2.620290756225586, "learning_rate": 5.768369485706966e-05, "loss": 0.06, "step": 95670 }, { "epoch": 3.2347273403428107, "grad_norm": 2.545166492462158, "learning_rate": 5.767552560570865e-05, "loss": 0.0564, "step": 95680 }, { "epoch": 3.235065418033064, "grad_norm": 11.713826179504395, "learning_rate": 5.7667356144501675e-05, "loss": 0.0669, "step": 95690 }, { "epoch": 3.2354034957233173, "grad_norm": 2.6314799785614014, "learning_rate": 5.765918647367203e-05, "loss": 0.0633, "step": 95700 }, { "epoch": 3.2357415734135704, "grad_norm": 2.4309213161468506, "learning_rate": 5.765101659344313e-05, "loss": 0.071, "step": 95710 }, { "epoch": 3.2360796511038235, "grad_norm": 2.350098133087158, "learning_rate": 5.764284650403831e-05, "loss": 0.065, "step": 95720 }, { "epoch": 3.236417728794077, "grad_norm": 3.6415317058563232, "learning_rate": 5.7634676205680924e-05, "loss": 0.0676, "step": 95730 }, { "epoch": 3.23675580648433, "grad_norm": 2.28572678565979, "learning_rate": 5.762650569859438e-05, "loss": 0.0694, "step": 95740 }, { "epoch": 3.237093884174583, "grad_norm": 3.3288168907165527, "learning_rate": 5.7618334983002006e-05, "loss": 0.0744, "step": 95750 }, { "epoch": 3.2374319618648366, "grad_norm": 1.4737329483032227, "learning_rate": 5.761016405912724e-05, "loss": 0.0685, "step": 95760 }, { "epoch": 3.2377700395550897, "grad_norm": 2.2779221534729004, "learning_rate": 5.7601992927193435e-05, "loss": 0.061, "step": 95770 }, { "epoch": 3.2381081172453428, "grad_norm": 3.888794422149658, "learning_rate": 5.759382158742401e-05, "loss": 0.0697, "step": 95780 }, { "epoch": 3.2384461949355963, "grad_norm": 1.915069580078125, "learning_rate": 5.758565004004235e-05, "loss": 0.07, "step": 95790 }, { "epoch": 3.2387842726258493, "grad_norm": 10.727027893066406, "learning_rate": 5.757747828527187e-05, "loss": 0.0643, "step": 95800 }, { "epoch": 3.2391223503161024, "grad_norm": 2.8699114322662354, "learning_rate": 5.756930632333596e-05, "loss": 0.0643, "step": 95810 }, { "epoch": 3.239460428006356, "grad_norm": 3.1592066287994385, "learning_rate": 5.7561134154458084e-05, "loss": 0.0617, "step": 95820 }, { "epoch": 3.239798505696609, "grad_norm": 3.0991716384887695, "learning_rate": 5.7552961778861624e-05, "loss": 0.0596, "step": 95830 }, { "epoch": 3.2401365833868625, "grad_norm": 2.2728004455566406, "learning_rate": 5.7544789196770045e-05, "loss": 0.0699, "step": 95840 }, { "epoch": 3.2404746610771156, "grad_norm": 4.009626388549805, "learning_rate": 5.753661640840675e-05, "loss": 0.0691, "step": 95850 }, { "epoch": 3.2408127387673686, "grad_norm": 3.3359029293060303, "learning_rate": 5.75284434139952e-05, "loss": 0.069, "step": 95860 }, { "epoch": 3.241150816457622, "grad_norm": 2.1368892192840576, "learning_rate": 5.752027021375883e-05, "loss": 0.0717, "step": 95870 }, { "epoch": 3.241488894147875, "grad_norm": 1.999482274055481, "learning_rate": 5.75120968079211e-05, "loss": 0.0676, "step": 95880 }, { "epoch": 3.2418269718381283, "grad_norm": 1.5880955457687378, "learning_rate": 5.7503923196705465e-05, "loss": 0.0765, "step": 95890 }, { "epoch": 3.242165049528382, "grad_norm": 3.689410924911499, "learning_rate": 5.74957493803354e-05, "loss": 0.0651, "step": 95900 }, { "epoch": 3.242503127218635, "grad_norm": 3.3537259101867676, "learning_rate": 5.748757535903435e-05, "loss": 0.0578, "step": 95910 }, { "epoch": 3.242841204908888, "grad_norm": 1.539813756942749, "learning_rate": 5.74794011330258e-05, "loss": 0.0688, "step": 95920 }, { "epoch": 3.2431792825991415, "grad_norm": 1.3135875463485718, "learning_rate": 5.7471226702533244e-05, "loss": 0.0621, "step": 95930 }, { "epoch": 3.2435173602893945, "grad_norm": 2.147900342941284, "learning_rate": 5.746305206778016e-05, "loss": 0.0633, "step": 95940 }, { "epoch": 3.2438554379796476, "grad_norm": 1.4040461778640747, "learning_rate": 5.745487722899001e-05, "loss": 0.0611, "step": 95950 }, { "epoch": 3.244193515669901, "grad_norm": 1.475217342376709, "learning_rate": 5.744670218638634e-05, "loss": 0.0594, "step": 95960 }, { "epoch": 3.244531593360154, "grad_norm": 1.5602377653121948, "learning_rate": 5.7438526940192626e-05, "loss": 0.0604, "step": 95970 }, { "epoch": 3.2448696710504072, "grad_norm": 1.8678160905838013, "learning_rate": 5.7430351490632374e-05, "loss": 0.0607, "step": 95980 }, { "epoch": 3.2452077487406608, "grad_norm": 1.1749186515808105, "learning_rate": 5.742217583792912e-05, "loss": 0.0579, "step": 95990 }, { "epoch": 3.245545826430914, "grad_norm": 1.002355694770813, "learning_rate": 5.7413999982306345e-05, "loss": 0.0659, "step": 96000 }, { "epoch": 3.245883904121167, "grad_norm": 1.3920671939849854, "learning_rate": 5.740582392398761e-05, "loss": 0.0585, "step": 96010 }, { "epoch": 3.2462219818114204, "grad_norm": 1.5348734855651855, "learning_rate": 5.7397647663196416e-05, "loss": 0.0624, "step": 96020 }, { "epoch": 3.2465600595016735, "grad_norm": 1.2302194833755493, "learning_rate": 5.738947120015634e-05, "loss": 0.0594, "step": 96030 }, { "epoch": 3.2468981371919265, "grad_norm": 0.8923642635345459, "learning_rate": 5.738129453509088e-05, "loss": 0.0574, "step": 96040 }, { "epoch": 3.24723621488218, "grad_norm": 2.0591089725494385, "learning_rate": 5.737311766822361e-05, "loss": 0.0593, "step": 96050 }, { "epoch": 3.247574292572433, "grad_norm": 3.113462209701538, "learning_rate": 5.736494059977806e-05, "loss": 0.0637, "step": 96060 }, { "epoch": 3.247912370262686, "grad_norm": 1.2139830589294434, "learning_rate": 5.7356763329977804e-05, "loss": 0.0628, "step": 96070 }, { "epoch": 3.2482504479529397, "grad_norm": 2.293139696121216, "learning_rate": 5.73485858590464e-05, "loss": 0.066, "step": 96080 }, { "epoch": 3.248588525643193, "grad_norm": 1.5748604536056519, "learning_rate": 5.7340408187207427e-05, "loss": 0.0647, "step": 96090 }, { "epoch": 3.248926603333446, "grad_norm": 1.7101129293441772, "learning_rate": 5.733223031468445e-05, "loss": 0.0673, "step": 96100 }, { "epoch": 3.2492646810236994, "grad_norm": 5.210004806518555, "learning_rate": 5.732405224170105e-05, "loss": 0.0626, "step": 96110 }, { "epoch": 3.2496027587139524, "grad_norm": 1.2294831275939941, "learning_rate": 5.731587396848081e-05, "loss": 0.0604, "step": 96120 }, { "epoch": 3.2499408364042055, "grad_norm": 2.250420331954956, "learning_rate": 5.730769549524733e-05, "loss": 0.0597, "step": 96130 }, { "epoch": 3.250278914094459, "grad_norm": 1.9290995597839355, "learning_rate": 5.72995168222242e-05, "loss": 0.0623, "step": 96140 }, { "epoch": 3.250616991784712, "grad_norm": 2.8472418785095215, "learning_rate": 5.729133794963502e-05, "loss": 0.0665, "step": 96150 }, { "epoch": 3.250955069474965, "grad_norm": 1.9950001239776611, "learning_rate": 5.72831588777034e-05, "loss": 0.0582, "step": 96160 }, { "epoch": 3.2512931471652187, "grad_norm": 3.793895721435547, "learning_rate": 5.7274979606652944e-05, "loss": 0.0668, "step": 96170 }, { "epoch": 3.2516312248554717, "grad_norm": 5.105072021484375, "learning_rate": 5.72668001367073e-05, "loss": 0.0634, "step": 96180 }, { "epoch": 3.2519693025457252, "grad_norm": 2.5446903705596924, "learning_rate": 5.7258620468090053e-05, "loss": 0.0673, "step": 96190 }, { "epoch": 3.2523073802359783, "grad_norm": 1.8184586763381958, "learning_rate": 5.7250440601024844e-05, "loss": 0.0681, "step": 96200 }, { "epoch": 3.2526454579262314, "grad_norm": 3.4494030475616455, "learning_rate": 5.7242260535735325e-05, "loss": 0.0743, "step": 96210 }, { "epoch": 3.252983535616485, "grad_norm": 8.446684837341309, "learning_rate": 5.723408027244513e-05, "loss": 0.0677, "step": 96220 }, { "epoch": 3.253321613306738, "grad_norm": 2.6078004837036133, "learning_rate": 5.722589981137788e-05, "loss": 0.0644, "step": 96230 }, { "epoch": 3.253659690996991, "grad_norm": 7.119508266448975, "learning_rate": 5.721771915275727e-05, "loss": 0.0668, "step": 96240 }, { "epoch": 3.2539977686872446, "grad_norm": 0.9632529020309448, "learning_rate": 5.720953829680691e-05, "loss": 0.0615, "step": 96250 }, { "epoch": 3.2543358463774976, "grad_norm": 1.6414827108383179, "learning_rate": 5.7201357243750475e-05, "loss": 0.068, "step": 96260 }, { "epoch": 3.2546739240677507, "grad_norm": 1.9557795524597168, "learning_rate": 5.7193175993811646e-05, "loss": 0.0648, "step": 96270 }, { "epoch": 3.255012001758004, "grad_norm": 2.228689432144165, "learning_rate": 5.71849945472141e-05, "loss": 0.0704, "step": 96280 }, { "epoch": 3.2553500794482573, "grad_norm": 4.455331325531006, "learning_rate": 5.7176812904181485e-05, "loss": 0.0639, "step": 96290 }, { "epoch": 3.2556881571385103, "grad_norm": 3.748002290725708, "learning_rate": 5.7168631064937515e-05, "loss": 0.0675, "step": 96300 }, { "epoch": 3.256026234828764, "grad_norm": 1.0282682180404663, "learning_rate": 5.7160449029705855e-05, "loss": 0.0652, "step": 96310 }, { "epoch": 3.256364312519017, "grad_norm": 1.0702266693115234, "learning_rate": 5.715226679871021e-05, "loss": 0.0636, "step": 96320 }, { "epoch": 3.25670239020927, "grad_norm": 2.408916473388672, "learning_rate": 5.714408437217428e-05, "loss": 0.0674, "step": 96330 }, { "epoch": 3.2570404678995235, "grad_norm": 1.3600754737854004, "learning_rate": 5.7135901750321774e-05, "loss": 0.0648, "step": 96340 }, { "epoch": 3.2573785455897766, "grad_norm": 5.3321428298950195, "learning_rate": 5.7127718933376384e-05, "loss": 0.0674, "step": 96350 }, { "epoch": 3.2577166232800296, "grad_norm": 2.1747825145721436, "learning_rate": 5.711953592156184e-05, "loss": 0.0627, "step": 96360 }, { "epoch": 3.258054700970283, "grad_norm": 1.2996946573257446, "learning_rate": 5.7111352715101866e-05, "loss": 0.066, "step": 96370 }, { "epoch": 3.2583927786605362, "grad_norm": 0.9803115129470825, "learning_rate": 5.710316931422016e-05, "loss": 0.0641, "step": 96380 }, { "epoch": 3.2587308563507893, "grad_norm": 35.48767852783203, "learning_rate": 5.709498571914048e-05, "loss": 0.0655, "step": 96390 }, { "epoch": 3.259068934041043, "grad_norm": 3.7677807807922363, "learning_rate": 5.7086801930086574e-05, "loss": 0.0676, "step": 96400 }, { "epoch": 3.259407011731296, "grad_norm": 1.708949089050293, "learning_rate": 5.7078617947282155e-05, "loss": 0.0666, "step": 96410 }, { "epoch": 3.259745089421549, "grad_norm": 2.1735968589782715, "learning_rate": 5.707043377095097e-05, "loss": 0.0628, "step": 96420 }, { "epoch": 3.2600831671118025, "grad_norm": 1.5026557445526123, "learning_rate": 5.70622494013168e-05, "loss": 0.0652, "step": 96430 }, { "epoch": 3.2604212448020555, "grad_norm": 3.6262338161468506, "learning_rate": 5.705406483860337e-05, "loss": 0.0592, "step": 96440 }, { "epoch": 3.2607593224923086, "grad_norm": 0.8317398428916931, "learning_rate": 5.7045880083034464e-05, "loss": 0.0596, "step": 96450 }, { "epoch": 3.261097400182562, "grad_norm": 1.6114541292190552, "learning_rate": 5.703769513483383e-05, "loss": 0.0634, "step": 96460 }, { "epoch": 3.261435477872815, "grad_norm": 2.2564704418182373, "learning_rate": 5.702950999422528e-05, "loss": 0.064, "step": 96470 }, { "epoch": 3.2617735555630682, "grad_norm": 1.56371009349823, "learning_rate": 5.702132466143255e-05, "loss": 0.0591, "step": 96480 }, { "epoch": 3.2621116332533218, "grad_norm": 1.6698427200317383, "learning_rate": 5.7013139136679464e-05, "loss": 0.0637, "step": 96490 }, { "epoch": 3.262449710943575, "grad_norm": 3.842864990234375, "learning_rate": 5.700495342018977e-05, "loss": 0.0622, "step": 96500 }, { "epoch": 3.262787788633828, "grad_norm": 12.08166790008545, "learning_rate": 5.699676751218728e-05, "loss": 0.0752, "step": 96510 }, { "epoch": 3.2631258663240814, "grad_norm": 1.6723833084106445, "learning_rate": 5.698858141289579e-05, "loss": 0.0578, "step": 96520 }, { "epoch": 3.2634639440143345, "grad_norm": 2.207643985748291, "learning_rate": 5.698039512253912e-05, "loss": 0.0664, "step": 96530 }, { "epoch": 3.2638020217045876, "grad_norm": 1.0954636335372925, "learning_rate": 5.6972208641341064e-05, "loss": 0.0535, "step": 96540 }, { "epoch": 3.264140099394841, "grad_norm": 2.196070909500122, "learning_rate": 5.696402196952546e-05, "loss": 0.0575, "step": 96550 }, { "epoch": 3.264478177085094, "grad_norm": 1.0474485158920288, "learning_rate": 5.695583510731609e-05, "loss": 0.0608, "step": 96560 }, { "epoch": 3.264816254775347, "grad_norm": 1.3202755451202393, "learning_rate": 5.694764805493682e-05, "loss": 0.058, "step": 96570 }, { "epoch": 3.2651543324656007, "grad_norm": 1.7880401611328125, "learning_rate": 5.6939460812611436e-05, "loss": 0.07, "step": 96580 }, { "epoch": 3.265492410155854, "grad_norm": 1.7756385803222656, "learning_rate": 5.693127338056382e-05, "loss": 0.061, "step": 96590 }, { "epoch": 3.265830487846107, "grad_norm": 3.1175506114959717, "learning_rate": 5.692308575901779e-05, "loss": 0.067, "step": 96600 }, { "epoch": 3.2661685655363604, "grad_norm": 1.3032766580581665, "learning_rate": 5.69148979481972e-05, "loss": 0.0669, "step": 96610 }, { "epoch": 3.2665066432266134, "grad_norm": 3.0174002647399902, "learning_rate": 5.690670994832589e-05, "loss": 0.0605, "step": 96620 }, { "epoch": 3.2668447209168665, "grad_norm": 1.2939804792404175, "learning_rate": 5.689852175962772e-05, "loss": 0.0547, "step": 96630 }, { "epoch": 3.26718279860712, "grad_norm": 1.8942632675170898, "learning_rate": 5.6890333382326566e-05, "loss": 0.0634, "step": 96640 }, { "epoch": 3.267520876297373, "grad_norm": 0.9499137997627258, "learning_rate": 5.6882144816646286e-05, "loss": 0.0581, "step": 96650 }, { "epoch": 3.267858953987626, "grad_norm": 1.3017417192459106, "learning_rate": 5.687395606281075e-05, "loss": 0.0593, "step": 96660 }, { "epoch": 3.2681970316778797, "grad_norm": 1.8347946405410767, "learning_rate": 5.686576712104383e-05, "loss": 0.0625, "step": 96670 }, { "epoch": 3.2685351093681327, "grad_norm": 4.120349884033203, "learning_rate": 5.685757799156944e-05, "loss": 0.0644, "step": 96680 }, { "epoch": 3.268873187058386, "grad_norm": 2.4423463344573975, "learning_rate": 5.684938867461143e-05, "loss": 0.0579, "step": 96690 }, { "epoch": 3.2692112647486393, "grad_norm": 2.267524242401123, "learning_rate": 5.68411991703937e-05, "loss": 0.0665, "step": 96700 }, { "epoch": 3.2695493424388924, "grad_norm": 2.0680015087127686, "learning_rate": 5.683300947914015e-05, "loss": 0.0698, "step": 96710 }, { "epoch": 3.2698874201291455, "grad_norm": 1.9479249715805054, "learning_rate": 5.682481960107471e-05, "loss": 0.0615, "step": 96720 }, { "epoch": 3.270225497819399, "grad_norm": 2.2357122898101807, "learning_rate": 5.681662953642125e-05, "loss": 0.066, "step": 96730 }, { "epoch": 3.270563575509652, "grad_norm": 1.2510079145431519, "learning_rate": 5.6808439285403716e-05, "loss": 0.0669, "step": 96740 }, { "epoch": 3.270901653199905, "grad_norm": 1.3091760873794556, "learning_rate": 5.6800248848246e-05, "loss": 0.0661, "step": 96750 }, { "epoch": 3.2712397308901586, "grad_norm": 1.018257975578308, "learning_rate": 5.679205822517204e-05, "loss": 0.0647, "step": 96760 }, { "epoch": 3.2715778085804117, "grad_norm": 1.2191476821899414, "learning_rate": 5.6783867416405734e-05, "loss": 0.0616, "step": 96770 }, { "epoch": 3.2719158862706648, "grad_norm": 1.2251863479614258, "learning_rate": 5.677567642217107e-05, "loss": 0.0576, "step": 96780 }, { "epoch": 3.2722539639609183, "grad_norm": 4.454538345336914, "learning_rate": 5.676748524269194e-05, "loss": 0.0581, "step": 96790 }, { "epoch": 3.2725920416511713, "grad_norm": 1.9925568103790283, "learning_rate": 5.675929387819232e-05, "loss": 0.0618, "step": 96800 }, { "epoch": 3.2729301193414244, "grad_norm": 1.5142433643341064, "learning_rate": 5.675110232889614e-05, "loss": 0.0639, "step": 96810 }, { "epoch": 3.273268197031678, "grad_norm": 0.8485532402992249, "learning_rate": 5.6742910595027366e-05, "loss": 0.0683, "step": 96820 }, { "epoch": 3.273606274721931, "grad_norm": 1.7036821842193604, "learning_rate": 5.6734718676809926e-05, "loss": 0.0652, "step": 96830 }, { "epoch": 3.2739443524121845, "grad_norm": 0.6747700572013855, "learning_rate": 5.672652657446783e-05, "loss": 0.0659, "step": 96840 }, { "epoch": 3.2742824301024376, "grad_norm": 1.1303452253341675, "learning_rate": 5.671833428822502e-05, "loss": 0.0668, "step": 96850 }, { "epoch": 3.2746205077926906, "grad_norm": 1.362564206123352, "learning_rate": 5.6710141818305476e-05, "loss": 0.0646, "step": 96860 }, { "epoch": 3.274958585482944, "grad_norm": 1.8264145851135254, "learning_rate": 5.6701949164933165e-05, "loss": 0.0595, "step": 96870 }, { "epoch": 3.2752966631731972, "grad_norm": 2.281526565551758, "learning_rate": 5.6693756328332084e-05, "loss": 0.0641, "step": 96880 }, { "epoch": 3.2756347408634503, "grad_norm": 1.219321370124817, "learning_rate": 5.668556330872622e-05, "loss": 0.0666, "step": 96890 }, { "epoch": 3.275972818553704, "grad_norm": 3.562302589416504, "learning_rate": 5.667737010633958e-05, "loss": 0.0631, "step": 96900 }, { "epoch": 3.276310896243957, "grad_norm": 1.5843507051467896, "learning_rate": 5.666917672139613e-05, "loss": 0.065, "step": 96910 }, { "epoch": 3.27664897393421, "grad_norm": 1.7624534368515015, "learning_rate": 5.666098315411991e-05, "loss": 0.0639, "step": 96920 }, { "epoch": 3.2769870516244635, "grad_norm": 2.243149518966675, "learning_rate": 5.6652789404734904e-05, "loss": 0.0644, "step": 96930 }, { "epoch": 3.2773251293147165, "grad_norm": 1.5525641441345215, "learning_rate": 5.664459547346514e-05, "loss": 0.0616, "step": 96940 }, { "epoch": 3.2776632070049696, "grad_norm": 0.9670230746269226, "learning_rate": 5.663640136053464e-05, "loss": 0.0696, "step": 96950 }, { "epoch": 3.278001284695223, "grad_norm": 1.9673084020614624, "learning_rate": 5.6628207066167395e-05, "loss": 0.0636, "step": 96960 }, { "epoch": 3.278339362385476, "grad_norm": 1.8452752828598022, "learning_rate": 5.662001259058748e-05, "loss": 0.0638, "step": 96970 }, { "epoch": 3.2786774400757293, "grad_norm": 1.0627812147140503, "learning_rate": 5.661181793401889e-05, "loss": 0.0715, "step": 96980 }, { "epoch": 3.2790155177659828, "grad_norm": 2.6712327003479004, "learning_rate": 5.660362309668571e-05, "loss": 0.0695, "step": 96990 }, { "epoch": 3.279353595456236, "grad_norm": 1.3288127183914185, "learning_rate": 5.659542807881193e-05, "loss": 0.0543, "step": 97000 }, { "epoch": 3.279691673146489, "grad_norm": 4.815249443054199, "learning_rate": 5.6587232880621646e-05, "loss": 0.0641, "step": 97010 }, { "epoch": 3.2800297508367424, "grad_norm": 2.7588448524475098, "learning_rate": 5.657903750233886e-05, "loss": 0.0654, "step": 97020 }, { "epoch": 3.2803678285269955, "grad_norm": 0.8288488984107971, "learning_rate": 5.657084194418768e-05, "loss": 0.0602, "step": 97030 }, { "epoch": 3.2807059062172486, "grad_norm": 1.9165922403335571, "learning_rate": 5.6562646206392146e-05, "loss": 0.0699, "step": 97040 }, { "epoch": 3.281043983907502, "grad_norm": 1.4596234560012817, "learning_rate": 5.655445028917634e-05, "loss": 0.0605, "step": 97050 }, { "epoch": 3.281382061597755, "grad_norm": 1.9794437885284424, "learning_rate": 5.6546254192764314e-05, "loss": 0.0641, "step": 97060 }, { "epoch": 3.281720139288008, "grad_norm": 1.6059691905975342, "learning_rate": 5.6538057917380174e-05, "loss": 0.0651, "step": 97070 }, { "epoch": 3.2820582169782617, "grad_norm": 1.3281593322753906, "learning_rate": 5.6529861463247966e-05, "loss": 0.0613, "step": 97080 }, { "epoch": 3.282396294668515, "grad_norm": 1.0528875589370728, "learning_rate": 5.65216648305918e-05, "loss": 0.0643, "step": 97090 }, { "epoch": 3.282734372358768, "grad_norm": 1.7811717987060547, "learning_rate": 5.6513468019635775e-05, "loss": 0.069, "step": 97100 }, { "epoch": 3.2830724500490214, "grad_norm": 2.0822594165802, "learning_rate": 5.650527103060398e-05, "loss": 0.0617, "step": 97110 }, { "epoch": 3.2834105277392744, "grad_norm": 1.749049425125122, "learning_rate": 5.649707386372052e-05, "loss": 0.0645, "step": 97120 }, { "epoch": 3.2837486054295275, "grad_norm": 1.826456069946289, "learning_rate": 5.648887651920949e-05, "loss": 0.0672, "step": 97130 }, { "epoch": 3.284086683119781, "grad_norm": 2.3343639373779297, "learning_rate": 5.6480678997295036e-05, "loss": 0.0695, "step": 97140 }, { "epoch": 3.284424760810034, "grad_norm": 3.031651020050049, "learning_rate": 5.647248129820122e-05, "loss": 0.0706, "step": 97150 }, { "epoch": 3.2847628385002876, "grad_norm": 2.084500551223755, "learning_rate": 5.646428342215223e-05, "loss": 0.0769, "step": 97160 }, { "epoch": 3.2851009161905407, "grad_norm": 3.470562696456909, "learning_rate": 5.6456085369372144e-05, "loss": 0.0753, "step": 97170 }, { "epoch": 3.2854389938807937, "grad_norm": 2.557492733001709, "learning_rate": 5.644788714008512e-05, "loss": 0.0674, "step": 97180 }, { "epoch": 3.2857770715710473, "grad_norm": 3.800447463989258, "learning_rate": 5.6439688734515274e-05, "loss": 0.0727, "step": 97190 }, { "epoch": 3.2861151492613003, "grad_norm": 4.107000350952148, "learning_rate": 5.6431490152886766e-05, "loss": 0.0691, "step": 97200 }, { "epoch": 3.2864532269515534, "grad_norm": 6.056979656219482, "learning_rate": 5.642329139542372e-05, "loss": 0.07, "step": 97210 }, { "epoch": 3.286791304641807, "grad_norm": 7.529890060424805, "learning_rate": 5.6415092462350304e-05, "loss": 0.0702, "step": 97220 }, { "epoch": 3.28712938233206, "grad_norm": 1.5004383325576782, "learning_rate": 5.6406893353890675e-05, "loss": 0.0631, "step": 97230 }, { "epoch": 3.287467460022313, "grad_norm": 1.5409601926803589, "learning_rate": 5.6398694070268996e-05, "loss": 0.0714, "step": 97240 }, { "epoch": 3.2878055377125666, "grad_norm": 2.781672716140747, "learning_rate": 5.639049461170941e-05, "loss": 0.0602, "step": 97250 }, { "epoch": 3.2881436154028196, "grad_norm": 2.307940721511841, "learning_rate": 5.6382294978436124e-05, "loss": 0.066, "step": 97260 }, { "epoch": 3.2884816930930727, "grad_norm": 6.474532604217529, "learning_rate": 5.637409517067327e-05, "loss": 0.0763, "step": 97270 }, { "epoch": 3.288819770783326, "grad_norm": 3.976248025894165, "learning_rate": 5.636589518864505e-05, "loss": 0.0717, "step": 97280 }, { "epoch": 3.2891578484735793, "grad_norm": 3.0820109844207764, "learning_rate": 5.6357695032575644e-05, "loss": 0.0663, "step": 97290 }, { "epoch": 3.2894959261638324, "grad_norm": 2.754758358001709, "learning_rate": 5.6349494702689267e-05, "loss": 0.0671, "step": 97300 }, { "epoch": 3.289834003854086, "grad_norm": 2.4515578746795654, "learning_rate": 5.634129419921006e-05, "loss": 0.0681, "step": 97310 }, { "epoch": 3.290172081544339, "grad_norm": 3.2594311237335205, "learning_rate": 5.6333093522362276e-05, "loss": 0.0685, "step": 97320 }, { "epoch": 3.290510159234592, "grad_norm": 2.094761371612549, "learning_rate": 5.632489267237009e-05, "loss": 0.0734, "step": 97330 }, { "epoch": 3.2908482369248455, "grad_norm": 3.3675756454467773, "learning_rate": 5.63166916494577e-05, "loss": 0.0636, "step": 97340 }, { "epoch": 3.2911863146150986, "grad_norm": 1.8822412490844727, "learning_rate": 5.630849045384934e-05, "loss": 0.0603, "step": 97350 }, { "epoch": 3.2915243923053517, "grad_norm": 3.8828155994415283, "learning_rate": 5.630028908576922e-05, "loss": 0.0681, "step": 97360 }, { "epoch": 3.291862469995605, "grad_norm": 2.5095441341400146, "learning_rate": 5.6292087545441575e-05, "loss": 0.0696, "step": 97370 }, { "epoch": 3.2922005476858582, "grad_norm": 2.1830668449401855, "learning_rate": 5.628388583309061e-05, "loss": 0.0708, "step": 97380 }, { "epoch": 3.2925386253761113, "grad_norm": 14.715867042541504, "learning_rate": 5.627568394894057e-05, "loss": 0.0657, "step": 97390 }, { "epoch": 3.292876703066365, "grad_norm": 1.2991981506347656, "learning_rate": 5.626748189321568e-05, "loss": 0.0639, "step": 97400 }, { "epoch": 3.293214780756618, "grad_norm": 6.566405296325684, "learning_rate": 5.6259279666140206e-05, "loss": 0.0645, "step": 97410 }, { "epoch": 3.293552858446871, "grad_norm": 5.819421291351318, "learning_rate": 5.6251077267938366e-05, "loss": 0.0654, "step": 97420 }, { "epoch": 3.2938909361371245, "grad_norm": 2.1592814922332764, "learning_rate": 5.6242874698834435e-05, "loss": 0.0575, "step": 97430 }, { "epoch": 3.2942290138273775, "grad_norm": 1.590632677078247, "learning_rate": 5.623467195905264e-05, "loss": 0.0756, "step": 97440 }, { "epoch": 3.2945670915176306, "grad_norm": 2.0684118270874023, "learning_rate": 5.622646904881726e-05, "loss": 0.0692, "step": 97450 }, { "epoch": 3.294905169207884, "grad_norm": 8.213935852050781, "learning_rate": 5.6218265968352556e-05, "loss": 0.0635, "step": 97460 }, { "epoch": 3.295243246898137, "grad_norm": 2.249898910522461, "learning_rate": 5.621006271788279e-05, "loss": 0.0612, "step": 97470 }, { "epoch": 3.2955813245883903, "grad_norm": 1.050568699836731, "learning_rate": 5.6201859297632256e-05, "loss": 0.0616, "step": 97480 }, { "epoch": 3.2959194022786438, "grad_norm": 2.433480978012085, "learning_rate": 5.6193655707825224e-05, "loss": 0.0594, "step": 97490 }, { "epoch": 3.296257479968897, "grad_norm": 1.7860223054885864, "learning_rate": 5.618545194868595e-05, "loss": 0.0598, "step": 97500 }, { "epoch": 3.29659555765915, "grad_norm": 2.441898822784424, "learning_rate": 5.6177248020438755e-05, "loss": 0.0618, "step": 97510 }, { "epoch": 3.2969336353494034, "grad_norm": 2.1118738651275635, "learning_rate": 5.616904392330792e-05, "loss": 0.0647, "step": 97520 }, { "epoch": 3.2972717130396565, "grad_norm": 1.5180317163467407, "learning_rate": 5.616083965751774e-05, "loss": 0.0608, "step": 97530 }, { "epoch": 3.2976097907299096, "grad_norm": 2.1436736583709717, "learning_rate": 5.6152635223292505e-05, "loss": 0.0678, "step": 97540 }, { "epoch": 3.297947868420163, "grad_norm": 1.3464488983154297, "learning_rate": 5.6144430620856556e-05, "loss": 0.0754, "step": 97550 }, { "epoch": 3.298285946110416, "grad_norm": 3.109997510910034, "learning_rate": 5.613622585043418e-05, "loss": 0.066, "step": 97560 }, { "epoch": 3.298624023800669, "grad_norm": 2.709782838821411, "learning_rate": 5.6128020912249676e-05, "loss": 0.0661, "step": 97570 }, { "epoch": 3.2989621014909227, "grad_norm": 2.548752784729004, "learning_rate": 5.611981580652741e-05, "loss": 0.0594, "step": 97580 }, { "epoch": 3.299300179181176, "grad_norm": 1.64768385887146, "learning_rate": 5.6111610533491654e-05, "loss": 0.0669, "step": 97590 }, { "epoch": 3.299638256871429, "grad_norm": 1.9225680828094482, "learning_rate": 5.610340509336677e-05, "loss": 0.0687, "step": 97600 }, { "epoch": 3.2999763345616824, "grad_norm": 1.6988402605056763, "learning_rate": 5.609519948637708e-05, "loss": 0.0661, "step": 97610 }, { "epoch": 3.3003144122519354, "grad_norm": 2.3465185165405273, "learning_rate": 5.608699371274694e-05, "loss": 0.0567, "step": 97620 }, { "epoch": 3.3006524899421885, "grad_norm": 3.2863941192626953, "learning_rate": 5.607878777270067e-05, "loss": 0.0615, "step": 97630 }, { "epoch": 3.300990567632442, "grad_norm": 3.4185469150543213, "learning_rate": 5.607058166646263e-05, "loss": 0.0639, "step": 97640 }, { "epoch": 3.301328645322695, "grad_norm": 1.4929194450378418, "learning_rate": 5.606237539425716e-05, "loss": 0.0659, "step": 97650 }, { "epoch": 3.301666723012948, "grad_norm": 2.685325860977173, "learning_rate": 5.605416895630863e-05, "loss": 0.0679, "step": 97660 }, { "epoch": 3.3020048007032017, "grad_norm": 1.6053504943847656, "learning_rate": 5.604596235284139e-05, "loss": 0.0638, "step": 97670 }, { "epoch": 3.3023428783934548, "grad_norm": 1.5962586402893066, "learning_rate": 5.603775558407982e-05, "loss": 0.062, "step": 97680 }, { "epoch": 3.302680956083708, "grad_norm": 3.0004618167877197, "learning_rate": 5.602954865024826e-05, "loss": 0.0665, "step": 97690 }, { "epoch": 3.3030190337739613, "grad_norm": 1.4405159950256348, "learning_rate": 5.6021341551571135e-05, "loss": 0.0606, "step": 97700 }, { "epoch": 3.3033571114642144, "grad_norm": 0.9323379397392273, "learning_rate": 5.601313428827277e-05, "loss": 0.0628, "step": 97710 }, { "epoch": 3.3036951891544675, "grad_norm": 1.7434678077697754, "learning_rate": 5.600492686057759e-05, "loss": 0.0711, "step": 97720 }, { "epoch": 3.304033266844721, "grad_norm": 3.693145751953125, "learning_rate": 5.599671926870995e-05, "loss": 0.0622, "step": 97730 }, { "epoch": 3.304371344534974, "grad_norm": 2.426215171813965, "learning_rate": 5.598851151289427e-05, "loss": 0.0658, "step": 97740 }, { "epoch": 3.304709422225227, "grad_norm": 1.4690289497375488, "learning_rate": 5.598030359335492e-05, "loss": 0.0647, "step": 97750 }, { "epoch": 3.3050474999154806, "grad_norm": 1.5999475717544556, "learning_rate": 5.5972095510316336e-05, "loss": 0.0704, "step": 97760 }, { "epoch": 3.3053855776057337, "grad_norm": 3.808518886566162, "learning_rate": 5.59638872640029e-05, "loss": 0.0669, "step": 97770 }, { "epoch": 3.3057236552959868, "grad_norm": 1.0993900299072266, "learning_rate": 5.5955678854639015e-05, "loss": 0.0684, "step": 97780 }, { "epoch": 3.3060617329862403, "grad_norm": 2.467857599258423, "learning_rate": 5.594747028244911e-05, "loss": 0.0692, "step": 97790 }, { "epoch": 3.3063998106764934, "grad_norm": 1.8893866539001465, "learning_rate": 5.5939261547657604e-05, "loss": 0.0655, "step": 97800 }, { "epoch": 3.306737888366747, "grad_norm": 2.7209107875823975, "learning_rate": 5.5931052650488926e-05, "loss": 0.066, "step": 97810 }, { "epoch": 3.307075966057, "grad_norm": 3.5129318237304688, "learning_rate": 5.5922843591167486e-05, "loss": 0.0733, "step": 97820 }, { "epoch": 3.307414043747253, "grad_norm": 1.2200672626495361, "learning_rate": 5.591463436991774e-05, "loss": 0.0681, "step": 97830 }, { "epoch": 3.3077521214375065, "grad_norm": 2.8572189807891846, "learning_rate": 5.59064249869641e-05, "loss": 0.0704, "step": 97840 }, { "epoch": 3.3080901991277596, "grad_norm": 1.323327898979187, "learning_rate": 5.589821544253102e-05, "loss": 0.0656, "step": 97850 }, { "epoch": 3.3084282768180127, "grad_norm": 1.2734920978546143, "learning_rate": 5.5890005736842955e-05, "loss": 0.0709, "step": 97860 }, { "epoch": 3.308766354508266, "grad_norm": 1.5687850713729858, "learning_rate": 5.588179587012434e-05, "loss": 0.0662, "step": 97870 }, { "epoch": 3.3091044321985192, "grad_norm": 1.7976855039596558, "learning_rate": 5.587358584259964e-05, "loss": 0.0635, "step": 97880 }, { "epoch": 3.3094425098887723, "grad_norm": 23.859661102294922, "learning_rate": 5.586537565449331e-05, "loss": 0.0718, "step": 97890 }, { "epoch": 3.309780587579026, "grad_norm": 1.5146172046661377, "learning_rate": 5.58571653060298e-05, "loss": 0.0626, "step": 97900 }, { "epoch": 3.310118665269279, "grad_norm": 6.666080474853516, "learning_rate": 5.5848954797433606e-05, "loss": 0.0635, "step": 97910 }, { "epoch": 3.310456742959532, "grad_norm": 2.520810604095459, "learning_rate": 5.584074412892918e-05, "loss": 0.0695, "step": 97920 }, { "epoch": 3.3107948206497855, "grad_norm": 1.285630702972412, "learning_rate": 5.583253330074101e-05, "loss": 0.063, "step": 97930 }, { "epoch": 3.3111328983400385, "grad_norm": 2.8788678646087646, "learning_rate": 5.582432231309356e-05, "loss": 0.0656, "step": 97940 }, { "epoch": 3.3114709760302916, "grad_norm": 1.6253224611282349, "learning_rate": 5.5816111166211336e-05, "loss": 0.0676, "step": 97950 }, { "epoch": 3.311809053720545, "grad_norm": 4.261768817901611, "learning_rate": 5.5807899860318803e-05, "loss": 0.0709, "step": 97960 }, { "epoch": 3.312147131410798, "grad_norm": 1.166587233543396, "learning_rate": 5.579968839564049e-05, "loss": 0.0701, "step": 97970 }, { "epoch": 3.3124852091010513, "grad_norm": 1.404114842414856, "learning_rate": 5.579147677240085e-05, "loss": 0.0602, "step": 97980 }, { "epoch": 3.3128232867913048, "grad_norm": 1.3163731098175049, "learning_rate": 5.5783264990824434e-05, "loss": 0.0591, "step": 97990 }, { "epoch": 3.313161364481558, "grad_norm": 2.414443016052246, "learning_rate": 5.57750530511357e-05, "loss": 0.0592, "step": 98000 }, { "epoch": 3.313499442171811, "grad_norm": 3.851609468460083, "learning_rate": 5.576684095355922e-05, "loss": 0.0645, "step": 98010 }, { "epoch": 3.3138375198620644, "grad_norm": 1.3382335901260376, "learning_rate": 5.575862869831945e-05, "loss": 0.0655, "step": 98020 }, { "epoch": 3.3141755975523175, "grad_norm": 1.381211280822754, "learning_rate": 5.575041628564094e-05, "loss": 0.0613, "step": 98030 }, { "epoch": 3.3145136752425706, "grad_norm": 1.3151304721832275, "learning_rate": 5.574220371574821e-05, "loss": 0.0625, "step": 98040 }, { "epoch": 3.314851752932824, "grad_norm": 1.8085627555847168, "learning_rate": 5.573399098886579e-05, "loss": 0.0641, "step": 98050 }, { "epoch": 3.315189830623077, "grad_norm": 1.19582200050354, "learning_rate": 5.572577810521821e-05, "loss": 0.0585, "step": 98060 }, { "epoch": 3.31552790831333, "grad_norm": 0.8053649663925171, "learning_rate": 5.571756506503e-05, "loss": 0.0614, "step": 98070 }, { "epoch": 3.3158659860035837, "grad_norm": 4.193192005157471, "learning_rate": 5.570935186852573e-05, "loss": 0.0636, "step": 98080 }, { "epoch": 3.316204063693837, "grad_norm": 1.6857115030288696, "learning_rate": 5.570113851592991e-05, "loss": 0.0651, "step": 98090 }, { "epoch": 3.31654214138409, "grad_norm": 1.431353211402893, "learning_rate": 5.569292500746709e-05, "loss": 0.0618, "step": 98100 }, { "epoch": 3.3168802190743434, "grad_norm": 2.3436951637268066, "learning_rate": 5.568471134336185e-05, "loss": 0.0677, "step": 98110 }, { "epoch": 3.3172182967645965, "grad_norm": 1.2256040573120117, "learning_rate": 5.567649752383876e-05, "loss": 0.063, "step": 98120 }, { "epoch": 3.31755637445485, "grad_norm": 1.1955233812332153, "learning_rate": 5.5668283549122336e-05, "loss": 0.0656, "step": 98130 }, { "epoch": 3.317894452145103, "grad_norm": 1.233873963356018, "learning_rate": 5.566006941943718e-05, "loss": 0.0616, "step": 98140 }, { "epoch": 3.318232529835356, "grad_norm": 1.0065991878509521, "learning_rate": 5.565185513500785e-05, "loss": 0.0576, "step": 98150 }, { "epoch": 3.3185706075256096, "grad_norm": 1.7173558473587036, "learning_rate": 5.564364069605892e-05, "loss": 0.0608, "step": 98160 }, { "epoch": 3.3189086852158627, "grad_norm": 1.9637893438339233, "learning_rate": 5.563542610281496e-05, "loss": 0.0578, "step": 98170 }, { "epoch": 3.3192467629061158, "grad_norm": 1.2180333137512207, "learning_rate": 5.5627211355500594e-05, "loss": 0.0616, "step": 98180 }, { "epoch": 3.3195848405963693, "grad_norm": 1.07567298412323, "learning_rate": 5.561899645434037e-05, "loss": 0.0614, "step": 98190 }, { "epoch": 3.3199229182866223, "grad_norm": 8.434762954711914, "learning_rate": 5.56107813995589e-05, "loss": 0.066, "step": 98200 }, { "epoch": 3.3202609959768754, "grad_norm": 1.1414870023727417, "learning_rate": 5.560256619138077e-05, "loss": 0.0699, "step": 98210 }, { "epoch": 3.320599073667129, "grad_norm": 8.704984664916992, "learning_rate": 5.55943508300306e-05, "loss": 0.0629, "step": 98220 }, { "epoch": 3.320937151357382, "grad_norm": 1.2802269458770752, "learning_rate": 5.558613531573294e-05, "loss": 0.0604, "step": 98230 }, { "epoch": 3.321275229047635, "grad_norm": 6.4401960372924805, "learning_rate": 5.5577919648712475e-05, "loss": 0.0586, "step": 98240 }, { "epoch": 3.3216133067378886, "grad_norm": 2.8210761547088623, "learning_rate": 5.556970382919376e-05, "loss": 0.0652, "step": 98250 }, { "epoch": 3.3219513844281416, "grad_norm": 2.455113172531128, "learning_rate": 5.556148785740146e-05, "loss": 0.0641, "step": 98260 }, { "epoch": 3.3222894621183947, "grad_norm": 2.70147442817688, "learning_rate": 5.5553271733560154e-05, "loss": 0.0578, "step": 98270 }, { "epoch": 3.3226275398086482, "grad_norm": 1.616568684577942, "learning_rate": 5.554505545789449e-05, "loss": 0.0561, "step": 98280 }, { "epoch": 3.3229656174989013, "grad_norm": 3.567328691482544, "learning_rate": 5.5536839030629084e-05, "loss": 0.0646, "step": 98290 }, { "epoch": 3.3233036951891544, "grad_norm": 1.6112176179885864, "learning_rate": 5.55286224519886e-05, "loss": 0.0655, "step": 98300 }, { "epoch": 3.323641772879408, "grad_norm": 2.6971030235290527, "learning_rate": 5.552040572219763e-05, "loss": 0.0571, "step": 98310 }, { "epoch": 3.323979850569661, "grad_norm": 1.659011960029602, "learning_rate": 5.5512188841480864e-05, "loss": 0.0646, "step": 98320 }, { "epoch": 3.324317928259914, "grad_norm": 1.7424734830856323, "learning_rate": 5.550397181006292e-05, "loss": 0.0576, "step": 98330 }, { "epoch": 3.3246560059501675, "grad_norm": 1.0816357135772705, "learning_rate": 5.549575462816844e-05, "loss": 0.0597, "step": 98340 }, { "epoch": 3.3249940836404206, "grad_norm": 1.1233949661254883, "learning_rate": 5.5487537296022106e-05, "loss": 0.061, "step": 98350 }, { "epoch": 3.3253321613306737, "grad_norm": 1.077933430671692, "learning_rate": 5.547931981384855e-05, "loss": 0.0551, "step": 98360 }, { "epoch": 3.325670239020927, "grad_norm": 1.178673505783081, "learning_rate": 5.547110218187247e-05, "loss": 0.0532, "step": 98370 }, { "epoch": 3.3260083167111802, "grad_norm": 1.1448910236358643, "learning_rate": 5.5462884400318494e-05, "loss": 0.0655, "step": 98380 }, { "epoch": 3.3263463944014333, "grad_norm": 1.1593159437179565, "learning_rate": 5.545466646941133e-05, "loss": 0.0698, "step": 98390 }, { "epoch": 3.326684472091687, "grad_norm": 2.092484951019287, "learning_rate": 5.5446448389375614e-05, "loss": 0.0591, "step": 98400 }, { "epoch": 3.32702254978194, "grad_norm": 0.6543875336647034, "learning_rate": 5.5438230160436066e-05, "loss": 0.0584, "step": 98410 }, { "epoch": 3.327360627472193, "grad_norm": 0.9196240901947021, "learning_rate": 5.5430011782817326e-05, "loss": 0.0625, "step": 98420 }, { "epoch": 3.3276987051624465, "grad_norm": 1.894120454788208, "learning_rate": 5.5421793256744115e-05, "loss": 0.0617, "step": 98430 }, { "epoch": 3.3280367828526995, "grad_norm": 4.694197654724121, "learning_rate": 5.541357458244111e-05, "loss": 0.064, "step": 98440 }, { "epoch": 3.3283748605429526, "grad_norm": 0.8386731147766113, "learning_rate": 5.540535576013302e-05, "loss": 0.0584, "step": 98450 }, { "epoch": 3.328712938233206, "grad_norm": 0.9906351566314697, "learning_rate": 5.539713679004453e-05, "loss": 0.0591, "step": 98460 }, { "epoch": 3.329051015923459, "grad_norm": 1.2627891302108765, "learning_rate": 5.538891767240034e-05, "loss": 0.0599, "step": 98470 }, { "epoch": 3.3293890936137123, "grad_norm": 2.097404956817627, "learning_rate": 5.538069840742517e-05, "loss": 0.0609, "step": 98480 }, { "epoch": 3.329727171303966, "grad_norm": 1.5958633422851562, "learning_rate": 5.5372478995343726e-05, "loss": 0.0624, "step": 98490 }, { "epoch": 3.330065248994219, "grad_norm": 1.650900959968567, "learning_rate": 5.536425943638072e-05, "loss": 0.0616, "step": 98500 }, { "epoch": 3.330403326684472, "grad_norm": 0.9872956275939941, "learning_rate": 5.535603973076089e-05, "loss": 0.0665, "step": 98510 }, { "epoch": 3.3307414043747254, "grad_norm": 1.3217978477478027, "learning_rate": 5.534781987870893e-05, "loss": 0.0579, "step": 98520 }, { "epoch": 3.3310794820649785, "grad_norm": 0.7950149178504944, "learning_rate": 5.533959988044959e-05, "loss": 0.0573, "step": 98530 }, { "epoch": 3.3314175597552316, "grad_norm": 2.3362936973571777, "learning_rate": 5.53313797362076e-05, "loss": 0.0613, "step": 98540 }, { "epoch": 3.331755637445485, "grad_norm": 1.0641798973083496, "learning_rate": 5.532315944620769e-05, "loss": 0.062, "step": 98550 }, { "epoch": 3.332093715135738, "grad_norm": 1.3567453622817993, "learning_rate": 5.531493901067458e-05, "loss": 0.0624, "step": 98560 }, { "epoch": 3.3324317928259912, "grad_norm": 0.9070695042610168, "learning_rate": 5.530671842983305e-05, "loss": 0.0593, "step": 98570 }, { "epoch": 3.3327698705162447, "grad_norm": 0.9933878779411316, "learning_rate": 5.529849770390785e-05, "loss": 0.0641, "step": 98580 }, { "epoch": 3.333107948206498, "grad_norm": 1.8080815076828003, "learning_rate": 5.5290276833123686e-05, "loss": 0.0608, "step": 98590 }, { "epoch": 3.333446025896751, "grad_norm": 1.055767297744751, "learning_rate": 5.528205581770536e-05, "loss": 0.0627, "step": 98600 }, { "epoch": 3.3337841035870044, "grad_norm": 1.3426287174224854, "learning_rate": 5.527383465787759e-05, "loss": 0.0589, "step": 98610 }, { "epoch": 3.3341221812772575, "grad_norm": 0.48707661032676697, "learning_rate": 5.526561335386519e-05, "loss": 0.0555, "step": 98620 }, { "epoch": 3.3344602589675105, "grad_norm": 1.6131486892700195, "learning_rate": 5.525739190589287e-05, "loss": 0.0535, "step": 98630 }, { "epoch": 3.334798336657764, "grad_norm": 1.1770310401916504, "learning_rate": 5.524917031418546e-05, "loss": 0.0608, "step": 98640 }, { "epoch": 3.335136414348017, "grad_norm": 1.577340006828308, "learning_rate": 5.524094857896769e-05, "loss": 0.0561, "step": 98650 }, { "epoch": 3.33547449203827, "grad_norm": 1.8079466819763184, "learning_rate": 5.523272670046435e-05, "loss": 0.064, "step": 98660 }, { "epoch": 3.3358125697285237, "grad_norm": 0.9826188683509827, "learning_rate": 5.522450467890024e-05, "loss": 0.0704, "step": 98670 }, { "epoch": 3.3361506474187768, "grad_norm": 2.8956637382507324, "learning_rate": 5.521628251450012e-05, "loss": 0.0593, "step": 98680 }, { "epoch": 3.33648872510903, "grad_norm": 1.4427381753921509, "learning_rate": 5.52080602074888e-05, "loss": 0.0629, "step": 98690 }, { "epoch": 3.3368268027992833, "grad_norm": 0.5250252485275269, "learning_rate": 5.519983775809108e-05, "loss": 0.0605, "step": 98700 }, { "epoch": 3.3371648804895364, "grad_norm": 3.759683132171631, "learning_rate": 5.519161516653174e-05, "loss": 0.0638, "step": 98710 }, { "epoch": 3.3375029581797895, "grad_norm": 1.4286775588989258, "learning_rate": 5.518339243303561e-05, "loss": 0.061, "step": 98720 }, { "epoch": 3.337841035870043, "grad_norm": 1.1661580801010132, "learning_rate": 5.5175169557827454e-05, "loss": 0.0625, "step": 98730 }, { "epoch": 3.338179113560296, "grad_norm": 0.8101580142974854, "learning_rate": 5.5166946541132116e-05, "loss": 0.0602, "step": 98740 }, { "epoch": 3.338517191250549, "grad_norm": 17.98740577697754, "learning_rate": 5.5158723383174396e-05, "loss": 0.066, "step": 98750 }, { "epoch": 3.3388552689408026, "grad_norm": 0.9666897654533386, "learning_rate": 5.5150500084179124e-05, "loss": 0.0614, "step": 98760 }, { "epoch": 3.3391933466310557, "grad_norm": 3.368436574935913, "learning_rate": 5.5142276644371116e-05, "loss": 0.0646, "step": 98770 }, { "epoch": 3.3395314243213092, "grad_norm": 1.2032045125961304, "learning_rate": 5.513405306397519e-05, "loss": 0.0563, "step": 98780 }, { "epoch": 3.3398695020115623, "grad_norm": 1.0724600553512573, "learning_rate": 5.5125829343216184e-05, "loss": 0.0621, "step": 98790 }, { "epoch": 3.3402075797018154, "grad_norm": 1.7322357892990112, "learning_rate": 5.5117605482318927e-05, "loss": 0.0616, "step": 98800 }, { "epoch": 3.340545657392069, "grad_norm": 2.9744043350219727, "learning_rate": 5.510938148150826e-05, "loss": 0.069, "step": 98810 }, { "epoch": 3.340883735082322, "grad_norm": 1.133407711982727, "learning_rate": 5.510115734100902e-05, "loss": 0.0643, "step": 98820 }, { "epoch": 3.341221812772575, "grad_norm": 0.6009436845779419, "learning_rate": 5.509293306104607e-05, "loss": 0.0641, "step": 98830 }, { "epoch": 3.3415598904628285, "grad_norm": 1.4913092851638794, "learning_rate": 5.508470864184423e-05, "loss": 0.0587, "step": 98840 }, { "epoch": 3.3418979681530816, "grad_norm": 0.7142014503479004, "learning_rate": 5.507648408362838e-05, "loss": 0.0631, "step": 98850 }, { "epoch": 3.3422360458433347, "grad_norm": 0.8383044004440308, "learning_rate": 5.506825938662335e-05, "loss": 0.0579, "step": 98860 }, { "epoch": 3.342574123533588, "grad_norm": 2.7797420024871826, "learning_rate": 5.5060034551054e-05, "loss": 0.072, "step": 98870 }, { "epoch": 3.3429122012238413, "grad_norm": 1.03565514087677, "learning_rate": 5.5051809577145216e-05, "loss": 0.0639, "step": 98880 }, { "epoch": 3.3432502789140943, "grad_norm": 1.0639276504516602, "learning_rate": 5.5043584465121863e-05, "loss": 0.0677, "step": 98890 }, { "epoch": 3.343588356604348, "grad_norm": 0.9971518516540527, "learning_rate": 5.5035359215208805e-05, "loss": 0.0676, "step": 98900 }, { "epoch": 3.343926434294601, "grad_norm": 1.608657956123352, "learning_rate": 5.5027133827630916e-05, "loss": 0.0691, "step": 98910 }, { "epoch": 3.344264511984854, "grad_norm": 1.5194145441055298, "learning_rate": 5.501890830261307e-05, "loss": 0.0604, "step": 98920 }, { "epoch": 3.3446025896751075, "grad_norm": 0.8564951419830322, "learning_rate": 5.501068264038016e-05, "loss": 0.0608, "step": 98930 }, { "epoch": 3.3449406673653606, "grad_norm": 0.7913309335708618, "learning_rate": 5.5002456841157066e-05, "loss": 0.0615, "step": 98940 }, { "epoch": 3.3452787450556136, "grad_norm": 1.1926214694976807, "learning_rate": 5.49942309051687e-05, "loss": 0.0716, "step": 98950 }, { "epoch": 3.345616822745867, "grad_norm": 0.8600588440895081, "learning_rate": 5.4986004832639915e-05, "loss": 0.0589, "step": 98960 }, { "epoch": 3.34595490043612, "grad_norm": 2.3035504817962646, "learning_rate": 5.497777862379565e-05, "loss": 0.0614, "step": 98970 }, { "epoch": 3.3462929781263733, "grad_norm": 1.2198892831802368, "learning_rate": 5.496955227886078e-05, "loss": 0.0607, "step": 98980 }, { "epoch": 3.346631055816627, "grad_norm": 1.4284433126449585, "learning_rate": 5.4961325798060206e-05, "loss": 0.0623, "step": 98990 }, { "epoch": 3.34696913350688, "grad_norm": 1.5797526836395264, "learning_rate": 5.495309918161885e-05, "loss": 0.0546, "step": 99000 }, { "epoch": 3.347307211197133, "grad_norm": 2.475074291229248, "learning_rate": 5.494487242976164e-05, "loss": 0.0609, "step": 99010 }, { "epoch": 3.3476452888873864, "grad_norm": 0.9147442579269409, "learning_rate": 5.493664554271347e-05, "loss": 0.0609, "step": 99020 }, { "epoch": 3.3479833665776395, "grad_norm": 0.8150094151496887, "learning_rate": 5.492841852069926e-05, "loss": 0.0598, "step": 99030 }, { "epoch": 3.3483214442678926, "grad_norm": 1.0450268983840942, "learning_rate": 5.492019136394394e-05, "loss": 0.0575, "step": 99040 }, { "epoch": 3.348659521958146, "grad_norm": 1.3569010496139526, "learning_rate": 5.491196407267244e-05, "loss": 0.0558, "step": 99050 }, { "epoch": 3.348997599648399, "grad_norm": 0.9850186109542847, "learning_rate": 5.4903736647109684e-05, "loss": 0.0578, "step": 99060 }, { "epoch": 3.3493356773386522, "grad_norm": 0.6451071500778198, "learning_rate": 5.48955090874806e-05, "loss": 0.0574, "step": 99070 }, { "epoch": 3.3496737550289057, "grad_norm": 4.823431015014648, "learning_rate": 5.488728139401016e-05, "loss": 0.0687, "step": 99080 }, { "epoch": 3.350011832719159, "grad_norm": 0.9120978116989136, "learning_rate": 5.4879053566923264e-05, "loss": 0.0633, "step": 99090 }, { "epoch": 3.3503499104094123, "grad_norm": 0.6788836717605591, "learning_rate": 5.4870825606444885e-05, "loss": 0.0652, "step": 99100 }, { "epoch": 3.3506879880996654, "grad_norm": 1.1300824880599976, "learning_rate": 5.4862597512799964e-05, "loss": 0.0587, "step": 99110 }, { "epoch": 3.3510260657899185, "grad_norm": 0.8244874477386475, "learning_rate": 5.4854369286213446e-05, "loss": 0.0667, "step": 99120 }, { "epoch": 3.351364143480172, "grad_norm": 0.7227804660797119, "learning_rate": 5.4846140926910295e-05, "loss": 0.06, "step": 99130 }, { "epoch": 3.351702221170425, "grad_norm": 1.8858245611190796, "learning_rate": 5.483791243511549e-05, "loss": 0.0583, "step": 99140 }, { "epoch": 3.352040298860678, "grad_norm": 0.7010192275047302, "learning_rate": 5.482968381105396e-05, "loss": 0.0614, "step": 99150 }, { "epoch": 3.3523783765509316, "grad_norm": 0.6618947982788086, "learning_rate": 5.48214550549507e-05, "loss": 0.0647, "step": 99160 }, { "epoch": 3.3527164542411847, "grad_norm": 0.6100813746452332, "learning_rate": 5.4813226167030665e-05, "loss": 0.0537, "step": 99170 }, { "epoch": 3.3530545319314378, "grad_norm": 0.7836237549781799, "learning_rate": 5.480499714751883e-05, "loss": 0.0656, "step": 99180 }, { "epoch": 3.3533926096216913, "grad_norm": 0.9301638007164001, "learning_rate": 5.4796767996640175e-05, "loss": 0.0581, "step": 99190 }, { "epoch": 3.3537306873119443, "grad_norm": 0.9628056883811951, "learning_rate": 5.47885387146197e-05, "loss": 0.0585, "step": 99200 }, { "epoch": 3.3540687650021974, "grad_norm": 1.5688115358352661, "learning_rate": 5.4780309301682364e-05, "loss": 0.0573, "step": 99210 }, { "epoch": 3.354406842692451, "grad_norm": 0.9325270056724548, "learning_rate": 5.477207975805317e-05, "loss": 0.0529, "step": 99220 }, { "epoch": 3.354744920382704, "grad_norm": 3.6253178119659424, "learning_rate": 5.4763850083957115e-05, "loss": 0.0633, "step": 99230 }, { "epoch": 3.355082998072957, "grad_norm": 0.884874701499939, "learning_rate": 5.475562027961918e-05, "loss": 0.0607, "step": 99240 }, { "epoch": 3.3554210757632106, "grad_norm": 5.146923065185547, "learning_rate": 5.474739034526437e-05, "loss": 0.0613, "step": 99250 }, { "epoch": 3.3557591534534637, "grad_norm": 1.0427762269973755, "learning_rate": 5.473916028111771e-05, "loss": 0.0609, "step": 99260 }, { "epoch": 3.3560972311437167, "grad_norm": 2.964020252227783, "learning_rate": 5.473093008740416e-05, "loss": 0.0614, "step": 99270 }, { "epoch": 3.3564353088339702, "grad_norm": 1.780800223350525, "learning_rate": 5.472269976434877e-05, "loss": 0.0606, "step": 99280 }, { "epoch": 3.3567733865242233, "grad_norm": 1.0880557298660278, "learning_rate": 5.471446931217656e-05, "loss": 0.0653, "step": 99290 }, { "epoch": 3.3571114642144764, "grad_norm": 1.6081067323684692, "learning_rate": 5.4706238731112516e-05, "loss": 0.0621, "step": 99300 }, { "epoch": 3.35744954190473, "grad_norm": 3.1753101348876953, "learning_rate": 5.4698008021381665e-05, "loss": 0.0565, "step": 99310 }, { "epoch": 3.357787619594983, "grad_norm": 1.0172673463821411, "learning_rate": 5.468977718320905e-05, "loss": 0.0583, "step": 99320 }, { "epoch": 3.358125697285236, "grad_norm": 0.6092849373817444, "learning_rate": 5.468154621681969e-05, "loss": 0.0556, "step": 99330 }, { "epoch": 3.3584637749754895, "grad_norm": 1.1152116060256958, "learning_rate": 5.4673315122438604e-05, "loss": 0.0608, "step": 99340 }, { "epoch": 3.3588018526657426, "grad_norm": 1.2523905038833618, "learning_rate": 5.466508390029086e-05, "loss": 0.0621, "step": 99350 }, { "epoch": 3.3591399303559957, "grad_norm": 1.5459661483764648, "learning_rate": 5.465685255060146e-05, "loss": 0.0648, "step": 99360 }, { "epoch": 3.359478008046249, "grad_norm": 1.049788236618042, "learning_rate": 5.464862107359546e-05, "loss": 0.0618, "step": 99370 }, { "epoch": 3.3598160857365023, "grad_norm": 4.064384460449219, "learning_rate": 5.46403894694979e-05, "loss": 0.0624, "step": 99380 }, { "epoch": 3.3601541634267553, "grad_norm": 1.6157997846603394, "learning_rate": 5.463215773853385e-05, "loss": 0.062, "step": 99390 }, { "epoch": 3.360492241117009, "grad_norm": 0.7024680376052856, "learning_rate": 5.462392588092834e-05, "loss": 0.0557, "step": 99400 }, { "epoch": 3.360830318807262, "grad_norm": 0.7352449893951416, "learning_rate": 5.461569389690644e-05, "loss": 0.059, "step": 99410 }, { "epoch": 3.361168396497515, "grad_norm": 0.902134895324707, "learning_rate": 5.4607461786693195e-05, "loss": 0.0652, "step": 99420 }, { "epoch": 3.3615064741877685, "grad_norm": 1.2900362014770508, "learning_rate": 5.4599229550513685e-05, "loss": 0.0616, "step": 99430 }, { "epoch": 3.3618445518780216, "grad_norm": 1.2748380899429321, "learning_rate": 5.4590997188592954e-05, "loss": 0.063, "step": 99440 }, { "epoch": 3.3621826295682746, "grad_norm": 0.7331117987632751, "learning_rate": 5.4582764701156096e-05, "loss": 0.0685, "step": 99450 }, { "epoch": 3.362520707258528, "grad_norm": 0.8318500518798828, "learning_rate": 5.4574532088428164e-05, "loss": 0.0607, "step": 99460 }, { "epoch": 3.362858784948781, "grad_norm": 1.3348973989486694, "learning_rate": 5.456629935063425e-05, "loss": 0.0643, "step": 99470 }, { "epoch": 3.3631968626390343, "grad_norm": 0.7948240041732788, "learning_rate": 5.455806648799943e-05, "loss": 0.0601, "step": 99480 }, { "epoch": 3.363534940329288, "grad_norm": 0.7704846262931824, "learning_rate": 5.454983350074877e-05, "loss": 0.0664, "step": 99490 }, { "epoch": 3.363873018019541, "grad_norm": 0.7084330916404724, "learning_rate": 5.454160038910737e-05, "loss": 0.0639, "step": 99500 }, { "epoch": 3.364211095709794, "grad_norm": 1.0166548490524292, "learning_rate": 5.453336715330034e-05, "loss": 0.0615, "step": 99510 }, { "epoch": 3.3645491734000474, "grad_norm": 3.022005796432495, "learning_rate": 5.452513379355275e-05, "loss": 0.0561, "step": 99520 }, { "epoch": 3.3648872510903005, "grad_norm": 1.6668888330459595, "learning_rate": 5.451690031008969e-05, "loss": 0.0634, "step": 99530 }, { "epoch": 3.3652253287805536, "grad_norm": 1.1576074361801147, "learning_rate": 5.450866670313629e-05, "loss": 0.0583, "step": 99540 }, { "epoch": 3.365563406470807, "grad_norm": 0.9064756035804749, "learning_rate": 5.450043297291763e-05, "loss": 0.0594, "step": 99550 }, { "epoch": 3.36590148416106, "grad_norm": 1.502579689025879, "learning_rate": 5.4492199119658824e-05, "loss": 0.0629, "step": 99560 }, { "epoch": 3.3662395618513132, "grad_norm": 1.6428254842758179, "learning_rate": 5.448396514358497e-05, "loss": 0.0624, "step": 99570 }, { "epoch": 3.3665776395415667, "grad_norm": 1.5330668687820435, "learning_rate": 5.447573104492121e-05, "loss": 0.0607, "step": 99580 }, { "epoch": 3.36691571723182, "grad_norm": 0.4902097284793854, "learning_rate": 5.4467496823892626e-05, "loss": 0.0629, "step": 99590 }, { "epoch": 3.367253794922073, "grad_norm": 1.417394757270813, "learning_rate": 5.445926248072438e-05, "loss": 0.0652, "step": 99600 }, { "epoch": 3.3675918726123264, "grad_norm": 0.9056054949760437, "learning_rate": 5.445102801564156e-05, "loss": 0.0644, "step": 99610 }, { "epoch": 3.3679299503025795, "grad_norm": 0.9934442043304443, "learning_rate": 5.444279342886931e-05, "loss": 0.0692, "step": 99620 }, { "epoch": 3.3682680279928325, "grad_norm": 0.7076150178909302, "learning_rate": 5.443455872063274e-05, "loss": 0.0621, "step": 99630 }, { "epoch": 3.368606105683086, "grad_norm": 0.8934069275856018, "learning_rate": 5.442632389115703e-05, "loss": 0.0625, "step": 99640 }, { "epoch": 3.368944183373339, "grad_norm": 0.782667875289917, "learning_rate": 5.441808894066727e-05, "loss": 0.0566, "step": 99650 }, { "epoch": 3.369282261063592, "grad_norm": 2.0185723304748535, "learning_rate": 5.4409853869388614e-05, "loss": 0.0591, "step": 99660 }, { "epoch": 3.3696203387538457, "grad_norm": 1.8740538358688354, "learning_rate": 5.4401618677546215e-05, "loss": 0.0589, "step": 99670 }, { "epoch": 3.3699584164440988, "grad_norm": 0.6569246649742126, "learning_rate": 5.439338336536522e-05, "loss": 0.0564, "step": 99680 }, { "epoch": 3.370296494134352, "grad_norm": 0.9121879935264587, "learning_rate": 5.438514793307077e-05, "loss": 0.0612, "step": 99690 }, { "epoch": 3.3706345718246054, "grad_norm": 2.0323071479797363, "learning_rate": 5.437691238088802e-05, "loss": 0.0647, "step": 99700 }, { "epoch": 3.3709726495148584, "grad_norm": 1.0224132537841797, "learning_rate": 5.436867670904213e-05, "loss": 0.0557, "step": 99710 }, { "epoch": 3.3713107272051115, "grad_norm": 0.691131055355072, "learning_rate": 5.436044091775827e-05, "loss": 0.0589, "step": 99720 }, { "epoch": 3.371648804895365, "grad_norm": 1.4016577005386353, "learning_rate": 5.435220500726159e-05, "loss": 0.0626, "step": 99730 }, { "epoch": 3.371986882585618, "grad_norm": 9.19621467590332, "learning_rate": 5.4343968977777246e-05, "loss": 0.0546, "step": 99740 }, { "epoch": 3.3723249602758716, "grad_norm": 1.8466644287109375, "learning_rate": 5.433573282953045e-05, "loss": 0.0547, "step": 99750 }, { "epoch": 3.3726630379661247, "grad_norm": 1.4186081886291504, "learning_rate": 5.432749656274633e-05, "loss": 0.0575, "step": 99760 }, { "epoch": 3.3730011156563777, "grad_norm": 1.8948012590408325, "learning_rate": 5.431926017765007e-05, "loss": 0.0626, "step": 99770 }, { "epoch": 3.3733391933466312, "grad_norm": 2.3753178119659424, "learning_rate": 5.431102367446686e-05, "loss": 0.0563, "step": 99780 }, { "epoch": 3.3736772710368843, "grad_norm": 0.8655813932418823, "learning_rate": 5.430278705342191e-05, "loss": 0.0617, "step": 99790 }, { "epoch": 3.3740153487271374, "grad_norm": 2.0213782787323, "learning_rate": 5.429455031474036e-05, "loss": 0.0676, "step": 99800 }, { "epoch": 3.374353426417391, "grad_norm": 2.1829373836517334, "learning_rate": 5.428631345864742e-05, "loss": 0.0654, "step": 99810 }, { "epoch": 3.374691504107644, "grad_norm": 1.032371163368225, "learning_rate": 5.4278076485368255e-05, "loss": 0.0661, "step": 99820 }, { "epoch": 3.375029581797897, "grad_norm": 1.6270878314971924, "learning_rate": 5.426983939512812e-05, "loss": 0.0611, "step": 99830 }, { "epoch": 3.3753676594881505, "grad_norm": 0.9289358258247375, "learning_rate": 5.4261602188152175e-05, "loss": 0.0639, "step": 99840 }, { "epoch": 3.3757057371784036, "grad_norm": 4.607096195220947, "learning_rate": 5.4253364864665625e-05, "loss": 0.0615, "step": 99850 }, { "epoch": 3.3760438148686567, "grad_norm": 1.8468571901321411, "learning_rate": 5.424512742489367e-05, "loss": 0.0629, "step": 99860 }, { "epoch": 3.37638189255891, "grad_norm": 2.6912639141082764, "learning_rate": 5.4236889869061536e-05, "loss": 0.0587, "step": 99870 }, { "epoch": 3.3767199702491633, "grad_norm": 1.0391805171966553, "learning_rate": 5.4228652197394415e-05, "loss": 0.0745, "step": 99880 }, { "epoch": 3.3770580479394163, "grad_norm": 1.2653173208236694, "learning_rate": 5.4220414410117536e-05, "loss": 0.0668, "step": 99890 }, { "epoch": 3.37739612562967, "grad_norm": 1.3555361032485962, "learning_rate": 5.42121765074561e-05, "loss": 0.0674, "step": 99900 }, { "epoch": 3.377734203319923, "grad_norm": 0.6595315933227539, "learning_rate": 5.4203938489635366e-05, "loss": 0.0584, "step": 99910 }, { "epoch": 3.378072281010176, "grad_norm": 0.7578237652778625, "learning_rate": 5.419570035688052e-05, "loss": 0.0618, "step": 99920 }, { "epoch": 3.3784103587004295, "grad_norm": 1.4493598937988281, "learning_rate": 5.418746210941681e-05, "loss": 0.0609, "step": 99930 }, { "epoch": 3.3787484363906826, "grad_norm": 2.192361354827881, "learning_rate": 5.417922374746944e-05, "loss": 0.0573, "step": 99940 }, { "epoch": 3.3790865140809356, "grad_norm": 1.0584887266159058, "learning_rate": 5.4170985271263676e-05, "loss": 0.0629, "step": 99950 }, { "epoch": 3.379424591771189, "grad_norm": 1.378668189048767, "learning_rate": 5.416274668102474e-05, "loss": 0.0599, "step": 99960 }, { "epoch": 3.379762669461442, "grad_norm": 0.7906145453453064, "learning_rate": 5.4154507976977875e-05, "loss": 0.0524, "step": 99970 }, { "epoch": 3.3801007471516953, "grad_norm": 1.1873345375061035, "learning_rate": 5.414626915934832e-05, "loss": 0.0536, "step": 99980 }, { "epoch": 3.380438824841949, "grad_norm": 2.379396438598633, "learning_rate": 5.413803022836133e-05, "loss": 0.0621, "step": 99990 }, { "epoch": 3.380776902532202, "grad_norm": 1.183734655380249, "learning_rate": 5.412979118424214e-05, "loss": 0.0611, "step": 100000 }, { "epoch": 3.381114980222455, "grad_norm": 1.0047322511672974, "learning_rate": 5.412155202721602e-05, "loss": 0.0626, "step": 100010 }, { "epoch": 3.3814530579127084, "grad_norm": 3.0004870891571045, "learning_rate": 5.4113312757508194e-05, "loss": 0.0611, "step": 100020 }, { "epoch": 3.3817911356029615, "grad_norm": 1.6726348400115967, "learning_rate": 5.410507337534396e-05, "loss": 0.06, "step": 100030 }, { "epoch": 3.3821292132932146, "grad_norm": 3.1490938663482666, "learning_rate": 5.409683388094857e-05, "loss": 0.0625, "step": 100040 }, { "epoch": 3.382467290983468, "grad_norm": 1.5171141624450684, "learning_rate": 5.4088594274547264e-05, "loss": 0.0683, "step": 100050 }, { "epoch": 3.382805368673721, "grad_norm": 0.8746604323387146, "learning_rate": 5.4080354556365344e-05, "loss": 0.057, "step": 100060 }, { "epoch": 3.3831434463639747, "grad_norm": 2.537346839904785, "learning_rate": 5.407211472662805e-05, "loss": 0.0638, "step": 100070 }, { "epoch": 3.3834815240542278, "grad_norm": 0.9441570043563843, "learning_rate": 5.406387478556067e-05, "loss": 0.0564, "step": 100080 }, { "epoch": 3.383819601744481, "grad_norm": 1.2142027616500854, "learning_rate": 5.405563473338847e-05, "loss": 0.0611, "step": 100090 }, { "epoch": 3.3841576794347343, "grad_norm": 1.7621787786483765, "learning_rate": 5.4047394570336764e-05, "loss": 0.0693, "step": 100100 }, { "epoch": 3.3844957571249874, "grad_norm": 1.3876140117645264, "learning_rate": 5.40391542966308e-05, "loss": 0.0576, "step": 100110 }, { "epoch": 3.3848338348152405, "grad_norm": 3.0864789485931396, "learning_rate": 5.403091391249588e-05, "loss": 0.0667, "step": 100120 }, { "epoch": 3.385171912505494, "grad_norm": 2.091062068939209, "learning_rate": 5.4022673418157276e-05, "loss": 0.0687, "step": 100130 }, { "epoch": 3.385509990195747, "grad_norm": 0.755165159702301, "learning_rate": 5.40144328138403e-05, "loss": 0.0562, "step": 100140 }, { "epoch": 3.385848067886, "grad_norm": 0.9931056499481201, "learning_rate": 5.4006192099770225e-05, "loss": 0.064, "step": 100150 }, { "epoch": 3.3861861455762536, "grad_norm": 4.875959873199463, "learning_rate": 5.399795127617239e-05, "loss": 0.0651, "step": 100160 }, { "epoch": 3.3865242232665067, "grad_norm": 1.1400058269500732, "learning_rate": 5.398971034327205e-05, "loss": 0.068, "step": 100170 }, { "epoch": 3.3868623009567598, "grad_norm": 1.1001631021499634, "learning_rate": 5.398146930129454e-05, "loss": 0.063, "step": 100180 }, { "epoch": 3.3872003786470133, "grad_norm": 2.913877010345459, "learning_rate": 5.397322815046515e-05, "loss": 0.0645, "step": 100190 }, { "epoch": 3.3875384563372664, "grad_norm": 1.1902614831924438, "learning_rate": 5.396498689100918e-05, "loss": 0.061, "step": 100200 }, { "epoch": 3.3878765340275194, "grad_norm": 3.4552276134490967, "learning_rate": 5.395674552315198e-05, "loss": 0.0671, "step": 100210 }, { "epoch": 3.388214611717773, "grad_norm": 0.9851881861686707, "learning_rate": 5.394850404711883e-05, "loss": 0.0668, "step": 100220 }, { "epoch": 3.388552689408026, "grad_norm": 1.338441252708435, "learning_rate": 5.394026246313506e-05, "loss": 0.0608, "step": 100230 }, { "epoch": 3.388890767098279, "grad_norm": 2.0430829524993896, "learning_rate": 5.3932020771426006e-05, "loss": 0.0776, "step": 100240 }, { "epoch": 3.3892288447885326, "grad_norm": 3.335960626602173, "learning_rate": 5.392377897221699e-05, "loss": 0.069, "step": 100250 }, { "epoch": 3.3895669224787857, "grad_norm": 3.7847795486450195, "learning_rate": 5.391553706573331e-05, "loss": 0.0638, "step": 100260 }, { "epoch": 3.3899050001690387, "grad_norm": 2.0180628299713135, "learning_rate": 5.390729505220032e-05, "loss": 0.0619, "step": 100270 }, { "epoch": 3.3902430778592922, "grad_norm": 1.2812535762786865, "learning_rate": 5.389905293184335e-05, "loss": 0.0689, "step": 100280 }, { "epoch": 3.3905811555495453, "grad_norm": 1.2447260618209839, "learning_rate": 5.389081070488775e-05, "loss": 0.0649, "step": 100290 }, { "epoch": 3.3909192332397984, "grad_norm": 1.7127420902252197, "learning_rate": 5.388256837155884e-05, "loss": 0.0641, "step": 100300 }, { "epoch": 3.391257310930052, "grad_norm": 1.3652935028076172, "learning_rate": 5.387432593208198e-05, "loss": 0.0596, "step": 100310 }, { "epoch": 3.391595388620305, "grad_norm": 0.6811684370040894, "learning_rate": 5.3866083386682496e-05, "loss": 0.0581, "step": 100320 }, { "epoch": 3.391933466310558, "grad_norm": 1.2160907983779907, "learning_rate": 5.385784073558574e-05, "loss": 0.0576, "step": 100330 }, { "epoch": 3.3922715440008115, "grad_norm": 3.0763657093048096, "learning_rate": 5.3849597979017064e-05, "loss": 0.0574, "step": 100340 }, { "epoch": 3.3926096216910646, "grad_norm": 1.2074074745178223, "learning_rate": 5.384135511720184e-05, "loss": 0.0633, "step": 100350 }, { "epoch": 3.3929476993813177, "grad_norm": 0.7678217887878418, "learning_rate": 5.383311215036541e-05, "loss": 0.0554, "step": 100360 }, { "epoch": 3.393285777071571, "grad_norm": 0.9215506911277771, "learning_rate": 5.382486907873313e-05, "loss": 0.0607, "step": 100370 }, { "epoch": 3.3936238547618243, "grad_norm": 1.198360562324524, "learning_rate": 5.381662590253036e-05, "loss": 0.0583, "step": 100380 }, { "epoch": 3.3939619324520773, "grad_norm": 3.5492403507232666, "learning_rate": 5.3808382621982475e-05, "loss": 0.0576, "step": 100390 }, { "epoch": 3.394300010142331, "grad_norm": 2.5958428382873535, "learning_rate": 5.380013923731483e-05, "loss": 0.0613, "step": 100400 }, { "epoch": 3.394638087832584, "grad_norm": 2.023578643798828, "learning_rate": 5.379189574875282e-05, "loss": 0.0632, "step": 100410 }, { "epoch": 3.394976165522837, "grad_norm": 1.2532457113265991, "learning_rate": 5.37836521565218e-05, "loss": 0.064, "step": 100420 }, { "epoch": 3.3953142432130905, "grad_norm": 1.0228626728057861, "learning_rate": 5.377540846084716e-05, "loss": 0.0633, "step": 100430 }, { "epoch": 3.3956523209033436, "grad_norm": 1.0826900005340576, "learning_rate": 5.376716466195426e-05, "loss": 0.0611, "step": 100440 }, { "epoch": 3.3959903985935966, "grad_norm": 2.3361597061157227, "learning_rate": 5.37589207600685e-05, "loss": 0.0618, "step": 100450 }, { "epoch": 3.39632847628385, "grad_norm": 0.6289465427398682, "learning_rate": 5.375067675541524e-05, "loss": 0.0587, "step": 100460 }, { "epoch": 3.396666553974103, "grad_norm": 1.305018663406372, "learning_rate": 5.374243264821991e-05, "loss": 0.063, "step": 100470 }, { "epoch": 3.3970046316643563, "grad_norm": 2.817107915878296, "learning_rate": 5.3734188438707866e-05, "loss": 0.059, "step": 100480 }, { "epoch": 3.39734270935461, "grad_norm": 1.3559017181396484, "learning_rate": 5.3725944127104524e-05, "loss": 0.0644, "step": 100490 }, { "epoch": 3.397680787044863, "grad_norm": 0.7102125883102417, "learning_rate": 5.371769971363526e-05, "loss": 0.0606, "step": 100500 }, { "epoch": 3.398018864735116, "grad_norm": 1.4405906200408936, "learning_rate": 5.370945519852548e-05, "loss": 0.0633, "step": 100510 }, { "epoch": 3.3983569424253695, "grad_norm": 0.8121451139450073, "learning_rate": 5.3701210582000594e-05, "loss": 0.0672, "step": 100520 }, { "epoch": 3.3986950201156225, "grad_norm": 3.438530445098877, "learning_rate": 5.369296586428599e-05, "loss": 0.0644, "step": 100530 }, { "epoch": 3.3990330978058756, "grad_norm": 1.1102131605148315, "learning_rate": 5.368472104560711e-05, "loss": 0.0565, "step": 100540 }, { "epoch": 3.399371175496129, "grad_norm": 1.0244125127792358, "learning_rate": 5.367647612618932e-05, "loss": 0.057, "step": 100550 }, { "epoch": 3.399709253186382, "grad_norm": 0.8569086790084839, "learning_rate": 5.366823110625807e-05, "loss": 0.0595, "step": 100560 }, { "epoch": 3.4000473308766352, "grad_norm": 1.3592125177383423, "learning_rate": 5.365998598603874e-05, "loss": 0.0564, "step": 100570 }, { "epoch": 3.4003854085668888, "grad_norm": 3.9760289192199707, "learning_rate": 5.36517407657568e-05, "loss": 0.0585, "step": 100580 }, { "epoch": 3.400723486257142, "grad_norm": 0.8758002519607544, "learning_rate": 5.3643495445637595e-05, "loss": 0.0617, "step": 100590 }, { "epoch": 3.401061563947395, "grad_norm": 0.9897269010543823, "learning_rate": 5.3635250025906626e-05, "loss": 0.0598, "step": 100600 }, { "epoch": 3.4013996416376484, "grad_norm": 2.312039375305176, "learning_rate": 5.362700450678927e-05, "loss": 0.0704, "step": 100610 }, { "epoch": 3.4017377193279015, "grad_norm": 0.671046793460846, "learning_rate": 5.3618758888510986e-05, "loss": 0.0618, "step": 100620 }, { "epoch": 3.4020757970181545, "grad_norm": 0.9826616048812866, "learning_rate": 5.3610513171297194e-05, "loss": 0.0627, "step": 100630 }, { "epoch": 3.402413874708408, "grad_norm": 0.5398849844932556, "learning_rate": 5.360226735537333e-05, "loss": 0.0583, "step": 100640 }, { "epoch": 3.402751952398661, "grad_norm": 1.0765817165374756, "learning_rate": 5.35940214409648e-05, "loss": 0.0643, "step": 100650 }, { "epoch": 3.403090030088914, "grad_norm": 1.5245044231414795, "learning_rate": 5.35857754282971e-05, "loss": 0.0593, "step": 100660 }, { "epoch": 3.4034281077791677, "grad_norm": 1.5634161233901978, "learning_rate": 5.357752931759565e-05, "loss": 0.0554, "step": 100670 }, { "epoch": 3.403766185469421, "grad_norm": 1.5205198526382446, "learning_rate": 5.3569283109085886e-05, "loss": 0.0601, "step": 100680 }, { "epoch": 3.404104263159674, "grad_norm": 1.0168434381484985, "learning_rate": 5.356103680299327e-05, "loss": 0.0598, "step": 100690 }, { "epoch": 3.4044423408499274, "grad_norm": 0.7326715588569641, "learning_rate": 5.355279039954323e-05, "loss": 0.0635, "step": 100700 }, { "epoch": 3.4047804185401804, "grad_norm": 0.9646532535552979, "learning_rate": 5.3544543898961244e-05, "loss": 0.0639, "step": 100710 }, { "epoch": 3.405118496230434, "grad_norm": 1.373719334602356, "learning_rate": 5.353629730147276e-05, "loss": 0.0589, "step": 100720 }, { "epoch": 3.405456573920687, "grad_norm": 2.0859739780426025, "learning_rate": 5.352805060730324e-05, "loss": 0.069, "step": 100730 }, { "epoch": 3.40579465161094, "grad_norm": 0.9966009259223938, "learning_rate": 5.351980381667813e-05, "loss": 0.0623, "step": 100740 }, { "epoch": 3.4061327293011936, "grad_norm": 0.9671099781990051, "learning_rate": 5.3511556929822914e-05, "loss": 0.059, "step": 100750 }, { "epoch": 3.4064708069914467, "grad_norm": 1.26337730884552, "learning_rate": 5.350330994696305e-05, "loss": 0.0605, "step": 100760 }, { "epoch": 3.4068088846816997, "grad_norm": 1.1829313039779663, "learning_rate": 5.349506286832402e-05, "loss": 0.0589, "step": 100770 }, { "epoch": 3.4071469623719532, "grad_norm": 0.9825907349586487, "learning_rate": 5.3486815694131254e-05, "loss": 0.0595, "step": 100780 }, { "epoch": 3.4074850400622063, "grad_norm": 1.1301332712173462, "learning_rate": 5.347856842461028e-05, "loss": 0.0607, "step": 100790 }, { "epoch": 3.4078231177524594, "grad_norm": 1.506287693977356, "learning_rate": 5.347032105998654e-05, "loss": 0.0672, "step": 100800 }, { "epoch": 3.408161195442713, "grad_norm": 1.6749426126480103, "learning_rate": 5.346207360048554e-05, "loss": 0.054, "step": 100810 }, { "epoch": 3.408499273132966, "grad_norm": 0.6784353256225586, "learning_rate": 5.345382604633275e-05, "loss": 0.0578, "step": 100820 }, { "epoch": 3.408837350823219, "grad_norm": 1.5712330341339111, "learning_rate": 5.344557839775365e-05, "loss": 0.0665, "step": 100830 }, { "epoch": 3.4091754285134725, "grad_norm": 1.4087079763412476, "learning_rate": 5.343733065497371e-05, "loss": 0.0655, "step": 100840 }, { "epoch": 3.4095135062037256, "grad_norm": 1.6339155435562134, "learning_rate": 5.3429082818218456e-05, "loss": 0.0607, "step": 100850 }, { "epoch": 3.4098515838939787, "grad_norm": 1.2029180526733398, "learning_rate": 5.3420834887713366e-05, "loss": 0.0634, "step": 100860 }, { "epoch": 3.410189661584232, "grad_norm": 0.6439626812934875, "learning_rate": 5.341258686368395e-05, "loss": 0.0524, "step": 100870 }, { "epoch": 3.4105277392744853, "grad_norm": 2.4907755851745605, "learning_rate": 5.340433874635568e-05, "loss": 0.0727, "step": 100880 }, { "epoch": 3.4108658169647383, "grad_norm": 1.7644331455230713, "learning_rate": 5.339609053595407e-05, "loss": 0.0593, "step": 100890 }, { "epoch": 3.411203894654992, "grad_norm": 1.4466272592544556, "learning_rate": 5.338784223270461e-05, "loss": 0.0558, "step": 100900 }, { "epoch": 3.411541972345245, "grad_norm": 3.0825934410095215, "learning_rate": 5.337959383683282e-05, "loss": 0.0624, "step": 100910 }, { "epoch": 3.411880050035498, "grad_norm": 1.3270817995071411, "learning_rate": 5.337134534856419e-05, "loss": 0.0652, "step": 100920 }, { "epoch": 3.4122181277257515, "grad_norm": 3.026658773422241, "learning_rate": 5.336309676812428e-05, "loss": 0.0595, "step": 100930 }, { "epoch": 3.4125562054160046, "grad_norm": 3.0381417274475098, "learning_rate": 5.335484809573854e-05, "loss": 0.0629, "step": 100940 }, { "epoch": 3.4128942831062576, "grad_norm": 2.1820051670074463, "learning_rate": 5.334659933163252e-05, "loss": 0.0621, "step": 100950 }, { "epoch": 3.413232360796511, "grad_norm": 2.332335948944092, "learning_rate": 5.333835047603174e-05, "loss": 0.0621, "step": 100960 }, { "epoch": 3.4135704384867642, "grad_norm": 0.9564212560653687, "learning_rate": 5.3330101529161705e-05, "loss": 0.0644, "step": 100970 }, { "epoch": 3.4139085161770173, "grad_norm": 1.0421782732009888, "learning_rate": 5.3321852491247936e-05, "loss": 0.0617, "step": 100980 }, { "epoch": 3.414246593867271, "grad_norm": 0.9673023819923401, "learning_rate": 5.331360336251597e-05, "loss": 0.063, "step": 100990 }, { "epoch": 3.414584671557524, "grad_norm": 1.6049546003341675, "learning_rate": 5.3305354143191354e-05, "loss": 0.0676, "step": 101000 }, { "epoch": 3.414922749247777, "grad_norm": 2.1406590938568115, "learning_rate": 5.329710483349959e-05, "loss": 0.063, "step": 101010 }, { "epoch": 3.4152608269380305, "grad_norm": 1.6978720426559448, "learning_rate": 5.328885543366622e-05, "loss": 0.0603, "step": 101020 }, { "epoch": 3.4155989046282835, "grad_norm": 1.2864402532577515, "learning_rate": 5.328060594391675e-05, "loss": 0.0563, "step": 101030 }, { "epoch": 3.415936982318537, "grad_norm": 1.3830987215042114, "learning_rate": 5.3272356364476774e-05, "loss": 0.0568, "step": 101040 }, { "epoch": 3.41627506000879, "grad_norm": 2.3737123012542725, "learning_rate": 5.3264106695571795e-05, "loss": 0.064, "step": 101050 }, { "epoch": 3.416613137699043, "grad_norm": 1.0990642309188843, "learning_rate": 5.3255856937427374e-05, "loss": 0.0592, "step": 101060 }, { "epoch": 3.4169512153892967, "grad_norm": 1.646950125694275, "learning_rate": 5.3247607090269034e-05, "loss": 0.0587, "step": 101070 }, { "epoch": 3.4172892930795498, "grad_norm": 2.4214015007019043, "learning_rate": 5.3239357154322355e-05, "loss": 0.0549, "step": 101080 }, { "epoch": 3.417627370769803, "grad_norm": 0.9697939157485962, "learning_rate": 5.323110712981285e-05, "loss": 0.0669, "step": 101090 }, { "epoch": 3.4179654484600563, "grad_norm": 1.4417309761047363, "learning_rate": 5.32228570169661e-05, "loss": 0.0553, "step": 101100 }, { "epoch": 3.4183035261503094, "grad_norm": 0.7497163414955139, "learning_rate": 5.321460681600764e-05, "loss": 0.0646, "step": 101110 }, { "epoch": 3.4186416038405625, "grad_norm": 1.3526428937911987, "learning_rate": 5.320635652716305e-05, "loss": 0.0632, "step": 101120 }, { "epoch": 3.418979681530816, "grad_norm": 0.5944687724113464, "learning_rate": 5.3198106150657865e-05, "loss": 0.062, "step": 101130 }, { "epoch": 3.419317759221069, "grad_norm": 3.061255931854248, "learning_rate": 5.318985568671767e-05, "loss": 0.0618, "step": 101140 }, { "epoch": 3.419655836911322, "grad_norm": 1.2950823307037354, "learning_rate": 5.3181605135568016e-05, "loss": 0.0543, "step": 101150 }, { "epoch": 3.4199939146015756, "grad_norm": 0.7170599102973938, "learning_rate": 5.3173354497434466e-05, "loss": 0.0571, "step": 101160 }, { "epoch": 3.4203319922918287, "grad_norm": 1.397627353668213, "learning_rate": 5.31651037725426e-05, "loss": 0.063, "step": 101170 }, { "epoch": 3.420670069982082, "grad_norm": 1.4031733274459839, "learning_rate": 5.315685296111799e-05, "loss": 0.0632, "step": 101180 }, { "epoch": 3.4210081476723353, "grad_norm": 0.8074166178703308, "learning_rate": 5.31486020633862e-05, "loss": 0.0634, "step": 101190 }, { "epoch": 3.4213462253625884, "grad_norm": 7.100984573364258, "learning_rate": 5.314035107957281e-05, "loss": 0.0626, "step": 101200 }, { "epoch": 3.4216843030528414, "grad_norm": 2.3350002765655518, "learning_rate": 5.313210000990342e-05, "loss": 0.0665, "step": 101210 }, { "epoch": 3.422022380743095, "grad_norm": 1.2332714796066284, "learning_rate": 5.312384885460358e-05, "loss": 0.0623, "step": 101220 }, { "epoch": 3.422360458433348, "grad_norm": 13.982069969177246, "learning_rate": 5.3115597613898874e-05, "loss": 0.0644, "step": 101230 }, { "epoch": 3.422698536123601, "grad_norm": 2.0383288860321045, "learning_rate": 5.3107346288014903e-05, "loss": 0.0555, "step": 101240 }, { "epoch": 3.4230366138138546, "grad_norm": 1.9077157974243164, "learning_rate": 5.309909487717727e-05, "loss": 0.059, "step": 101250 }, { "epoch": 3.4233746915041077, "grad_norm": 1.850370168685913, "learning_rate": 5.309084338161153e-05, "loss": 0.0598, "step": 101260 }, { "epoch": 3.4237127691943607, "grad_norm": 0.9121135473251343, "learning_rate": 5.3082591801543315e-05, "loss": 0.0562, "step": 101270 }, { "epoch": 3.4240508468846143, "grad_norm": 0.8771483898162842, "learning_rate": 5.307434013719817e-05, "loss": 0.059, "step": 101280 }, { "epoch": 3.4243889245748673, "grad_norm": 2.2658462524414062, "learning_rate": 5.306608838880174e-05, "loss": 0.0604, "step": 101290 }, { "epoch": 3.4247270022651204, "grad_norm": 0.9054294228553772, "learning_rate": 5.30578365565796e-05, "loss": 0.0661, "step": 101300 }, { "epoch": 3.425065079955374, "grad_norm": 2.6727256774902344, "learning_rate": 5.3049584640757355e-05, "loss": 0.0658, "step": 101310 }, { "epoch": 3.425403157645627, "grad_norm": 0.7842798829078674, "learning_rate": 5.3041332641560614e-05, "loss": 0.0648, "step": 101320 }, { "epoch": 3.42574123533588, "grad_norm": 1.755202293395996, "learning_rate": 5.3033080559214985e-05, "loss": 0.0619, "step": 101330 }, { "epoch": 3.4260793130261336, "grad_norm": 1.7482551336288452, "learning_rate": 5.302482839394607e-05, "loss": 0.0573, "step": 101340 }, { "epoch": 3.4264173907163866, "grad_norm": 2.5588557720184326, "learning_rate": 5.301657614597947e-05, "loss": 0.0656, "step": 101350 }, { "epoch": 3.4267554684066397, "grad_norm": 2.4428958892822266, "learning_rate": 5.3008323815540826e-05, "loss": 0.07, "step": 101360 }, { "epoch": 3.427093546096893, "grad_norm": 0.9401923418045044, "learning_rate": 5.300007140285574e-05, "loss": 0.0606, "step": 101370 }, { "epoch": 3.4274316237871463, "grad_norm": 1.808182716369629, "learning_rate": 5.299181890814983e-05, "loss": 0.061, "step": 101380 }, { "epoch": 3.4277697014773993, "grad_norm": 3.8847031593322754, "learning_rate": 5.29835663316487e-05, "loss": 0.0572, "step": 101390 }, { "epoch": 3.428107779167653, "grad_norm": 1.4500212669372559, "learning_rate": 5.2975313673578e-05, "loss": 0.0629, "step": 101400 }, { "epoch": 3.428445856857906, "grad_norm": 2.621788501739502, "learning_rate": 5.296706093416334e-05, "loss": 0.0557, "step": 101410 }, { "epoch": 3.428783934548159, "grad_norm": 1.2717444896697998, "learning_rate": 5.2958808113630344e-05, "loss": 0.0652, "step": 101420 }, { "epoch": 3.4291220122384125, "grad_norm": 1.8459696769714355, "learning_rate": 5.2950555212204655e-05, "loss": 0.06, "step": 101430 }, { "epoch": 3.4294600899286656, "grad_norm": 1.2739222049713135, "learning_rate": 5.2942302230111884e-05, "loss": 0.0615, "step": 101440 }, { "epoch": 3.4297981676189186, "grad_norm": 3.344219923019409, "learning_rate": 5.293404916757768e-05, "loss": 0.0658, "step": 101450 }, { "epoch": 3.430136245309172, "grad_norm": 2.4220845699310303, "learning_rate": 5.292579602482768e-05, "loss": 0.0634, "step": 101460 }, { "epoch": 3.4304743229994252, "grad_norm": 2.1057393550872803, "learning_rate": 5.2917542802087506e-05, "loss": 0.062, "step": 101470 }, { "epoch": 3.4308124006896783, "grad_norm": 2.644641399383545, "learning_rate": 5.290928949958281e-05, "loss": 0.0581, "step": 101480 }, { "epoch": 3.431150478379932, "grad_norm": 4.600888252258301, "learning_rate": 5.290103611753924e-05, "loss": 0.0714, "step": 101490 }, { "epoch": 3.431488556070185, "grad_norm": 3.086080312728882, "learning_rate": 5.2892782656182425e-05, "loss": 0.0623, "step": 101500 }, { "epoch": 3.431826633760438, "grad_norm": 1.7131845951080322, "learning_rate": 5.2884529115738024e-05, "loss": 0.0772, "step": 101510 }, { "epoch": 3.4321647114506915, "grad_norm": 2.9090936183929443, "learning_rate": 5.2876275496431684e-05, "loss": 0.0624, "step": 101520 }, { "epoch": 3.4325027891409445, "grad_norm": 2.479135274887085, "learning_rate": 5.286802179848904e-05, "loss": 0.0654, "step": 101530 }, { "epoch": 3.4328408668311976, "grad_norm": 3.6384551525115967, "learning_rate": 5.285976802213576e-05, "loss": 0.0616, "step": 101540 }, { "epoch": 3.433178944521451, "grad_norm": 0.8800154328346252, "learning_rate": 5.28515141675975e-05, "loss": 0.0661, "step": 101550 }, { "epoch": 3.433517022211704, "grad_norm": 2.66872501373291, "learning_rate": 5.2843260235099924e-05, "loss": 0.0638, "step": 101560 }, { "epoch": 3.4338550999019573, "grad_norm": 2.0957205295562744, "learning_rate": 5.283500622486867e-05, "loss": 0.0661, "step": 101570 }, { "epoch": 3.4341931775922108, "grad_norm": 1.5752109289169312, "learning_rate": 5.282675213712942e-05, "loss": 0.0628, "step": 101580 }, { "epoch": 3.434531255282464, "grad_norm": 1.1879477500915527, "learning_rate": 5.281849797210782e-05, "loss": 0.0577, "step": 101590 }, { "epoch": 3.434869332972717, "grad_norm": 1.3111549615859985, "learning_rate": 5.281024373002954e-05, "loss": 0.0591, "step": 101600 }, { "epoch": 3.4352074106629704, "grad_norm": 1.2656153440475464, "learning_rate": 5.2801989411120245e-05, "loss": 0.0667, "step": 101610 }, { "epoch": 3.4355454883532235, "grad_norm": 2.6708457469940186, "learning_rate": 5.279373501560564e-05, "loss": 0.0611, "step": 101620 }, { "epoch": 3.4358835660434766, "grad_norm": 1.321882724761963, "learning_rate": 5.2785480543711353e-05, "loss": 0.0689, "step": 101630 }, { "epoch": 3.43622164373373, "grad_norm": 1.4952797889709473, "learning_rate": 5.2777225995663074e-05, "loss": 0.0661, "step": 101640 }, { "epoch": 3.436559721423983, "grad_norm": 5.102287769317627, "learning_rate": 5.276897137168648e-05, "loss": 0.0662, "step": 101650 }, { "epoch": 3.436897799114236, "grad_norm": 1.4529502391815186, "learning_rate": 5.276071667200725e-05, "loss": 0.0578, "step": 101660 }, { "epoch": 3.4372358768044897, "grad_norm": 1.9158016443252563, "learning_rate": 5.2752461896851055e-05, "loss": 0.069, "step": 101670 }, { "epoch": 3.437573954494743, "grad_norm": 18.338802337646484, "learning_rate": 5.2744207046443604e-05, "loss": 0.0619, "step": 101680 }, { "epoch": 3.4379120321849963, "grad_norm": 3.2123048305511475, "learning_rate": 5.273595212101055e-05, "loss": 0.0686, "step": 101690 }, { "epoch": 3.4382501098752494, "grad_norm": 2.268282651901245, "learning_rate": 5.2727697120777586e-05, "loss": 0.0623, "step": 101700 }, { "epoch": 3.4385881875655024, "grad_norm": 2.0527875423431396, "learning_rate": 5.271944204597043e-05, "loss": 0.0647, "step": 101710 }, { "epoch": 3.438926265255756, "grad_norm": 2.656902551651001, "learning_rate": 5.271118689681473e-05, "loss": 0.0628, "step": 101720 }, { "epoch": 3.439264342946009, "grad_norm": 1.6612615585327148, "learning_rate": 5.27029316735362e-05, "loss": 0.0612, "step": 101730 }, { "epoch": 3.439602420636262, "grad_norm": 3.6640372276306152, "learning_rate": 5.269467637636054e-05, "loss": 0.0612, "step": 101740 }, { "epoch": 3.4399404983265156, "grad_norm": 2.602964162826538, "learning_rate": 5.2686421005513454e-05, "loss": 0.0641, "step": 101750 }, { "epoch": 3.4402785760167687, "grad_norm": 3.2049813270568848, "learning_rate": 5.26781655612206e-05, "loss": 0.0694, "step": 101760 }, { "epoch": 3.4406166537070217, "grad_norm": 1.2191317081451416, "learning_rate": 5.2669910043707736e-05, "loss": 0.0622, "step": 101770 }, { "epoch": 3.4409547313972753, "grad_norm": 2.247993230819702, "learning_rate": 5.266165445320052e-05, "loss": 0.0634, "step": 101780 }, { "epoch": 3.4412928090875283, "grad_norm": 2.8704004287719727, "learning_rate": 5.2653398789924666e-05, "loss": 0.0668, "step": 101790 }, { "epoch": 3.4416308867777814, "grad_norm": 3.1456377506256104, "learning_rate": 5.264514305410587e-05, "loss": 0.0682, "step": 101800 }, { "epoch": 3.441968964468035, "grad_norm": 1.9698196649551392, "learning_rate": 5.263688724596989e-05, "loss": 0.0692, "step": 101810 }, { "epoch": 3.442307042158288, "grad_norm": 2.9976532459259033, "learning_rate": 5.262863136574239e-05, "loss": 0.0647, "step": 101820 }, { "epoch": 3.442645119848541, "grad_norm": 1.0999568700790405, "learning_rate": 5.2620375413649096e-05, "loss": 0.0687, "step": 101830 }, { "epoch": 3.4429831975387946, "grad_norm": 2.899568796157837, "learning_rate": 5.2612119389915716e-05, "loss": 0.0647, "step": 101840 }, { "epoch": 3.4433212752290476, "grad_norm": 2.052889585494995, "learning_rate": 5.2603863294767986e-05, "loss": 0.0602, "step": 101850 }, { "epoch": 3.4436593529193007, "grad_norm": 9.429327011108398, "learning_rate": 5.259560712843159e-05, "loss": 0.0613, "step": 101860 }, { "epoch": 3.443997430609554, "grad_norm": 2.1584389209747314, "learning_rate": 5.2587350891132295e-05, "loss": 0.0655, "step": 101870 }, { "epoch": 3.4443355082998073, "grad_norm": 1.8063075542449951, "learning_rate": 5.2579094583095776e-05, "loss": 0.0628, "step": 101880 }, { "epoch": 3.4446735859900603, "grad_norm": 3.002178430557251, "learning_rate": 5.25708382045478e-05, "loss": 0.063, "step": 101890 }, { "epoch": 3.445011663680314, "grad_norm": 2.376725435256958, "learning_rate": 5.2562581755714066e-05, "loss": 0.0665, "step": 101900 }, { "epoch": 3.445349741370567, "grad_norm": 1.4971251487731934, "learning_rate": 5.25543252368203e-05, "loss": 0.0705, "step": 101910 }, { "epoch": 3.44568781906082, "grad_norm": 1.3736248016357422, "learning_rate": 5.254606864809225e-05, "loss": 0.0605, "step": 101920 }, { "epoch": 3.4460258967510735, "grad_norm": 11.894665718078613, "learning_rate": 5.253781198975565e-05, "loss": 0.0641, "step": 101930 }, { "epoch": 3.4463639744413266, "grad_norm": 1.3453446626663208, "learning_rate": 5.252955526203621e-05, "loss": 0.0682, "step": 101940 }, { "epoch": 3.4467020521315797, "grad_norm": 1.2217437028884888, "learning_rate": 5.252129846515967e-05, "loss": 0.0654, "step": 101950 }, { "epoch": 3.447040129821833, "grad_norm": 1.333240270614624, "learning_rate": 5.251304159935181e-05, "loss": 0.0602, "step": 101960 }, { "epoch": 3.4473782075120862, "grad_norm": 1.7980833053588867, "learning_rate": 5.250478466483831e-05, "loss": 0.0641, "step": 101970 }, { "epoch": 3.4477162852023393, "grad_norm": 7.675108432769775, "learning_rate": 5.249652766184495e-05, "loss": 0.0728, "step": 101980 }, { "epoch": 3.448054362892593, "grad_norm": 2.382633686065674, "learning_rate": 5.2488270590597444e-05, "loss": 0.0609, "step": 101990 }, { "epoch": 3.448392440582846, "grad_norm": 1.8788601160049438, "learning_rate": 5.248001345132157e-05, "loss": 0.0701, "step": 102000 }, { "epoch": 3.4487305182730994, "grad_norm": 1.7410086393356323, "learning_rate": 5.247175624424306e-05, "loss": 0.061, "step": 102010 }, { "epoch": 3.4490685959633525, "grad_norm": 2.9508631229400635, "learning_rate": 5.246349896958766e-05, "loss": 0.0551, "step": 102020 }, { "epoch": 3.4494066736536055, "grad_norm": 2.5009891986846924, "learning_rate": 5.245524162758113e-05, "loss": 0.062, "step": 102030 }, { "epoch": 3.449744751343859, "grad_norm": 1.4033480882644653, "learning_rate": 5.2446984218449225e-05, "loss": 0.0692, "step": 102040 }, { "epoch": 3.450082829034112, "grad_norm": 4.087813377380371, "learning_rate": 5.243872674241766e-05, "loss": 0.0578, "step": 102050 }, { "epoch": 3.450420906724365, "grad_norm": 2.7563064098358154, "learning_rate": 5.243046919971224e-05, "loss": 0.0705, "step": 102060 }, { "epoch": 3.4507589844146187, "grad_norm": 1.1453906297683716, "learning_rate": 5.24222115905587e-05, "loss": 0.0651, "step": 102070 }, { "epoch": 3.4510970621048718, "grad_norm": 1.6556134223937988, "learning_rate": 5.241395391518282e-05, "loss": 0.0606, "step": 102080 }, { "epoch": 3.451435139795125, "grad_norm": 1.4581044912338257, "learning_rate": 5.240569617381033e-05, "loss": 0.0637, "step": 102090 }, { "epoch": 3.4517732174853784, "grad_norm": 2.0063822269439697, "learning_rate": 5.2397438366667015e-05, "loss": 0.0668, "step": 102100 }, { "epoch": 3.4521112951756314, "grad_norm": 1.4081069231033325, "learning_rate": 5.238918049397864e-05, "loss": 0.0636, "step": 102110 }, { "epoch": 3.4524493728658845, "grad_norm": 1.5610344409942627, "learning_rate": 5.238092255597096e-05, "loss": 0.0582, "step": 102120 }, { "epoch": 3.452787450556138, "grad_norm": 3.382662534713745, "learning_rate": 5.237266455286976e-05, "loss": 0.0607, "step": 102130 }, { "epoch": 3.453125528246391, "grad_norm": 7.163872718811035, "learning_rate": 5.2364406484900795e-05, "loss": 0.0653, "step": 102140 }, { "epoch": 3.453463605936644, "grad_norm": 2.191342830657959, "learning_rate": 5.235614835228985e-05, "loss": 0.0582, "step": 102150 }, { "epoch": 3.4538016836268977, "grad_norm": 3.8017990589141846, "learning_rate": 5.234789015526269e-05, "loss": 0.0662, "step": 102160 }, { "epoch": 3.4541397613171507, "grad_norm": 2.0822689533233643, "learning_rate": 5.2339631894045106e-05, "loss": 0.0512, "step": 102170 }, { "epoch": 3.454477839007404, "grad_norm": 4.426382541656494, "learning_rate": 5.233137356886285e-05, "loss": 0.0651, "step": 102180 }, { "epoch": 3.4548159166976573, "grad_norm": 3.455458402633667, "learning_rate": 5.232311517994173e-05, "loss": 0.0695, "step": 102190 }, { "epoch": 3.4551539943879104, "grad_norm": 7.691814422607422, "learning_rate": 5.23148567275075e-05, "loss": 0.0672, "step": 102200 }, { "epoch": 3.4554920720781634, "grad_norm": 1.8420906066894531, "learning_rate": 5.230659821178597e-05, "loss": 0.0618, "step": 102210 }, { "epoch": 3.455830149768417, "grad_norm": 1.9041060209274292, "learning_rate": 5.22983396330029e-05, "loss": 0.0628, "step": 102220 }, { "epoch": 3.45616822745867, "grad_norm": 2.078371047973633, "learning_rate": 5.229008099138411e-05, "loss": 0.0565, "step": 102230 }, { "epoch": 3.456506305148923, "grad_norm": 2.1330089569091797, "learning_rate": 5.228182228715532e-05, "loss": 0.0619, "step": 102240 }, { "epoch": 3.4568443828391766, "grad_norm": 1.3824621438980103, "learning_rate": 5.227356352054241e-05, "loss": 0.0601, "step": 102250 }, { "epoch": 3.4571824605294297, "grad_norm": 1.3392040729522705, "learning_rate": 5.226530469177111e-05, "loss": 0.0643, "step": 102260 }, { "epoch": 3.4575205382196827, "grad_norm": 1.8239881992340088, "learning_rate": 5.225704580106724e-05, "loss": 0.0593, "step": 102270 }, { "epoch": 3.4578586159099363, "grad_norm": 1.859138011932373, "learning_rate": 5.224878684865657e-05, "loss": 0.0595, "step": 102280 }, { "epoch": 3.4581966936001893, "grad_norm": 1.843689203262329, "learning_rate": 5.224052783476494e-05, "loss": 0.0632, "step": 102290 }, { "epoch": 3.4585347712904424, "grad_norm": 5.653196811676025, "learning_rate": 5.2232268759618094e-05, "loss": 0.0652, "step": 102300 }, { "epoch": 3.458872848980696, "grad_norm": 3.733916759490967, "learning_rate": 5.222400962344186e-05, "loss": 0.0611, "step": 102310 }, { "epoch": 3.459210926670949, "grad_norm": 3.1127395629882812, "learning_rate": 5.221575042646205e-05, "loss": 0.0645, "step": 102320 }, { "epoch": 3.459549004361202, "grad_norm": 3.4901340007781982, "learning_rate": 5.2207491168904455e-05, "loss": 0.0595, "step": 102330 }, { "epoch": 3.4598870820514556, "grad_norm": 2.1514768600463867, "learning_rate": 5.219923185099488e-05, "loss": 0.0674, "step": 102340 }, { "epoch": 3.4602251597417086, "grad_norm": 3.4963674545288086, "learning_rate": 5.2190972472959134e-05, "loss": 0.0615, "step": 102350 }, { "epoch": 3.4605632374319617, "grad_norm": 1.8689056634902954, "learning_rate": 5.218271303502301e-05, "loss": 0.0641, "step": 102360 }, { "epoch": 3.460901315122215, "grad_norm": 1.8313182592391968, "learning_rate": 5.217445353741234e-05, "loss": 0.06, "step": 102370 }, { "epoch": 3.4612393928124683, "grad_norm": 1.3884893655776978, "learning_rate": 5.216619398035293e-05, "loss": 0.0607, "step": 102380 }, { "epoch": 3.4615774705027214, "grad_norm": 5.313385963439941, "learning_rate": 5.2157934364070594e-05, "loss": 0.0651, "step": 102390 }, { "epoch": 3.461915548192975, "grad_norm": 2.814401865005493, "learning_rate": 5.214967468879113e-05, "loss": 0.0556, "step": 102400 }, { "epoch": 3.462253625883228, "grad_norm": 1.2316612005233765, "learning_rate": 5.2141414954740376e-05, "loss": 0.0687, "step": 102410 }, { "epoch": 3.462591703573481, "grad_norm": 1.716015338897705, "learning_rate": 5.2133155162144144e-05, "loss": 0.0595, "step": 102420 }, { "epoch": 3.4629297812637345, "grad_norm": 2.1509523391723633, "learning_rate": 5.212489531122825e-05, "loss": 0.064, "step": 102430 }, { "epoch": 3.4632678589539876, "grad_norm": 12.209434509277344, "learning_rate": 5.2116635402218505e-05, "loss": 0.0578, "step": 102440 }, { "epoch": 3.4636059366442407, "grad_norm": 1.3976184129714966, "learning_rate": 5.210837543534075e-05, "loss": 0.0668, "step": 102450 }, { "epoch": 3.463944014334494, "grad_norm": 3.365429162979126, "learning_rate": 5.210011541082081e-05, "loss": 0.061, "step": 102460 }, { "epoch": 3.4642820920247472, "grad_norm": 1.1238758563995361, "learning_rate": 5.20918553288845e-05, "loss": 0.0671, "step": 102470 }, { "epoch": 3.4646201697150003, "grad_norm": 0.777164101600647, "learning_rate": 5.2083595189757664e-05, "loss": 0.0648, "step": 102480 }, { "epoch": 3.464958247405254, "grad_norm": 12.68887996673584, "learning_rate": 5.207533499366609e-05, "loss": 0.0637, "step": 102490 }, { "epoch": 3.465296325095507, "grad_norm": 5.33276891708374, "learning_rate": 5.206707474083565e-05, "loss": 0.0672, "step": 102500 }, { "epoch": 3.46563440278576, "grad_norm": 1.766762375831604, "learning_rate": 5.2058814431492165e-05, "loss": 0.0599, "step": 102510 }, { "epoch": 3.4659724804760135, "grad_norm": 2.0916695594787598, "learning_rate": 5.205055406586147e-05, "loss": 0.0616, "step": 102520 }, { "epoch": 3.4663105581662665, "grad_norm": 1.7194666862487793, "learning_rate": 5.204229364416939e-05, "loss": 0.0657, "step": 102530 }, { "epoch": 3.4666486358565196, "grad_norm": 1.5036969184875488, "learning_rate": 5.203403316664178e-05, "loss": 0.0632, "step": 102540 }, { "epoch": 3.466986713546773, "grad_norm": 1.2886528968811035, "learning_rate": 5.2025772633504455e-05, "loss": 0.051, "step": 102550 }, { "epoch": 3.467324791237026, "grad_norm": 1.3920081853866577, "learning_rate": 5.201751204498327e-05, "loss": 0.057, "step": 102560 }, { "epoch": 3.4676628689272793, "grad_norm": 1.8898875713348389, "learning_rate": 5.2009251401304074e-05, "loss": 0.0634, "step": 102570 }, { "epoch": 3.4680009466175328, "grad_norm": 2.4770123958587646, "learning_rate": 5.2000990702692683e-05, "loss": 0.0617, "step": 102580 }, { "epoch": 3.468339024307786, "grad_norm": 5.260518550872803, "learning_rate": 5.199272994937497e-05, "loss": 0.0636, "step": 102590 }, { "epoch": 3.468677101998039, "grad_norm": 2.1796035766601562, "learning_rate": 5.198446914157677e-05, "loss": 0.0645, "step": 102600 }, { "epoch": 3.4690151796882924, "grad_norm": 5.734726905822754, "learning_rate": 5.197620827952393e-05, "loss": 0.0678, "step": 102610 }, { "epoch": 3.4693532573785455, "grad_norm": 1.8176629543304443, "learning_rate": 5.1967947363442284e-05, "loss": 0.0663, "step": 102620 }, { "epoch": 3.4696913350687986, "grad_norm": 3.257878541946411, "learning_rate": 5.19596863935577e-05, "loss": 0.0645, "step": 102630 }, { "epoch": 3.470029412759052, "grad_norm": 4.835374355316162, "learning_rate": 5.1951425370096044e-05, "loss": 0.0677, "step": 102640 }, { "epoch": 3.470367490449305, "grad_norm": 2.3821523189544678, "learning_rate": 5.194316429328313e-05, "loss": 0.0629, "step": 102650 }, { "epoch": 3.4707055681395587, "grad_norm": 3.297795295715332, "learning_rate": 5.193490316334484e-05, "loss": 0.0668, "step": 102660 }, { "epoch": 3.4710436458298117, "grad_norm": 1.4374847412109375, "learning_rate": 5.192664198050704e-05, "loss": 0.0629, "step": 102670 }, { "epoch": 3.471381723520065, "grad_norm": 2.591156482696533, "learning_rate": 5.1918380744995556e-05, "loss": 0.0624, "step": 102680 }, { "epoch": 3.4717198012103183, "grad_norm": 6.7392191886901855, "learning_rate": 5.191011945703625e-05, "loss": 0.0656, "step": 102690 }, { "epoch": 3.4720578789005714, "grad_norm": 5.618735313415527, "learning_rate": 5.1901858116855016e-05, "loss": 0.0675, "step": 102700 }, { "epoch": 3.4723959565908245, "grad_norm": 0.972935676574707, "learning_rate": 5.1893596724677706e-05, "loss": 0.0539, "step": 102710 }, { "epoch": 3.472734034281078, "grad_norm": 2.02494215965271, "learning_rate": 5.188533528073015e-05, "loss": 0.0637, "step": 102720 }, { "epoch": 3.473072111971331, "grad_norm": 3.2054173946380615, "learning_rate": 5.187707378523825e-05, "loss": 0.0707, "step": 102730 }, { "epoch": 3.473410189661584, "grad_norm": 4.076421737670898, "learning_rate": 5.1868812238427855e-05, "loss": 0.0637, "step": 102740 }, { "epoch": 3.4737482673518376, "grad_norm": 2.661773681640625, "learning_rate": 5.186055064052483e-05, "loss": 0.064, "step": 102750 }, { "epoch": 3.4740863450420907, "grad_norm": 4.354435920715332, "learning_rate": 5.185228899175505e-05, "loss": 0.0647, "step": 102760 }, { "epoch": 3.4744244227323438, "grad_norm": 1.5932495594024658, "learning_rate": 5.184402729234439e-05, "loss": 0.0622, "step": 102770 }, { "epoch": 3.4747625004225973, "grad_norm": 1.9421488046646118, "learning_rate": 5.183576554251871e-05, "loss": 0.0656, "step": 102780 }, { "epoch": 3.4751005781128503, "grad_norm": 13.354283332824707, "learning_rate": 5.1827503742503904e-05, "loss": 0.0665, "step": 102790 }, { "epoch": 3.4754386558031034, "grad_norm": 5.1198601722717285, "learning_rate": 5.181924189252582e-05, "loss": 0.068, "step": 102800 }, { "epoch": 3.475776733493357, "grad_norm": 2.558166027069092, "learning_rate": 5.181097999281035e-05, "loss": 0.0644, "step": 102810 }, { "epoch": 3.47611481118361, "grad_norm": 3.5718424320220947, "learning_rate": 5.1802718043583356e-05, "loss": 0.0702, "step": 102820 }, { "epoch": 3.476452888873863, "grad_norm": 2.259601593017578, "learning_rate": 5.179445604507074e-05, "loss": 0.062, "step": 102830 }, { "epoch": 3.4767909665641166, "grad_norm": 1.9076590538024902, "learning_rate": 5.1786193997498365e-05, "loss": 0.0584, "step": 102840 }, { "epoch": 3.4771290442543696, "grad_norm": 0.9943801164627075, "learning_rate": 5.1777931901092124e-05, "loss": 0.0608, "step": 102850 }, { "epoch": 3.4774671219446227, "grad_norm": 9.670793533325195, "learning_rate": 5.176966975607789e-05, "loss": 0.0681, "step": 102860 }, { "epoch": 3.477805199634876, "grad_norm": 6.194944858551025, "learning_rate": 5.176140756268154e-05, "loss": 0.0631, "step": 102870 }, { "epoch": 3.4781432773251293, "grad_norm": 2.4061968326568604, "learning_rate": 5.175314532112897e-05, "loss": 0.0608, "step": 102880 }, { "epoch": 3.4784813550153824, "grad_norm": 2.613380193710327, "learning_rate": 5.1744883031646086e-05, "loss": 0.0691, "step": 102890 }, { "epoch": 3.478819432705636, "grad_norm": 1.579450249671936, "learning_rate": 5.1736620694458735e-05, "loss": 0.0621, "step": 102900 }, { "epoch": 3.479157510395889, "grad_norm": 3.388448715209961, "learning_rate": 5.172835830979284e-05, "loss": 0.0637, "step": 102910 }, { "epoch": 3.479495588086142, "grad_norm": 1.8383105993270874, "learning_rate": 5.172009587787426e-05, "loss": 0.0589, "step": 102920 }, { "epoch": 3.4798336657763955, "grad_norm": 1.5207293033599854, "learning_rate": 5.171183339892892e-05, "loss": 0.0629, "step": 102930 }, { "epoch": 3.4801717434666486, "grad_norm": 2.019143581390381, "learning_rate": 5.170357087318269e-05, "loss": 0.0685, "step": 102940 }, { "epoch": 3.4805098211569017, "grad_norm": 1.3583543300628662, "learning_rate": 5.1695308300861476e-05, "loss": 0.0656, "step": 102950 }, { "epoch": 3.480847898847155, "grad_norm": 1.9494048357009888, "learning_rate": 5.168704568219118e-05, "loss": 0.0693, "step": 102960 }, { "epoch": 3.4811859765374082, "grad_norm": 5.087681770324707, "learning_rate": 5.167878301739767e-05, "loss": 0.0667, "step": 102970 }, { "epoch": 3.4815240542276618, "grad_norm": 1.8628263473510742, "learning_rate": 5.1670520306706884e-05, "loss": 0.0665, "step": 102980 }, { "epoch": 3.481862131917915, "grad_norm": 2.0751540660858154, "learning_rate": 5.1662257550344686e-05, "loss": 0.0681, "step": 102990 }, { "epoch": 3.482200209608168, "grad_norm": 1.9119170904159546, "learning_rate": 5.1653994748536996e-05, "loss": 0.0701, "step": 103000 }, { "epoch": 3.4825382872984214, "grad_norm": 1.4173564910888672, "learning_rate": 5.16457319015097e-05, "loss": 0.0659, "step": 103010 }, { "epoch": 3.4828763649886745, "grad_norm": 1.019335389137268, "learning_rate": 5.1637469009488724e-05, "loss": 0.0637, "step": 103020 }, { "epoch": 3.4832144426789275, "grad_norm": 4.194655418395996, "learning_rate": 5.162920607269996e-05, "loss": 0.0674, "step": 103030 }, { "epoch": 3.483552520369181, "grad_norm": 0.9442597031593323, "learning_rate": 5.162094309136931e-05, "loss": 0.0613, "step": 103040 }, { "epoch": 3.483890598059434, "grad_norm": 0.8098688721656799, "learning_rate": 5.161268006572268e-05, "loss": 0.0633, "step": 103050 }, { "epoch": 3.484228675749687, "grad_norm": 2.704733371734619, "learning_rate": 5.1604416995986e-05, "loss": 0.068, "step": 103060 }, { "epoch": 3.4845667534399407, "grad_norm": 1.6527308225631714, "learning_rate": 5.1596153882385145e-05, "loss": 0.0655, "step": 103070 }, { "epoch": 3.484904831130194, "grad_norm": 0.9754257202148438, "learning_rate": 5.158789072514605e-05, "loss": 0.0687, "step": 103080 }, { "epoch": 3.485242908820447, "grad_norm": 0.9444442391395569, "learning_rate": 5.157962752449461e-05, "loss": 0.0624, "step": 103090 }, { "epoch": 3.4855809865107004, "grad_norm": 1.0291695594787598, "learning_rate": 5.1571364280656765e-05, "loss": 0.0606, "step": 103100 }, { "epoch": 3.4859190642009534, "grad_norm": 3.9965696334838867, "learning_rate": 5.1563100993858405e-05, "loss": 0.0707, "step": 103110 }, { "epoch": 3.4862571418912065, "grad_norm": 2.607778549194336, "learning_rate": 5.155483766432544e-05, "loss": 0.0651, "step": 103120 }, { "epoch": 3.48659521958146, "grad_norm": 2.0254054069519043, "learning_rate": 5.1546574292283803e-05, "loss": 0.066, "step": 103130 }, { "epoch": 3.486933297271713, "grad_norm": 1.786616325378418, "learning_rate": 5.1538310877959396e-05, "loss": 0.068, "step": 103140 }, { "epoch": 3.487271374961966, "grad_norm": 1.1132668256759644, "learning_rate": 5.153004742157816e-05, "loss": 0.0653, "step": 103150 }, { "epoch": 3.4876094526522197, "grad_norm": 1.9276080131530762, "learning_rate": 5.152178392336601e-05, "loss": 0.0707, "step": 103160 }, { "epoch": 3.4879475303424727, "grad_norm": 2.2565815448760986, "learning_rate": 5.1513520383548853e-05, "loss": 0.0691, "step": 103170 }, { "epoch": 3.488285608032726, "grad_norm": 1.0910719633102417, "learning_rate": 5.1505256802352606e-05, "loss": 0.0648, "step": 103180 }, { "epoch": 3.4886236857229793, "grad_norm": 1.114289402961731, "learning_rate": 5.149699318000323e-05, "loss": 0.069, "step": 103190 }, { "epoch": 3.4889617634132324, "grad_norm": 1.1974204778671265, "learning_rate": 5.148872951672659e-05, "loss": 0.0634, "step": 103200 }, { "epoch": 3.4892998411034855, "grad_norm": 2.7593188285827637, "learning_rate": 5.148046581274866e-05, "loss": 0.0653, "step": 103210 }, { "epoch": 3.489637918793739, "grad_norm": 1.2484052181243896, "learning_rate": 5.147220206829535e-05, "loss": 0.0641, "step": 103220 }, { "epoch": 3.489975996483992, "grad_norm": 2.1834094524383545, "learning_rate": 5.146393828359259e-05, "loss": 0.061, "step": 103230 }, { "epoch": 3.490314074174245, "grad_norm": 1.4983865022659302, "learning_rate": 5.14556744588663e-05, "loss": 0.0649, "step": 103240 }, { "epoch": 3.4906521518644986, "grad_norm": 1.1251574754714966, "learning_rate": 5.144741059434244e-05, "loss": 0.0651, "step": 103250 }, { "epoch": 3.4909902295547517, "grad_norm": 2.4825544357299805, "learning_rate": 5.143914669024688e-05, "loss": 0.0675, "step": 103260 }, { "epoch": 3.4913283072450048, "grad_norm": 1.0769624710083008, "learning_rate": 5.143088274680562e-05, "loss": 0.0647, "step": 103270 }, { "epoch": 3.4916663849352583, "grad_norm": 1.2054580450057983, "learning_rate": 5.1422618764244544e-05, "loss": 0.0598, "step": 103280 }, { "epoch": 3.4920044626255113, "grad_norm": 1.8505719900131226, "learning_rate": 5.141435474278963e-05, "loss": 0.0734, "step": 103290 }, { "epoch": 3.4923425403157644, "grad_norm": 1.576208233833313, "learning_rate": 5.1406090682666775e-05, "loss": 0.0636, "step": 103300 }, { "epoch": 3.492680618006018, "grad_norm": 2.9452085494995117, "learning_rate": 5.139782658410193e-05, "loss": 0.0621, "step": 103310 }, { "epoch": 3.493018695696271, "grad_norm": 1.290029764175415, "learning_rate": 5.138956244732103e-05, "loss": 0.0649, "step": 103320 }, { "epoch": 3.493356773386524, "grad_norm": 2.0466244220733643, "learning_rate": 5.1381298272550005e-05, "loss": 0.0686, "step": 103330 }, { "epoch": 3.4936948510767776, "grad_norm": 0.9332560300827026, "learning_rate": 5.1373034060014816e-05, "loss": 0.0618, "step": 103340 }, { "epoch": 3.4940329287670306, "grad_norm": 1.3777481317520142, "learning_rate": 5.136476980994138e-05, "loss": 0.0628, "step": 103350 }, { "epoch": 3.4943710064572837, "grad_norm": 5.261030197143555, "learning_rate": 5.135650552255567e-05, "loss": 0.0579, "step": 103360 }, { "epoch": 3.4947090841475372, "grad_norm": 3.649383783340454, "learning_rate": 5.134824119808359e-05, "loss": 0.0625, "step": 103370 }, { "epoch": 3.4950471618377903, "grad_norm": 2.038308620452881, "learning_rate": 5.133997683675113e-05, "loss": 0.0639, "step": 103380 }, { "epoch": 3.4953852395280434, "grad_norm": 4.808132648468018, "learning_rate": 5.133171243878416e-05, "loss": 0.0622, "step": 103390 }, { "epoch": 3.495723317218297, "grad_norm": 1.6950417757034302, "learning_rate": 5.1323448004408714e-05, "loss": 0.0651, "step": 103400 }, { "epoch": 3.49606139490855, "grad_norm": 1.714301586151123, "learning_rate": 5.131518353385067e-05, "loss": 0.0641, "step": 103410 }, { "epoch": 3.496399472598803, "grad_norm": 1.4431780576705933, "learning_rate": 5.130691902733603e-05, "loss": 0.0569, "step": 103420 }, { "epoch": 3.4967375502890565, "grad_norm": 2.5958573818206787, "learning_rate": 5.129865448509069e-05, "loss": 0.0617, "step": 103430 }, { "epoch": 3.4970756279793096, "grad_norm": 1.301926851272583, "learning_rate": 5.129038990734064e-05, "loss": 0.0613, "step": 103440 }, { "epoch": 3.4974137056695627, "grad_norm": 2.114413261413574, "learning_rate": 5.12821252943118e-05, "loss": 0.0646, "step": 103450 }, { "epoch": 3.497751783359816, "grad_norm": 1.682897925376892, "learning_rate": 5.127386064623014e-05, "loss": 0.0583, "step": 103460 }, { "epoch": 3.4980898610500692, "grad_norm": 2.061328172683716, "learning_rate": 5.1265595963321625e-05, "loss": 0.0594, "step": 103470 }, { "epoch": 3.4984279387403223, "grad_norm": 1.8932371139526367, "learning_rate": 5.125733124581218e-05, "loss": 0.0644, "step": 103480 }, { "epoch": 3.498766016430576, "grad_norm": 36.428680419921875, "learning_rate": 5.1249066493927765e-05, "loss": 0.0642, "step": 103490 }, { "epoch": 3.499104094120829, "grad_norm": 2.273787498474121, "learning_rate": 5.124080170789436e-05, "loss": 0.0592, "step": 103500 }, { "epoch": 3.499442171811082, "grad_norm": 35.32795333862305, "learning_rate": 5.1232536887937876e-05, "loss": 0.0613, "step": 103510 }, { "epoch": 3.4997802495013355, "grad_norm": 1.689509630203247, "learning_rate": 5.1224272034284316e-05, "loss": 0.0652, "step": 103520 }, { "epoch": 3.5001183271915886, "grad_norm": 1.3703978061676025, "learning_rate": 5.1216007147159606e-05, "loss": 0.0678, "step": 103530 }, { "epoch": 3.5004564048818416, "grad_norm": 1.668554663658142, "learning_rate": 5.120774222678972e-05, "loss": 0.0706, "step": 103540 }, { "epoch": 3.500794482572095, "grad_norm": 5.154171943664551, "learning_rate": 5.119947727340062e-05, "loss": 0.0669, "step": 103550 }, { "epoch": 3.501132560262348, "grad_norm": 1.874027967453003, "learning_rate": 5.119121228721826e-05, "loss": 0.0639, "step": 103560 }, { "epoch": 3.5014706379526013, "grad_norm": 1.7249243259429932, "learning_rate": 5.118294726846862e-05, "loss": 0.0623, "step": 103570 }, { "epoch": 3.501808715642855, "grad_norm": 1.4819366931915283, "learning_rate": 5.117468221737762e-05, "loss": 0.0648, "step": 103580 }, { "epoch": 3.502146793333108, "grad_norm": 3.8719522953033447, "learning_rate": 5.116641713417125e-05, "loss": 0.0641, "step": 103590 }, { "epoch": 3.502484871023361, "grad_norm": 1.5654350519180298, "learning_rate": 5.115815201907549e-05, "loss": 0.0635, "step": 103600 }, { "epoch": 3.5028229487136144, "grad_norm": 2.0096595287323, "learning_rate": 5.114988687231629e-05, "loss": 0.0672, "step": 103610 }, { "epoch": 3.5031610264038675, "grad_norm": 6.224466323852539, "learning_rate": 5.11416216941196e-05, "loss": 0.0593, "step": 103620 }, { "epoch": 3.5034991040941206, "grad_norm": 2.099719524383545, "learning_rate": 5.113335648471141e-05, "loss": 0.0651, "step": 103630 }, { "epoch": 3.503837181784374, "grad_norm": 1.7893173694610596, "learning_rate": 5.1125091244317666e-05, "loss": 0.0646, "step": 103640 }, { "epoch": 3.504175259474627, "grad_norm": 1.1080514192581177, "learning_rate": 5.111682597316437e-05, "loss": 0.0704, "step": 103650 }, { "epoch": 3.5045133371648802, "grad_norm": 1.6569308042526245, "learning_rate": 5.1108560671477466e-05, "loss": 0.0635, "step": 103660 }, { "epoch": 3.5048514148551337, "grad_norm": 3.763532876968384, "learning_rate": 5.110029533948294e-05, "loss": 0.0593, "step": 103670 }, { "epoch": 3.505189492545387, "grad_norm": 1.5077135562896729, "learning_rate": 5.109202997740673e-05, "loss": 0.0678, "step": 103680 }, { "epoch": 3.50552757023564, "grad_norm": 2.7950398921966553, "learning_rate": 5.108376458547486e-05, "loss": 0.0603, "step": 103690 }, { "epoch": 3.5058656479258934, "grad_norm": 2.968040704727173, "learning_rate": 5.107549916391324e-05, "loss": 0.0659, "step": 103700 }, { "epoch": 3.5062037256161465, "grad_norm": 6.83079195022583, "learning_rate": 5.106723371294789e-05, "loss": 0.0684, "step": 103710 }, { "epoch": 3.5065418033064, "grad_norm": 2.4254252910614014, "learning_rate": 5.105896823280477e-05, "loss": 0.064, "step": 103720 }, { "epoch": 3.506879880996653, "grad_norm": 1.7975070476531982, "learning_rate": 5.1050702723709864e-05, "loss": 0.0708, "step": 103730 }, { "epoch": 3.507217958686906, "grad_norm": 0.9428207278251648, "learning_rate": 5.104243718588914e-05, "loss": 0.0654, "step": 103740 }, { "epoch": 3.5075560363771596, "grad_norm": 1.3583565950393677, "learning_rate": 5.103417161956857e-05, "loss": 0.0702, "step": 103750 }, { "epoch": 3.5078941140674127, "grad_norm": 1.356075406074524, "learning_rate": 5.1025906024974135e-05, "loss": 0.0624, "step": 103760 }, { "epoch": 3.5082321917576658, "grad_norm": 2.447633981704712, "learning_rate": 5.1017640402331814e-05, "loss": 0.0617, "step": 103770 }, { "epoch": 3.5085702694479193, "grad_norm": 1.6239982843399048, "learning_rate": 5.100937475186759e-05, "loss": 0.0678, "step": 103780 }, { "epoch": 3.5089083471381723, "grad_norm": 1.3448911905288696, "learning_rate": 5.100110907380744e-05, "loss": 0.0631, "step": 103790 }, { "epoch": 3.5092464248284254, "grad_norm": 1.818616509437561, "learning_rate": 5.099284336837734e-05, "loss": 0.0668, "step": 103800 }, { "epoch": 3.509584502518679, "grad_norm": 1.5979799032211304, "learning_rate": 5.0984577635803276e-05, "loss": 0.0672, "step": 103810 }, { "epoch": 3.509922580208932, "grad_norm": 1.8256144523620605, "learning_rate": 5.097631187631124e-05, "loss": 0.0614, "step": 103820 }, { "epoch": 3.5102606578991855, "grad_norm": 1.9703173637390137, "learning_rate": 5.0968046090127195e-05, "loss": 0.0643, "step": 103830 }, { "epoch": 3.5105987355894386, "grad_norm": 1.6953915357589722, "learning_rate": 5.095978027747713e-05, "loss": 0.0675, "step": 103840 }, { "epoch": 3.5109368132796916, "grad_norm": 8.22176456451416, "learning_rate": 5.095151443858703e-05, "loss": 0.0646, "step": 103850 }, { "epoch": 3.511274890969945, "grad_norm": 1.9097442626953125, "learning_rate": 5.09432485736829e-05, "loss": 0.0636, "step": 103860 }, { "epoch": 3.5116129686601982, "grad_norm": 1.3886568546295166, "learning_rate": 5.0934982682990696e-05, "loss": 0.0563, "step": 103870 }, { "epoch": 3.5119510463504513, "grad_norm": 0.9883968234062195, "learning_rate": 5.0926716766736425e-05, "loss": 0.0599, "step": 103880 }, { "epoch": 3.512289124040705, "grad_norm": 1.4636452198028564, "learning_rate": 5.091845082514606e-05, "loss": 0.0594, "step": 103890 }, { "epoch": 3.512627201730958, "grad_norm": 1.1472954750061035, "learning_rate": 5.091018485844559e-05, "loss": 0.0651, "step": 103900 }, { "epoch": 3.512965279421211, "grad_norm": 1.2921504974365234, "learning_rate": 5.090191886686101e-05, "loss": 0.0619, "step": 103910 }, { "epoch": 3.5133033571114645, "grad_norm": 1.705068588256836, "learning_rate": 5.0893652850618315e-05, "loss": 0.0698, "step": 103920 }, { "epoch": 3.5136414348017175, "grad_norm": 1.3401274681091309, "learning_rate": 5.088538680994347e-05, "loss": 0.0665, "step": 103930 }, { "epoch": 3.5139795124919706, "grad_norm": 1.0801794528961182, "learning_rate": 5.08771207450625e-05, "loss": 0.0601, "step": 103940 }, { "epoch": 3.514317590182224, "grad_norm": 1.2289185523986816, "learning_rate": 5.086885465620137e-05, "loss": 0.0689, "step": 103950 }, { "epoch": 3.514655667872477, "grad_norm": 1.5275261402130127, "learning_rate": 5.0860588543586083e-05, "loss": 0.0609, "step": 103960 }, { "epoch": 3.5149937455627303, "grad_norm": 0.9601311683654785, "learning_rate": 5.085232240744262e-05, "loss": 0.0644, "step": 103970 }, { "epoch": 3.5153318232529838, "grad_norm": 0.6879487037658691, "learning_rate": 5.0844056247997e-05, "loss": 0.0621, "step": 103980 }, { "epoch": 3.515669900943237, "grad_norm": 1.7482876777648926, "learning_rate": 5.0835790065475184e-05, "loss": 0.058, "step": 103990 }, { "epoch": 3.51600797863349, "grad_norm": 1.7035568952560425, "learning_rate": 5.08275238601032e-05, "loss": 0.0515, "step": 104000 }, { "epoch": 3.5163460563237434, "grad_norm": 1.3458620309829712, "learning_rate": 5.0819257632106996e-05, "loss": 0.0546, "step": 104010 }, { "epoch": 3.5166841340139965, "grad_norm": 1.4947787523269653, "learning_rate": 5.081099138171261e-05, "loss": 0.0624, "step": 104020 }, { "epoch": 3.5170222117042496, "grad_norm": 0.9705289602279663, "learning_rate": 5.080272510914602e-05, "loss": 0.0609, "step": 104030 }, { "epoch": 3.517360289394503, "grad_norm": 0.8803079724311829, "learning_rate": 5.079445881463325e-05, "loss": 0.0589, "step": 104040 }, { "epoch": 3.517698367084756, "grad_norm": 1.735231876373291, "learning_rate": 5.0786192498400254e-05, "loss": 0.061, "step": 104050 }, { "epoch": 3.518036444775009, "grad_norm": 1.5228748321533203, "learning_rate": 5.077792616067304e-05, "loss": 0.0623, "step": 104060 }, { "epoch": 3.5183745224652627, "grad_norm": 5.947436332702637, "learning_rate": 5.076965980167764e-05, "loss": 0.0615, "step": 104070 }, { "epoch": 3.518712600155516, "grad_norm": 4.4704670906066895, "learning_rate": 5.0761393421640024e-05, "loss": 0.0693, "step": 104080 }, { "epoch": 3.519050677845769, "grad_norm": 0.6998491287231445, "learning_rate": 5.075312702078618e-05, "loss": 0.0582, "step": 104090 }, { "epoch": 3.5193887555360224, "grad_norm": 1.388789176940918, "learning_rate": 5.074486059934215e-05, "loss": 0.0586, "step": 104100 }, { "epoch": 3.5197268332262754, "grad_norm": 1.2560545206069946, "learning_rate": 5.07365941575339e-05, "loss": 0.0631, "step": 104110 }, { "epoch": 3.5200649109165285, "grad_norm": 0.9136276245117188, "learning_rate": 5.072832769558744e-05, "loss": 0.0566, "step": 104120 }, { "epoch": 3.520402988606782, "grad_norm": 0.6994821429252625, "learning_rate": 5.0720061213728784e-05, "loss": 0.0564, "step": 104130 }, { "epoch": 3.520741066297035, "grad_norm": 1.3873859643936157, "learning_rate": 5.071179471218392e-05, "loss": 0.0548, "step": 104140 }, { "epoch": 3.521079143987288, "grad_norm": 1.0898828506469727, "learning_rate": 5.070352819117885e-05, "loss": 0.0602, "step": 104150 }, { "epoch": 3.5214172216775417, "grad_norm": 1.122736930847168, "learning_rate": 5.0695261650939584e-05, "loss": 0.0626, "step": 104160 }, { "epoch": 3.5217552993677947, "grad_norm": 3.2774198055267334, "learning_rate": 5.068699509169215e-05, "loss": 0.0655, "step": 104170 }, { "epoch": 3.522093377058048, "grad_norm": 0.6268537640571594, "learning_rate": 5.06787285136625e-05, "loss": 0.066, "step": 104180 }, { "epoch": 3.5224314547483013, "grad_norm": 1.319205641746521, "learning_rate": 5.0670461917076686e-05, "loss": 0.0656, "step": 104190 }, { "epoch": 3.5227695324385544, "grad_norm": 0.679235577583313, "learning_rate": 5.066219530216068e-05, "loss": 0.0588, "step": 104200 }, { "epoch": 3.5231076101288075, "grad_norm": 1.3431532382965088, "learning_rate": 5.0653928669140514e-05, "loss": 0.0589, "step": 104210 }, { "epoch": 3.523445687819061, "grad_norm": 1.2329453229904175, "learning_rate": 5.064566201824217e-05, "loss": 0.0684, "step": 104220 }, { "epoch": 3.523783765509314, "grad_norm": 1.9789495468139648, "learning_rate": 5.063739534969169e-05, "loss": 0.0688, "step": 104230 }, { "epoch": 3.524121843199567, "grad_norm": 0.9484276175498962, "learning_rate": 5.062912866371503e-05, "loss": 0.0581, "step": 104240 }, { "epoch": 3.5244599208898206, "grad_norm": 0.6489514708518982, "learning_rate": 5.062086196053826e-05, "loss": 0.0704, "step": 104250 }, { "epoch": 3.5247979985800737, "grad_norm": 2.069690704345703, "learning_rate": 5.061259524038734e-05, "loss": 0.0686, "step": 104260 }, { "epoch": 3.5251360762703268, "grad_norm": 0.7184461951255798, "learning_rate": 5.06043285034883e-05, "loss": 0.0622, "step": 104270 }, { "epoch": 3.5254741539605803, "grad_norm": 1.4495562314987183, "learning_rate": 5.059606175006714e-05, "loss": 0.0664, "step": 104280 }, { "epoch": 3.5258122316508334, "grad_norm": 1.3750196695327759, "learning_rate": 5.058779498034989e-05, "loss": 0.0685, "step": 104290 }, { "epoch": 3.5261503093410864, "grad_norm": 2.0791947841644287, "learning_rate": 5.057952819456253e-05, "loss": 0.0624, "step": 104300 }, { "epoch": 3.52648838703134, "grad_norm": 1.4061015844345093, "learning_rate": 5.057126139293109e-05, "loss": 0.0691, "step": 104310 }, { "epoch": 3.526826464721593, "grad_norm": 1.9324437379837036, "learning_rate": 5.05629945756816e-05, "loss": 0.0642, "step": 104320 }, { "epoch": 3.527164542411846, "grad_norm": 0.9698336124420166, "learning_rate": 5.055472774304002e-05, "loss": 0.0558, "step": 104330 }, { "epoch": 3.5275026201020996, "grad_norm": 2.057739734649658, "learning_rate": 5.054646089523241e-05, "loss": 0.0651, "step": 104340 }, { "epoch": 3.5278406977923527, "grad_norm": 2.110168933868408, "learning_rate": 5.053819403248475e-05, "loss": 0.0671, "step": 104350 }, { "epoch": 3.5281787754826057, "grad_norm": 2.915313720703125, "learning_rate": 5.052992715502308e-05, "loss": 0.059, "step": 104360 }, { "epoch": 3.5285168531728592, "grad_norm": 2.178311586380005, "learning_rate": 5.05216602630734e-05, "loss": 0.0677, "step": 104370 }, { "epoch": 3.5288549308631123, "grad_norm": 1.0661126375198364, "learning_rate": 5.051339335686173e-05, "loss": 0.0631, "step": 104380 }, { "epoch": 3.5291930085533654, "grad_norm": 2.6777775287628174, "learning_rate": 5.050512643661406e-05, "loss": 0.0623, "step": 104390 }, { "epoch": 3.529531086243619, "grad_norm": 1.4854848384857178, "learning_rate": 5.049685950255645e-05, "loss": 0.0586, "step": 104400 }, { "epoch": 3.529869163933872, "grad_norm": 1.2998688220977783, "learning_rate": 5.0488592554914846e-05, "loss": 0.0662, "step": 104410 }, { "epoch": 3.530207241624125, "grad_norm": 1.3916196823120117, "learning_rate": 5.0480325593915345e-05, "loss": 0.0646, "step": 104420 }, { "epoch": 3.5305453193143785, "grad_norm": 4.111523628234863, "learning_rate": 5.04720586197839e-05, "loss": 0.0555, "step": 104430 }, { "epoch": 3.5308833970046316, "grad_norm": 2.219383478164673, "learning_rate": 5.046379163274656e-05, "loss": 0.0616, "step": 104440 }, { "epoch": 3.5312214746948847, "grad_norm": 2.611536741256714, "learning_rate": 5.0455524633029325e-05, "loss": 0.0606, "step": 104450 }, { "epoch": 3.531559552385138, "grad_norm": 1.5599699020385742, "learning_rate": 5.044725762085823e-05, "loss": 0.0644, "step": 104460 }, { "epoch": 3.5318976300753913, "grad_norm": 1.7110557556152344, "learning_rate": 5.043899059645926e-05, "loss": 0.0606, "step": 104470 }, { "epoch": 3.5322357077656443, "grad_norm": 0.7900499105453491, "learning_rate": 5.043072356005847e-05, "loss": 0.0648, "step": 104480 }, { "epoch": 3.532573785455898, "grad_norm": 0.9459646344184875, "learning_rate": 5.0422456511881844e-05, "loss": 0.0604, "step": 104490 }, { "epoch": 3.532911863146151, "grad_norm": 0.8316420316696167, "learning_rate": 5.0414189452155436e-05, "loss": 0.0657, "step": 104500 }, { "epoch": 3.533249940836404, "grad_norm": 0.8989819288253784, "learning_rate": 5.0405922381105234e-05, "loss": 0.0657, "step": 104510 }, { "epoch": 3.5335880185266575, "grad_norm": 7.476743221282959, "learning_rate": 5.039765529895726e-05, "loss": 0.0616, "step": 104520 }, { "epoch": 3.5339260962169106, "grad_norm": 1.3385238647460938, "learning_rate": 5.038938820593754e-05, "loss": 0.0639, "step": 104530 }, { "epoch": 3.5342641739071636, "grad_norm": 1.8579139709472656, "learning_rate": 5.03811211022721e-05, "loss": 0.058, "step": 104540 }, { "epoch": 3.534602251597417, "grad_norm": 2.088721513748169, "learning_rate": 5.037285398818694e-05, "loss": 0.0679, "step": 104550 }, { "epoch": 3.53494032928767, "grad_norm": 1.1896393299102783, "learning_rate": 5.0364586863908104e-05, "loss": 0.0721, "step": 104560 }, { "epoch": 3.5352784069779233, "grad_norm": 0.9000208973884583, "learning_rate": 5.035631972966161e-05, "loss": 0.0667, "step": 104570 }, { "epoch": 3.535616484668177, "grad_norm": 1.5805096626281738, "learning_rate": 5.0348052585673446e-05, "loss": 0.0635, "step": 104580 }, { "epoch": 3.53595456235843, "grad_norm": 1.411474347114563, "learning_rate": 5.0339785432169675e-05, "loss": 0.0645, "step": 104590 }, { "epoch": 3.536292640048683, "grad_norm": 2.00775408744812, "learning_rate": 5.0331518269376264e-05, "loss": 0.0552, "step": 104600 }, { "epoch": 3.5366307177389364, "grad_norm": 1.8138374090194702, "learning_rate": 5.032325109751931e-05, "loss": 0.0658, "step": 104610 }, { "epoch": 3.5369687954291895, "grad_norm": 2.352395534515381, "learning_rate": 5.0314983916824764e-05, "loss": 0.0606, "step": 104620 }, { "epoch": 3.5373068731194426, "grad_norm": 2.637599229812622, "learning_rate": 5.03067167275187e-05, "loss": 0.0668, "step": 104630 }, { "epoch": 3.537644950809696, "grad_norm": 4.939246654510498, "learning_rate": 5.02984495298271e-05, "loss": 0.0648, "step": 104640 }, { "epoch": 3.537983028499949, "grad_norm": 3.0656206607818604, "learning_rate": 5.029018232397601e-05, "loss": 0.068, "step": 104650 }, { "epoch": 3.5383211061902022, "grad_norm": 1.4390748739242554, "learning_rate": 5.028191511019142e-05, "loss": 0.0664, "step": 104660 }, { "epoch": 3.5386591838804557, "grad_norm": 2.3310201168060303, "learning_rate": 5.027364788869941e-05, "loss": 0.0751, "step": 104670 }, { "epoch": 3.538997261570709, "grad_norm": 2.6150600910186768, "learning_rate": 5.026538065972595e-05, "loss": 0.0709, "step": 104680 }, { "epoch": 3.5393353392609623, "grad_norm": 2.750216007232666, "learning_rate": 5.025711342349709e-05, "loss": 0.0647, "step": 104690 }, { "epoch": 3.5396734169512154, "grad_norm": 3.1972033977508545, "learning_rate": 5.024884618023884e-05, "loss": 0.0655, "step": 104700 }, { "epoch": 3.5400114946414685, "grad_norm": 2.525118112564087, "learning_rate": 5.024057893017724e-05, "loss": 0.0676, "step": 104710 }, { "epoch": 3.540349572331722, "grad_norm": 2.092703104019165, "learning_rate": 5.023231167353829e-05, "loss": 0.0703, "step": 104720 }, { "epoch": 3.540687650021975, "grad_norm": 10.158061027526855, "learning_rate": 5.022404441054802e-05, "loss": 0.0683, "step": 104730 }, { "epoch": 3.541025727712228, "grad_norm": 4.935099124908447, "learning_rate": 5.021577714143247e-05, "loss": 0.0706, "step": 104740 }, { "epoch": 3.5413638054024816, "grad_norm": 2.2987892627716064, "learning_rate": 5.020750986641766e-05, "loss": 0.0568, "step": 104750 }, { "epoch": 3.5417018830927347, "grad_norm": 3.635803461074829, "learning_rate": 5.01992425857296e-05, "loss": 0.0665, "step": 104760 }, { "epoch": 3.5420399607829878, "grad_norm": 1.920886754989624, "learning_rate": 5.019097529959432e-05, "loss": 0.0661, "step": 104770 }, { "epoch": 3.5423780384732413, "grad_norm": 2.185868501663208, "learning_rate": 5.018270800823787e-05, "loss": 0.0667, "step": 104780 }, { "epoch": 3.5427161161634944, "grad_norm": 1.2715482711791992, "learning_rate": 5.017444071188622e-05, "loss": 0.0626, "step": 104790 }, { "epoch": 3.543054193853748, "grad_norm": 2.129225492477417, "learning_rate": 5.0166173410765437e-05, "loss": 0.0673, "step": 104800 }, { "epoch": 3.543392271544001, "grad_norm": 2.5125551223754883, "learning_rate": 5.0157906105101535e-05, "loss": 0.0636, "step": 104810 }, { "epoch": 3.543730349234254, "grad_norm": 2.1584694385528564, "learning_rate": 5.014963879512056e-05, "loss": 0.0701, "step": 104820 }, { "epoch": 3.5440684269245075, "grad_norm": 2.497425079345703, "learning_rate": 5.014137148104849e-05, "loss": 0.0658, "step": 104830 }, { "epoch": 3.5444065046147606, "grad_norm": 2.942995071411133, "learning_rate": 5.013310416311139e-05, "loss": 0.0629, "step": 104840 }, { "epoch": 3.5447445823050137, "grad_norm": 2.3621058464050293, "learning_rate": 5.012483684153526e-05, "loss": 0.0637, "step": 104850 }, { "epoch": 3.545082659995267, "grad_norm": 1.1822012662887573, "learning_rate": 5.011656951654615e-05, "loss": 0.0626, "step": 104860 }, { "epoch": 3.5454207376855202, "grad_norm": 4.028791427612305, "learning_rate": 5.010830218837007e-05, "loss": 0.068, "step": 104870 }, { "epoch": 3.5457588153757733, "grad_norm": 1.524646282196045, "learning_rate": 5.0100034857233044e-05, "loss": 0.0705, "step": 104880 }, { "epoch": 3.546096893066027, "grad_norm": 5.442321300506592, "learning_rate": 5.009176752336111e-05, "loss": 0.069, "step": 104890 }, { "epoch": 3.54643497075628, "grad_norm": 3.9477553367614746, "learning_rate": 5.008350018698029e-05, "loss": 0.0699, "step": 104900 }, { "epoch": 3.546773048446533, "grad_norm": 2.823819160461426, "learning_rate": 5.00752328483166e-05, "loss": 0.0682, "step": 104910 }, { "epoch": 3.5471111261367865, "grad_norm": 2.0477545261383057, "learning_rate": 5.0066965507596065e-05, "loss": 0.0673, "step": 104920 }, { "epoch": 3.5474492038270395, "grad_norm": 1.9461774826049805, "learning_rate": 5.0058698165044726e-05, "loss": 0.0638, "step": 104930 }, { "epoch": 3.5477872815172926, "grad_norm": 2.579251527786255, "learning_rate": 5.0050430820888604e-05, "loss": 0.0606, "step": 104940 }, { "epoch": 3.548125359207546, "grad_norm": 14.456727027893066, "learning_rate": 5.004216347535371e-05, "loss": 0.0626, "step": 104950 }, { "epoch": 3.548463436897799, "grad_norm": 4.345439910888672, "learning_rate": 5.0033896128666105e-05, "loss": 0.0689, "step": 104960 }, { "epoch": 3.5488015145880523, "grad_norm": 2.227954626083374, "learning_rate": 5.002562878105177e-05, "loss": 0.058, "step": 104970 }, { "epoch": 3.5491395922783058, "grad_norm": 3.342468500137329, "learning_rate": 5.0017361432736766e-05, "loss": 0.0638, "step": 104980 }, { "epoch": 3.549477669968559, "grad_norm": 4.679305076599121, "learning_rate": 5.0009094083947096e-05, "loss": 0.0706, "step": 104990 }, { "epoch": 3.549815747658812, "grad_norm": 2.2723898887634277, "learning_rate": 5.000082673490881e-05, "loss": 0.0684, "step": 105000 }, { "epoch": 3.5501538253490654, "grad_norm": 2.6900057792663574, "learning_rate": 4.99925593858479e-05, "loss": 0.0636, "step": 105010 }, { "epoch": 3.5504919030393185, "grad_norm": 3.6717798709869385, "learning_rate": 4.998429203699044e-05, "loss": 0.0644, "step": 105020 }, { "epoch": 3.5508299807295716, "grad_norm": 1.5699354410171509, "learning_rate": 4.997602468856241e-05, "loss": 0.0685, "step": 105030 }, { "epoch": 3.551168058419825, "grad_norm": 3.4022483825683594, "learning_rate": 4.996775734078988e-05, "loss": 0.0628, "step": 105040 }, { "epoch": 3.551506136110078, "grad_norm": 2.5880587100982666, "learning_rate": 4.995948999389883e-05, "loss": 0.0712, "step": 105050 }, { "epoch": 3.551844213800331, "grad_norm": 5.2957377433776855, "learning_rate": 4.995122264811533e-05, "loss": 0.0731, "step": 105060 }, { "epoch": 3.5521822914905847, "grad_norm": 2.35514760017395, "learning_rate": 4.994295530366536e-05, "loss": 0.0704, "step": 105070 }, { "epoch": 3.552520369180838, "grad_norm": 14.843917846679688, "learning_rate": 4.993468796077498e-05, "loss": 0.0664, "step": 105080 }, { "epoch": 3.552858446871091, "grad_norm": 2.305478572845459, "learning_rate": 4.9926420619670214e-05, "loss": 0.0736, "step": 105090 }, { "epoch": 3.5531965245613444, "grad_norm": 4.849583148956299, "learning_rate": 4.991815328057706e-05, "loss": 0.0703, "step": 105100 }, { "epoch": 3.5535346022515975, "grad_norm": 6.237448692321777, "learning_rate": 4.9909885943721596e-05, "loss": 0.0629, "step": 105110 }, { "epoch": 3.5538726799418505, "grad_norm": 18.523380279541016, "learning_rate": 4.990161860932978e-05, "loss": 0.0693, "step": 105120 }, { "epoch": 3.554210757632104, "grad_norm": 34.30582046508789, "learning_rate": 4.989335127762771e-05, "loss": 0.0604, "step": 105130 }, { "epoch": 3.554548835322357, "grad_norm": 5.257853031158447, "learning_rate": 4.988508394884135e-05, "loss": 0.0721, "step": 105140 }, { "epoch": 3.55488691301261, "grad_norm": 4.969095706939697, "learning_rate": 4.987681662319675e-05, "loss": 0.0675, "step": 105150 }, { "epoch": 3.5552249907028637, "grad_norm": 5.89286470413208, "learning_rate": 4.986854930091996e-05, "loss": 0.0706, "step": 105160 }, { "epoch": 3.5555630683931168, "grad_norm": 4.616849422454834, "learning_rate": 4.9860281982236965e-05, "loss": 0.0774, "step": 105170 }, { "epoch": 3.55590114608337, "grad_norm": 4.238641738891602, "learning_rate": 4.9852014667373825e-05, "loss": 0.0662, "step": 105180 }, { "epoch": 3.5562392237736233, "grad_norm": 4.450649738311768, "learning_rate": 4.9843747356556534e-05, "loss": 0.0703, "step": 105190 }, { "epoch": 3.5565773014638764, "grad_norm": 6.588795185089111, "learning_rate": 4.983548005001112e-05, "loss": 0.0713, "step": 105200 }, { "epoch": 3.5569153791541295, "grad_norm": 3.4268605709075928, "learning_rate": 4.982721274796365e-05, "loss": 0.0692, "step": 105210 }, { "epoch": 3.557253456844383, "grad_norm": 3.2307910919189453, "learning_rate": 4.981894545064009e-05, "loss": 0.0592, "step": 105220 }, { "epoch": 3.557591534534636, "grad_norm": 18.702497482299805, "learning_rate": 4.9810678158266514e-05, "loss": 0.0683, "step": 105230 }, { "epoch": 3.557929612224889, "grad_norm": 5.178793430328369, "learning_rate": 4.980241087106891e-05, "loss": 0.0677, "step": 105240 }, { "epoch": 3.5582676899151426, "grad_norm": 3.4966492652893066, "learning_rate": 4.979414358927334e-05, "loss": 0.0658, "step": 105250 }, { "epoch": 3.5586057676053957, "grad_norm": 4.12351655960083, "learning_rate": 4.978587631310579e-05, "loss": 0.0683, "step": 105260 }, { "epoch": 3.5589438452956488, "grad_norm": 5.6645188331604, "learning_rate": 4.97776090427923e-05, "loss": 0.0719, "step": 105270 }, { "epoch": 3.5592819229859023, "grad_norm": 4.742364883422852, "learning_rate": 4.976934177855891e-05, "loss": 0.0696, "step": 105280 }, { "epoch": 3.5596200006761554, "grad_norm": 4.511775493621826, "learning_rate": 4.9761074520631616e-05, "loss": 0.0666, "step": 105290 }, { "epoch": 3.5599580783664084, "grad_norm": 7.246629238128662, "learning_rate": 4.9752807269236464e-05, "loss": 0.0633, "step": 105300 }, { "epoch": 3.560296156056662, "grad_norm": 4.325153350830078, "learning_rate": 4.974454002459945e-05, "loss": 0.0724, "step": 105310 }, { "epoch": 3.560634233746915, "grad_norm": 4.743143558502197, "learning_rate": 4.973627278694665e-05, "loss": 0.0622, "step": 105320 }, { "epoch": 3.560972311437168, "grad_norm": 3.590843915939331, "learning_rate": 4.972800555650402e-05, "loss": 0.0748, "step": 105330 }, { "epoch": 3.5613103891274216, "grad_norm": 3.944356918334961, "learning_rate": 4.9719738333497615e-05, "loss": 0.0691, "step": 105340 }, { "epoch": 3.5616484668176747, "grad_norm": 8.793462753295898, "learning_rate": 4.971147111815348e-05, "loss": 0.0698, "step": 105350 }, { "epoch": 3.5619865445079277, "grad_norm": 4.329751491546631, "learning_rate": 4.970320391069761e-05, "loss": 0.0697, "step": 105360 }, { "epoch": 3.5623246221981812, "grad_norm": 5.624368667602539, "learning_rate": 4.969493671135603e-05, "loss": 0.0657, "step": 105370 }, { "epoch": 3.5626626998884343, "grad_norm": 5.900984287261963, "learning_rate": 4.9686669520354764e-05, "loss": 0.0702, "step": 105380 }, { "epoch": 3.5630007775786874, "grad_norm": 7.878297805786133, "learning_rate": 4.9678402337919824e-05, "loss": 0.075, "step": 105390 }, { "epoch": 3.563338855268941, "grad_norm": 8.499402046203613, "learning_rate": 4.9670135164277276e-05, "loss": 0.0796, "step": 105400 }, { "epoch": 3.563676932959194, "grad_norm": 5.496952533721924, "learning_rate": 4.966186799965308e-05, "loss": 0.0666, "step": 105410 }, { "epoch": 3.564015010649447, "grad_norm": 8.471429824829102, "learning_rate": 4.9653600844273316e-05, "loss": 0.0733, "step": 105420 }, { "epoch": 3.5643530883397005, "grad_norm": 4.736939430236816, "learning_rate": 4.9645333698363934e-05, "loss": 0.0749, "step": 105430 }, { "epoch": 3.5646911660299536, "grad_norm": 4.52167272567749, "learning_rate": 4.963706656215103e-05, "loss": 0.0743, "step": 105440 }, { "epoch": 3.5650292437202067, "grad_norm": 4.640227794647217, "learning_rate": 4.962879943586057e-05, "loss": 0.0713, "step": 105450 }, { "epoch": 3.56536732141046, "grad_norm": 7.938912868499756, "learning_rate": 4.9620532319718594e-05, "loss": 0.0647, "step": 105460 }, { "epoch": 3.5657053991007133, "grad_norm": 18.080995559692383, "learning_rate": 4.9612265213951134e-05, "loss": 0.0688, "step": 105470 }, { "epoch": 3.5660434767909663, "grad_norm": 8.023988723754883, "learning_rate": 4.960399811878419e-05, "loss": 0.0721, "step": 105480 }, { "epoch": 3.56638155448122, "grad_norm": 3.6901025772094727, "learning_rate": 4.959573103444379e-05, "loss": 0.0718, "step": 105490 }, { "epoch": 3.566719632171473, "grad_norm": 11.735868453979492, "learning_rate": 4.958746396115596e-05, "loss": 0.068, "step": 105500 }, { "epoch": 3.567057709861726, "grad_norm": 8.033514022827148, "learning_rate": 4.95791968991467e-05, "loss": 0.0696, "step": 105510 }, { "epoch": 3.5673957875519795, "grad_norm": 10.198503494262695, "learning_rate": 4.957092984864203e-05, "loss": 0.0706, "step": 105520 }, { "epoch": 3.5677338652422326, "grad_norm": 6.9718708992004395, "learning_rate": 4.956266280986798e-05, "loss": 0.0688, "step": 105530 }, { "epoch": 3.5680719429324856, "grad_norm": 3.4447004795074463, "learning_rate": 4.9554395783050586e-05, "loss": 0.0685, "step": 105540 }, { "epoch": 3.568410020622739, "grad_norm": 4.0475592613220215, "learning_rate": 4.954612876841582e-05, "loss": 0.0639, "step": 105550 }, { "epoch": 3.5687480983129922, "grad_norm": 3.8510148525238037, "learning_rate": 4.953786176618974e-05, "loss": 0.0664, "step": 105560 }, { "epoch": 3.5690861760032453, "grad_norm": 2.8470661640167236, "learning_rate": 4.9529594776598334e-05, "loss": 0.0641, "step": 105570 }, { "epoch": 3.569424253693499, "grad_norm": 3.6841952800750732, "learning_rate": 4.952132779986764e-05, "loss": 0.0687, "step": 105580 }, { "epoch": 3.569762331383752, "grad_norm": 15.087964057922363, "learning_rate": 4.951306083622366e-05, "loss": 0.0753, "step": 105590 }, { "epoch": 3.570100409074005, "grad_norm": 5.923224925994873, "learning_rate": 4.9504793885892395e-05, "loss": 0.0599, "step": 105600 }, { "epoch": 3.5704384867642585, "grad_norm": 2.8217530250549316, "learning_rate": 4.9496526949099917e-05, "loss": 0.0672, "step": 105610 }, { "epoch": 3.5707765644545115, "grad_norm": 4.214160919189453, "learning_rate": 4.9488260026072164e-05, "loss": 0.0688, "step": 105620 }, { "epoch": 3.5711146421447646, "grad_norm": 6.279852867126465, "learning_rate": 4.947999311703523e-05, "loss": 0.068, "step": 105630 }, { "epoch": 3.571452719835018, "grad_norm": 15.099431991577148, "learning_rate": 4.9471726222215056e-05, "loss": 0.0653, "step": 105640 }, { "epoch": 3.571790797525271, "grad_norm": 4.576601982116699, "learning_rate": 4.946345934183769e-05, "loss": 0.0684, "step": 105650 }, { "epoch": 3.5721288752155247, "grad_norm": 6.563035488128662, "learning_rate": 4.9455192476129165e-05, "loss": 0.0726, "step": 105660 }, { "epoch": 3.5724669529057778, "grad_norm": 7.201695919036865, "learning_rate": 4.944692562531546e-05, "loss": 0.0671, "step": 105670 }, { "epoch": 3.572805030596031, "grad_norm": 10.706412315368652, "learning_rate": 4.94386587896226e-05, "loss": 0.076, "step": 105680 }, { "epoch": 3.5731431082862843, "grad_norm": 15.137368202209473, "learning_rate": 4.94303919692766e-05, "loss": 0.0724, "step": 105690 }, { "epoch": 3.5734811859765374, "grad_norm": 15.87414836883545, "learning_rate": 4.942212516450348e-05, "loss": 0.0778, "step": 105700 }, { "epoch": 3.5738192636667905, "grad_norm": 26.077022552490234, "learning_rate": 4.941385837552923e-05, "loss": 0.0691, "step": 105710 }, { "epoch": 3.574157341357044, "grad_norm": 6.23069429397583, "learning_rate": 4.940559160257985e-05, "loss": 0.0712, "step": 105720 }, { "epoch": 3.574495419047297, "grad_norm": 7.085961818695068, "learning_rate": 4.9397324845881404e-05, "loss": 0.0706, "step": 105730 }, { "epoch": 3.57483349673755, "grad_norm": 16.3780574798584, "learning_rate": 4.9389058105659855e-05, "loss": 0.072, "step": 105740 }, { "epoch": 3.5751715744278036, "grad_norm": 6.983273506164551, "learning_rate": 4.938079138214125e-05, "loss": 0.0745, "step": 105750 }, { "epoch": 3.5755096521180567, "grad_norm": 2.9431679248809814, "learning_rate": 4.937252467555155e-05, "loss": 0.0638, "step": 105760 }, { "epoch": 3.5758477298083102, "grad_norm": 4.204875469207764, "learning_rate": 4.936425798611681e-05, "loss": 0.0718, "step": 105770 }, { "epoch": 3.5761858074985633, "grad_norm": 3.414668321609497, "learning_rate": 4.9355991314063e-05, "loss": 0.0764, "step": 105780 }, { "epoch": 3.5765238851888164, "grad_norm": 6.197792053222656, "learning_rate": 4.9347724659616155e-05, "loss": 0.0696, "step": 105790 }, { "epoch": 3.57686196287907, "grad_norm": 11.038006782531738, "learning_rate": 4.933945802300227e-05, "loss": 0.0606, "step": 105800 }, { "epoch": 3.577200040569323, "grad_norm": 38.12348175048828, "learning_rate": 4.933119140444734e-05, "loss": 0.0677, "step": 105810 }, { "epoch": 3.577538118259576, "grad_norm": 17.048446655273438, "learning_rate": 4.9322924804177424e-05, "loss": 0.0705, "step": 105820 }, { "epoch": 3.5778761959498295, "grad_norm": 18.297836303710938, "learning_rate": 4.931465822241845e-05, "loss": 0.0625, "step": 105830 }, { "epoch": 3.5782142736400826, "grad_norm": 3.931387186050415, "learning_rate": 4.9306391659396476e-05, "loss": 0.0639, "step": 105840 }, { "epoch": 3.5785523513303357, "grad_norm": 6.1140947341918945, "learning_rate": 4.92981251153375e-05, "loss": 0.0704, "step": 105850 }, { "epoch": 3.578890429020589, "grad_norm": 2.9943556785583496, "learning_rate": 4.928985859046752e-05, "loss": 0.0647, "step": 105860 }, { "epoch": 3.5792285067108423, "grad_norm": 17.4770450592041, "learning_rate": 4.928159208501254e-05, "loss": 0.0714, "step": 105870 }, { "epoch": 3.5795665844010953, "grad_norm": 4.492607116699219, "learning_rate": 4.927332559919855e-05, "loss": 0.0731, "step": 105880 }, { "epoch": 3.579904662091349, "grad_norm": 4.4372239112854, "learning_rate": 4.926505913325158e-05, "loss": 0.0768, "step": 105890 }, { "epoch": 3.580242739781602, "grad_norm": 7.922103404998779, "learning_rate": 4.9256792687397605e-05, "loss": 0.0691, "step": 105900 }, { "epoch": 3.580580817471855, "grad_norm": 18.373619079589844, "learning_rate": 4.9248526261862633e-05, "loss": 0.0718, "step": 105910 }, { "epoch": 3.5809188951621085, "grad_norm": 4.9359025955200195, "learning_rate": 4.92402598568727e-05, "loss": 0.0641, "step": 105920 }, { "epoch": 3.5812569728523616, "grad_norm": 3.5395050048828125, "learning_rate": 4.923199347265374e-05, "loss": 0.0651, "step": 105930 }, { "epoch": 3.5815950505426146, "grad_norm": 3.3201851844787598, "learning_rate": 4.922372710943182e-05, "loss": 0.0695, "step": 105940 }, { "epoch": 3.581933128232868, "grad_norm": 1.996606707572937, "learning_rate": 4.92154607674329e-05, "loss": 0.0601, "step": 105950 }, { "epoch": 3.582271205923121, "grad_norm": 3.513019561767578, "learning_rate": 4.9207194446882984e-05, "loss": 0.0734, "step": 105960 }, { "epoch": 3.5826092836133743, "grad_norm": 5.101959705352783, "learning_rate": 4.9198928148008075e-05, "loss": 0.0608, "step": 105970 }, { "epoch": 3.582947361303628, "grad_norm": 2.832906484603882, "learning_rate": 4.9190661871034164e-05, "loss": 0.0651, "step": 105980 }, { "epoch": 3.583285438993881, "grad_norm": 6.812157154083252, "learning_rate": 4.918239561618727e-05, "loss": 0.0638, "step": 105990 }, { "epoch": 3.583623516684134, "grad_norm": 1.3132238388061523, "learning_rate": 4.9174129383693355e-05, "loss": 0.0559, "step": 106000 }, { "epoch": 3.5839615943743874, "grad_norm": 1.6178721189498901, "learning_rate": 4.916586317377845e-05, "loss": 0.0639, "step": 106010 }, { "epoch": 3.5842996720646405, "grad_norm": 2.8727471828460693, "learning_rate": 4.915759698666851e-05, "loss": 0.0666, "step": 106020 }, { "epoch": 3.5846377497548936, "grad_norm": 3.3204476833343506, "learning_rate": 4.914933082258956e-05, "loss": 0.068, "step": 106030 }, { "epoch": 3.584975827445147, "grad_norm": 3.615323781967163, "learning_rate": 4.9141064681767594e-05, "loss": 0.0657, "step": 106040 }, { "epoch": 3.5853139051354, "grad_norm": 1.958910584449768, "learning_rate": 4.913279856442859e-05, "loss": 0.0702, "step": 106050 }, { "epoch": 3.5856519828256532, "grad_norm": 4.368844509124756, "learning_rate": 4.912453247079856e-05, "loss": 0.0632, "step": 106060 }, { "epoch": 3.5859900605159067, "grad_norm": 6.714580059051514, "learning_rate": 4.911626640110347e-05, "loss": 0.0657, "step": 106070 }, { "epoch": 3.58632813820616, "grad_norm": 4.248196125030518, "learning_rate": 4.910800035556933e-05, "loss": 0.0652, "step": 106080 }, { "epoch": 3.586666215896413, "grad_norm": 4.961484909057617, "learning_rate": 4.909973433442212e-05, "loss": 0.068, "step": 106090 }, { "epoch": 3.5870042935866664, "grad_norm": 2.6009533405303955, "learning_rate": 4.909146833788783e-05, "loss": 0.067, "step": 106100 }, { "epoch": 3.5873423712769195, "grad_norm": 2.23394513130188, "learning_rate": 4.908320236619248e-05, "loss": 0.0648, "step": 106110 }, { "epoch": 3.5876804489671725, "grad_norm": 2.4956655502319336, "learning_rate": 4.9074936419562e-05, "loss": 0.0699, "step": 106120 }, { "epoch": 3.588018526657426, "grad_norm": 2.118515968322754, "learning_rate": 4.906667049822244e-05, "loss": 0.0618, "step": 106130 }, { "epoch": 3.588356604347679, "grad_norm": 1.8200336694717407, "learning_rate": 4.905840460239973e-05, "loss": 0.0602, "step": 106140 }, { "epoch": 3.588694682037932, "grad_norm": 6.429365158081055, "learning_rate": 4.905013873231991e-05, "loss": 0.0611, "step": 106150 }, { "epoch": 3.5890327597281857, "grad_norm": 4.965824604034424, "learning_rate": 4.9041872888208925e-05, "loss": 0.0659, "step": 106160 }, { "epoch": 3.5893708374184388, "grad_norm": 3.9418835639953613, "learning_rate": 4.9033607070292774e-05, "loss": 0.0594, "step": 106170 }, { "epoch": 3.589708915108692, "grad_norm": 2.497528076171875, "learning_rate": 4.902534127879746e-05, "loss": 0.0627, "step": 106180 }, { "epoch": 3.5900469927989453, "grad_norm": 2.7004013061523438, "learning_rate": 4.901707551394893e-05, "loss": 0.0707, "step": 106190 }, { "epoch": 3.5903850704891984, "grad_norm": 11.807015419006348, "learning_rate": 4.90088097759732e-05, "loss": 0.0668, "step": 106200 }, { "epoch": 3.5907231481794515, "grad_norm": 1.3964539766311646, "learning_rate": 4.900054406509622e-05, "loss": 0.0654, "step": 106210 }, { "epoch": 3.591061225869705, "grad_norm": 3.008741855621338, "learning_rate": 4.8992278381543994e-05, "loss": 0.0616, "step": 106220 }, { "epoch": 3.591399303559958, "grad_norm": 1.4032620191574097, "learning_rate": 4.8984012725542514e-05, "loss": 0.0593, "step": 106230 }, { "epoch": 3.591737381250211, "grad_norm": 2.558790922164917, "learning_rate": 4.897574709731773e-05, "loss": 0.0685, "step": 106240 }, { "epoch": 3.5920754589404646, "grad_norm": 2.168389320373535, "learning_rate": 4.8967481497095654e-05, "loss": 0.0684, "step": 106250 }, { "epoch": 3.5924135366307177, "grad_norm": 9.694836616516113, "learning_rate": 4.895921592510223e-05, "loss": 0.0734, "step": 106260 }, { "epoch": 3.592751614320971, "grad_norm": 3.2741591930389404, "learning_rate": 4.8950950381563456e-05, "loss": 0.0698, "step": 106270 }, { "epoch": 3.5930896920112243, "grad_norm": 2.5359597206115723, "learning_rate": 4.8942684866705305e-05, "loss": 0.0703, "step": 106280 }, { "epoch": 3.5934277697014774, "grad_norm": 7.308677673339844, "learning_rate": 4.8934419380753747e-05, "loss": 0.0655, "step": 106290 }, { "epoch": 3.5937658473917304, "grad_norm": 1.6901594400405884, "learning_rate": 4.8926153923934774e-05, "loss": 0.0659, "step": 106300 }, { "epoch": 3.594103925081984, "grad_norm": 2.212611675262451, "learning_rate": 4.891788849647433e-05, "loss": 0.0629, "step": 106310 }, { "epoch": 3.594442002772237, "grad_norm": 6.152832508087158, "learning_rate": 4.8909623098598436e-05, "loss": 0.0691, "step": 106320 }, { "epoch": 3.59478008046249, "grad_norm": 3.4655070304870605, "learning_rate": 4.8901357730533006e-05, "loss": 0.0674, "step": 106330 }, { "epoch": 3.5951181581527436, "grad_norm": 2.8162198066711426, "learning_rate": 4.889309239250408e-05, "loss": 0.0645, "step": 106340 }, { "epoch": 3.5954562358429967, "grad_norm": 2.5378832817077637, "learning_rate": 4.888482708473755e-05, "loss": 0.0625, "step": 106350 }, { "epoch": 3.5957943135332497, "grad_norm": 4.75209903717041, "learning_rate": 4.8876561807459445e-05, "loss": 0.0683, "step": 106360 }, { "epoch": 3.5961323912235033, "grad_norm": 1.8280160427093506, "learning_rate": 4.8868296560895734e-05, "loss": 0.0733, "step": 106370 }, { "epoch": 3.5964704689137563, "grad_norm": 1.5053794384002686, "learning_rate": 4.8860031345272355e-05, "loss": 0.0776, "step": 106380 }, { "epoch": 3.5968085466040094, "grad_norm": 2.0953757762908936, "learning_rate": 4.8851766160815296e-05, "loss": 0.0613, "step": 106390 }, { "epoch": 3.597146624294263, "grad_norm": 1.6431630849838257, "learning_rate": 4.884350100775052e-05, "loss": 0.0559, "step": 106400 }, { "epoch": 3.597484701984516, "grad_norm": 4.7038187980651855, "learning_rate": 4.883523588630397e-05, "loss": 0.0663, "step": 106410 }, { "epoch": 3.597822779674769, "grad_norm": 1.194467306137085, "learning_rate": 4.882697079670167e-05, "loss": 0.0628, "step": 106420 }, { "epoch": 3.5981608573650226, "grad_norm": 1.9380306005477905, "learning_rate": 4.8818705739169516e-05, "loss": 0.0598, "step": 106430 }, { "epoch": 3.5984989350552756, "grad_norm": 1.7908939123153687, "learning_rate": 4.881044071393352e-05, "loss": 0.0634, "step": 106440 }, { "epoch": 3.5988370127455287, "grad_norm": 2.5328381061553955, "learning_rate": 4.880217572121962e-05, "loss": 0.0654, "step": 106450 }, { "epoch": 3.599175090435782, "grad_norm": 1.372521996498108, "learning_rate": 4.87939107612538e-05, "loss": 0.0663, "step": 106460 }, { "epoch": 3.5995131681260353, "grad_norm": 2.1954901218414307, "learning_rate": 4.878564583426199e-05, "loss": 0.0597, "step": 106470 }, { "epoch": 3.5998512458162883, "grad_norm": 3.427112579345703, "learning_rate": 4.877738094047018e-05, "loss": 0.067, "step": 106480 }, { "epoch": 3.600189323506542, "grad_norm": 5.1319899559021, "learning_rate": 4.8769116080104317e-05, "loss": 0.0731, "step": 106490 }, { "epoch": 3.600527401196795, "grad_norm": 1.9348217248916626, "learning_rate": 4.876085125339034e-05, "loss": 0.0591, "step": 106500 }, { "epoch": 3.600865478887048, "grad_norm": 2.2586591243743896, "learning_rate": 4.875258646055424e-05, "loss": 0.0596, "step": 106510 }, { "epoch": 3.6012035565773015, "grad_norm": 3.6260879039764404, "learning_rate": 4.8744321701821946e-05, "loss": 0.0668, "step": 106520 }, { "epoch": 3.6015416342675546, "grad_norm": 1.7936800718307495, "learning_rate": 4.873605697741945e-05, "loss": 0.0634, "step": 106530 }, { "epoch": 3.6018797119578077, "grad_norm": 2.2704875469207764, "learning_rate": 4.872779228757264e-05, "loss": 0.0648, "step": 106540 }, { "epoch": 3.602217789648061, "grad_norm": 3.0154354572296143, "learning_rate": 4.8719527632507536e-05, "loss": 0.0686, "step": 106550 }, { "epoch": 3.6025558673383142, "grad_norm": 1.7889742851257324, "learning_rate": 4.871126301245007e-05, "loss": 0.068, "step": 106560 }, { "epoch": 3.6028939450285673, "grad_norm": 2.2823591232299805, "learning_rate": 4.870299842762618e-05, "loss": 0.0674, "step": 106570 }, { "epoch": 3.603232022718821, "grad_norm": 1.2354249954223633, "learning_rate": 4.8694733878261836e-05, "loss": 0.0701, "step": 106580 }, { "epoch": 3.603570100409074, "grad_norm": 2.7170426845550537, "learning_rate": 4.868646936458297e-05, "loss": 0.0608, "step": 106590 }, { "epoch": 3.603908178099327, "grad_norm": 4.0688157081604, "learning_rate": 4.867820488681551e-05, "loss": 0.0618, "step": 106600 }, { "epoch": 3.6042462557895805, "grad_norm": 1.6570146083831787, "learning_rate": 4.866994044518548e-05, "loss": 0.0653, "step": 106610 }, { "epoch": 3.6045843334798335, "grad_norm": 2.2087759971618652, "learning_rate": 4.866167603991875e-05, "loss": 0.0612, "step": 106620 }, { "epoch": 3.604922411170087, "grad_norm": 2.261009693145752, "learning_rate": 4.8653411671241316e-05, "loss": 0.0656, "step": 106630 }, { "epoch": 3.60526048886034, "grad_norm": 1.401828646659851, "learning_rate": 4.864514733937907e-05, "loss": 0.0604, "step": 106640 }, { "epoch": 3.605598566550593, "grad_norm": 17.194826126098633, "learning_rate": 4.8636883044558014e-05, "loss": 0.0636, "step": 106650 }, { "epoch": 3.6059366442408467, "grad_norm": 2.1118500232696533, "learning_rate": 4.862861878700405e-05, "loss": 0.0658, "step": 106660 }, { "epoch": 3.6062747219310998, "grad_norm": 2.540239095687866, "learning_rate": 4.862035456694313e-05, "loss": 0.0677, "step": 106670 }, { "epoch": 3.606612799621353, "grad_norm": 1.358582615852356, "learning_rate": 4.861209038460122e-05, "loss": 0.0702, "step": 106680 }, { "epoch": 3.6069508773116064, "grad_norm": 2.6101560592651367, "learning_rate": 4.8603826240204214e-05, "loss": 0.0697, "step": 106690 }, { "epoch": 3.6072889550018594, "grad_norm": 3.03627872467041, "learning_rate": 4.859556213397809e-05, "loss": 0.0741, "step": 106700 }, { "epoch": 3.6076270326921125, "grad_norm": 2.005174160003662, "learning_rate": 4.8587298066148756e-05, "loss": 0.0646, "step": 106710 }, { "epoch": 3.607965110382366, "grad_norm": 2.4196362495422363, "learning_rate": 4.857903403694217e-05, "loss": 0.0723, "step": 106720 }, { "epoch": 3.608303188072619, "grad_norm": 1.6511437892913818, "learning_rate": 4.857077004658425e-05, "loss": 0.0659, "step": 106730 }, { "epoch": 3.6086412657628726, "grad_norm": 1.294138789176941, "learning_rate": 4.856250609530094e-05, "loss": 0.0709, "step": 106740 }, { "epoch": 3.6089793434531257, "grad_norm": 2.000208616256714, "learning_rate": 4.855424218331819e-05, "loss": 0.0656, "step": 106750 }, { "epoch": 3.6093174211433787, "grad_norm": 11.441970825195312, "learning_rate": 4.8545978310861904e-05, "loss": 0.0707, "step": 106760 }, { "epoch": 3.6096554988336322, "grad_norm": 3.082404136657715, "learning_rate": 4.853771447815803e-05, "loss": 0.0634, "step": 106770 }, { "epoch": 3.6099935765238853, "grad_norm": 1.3519635200500488, "learning_rate": 4.8529450685432485e-05, "loss": 0.0659, "step": 106780 }, { "epoch": 3.6103316542141384, "grad_norm": 1.9251829385757446, "learning_rate": 4.852118693291122e-05, "loss": 0.0661, "step": 106790 }, { "epoch": 3.610669731904392, "grad_norm": 2.7544920444488525, "learning_rate": 4.851292322082013e-05, "loss": 0.0623, "step": 106800 }, { "epoch": 3.611007809594645, "grad_norm": 2.0844967365264893, "learning_rate": 4.850465954938516e-05, "loss": 0.0682, "step": 106810 }, { "epoch": 3.611345887284898, "grad_norm": 2.2608015537261963, "learning_rate": 4.849639591883226e-05, "loss": 0.0649, "step": 106820 }, { "epoch": 3.6116839649751515, "grad_norm": 1.45806884765625, "learning_rate": 4.84881323293873e-05, "loss": 0.062, "step": 106830 }, { "epoch": 3.6120220426654046, "grad_norm": 1.537959098815918, "learning_rate": 4.847986878127627e-05, "loss": 0.058, "step": 106840 }, { "epoch": 3.6123601203556577, "grad_norm": 2.327103614807129, "learning_rate": 4.847160527472503e-05, "loss": 0.071, "step": 106850 }, { "epoch": 3.612698198045911, "grad_norm": 6.408167839050293, "learning_rate": 4.846334180995953e-05, "loss": 0.0679, "step": 106860 }, { "epoch": 3.6130362757361643, "grad_norm": 4.95030403137207, "learning_rate": 4.84550783872057e-05, "loss": 0.0612, "step": 106870 }, { "epoch": 3.6133743534264173, "grad_norm": 1.735991358757019, "learning_rate": 4.844681500668943e-05, "loss": 0.0633, "step": 106880 }, { "epoch": 3.613712431116671, "grad_norm": 4.28515625, "learning_rate": 4.8438551668636675e-05, "loss": 0.0687, "step": 106890 }, { "epoch": 3.614050508806924, "grad_norm": 6.643544673919678, "learning_rate": 4.8430288373273315e-05, "loss": 0.0619, "step": 106900 }, { "epoch": 3.614388586497177, "grad_norm": 1.0345767736434937, "learning_rate": 4.8422025120825286e-05, "loss": 0.0653, "step": 106910 }, { "epoch": 3.6147266641874305, "grad_norm": 2.1589179039001465, "learning_rate": 4.8413761911518494e-05, "loss": 0.061, "step": 106920 }, { "epoch": 3.6150647418776836, "grad_norm": 3.2049758434295654, "learning_rate": 4.840549874557885e-05, "loss": 0.0743, "step": 106930 }, { "epoch": 3.6154028195679366, "grad_norm": 3.3527672290802, "learning_rate": 4.839723562323228e-05, "loss": 0.0652, "step": 106940 }, { "epoch": 3.61574089725819, "grad_norm": 1.031941533088684, "learning_rate": 4.838897254470468e-05, "loss": 0.066, "step": 106950 }, { "epoch": 3.616078974948443, "grad_norm": 1.2717578411102295, "learning_rate": 4.838070951022198e-05, "loss": 0.065, "step": 106960 }, { "epoch": 3.6164170526386963, "grad_norm": 7.842342853546143, "learning_rate": 4.8372446520010064e-05, "loss": 0.0646, "step": 106970 }, { "epoch": 3.61675513032895, "grad_norm": 1.3220337629318237, "learning_rate": 4.836418357429485e-05, "loss": 0.0584, "step": 106980 }, { "epoch": 3.617093208019203, "grad_norm": 1.5923563241958618, "learning_rate": 4.835592067330224e-05, "loss": 0.0679, "step": 106990 }, { "epoch": 3.617431285709456, "grad_norm": 1.4751949310302734, "learning_rate": 4.834765781725814e-05, "loss": 0.0635, "step": 107000 }, { "epoch": 3.6177693633997094, "grad_norm": 2.3579554557800293, "learning_rate": 4.833939500638846e-05, "loss": 0.0631, "step": 107010 }, { "epoch": 3.6181074410899625, "grad_norm": 3.7459287643432617, "learning_rate": 4.833113224091908e-05, "loss": 0.0604, "step": 107020 }, { "epoch": 3.6184455187802156, "grad_norm": 7.935959339141846, "learning_rate": 4.832286952107595e-05, "loss": 0.0644, "step": 107030 }, { "epoch": 3.618783596470469, "grad_norm": 2.0672364234924316, "learning_rate": 4.8314606847084904e-05, "loss": 0.0671, "step": 107040 }, { "epoch": 3.619121674160722, "grad_norm": 6.088987827301025, "learning_rate": 4.8306344219171885e-05, "loss": 0.065, "step": 107050 }, { "epoch": 3.6194597518509752, "grad_norm": 1.3883953094482422, "learning_rate": 4.829808163756279e-05, "loss": 0.0696, "step": 107060 }, { "epoch": 3.6197978295412288, "grad_norm": 3.6061835289001465, "learning_rate": 4.828981910248349e-05, "loss": 0.0693, "step": 107070 }, { "epoch": 3.620135907231482, "grad_norm": 2.2433102130889893, "learning_rate": 4.828155661415991e-05, "loss": 0.0676, "step": 107080 }, { "epoch": 3.620473984921735, "grad_norm": 1.723670482635498, "learning_rate": 4.8273294172817904e-05, "loss": 0.0698, "step": 107090 }, { "epoch": 3.6208120626119884, "grad_norm": 1.4056203365325928, "learning_rate": 4.82650317786834e-05, "loss": 0.0673, "step": 107100 }, { "epoch": 3.6211501403022415, "grad_norm": 1.4388151168823242, "learning_rate": 4.825676943198228e-05, "loss": 0.0622, "step": 107110 }, { "epoch": 3.6214882179924945, "grad_norm": 3.2229349613189697, "learning_rate": 4.824850713294041e-05, "loss": 0.0622, "step": 107120 }, { "epoch": 3.621826295682748, "grad_norm": 3.2425472736358643, "learning_rate": 4.824024488178372e-05, "loss": 0.0629, "step": 107130 }, { "epoch": 3.622164373373001, "grad_norm": 2.1558377742767334, "learning_rate": 4.823198267873805e-05, "loss": 0.0685, "step": 107140 }, { "epoch": 3.622502451063254, "grad_norm": 3.326432704925537, "learning_rate": 4.822372052402933e-05, "loss": 0.066, "step": 107150 }, { "epoch": 3.6228405287535077, "grad_norm": 1.2116422653198242, "learning_rate": 4.821545841788341e-05, "loss": 0.0611, "step": 107160 }, { "epoch": 3.6231786064437608, "grad_norm": 6.366746425628662, "learning_rate": 4.82071963605262e-05, "loss": 0.0649, "step": 107170 }, { "epoch": 3.623516684134014, "grad_norm": 2.4042298793792725, "learning_rate": 4.819893435218356e-05, "loss": 0.0669, "step": 107180 }, { "epoch": 3.6238547618242674, "grad_norm": 1.3858133554458618, "learning_rate": 4.819067239308137e-05, "loss": 0.0622, "step": 107190 }, { "epoch": 3.6241928395145204, "grad_norm": 2.250223159790039, "learning_rate": 4.8182410483445534e-05, "loss": 0.066, "step": 107200 }, { "epoch": 3.6245309172047735, "grad_norm": 3.0243117809295654, "learning_rate": 4.81741486235019e-05, "loss": 0.0642, "step": 107210 }, { "epoch": 3.624868994895027, "grad_norm": 1.2556116580963135, "learning_rate": 4.816588681347637e-05, "loss": 0.0749, "step": 107220 }, { "epoch": 3.62520707258528, "grad_norm": 5.263539791107178, "learning_rate": 4.815762505359478e-05, "loss": 0.0657, "step": 107230 }, { "epoch": 3.625545150275533, "grad_norm": 1.8025195598602295, "learning_rate": 4.814936334408304e-05, "loss": 0.0652, "step": 107240 }, { "epoch": 3.6258832279657867, "grad_norm": 5.382535934448242, "learning_rate": 4.814110168516702e-05, "loss": 0.0647, "step": 107250 }, { "epoch": 3.6262213056560397, "grad_norm": 2.678433656692505, "learning_rate": 4.8132840077072575e-05, "loss": 0.057, "step": 107260 }, { "epoch": 3.626559383346293, "grad_norm": 1.260231852531433, "learning_rate": 4.8124578520025595e-05, "loss": 0.0605, "step": 107270 }, { "epoch": 3.6268974610365463, "grad_norm": 2.548243999481201, "learning_rate": 4.811631701425192e-05, "loss": 0.0599, "step": 107280 }, { "epoch": 3.6272355387267994, "grad_norm": 5.065525054931641, "learning_rate": 4.810805555997744e-05, "loss": 0.0646, "step": 107290 }, { "epoch": 3.6275736164170524, "grad_norm": 1.8478789329528809, "learning_rate": 4.8099794157428e-05, "loss": 0.0611, "step": 107300 }, { "epoch": 3.627911694107306, "grad_norm": 5.97162389755249, "learning_rate": 4.8091532806829456e-05, "loss": 0.07, "step": 107310 }, { "epoch": 3.628249771797559, "grad_norm": 3.358431339263916, "learning_rate": 4.808327150840772e-05, "loss": 0.0688, "step": 107320 }, { "epoch": 3.628587849487812, "grad_norm": 65.17982482910156, "learning_rate": 4.80750102623886e-05, "loss": 0.0744, "step": 107330 }, { "epoch": 3.6289259271780656, "grad_norm": 2.361724853515625, "learning_rate": 4.806674906899801e-05, "loss": 0.0601, "step": 107340 }, { "epoch": 3.6292640048683187, "grad_norm": 4.077786922454834, "learning_rate": 4.805848792846173e-05, "loss": 0.06, "step": 107350 }, { "epoch": 3.6296020825585718, "grad_norm": 2.254289388656616, "learning_rate": 4.8050226841005685e-05, "loss": 0.0675, "step": 107360 }, { "epoch": 3.6299401602488253, "grad_norm": 12.774996757507324, "learning_rate": 4.80419658068557e-05, "loss": 0.0703, "step": 107370 }, { "epoch": 3.6302782379390783, "grad_norm": 1.1469333171844482, "learning_rate": 4.803370482623763e-05, "loss": 0.0647, "step": 107380 }, { "epoch": 3.6306163156293314, "grad_norm": 1.7790287733078003, "learning_rate": 4.802544389937734e-05, "loss": 0.0672, "step": 107390 }, { "epoch": 3.630954393319585, "grad_norm": 3.903550148010254, "learning_rate": 4.8017183026500664e-05, "loss": 0.0668, "step": 107400 }, { "epoch": 3.631292471009838, "grad_norm": 2.560912847518921, "learning_rate": 4.800892220783347e-05, "loss": 0.061, "step": 107410 }, { "epoch": 3.631630548700091, "grad_norm": 2.2881224155426025, "learning_rate": 4.800066144360159e-05, "loss": 0.06, "step": 107420 }, { "epoch": 3.6319686263903446, "grad_norm": 5.16703987121582, "learning_rate": 4.7992400734030856e-05, "loss": 0.0709, "step": 107430 }, { "epoch": 3.6323067040805976, "grad_norm": 2.378143787384033, "learning_rate": 4.798414007934716e-05, "loss": 0.0613, "step": 107440 }, { "epoch": 3.6326447817708507, "grad_norm": 5.1581292152404785, "learning_rate": 4.7975879479776306e-05, "loss": 0.0659, "step": 107450 }, { "epoch": 3.632982859461104, "grad_norm": 3.5443196296691895, "learning_rate": 4.796761893554415e-05, "loss": 0.0679, "step": 107460 }, { "epoch": 3.6333209371513573, "grad_norm": 1.900335669517517, "learning_rate": 4.795935844687652e-05, "loss": 0.0648, "step": 107470 }, { "epoch": 3.6336590148416104, "grad_norm": 3.3374736309051514, "learning_rate": 4.7951098013999276e-05, "loss": 0.0697, "step": 107480 }, { "epoch": 3.633997092531864, "grad_norm": 6.1485700607299805, "learning_rate": 4.794283763713823e-05, "loss": 0.0591, "step": 107490 }, { "epoch": 3.634335170222117, "grad_norm": 1.8217358589172363, "learning_rate": 4.7934577316519235e-05, "loss": 0.0651, "step": 107500 }, { "epoch": 3.63467324791237, "grad_norm": 2.8227901458740234, "learning_rate": 4.792631705236813e-05, "loss": 0.0622, "step": 107510 }, { "epoch": 3.6350113256026235, "grad_norm": 1.8512499332427979, "learning_rate": 4.7918056844910714e-05, "loss": 0.0651, "step": 107520 }, { "epoch": 3.6353494032928766, "grad_norm": 2.120081901550293, "learning_rate": 4.790979669437287e-05, "loss": 0.0617, "step": 107530 }, { "epoch": 3.6356874809831297, "grad_norm": 1.4684449434280396, "learning_rate": 4.790153660098037e-05, "loss": 0.0649, "step": 107540 }, { "epoch": 3.636025558673383, "grad_norm": 2.173030376434326, "learning_rate": 4.78932765649591e-05, "loss": 0.0642, "step": 107550 }, { "epoch": 3.6363636363636362, "grad_norm": 1.326777458190918, "learning_rate": 4.788501658653484e-05, "loss": 0.0651, "step": 107560 }, { "epoch": 3.6367017140538893, "grad_norm": 1.5776352882385254, "learning_rate": 4.787675666593343e-05, "loss": 0.0697, "step": 107570 }, { "epoch": 3.637039791744143, "grad_norm": 1.1830681562423706, "learning_rate": 4.786849680338071e-05, "loss": 0.0698, "step": 107580 }, { "epoch": 3.637377869434396, "grad_norm": 7.035912990570068, "learning_rate": 4.786023699910247e-05, "loss": 0.0695, "step": 107590 }, { "epoch": 3.6377159471246494, "grad_norm": 0.9220815300941467, "learning_rate": 4.785197725332456e-05, "loss": 0.0663, "step": 107600 }, { "epoch": 3.6380540248149025, "grad_norm": 1.3767387866973877, "learning_rate": 4.7843717566272775e-05, "loss": 0.0622, "step": 107610 }, { "epoch": 3.6383921025051555, "grad_norm": 1.4703905582427979, "learning_rate": 4.7835457938172936e-05, "loss": 0.0567, "step": 107620 }, { "epoch": 3.638730180195409, "grad_norm": 0.9543266892433167, "learning_rate": 4.7827198369250884e-05, "loss": 0.055, "step": 107630 }, { "epoch": 3.639068257885662, "grad_norm": 1.8820902109146118, "learning_rate": 4.7818938859732385e-05, "loss": 0.0625, "step": 107640 }, { "epoch": 3.639406335575915, "grad_norm": 1.1291837692260742, "learning_rate": 4.7810679409843304e-05, "loss": 0.0676, "step": 107650 }, { "epoch": 3.6397444132661687, "grad_norm": 1.4177796840667725, "learning_rate": 4.7802420019809413e-05, "loss": 0.0616, "step": 107660 }, { "epoch": 3.640082490956422, "grad_norm": 5.01289701461792, "learning_rate": 4.7794160689856546e-05, "loss": 0.0662, "step": 107670 }, { "epoch": 3.640420568646675, "grad_norm": 1.5109155178070068, "learning_rate": 4.7785901420210486e-05, "loss": 0.0603, "step": 107680 }, { "epoch": 3.6407586463369284, "grad_norm": 3.2556605339050293, "learning_rate": 4.7777642211097046e-05, "loss": 0.0712, "step": 107690 }, { "epoch": 3.6410967240271814, "grad_norm": 1.8069615364074707, "learning_rate": 4.7769383062742045e-05, "loss": 0.0634, "step": 107700 }, { "epoch": 3.641434801717435, "grad_norm": 1.6757056713104248, "learning_rate": 4.7761123975371255e-05, "loss": 0.0676, "step": 107710 }, { "epoch": 3.641772879407688, "grad_norm": 2.612985849380493, "learning_rate": 4.775286494921052e-05, "loss": 0.0654, "step": 107720 }, { "epoch": 3.642110957097941, "grad_norm": 1.5867975950241089, "learning_rate": 4.774460598448558e-05, "loss": 0.0634, "step": 107730 }, { "epoch": 3.6424490347881946, "grad_norm": 4.495205402374268, "learning_rate": 4.77363470814223e-05, "loss": 0.0635, "step": 107740 }, { "epoch": 3.6427871124784477, "grad_norm": 1.259891390800476, "learning_rate": 4.772808824024641e-05, "loss": 0.0671, "step": 107750 }, { "epoch": 3.6431251901687007, "grad_norm": 0.7793198823928833, "learning_rate": 4.771982946118375e-05, "loss": 0.0626, "step": 107760 }, { "epoch": 3.6434632678589542, "grad_norm": 1.727860450744629, "learning_rate": 4.77115707444601e-05, "loss": 0.0657, "step": 107770 }, { "epoch": 3.6438013455492073, "grad_norm": 12.194340705871582, "learning_rate": 4.7703312090301234e-05, "loss": 0.0643, "step": 107780 }, { "epoch": 3.6441394232394604, "grad_norm": 2.127976179122925, "learning_rate": 4.769505349893296e-05, "loss": 0.0661, "step": 107790 }, { "epoch": 3.644477500929714, "grad_norm": 1.21916663646698, "learning_rate": 4.7686794970581053e-05, "loss": 0.0653, "step": 107800 }, { "epoch": 3.644815578619967, "grad_norm": 4.460212230682373, "learning_rate": 4.7678536505471295e-05, "loss": 0.0659, "step": 107810 }, { "epoch": 3.64515365631022, "grad_norm": 1.5603550672531128, "learning_rate": 4.7670278103829505e-05, "loss": 0.0633, "step": 107820 }, { "epoch": 3.6454917340004735, "grad_norm": 0.9577100276947021, "learning_rate": 4.766201976588141e-05, "loss": 0.0679, "step": 107830 }, { "epoch": 3.6458298116907266, "grad_norm": 3.801269292831421, "learning_rate": 4.7653761491852836e-05, "loss": 0.0655, "step": 107840 }, { "epoch": 3.6461678893809797, "grad_norm": 2.145113706588745, "learning_rate": 4.7645503281969514e-05, "loss": 0.0576, "step": 107850 }, { "epoch": 3.646505967071233, "grad_norm": 1.3547024726867676, "learning_rate": 4.763724513645727e-05, "loss": 0.066, "step": 107860 }, { "epoch": 3.6468440447614863, "grad_norm": 0.963437557220459, "learning_rate": 4.7628987055541846e-05, "loss": 0.0682, "step": 107870 }, { "epoch": 3.6471821224517393, "grad_norm": 1.1565303802490234, "learning_rate": 4.762072903944903e-05, "loss": 0.0644, "step": 107880 }, { "epoch": 3.647520200141993, "grad_norm": 0.9939778447151184, "learning_rate": 4.761247108840459e-05, "loss": 0.0627, "step": 107890 }, { "epoch": 3.647858277832246, "grad_norm": 0.8311955332756042, "learning_rate": 4.760421320263429e-05, "loss": 0.0618, "step": 107900 }, { "epoch": 3.648196355522499, "grad_norm": 2.2311360836029053, "learning_rate": 4.759595538236391e-05, "loss": 0.0644, "step": 107910 }, { "epoch": 3.6485344332127525, "grad_norm": 1.697270154953003, "learning_rate": 4.758769762781919e-05, "loss": 0.0623, "step": 107920 }, { "epoch": 3.6488725109030056, "grad_norm": 0.7918906807899475, "learning_rate": 4.757943993922593e-05, "loss": 0.0629, "step": 107930 }, { "epoch": 3.6492105885932586, "grad_norm": 1.4362858533859253, "learning_rate": 4.7571182316809845e-05, "loss": 0.0603, "step": 107940 }, { "epoch": 3.649548666283512, "grad_norm": 1.1700472831726074, "learning_rate": 4.756292476079674e-05, "loss": 0.0681, "step": 107950 }, { "epoch": 3.6498867439737652, "grad_norm": 2.0045006275177, "learning_rate": 4.755466727141236e-05, "loss": 0.0687, "step": 107960 }, { "epoch": 3.6502248216640183, "grad_norm": 1.5663437843322754, "learning_rate": 4.7546409848882454e-05, "loss": 0.0685, "step": 107970 }, { "epoch": 3.650562899354272, "grad_norm": 3.512446403503418, "learning_rate": 4.753815249343279e-05, "loss": 0.0662, "step": 107980 }, { "epoch": 3.650900977044525, "grad_norm": 0.9040577411651611, "learning_rate": 4.75298952052891e-05, "loss": 0.0696, "step": 107990 }, { "epoch": 3.651239054734778, "grad_norm": 1.3045555353164673, "learning_rate": 4.752163798467715e-05, "loss": 0.0631, "step": 108000 }, { "epoch": 3.6515771324250315, "grad_norm": 0.973569929599762, "learning_rate": 4.751338083182269e-05, "loss": 0.0671, "step": 108010 }, { "epoch": 3.6519152101152845, "grad_norm": 7.344849109649658, "learning_rate": 4.750512374695145e-05, "loss": 0.0635, "step": 108020 }, { "epoch": 3.6522532878055376, "grad_norm": 0.9166580438613892, "learning_rate": 4.749686673028921e-05, "loss": 0.0662, "step": 108030 }, { "epoch": 3.652591365495791, "grad_norm": 4.838881492614746, "learning_rate": 4.7488609782061674e-05, "loss": 0.0657, "step": 108040 }, { "epoch": 3.652929443186044, "grad_norm": 1.289475679397583, "learning_rate": 4.748035290249463e-05, "loss": 0.0661, "step": 108050 }, { "epoch": 3.6532675208762972, "grad_norm": 1.0144668817520142, "learning_rate": 4.7472096091813755e-05, "loss": 0.0635, "step": 108060 }, { "epoch": 3.6536055985665508, "grad_norm": 4.028223514556885, "learning_rate": 4.746383935024484e-05, "loss": 0.0712, "step": 108070 }, { "epoch": 3.653943676256804, "grad_norm": 1.2935690879821777, "learning_rate": 4.745558267801362e-05, "loss": 0.0633, "step": 108080 }, { "epoch": 3.654281753947057, "grad_norm": 1.772064447402954, "learning_rate": 4.7447326075345797e-05, "loss": 0.0609, "step": 108090 }, { "epoch": 3.6546198316373104, "grad_norm": 1.0791913270950317, "learning_rate": 4.7439069542467126e-05, "loss": 0.0673, "step": 108100 }, { "epoch": 3.6549579093275635, "grad_norm": 1.2632246017456055, "learning_rate": 4.7430813079603325e-05, "loss": 0.0618, "step": 108110 }, { "epoch": 3.6552959870178166, "grad_norm": 1.4222548007965088, "learning_rate": 4.742255668698014e-05, "loss": 0.0596, "step": 108120 }, { "epoch": 3.65563406470807, "grad_norm": 1.5796537399291992, "learning_rate": 4.7414300364823275e-05, "loss": 0.0623, "step": 108130 }, { "epoch": 3.655972142398323, "grad_norm": 14.026453018188477, "learning_rate": 4.740604411335845e-05, "loss": 0.0592, "step": 108140 }, { "epoch": 3.656310220088576, "grad_norm": 2.4993832111358643, "learning_rate": 4.739778793281144e-05, "loss": 0.0622, "step": 108150 }, { "epoch": 3.6566482977788297, "grad_norm": 1.1227725744247437, "learning_rate": 4.738953182340791e-05, "loss": 0.0632, "step": 108160 }, { "epoch": 3.656986375469083, "grad_norm": 0.6579683423042297, "learning_rate": 4.7381275785373616e-05, "loss": 0.0625, "step": 108170 }, { "epoch": 3.657324453159336, "grad_norm": 1.1629635095596313, "learning_rate": 4.737301981893424e-05, "loss": 0.0645, "step": 108180 }, { "epoch": 3.6576625308495894, "grad_norm": 1.0331872701644897, "learning_rate": 4.736476392431553e-05, "loss": 0.065, "step": 108190 }, { "epoch": 3.6580006085398424, "grad_norm": 3.2947661876678467, "learning_rate": 4.735650810174317e-05, "loss": 0.0661, "step": 108200 }, { "epoch": 3.6583386862300955, "grad_norm": 1.018242597579956, "learning_rate": 4.73482523514429e-05, "loss": 0.0616, "step": 108210 }, { "epoch": 3.658676763920349, "grad_norm": 1.055078387260437, "learning_rate": 4.733999667364041e-05, "loss": 0.0624, "step": 108220 }, { "epoch": 3.659014841610602, "grad_norm": 1.0052649974822998, "learning_rate": 4.733174106856139e-05, "loss": 0.0574, "step": 108230 }, { "epoch": 3.659352919300855, "grad_norm": 2.983088254928589, "learning_rate": 4.7323485536431604e-05, "loss": 0.055, "step": 108240 }, { "epoch": 3.6596909969911087, "grad_norm": 0.885204553604126, "learning_rate": 4.731523007747669e-05, "loss": 0.0679, "step": 108250 }, { "epoch": 3.6600290746813617, "grad_norm": 1.1395893096923828, "learning_rate": 4.730697469192238e-05, "loss": 0.0535, "step": 108260 }, { "epoch": 3.660367152371615, "grad_norm": 1.8821929693222046, "learning_rate": 4.729871937999438e-05, "loss": 0.0611, "step": 108270 }, { "epoch": 3.6607052300618683, "grad_norm": 1.3700997829437256, "learning_rate": 4.729046414191837e-05, "loss": 0.0607, "step": 108280 }, { "epoch": 3.6610433077521214, "grad_norm": 1.2623131275177002, "learning_rate": 4.7282208977920064e-05, "loss": 0.0669, "step": 108290 }, { "epoch": 3.6613813854423745, "grad_norm": 2.6287038326263428, "learning_rate": 4.7273953888225135e-05, "loss": 0.0631, "step": 108300 }, { "epoch": 3.661719463132628, "grad_norm": 1.137067198753357, "learning_rate": 4.7265698873059286e-05, "loss": 0.0606, "step": 108310 }, { "epoch": 3.662057540822881, "grad_norm": 1.139382243156433, "learning_rate": 4.72574439326482e-05, "loss": 0.0625, "step": 108320 }, { "epoch": 3.662395618513134, "grad_norm": 1.324195146560669, "learning_rate": 4.724918906721755e-05, "loss": 0.0622, "step": 108330 }, { "epoch": 3.6627336962033876, "grad_norm": 0.8181565999984741, "learning_rate": 4.724093427699308e-05, "loss": 0.0608, "step": 108340 }, { "epoch": 3.6630717738936407, "grad_norm": 1.0824313163757324, "learning_rate": 4.723267956220038e-05, "loss": 0.0637, "step": 108350 }, { "epoch": 3.6634098515838938, "grad_norm": 3.9382247924804688, "learning_rate": 4.722442492306522e-05, "loss": 0.061, "step": 108360 }, { "epoch": 3.6637479292741473, "grad_norm": 0.7526648640632629, "learning_rate": 4.7216170359813214e-05, "loss": 0.0559, "step": 108370 }, { "epoch": 3.6640860069644003, "grad_norm": 2.308943510055542, "learning_rate": 4.720791587267009e-05, "loss": 0.0632, "step": 108380 }, { "epoch": 3.6644240846546534, "grad_norm": 1.9096596240997314, "learning_rate": 4.719966146186148e-05, "loss": 0.0596, "step": 108390 }, { "epoch": 3.664762162344907, "grad_norm": 4.284951210021973, "learning_rate": 4.719140712761307e-05, "loss": 0.0604, "step": 108400 }, { "epoch": 3.66510024003516, "grad_norm": 7.680150032043457, "learning_rate": 4.718315287015055e-05, "loss": 0.0622, "step": 108410 }, { "epoch": 3.665438317725413, "grad_norm": 1.0107710361480713, "learning_rate": 4.717489868969956e-05, "loss": 0.0645, "step": 108420 }, { "epoch": 3.6657763954156666, "grad_norm": 0.7599583268165588, "learning_rate": 4.7166644586485786e-05, "loss": 0.0614, "step": 108430 }, { "epoch": 3.6661144731059196, "grad_norm": 1.5503387451171875, "learning_rate": 4.715839056073487e-05, "loss": 0.0621, "step": 108440 }, { "epoch": 3.6664525507961727, "grad_norm": 1.233689785003662, "learning_rate": 4.715013661267249e-05, "loss": 0.0654, "step": 108450 }, { "epoch": 3.6667906284864262, "grad_norm": 21.20206069946289, "learning_rate": 4.7141882742524315e-05, "loss": 0.0667, "step": 108460 }, { "epoch": 3.6671287061766793, "grad_norm": 1.0392568111419678, "learning_rate": 4.713362895051599e-05, "loss": 0.0645, "step": 108470 }, { "epoch": 3.6674667838669324, "grad_norm": 0.9708513021469116, "learning_rate": 4.712537523687318e-05, "loss": 0.0616, "step": 108480 }, { "epoch": 3.667804861557186, "grad_norm": 2.376241683959961, "learning_rate": 4.711712160182152e-05, "loss": 0.0666, "step": 108490 }, { "epoch": 3.668142939247439, "grad_norm": 1.2140306234359741, "learning_rate": 4.710886804558669e-05, "loss": 0.0675, "step": 108500 }, { "epoch": 3.668481016937692, "grad_norm": 0.9762074947357178, "learning_rate": 4.7100614568394296e-05, "loss": 0.0616, "step": 108510 }, { "epoch": 3.6688190946279455, "grad_norm": 2.2566566467285156, "learning_rate": 4.7092361170470005e-05, "loss": 0.0571, "step": 108520 }, { "epoch": 3.6691571723181986, "grad_norm": 1.6611891984939575, "learning_rate": 4.7084107852039506e-05, "loss": 0.0584, "step": 108530 }, { "epoch": 3.6694952500084517, "grad_norm": 0.9104931354522705, "learning_rate": 4.707585461332836e-05, "loss": 0.0581, "step": 108540 }, { "epoch": 3.669833327698705, "grad_norm": 1.3449504375457764, "learning_rate": 4.706760145456229e-05, "loss": 0.0648, "step": 108550 }, { "epoch": 3.6701714053889583, "grad_norm": 3.1487607955932617, "learning_rate": 4.7059348375966866e-05, "loss": 0.0597, "step": 108560 }, { "epoch": 3.6705094830792118, "grad_norm": 2.4064574241638184, "learning_rate": 4.705109537776777e-05, "loss": 0.0634, "step": 108570 }, { "epoch": 3.670847560769465, "grad_norm": 0.6032615900039673, "learning_rate": 4.7042842460190606e-05, "loss": 0.0626, "step": 108580 }, { "epoch": 3.671185638459718, "grad_norm": 4.703935146331787, "learning_rate": 4.703458962346102e-05, "loss": 0.0575, "step": 108590 }, { "epoch": 3.6715237161499714, "grad_norm": 1.7838107347488403, "learning_rate": 4.702633686780465e-05, "loss": 0.0631, "step": 108600 }, { "epoch": 3.6718617938402245, "grad_norm": 1.6517311334609985, "learning_rate": 4.7018084193447105e-05, "loss": 0.0646, "step": 108610 }, { "epoch": 3.6721998715304776, "grad_norm": 0.9671050906181335, "learning_rate": 4.7009831600614024e-05, "loss": 0.0651, "step": 108620 }, { "epoch": 3.672537949220731, "grad_norm": 1.1486444473266602, "learning_rate": 4.700157908953102e-05, "loss": 0.0609, "step": 108630 }, { "epoch": 3.672876026910984, "grad_norm": 6.699403762817383, "learning_rate": 4.6993326660423695e-05, "loss": 0.0607, "step": 108640 }, { "epoch": 3.673214104601237, "grad_norm": 4.320432186126709, "learning_rate": 4.698507431351772e-05, "loss": 0.0589, "step": 108650 }, { "epoch": 3.6735521822914907, "grad_norm": 2.127652883529663, "learning_rate": 4.697682204903867e-05, "loss": 0.0687, "step": 108660 }, { "epoch": 3.673890259981744, "grad_norm": 0.46509063243865967, "learning_rate": 4.696856986721217e-05, "loss": 0.0628, "step": 108670 }, { "epoch": 3.6742283376719973, "grad_norm": 1.638274073600769, "learning_rate": 4.696031776826383e-05, "loss": 0.0579, "step": 108680 }, { "epoch": 3.6745664153622504, "grad_norm": 1.8648101091384888, "learning_rate": 4.6952065752419264e-05, "loss": 0.0634, "step": 108690 }, { "epoch": 3.6749044930525034, "grad_norm": 0.8051987290382385, "learning_rate": 4.694381381990407e-05, "loss": 0.0592, "step": 108700 }, { "epoch": 3.675242570742757, "grad_norm": 0.7522819638252258, "learning_rate": 4.693556197094386e-05, "loss": 0.0585, "step": 108710 }, { "epoch": 3.67558064843301, "grad_norm": 0.8444926142692566, "learning_rate": 4.692731020576424e-05, "loss": 0.0561, "step": 108720 }, { "epoch": 3.675918726123263, "grad_norm": 0.9760401248931885, "learning_rate": 4.691905852459079e-05, "loss": 0.0676, "step": 108730 }, { "epoch": 3.6762568038135166, "grad_norm": 3.081507921218872, "learning_rate": 4.691080692764915e-05, "loss": 0.06, "step": 108740 }, { "epoch": 3.6765948815037697, "grad_norm": 1.1152033805847168, "learning_rate": 4.6902555415164855e-05, "loss": 0.0635, "step": 108750 }, { "epoch": 3.6769329591940227, "grad_norm": 1.0941447019577026, "learning_rate": 4.689430398736356e-05, "loss": 0.0615, "step": 108760 }, { "epoch": 3.6772710368842763, "grad_norm": 1.7957185506820679, "learning_rate": 4.68860526444708e-05, "loss": 0.0595, "step": 108770 }, { "epoch": 3.6776091145745293, "grad_norm": 4.819155693054199, "learning_rate": 4.6877801386712204e-05, "loss": 0.0654, "step": 108780 }, { "epoch": 3.6779471922647824, "grad_norm": 0.7030797004699707, "learning_rate": 4.686955021431335e-05, "loss": 0.0633, "step": 108790 }, { "epoch": 3.678285269955036, "grad_norm": 1.3596558570861816, "learning_rate": 4.686129912749981e-05, "loss": 0.0614, "step": 108800 }, { "epoch": 3.678623347645289, "grad_norm": 1.3785547018051147, "learning_rate": 4.6853048126497186e-05, "loss": 0.0639, "step": 108810 }, { "epoch": 3.678961425335542, "grad_norm": 2.4575893878936768, "learning_rate": 4.684479721153103e-05, "loss": 0.063, "step": 108820 }, { "epoch": 3.6792995030257956, "grad_norm": 2.7704808712005615, "learning_rate": 4.683654638282692e-05, "loss": 0.0635, "step": 108830 }, { "epoch": 3.6796375807160486, "grad_norm": 2.818551540374756, "learning_rate": 4.6828295640610476e-05, "loss": 0.0601, "step": 108840 }, { "epoch": 3.6799756584063017, "grad_norm": 3.805840015411377, "learning_rate": 4.682004498510721e-05, "loss": 0.0591, "step": 108850 }, { "epoch": 3.680313736096555, "grad_norm": 0.9171261191368103, "learning_rate": 4.681179441654274e-05, "loss": 0.0608, "step": 108860 }, { "epoch": 3.6806518137868083, "grad_norm": 0.737694501876831, "learning_rate": 4.68035439351426e-05, "loss": 0.0557, "step": 108870 }, { "epoch": 3.6809898914770613, "grad_norm": 1.334388017654419, "learning_rate": 4.679529354113238e-05, "loss": 0.0637, "step": 108880 }, { "epoch": 3.681327969167315, "grad_norm": 2.713714361190796, "learning_rate": 4.678704323473761e-05, "loss": 0.0606, "step": 108890 }, { "epoch": 3.681666046857568, "grad_norm": 2.8194544315338135, "learning_rate": 4.677879301618389e-05, "loss": 0.0606, "step": 108900 }, { "epoch": 3.682004124547821, "grad_norm": 1.7248961925506592, "learning_rate": 4.677054288569675e-05, "loss": 0.0632, "step": 108910 }, { "epoch": 3.6823422022380745, "grad_norm": 1.7044109106063843, "learning_rate": 4.676229284350176e-05, "loss": 0.0631, "step": 108920 }, { "epoch": 3.6826802799283276, "grad_norm": 1.6511465311050415, "learning_rate": 4.675404288982447e-05, "loss": 0.0594, "step": 108930 }, { "epoch": 3.6830183576185807, "grad_norm": 10.860651969909668, "learning_rate": 4.6745793024890405e-05, "loss": 0.0645, "step": 108940 }, { "epoch": 3.683356435308834, "grad_norm": 2.1877572536468506, "learning_rate": 4.673754324892517e-05, "loss": 0.0597, "step": 108950 }, { "epoch": 3.6836945129990872, "grad_norm": 11.436051368713379, "learning_rate": 4.6729293562154247e-05, "loss": 0.0703, "step": 108960 }, { "epoch": 3.6840325906893403, "grad_norm": 1.5695545673370361, "learning_rate": 4.6721043964803215e-05, "loss": 0.0678, "step": 108970 }, { "epoch": 3.684370668379594, "grad_norm": 2.512033224105835, "learning_rate": 4.6712794457097625e-05, "loss": 0.0596, "step": 108980 }, { "epoch": 3.684708746069847, "grad_norm": 2.381847858428955, "learning_rate": 4.670454503926299e-05, "loss": 0.0582, "step": 108990 }, { "epoch": 3.6850468237601, "grad_norm": 3.5323538780212402, "learning_rate": 4.669629571152486e-05, "loss": 0.0587, "step": 109000 }, { "epoch": 3.6853849014503535, "grad_norm": 2.765010356903076, "learning_rate": 4.668804647410876e-05, "loss": 0.063, "step": 109010 }, { "epoch": 3.6857229791406065, "grad_norm": 2.6303470134735107, "learning_rate": 4.667979732724021e-05, "loss": 0.0574, "step": 109020 }, { "epoch": 3.6860610568308596, "grad_norm": 3.713658332824707, "learning_rate": 4.667154827114479e-05, "loss": 0.069, "step": 109030 }, { "epoch": 3.686399134521113, "grad_norm": 1.5976380109786987, "learning_rate": 4.6663299306047954e-05, "loss": 0.0655, "step": 109040 }, { "epoch": 3.686737212211366, "grad_norm": 3.205610990524292, "learning_rate": 4.665505043217529e-05, "loss": 0.0637, "step": 109050 }, { "epoch": 3.6870752899016193, "grad_norm": 4.963067054748535, "learning_rate": 4.664680164975226e-05, "loss": 0.0616, "step": 109060 }, { "epoch": 3.6874133675918728, "grad_norm": 1.2750462293624878, "learning_rate": 4.663855295900444e-05, "loss": 0.0687, "step": 109070 }, { "epoch": 3.687751445282126, "grad_norm": 1.0773837566375732, "learning_rate": 4.663030436015731e-05, "loss": 0.0578, "step": 109080 }, { "epoch": 3.688089522972379, "grad_norm": 1.422001838684082, "learning_rate": 4.6622055853436387e-05, "loss": 0.0614, "step": 109090 }, { "epoch": 3.6884276006626324, "grad_norm": 3.860261917114258, "learning_rate": 4.6613807439067196e-05, "loss": 0.0638, "step": 109100 }, { "epoch": 3.6887656783528855, "grad_norm": 1.2994886636734009, "learning_rate": 4.660555911727523e-05, "loss": 0.0607, "step": 109110 }, { "epoch": 3.6891037560431386, "grad_norm": 0.9951798915863037, "learning_rate": 4.6597310888286016e-05, "loss": 0.0602, "step": 109120 }, { "epoch": 3.689441833733392, "grad_norm": 0.8450618982315063, "learning_rate": 4.658906275232502e-05, "loss": 0.0626, "step": 109130 }, { "epoch": 3.689779911423645, "grad_norm": 1.5152946710586548, "learning_rate": 4.65808147096178e-05, "loss": 0.0627, "step": 109140 }, { "epoch": 3.690117989113898, "grad_norm": 0.9583866596221924, "learning_rate": 4.6572566760389794e-05, "loss": 0.0639, "step": 109150 }, { "epoch": 3.6904560668041517, "grad_norm": 2.916032552719116, "learning_rate": 4.6564318904866534e-05, "loss": 0.0619, "step": 109160 }, { "epoch": 3.690794144494405, "grad_norm": 1.799465298652649, "learning_rate": 4.655607114327351e-05, "loss": 0.0592, "step": 109170 }, { "epoch": 3.691132222184658, "grad_norm": 1.7058407068252563, "learning_rate": 4.654782347583619e-05, "loss": 0.0636, "step": 109180 }, { "epoch": 3.6914702998749114, "grad_norm": 5.281704425811768, "learning_rate": 4.653957590278011e-05, "loss": 0.062, "step": 109190 }, { "epoch": 3.6918083775651644, "grad_norm": 2.012798309326172, "learning_rate": 4.65313284243307e-05, "loss": 0.0621, "step": 109200 }, { "epoch": 3.6921464552554175, "grad_norm": 1.5228112936019897, "learning_rate": 4.6523081040713455e-05, "loss": 0.063, "step": 109210 }, { "epoch": 3.692484532945671, "grad_norm": 1.1852977275848389, "learning_rate": 4.651483375215391e-05, "loss": 0.0675, "step": 109220 }, { "epoch": 3.692822610635924, "grad_norm": 1.037857174873352, "learning_rate": 4.650658655887746e-05, "loss": 0.062, "step": 109230 }, { "epoch": 3.693160688326177, "grad_norm": 0.9756142497062683, "learning_rate": 4.649833946110966e-05, "loss": 0.0611, "step": 109240 }, { "epoch": 3.6934987660164307, "grad_norm": 1.1958348751068115, "learning_rate": 4.649009245907591e-05, "loss": 0.0594, "step": 109250 }, { "epoch": 3.6938368437066837, "grad_norm": 1.3526256084442139, "learning_rate": 4.648184555300174e-05, "loss": 0.0623, "step": 109260 }, { "epoch": 3.694174921396937, "grad_norm": 1.47780442237854, "learning_rate": 4.6473598743112565e-05, "loss": 0.0563, "step": 109270 }, { "epoch": 3.6945129990871903, "grad_norm": 1.0758941173553467, "learning_rate": 4.646535202963388e-05, "loss": 0.0645, "step": 109280 }, { "epoch": 3.6948510767774434, "grad_norm": 1.1585116386413574, "learning_rate": 4.645710541279116e-05, "loss": 0.0616, "step": 109290 }, { "epoch": 3.6951891544676965, "grad_norm": 3.3717098236083984, "learning_rate": 4.6448858892809833e-05, "loss": 0.0628, "step": 109300 }, { "epoch": 3.69552723215795, "grad_norm": 0.8585448861122131, "learning_rate": 4.644061246991538e-05, "loss": 0.0616, "step": 109310 }, { "epoch": 3.695865309848203, "grad_norm": 0.8640837669372559, "learning_rate": 4.643236614433324e-05, "loss": 0.0667, "step": 109320 }, { "epoch": 3.696203387538456, "grad_norm": 1.9853546619415283, "learning_rate": 4.642411991628887e-05, "loss": 0.0599, "step": 109330 }, { "epoch": 3.6965414652287096, "grad_norm": 2.0737736225128174, "learning_rate": 4.64158737860077e-05, "loss": 0.0653, "step": 109340 }, { "epoch": 3.6968795429189627, "grad_norm": 2.0112016201019287, "learning_rate": 4.640762775371521e-05, "loss": 0.0618, "step": 109350 }, { "epoch": 3.6972176206092158, "grad_norm": 0.6278846263885498, "learning_rate": 4.639938181963684e-05, "loss": 0.0599, "step": 109360 }, { "epoch": 3.6975556982994693, "grad_norm": 1.2297799587249756, "learning_rate": 4.6391135983998005e-05, "loss": 0.0644, "step": 109370 }, { "epoch": 3.6978937759897224, "grad_norm": 0.8798813223838806, "learning_rate": 4.638289024702416e-05, "loss": 0.0576, "step": 109380 }, { "epoch": 3.6982318536799754, "grad_norm": 2.154568910598755, "learning_rate": 4.637464460894073e-05, "loss": 0.0596, "step": 109390 }, { "epoch": 3.698569931370229, "grad_norm": 0.7284610867500305, "learning_rate": 4.636639906997316e-05, "loss": 0.0531, "step": 109400 }, { "epoch": 3.698908009060482, "grad_norm": 1.186790943145752, "learning_rate": 4.635815363034688e-05, "loss": 0.0619, "step": 109410 }, { "epoch": 3.699246086750735, "grad_norm": 2.287667989730835, "learning_rate": 4.634990829028729e-05, "loss": 0.0602, "step": 109420 }, { "epoch": 3.6995841644409886, "grad_norm": 1.2946091890335083, "learning_rate": 4.6341663050019866e-05, "loss": 0.0686, "step": 109430 }, { "epoch": 3.6999222421312417, "grad_norm": 1.0913196802139282, "learning_rate": 4.633341790976997e-05, "loss": 0.0653, "step": 109440 }, { "epoch": 3.7002603198214947, "grad_norm": 0.5544339418411255, "learning_rate": 4.632517286976308e-05, "loss": 0.0617, "step": 109450 }, { "epoch": 3.7005983975117482, "grad_norm": 1.478863000869751, "learning_rate": 4.631692793022456e-05, "loss": 0.0588, "step": 109460 }, { "epoch": 3.7009364752020013, "grad_norm": 1.3683419227600098, "learning_rate": 4.630868309137985e-05, "loss": 0.0679, "step": 109470 }, { "epoch": 3.7012745528922544, "grad_norm": 1.758473515510559, "learning_rate": 4.630043835345437e-05, "loss": 0.065, "step": 109480 }, { "epoch": 3.701612630582508, "grad_norm": 1.2205016613006592, "learning_rate": 4.6292193716673504e-05, "loss": 0.0669, "step": 109490 }, { "epoch": 3.701950708272761, "grad_norm": 1.2312939167022705, "learning_rate": 4.628394918126268e-05, "loss": 0.0661, "step": 109500 }, { "epoch": 3.702288785963014, "grad_norm": 1.4265213012695312, "learning_rate": 4.627570474744728e-05, "loss": 0.0546, "step": 109510 }, { "epoch": 3.7026268636532675, "grad_norm": 1.5155640840530396, "learning_rate": 4.626746041545272e-05, "loss": 0.0682, "step": 109520 }, { "epoch": 3.7029649413435206, "grad_norm": 1.2135790586471558, "learning_rate": 4.625921618550439e-05, "loss": 0.0628, "step": 109530 }, { "epoch": 3.703303019033774, "grad_norm": 1.2625386714935303, "learning_rate": 4.625097205782766e-05, "loss": 0.0557, "step": 109540 }, { "epoch": 3.703641096724027, "grad_norm": 1.1860960721969604, "learning_rate": 4.624272803264798e-05, "loss": 0.0648, "step": 109550 }, { "epoch": 3.7039791744142803, "grad_norm": 1.157263994216919, "learning_rate": 4.623448411019069e-05, "loss": 0.0625, "step": 109560 }, { "epoch": 3.7043172521045338, "grad_norm": 0.965114951133728, "learning_rate": 4.6226240290681186e-05, "loss": 0.0599, "step": 109570 }, { "epoch": 3.704655329794787, "grad_norm": 0.8889259099960327, "learning_rate": 4.621799657434486e-05, "loss": 0.0632, "step": 109580 }, { "epoch": 3.70499340748504, "grad_norm": 2.189662456512451, "learning_rate": 4.620975296140709e-05, "loss": 0.0613, "step": 109590 }, { "epoch": 3.7053314851752934, "grad_norm": 1.4022938013076782, "learning_rate": 4.620150945209324e-05, "loss": 0.0591, "step": 109600 }, { "epoch": 3.7056695628655465, "grad_norm": 1.265517234802246, "learning_rate": 4.61932660466287e-05, "loss": 0.0636, "step": 109610 }, { "epoch": 3.7060076405557996, "grad_norm": 1.6093896627426147, "learning_rate": 4.6185022745238844e-05, "loss": 0.0619, "step": 109620 }, { "epoch": 3.706345718246053, "grad_norm": 1.1016137599945068, "learning_rate": 4.617677954814901e-05, "loss": 0.0704, "step": 109630 }, { "epoch": 3.706683795936306, "grad_norm": 0.7237205505371094, "learning_rate": 4.6168536455584615e-05, "loss": 0.058, "step": 109640 }, { "epoch": 3.7070218736265597, "grad_norm": 1.0201754570007324, "learning_rate": 4.6160293467770964e-05, "loss": 0.0593, "step": 109650 }, { "epoch": 3.7073599513168127, "grad_norm": 1.9624719619750977, "learning_rate": 4.615205058493347e-05, "loss": 0.0633, "step": 109660 }, { "epoch": 3.707698029007066, "grad_norm": 3.8806440830230713, "learning_rate": 4.614380780729746e-05, "loss": 0.0576, "step": 109670 }, { "epoch": 3.7080361066973193, "grad_norm": 0.9826555252075195, "learning_rate": 4.6135565135088294e-05, "loss": 0.0612, "step": 109680 }, { "epoch": 3.7083741843875724, "grad_norm": 1.017332911491394, "learning_rate": 4.6127322568531335e-05, "loss": 0.056, "step": 109690 }, { "epoch": 3.7087122620778255, "grad_norm": 1.3787767887115479, "learning_rate": 4.6119080107851915e-05, "loss": 0.0607, "step": 109700 }, { "epoch": 3.709050339768079, "grad_norm": 1.0208280086517334, "learning_rate": 4.611083775327539e-05, "loss": 0.0593, "step": 109710 }, { "epoch": 3.709388417458332, "grad_norm": 0.6150225400924683, "learning_rate": 4.61025955050271e-05, "loss": 0.0577, "step": 109720 }, { "epoch": 3.709726495148585, "grad_norm": 0.6049972176551819, "learning_rate": 4.609435336333236e-05, "loss": 0.0665, "step": 109730 }, { "epoch": 3.7100645728388386, "grad_norm": 1.4655475616455078, "learning_rate": 4.608611132841658e-05, "loss": 0.0613, "step": 109740 }, { "epoch": 3.7104026505290917, "grad_norm": 1.0078068971633911, "learning_rate": 4.6077869400505e-05, "loss": 0.0614, "step": 109750 }, { "epoch": 3.7107407282193448, "grad_norm": 0.7650166153907776, "learning_rate": 4.606962757982302e-05, "loss": 0.0619, "step": 109760 }, { "epoch": 3.7110788059095983, "grad_norm": 0.8995791077613831, "learning_rate": 4.6061385866595944e-05, "loss": 0.0656, "step": 109770 }, { "epoch": 3.7114168835998513, "grad_norm": 1.0002256631851196, "learning_rate": 4.60531442610491e-05, "loss": 0.0651, "step": 109780 }, { "epoch": 3.7117549612901044, "grad_norm": 0.8291429877281189, "learning_rate": 4.6044902763407806e-05, "loss": 0.0598, "step": 109790 }, { "epoch": 3.712093038980358, "grad_norm": 1.7608164548873901, "learning_rate": 4.603666137389738e-05, "loss": 0.0588, "step": 109800 }, { "epoch": 3.712431116670611, "grad_norm": 0.9974222779273987, "learning_rate": 4.602842009274315e-05, "loss": 0.0585, "step": 109810 }, { "epoch": 3.712769194360864, "grad_norm": 0.6659156084060669, "learning_rate": 4.602017892017042e-05, "loss": 0.0646, "step": 109820 }, { "epoch": 3.7131072720511176, "grad_norm": 0.539695143699646, "learning_rate": 4.6011937856404506e-05, "loss": 0.0639, "step": 109830 }, { "epoch": 3.7134453497413706, "grad_norm": 0.6797466278076172, "learning_rate": 4.6003696901670696e-05, "loss": 0.0561, "step": 109840 }, { "epoch": 3.7137834274316237, "grad_norm": 0.9531145095825195, "learning_rate": 4.599545605619432e-05, "loss": 0.0642, "step": 109850 }, { "epoch": 3.714121505121877, "grad_norm": 2.234199285507202, "learning_rate": 4.598721532020068e-05, "loss": 0.0635, "step": 109860 }, { "epoch": 3.7144595828121303, "grad_norm": 0.7525494694709778, "learning_rate": 4.597897469391506e-05, "loss": 0.0574, "step": 109870 }, { "epoch": 3.7147976605023834, "grad_norm": 0.7849600911140442, "learning_rate": 4.5970734177562775e-05, "loss": 0.0597, "step": 109880 }, { "epoch": 3.715135738192637, "grad_norm": 2.4505488872528076, "learning_rate": 4.5962493771369086e-05, "loss": 0.0601, "step": 109890 }, { "epoch": 3.71547381588289, "grad_norm": 1.6100537776947021, "learning_rate": 4.5954253475559316e-05, "loss": 0.0633, "step": 109900 }, { "epoch": 3.715811893573143, "grad_norm": 0.9038115739822388, "learning_rate": 4.594601329035872e-05, "loss": 0.0639, "step": 109910 }, { "epoch": 3.7161499712633965, "grad_norm": 0.826043426990509, "learning_rate": 4.5937773215992594e-05, "loss": 0.0607, "step": 109920 }, { "epoch": 3.7164880489536496, "grad_norm": 0.6689732670783997, "learning_rate": 4.5929533252686244e-05, "loss": 0.0646, "step": 109930 }, { "epoch": 3.7168261266439027, "grad_norm": 1.0851939916610718, "learning_rate": 4.59212934006649e-05, "loss": 0.0628, "step": 109940 }, { "epoch": 3.717164204334156, "grad_norm": 1.0072067975997925, "learning_rate": 4.59130536601539e-05, "loss": 0.0577, "step": 109950 }, { "epoch": 3.7175022820244092, "grad_norm": 0.8443906903266907, "learning_rate": 4.590481403137844e-05, "loss": 0.0628, "step": 109960 }, { "epoch": 3.7178403597146623, "grad_norm": 2.2766056060791016, "learning_rate": 4.589657451456384e-05, "loss": 0.0593, "step": 109970 }, { "epoch": 3.718178437404916, "grad_norm": 1.0453051328659058, "learning_rate": 4.588833510993535e-05, "loss": 0.0575, "step": 109980 }, { "epoch": 3.718516515095169, "grad_norm": 0.6822806596755981, "learning_rate": 4.588009581771823e-05, "loss": 0.0645, "step": 109990 }, { "epoch": 3.718854592785422, "grad_norm": 0.9257911443710327, "learning_rate": 4.587185663813775e-05, "loss": 0.065, "step": 110000 }, { "epoch": 3.7191926704756755, "grad_norm": 0.9522519707679749, "learning_rate": 4.586361757141915e-05, "loss": 0.0477, "step": 110010 }, { "epoch": 3.7195307481659285, "grad_norm": 1.87443208694458, "learning_rate": 4.585537861778771e-05, "loss": 0.0593, "step": 110020 }, { "epoch": 3.7198688258561816, "grad_norm": 0.875952422618866, "learning_rate": 4.584713977746864e-05, "loss": 0.0631, "step": 110030 }, { "epoch": 3.720206903546435, "grad_norm": 0.8690701723098755, "learning_rate": 4.5838901050687206e-05, "loss": 0.058, "step": 110040 }, { "epoch": 3.720544981236688, "grad_norm": 0.8288443088531494, "learning_rate": 4.583066243766867e-05, "loss": 0.0585, "step": 110050 }, { "epoch": 3.7208830589269413, "grad_norm": 0.9042144417762756, "learning_rate": 4.5822423938638246e-05, "loss": 0.0666, "step": 110060 }, { "epoch": 3.721221136617195, "grad_norm": 0.7848612070083618, "learning_rate": 4.58141855538212e-05, "loss": 0.0636, "step": 110070 }, { "epoch": 3.721559214307448, "grad_norm": 0.6397303938865662, "learning_rate": 4.5805947283442726e-05, "loss": 0.0645, "step": 110080 }, { "epoch": 3.721897291997701, "grad_norm": 0.7517744898796082, "learning_rate": 4.57977091277281e-05, "loss": 0.063, "step": 110090 }, { "epoch": 3.7222353696879544, "grad_norm": 0.8248890042304993, "learning_rate": 4.578947108690251e-05, "loss": 0.0676, "step": 110100 }, { "epoch": 3.7225734473782075, "grad_norm": 1.0503089427947998, "learning_rate": 4.57812331611912e-05, "loss": 0.0577, "step": 110110 }, { "epoch": 3.7229115250684606, "grad_norm": 0.921323835849762, "learning_rate": 4.57729953508194e-05, "loss": 0.0657, "step": 110120 }, { "epoch": 3.723249602758714, "grad_norm": 1.600253939628601, "learning_rate": 4.57647576560123e-05, "loss": 0.0639, "step": 110130 }, { "epoch": 3.723587680448967, "grad_norm": 1.6348230838775635, "learning_rate": 4.575652007699517e-05, "loss": 0.0624, "step": 110140 }, { "epoch": 3.72392575813922, "grad_norm": 1.0790728330612183, "learning_rate": 4.574828261399315e-05, "loss": 0.0606, "step": 110150 }, { "epoch": 3.7242638358294737, "grad_norm": 1.7128280401229858, "learning_rate": 4.574004526723152e-05, "loss": 0.0605, "step": 110160 }, { "epoch": 3.724601913519727, "grad_norm": 1.54130220413208, "learning_rate": 4.573180803693542e-05, "loss": 0.0635, "step": 110170 }, { "epoch": 3.72493999120998, "grad_norm": 1.2626506090164185, "learning_rate": 4.572357092333009e-05, "loss": 0.0532, "step": 110180 }, { "epoch": 3.7252780689002334, "grad_norm": 2.0094411373138428, "learning_rate": 4.571533392664074e-05, "loss": 0.0593, "step": 110190 }, { "epoch": 3.7256161465904865, "grad_norm": 1.7689950466156006, "learning_rate": 4.570709704709254e-05, "loss": 0.059, "step": 110200 }, { "epoch": 3.7259542242807395, "grad_norm": 0.736066997051239, "learning_rate": 4.569886028491071e-05, "loss": 0.0608, "step": 110210 }, { "epoch": 3.726292301970993, "grad_norm": 1.5120964050292969, "learning_rate": 4.56906236403204e-05, "loss": 0.0633, "step": 110220 }, { "epoch": 3.726630379661246, "grad_norm": 1.920481562614441, "learning_rate": 4.568238711354681e-05, "loss": 0.0625, "step": 110230 }, { "epoch": 3.726968457351499, "grad_norm": 1.613139271736145, "learning_rate": 4.5674150704815174e-05, "loss": 0.0575, "step": 110240 }, { "epoch": 3.7273065350417527, "grad_norm": 0.8426147103309631, "learning_rate": 4.5665914414350606e-05, "loss": 0.0665, "step": 110250 }, { "epoch": 3.7276446127320058, "grad_norm": 1.5029144287109375, "learning_rate": 4.5657678242378323e-05, "loss": 0.0577, "step": 110260 }, { "epoch": 3.727982690422259, "grad_norm": 0.9354835748672485, "learning_rate": 4.564944218912348e-05, "loss": 0.0632, "step": 110270 }, { "epoch": 3.7283207681125123, "grad_norm": 4.767247200012207, "learning_rate": 4.564120625481126e-05, "loss": 0.0635, "step": 110280 }, { "epoch": 3.7286588458027654, "grad_norm": 0.9803562164306641, "learning_rate": 4.563297043966681e-05, "loss": 0.0612, "step": 110290 }, { "epoch": 3.7289969234930185, "grad_norm": 1.9589687585830688, "learning_rate": 4.5624734743915316e-05, "loss": 0.0644, "step": 110300 }, { "epoch": 3.729335001183272, "grad_norm": 1.66395902633667, "learning_rate": 4.5616499167781936e-05, "loss": 0.0596, "step": 110310 }, { "epoch": 3.729673078873525, "grad_norm": 1.0999563932418823, "learning_rate": 4.560826371149181e-05, "loss": 0.0632, "step": 110320 }, { "epoch": 3.730011156563778, "grad_norm": 1.8487735986709595, "learning_rate": 4.5600028375270114e-05, "loss": 0.061, "step": 110330 }, { "epoch": 3.7303492342540316, "grad_norm": 0.9440431594848633, "learning_rate": 4.5591793159341975e-05, "loss": 0.0629, "step": 110340 }, { "epoch": 3.7306873119442847, "grad_norm": 6.336455821990967, "learning_rate": 4.558355806393258e-05, "loss": 0.0636, "step": 110350 }, { "epoch": 3.731025389634538, "grad_norm": 1.2254103422164917, "learning_rate": 4.557532308926702e-05, "loss": 0.062, "step": 110360 }, { "epoch": 3.7313634673247913, "grad_norm": 1.4310795068740845, "learning_rate": 4.5567088235570474e-05, "loss": 0.0521, "step": 110370 }, { "epoch": 3.7317015450150444, "grad_norm": 1.2754673957824707, "learning_rate": 4.555885350306808e-05, "loss": 0.0629, "step": 110380 }, { "epoch": 3.7320396227052974, "grad_norm": 1.9369828701019287, "learning_rate": 4.555061889198495e-05, "loss": 0.0646, "step": 110390 }, { "epoch": 3.732377700395551, "grad_norm": 1.5784705877304077, "learning_rate": 4.5542384402546244e-05, "loss": 0.0638, "step": 110400 }, { "epoch": 3.732715778085804, "grad_norm": 1.2849230766296387, "learning_rate": 4.553415003497706e-05, "loss": 0.06, "step": 110410 }, { "epoch": 3.733053855776057, "grad_norm": 1.8527462482452393, "learning_rate": 4.552591578950253e-05, "loss": 0.0658, "step": 110420 }, { "epoch": 3.7333919334663106, "grad_norm": 1.291481375694275, "learning_rate": 4.55176816663478e-05, "loss": 0.0592, "step": 110430 }, { "epoch": 3.7337300111565637, "grad_norm": 1.044047236442566, "learning_rate": 4.5509447665737956e-05, "loss": 0.0552, "step": 110440 }, { "epoch": 3.7340680888468167, "grad_norm": 1.9346182346343994, "learning_rate": 4.550121378789815e-05, "loss": 0.0646, "step": 110450 }, { "epoch": 3.7344061665370702, "grad_norm": 1.479601263999939, "learning_rate": 4.5492980033053426e-05, "loss": 0.0681, "step": 110460 }, { "epoch": 3.7347442442273233, "grad_norm": 0.9312806129455566, "learning_rate": 4.5484746401428966e-05, "loss": 0.0634, "step": 110470 }, { "epoch": 3.7350823219175764, "grad_norm": 2.7164762020111084, "learning_rate": 4.547651289324983e-05, "loss": 0.0641, "step": 110480 }, { "epoch": 3.73542039960783, "grad_norm": 0.8428694605827332, "learning_rate": 4.546827950874114e-05, "loss": 0.0592, "step": 110490 }, { "epoch": 3.735758477298083, "grad_norm": 1.0447901487350464, "learning_rate": 4.546004624812799e-05, "loss": 0.0637, "step": 110500 }, { "epoch": 3.7360965549883365, "grad_norm": 0.9988545179367065, "learning_rate": 4.545181311163546e-05, "loss": 0.0597, "step": 110510 }, { "epoch": 3.7364346326785896, "grad_norm": 5.33104944229126, "learning_rate": 4.5443580099488667e-05, "loss": 0.0589, "step": 110520 }, { "epoch": 3.7367727103688426, "grad_norm": 1.3052436113357544, "learning_rate": 4.543534721191266e-05, "loss": 0.0686, "step": 110530 }, { "epoch": 3.737110788059096, "grad_norm": 1.616384506225586, "learning_rate": 4.542711444913256e-05, "loss": 0.0567, "step": 110540 }, { "epoch": 3.737448865749349, "grad_norm": 2.7104597091674805, "learning_rate": 4.541888181137342e-05, "loss": 0.0608, "step": 110550 }, { "epoch": 3.7377869434396023, "grad_norm": 1.438319444656372, "learning_rate": 4.541064929886033e-05, "loss": 0.0601, "step": 110560 }, { "epoch": 3.738125021129856, "grad_norm": 1.6367450952529907, "learning_rate": 4.540241691181838e-05, "loss": 0.0693, "step": 110570 }, { "epoch": 3.738463098820109, "grad_norm": 2.9625449180603027, "learning_rate": 4.539418465047261e-05, "loss": 0.0655, "step": 110580 }, { "epoch": 3.738801176510362, "grad_norm": 1.564184546470642, "learning_rate": 4.538595251504812e-05, "loss": 0.0589, "step": 110590 }, { "epoch": 3.7391392542006154, "grad_norm": 1.5920989513397217, "learning_rate": 4.537772050576994e-05, "loss": 0.0592, "step": 110600 }, { "epoch": 3.7394773318908685, "grad_norm": 1.1564130783081055, "learning_rate": 4.536948862286316e-05, "loss": 0.0618, "step": 110610 }, { "epoch": 3.739815409581122, "grad_norm": 2.280860662460327, "learning_rate": 4.5361256866552796e-05, "loss": 0.0664, "step": 110620 }, { "epoch": 3.740153487271375, "grad_norm": 1.2893680334091187, "learning_rate": 4.535302523706393e-05, "loss": 0.0581, "step": 110630 }, { "epoch": 3.740491564961628, "grad_norm": 1.6362026929855347, "learning_rate": 4.5344793734621634e-05, "loss": 0.0646, "step": 110640 }, { "epoch": 3.7408296426518817, "grad_norm": 1.150871753692627, "learning_rate": 4.53365623594509e-05, "loss": 0.0605, "step": 110650 }, { "epoch": 3.7411677203421347, "grad_norm": 1.3030786514282227, "learning_rate": 4.532833111177683e-05, "loss": 0.0566, "step": 110660 }, { "epoch": 3.741505798032388, "grad_norm": 3.2562742233276367, "learning_rate": 4.53200999918244e-05, "loss": 0.0561, "step": 110670 }, { "epoch": 3.7418438757226413, "grad_norm": 1.6924052238464355, "learning_rate": 4.5311868999818685e-05, "loss": 0.0609, "step": 110680 }, { "epoch": 3.7421819534128944, "grad_norm": 2.1132469177246094, "learning_rate": 4.5303638135984724e-05, "loss": 0.0616, "step": 110690 }, { "epoch": 3.7425200311031475, "grad_norm": 0.904741108417511, "learning_rate": 4.529540740054752e-05, "loss": 0.0643, "step": 110700 }, { "epoch": 3.742858108793401, "grad_norm": 3.200824022293091, "learning_rate": 4.528717679373212e-05, "loss": 0.0613, "step": 110710 }, { "epoch": 3.743196186483654, "grad_norm": 1.153307318687439, "learning_rate": 4.527894631576353e-05, "loss": 0.0569, "step": 110720 }, { "epoch": 3.743534264173907, "grad_norm": 1.8056402206420898, "learning_rate": 4.5270715966866786e-05, "loss": 0.0615, "step": 110730 }, { "epoch": 3.7438723418641606, "grad_norm": 1.3200544118881226, "learning_rate": 4.5262485747266874e-05, "loss": 0.0664, "step": 110740 }, { "epoch": 3.7442104195544137, "grad_norm": 2.508749485015869, "learning_rate": 4.5254255657188816e-05, "loss": 0.0606, "step": 110750 }, { "epoch": 3.7445484972446668, "grad_norm": 2.0335662364959717, "learning_rate": 4.524602569685764e-05, "loss": 0.0633, "step": 110760 }, { "epoch": 3.7448865749349203, "grad_norm": 5.998570919036865, "learning_rate": 4.5237795866498336e-05, "loss": 0.06, "step": 110770 }, { "epoch": 3.7452246526251733, "grad_norm": 1.0695316791534424, "learning_rate": 4.5229566166335904e-05, "loss": 0.0573, "step": 110780 }, { "epoch": 3.7455627303154264, "grad_norm": 0.9113665223121643, "learning_rate": 4.522133659659533e-05, "loss": 0.0596, "step": 110790 }, { "epoch": 3.74590080800568, "grad_norm": 1.4008349180221558, "learning_rate": 4.521310715750164e-05, "loss": 0.06, "step": 110800 }, { "epoch": 3.746238885695933, "grad_norm": 6.124046802520752, "learning_rate": 4.520487784927978e-05, "loss": 0.0645, "step": 110810 }, { "epoch": 3.746576963386186, "grad_norm": 2.006364583969116, "learning_rate": 4.519664867215476e-05, "loss": 0.0615, "step": 110820 }, { "epoch": 3.7469150410764396, "grad_norm": 2.033937692642212, "learning_rate": 4.518841962635158e-05, "loss": 0.0641, "step": 110830 }, { "epoch": 3.7472531187666926, "grad_norm": 0.9254139065742493, "learning_rate": 4.518019071209517e-05, "loss": 0.0606, "step": 110840 }, { "epoch": 3.7475911964569457, "grad_norm": 1.4087553024291992, "learning_rate": 4.517196192961057e-05, "loss": 0.0655, "step": 110850 }, { "epoch": 3.7479292741471992, "grad_norm": 1.3205938339233398, "learning_rate": 4.5163733279122685e-05, "loss": 0.0593, "step": 110860 }, { "epoch": 3.7482673518374523, "grad_norm": 1.095894694328308, "learning_rate": 4.515550476085653e-05, "loss": 0.07, "step": 110870 }, { "epoch": 3.7486054295277054, "grad_norm": 0.8217483162879944, "learning_rate": 4.514727637503706e-05, "loss": 0.062, "step": 110880 }, { "epoch": 3.748943507217959, "grad_norm": 2.202166795730591, "learning_rate": 4.5139048121889214e-05, "loss": 0.0683, "step": 110890 }, { "epoch": 3.749281584908212, "grad_norm": 2.490753412246704, "learning_rate": 4.5130820001637985e-05, "loss": 0.0614, "step": 110900 }, { "epoch": 3.749619662598465, "grad_norm": 4.358832836151123, "learning_rate": 4.51225920145083e-05, "loss": 0.0612, "step": 110910 }, { "epoch": 3.7499577402887185, "grad_norm": 1.214651346206665, "learning_rate": 4.511436416072512e-05, "loss": 0.0586, "step": 110920 }, { "epoch": 3.7502958179789716, "grad_norm": 2.368332624435425, "learning_rate": 4.5106136440513386e-05, "loss": 0.0614, "step": 110930 }, { "epoch": 3.7506338956692247, "grad_norm": 1.2862441539764404, "learning_rate": 4.5097908854098025e-05, "loss": 0.0589, "step": 110940 }, { "epoch": 3.750971973359478, "grad_norm": 2.007708787918091, "learning_rate": 4.508968140170403e-05, "loss": 0.0638, "step": 110950 }, { "epoch": 3.7513100510497313, "grad_norm": 1.352982521057129, "learning_rate": 4.508145408355627e-05, "loss": 0.0598, "step": 110960 }, { "epoch": 3.7516481287399843, "grad_norm": 1.8629335165023804, "learning_rate": 4.507322689987973e-05, "loss": 0.0612, "step": 110970 }, { "epoch": 3.751986206430238, "grad_norm": 1.2428553104400635, "learning_rate": 4.50649998508993e-05, "loss": 0.0658, "step": 110980 }, { "epoch": 3.752324284120491, "grad_norm": 1.7414990663528442, "learning_rate": 4.505677293683993e-05, "loss": 0.0589, "step": 110990 }, { "epoch": 3.752662361810744, "grad_norm": 0.8628716468811035, "learning_rate": 4.504854615792652e-05, "loss": 0.0636, "step": 111000 }, { "epoch": 3.7530004395009975, "grad_norm": 1.0823402404785156, "learning_rate": 4.504031951438401e-05, "loss": 0.06, "step": 111010 }, { "epoch": 3.7533385171912506, "grad_norm": 1.478124976158142, "learning_rate": 4.50320930064373e-05, "loss": 0.062, "step": 111020 }, { "epoch": 3.7536765948815036, "grad_norm": 2.4284443855285645, "learning_rate": 4.502386663431129e-05, "loss": 0.056, "step": 111030 }, { "epoch": 3.754014672571757, "grad_norm": 1.05971097946167, "learning_rate": 4.501564039823091e-05, "loss": 0.0615, "step": 111040 }, { "epoch": 3.75435275026201, "grad_norm": 2.0477187633514404, "learning_rate": 4.500741429842103e-05, "loss": 0.0687, "step": 111050 }, { "epoch": 3.7546908279522633, "grad_norm": 2.097534418106079, "learning_rate": 4.499918833510657e-05, "loss": 0.0602, "step": 111060 }, { "epoch": 3.755028905642517, "grad_norm": 1.267223596572876, "learning_rate": 4.499096250851244e-05, "loss": 0.0591, "step": 111070 }, { "epoch": 3.75536698333277, "grad_norm": 1.6436842679977417, "learning_rate": 4.49827368188635e-05, "loss": 0.0581, "step": 111080 }, { "epoch": 3.755705061023023, "grad_norm": 1.82804274559021, "learning_rate": 4.497451126638467e-05, "loss": 0.0551, "step": 111090 }, { "epoch": 3.7560431387132764, "grad_norm": 1.6066745519638062, "learning_rate": 4.4966285851300796e-05, "loss": 0.0591, "step": 111100 }, { "epoch": 3.7563812164035295, "grad_norm": 1.4181545972824097, "learning_rate": 4.4958060573836785e-05, "loss": 0.0584, "step": 111110 }, { "epoch": 3.7567192940937826, "grad_norm": 1.0066452026367188, "learning_rate": 4.494983543421749e-05, "loss": 0.0611, "step": 111120 }, { "epoch": 3.757057371784036, "grad_norm": 2.3462891578674316, "learning_rate": 4.494161043266779e-05, "loss": 0.0651, "step": 111130 }, { "epoch": 3.757395449474289, "grad_norm": 1.4955613613128662, "learning_rate": 4.49333855694126e-05, "loss": 0.0591, "step": 111140 }, { "epoch": 3.7577335271645422, "grad_norm": 2.858873128890991, "learning_rate": 4.4925160844676704e-05, "loss": 0.0563, "step": 111150 }, { "epoch": 3.7580716048547957, "grad_norm": 1.9245386123657227, "learning_rate": 4.4916936258685036e-05, "loss": 0.0671, "step": 111160 }, { "epoch": 3.758409682545049, "grad_norm": 0.7813875675201416, "learning_rate": 4.490871181166239e-05, "loss": 0.0612, "step": 111170 }, { "epoch": 3.758747760235302, "grad_norm": 2.8845577239990234, "learning_rate": 4.4900487503833675e-05, "loss": 0.062, "step": 111180 }, { "epoch": 3.7590858379255554, "grad_norm": 0.9256322979927063, "learning_rate": 4.48922633354237e-05, "loss": 0.059, "step": 111190 }, { "epoch": 3.7594239156158085, "grad_norm": 2.4331881999969482, "learning_rate": 4.4884039306657334e-05, "loss": 0.0591, "step": 111200 }, { "epoch": 3.7597619933060615, "grad_norm": 1.0292880535125732, "learning_rate": 4.487581541775942e-05, "loss": 0.0586, "step": 111210 }, { "epoch": 3.760100070996315, "grad_norm": 1.1826382875442505, "learning_rate": 4.486759166895477e-05, "loss": 0.0599, "step": 111220 }, { "epoch": 3.760438148686568, "grad_norm": 1.3584846258163452, "learning_rate": 4.485936806046825e-05, "loss": 0.0656, "step": 111230 }, { "epoch": 3.760776226376821, "grad_norm": 3.463423013687134, "learning_rate": 4.485114459252467e-05, "loss": 0.0601, "step": 111240 }, { "epoch": 3.7611143040670747, "grad_norm": 3.019251585006714, "learning_rate": 4.484292126534885e-05, "loss": 0.0624, "step": 111250 }, { "epoch": 3.7614523817573278, "grad_norm": 2.1828880310058594, "learning_rate": 4.4834698079165645e-05, "loss": 0.0579, "step": 111260 }, { "epoch": 3.761790459447581, "grad_norm": 1.4029204845428467, "learning_rate": 4.4826475034199836e-05, "loss": 0.0599, "step": 111270 }, { "epoch": 3.7621285371378343, "grad_norm": 1.7927674055099487, "learning_rate": 4.481825213067627e-05, "loss": 0.0612, "step": 111280 }, { "epoch": 3.7624666148280874, "grad_norm": 2.639110565185547, "learning_rate": 4.481002936881975e-05, "loss": 0.0601, "step": 111290 }, { "epoch": 3.7628046925183405, "grad_norm": 1.056854009628296, "learning_rate": 4.480180674885507e-05, "loss": 0.0627, "step": 111300 }, { "epoch": 3.763142770208594, "grad_norm": 1.5908985137939453, "learning_rate": 4.479358427100703e-05, "loss": 0.0679, "step": 111310 }, { "epoch": 3.763480847898847, "grad_norm": 2.8354673385620117, "learning_rate": 4.478536193550045e-05, "loss": 0.0649, "step": 111320 }, { "epoch": 3.7638189255891, "grad_norm": 1.491284966468811, "learning_rate": 4.477713974256012e-05, "loss": 0.0579, "step": 111330 }, { "epoch": 3.7641570032793537, "grad_norm": 3.0694828033447266, "learning_rate": 4.476891769241081e-05, "loss": 0.0623, "step": 111340 }, { "epoch": 3.7644950809696067, "grad_norm": 0.996130108833313, "learning_rate": 4.4760695785277353e-05, "loss": 0.063, "step": 111350 }, { "epoch": 3.76483315865986, "grad_norm": 1.8468667268753052, "learning_rate": 4.475247402138447e-05, "loss": 0.0575, "step": 111360 }, { "epoch": 3.7651712363501133, "grad_norm": 1.1564421653747559, "learning_rate": 4.474425240095702e-05, "loss": 0.059, "step": 111370 }, { "epoch": 3.7655093140403664, "grad_norm": 1.569088101387024, "learning_rate": 4.473603092421969e-05, "loss": 0.0599, "step": 111380 }, { "epoch": 3.7658473917306194, "grad_norm": 1.0940710306167603, "learning_rate": 4.4727809591397306e-05, "loss": 0.0623, "step": 111390 }, { "epoch": 3.766185469420873, "grad_norm": 1.8901540040969849, "learning_rate": 4.4719588402714644e-05, "loss": 0.0579, "step": 111400 }, { "epoch": 3.766523547111126, "grad_norm": 0.9958495497703552, "learning_rate": 4.471136735839644e-05, "loss": 0.0724, "step": 111410 }, { "epoch": 3.766861624801379, "grad_norm": 3.1509180068969727, "learning_rate": 4.470314645866748e-05, "loss": 0.0639, "step": 111420 }, { "epoch": 3.7671997024916326, "grad_norm": 0.8991385698318481, "learning_rate": 4.469492570375249e-05, "loss": 0.0636, "step": 111430 }, { "epoch": 3.7675377801818857, "grad_norm": 1.2803478240966797, "learning_rate": 4.4686705093876235e-05, "loss": 0.0635, "step": 111440 }, { "epoch": 3.7678758578721387, "grad_norm": 7.278514385223389, "learning_rate": 4.467848462926349e-05, "loss": 0.0641, "step": 111450 }, { "epoch": 3.7682139355623923, "grad_norm": 2.6011946201324463, "learning_rate": 4.467026431013895e-05, "loss": 0.0609, "step": 111460 }, { "epoch": 3.7685520132526453, "grad_norm": 4.368264675140381, "learning_rate": 4.46620441367274e-05, "loss": 0.0588, "step": 111470 }, { "epoch": 3.768890090942899, "grad_norm": 2.2699289321899414, "learning_rate": 4.4653824109253543e-05, "loss": 0.062, "step": 111480 }, { "epoch": 3.769228168633152, "grad_norm": 2.176730155944824, "learning_rate": 4.4645604227942146e-05, "loss": 0.061, "step": 111490 }, { "epoch": 3.769566246323405, "grad_norm": 0.8017515540122986, "learning_rate": 4.46373844930179e-05, "loss": 0.0514, "step": 111500 }, { "epoch": 3.7699043240136585, "grad_norm": 1.8370988368988037, "learning_rate": 4.462916490470555e-05, "loss": 0.0663, "step": 111510 }, { "epoch": 3.7702424017039116, "grad_norm": 1.4928451776504517, "learning_rate": 4.4620945463229825e-05, "loss": 0.067, "step": 111520 }, { "epoch": 3.7705804793941646, "grad_norm": 1.350793480873108, "learning_rate": 4.461272616881542e-05, "loss": 0.0665, "step": 111530 }, { "epoch": 3.770918557084418, "grad_norm": 2.3918886184692383, "learning_rate": 4.460450702168707e-05, "loss": 0.0615, "step": 111540 }, { "epoch": 3.771256634774671, "grad_norm": 1.62542724609375, "learning_rate": 4.459628802206944e-05, "loss": 0.0714, "step": 111550 }, { "epoch": 3.7715947124649243, "grad_norm": 1.3677936792373657, "learning_rate": 4.4588069170187306e-05, "loss": 0.0662, "step": 111560 }, { "epoch": 3.771932790155178, "grad_norm": 1.4664785861968994, "learning_rate": 4.45798504662653e-05, "loss": 0.0578, "step": 111570 }, { "epoch": 3.772270867845431, "grad_norm": 2.722813606262207, "learning_rate": 4.457163191052815e-05, "loss": 0.069, "step": 111580 }, { "epoch": 3.7726089455356844, "grad_norm": 1.6288838386535645, "learning_rate": 4.456341350320056e-05, "loss": 0.0668, "step": 111590 }, { "epoch": 3.7729470232259374, "grad_norm": 1.473860740661621, "learning_rate": 4.455519524450718e-05, "loss": 0.0586, "step": 111600 }, { "epoch": 3.7732851009161905, "grad_norm": 1.996543526649475, "learning_rate": 4.454697713467273e-05, "loss": 0.0612, "step": 111610 }, { "epoch": 3.773623178606444, "grad_norm": 1.8929765224456787, "learning_rate": 4.453875917392188e-05, "loss": 0.0607, "step": 111620 }, { "epoch": 3.773961256296697, "grad_norm": 5.040283679962158, "learning_rate": 4.4530541362479275e-05, "loss": 0.0597, "step": 111630 }, { "epoch": 3.77429933398695, "grad_norm": 0.9600184559822083, "learning_rate": 4.4522323700569657e-05, "loss": 0.0558, "step": 111640 }, { "epoch": 3.7746374116772037, "grad_norm": 3.7851569652557373, "learning_rate": 4.451410618841761e-05, "loss": 0.0564, "step": 111650 }, { "epoch": 3.7749754893674567, "grad_norm": 1.6572998762130737, "learning_rate": 4.450588882624788e-05, "loss": 0.0647, "step": 111660 }, { "epoch": 3.77531356705771, "grad_norm": 1.2296291589736938, "learning_rate": 4.449767161428505e-05, "loss": 0.0621, "step": 111670 }, { "epoch": 3.7756516447479633, "grad_norm": 5.7452168464660645, "learning_rate": 4.4489454552753836e-05, "loss": 0.0635, "step": 111680 }, { "epoch": 3.7759897224382164, "grad_norm": 1.0976481437683105, "learning_rate": 4.448123764187885e-05, "loss": 0.0605, "step": 111690 }, { "epoch": 3.7763278001284695, "grad_norm": 1.278126835823059, "learning_rate": 4.4473020881884755e-05, "loss": 0.0554, "step": 111700 }, { "epoch": 3.776665877818723, "grad_norm": 1.9973872900009155, "learning_rate": 4.4464804272996204e-05, "loss": 0.0647, "step": 111710 }, { "epoch": 3.777003955508976, "grad_norm": 1.1102333068847656, "learning_rate": 4.4456587815437814e-05, "loss": 0.0684, "step": 111720 }, { "epoch": 3.777342033199229, "grad_norm": 1.4651072025299072, "learning_rate": 4.444837150943423e-05, "loss": 0.0595, "step": 111730 }, { "epoch": 3.7776801108894826, "grad_norm": 2.4836134910583496, "learning_rate": 4.444015535521009e-05, "loss": 0.0631, "step": 111740 }, { "epoch": 3.7780181885797357, "grad_norm": 1.453346610069275, "learning_rate": 4.4431939352990014e-05, "loss": 0.0609, "step": 111750 }, { "epoch": 3.7783562662699888, "grad_norm": 3.0824999809265137, "learning_rate": 4.4423723502998606e-05, "loss": 0.0609, "step": 111760 }, { "epoch": 3.7786943439602423, "grad_norm": 1.510188341140747, "learning_rate": 4.441550780546051e-05, "loss": 0.0557, "step": 111770 }, { "epoch": 3.7790324216504954, "grad_norm": 0.8315591812133789, "learning_rate": 4.440729226060035e-05, "loss": 0.0578, "step": 111780 }, { "epoch": 3.7793704993407484, "grad_norm": 4.159554481506348, "learning_rate": 4.4399076868642697e-05, "loss": 0.0535, "step": 111790 }, { "epoch": 3.779708577031002, "grad_norm": 7.814333915710449, "learning_rate": 4.43908616298122e-05, "loss": 0.0651, "step": 111800 }, { "epoch": 3.780046654721255, "grad_norm": 1.8992786407470703, "learning_rate": 4.438264654433342e-05, "loss": 0.0613, "step": 111810 }, { "epoch": 3.780384732411508, "grad_norm": 1.0661470890045166, "learning_rate": 4.437443161243099e-05, "loss": 0.0576, "step": 111820 }, { "epoch": 3.7807228101017616, "grad_norm": 0.9314042329788208, "learning_rate": 4.436621683432947e-05, "loss": 0.0582, "step": 111830 }, { "epoch": 3.7810608877920147, "grad_norm": 2.166646957397461, "learning_rate": 4.4358002210253456e-05, "loss": 0.0645, "step": 111840 }, { "epoch": 3.7813989654822677, "grad_norm": 1.718641757965088, "learning_rate": 4.434978774042758e-05, "loss": 0.0571, "step": 111850 }, { "epoch": 3.7817370431725212, "grad_norm": 0.790007472038269, "learning_rate": 4.434157342507635e-05, "loss": 0.06, "step": 111860 }, { "epoch": 3.7820751208627743, "grad_norm": 1.4016077518463135, "learning_rate": 4.433335926442441e-05, "loss": 0.0655, "step": 111870 }, { "epoch": 3.7824131985530274, "grad_norm": 2.292377233505249, "learning_rate": 4.432514525869626e-05, "loss": 0.0568, "step": 111880 }, { "epoch": 3.782751276243281, "grad_norm": 2.937739610671997, "learning_rate": 4.431693140811652e-05, "loss": 0.0567, "step": 111890 }, { "epoch": 3.783089353933534, "grad_norm": 3.927989959716797, "learning_rate": 4.430871771290975e-05, "loss": 0.0654, "step": 111900 }, { "epoch": 3.783427431623787, "grad_norm": 5.705895900726318, "learning_rate": 4.430050417330049e-05, "loss": 0.0592, "step": 111910 }, { "epoch": 3.7837655093140405, "grad_norm": 1.3437989950180054, "learning_rate": 4.429229078951332e-05, "loss": 0.0622, "step": 111920 }, { "epoch": 3.7841035870042936, "grad_norm": 3.3756959438323975, "learning_rate": 4.4284077561772765e-05, "loss": 0.0584, "step": 111930 }, { "epoch": 3.7844416646945467, "grad_norm": 1.6031334400177002, "learning_rate": 4.4275864490303384e-05, "loss": 0.0674, "step": 111940 }, { "epoch": 3.7847797423848, "grad_norm": 2.278642177581787, "learning_rate": 4.426765157532971e-05, "loss": 0.0633, "step": 111950 }, { "epoch": 3.7851178200750533, "grad_norm": 2.121119260787964, "learning_rate": 4.425943881707628e-05, "loss": 0.0594, "step": 111960 }, { "epoch": 3.7854558977653063, "grad_norm": 3.1312572956085205, "learning_rate": 4.425122621576766e-05, "loss": 0.0576, "step": 111970 }, { "epoch": 3.78579397545556, "grad_norm": 5.833272933959961, "learning_rate": 4.424301377162834e-05, "loss": 0.0653, "step": 111980 }, { "epoch": 3.786132053145813, "grad_norm": 1.0621017217636108, "learning_rate": 4.4234801484882874e-05, "loss": 0.0689, "step": 111990 }, { "epoch": 3.786470130836066, "grad_norm": 2.211038112640381, "learning_rate": 4.422658935575575e-05, "loss": 0.0643, "step": 112000 }, { "epoch": 3.7868082085263195, "grad_norm": 6.171518325805664, "learning_rate": 4.4218377384471535e-05, "loss": 0.0644, "step": 112010 }, { "epoch": 3.7871462862165726, "grad_norm": 1.542117714881897, "learning_rate": 4.421016557125468e-05, "loss": 0.0624, "step": 112020 }, { "epoch": 3.7874843639068256, "grad_norm": 0.9927784204483032, "learning_rate": 4.420195391632973e-05, "loss": 0.0586, "step": 112030 }, { "epoch": 3.787822441597079, "grad_norm": 1.6535121202468872, "learning_rate": 4.4193742419921196e-05, "loss": 0.0657, "step": 112040 }, { "epoch": 3.788160519287332, "grad_norm": 3.379221200942993, "learning_rate": 4.418553108225354e-05, "loss": 0.0627, "step": 112050 }, { "epoch": 3.7884985969775853, "grad_norm": 5.554274559020996, "learning_rate": 4.417731990355131e-05, "loss": 0.0611, "step": 112060 }, { "epoch": 3.788836674667839, "grad_norm": 0.9689283967018127, "learning_rate": 4.4169108884038935e-05, "loss": 0.0539, "step": 112070 }, { "epoch": 3.789174752358092, "grad_norm": 2.3493947982788086, "learning_rate": 4.416089802394094e-05, "loss": 0.0615, "step": 112080 }, { "epoch": 3.789512830048345, "grad_norm": 2.868415355682373, "learning_rate": 4.4152687323481814e-05, "loss": 0.0595, "step": 112090 }, { "epoch": 3.7898509077385985, "grad_norm": 1.2886934280395508, "learning_rate": 4.414447678288601e-05, "loss": 0.0592, "step": 112100 }, { "epoch": 3.7901889854288515, "grad_norm": 1.3055455684661865, "learning_rate": 4.413626640237802e-05, "loss": 0.0618, "step": 112110 }, { "epoch": 3.7905270631191046, "grad_norm": 1.90300452709198, "learning_rate": 4.412805618218229e-05, "loss": 0.0613, "step": 112120 }, { "epoch": 3.790865140809358, "grad_norm": 4.78623628616333, "learning_rate": 4.411984612252331e-05, "loss": 0.0627, "step": 112130 }, { "epoch": 3.791203218499611, "grad_norm": 2.2448360919952393, "learning_rate": 4.411163622362552e-05, "loss": 0.067, "step": 112140 }, { "epoch": 3.7915412961898642, "grad_norm": 1.8328152894973755, "learning_rate": 4.4103426485713364e-05, "loss": 0.0513, "step": 112150 }, { "epoch": 3.7918793738801178, "grad_norm": 2.3683178424835205, "learning_rate": 4.409521690901136e-05, "loss": 0.0603, "step": 112160 }, { "epoch": 3.792217451570371, "grad_norm": 1.288934588432312, "learning_rate": 4.4087007493743866e-05, "loss": 0.0575, "step": 112170 }, { "epoch": 3.792555529260624, "grad_norm": 2.4638612270355225, "learning_rate": 4.407879824013539e-05, "loss": 0.0666, "step": 112180 }, { "epoch": 3.7928936069508774, "grad_norm": 2.877200126647949, "learning_rate": 4.407058914841033e-05, "loss": 0.0635, "step": 112190 }, { "epoch": 3.7932316846411305, "grad_norm": 11.14331340789795, "learning_rate": 4.406238021879315e-05, "loss": 0.0664, "step": 112200 }, { "epoch": 3.7935697623313835, "grad_norm": 2.20676851272583, "learning_rate": 4.405417145150825e-05, "loss": 0.0549, "step": 112210 }, { "epoch": 3.793907840021637, "grad_norm": 3.0150063037872314, "learning_rate": 4.404596284678008e-05, "loss": 0.0574, "step": 112220 }, { "epoch": 3.79424591771189, "grad_norm": 2.2478790283203125, "learning_rate": 4.403775440483304e-05, "loss": 0.0666, "step": 112230 }, { "epoch": 3.794583995402143, "grad_norm": 14.198627471923828, "learning_rate": 4.4029546125891554e-05, "loss": 0.0625, "step": 112240 }, { "epoch": 3.7949220730923967, "grad_norm": 2.0884552001953125, "learning_rate": 4.402133801018005e-05, "loss": 0.0649, "step": 112250 }, { "epoch": 3.7952601507826498, "grad_norm": 2.933229684829712, "learning_rate": 4.4013130057922894e-05, "loss": 0.0607, "step": 112260 }, { "epoch": 3.795598228472903, "grad_norm": 6.633394718170166, "learning_rate": 4.400492226934452e-05, "loss": 0.0654, "step": 112270 }, { "epoch": 3.7959363061631564, "grad_norm": 70.00518798828125, "learning_rate": 4.399671464466933e-05, "loss": 0.0634, "step": 112280 }, { "epoch": 3.7962743838534094, "grad_norm": 3.80316162109375, "learning_rate": 4.39885071841217e-05, "loss": 0.0585, "step": 112290 }, { "epoch": 3.7966124615436625, "grad_norm": 3.4351534843444824, "learning_rate": 4.398029988792604e-05, "loss": 0.0616, "step": 112300 }, { "epoch": 3.796950539233916, "grad_norm": 1.6542279720306396, "learning_rate": 4.3972092756306696e-05, "loss": 0.0605, "step": 112310 }, { "epoch": 3.797288616924169, "grad_norm": 1.7480169534683228, "learning_rate": 4.396388578948809e-05, "loss": 0.0619, "step": 112320 }, { "epoch": 3.797626694614422, "grad_norm": 3.0160183906555176, "learning_rate": 4.395567898769457e-05, "loss": 0.0612, "step": 112330 }, { "epoch": 3.7979647723046757, "grad_norm": 6.408498287200928, "learning_rate": 4.39474723511505e-05, "loss": 0.0579, "step": 112340 }, { "epoch": 3.7983028499949287, "grad_norm": 2.719085454940796, "learning_rate": 4.39392658800803e-05, "loss": 0.0672, "step": 112350 }, { "epoch": 3.798640927685182, "grad_norm": 3.925281286239624, "learning_rate": 4.3931059574708255e-05, "loss": 0.0666, "step": 112360 }, { "epoch": 3.7989790053754353, "grad_norm": 2.470367908477783, "learning_rate": 4.39228534352588e-05, "loss": 0.063, "step": 112370 }, { "epoch": 3.7993170830656884, "grad_norm": 1.2045258283615112, "learning_rate": 4.391464746195621e-05, "loss": 0.0607, "step": 112380 }, { "epoch": 3.7996551607559415, "grad_norm": 9.113961219787598, "learning_rate": 4.39064416550249e-05, "loss": 0.0641, "step": 112390 }, { "epoch": 3.799993238446195, "grad_norm": 1.7750723361968994, "learning_rate": 4.389823601468917e-05, "loss": 0.068, "step": 112400 }, { "epoch": 3.800331316136448, "grad_norm": 2.9719455242156982, "learning_rate": 4.389003054117338e-05, "loss": 0.0595, "step": 112410 }, { "epoch": 3.800669393826701, "grad_norm": 2.6506218910217285, "learning_rate": 4.3881825234701865e-05, "loss": 0.0637, "step": 112420 }, { "epoch": 3.8010074715169546, "grad_norm": 2.8076250553131104, "learning_rate": 4.387362009549894e-05, "loss": 0.0656, "step": 112430 }, { "epoch": 3.8013455492072077, "grad_norm": 1.5629985332489014, "learning_rate": 4.386541512378895e-05, "loss": 0.0635, "step": 112440 }, { "epoch": 3.801683626897461, "grad_norm": 1.362733244895935, "learning_rate": 4.3857210319796196e-05, "loss": 0.0624, "step": 112450 }, { "epoch": 3.8020217045877143, "grad_norm": 2.309009313583374, "learning_rate": 4.3849005683745e-05, "loss": 0.0651, "step": 112460 }, { "epoch": 3.8023597822779673, "grad_norm": 1.6020152568817139, "learning_rate": 4.384080121585969e-05, "loss": 0.0626, "step": 112470 }, { "epoch": 3.802697859968221, "grad_norm": 3.8629274368286133, "learning_rate": 4.383259691636455e-05, "loss": 0.0604, "step": 112480 }, { "epoch": 3.803035937658474, "grad_norm": 1.6422148942947388, "learning_rate": 4.382439278548391e-05, "loss": 0.0661, "step": 112490 }, { "epoch": 3.803374015348727, "grad_norm": 3.810468912124634, "learning_rate": 4.381618882344204e-05, "loss": 0.0644, "step": 112500 }, { "epoch": 3.8037120930389805, "grad_norm": 2.2707390785217285, "learning_rate": 4.380798503046325e-05, "loss": 0.0678, "step": 112510 }, { "epoch": 3.8040501707292336, "grad_norm": 5.407049655914307, "learning_rate": 4.379978140677182e-05, "loss": 0.0645, "step": 112520 }, { "epoch": 3.8043882484194866, "grad_norm": 2.383376121520996, "learning_rate": 4.3791577952592036e-05, "loss": 0.0623, "step": 112530 }, { "epoch": 3.80472632610974, "grad_norm": 1.5837138891220093, "learning_rate": 4.378337466814818e-05, "loss": 0.0624, "step": 112540 }, { "epoch": 3.8050644037999932, "grad_norm": 1.4192936420440674, "learning_rate": 4.377517155366451e-05, "loss": 0.0574, "step": 112550 }, { "epoch": 3.8054024814902467, "grad_norm": 1.9347302913665771, "learning_rate": 4.3766968609365343e-05, "loss": 0.0646, "step": 112560 }, { "epoch": 3.8057405591805, "grad_norm": 4.9371724128723145, "learning_rate": 4.375876583547488e-05, "loss": 0.0594, "step": 112570 }, { "epoch": 3.806078636870753, "grad_norm": 2.6087071895599365, "learning_rate": 4.375056323221745e-05, "loss": 0.0624, "step": 112580 }, { "epoch": 3.8064167145610064, "grad_norm": 3.107998847961426, "learning_rate": 4.374236079981724e-05, "loss": 0.0623, "step": 112590 }, { "epoch": 3.8067547922512595, "grad_norm": 7.0710625648498535, "learning_rate": 4.373415853849854e-05, "loss": 0.0707, "step": 112600 }, { "epoch": 3.8070928699415125, "grad_norm": 1.5865943431854248, "learning_rate": 4.37259564484856e-05, "loss": 0.0664, "step": 112610 }, { "epoch": 3.807430947631766, "grad_norm": 1.3914055824279785, "learning_rate": 4.3717754530002655e-05, "loss": 0.0569, "step": 112620 }, { "epoch": 3.807769025322019, "grad_norm": 2.46968674659729, "learning_rate": 4.3709552783273946e-05, "loss": 0.0585, "step": 112630 }, { "epoch": 3.808107103012272, "grad_norm": 3.7677395343780518, "learning_rate": 4.370135120852369e-05, "loss": 0.0643, "step": 112640 }, { "epoch": 3.8084451807025257, "grad_norm": 2.3222155570983887, "learning_rate": 4.369314980597611e-05, "loss": 0.0598, "step": 112650 }, { "epoch": 3.8087832583927788, "grad_norm": 1.858684778213501, "learning_rate": 4.368494857585549e-05, "loss": 0.063, "step": 112660 }, { "epoch": 3.809121336083032, "grad_norm": 2.117152452468872, "learning_rate": 4.3676747518385956e-05, "loss": 0.0663, "step": 112670 }, { "epoch": 3.8094594137732853, "grad_norm": 1.4307433366775513, "learning_rate": 4.36685466337918e-05, "loss": 0.0624, "step": 112680 }, { "epoch": 3.8097974914635384, "grad_norm": 1.5523357391357422, "learning_rate": 4.366034592229719e-05, "loss": 0.068, "step": 112690 }, { "epoch": 3.8101355691537915, "grad_norm": 1.7481759786605835, "learning_rate": 4.365214538412635e-05, "loss": 0.0641, "step": 112700 }, { "epoch": 3.810473646844045, "grad_norm": 2.1905553340911865, "learning_rate": 4.3643945019503456e-05, "loss": 0.0644, "step": 112710 }, { "epoch": 3.810811724534298, "grad_norm": 1.605789065361023, "learning_rate": 4.363574482865272e-05, "loss": 0.0713, "step": 112720 }, { "epoch": 3.811149802224551, "grad_norm": 1.5298798084259033, "learning_rate": 4.362754481179834e-05, "loss": 0.0633, "step": 112730 }, { "epoch": 3.8114878799148046, "grad_norm": 1.7655049562454224, "learning_rate": 4.361934496916448e-05, "loss": 0.0714, "step": 112740 }, { "epoch": 3.8118259576050577, "grad_norm": 39.0708122253418, "learning_rate": 4.361114530097534e-05, "loss": 0.0689, "step": 112750 }, { "epoch": 3.812164035295311, "grad_norm": 0.8909146189689636, "learning_rate": 4.360294580745507e-05, "loss": 0.0665, "step": 112760 }, { "epoch": 3.8125021129855643, "grad_norm": 40.27785110473633, "learning_rate": 4.3594746488827895e-05, "loss": 0.0705, "step": 112770 }, { "epoch": 3.8128401906758174, "grad_norm": 1.3456772565841675, "learning_rate": 4.35865473453179e-05, "loss": 0.0622, "step": 112780 }, { "epoch": 3.8131782683660704, "grad_norm": 1.3966892957687378, "learning_rate": 4.357834837714931e-05, "loss": 0.0642, "step": 112790 }, { "epoch": 3.813516346056324, "grad_norm": 2.291923761367798, "learning_rate": 4.357014958454627e-05, "loss": 0.0621, "step": 112800 }, { "epoch": 3.813854423746577, "grad_norm": 1.8673629760742188, "learning_rate": 4.356195096773292e-05, "loss": 0.0694, "step": 112810 }, { "epoch": 3.81419250143683, "grad_norm": 2.504164218902588, "learning_rate": 4.3553752526933415e-05, "loss": 0.0599, "step": 112820 }, { "epoch": 3.8145305791270836, "grad_norm": 2.5577027797698975, "learning_rate": 4.354555426237189e-05, "loss": 0.0661, "step": 112830 }, { "epoch": 3.8148686568173367, "grad_norm": 2.2803008556365967, "learning_rate": 4.353735617427247e-05, "loss": 0.0593, "step": 112840 }, { "epoch": 3.8152067345075897, "grad_norm": 2.538043260574341, "learning_rate": 4.3529158262859336e-05, "loss": 0.0704, "step": 112850 }, { "epoch": 3.8155448121978432, "grad_norm": 6.59485387802124, "learning_rate": 4.3520960528356557e-05, "loss": 0.0684, "step": 112860 }, { "epoch": 3.8158828898880963, "grad_norm": 2.678082227706909, "learning_rate": 4.3512762970988315e-05, "loss": 0.0587, "step": 112870 }, { "epoch": 3.8162209675783494, "grad_norm": 2.754399061203003, "learning_rate": 4.350456559097866e-05, "loss": 0.0736, "step": 112880 }, { "epoch": 3.816559045268603, "grad_norm": 1.812615990638733, "learning_rate": 4.3496368388551765e-05, "loss": 0.0602, "step": 112890 }, { "epoch": 3.816897122958856, "grad_norm": 2.131197214126587, "learning_rate": 4.3488171363931706e-05, "loss": 0.0593, "step": 112900 }, { "epoch": 3.817235200649109, "grad_norm": 1.9182922840118408, "learning_rate": 4.347997451734259e-05, "loss": 0.0601, "step": 112910 }, { "epoch": 3.8175732783393626, "grad_norm": 3.669231414794922, "learning_rate": 4.347177784900854e-05, "loss": 0.06, "step": 112920 }, { "epoch": 3.8179113560296156, "grad_norm": 2.1095380783081055, "learning_rate": 4.3463581359153614e-05, "loss": 0.0647, "step": 112930 }, { "epoch": 3.8182494337198687, "grad_norm": 1.3750447034835815, "learning_rate": 4.3455385048001936e-05, "loss": 0.0619, "step": 112940 }, { "epoch": 3.818587511410122, "grad_norm": 1.3436379432678223, "learning_rate": 4.344718891577756e-05, "loss": 0.0615, "step": 112950 }, { "epoch": 3.8189255891003753, "grad_norm": 3.2504327297210693, "learning_rate": 4.3438992962704587e-05, "loss": 0.0711, "step": 112960 }, { "epoch": 3.8192636667906283, "grad_norm": 15.579222679138184, "learning_rate": 4.343079718900705e-05, "loss": 0.0603, "step": 112970 }, { "epoch": 3.819601744480882, "grad_norm": 6.840324401855469, "learning_rate": 4.342260159490907e-05, "loss": 0.0723, "step": 112980 }, { "epoch": 3.819939822171135, "grad_norm": 1.3878823518753052, "learning_rate": 4.34144061806347e-05, "loss": 0.0664, "step": 112990 }, { "epoch": 3.820277899861388, "grad_norm": 6.2705979347229, "learning_rate": 4.3406210946407976e-05, "loss": 0.0679, "step": 113000 }, { "epoch": 3.8206159775516415, "grad_norm": 2.7546210289001465, "learning_rate": 4.339801589245299e-05, "loss": 0.0579, "step": 113010 }, { "epoch": 3.8209540552418946, "grad_norm": 2.049268960952759, "learning_rate": 4.3389821018993745e-05, "loss": 0.0611, "step": 113020 }, { "epoch": 3.8212921329321476, "grad_norm": 2.9408273696899414, "learning_rate": 4.338162632625433e-05, "loss": 0.0704, "step": 113030 }, { "epoch": 3.821630210622401, "grad_norm": 2.1583385467529297, "learning_rate": 4.3373431814458756e-05, "loss": 0.06, "step": 113040 }, { "epoch": 3.8219682883126542, "grad_norm": 1.3840410709381104, "learning_rate": 4.3365237483831044e-05, "loss": 0.0565, "step": 113050 }, { "epoch": 3.8223063660029073, "grad_norm": 1.2359386682510376, "learning_rate": 4.3357043334595296e-05, "loss": 0.0635, "step": 113060 }, { "epoch": 3.822644443693161, "grad_norm": 1.4098104238510132, "learning_rate": 4.334884936697544e-05, "loss": 0.0676, "step": 113070 }, { "epoch": 3.822982521383414, "grad_norm": 5.471522808074951, "learning_rate": 4.334065558119558e-05, "loss": 0.0661, "step": 113080 }, { "epoch": 3.823320599073667, "grad_norm": 1.72611403465271, "learning_rate": 4.3332461977479666e-05, "loss": 0.0628, "step": 113090 }, { "epoch": 3.8236586767639205, "grad_norm": 2.772602081298828, "learning_rate": 4.3324268556051746e-05, "loss": 0.0559, "step": 113100 }, { "epoch": 3.8239967544541735, "grad_norm": 2.762115478515625, "learning_rate": 4.3316075317135825e-05, "loss": 0.0633, "step": 113110 }, { "epoch": 3.8243348321444266, "grad_norm": 1.1156322956085205, "learning_rate": 4.330788226095588e-05, "loss": 0.0686, "step": 113120 }, { "epoch": 3.82467290983468, "grad_norm": 3.6491456031799316, "learning_rate": 4.3299689387735936e-05, "loss": 0.0676, "step": 113130 }, { "epoch": 3.825010987524933, "grad_norm": 3.532158374786377, "learning_rate": 4.3291496697699954e-05, "loss": 0.0707, "step": 113140 }, { "epoch": 3.8253490652151863, "grad_norm": 1.8077985048294067, "learning_rate": 4.328330419107194e-05, "loss": 0.0633, "step": 113150 }, { "epoch": 3.8256871429054398, "grad_norm": 1.441625714302063, "learning_rate": 4.327511186807586e-05, "loss": 0.0693, "step": 113160 }, { "epoch": 3.826025220595693, "grad_norm": 2.2299623489379883, "learning_rate": 4.3266919728935686e-05, "loss": 0.0642, "step": 113170 }, { "epoch": 3.826363298285946, "grad_norm": 2.4651389122009277, "learning_rate": 4.325872777387541e-05, "loss": 0.0565, "step": 113180 }, { "epoch": 3.8267013759761994, "grad_norm": 3.4163427352905273, "learning_rate": 4.325053600311898e-05, "loss": 0.0672, "step": 113190 }, { "epoch": 3.8270394536664525, "grad_norm": 1.4035981893539429, "learning_rate": 4.324234441689038e-05, "loss": 0.0653, "step": 113200 }, { "epoch": 3.8273775313567056, "grad_norm": 1.370800256729126, "learning_rate": 4.323415301541351e-05, "loss": 0.0551, "step": 113210 }, { "epoch": 3.827715609046959, "grad_norm": 1.4721200466156006, "learning_rate": 4.3225961798912385e-05, "loss": 0.0656, "step": 113220 }, { "epoch": 3.828053686737212, "grad_norm": 27.870798110961914, "learning_rate": 4.3217770767610906e-05, "loss": 0.0658, "step": 113230 }, { "epoch": 3.828391764427465, "grad_norm": 1.3949551582336426, "learning_rate": 4.320957992173302e-05, "loss": 0.0657, "step": 113240 }, { "epoch": 3.8287298421177187, "grad_norm": 2.8326151371002197, "learning_rate": 4.3201389261502685e-05, "loss": 0.0711, "step": 113250 }, { "epoch": 3.829067919807972, "grad_norm": 1.0232810974121094, "learning_rate": 4.319319878714379e-05, "loss": 0.0651, "step": 113260 }, { "epoch": 3.829405997498225, "grad_norm": 1.5256539583206177, "learning_rate": 4.318500849888031e-05, "loss": 0.0599, "step": 113270 }, { "epoch": 3.8297440751884784, "grad_norm": 1.4856802225112915, "learning_rate": 4.317681839693611e-05, "loss": 0.0729, "step": 113280 }, { "epoch": 3.8300821528787314, "grad_norm": 5.889832496643066, "learning_rate": 4.316862848153514e-05, "loss": 0.0647, "step": 113290 }, { "epoch": 3.8304202305689845, "grad_norm": 1.3878462314605713, "learning_rate": 4.3160438752901313e-05, "loss": 0.0619, "step": 113300 }, { "epoch": 3.830758308259238, "grad_norm": 1.383495569229126, "learning_rate": 4.315224921125851e-05, "loss": 0.0605, "step": 113310 }, { "epoch": 3.831096385949491, "grad_norm": 1.478801965713501, "learning_rate": 4.314405985683065e-05, "loss": 0.061, "step": 113320 }, { "epoch": 3.831434463639744, "grad_norm": 1.2768858671188354, "learning_rate": 4.313587068984161e-05, "loss": 0.06, "step": 113330 }, { "epoch": 3.8317725413299977, "grad_norm": 2.0565171241760254, "learning_rate": 4.31276817105153e-05, "loss": 0.0626, "step": 113340 }, { "epoch": 3.8321106190202507, "grad_norm": 6.724937438964844, "learning_rate": 4.311949291907558e-05, "loss": 0.0699, "step": 113350 }, { "epoch": 3.832448696710504, "grad_norm": 2.0528717041015625, "learning_rate": 4.311130431574632e-05, "loss": 0.0669, "step": 113360 }, { "epoch": 3.8327867744007573, "grad_norm": 9.534130096435547, "learning_rate": 4.310311590075145e-05, "loss": 0.0576, "step": 113370 }, { "epoch": 3.8331248520910104, "grad_norm": 1.7775241136550903, "learning_rate": 4.3094927674314764e-05, "loss": 0.0584, "step": 113380 }, { "epoch": 3.8334629297812635, "grad_norm": 3.1552133560180664, "learning_rate": 4.308673963666019e-05, "loss": 0.0647, "step": 113390 }, { "epoch": 3.833801007471517, "grad_norm": 5.029253005981445, "learning_rate": 4.307855178801154e-05, "loss": 0.0636, "step": 113400 }, { "epoch": 3.83413908516177, "grad_norm": 2.284749984741211, "learning_rate": 4.3070364128592695e-05, "loss": 0.0651, "step": 113410 }, { "epoch": 3.8344771628520236, "grad_norm": 1.233704924583435, "learning_rate": 4.306217665862749e-05, "loss": 0.0641, "step": 113420 }, { "epoch": 3.8348152405422766, "grad_norm": 1.9082523584365845, "learning_rate": 4.3053989378339756e-05, "loss": 0.0707, "step": 113430 }, { "epoch": 3.8351533182325297, "grad_norm": 3.4705235958099365, "learning_rate": 4.304580228795335e-05, "loss": 0.0667, "step": 113440 }, { "epoch": 3.835491395922783, "grad_norm": 3.2756385803222656, "learning_rate": 4.30376153876921e-05, "loss": 0.0647, "step": 113450 }, { "epoch": 3.8358294736130363, "grad_norm": 1.8864182233810425, "learning_rate": 4.302942867777983e-05, "loss": 0.0599, "step": 113460 }, { "epoch": 3.8361675513032893, "grad_norm": 2.8580539226531982, "learning_rate": 4.302124215844035e-05, "loss": 0.0635, "step": 113470 }, { "epoch": 3.836505628993543, "grad_norm": 2.697094678878784, "learning_rate": 4.301305582989749e-05, "loss": 0.0653, "step": 113480 }, { "epoch": 3.836843706683796, "grad_norm": 4.241600036621094, "learning_rate": 4.300486969237507e-05, "loss": 0.0647, "step": 113490 }, { "epoch": 3.837181784374049, "grad_norm": 5.740230083465576, "learning_rate": 4.299668374609688e-05, "loss": 0.0674, "step": 113500 }, { "epoch": 3.8375198620643025, "grad_norm": 2.2570226192474365, "learning_rate": 4.298849799128673e-05, "loss": 0.065, "step": 113510 }, { "epoch": 3.8378579397545556, "grad_norm": 2.0264480113983154, "learning_rate": 4.2980312428168404e-05, "loss": 0.0699, "step": 113520 }, { "epoch": 3.838196017444809, "grad_norm": 2.4515223503112793, "learning_rate": 4.2972127056965705e-05, "loss": 0.0647, "step": 113530 }, { "epoch": 3.838534095135062, "grad_norm": 1.7826595306396484, "learning_rate": 4.296394187790241e-05, "loss": 0.056, "step": 113540 }, { "epoch": 3.8388721728253152, "grad_norm": 4.233926296234131, "learning_rate": 4.295575689120228e-05, "loss": 0.0682, "step": 113550 }, { "epoch": 3.8392102505155687, "grad_norm": 1.0544135570526123, "learning_rate": 4.294757209708915e-05, "loss": 0.0617, "step": 113560 }, { "epoch": 3.839548328205822, "grad_norm": 1.7609177827835083, "learning_rate": 4.293938749578671e-05, "loss": 0.0596, "step": 113570 }, { "epoch": 3.839886405896075, "grad_norm": 1.6400138139724731, "learning_rate": 4.29312030875188e-05, "loss": 0.0678, "step": 113580 }, { "epoch": 3.8402244835863284, "grad_norm": 3.3688433170318604, "learning_rate": 4.2923018872509115e-05, "loss": 0.0607, "step": 113590 }, { "epoch": 3.8405625612765815, "grad_norm": 2.2958381175994873, "learning_rate": 4.2914834850981446e-05, "loss": 0.0668, "step": 113600 }, { "epoch": 3.8409006389668345, "grad_norm": 1.5684781074523926, "learning_rate": 4.290665102315953e-05, "loss": 0.0627, "step": 113610 }, { "epoch": 3.841238716657088, "grad_norm": 2.4357974529266357, "learning_rate": 4.28984673892671e-05, "loss": 0.0603, "step": 113620 }, { "epoch": 3.841576794347341, "grad_norm": 1.842341423034668, "learning_rate": 4.2890283949527924e-05, "loss": 0.0695, "step": 113630 }, { "epoch": 3.841914872037594, "grad_norm": 1.004607915878296, "learning_rate": 4.2882100704165696e-05, "loss": 0.0637, "step": 113640 }, { "epoch": 3.8422529497278477, "grad_norm": 1.3189451694488525, "learning_rate": 4.287391765340417e-05, "loss": 0.0639, "step": 113650 }, { "epoch": 3.8425910274181008, "grad_norm": 2.6235299110412598, "learning_rate": 4.286573479746706e-05, "loss": 0.0636, "step": 113660 }, { "epoch": 3.842929105108354, "grad_norm": 6.476500034332275, "learning_rate": 4.2857552136578055e-05, "loss": 0.0641, "step": 113670 }, { "epoch": 3.8432671827986074, "grad_norm": 4.308453559875488, "learning_rate": 4.284936967096091e-05, "loss": 0.0638, "step": 113680 }, { "epoch": 3.8436052604888604, "grad_norm": 1.1259263753890991, "learning_rate": 4.284118740083932e-05, "loss": 0.0591, "step": 113690 }, { "epoch": 3.8439433381791135, "grad_norm": 6.761944770812988, "learning_rate": 4.283300532643698e-05, "loss": 0.0655, "step": 113700 }, { "epoch": 3.844281415869367, "grad_norm": 1.5009492635726929, "learning_rate": 4.2824823447977565e-05, "loss": 0.0705, "step": 113710 }, { "epoch": 3.84461949355962, "grad_norm": 0.9492446184158325, "learning_rate": 4.2816641765684796e-05, "loss": 0.0619, "step": 113720 }, { "epoch": 3.844957571249873, "grad_norm": 1.7546552419662476, "learning_rate": 4.2808460279782334e-05, "loss": 0.0631, "step": 113730 }, { "epoch": 3.8452956489401267, "grad_norm": 3.4096827507019043, "learning_rate": 4.280027899049386e-05, "loss": 0.0691, "step": 113740 }, { "epoch": 3.8456337266303797, "grad_norm": 3.4249892234802246, "learning_rate": 4.279209789804307e-05, "loss": 0.0637, "step": 113750 }, { "epoch": 3.845971804320633, "grad_norm": 1.4237632751464844, "learning_rate": 4.278391700265359e-05, "loss": 0.0599, "step": 113760 }, { "epoch": 3.8463098820108863, "grad_norm": 1.4224073886871338, "learning_rate": 4.277573630454914e-05, "loss": 0.0653, "step": 113770 }, { "epoch": 3.8466479597011394, "grad_norm": 4.196549415588379, "learning_rate": 4.2767555803953316e-05, "loss": 0.0723, "step": 113780 }, { "epoch": 3.8469860373913924, "grad_norm": 5.701002597808838, "learning_rate": 4.275937550108984e-05, "loss": 0.068, "step": 113790 }, { "epoch": 3.847324115081646, "grad_norm": 0.730865478515625, "learning_rate": 4.275119539618227e-05, "loss": 0.0639, "step": 113800 }, { "epoch": 3.847662192771899, "grad_norm": 1.5601381063461304, "learning_rate": 4.274301548945431e-05, "loss": 0.0566, "step": 113810 }, { "epoch": 3.848000270462152, "grad_norm": 1.359798550605774, "learning_rate": 4.2734835781129595e-05, "loss": 0.0556, "step": 113820 }, { "epoch": 3.8483383481524056, "grad_norm": 2.2675793170928955, "learning_rate": 4.272665627143173e-05, "loss": 0.0624, "step": 113830 }, { "epoch": 3.8486764258426587, "grad_norm": 2.1980254650115967, "learning_rate": 4.271847696058436e-05, "loss": 0.0593, "step": 113840 }, { "epoch": 3.8490145035329117, "grad_norm": 1.0359960794448853, "learning_rate": 4.2710297848811075e-05, "loss": 0.0661, "step": 113850 }, { "epoch": 3.8493525812231653, "grad_norm": 5.610793113708496, "learning_rate": 4.270211893633551e-05, "loss": 0.0637, "step": 113860 }, { "epoch": 3.8496906589134183, "grad_norm": 1.3814496994018555, "learning_rate": 4.2693940223381295e-05, "loss": 0.066, "step": 113870 }, { "epoch": 3.8500287366036714, "grad_norm": 1.3047258853912354, "learning_rate": 4.2685761710171976e-05, "loss": 0.0588, "step": 113880 }, { "epoch": 3.850366814293925, "grad_norm": 1.3905518054962158, "learning_rate": 4.267758339693121e-05, "loss": 0.064, "step": 113890 }, { "epoch": 3.850704891984178, "grad_norm": 0.8351117372512817, "learning_rate": 4.266940528388256e-05, "loss": 0.0631, "step": 113900 }, { "epoch": 3.851042969674431, "grad_norm": 0.6907574534416199, "learning_rate": 4.266122737124962e-05, "loss": 0.0613, "step": 113910 }, { "epoch": 3.8513810473646846, "grad_norm": 2.911220073699951, "learning_rate": 4.265304965925595e-05, "loss": 0.067, "step": 113920 }, { "epoch": 3.8517191250549376, "grad_norm": 1.3979616165161133, "learning_rate": 4.264487214812515e-05, "loss": 0.0594, "step": 113930 }, { "epoch": 3.8520572027451907, "grad_norm": 1.017911434173584, "learning_rate": 4.2636694838080786e-05, "loss": 0.0644, "step": 113940 }, { "epoch": 3.852395280435444, "grad_norm": 1.0614577531814575, "learning_rate": 4.262851772934641e-05, "loss": 0.0613, "step": 113950 }, { "epoch": 3.8527333581256973, "grad_norm": 0.968782901763916, "learning_rate": 4.262034082214561e-05, "loss": 0.0604, "step": 113960 }, { "epoch": 3.8530714358159504, "grad_norm": 1.2222564220428467, "learning_rate": 4.261216411670189e-05, "loss": 0.0583, "step": 113970 }, { "epoch": 3.853409513506204, "grad_norm": 1.3703011274337769, "learning_rate": 4.260398761323886e-05, "loss": 0.0576, "step": 113980 }, { "epoch": 3.853747591196457, "grad_norm": 1.0113892555236816, "learning_rate": 4.2595811311979995e-05, "loss": 0.063, "step": 113990 }, { "epoch": 3.85408566888671, "grad_norm": 1.1505131721496582, "learning_rate": 4.258763521314888e-05, "loss": 0.0556, "step": 114000 }, { "epoch": 3.8544237465769635, "grad_norm": 3.6403114795684814, "learning_rate": 4.257945931696904e-05, "loss": 0.0547, "step": 114010 }, { "epoch": 3.8547618242672166, "grad_norm": 1.6539944410324097, "learning_rate": 4.257128362366399e-05, "loss": 0.0536, "step": 114020 }, { "epoch": 3.8550999019574697, "grad_norm": 2.4176177978515625, "learning_rate": 4.256310813345727e-05, "loss": 0.0629, "step": 114030 }, { "epoch": 3.855437979647723, "grad_norm": 0.7709589004516602, "learning_rate": 4.255493284657237e-05, "loss": 0.0657, "step": 114040 }, { "epoch": 3.8557760573379762, "grad_norm": 2.3830878734588623, "learning_rate": 4.2546757763232786e-05, "loss": 0.0627, "step": 114050 }, { "epoch": 3.8561141350282293, "grad_norm": 0.7279946208000183, "learning_rate": 4.2538582883662084e-05, "loss": 0.0645, "step": 114060 }, { "epoch": 3.856452212718483, "grad_norm": 0.8197188377380371, "learning_rate": 4.253040820808369e-05, "loss": 0.0598, "step": 114070 }, { "epoch": 3.856790290408736, "grad_norm": 1.4230458736419678, "learning_rate": 4.2522233736721164e-05, "loss": 0.0672, "step": 114080 }, { "epoch": 3.857128368098989, "grad_norm": 1.1099777221679688, "learning_rate": 4.2514059469797915e-05, "loss": 0.0647, "step": 114090 }, { "epoch": 3.8574664457892425, "grad_norm": 3.3998610973358154, "learning_rate": 4.2505885407537503e-05, "loss": 0.059, "step": 114100 }, { "epoch": 3.8578045234794955, "grad_norm": 1.2263962030410767, "learning_rate": 4.249771155016335e-05, "loss": 0.0589, "step": 114110 }, { "epoch": 3.8581426011697486, "grad_norm": 1.1896106004714966, "learning_rate": 4.2489537897898944e-05, "loss": 0.0638, "step": 114120 }, { "epoch": 3.858480678860002, "grad_norm": 0.9214761853218079, "learning_rate": 4.248136445096777e-05, "loss": 0.064, "step": 114130 }, { "epoch": 3.858818756550255, "grad_norm": 0.9662880301475525, "learning_rate": 4.247319120959325e-05, "loss": 0.0654, "step": 114140 }, { "epoch": 3.8591568342405083, "grad_norm": 2.3129701614379883, "learning_rate": 4.246501817399887e-05, "loss": 0.0657, "step": 114150 }, { "epoch": 3.8594949119307618, "grad_norm": 1.9116008281707764, "learning_rate": 4.245684534440804e-05, "loss": 0.061, "step": 114160 }, { "epoch": 3.859832989621015, "grad_norm": 2.176544666290283, "learning_rate": 4.2448672721044246e-05, "loss": 0.0612, "step": 114170 }, { "epoch": 3.860171067311268, "grad_norm": 2.308159828186035, "learning_rate": 4.244050030413088e-05, "loss": 0.0586, "step": 114180 }, { "epoch": 3.8605091450015214, "grad_norm": 1.9798177480697632, "learning_rate": 4.243232809389141e-05, "loss": 0.0584, "step": 114190 }, { "epoch": 3.8608472226917745, "grad_norm": 1.029910683631897, "learning_rate": 4.2424156090549244e-05, "loss": 0.0608, "step": 114200 }, { "epoch": 3.8611853003820276, "grad_norm": 1.3928565979003906, "learning_rate": 4.2415984294327806e-05, "loss": 0.0607, "step": 114210 }, { "epoch": 3.861523378072281, "grad_norm": 1.377877950668335, "learning_rate": 4.240781270545051e-05, "loss": 0.0535, "step": 114220 }, { "epoch": 3.861861455762534, "grad_norm": 0.6621464490890503, "learning_rate": 4.239964132414076e-05, "loss": 0.059, "step": 114230 }, { "epoch": 3.862199533452787, "grad_norm": 4.1730637550354, "learning_rate": 4.239147015062196e-05, "loss": 0.0601, "step": 114240 }, { "epoch": 3.8625376111430407, "grad_norm": 1.3526166677474976, "learning_rate": 4.2383299185117514e-05, "loss": 0.0603, "step": 114250 }, { "epoch": 3.862875688833294, "grad_norm": 2.220365285873413, "learning_rate": 4.2375128427850785e-05, "loss": 0.0572, "step": 114260 }, { "epoch": 3.863213766523547, "grad_norm": 1.7723004817962646, "learning_rate": 4.2366957879045214e-05, "loss": 0.0593, "step": 114270 }, { "epoch": 3.8635518442138004, "grad_norm": 1.7176929712295532, "learning_rate": 4.235878753892412e-05, "loss": 0.0606, "step": 114280 }, { "epoch": 3.8638899219040534, "grad_norm": 2.203126907348633, "learning_rate": 4.235061740771093e-05, "loss": 0.0596, "step": 114290 }, { "epoch": 3.8642279995943065, "grad_norm": 2.1834375858306885, "learning_rate": 4.234244748562896e-05, "loss": 0.0638, "step": 114300 }, { "epoch": 3.86456607728456, "grad_norm": 1.7662808895111084, "learning_rate": 4.2334277772901595e-05, "loss": 0.0628, "step": 114310 }, { "epoch": 3.864904154974813, "grad_norm": 4.017113208770752, "learning_rate": 4.232610826975222e-05, "loss": 0.0658, "step": 114320 }, { "epoch": 3.865242232665066, "grad_norm": 3.1790032386779785, "learning_rate": 4.2317938976404146e-05, "loss": 0.0617, "step": 114330 }, { "epoch": 3.8655803103553197, "grad_norm": 1.2636922597885132, "learning_rate": 4.2309769893080756e-05, "loss": 0.0627, "step": 114340 }, { "epoch": 3.8659183880455728, "grad_norm": 1.5412338972091675, "learning_rate": 4.2301601020005344e-05, "loss": 0.0576, "step": 114350 }, { "epoch": 3.866256465735826, "grad_norm": 11.617080688476562, "learning_rate": 4.229343235740129e-05, "loss": 0.059, "step": 114360 }, { "epoch": 3.8665945434260793, "grad_norm": 1.8949726819992065, "learning_rate": 4.228526390549188e-05, "loss": 0.0624, "step": 114370 }, { "epoch": 3.8669326211163324, "grad_norm": 1.886828899383545, "learning_rate": 4.2277095664500446e-05, "loss": 0.0555, "step": 114380 }, { "epoch": 3.867270698806586, "grad_norm": 1.3036390542984009, "learning_rate": 4.2268927634650344e-05, "loss": 0.0584, "step": 114390 }, { "epoch": 3.867608776496839, "grad_norm": 1.6442941427230835, "learning_rate": 4.226075981616484e-05, "loss": 0.0619, "step": 114400 }, { "epoch": 3.867946854187092, "grad_norm": 2.906980037689209, "learning_rate": 4.225259220926726e-05, "loss": 0.0631, "step": 114410 }, { "epoch": 3.8682849318773456, "grad_norm": 2.0090489387512207, "learning_rate": 4.2244424814180894e-05, "loss": 0.0535, "step": 114420 }, { "epoch": 3.8686230095675986, "grad_norm": 4.16766357421875, "learning_rate": 4.223625763112905e-05, "loss": 0.0594, "step": 114430 }, { "epoch": 3.8689610872578517, "grad_norm": 0.9858079552650452, "learning_rate": 4.2228090660334996e-05, "loss": 0.0658, "step": 114440 }, { "epoch": 3.869299164948105, "grad_norm": 1.724923014640808, "learning_rate": 4.221992390202202e-05, "loss": 0.0606, "step": 114450 }, { "epoch": 3.8696372426383583, "grad_norm": 2.026947021484375, "learning_rate": 4.221175735641342e-05, "loss": 0.0566, "step": 114460 }, { "epoch": 3.8699753203286114, "grad_norm": 1.6143720149993896, "learning_rate": 4.2203591023732426e-05, "loss": 0.0637, "step": 114470 }, { "epoch": 3.870313398018865, "grad_norm": 3.5261833667755127, "learning_rate": 4.219542490420235e-05, "loss": 0.0608, "step": 114480 }, { "epoch": 3.870651475709118, "grad_norm": 2.070971965789795, "learning_rate": 4.21872589980464e-05, "loss": 0.0585, "step": 114490 }, { "epoch": 3.8709895533993715, "grad_norm": 2.488333225250244, "learning_rate": 4.217909330548785e-05, "loss": 0.0656, "step": 114500 }, { "epoch": 3.8713276310896245, "grad_norm": 2.792027473449707, "learning_rate": 4.217092782674998e-05, "loss": 0.0621, "step": 114510 }, { "epoch": 3.8716657087798776, "grad_norm": 2.835695505142212, "learning_rate": 4.216276256205598e-05, "loss": 0.0597, "step": 114520 }, { "epoch": 3.872003786470131, "grad_norm": 1.2677000761032104, "learning_rate": 4.2154597511629126e-05, "loss": 0.0562, "step": 114530 }, { "epoch": 3.872341864160384, "grad_norm": 3.0951788425445557, "learning_rate": 4.214643267569261e-05, "loss": 0.0584, "step": 114540 }, { "epoch": 3.8726799418506372, "grad_norm": 4.595004558563232, "learning_rate": 4.213826805446969e-05, "loss": 0.0598, "step": 114550 }, { "epoch": 3.8730180195408908, "grad_norm": 3.7076306343078613, "learning_rate": 4.213010364818355e-05, "loss": 0.0639, "step": 114560 }, { "epoch": 3.873356097231144, "grad_norm": 1.603309988975525, "learning_rate": 4.2121939457057414e-05, "loss": 0.059, "step": 114570 }, { "epoch": 3.873694174921397, "grad_norm": 1.5803289413452148, "learning_rate": 4.211377548131453e-05, "loss": 0.0597, "step": 114580 }, { "epoch": 3.8740322526116504, "grad_norm": 6.328398704528809, "learning_rate": 4.210561172117802e-05, "loss": 0.0658, "step": 114590 }, { "epoch": 3.8743703303019035, "grad_norm": 2.0788369178771973, "learning_rate": 4.209744817687115e-05, "loss": 0.061, "step": 114600 }, { "epoch": 3.8747084079921565, "grad_norm": 2.0489513874053955, "learning_rate": 4.208928484861706e-05, "loss": 0.0599, "step": 114610 }, { "epoch": 3.87504648568241, "grad_norm": 5.63489294052124, "learning_rate": 4.208112173663897e-05, "loss": 0.063, "step": 114620 }, { "epoch": 3.875384563372663, "grad_norm": 2.045525312423706, "learning_rate": 4.207295884116002e-05, "loss": 0.0564, "step": 114630 }, { "epoch": 3.875722641062916, "grad_norm": 6.083023548126221, "learning_rate": 4.2064796162403395e-05, "loss": 0.0629, "step": 114640 }, { "epoch": 3.8760607187531697, "grad_norm": 1.2273480892181396, "learning_rate": 4.205663370059228e-05, "loss": 0.0652, "step": 114650 }, { "epoch": 3.876398796443423, "grad_norm": 2.7035915851593018, "learning_rate": 4.20484714559498e-05, "loss": 0.0676, "step": 114660 }, { "epoch": 3.876736874133676, "grad_norm": 1.981445074081421, "learning_rate": 4.204030942869913e-05, "loss": 0.0571, "step": 114670 }, { "epoch": 3.8770749518239294, "grad_norm": 3.4458391666412354, "learning_rate": 4.2032147619063397e-05, "loss": 0.066, "step": 114680 }, { "epoch": 3.8774130295141824, "grad_norm": 4.058063507080078, "learning_rate": 4.202398602726576e-05, "loss": 0.0573, "step": 114690 }, { "epoch": 3.8777511072044355, "grad_norm": 1.3294528722763062, "learning_rate": 4.201582465352936e-05, "loss": 0.0641, "step": 114700 }, { "epoch": 3.878089184894689, "grad_norm": 1.909532070159912, "learning_rate": 4.200766349807731e-05, "loss": 0.0656, "step": 114710 }, { "epoch": 3.878427262584942, "grad_norm": 3.4927473068237305, "learning_rate": 4.199950256113274e-05, "loss": 0.0626, "step": 114720 }, { "epoch": 3.878765340275195, "grad_norm": 3.049191474914551, "learning_rate": 4.199134184291875e-05, "loss": 0.0664, "step": 114730 }, { "epoch": 3.8791034179654487, "grad_norm": 4.006819725036621, "learning_rate": 4.198318134365848e-05, "loss": 0.0645, "step": 114740 }, { "epoch": 3.8794414956557017, "grad_norm": 0.7903273105621338, "learning_rate": 4.1975021063575015e-05, "loss": 0.0582, "step": 114750 }, { "epoch": 3.879779573345955, "grad_norm": 0.8264535665512085, "learning_rate": 4.1966861002891445e-05, "loss": 0.0596, "step": 114760 }, { "epoch": 3.8801176510362083, "grad_norm": 1.7893229722976685, "learning_rate": 4.195870116183091e-05, "loss": 0.0622, "step": 114770 }, { "epoch": 3.8804557287264614, "grad_norm": 2.4993765354156494, "learning_rate": 4.195054154061643e-05, "loss": 0.0588, "step": 114780 }, { "epoch": 3.8807938064167145, "grad_norm": 1.910301685333252, "learning_rate": 4.1942382139471155e-05, "loss": 0.0642, "step": 114790 }, { "epoch": 3.881131884106968, "grad_norm": 1.3371975421905518, "learning_rate": 4.193422295861809e-05, "loss": 0.0602, "step": 114800 }, { "epoch": 3.881469961797221, "grad_norm": 2.7512528896331787, "learning_rate": 4.192606399828036e-05, "loss": 0.0618, "step": 114810 }, { "epoch": 3.881808039487474, "grad_norm": 1.2014145851135254, "learning_rate": 4.1917905258681003e-05, "loss": 0.0623, "step": 114820 }, { "epoch": 3.8821461171777276, "grad_norm": 14.776516914367676, "learning_rate": 4.190974674004308e-05, "loss": 0.064, "step": 114830 }, { "epoch": 3.8824841948679807, "grad_norm": 1.9169021844863892, "learning_rate": 4.190158844258964e-05, "loss": 0.0629, "step": 114840 }, { "epoch": 3.8828222725582338, "grad_norm": 11.467548370361328, "learning_rate": 4.1893430366543726e-05, "loss": 0.0585, "step": 114850 }, { "epoch": 3.8831603502484873, "grad_norm": 1.411113977432251, "learning_rate": 4.18852725121284e-05, "loss": 0.0547, "step": 114860 }, { "epoch": 3.8834984279387403, "grad_norm": 1.2723720073699951, "learning_rate": 4.1877114879566654e-05, "loss": 0.0651, "step": 114870 }, { "epoch": 3.8838365056289934, "grad_norm": 4.266836643218994, "learning_rate": 4.1868957469081524e-05, "loss": 0.0569, "step": 114880 }, { "epoch": 3.884174583319247, "grad_norm": 0.8025494813919067, "learning_rate": 4.1860800280896065e-05, "loss": 0.0594, "step": 114890 }, { "epoch": 3.8845126610095, "grad_norm": 10.28865909576416, "learning_rate": 4.185264331523326e-05, "loss": 0.066, "step": 114900 }, { "epoch": 3.884850738699753, "grad_norm": 1.7119415998458862, "learning_rate": 4.184448657231614e-05, "loss": 0.0587, "step": 114910 }, { "epoch": 3.8851888163900066, "grad_norm": 16.30255699157715, "learning_rate": 4.1836330052367676e-05, "loss": 0.0556, "step": 114920 }, { "epoch": 3.8855268940802596, "grad_norm": 2.255727529525757, "learning_rate": 4.18281737556109e-05, "loss": 0.0685, "step": 114930 }, { "epoch": 3.8858649717705127, "grad_norm": 3.087082624435425, "learning_rate": 4.1820017682268765e-05, "loss": 0.065, "step": 114940 }, { "epoch": 3.8862030494607662, "grad_norm": 7.919627666473389, "learning_rate": 4.181186183256426e-05, "loss": 0.0661, "step": 114950 }, { "epoch": 3.8865411271510193, "grad_norm": 5.448674201965332, "learning_rate": 4.1803706206720414e-05, "loss": 0.0648, "step": 114960 }, { "epoch": 3.8868792048412724, "grad_norm": 4.871277332305908, "learning_rate": 4.179555080496013e-05, "loss": 0.0663, "step": 114970 }, { "epoch": 3.887217282531526, "grad_norm": 2.1684250831604004, "learning_rate": 4.178739562750644e-05, "loss": 0.0628, "step": 114980 }, { "epoch": 3.887555360221779, "grad_norm": 3.6644845008850098, "learning_rate": 4.1779240674582234e-05, "loss": 0.0695, "step": 114990 }, { "epoch": 3.887893437912032, "grad_norm": 4.2549943923950195, "learning_rate": 4.177108594641053e-05, "loss": 0.0665, "step": 115000 }, { "epoch": 3.8882315156022855, "grad_norm": 1.5661232471466064, "learning_rate": 4.176293144321421e-05, "loss": 0.0664, "step": 115010 }, { "epoch": 3.8885695932925386, "grad_norm": 2.499251127243042, "learning_rate": 4.175477716521626e-05, "loss": 0.0643, "step": 115020 }, { "epoch": 3.8889076709827917, "grad_norm": 5.009878158569336, "learning_rate": 4.174662311263962e-05, "loss": 0.0649, "step": 115030 }, { "epoch": 3.889245748673045, "grad_norm": 6.104309558868408, "learning_rate": 4.17384692857072e-05, "loss": 0.0596, "step": 115040 }, { "epoch": 3.8895838263632982, "grad_norm": 1.7779076099395752, "learning_rate": 4.1730315684641915e-05, "loss": 0.0658, "step": 115050 }, { "epoch": 3.8899219040535513, "grad_norm": 8.133495330810547, "learning_rate": 4.1722162309666695e-05, "loss": 0.0627, "step": 115060 }, { "epoch": 3.890259981743805, "grad_norm": 4.612181663513184, "learning_rate": 4.1714009161004435e-05, "loss": 0.0613, "step": 115070 }, { "epoch": 3.890598059434058, "grad_norm": 6.461801052093506, "learning_rate": 4.170585623887809e-05, "loss": 0.0685, "step": 115080 }, { "epoch": 3.890936137124311, "grad_norm": 2.170959949493408, "learning_rate": 4.169770354351048e-05, "loss": 0.0637, "step": 115090 }, { "epoch": 3.8912742148145645, "grad_norm": 3.5153493881225586, "learning_rate": 4.168955107512456e-05, "loss": 0.0673, "step": 115100 }, { "epoch": 3.8916122925048175, "grad_norm": 3.1906752586364746, "learning_rate": 4.168139883394318e-05, "loss": 0.0681, "step": 115110 }, { "epoch": 3.8919503701950706, "grad_norm": 1.7128174304962158, "learning_rate": 4.1673246820189245e-05, "loss": 0.0649, "step": 115120 }, { "epoch": 3.892288447885324, "grad_norm": 2.898172616958618, "learning_rate": 4.16650950340856e-05, "loss": 0.0582, "step": 115130 }, { "epoch": 3.892626525575577, "grad_norm": 6.973072052001953, "learning_rate": 4.165694347585512e-05, "loss": 0.0644, "step": 115140 }, { "epoch": 3.8929646032658303, "grad_norm": 3.66805362701416, "learning_rate": 4.164879214572069e-05, "loss": 0.0684, "step": 115150 }, { "epoch": 3.893302680956084, "grad_norm": 2.5748062133789062, "learning_rate": 4.164064104390512e-05, "loss": 0.0631, "step": 115160 }, { "epoch": 3.893640758646337, "grad_norm": 3.419846773147583, "learning_rate": 4.163249017063132e-05, "loss": 0.0638, "step": 115170 }, { "epoch": 3.89397883633659, "grad_norm": 1.176763653755188, "learning_rate": 4.162433952612206e-05, "loss": 0.0578, "step": 115180 }, { "epoch": 3.8943169140268434, "grad_norm": 1.2039732933044434, "learning_rate": 4.161618911060024e-05, "loss": 0.0617, "step": 115190 }, { "epoch": 3.8946549917170965, "grad_norm": 1.0584288835525513, "learning_rate": 4.160803892428863e-05, "loss": 0.0565, "step": 115200 }, { "epoch": 3.8949930694073496, "grad_norm": 1.4611974954605103, "learning_rate": 4.1599888967410093e-05, "loss": 0.0587, "step": 115210 }, { "epoch": 3.895331147097603, "grad_norm": 2.7444629669189453, "learning_rate": 4.159173924018745e-05, "loss": 0.0659, "step": 115220 }, { "epoch": 3.895669224787856, "grad_norm": 10.137028694152832, "learning_rate": 4.1583589742843486e-05, "loss": 0.0645, "step": 115230 }, { "epoch": 3.8960073024781092, "grad_norm": 1.1317086219787598, "learning_rate": 4.157544047560102e-05, "loss": 0.0624, "step": 115240 }, { "epoch": 3.8963453801683627, "grad_norm": 5.772334098815918, "learning_rate": 4.156729143868285e-05, "loss": 0.0585, "step": 115250 }, { "epoch": 3.896683457858616, "grad_norm": 1.4860379695892334, "learning_rate": 4.155914263231174e-05, "loss": 0.0637, "step": 115260 }, { "epoch": 3.897021535548869, "grad_norm": 2.5301833152770996, "learning_rate": 4.1550994056710536e-05, "loss": 0.0682, "step": 115270 }, { "epoch": 3.8973596132391224, "grad_norm": 1.4046660661697388, "learning_rate": 4.1542845712101955e-05, "loss": 0.0608, "step": 115280 }, { "epoch": 3.8976976909293755, "grad_norm": 1.8820481300354004, "learning_rate": 4.153469759870882e-05, "loss": 0.0587, "step": 115290 }, { "epoch": 3.8980357686196285, "grad_norm": 2.354969024658203, "learning_rate": 4.1526549716753836e-05, "loss": 0.0633, "step": 115300 }, { "epoch": 3.898373846309882, "grad_norm": 1.8183716535568237, "learning_rate": 4.151840206645983e-05, "loss": 0.0637, "step": 115310 }, { "epoch": 3.898711924000135, "grad_norm": 1.757181167602539, "learning_rate": 4.151025464804951e-05, "loss": 0.0581, "step": 115320 }, { "epoch": 3.899050001690388, "grad_norm": 2.2174534797668457, "learning_rate": 4.1502107461745635e-05, "loss": 0.0647, "step": 115330 }, { "epoch": 3.8993880793806417, "grad_norm": 3.938812255859375, "learning_rate": 4.1493960507770966e-05, "loss": 0.0611, "step": 115340 }, { "epoch": 3.8997261570708948, "grad_norm": 3.2561519145965576, "learning_rate": 4.1485813786348204e-05, "loss": 0.0633, "step": 115350 }, { "epoch": 3.9000642347611483, "grad_norm": 2.659827709197998, "learning_rate": 4.147766729770011e-05, "loss": 0.063, "step": 115360 }, { "epoch": 3.9004023124514013, "grad_norm": 1.3166333436965942, "learning_rate": 4.146952104204937e-05, "loss": 0.0592, "step": 115370 }, { "epoch": 3.9007403901416544, "grad_norm": 1.6097065210342407, "learning_rate": 4.1461375019618754e-05, "loss": 0.0611, "step": 115380 }, { "epoch": 3.901078467831908, "grad_norm": 1.665793538093567, "learning_rate": 4.1453229230630894e-05, "loss": 0.0673, "step": 115390 }, { "epoch": 3.901416545522161, "grad_norm": 2.0174295902252197, "learning_rate": 4.144508367530856e-05, "loss": 0.0673, "step": 115400 }, { "epoch": 3.901754623212414, "grad_norm": 1.2644643783569336, "learning_rate": 4.143693835387443e-05, "loss": 0.0585, "step": 115410 }, { "epoch": 3.9020927009026676, "grad_norm": 2.3333580493927, "learning_rate": 4.142879326655117e-05, "loss": 0.0613, "step": 115420 }, { "epoch": 3.9024307785929206, "grad_norm": 2.1766905784606934, "learning_rate": 4.142064841356151e-05, "loss": 0.0641, "step": 115430 }, { "epoch": 3.9027688562831737, "grad_norm": 2.6531364917755127, "learning_rate": 4.1412503795128075e-05, "loss": 0.0654, "step": 115440 }, { "epoch": 3.9031069339734272, "grad_norm": 1.8352960348129272, "learning_rate": 4.1404359411473556e-05, "loss": 0.0609, "step": 115450 }, { "epoch": 3.9034450116636803, "grad_norm": 1.9227625131607056, "learning_rate": 4.1396215262820645e-05, "loss": 0.0657, "step": 115460 }, { "epoch": 3.903783089353934, "grad_norm": 2.001573324203491, "learning_rate": 4.138807134939195e-05, "loss": 0.072, "step": 115470 }, { "epoch": 3.904121167044187, "grad_norm": 1.0395103693008423, "learning_rate": 4.137992767141018e-05, "loss": 0.0554, "step": 115480 }, { "epoch": 3.90445924473444, "grad_norm": 1.4359431266784668, "learning_rate": 4.137178422909792e-05, "loss": 0.058, "step": 115490 }, { "epoch": 3.9047973224246935, "grad_norm": 2.6354029178619385, "learning_rate": 4.1363641022677854e-05, "loss": 0.0566, "step": 115500 }, { "epoch": 3.9051354001149465, "grad_norm": 1.6950273513793945, "learning_rate": 4.135549805237259e-05, "loss": 0.0667, "step": 115510 }, { "epoch": 3.9054734778051996, "grad_norm": 1.3929479122161865, "learning_rate": 4.134735531840476e-05, "loss": 0.0579, "step": 115520 }, { "epoch": 3.905811555495453, "grad_norm": 1.5963650941848755, "learning_rate": 4.1339212820996996e-05, "loss": 0.0634, "step": 115530 }, { "epoch": 3.906149633185706, "grad_norm": 1.3290661573410034, "learning_rate": 4.13310705603719e-05, "loss": 0.0615, "step": 115540 }, { "epoch": 3.9064877108759593, "grad_norm": 1.4522159099578857, "learning_rate": 4.1322928536752073e-05, "loss": 0.0611, "step": 115550 }, { "epoch": 3.9068257885662128, "grad_norm": 2.137261152267456, "learning_rate": 4.1314786750360116e-05, "loss": 0.0598, "step": 115560 }, { "epoch": 3.907163866256466, "grad_norm": 31.834474563598633, "learning_rate": 4.130664520141864e-05, "loss": 0.0676, "step": 115570 }, { "epoch": 3.907501943946719, "grad_norm": 2.8890912532806396, "learning_rate": 4.129850389015019e-05, "loss": 0.0607, "step": 115580 }, { "epoch": 3.9078400216369724, "grad_norm": 1.4556187391281128, "learning_rate": 4.129036281677739e-05, "loss": 0.0622, "step": 115590 }, { "epoch": 3.9081780993272255, "grad_norm": 0.8816001415252686, "learning_rate": 4.1282221981522806e-05, "loss": 0.0683, "step": 115600 }, { "epoch": 3.9085161770174786, "grad_norm": 2.967991828918457, "learning_rate": 4.1274081384608984e-05, "loss": 0.0688, "step": 115610 }, { "epoch": 3.908854254707732, "grad_norm": 1.920699119567871, "learning_rate": 4.126594102625852e-05, "loss": 0.0633, "step": 115620 }, { "epoch": 3.909192332397985, "grad_norm": 3.673199415206909, "learning_rate": 4.125780090669393e-05, "loss": 0.0555, "step": 115630 }, { "epoch": 3.909530410088238, "grad_norm": 1.991718053817749, "learning_rate": 4.124966102613779e-05, "loss": 0.0614, "step": 115640 }, { "epoch": 3.9098684877784917, "grad_norm": 4.8353352546691895, "learning_rate": 4.124152138481262e-05, "loss": 0.0619, "step": 115650 }, { "epoch": 3.910206565468745, "grad_norm": 1.2450692653656006, "learning_rate": 4.1233381982940954e-05, "loss": 0.0619, "step": 115660 }, { "epoch": 3.910544643158998, "grad_norm": 1.8404244184494019, "learning_rate": 4.122524282074537e-05, "loss": 0.0644, "step": 115670 }, { "epoch": 3.9108827208492514, "grad_norm": 5.626251697540283, "learning_rate": 4.1217103898448305e-05, "loss": 0.0636, "step": 115680 }, { "epoch": 3.9112207985395044, "grad_norm": 1.922195315361023, "learning_rate": 4.120896521627237e-05, "loss": 0.0682, "step": 115690 }, { "epoch": 3.9115588762297575, "grad_norm": 1.3775619268417358, "learning_rate": 4.120082677443997e-05, "loss": 0.0629, "step": 115700 }, { "epoch": 3.911896953920011, "grad_norm": 1.5680031776428223, "learning_rate": 4.119268857317368e-05, "loss": 0.0609, "step": 115710 }, { "epoch": 3.912235031610264, "grad_norm": 1.248795747756958, "learning_rate": 4.118455061269599e-05, "loss": 0.0658, "step": 115720 }, { "epoch": 3.912573109300517, "grad_norm": 6.826006889343262, "learning_rate": 4.1176412893229366e-05, "loss": 0.0613, "step": 115730 }, { "epoch": 3.9129111869907707, "grad_norm": 7.5167036056518555, "learning_rate": 4.1168275414996307e-05, "loss": 0.0537, "step": 115740 }, { "epoch": 3.9132492646810237, "grad_norm": 1.186816930770874, "learning_rate": 4.1160138178219266e-05, "loss": 0.0617, "step": 115750 }, { "epoch": 3.913587342371277, "grad_norm": 1.252193570137024, "learning_rate": 4.1152001183120735e-05, "loss": 0.0619, "step": 115760 }, { "epoch": 3.9139254200615303, "grad_norm": 1.2202482223510742, "learning_rate": 4.1143864429923166e-05, "loss": 0.0615, "step": 115770 }, { "epoch": 3.9142634977517834, "grad_norm": 2.8832755088806152, "learning_rate": 4.1135727918848995e-05, "loss": 0.0678, "step": 115780 }, { "epoch": 3.9146015754420365, "grad_norm": 0.9524130821228027, "learning_rate": 4.1127591650120716e-05, "loss": 0.0593, "step": 115790 }, { "epoch": 3.91493965313229, "grad_norm": 0.6847655773162842, "learning_rate": 4.111945562396074e-05, "loss": 0.0709, "step": 115800 }, { "epoch": 3.915277730822543, "grad_norm": 1.9857233762741089, "learning_rate": 4.1111319840591525e-05, "loss": 0.0562, "step": 115810 }, { "epoch": 3.915615808512796, "grad_norm": 2.1392951011657715, "learning_rate": 4.110318430023547e-05, "loss": 0.0678, "step": 115820 }, { "epoch": 3.9159538862030496, "grad_norm": 2.767364025115967, "learning_rate": 4.109504900311503e-05, "loss": 0.0633, "step": 115830 }, { "epoch": 3.9162919638933027, "grad_norm": 1.3987656831741333, "learning_rate": 4.108691394945259e-05, "loss": 0.0592, "step": 115840 }, { "epoch": 3.9166300415835558, "grad_norm": 2.174443483352661, "learning_rate": 4.1078779139470575e-05, "loss": 0.0635, "step": 115850 }, { "epoch": 3.9169681192738093, "grad_norm": 1.6486722230911255, "learning_rate": 4.10706445733914e-05, "loss": 0.0655, "step": 115860 }, { "epoch": 3.9173061969640623, "grad_norm": 1.4936875104904175, "learning_rate": 4.1062510251437417e-05, "loss": 0.062, "step": 115870 }, { "epoch": 3.9176442746543154, "grad_norm": 1.7990206480026245, "learning_rate": 4.105437617383109e-05, "loss": 0.0612, "step": 115880 }, { "epoch": 3.917982352344569, "grad_norm": 3.1593048572540283, "learning_rate": 4.104624234079472e-05, "loss": 0.0668, "step": 115890 }, { "epoch": 3.918320430034822, "grad_norm": 1.4931811094284058, "learning_rate": 4.103810875255073e-05, "loss": 0.0616, "step": 115900 }, { "epoch": 3.918658507725075, "grad_norm": 0.9084632396697998, "learning_rate": 4.10299754093215e-05, "loss": 0.0576, "step": 115910 }, { "epoch": 3.9189965854153286, "grad_norm": 1.0223760604858398, "learning_rate": 4.102184231132935e-05, "loss": 0.0676, "step": 115920 }, { "epoch": 3.9193346631055817, "grad_norm": 1.0091203451156616, "learning_rate": 4.101370945879667e-05, "loss": 0.0625, "step": 115930 }, { "epoch": 3.9196727407958347, "grad_norm": 2.4773340225219727, "learning_rate": 4.100557685194579e-05, "loss": 0.0662, "step": 115940 }, { "epoch": 3.9200108184860882, "grad_norm": 1.1380809545516968, "learning_rate": 4.099744449099907e-05, "loss": 0.062, "step": 115950 }, { "epoch": 3.9203488961763413, "grad_norm": 1.5026309490203857, "learning_rate": 4.098931237617883e-05, "loss": 0.0593, "step": 115960 }, { "epoch": 3.9206869738665944, "grad_norm": 1.3929332494735718, "learning_rate": 4.0981180507707386e-05, "loss": 0.0626, "step": 115970 }, { "epoch": 3.921025051556848, "grad_norm": 0.8414669036865234, "learning_rate": 4.097304888580712e-05, "loss": 0.0565, "step": 115980 }, { "epoch": 3.921363129247101, "grad_norm": 0.9555842280387878, "learning_rate": 4.0964917510700265e-05, "loss": 0.066, "step": 115990 }, { "epoch": 3.921701206937354, "grad_norm": 1.2886767387390137, "learning_rate": 4.0956786382609196e-05, "loss": 0.0665, "step": 116000 }, { "epoch": 3.9220392846276075, "grad_norm": 1.709174394607544, "learning_rate": 4.0948655501756175e-05, "loss": 0.0608, "step": 116010 }, { "epoch": 3.9223773623178606, "grad_norm": 2.4484617710113525, "learning_rate": 4.0940524868363525e-05, "loss": 0.0651, "step": 116020 }, { "epoch": 3.9227154400081137, "grad_norm": 2.021101236343384, "learning_rate": 4.0932394482653514e-05, "loss": 0.0691, "step": 116030 }, { "epoch": 3.923053517698367, "grad_norm": 1.4780875444412231, "learning_rate": 4.0924264344848436e-05, "loss": 0.0583, "step": 116040 }, { "epoch": 3.9233915953886203, "grad_norm": 1.599968433380127, "learning_rate": 4.091613445517056e-05, "loss": 0.0538, "step": 116050 }, { "epoch": 3.9237296730788733, "grad_norm": 1.1324588060379028, "learning_rate": 4.090800481384215e-05, "loss": 0.0523, "step": 116060 }, { "epoch": 3.924067750769127, "grad_norm": 2.935490846633911, "learning_rate": 4.0899875421085485e-05, "loss": 0.0586, "step": 116070 }, { "epoch": 3.92440582845938, "grad_norm": 1.9590766429901123, "learning_rate": 4.0891746277122786e-05, "loss": 0.062, "step": 116080 }, { "epoch": 3.924743906149633, "grad_norm": 1.3608328104019165, "learning_rate": 4.088361738217633e-05, "loss": 0.0588, "step": 116090 }, { "epoch": 3.9250819838398865, "grad_norm": 1.4356930255889893, "learning_rate": 4.087548873646836e-05, "loss": 0.0597, "step": 116100 }, { "epoch": 3.9254200615301396, "grad_norm": 3.287623167037964, "learning_rate": 4.0867360340221095e-05, "loss": 0.065, "step": 116110 }, { "epoch": 3.9257581392203926, "grad_norm": 1.2026619911193848, "learning_rate": 4.0859232193656774e-05, "loss": 0.0608, "step": 116120 }, { "epoch": 3.926096216910646, "grad_norm": 4.836538314819336, "learning_rate": 4.08511042969976e-05, "loss": 0.0588, "step": 116130 }, { "epoch": 3.926434294600899, "grad_norm": 0.7168740630149841, "learning_rate": 4.084297665046581e-05, "loss": 0.0631, "step": 116140 }, { "epoch": 3.9267723722911523, "grad_norm": 1.8017158508300781, "learning_rate": 4.083484925428358e-05, "loss": 0.0603, "step": 116150 }, { "epoch": 3.927110449981406, "grad_norm": 2.6137592792510986, "learning_rate": 4.082672210867313e-05, "loss": 0.0612, "step": 116160 }, { "epoch": 3.927448527671659, "grad_norm": 1.381673812866211, "learning_rate": 4.081859521385668e-05, "loss": 0.0652, "step": 116170 }, { "epoch": 3.927786605361912, "grad_norm": 5.739594459533691, "learning_rate": 4.081046857005635e-05, "loss": 0.0638, "step": 116180 }, { "epoch": 3.9281246830521654, "grad_norm": 0.8522220849990845, "learning_rate": 4.080234217749439e-05, "loss": 0.0678, "step": 116190 }, { "epoch": 3.9284627607424185, "grad_norm": 1.891597032546997, "learning_rate": 4.0794216036392906e-05, "loss": 0.057, "step": 116200 }, { "epoch": 3.9288008384326716, "grad_norm": 0.9224226474761963, "learning_rate": 4.078609014697412e-05, "loss": 0.0644, "step": 116210 }, { "epoch": 3.929138916122925, "grad_norm": 2.3536102771759033, "learning_rate": 4.077796450946016e-05, "loss": 0.0532, "step": 116220 }, { "epoch": 3.929476993813178, "grad_norm": 1.2452129125595093, "learning_rate": 4.076983912407319e-05, "loss": 0.0586, "step": 116230 }, { "epoch": 3.9298150715034312, "grad_norm": 1.4615678787231445, "learning_rate": 4.076171399103535e-05, "loss": 0.0586, "step": 116240 }, { "epoch": 3.9301531491936847, "grad_norm": 0.6625009775161743, "learning_rate": 4.075358911056877e-05, "loss": 0.0604, "step": 116250 }, { "epoch": 3.930491226883938, "grad_norm": 1.696329951286316, "learning_rate": 4.0745464482895604e-05, "loss": 0.0588, "step": 116260 }, { "epoch": 3.930829304574191, "grad_norm": 1.4994370937347412, "learning_rate": 4.073734010823795e-05, "loss": 0.0586, "step": 116270 }, { "epoch": 3.9311673822644444, "grad_norm": 4.2487568855285645, "learning_rate": 4.072921598681794e-05, "loss": 0.0492, "step": 116280 }, { "epoch": 3.9315054599546975, "grad_norm": 2.47218918800354, "learning_rate": 4.0721092118857686e-05, "loss": 0.0645, "step": 116290 }, { "epoch": 3.9318435376449505, "grad_norm": 1.010928750038147, "learning_rate": 4.071296850457928e-05, "loss": 0.0541, "step": 116300 }, { "epoch": 3.932181615335204, "grad_norm": 1.504676103591919, "learning_rate": 4.0704845144204845e-05, "loss": 0.0575, "step": 116310 }, { "epoch": 3.932519693025457, "grad_norm": 0.9724406599998474, "learning_rate": 4.0696722037956434e-05, "loss": 0.0563, "step": 116320 }, { "epoch": 3.9328577707157106, "grad_norm": 2.2403225898742676, "learning_rate": 4.0688599186056167e-05, "loss": 0.066, "step": 116330 }, { "epoch": 3.9331958484059637, "grad_norm": 1.573083758354187, "learning_rate": 4.068047658872608e-05, "loss": 0.0618, "step": 116340 }, { "epoch": 3.9335339260962168, "grad_norm": 0.6343200206756592, "learning_rate": 4.067235424618826e-05, "loss": 0.0597, "step": 116350 }, { "epoch": 3.9338720037864703, "grad_norm": 3.068758010864258, "learning_rate": 4.066423215866479e-05, "loss": 0.0619, "step": 116360 }, { "epoch": 3.9342100814767234, "grad_norm": 0.9749755263328552, "learning_rate": 4.065611032637768e-05, "loss": 0.0673, "step": 116370 }, { "epoch": 3.9345481591669764, "grad_norm": 1.2848925590515137, "learning_rate": 4.064798874954903e-05, "loss": 0.0705, "step": 116380 }, { "epoch": 3.93488623685723, "grad_norm": 1.395567536354065, "learning_rate": 4.063986742840083e-05, "loss": 0.0631, "step": 116390 }, { "epoch": 3.935224314547483, "grad_norm": 0.7617850303649902, "learning_rate": 4.063174636315516e-05, "loss": 0.0569, "step": 116400 }, { "epoch": 3.935562392237736, "grad_norm": 1.434840202331543, "learning_rate": 4.062362555403401e-05, "loss": 0.0642, "step": 116410 }, { "epoch": 3.9359004699279896, "grad_norm": 1.3878116607666016, "learning_rate": 4.0615505001259414e-05, "loss": 0.0652, "step": 116420 }, { "epoch": 3.9362385476182427, "grad_norm": 1.2578845024108887, "learning_rate": 4.0607384705053396e-05, "loss": 0.0567, "step": 116430 }, { "epoch": 3.936576625308496, "grad_norm": 2.369778871536255, "learning_rate": 4.0599264665637946e-05, "loss": 0.0582, "step": 116440 }, { "epoch": 3.9369147029987492, "grad_norm": 3.0427582263946533, "learning_rate": 4.059114488323508e-05, "loss": 0.0618, "step": 116450 }, { "epoch": 3.9372527806890023, "grad_norm": 1.2417930364608765, "learning_rate": 4.058302535806676e-05, "loss": 0.0596, "step": 116460 }, { "epoch": 3.937590858379256, "grad_norm": 4.23177433013916, "learning_rate": 4.057490609035498e-05, "loss": 0.0625, "step": 116470 }, { "epoch": 3.937928936069509, "grad_norm": 3.8474600315093994, "learning_rate": 4.056678708032176e-05, "loss": 0.0587, "step": 116480 }, { "epoch": 3.938267013759762, "grad_norm": 1.9023445844650269, "learning_rate": 4.055866832818901e-05, "loss": 0.0639, "step": 116490 }, { "epoch": 3.9386050914500155, "grad_norm": 1.0986970663070679, "learning_rate": 4.055054983417874e-05, "loss": 0.0643, "step": 116500 }, { "epoch": 3.9389431691402685, "grad_norm": 2.74169921875, "learning_rate": 4.054243159851287e-05, "loss": 0.0584, "step": 116510 }, { "epoch": 3.9392812468305216, "grad_norm": 1.0487920045852661, "learning_rate": 4.053431362141338e-05, "loss": 0.0534, "step": 116520 }, { "epoch": 3.939619324520775, "grad_norm": 8.271795272827148, "learning_rate": 4.0526195903102196e-05, "loss": 0.0587, "step": 116530 }, { "epoch": 3.939957402211028, "grad_norm": 1.5188302993774414, "learning_rate": 4.051807844380125e-05, "loss": 0.066, "step": 116540 }, { "epoch": 3.9402954799012813, "grad_norm": 2.2959749698638916, "learning_rate": 4.0509961243732494e-05, "loss": 0.0668, "step": 116550 }, { "epoch": 3.9406335575915348, "grad_norm": 1.5308581590652466, "learning_rate": 4.050184430311782e-05, "loss": 0.0641, "step": 116560 }, { "epoch": 3.940971635281788, "grad_norm": 1.3223463296890259, "learning_rate": 4.049372762217916e-05, "loss": 0.0606, "step": 116570 }, { "epoch": 3.941309712972041, "grad_norm": 1.0546907186508179, "learning_rate": 4.0485611201138396e-05, "loss": 0.0701, "step": 116580 }, { "epoch": 3.9416477906622944, "grad_norm": 1.9813376665115356, "learning_rate": 4.047749504021748e-05, "loss": 0.0613, "step": 116590 }, { "epoch": 3.9419858683525475, "grad_norm": 2.247828722000122, "learning_rate": 4.046937913963824e-05, "loss": 0.0651, "step": 116600 }, { "epoch": 3.9423239460428006, "grad_norm": 3.4862277507781982, "learning_rate": 4.046126349962261e-05, "loss": 0.0628, "step": 116610 }, { "epoch": 3.942662023733054, "grad_norm": 2.442159652709961, "learning_rate": 4.045314812039246e-05, "loss": 0.0591, "step": 116620 }, { "epoch": 3.943000101423307, "grad_norm": 2.743415117263794, "learning_rate": 4.044503300216964e-05, "loss": 0.0608, "step": 116630 }, { "epoch": 3.94333817911356, "grad_norm": 2.713597536087036, "learning_rate": 4.043691814517604e-05, "loss": 0.0526, "step": 116640 }, { "epoch": 3.9436762568038137, "grad_norm": 1.1133434772491455, "learning_rate": 4.0428803549633496e-05, "loss": 0.0599, "step": 116650 }, { "epoch": 3.944014334494067, "grad_norm": 3.3525583744049072, "learning_rate": 4.042068921576385e-05, "loss": 0.0548, "step": 116660 }, { "epoch": 3.94435241218432, "grad_norm": 3.330676555633545, "learning_rate": 4.0412575143789e-05, "loss": 0.0594, "step": 116670 }, { "epoch": 3.9446904898745734, "grad_norm": 1.3973966836929321, "learning_rate": 4.040446133393071e-05, "loss": 0.0593, "step": 116680 }, { "epoch": 3.9450285675648264, "grad_norm": 2.013373613357544, "learning_rate": 4.039634778641088e-05, "loss": 0.06, "step": 116690 }, { "epoch": 3.9453666452550795, "grad_norm": 1.3188953399658203, "learning_rate": 4.038823450145126e-05, "loss": 0.0718, "step": 116700 }, { "epoch": 3.945704722945333, "grad_norm": 1.6691721677780151, "learning_rate": 4.038012147927371e-05, "loss": 0.062, "step": 116710 }, { "epoch": 3.946042800635586, "grad_norm": 3.356579542160034, "learning_rate": 4.0372008720100026e-05, "loss": 0.0537, "step": 116720 }, { "epoch": 3.946380878325839, "grad_norm": 6.146996974945068, "learning_rate": 4.0363896224152e-05, "loss": 0.0614, "step": 116730 }, { "epoch": 3.9467189560160927, "grad_norm": 2.5027191638946533, "learning_rate": 4.035578399165144e-05, "loss": 0.0619, "step": 116740 }, { "epoch": 3.9470570337063458, "grad_norm": 1.0511606931686401, "learning_rate": 4.0347672022820115e-05, "loss": 0.0648, "step": 116750 }, { "epoch": 3.947395111396599, "grad_norm": 2.040698528289795, "learning_rate": 4.033956031787982e-05, "loss": 0.0605, "step": 116760 }, { "epoch": 3.9477331890868523, "grad_norm": 2.922880172729492, "learning_rate": 4.0331448877052306e-05, "loss": 0.0593, "step": 116770 }, { "epoch": 3.9480712667771054, "grad_norm": 2.2137768268585205, "learning_rate": 4.0323337700559363e-05, "loss": 0.0604, "step": 116780 }, { "epoch": 3.9484093444673585, "grad_norm": 2.8698198795318604, "learning_rate": 4.03152267886227e-05, "loss": 0.0621, "step": 116790 }, { "epoch": 3.948747422157612, "grad_norm": 1.1100294589996338, "learning_rate": 4.030711614146412e-05, "loss": 0.0575, "step": 116800 }, { "epoch": 3.949085499847865, "grad_norm": 1.6990129947662354, "learning_rate": 4.029900575930535e-05, "loss": 0.0564, "step": 116810 }, { "epoch": 3.949423577538118, "grad_norm": 4.257694721221924, "learning_rate": 4.029089564236811e-05, "loss": 0.0629, "step": 116820 }, { "epoch": 3.9497616552283716, "grad_norm": 1.3301029205322266, "learning_rate": 4.028278579087413e-05, "loss": 0.0577, "step": 116830 }, { "epoch": 3.9500997329186247, "grad_norm": 1.6397464275360107, "learning_rate": 4.0274676205045145e-05, "loss": 0.0586, "step": 116840 }, { "epoch": 3.9504378106088778, "grad_norm": 1.5166839361190796, "learning_rate": 4.026656688510286e-05, "loss": 0.0631, "step": 116850 }, { "epoch": 3.9507758882991313, "grad_norm": 3.6863996982574463, "learning_rate": 4.025845783126897e-05, "loss": 0.0582, "step": 116860 }, { "epoch": 3.9511139659893844, "grad_norm": 2.47359299659729, "learning_rate": 4.025034904376518e-05, "loss": 0.0585, "step": 116870 }, { "epoch": 3.9514520436796374, "grad_norm": 3.4512929916381836, "learning_rate": 4.0242240522813206e-05, "loss": 0.06, "step": 116880 }, { "epoch": 3.951790121369891, "grad_norm": 7.8616862297058105, "learning_rate": 4.0234132268634686e-05, "loss": 0.0598, "step": 116890 }, { "epoch": 3.952128199060144, "grad_norm": 1.4293960332870483, "learning_rate": 4.0226024281451345e-05, "loss": 0.0617, "step": 116900 }, { "epoch": 3.952466276750397, "grad_norm": 1.374343752861023, "learning_rate": 4.02179165614848e-05, "loss": 0.0631, "step": 116910 }, { "epoch": 3.9528043544406506, "grad_norm": 2.1922121047973633, "learning_rate": 4.0209809108956755e-05, "loss": 0.061, "step": 116920 }, { "epoch": 3.9531424321309037, "grad_norm": 1.5185238122940063, "learning_rate": 4.0201701924088864e-05, "loss": 0.0548, "step": 116930 }, { "epoch": 3.9534805098211567, "grad_norm": 1.8866312503814697, "learning_rate": 4.0193595007102744e-05, "loss": 0.0671, "step": 116940 }, { "epoch": 3.9538185875114102, "grad_norm": 2.1737060546875, "learning_rate": 4.018548835822006e-05, "loss": 0.0618, "step": 116950 }, { "epoch": 3.9541566652016633, "grad_norm": 2.7580761909484863, "learning_rate": 4.017738197766244e-05, "loss": 0.0566, "step": 116960 }, { "epoch": 3.9544947428919164, "grad_norm": 1.678758144378662, "learning_rate": 4.0169275865651504e-05, "loss": 0.0586, "step": 116970 }, { "epoch": 3.95483282058217, "grad_norm": 1.476130723953247, "learning_rate": 4.016117002240887e-05, "loss": 0.0644, "step": 116980 }, { "epoch": 3.955170898272423, "grad_norm": 1.115346074104309, "learning_rate": 4.0153064448156135e-05, "loss": 0.0622, "step": 116990 }, { "epoch": 3.955508975962676, "grad_norm": 3.2242512702941895, "learning_rate": 4.014495914311494e-05, "loss": 0.059, "step": 117000 }, { "epoch": 3.9558470536529295, "grad_norm": 1.4328346252441406, "learning_rate": 4.013685410750686e-05, "loss": 0.0603, "step": 117010 }, { "epoch": 3.9561851313431826, "grad_norm": 1.3331247568130493, "learning_rate": 4.012874934155348e-05, "loss": 0.0615, "step": 117020 }, { "epoch": 3.9565232090334357, "grad_norm": 2.2005858421325684, "learning_rate": 4.012064484547638e-05, "loss": 0.0597, "step": 117030 }, { "epoch": 3.956861286723689, "grad_norm": 1.6870113611221313, "learning_rate": 4.0112540619497145e-05, "loss": 0.06, "step": 117040 }, { "epoch": 3.9571993644139423, "grad_norm": 1.3257734775543213, "learning_rate": 4.010443666383733e-05, "loss": 0.0691, "step": 117050 }, { "epoch": 3.9575374421041953, "grad_norm": 1.0440021753311157, "learning_rate": 4.0096332978718495e-05, "loss": 0.0683, "step": 117060 }, { "epoch": 3.957875519794449, "grad_norm": 2.5504846572875977, "learning_rate": 4.008822956436221e-05, "loss": 0.0657, "step": 117070 }, { "epoch": 3.958213597484702, "grad_norm": 0.8523148894309998, "learning_rate": 4.0080126420989985e-05, "loss": 0.0689, "step": 117080 }, { "epoch": 3.958551675174955, "grad_norm": 1.2909395694732666, "learning_rate": 4.00720235488234e-05, "loss": 0.0576, "step": 117090 }, { "epoch": 3.9588897528652085, "grad_norm": 2.515956163406372, "learning_rate": 4.006392094808393e-05, "loss": 0.0586, "step": 117100 }, { "epoch": 3.9592278305554616, "grad_norm": 1.9515366554260254, "learning_rate": 4.005581861899315e-05, "loss": 0.0586, "step": 117110 }, { "epoch": 3.9595659082457146, "grad_norm": 1.3194140195846558, "learning_rate": 4.0047716561772556e-05, "loss": 0.0609, "step": 117120 }, { "epoch": 3.959903985935968, "grad_norm": 1.7208406925201416, "learning_rate": 4.003961477664365e-05, "loss": 0.0591, "step": 117130 }, { "epoch": 3.960242063626221, "grad_norm": 1.7829411029815674, "learning_rate": 4.0031513263827934e-05, "loss": 0.0589, "step": 117140 }, { "epoch": 3.9605801413164743, "grad_norm": 1.0554389953613281, "learning_rate": 4.002341202354689e-05, "loss": 0.0609, "step": 117150 }, { "epoch": 3.960918219006728, "grad_norm": 1.0347368717193604, "learning_rate": 4.0015311056022034e-05, "loss": 0.0577, "step": 117160 }, { "epoch": 3.961256296696981, "grad_norm": 2.138986587524414, "learning_rate": 4.0007210361474806e-05, "loss": 0.0637, "step": 117170 }, { "epoch": 3.961594374387234, "grad_norm": 0.5598974227905273, "learning_rate": 3.999910994012668e-05, "loss": 0.06, "step": 117180 }, { "epoch": 3.9619324520774875, "grad_norm": 1.2108182907104492, "learning_rate": 3.999100979219917e-05, "loss": 0.0592, "step": 117190 }, { "epoch": 3.9622705297677405, "grad_norm": 1.4766974449157715, "learning_rate": 3.9982909917913644e-05, "loss": 0.0655, "step": 117200 }, { "epoch": 3.9626086074579936, "grad_norm": 1.590384840965271, "learning_rate": 3.9974810317491635e-05, "loss": 0.0639, "step": 117210 }, { "epoch": 3.962946685148247, "grad_norm": 2.3258073329925537, "learning_rate": 3.996671099115453e-05, "loss": 0.0658, "step": 117220 }, { "epoch": 3.9632847628385, "grad_norm": 1.3422715663909912, "learning_rate": 3.995861193912379e-05, "loss": 0.0634, "step": 117230 }, { "epoch": 3.9636228405287532, "grad_norm": 1.7089529037475586, "learning_rate": 3.995051316162081e-05, "loss": 0.0594, "step": 117240 }, { "epoch": 3.9639609182190068, "grad_norm": 1.6143118143081665, "learning_rate": 3.9942414658867035e-05, "loss": 0.0566, "step": 117250 }, { "epoch": 3.96429899590926, "grad_norm": 0.9483385682106018, "learning_rate": 3.993431643108387e-05, "loss": 0.0593, "step": 117260 }, { "epoch": 3.964637073599513, "grad_norm": 2.007972478866577, "learning_rate": 3.992621847849269e-05, "loss": 0.0555, "step": 117270 }, { "epoch": 3.9649751512897664, "grad_norm": 0.9639601111412048, "learning_rate": 3.991812080131494e-05, "loss": 0.0614, "step": 117280 }, { "epoch": 3.9653132289800195, "grad_norm": 1.0293610095977783, "learning_rate": 3.991002339977196e-05, "loss": 0.0586, "step": 117290 }, { "epoch": 3.965651306670273, "grad_norm": 1.0996071100234985, "learning_rate": 3.9901926274085146e-05, "loss": 0.0626, "step": 117300 }, { "epoch": 3.965989384360526, "grad_norm": 0.8784464597702026, "learning_rate": 3.989382942447589e-05, "loss": 0.0593, "step": 117310 }, { "epoch": 3.966327462050779, "grad_norm": 1.944823145866394, "learning_rate": 3.988573285116554e-05, "loss": 0.0573, "step": 117320 }, { "epoch": 3.9666655397410326, "grad_norm": 2.8427107334136963, "learning_rate": 3.987763655437545e-05, "loss": 0.0575, "step": 117330 }, { "epoch": 3.9670036174312857, "grad_norm": 0.8421029448509216, "learning_rate": 3.986954053432697e-05, "loss": 0.0588, "step": 117340 }, { "epoch": 3.967341695121539, "grad_norm": 2.34602689743042, "learning_rate": 3.986144479124145e-05, "loss": 0.0655, "step": 117350 }, { "epoch": 3.9676797728117923, "grad_norm": 4.17707633972168, "learning_rate": 3.985334932534022e-05, "loss": 0.0645, "step": 117360 }, { "epoch": 3.9680178505020454, "grad_norm": 1.0301843881607056, "learning_rate": 3.9845254136844593e-05, "loss": 0.064, "step": 117370 }, { "epoch": 3.9683559281922984, "grad_norm": 1.3456491231918335, "learning_rate": 3.9837159225975935e-05, "loss": 0.0587, "step": 117380 }, { "epoch": 3.968694005882552, "grad_norm": 1.221017837524414, "learning_rate": 3.98290645929555e-05, "loss": 0.0636, "step": 117390 }, { "epoch": 3.969032083572805, "grad_norm": 1.3540220260620117, "learning_rate": 3.982097023800464e-05, "loss": 0.0551, "step": 117400 }, { "epoch": 3.9693701612630585, "grad_norm": 0.9401470422744751, "learning_rate": 3.98128761613446e-05, "loss": 0.0546, "step": 117410 }, { "epoch": 3.9697082389533116, "grad_norm": 5.230093479156494, "learning_rate": 3.9804782363196716e-05, "loss": 0.064, "step": 117420 }, { "epoch": 3.9700463166435647, "grad_norm": 4.338653087615967, "learning_rate": 3.979668884378225e-05, "loss": 0.0701, "step": 117430 }, { "epoch": 3.970384394333818, "grad_norm": 1.736917495727539, "learning_rate": 3.978859560332247e-05, "loss": 0.0621, "step": 117440 }, { "epoch": 3.9707224720240712, "grad_norm": 1.552626132965088, "learning_rate": 3.978050264203867e-05, "loss": 0.0616, "step": 117450 }, { "epoch": 3.9710605497143243, "grad_norm": 1.9837778806686401, "learning_rate": 3.9772409960152066e-05, "loss": 0.0618, "step": 117460 }, { "epoch": 3.971398627404578, "grad_norm": 1.536336064338684, "learning_rate": 3.9764317557883954e-05, "loss": 0.059, "step": 117470 }, { "epoch": 3.971736705094831, "grad_norm": 1.9525970220565796, "learning_rate": 3.975622543545554e-05, "loss": 0.0641, "step": 117480 }, { "epoch": 3.972074782785084, "grad_norm": 1.0284106731414795, "learning_rate": 3.974813359308806e-05, "loss": 0.0608, "step": 117490 }, { "epoch": 3.9724128604753375, "grad_norm": 4.066507816314697, "learning_rate": 3.974004203100278e-05, "loss": 0.06, "step": 117500 }, { "epoch": 3.9727509381655906, "grad_norm": 3.4992382526397705, "learning_rate": 3.9731950749420886e-05, "loss": 0.064, "step": 117510 }, { "epoch": 3.9730890158558436, "grad_norm": 7.968722343444824, "learning_rate": 3.972385974856361e-05, "loss": 0.0623, "step": 117520 }, { "epoch": 3.973427093546097, "grad_norm": 1.8620096445083618, "learning_rate": 3.971576902865214e-05, "loss": 0.0666, "step": 117530 }, { "epoch": 3.97376517123635, "grad_norm": 5.814221382141113, "learning_rate": 3.9707678589907686e-05, "loss": 0.0594, "step": 117540 }, { "epoch": 3.9741032489266033, "grad_norm": 1.0943738222122192, "learning_rate": 3.9699588432551424e-05, "loss": 0.0574, "step": 117550 }, { "epoch": 3.974441326616857, "grad_norm": 4.838694095611572, "learning_rate": 3.9691498556804554e-05, "loss": 0.0621, "step": 117560 }, { "epoch": 3.97477940430711, "grad_norm": 1.4106543064117432, "learning_rate": 3.968340896288824e-05, "loss": 0.0586, "step": 117570 }, { "epoch": 3.975117481997363, "grad_norm": 2.136516571044922, "learning_rate": 3.9675319651023636e-05, "loss": 0.0591, "step": 117580 }, { "epoch": 3.9754555596876164, "grad_norm": 1.6500886678695679, "learning_rate": 3.9667230621431936e-05, "loss": 0.0557, "step": 117590 }, { "epoch": 3.9757936373778695, "grad_norm": 0.8144700527191162, "learning_rate": 3.965914187433424e-05, "loss": 0.0635, "step": 117600 }, { "epoch": 3.9761317150681226, "grad_norm": 1.1212866306304932, "learning_rate": 3.965105340995176e-05, "loss": 0.0659, "step": 117610 }, { "epoch": 3.976469792758376, "grad_norm": 1.5418757200241089, "learning_rate": 3.964296522850556e-05, "loss": 0.0573, "step": 117620 }, { "epoch": 3.976807870448629, "grad_norm": 1.311234951019287, "learning_rate": 3.9634877330216803e-05, "loss": 0.0613, "step": 117630 }, { "epoch": 3.9771459481388822, "grad_norm": 5.8320136070251465, "learning_rate": 3.962678971530662e-05, "loss": 0.0681, "step": 117640 }, { "epoch": 3.9774840258291357, "grad_norm": 1.3757647275924683, "learning_rate": 3.961870238399609e-05, "loss": 0.0687, "step": 117650 }, { "epoch": 3.977822103519389, "grad_norm": 7.850348949432373, "learning_rate": 3.961061533650636e-05, "loss": 0.0593, "step": 117660 }, { "epoch": 3.978160181209642, "grad_norm": 0.7885873913764954, "learning_rate": 3.960252857305848e-05, "loss": 0.0603, "step": 117670 }, { "epoch": 3.9784982588998954, "grad_norm": 1.817987084388733, "learning_rate": 3.959444209387356e-05, "loss": 0.059, "step": 117680 }, { "epoch": 3.9788363365901485, "grad_norm": 3.4228596687316895, "learning_rate": 3.9586355899172726e-05, "loss": 0.0652, "step": 117690 }, { "epoch": 3.9791744142804015, "grad_norm": 2.765326976776123, "learning_rate": 3.9578269989176963e-05, "loss": 0.0582, "step": 117700 }, { "epoch": 3.979512491970655, "grad_norm": 2.646453380584717, "learning_rate": 3.9570184364107404e-05, "loss": 0.0601, "step": 117710 }, { "epoch": 3.979850569660908, "grad_norm": 0.7094288468360901, "learning_rate": 3.9562099024185084e-05, "loss": 0.0561, "step": 117720 }, { "epoch": 3.980188647351161, "grad_norm": 0.8989772796630859, "learning_rate": 3.955401396963106e-05, "loss": 0.0634, "step": 117730 }, { "epoch": 3.9805267250414147, "grad_norm": 0.7521148324012756, "learning_rate": 3.954592920066636e-05, "loss": 0.057, "step": 117740 }, { "epoch": 3.9808648027316678, "grad_norm": 1.1397712230682373, "learning_rate": 3.953784471751203e-05, "loss": 0.0594, "step": 117750 }, { "epoch": 3.981202880421921, "grad_norm": 0.9336767196655273, "learning_rate": 3.9529760520389104e-05, "loss": 0.0583, "step": 117760 }, { "epoch": 3.9815409581121743, "grad_norm": 1.5927939414978027, "learning_rate": 3.952167660951858e-05, "loss": 0.0605, "step": 117770 }, { "epoch": 3.9818790358024274, "grad_norm": 2.7095985412597656, "learning_rate": 3.9513592985121495e-05, "loss": 0.057, "step": 117780 }, { "epoch": 3.9822171134926805, "grad_norm": 2.8671772480010986, "learning_rate": 3.950550964741881e-05, "loss": 0.0639, "step": 117790 }, { "epoch": 3.982555191182934, "grad_norm": 2.9149534702301025, "learning_rate": 3.9497426596631594e-05, "loss": 0.061, "step": 117800 }, { "epoch": 3.982893268873187, "grad_norm": 10.396185874938965, "learning_rate": 3.948934383298075e-05, "loss": 0.0604, "step": 117810 }, { "epoch": 3.98323134656344, "grad_norm": 2.4468746185302734, "learning_rate": 3.9481261356687306e-05, "loss": 0.0654, "step": 117820 }, { "epoch": 3.9835694242536936, "grad_norm": 21.29556655883789, "learning_rate": 3.947317916797224e-05, "loss": 0.0618, "step": 117830 }, { "epoch": 3.9839075019439467, "grad_norm": 2.4455385208129883, "learning_rate": 3.9465097267056486e-05, "loss": 0.0654, "step": 117840 }, { "epoch": 3.9842455796342, "grad_norm": 6.287729263305664, "learning_rate": 3.9457015654161024e-05, "loss": 0.0697, "step": 117850 }, { "epoch": 3.9845836573244533, "grad_norm": 1.4224861860275269, "learning_rate": 3.944893432950679e-05, "loss": 0.061, "step": 117860 }, { "epoch": 3.9849217350147064, "grad_norm": 2.930206537246704, "learning_rate": 3.9440853293314704e-05, "loss": 0.0626, "step": 117870 }, { "epoch": 3.9852598127049594, "grad_norm": 2.562354564666748, "learning_rate": 3.943277254580576e-05, "loss": 0.0596, "step": 117880 }, { "epoch": 3.985597890395213, "grad_norm": 1.5328311920166016, "learning_rate": 3.942469208720082e-05, "loss": 0.0617, "step": 117890 }, { "epoch": 3.985935968085466, "grad_norm": 4.987396240234375, "learning_rate": 3.941661191772085e-05, "loss": 0.0612, "step": 117900 }, { "epoch": 3.986274045775719, "grad_norm": 9.268474578857422, "learning_rate": 3.940853203758671e-05, "loss": 0.0631, "step": 117910 }, { "epoch": 3.9866121234659726, "grad_norm": 2.3257951736450195, "learning_rate": 3.940045244701934e-05, "loss": 0.0575, "step": 117920 }, { "epoch": 3.9869502011562257, "grad_norm": 2.779531240463257, "learning_rate": 3.93923731462396e-05, "loss": 0.0638, "step": 117930 }, { "epoch": 3.9872882788464787, "grad_norm": 1.755773663520813, "learning_rate": 3.9384294135468404e-05, "loss": 0.0665, "step": 117940 }, { "epoch": 3.9876263565367323, "grad_norm": 2.7862181663513184, "learning_rate": 3.937621541492662e-05, "loss": 0.063, "step": 117950 }, { "epoch": 3.9879644342269853, "grad_norm": 1.1476410627365112, "learning_rate": 3.9368136984835116e-05, "loss": 0.0632, "step": 117960 }, { "epoch": 3.9883025119172384, "grad_norm": 0.8583364486694336, "learning_rate": 3.936005884541476e-05, "loss": 0.0582, "step": 117970 }, { "epoch": 3.988640589607492, "grad_norm": 1.8151057958602905, "learning_rate": 3.935198099688638e-05, "loss": 0.0594, "step": 117980 }, { "epoch": 3.988978667297745, "grad_norm": 1.8743512630462646, "learning_rate": 3.934390343947086e-05, "loss": 0.0654, "step": 117990 }, { "epoch": 3.989316744987998, "grad_norm": 1.3924306631088257, "learning_rate": 3.933582617338899e-05, "loss": 0.0642, "step": 118000 }, { "epoch": 3.9896548226782516, "grad_norm": 1.3086727857589722, "learning_rate": 3.932774919886164e-05, "loss": 0.0634, "step": 118010 }, { "epoch": 3.9899929003685046, "grad_norm": 1.5459362268447876, "learning_rate": 3.931967251610962e-05, "loss": 0.058, "step": 118020 }, { "epoch": 3.9903309780587577, "grad_norm": 1.1506052017211914, "learning_rate": 3.931159612535374e-05, "loss": 0.0619, "step": 118030 }, { "epoch": 3.990669055749011, "grad_norm": 1.4732919931411743, "learning_rate": 3.930352002681481e-05, "loss": 0.0667, "step": 118040 }, { "epoch": 3.9910071334392643, "grad_norm": 1.6382255554199219, "learning_rate": 3.9295444220713615e-05, "loss": 0.0634, "step": 118050 }, { "epoch": 3.9913452111295173, "grad_norm": 11.148500442504883, "learning_rate": 3.9287368707270964e-05, "loss": 0.0627, "step": 118060 }, { "epoch": 3.991683288819771, "grad_norm": 2.9773364067077637, "learning_rate": 3.927929348670763e-05, "loss": 0.067, "step": 118070 }, { "epoch": 3.992021366510024, "grad_norm": 9.219109535217285, "learning_rate": 3.9271218559244356e-05, "loss": 0.0681, "step": 118080 }, { "epoch": 3.992359444200277, "grad_norm": 1.7940255403518677, "learning_rate": 3.926314392510197e-05, "loss": 0.0567, "step": 118090 }, { "epoch": 3.9926975218905305, "grad_norm": 4.802109718322754, "learning_rate": 3.925506958450117e-05, "loss": 0.0593, "step": 118100 }, { "epoch": 3.9930355995807836, "grad_norm": 1.5808073282241821, "learning_rate": 3.924699553766276e-05, "loss": 0.0646, "step": 118110 }, { "epoch": 3.9933736772710366, "grad_norm": 2.873236656188965, "learning_rate": 3.9238921784807426e-05, "loss": 0.0623, "step": 118120 }, { "epoch": 3.99371175496129, "grad_norm": 2.201322555541992, "learning_rate": 3.923084832615593e-05, "loss": 0.0673, "step": 118130 }, { "epoch": 3.9940498326515432, "grad_norm": 2.0237960815429688, "learning_rate": 3.922277516192901e-05, "loss": 0.0626, "step": 118140 }, { "epoch": 3.9943879103417963, "grad_norm": 6.478248119354248, "learning_rate": 3.921470229234736e-05, "loss": 0.0594, "step": 118150 }, { "epoch": 3.99472598803205, "grad_norm": 4.425425052642822, "learning_rate": 3.92066297176317e-05, "loss": 0.0554, "step": 118160 }, { "epoch": 3.995064065722303, "grad_norm": 2.1869261264801025, "learning_rate": 3.9198557438002726e-05, "loss": 0.0625, "step": 118170 }, { "epoch": 3.995402143412556, "grad_norm": 3.8298463821411133, "learning_rate": 3.9190485453681136e-05, "loss": 0.0605, "step": 118180 }, { "epoch": 3.9957402211028095, "grad_norm": 1.3987064361572266, "learning_rate": 3.918241376488761e-05, "loss": 0.0645, "step": 118190 }, { "epoch": 3.9960782987930625, "grad_norm": 4.512307643890381, "learning_rate": 3.917434237184281e-05, "loss": 0.0643, "step": 118200 }, { "epoch": 3.9964163764833156, "grad_norm": 2.192807197570801, "learning_rate": 3.916627127476745e-05, "loss": 0.0617, "step": 118210 }, { "epoch": 3.996754454173569, "grad_norm": 2.5840940475463867, "learning_rate": 3.915820047388215e-05, "loss": 0.0648, "step": 118220 }, { "epoch": 3.997092531863822, "grad_norm": 1.7571625709533691, "learning_rate": 3.915012996940759e-05, "loss": 0.0626, "step": 118230 }, { "epoch": 3.9974306095540753, "grad_norm": 0.8147624135017395, "learning_rate": 3.914205976156439e-05, "loss": 0.0623, "step": 118240 }, { "epoch": 3.9977686872443288, "grad_norm": 2.7527120113372803, "learning_rate": 3.9133989850573204e-05, "loss": 0.0639, "step": 118250 }, { "epoch": 3.998106764934582, "grad_norm": 1.2097628116607666, "learning_rate": 3.912592023665464e-05, "loss": 0.0615, "step": 118260 }, { "epoch": 3.9984448426248353, "grad_norm": 1.9860140085220337, "learning_rate": 3.9117850920029335e-05, "loss": 0.0637, "step": 118270 }, { "epoch": 3.9987829203150884, "grad_norm": 1.936922311782837, "learning_rate": 3.910978190091791e-05, "loss": 0.0585, "step": 118280 }, { "epoch": 3.9991209980053415, "grad_norm": 3.655179977416992, "learning_rate": 3.910171317954093e-05, "loss": 0.0582, "step": 118290 }, { "epoch": 3.999459075695595, "grad_norm": 3.342073440551758, "learning_rate": 3.909364475611906e-05, "loss": 0.0598, "step": 118300 }, { "epoch": 3.999797153385848, "grad_norm": 4.439705848693848, "learning_rate": 3.90855766308728e-05, "loss": 0.0601, "step": 118310 }, { "epoch": 4.000135231076102, "grad_norm": 16.83747673034668, "learning_rate": 3.907750880402279e-05, "loss": 0.0675, "step": 118320 }, { "epoch": 4.000473308766354, "grad_norm": 6.330966472625732, "learning_rate": 3.90694412757896e-05, "loss": 0.0628, "step": 118330 }, { "epoch": 4.000811386456608, "grad_norm": 2.109557628631592, "learning_rate": 3.906137404639377e-05, "loss": 0.0605, "step": 118340 }, { "epoch": 4.001149464146861, "grad_norm": 3.3838634490966797, "learning_rate": 3.9053307116055874e-05, "loss": 0.0611, "step": 118350 }, { "epoch": 4.001487541837114, "grad_norm": 6.572842121124268, "learning_rate": 3.904524048499643e-05, "loss": 0.0676, "step": 118360 }, { "epoch": 4.001825619527367, "grad_norm": 6.084491729736328, "learning_rate": 3.9037174153436015e-05, "loss": 0.0607, "step": 118370 }, { "epoch": 4.002163697217621, "grad_norm": 4.929693698883057, "learning_rate": 3.902910812159512e-05, "loss": 0.0572, "step": 118380 }, { "epoch": 4.0025017749078735, "grad_norm": 2.5617241859436035, "learning_rate": 3.9021042389694285e-05, "loss": 0.0615, "step": 118390 }, { "epoch": 4.002839852598127, "grad_norm": 9.647478103637695, "learning_rate": 3.9012976957954055e-05, "loss": 0.0667, "step": 118400 }, { "epoch": 4.0031779302883805, "grad_norm": 7.0751447677612305, "learning_rate": 3.9004911826594876e-05, "loss": 0.0588, "step": 118410 }, { "epoch": 4.003516007978633, "grad_norm": 8.338226318359375, "learning_rate": 3.89968469958373e-05, "loss": 0.0658, "step": 118420 }, { "epoch": 4.003854085668887, "grad_norm": 1.863383173942566, "learning_rate": 3.898878246590178e-05, "loss": 0.0606, "step": 118430 }, { "epoch": 4.00419216335914, "grad_norm": 2.583012342453003, "learning_rate": 3.8980718237008825e-05, "loss": 0.0632, "step": 118440 }, { "epoch": 4.004530241049393, "grad_norm": 1.61277437210083, "learning_rate": 3.897265430937888e-05, "loss": 0.0596, "step": 118450 }, { "epoch": 4.004868318739646, "grad_norm": 14.693275451660156, "learning_rate": 3.8964590683232424e-05, "loss": 0.0667, "step": 118460 }, { "epoch": 4.0052063964299, "grad_norm": 5.804349899291992, "learning_rate": 3.895652735878992e-05, "loss": 0.0755, "step": 118470 }, { "epoch": 4.0055444741201525, "grad_norm": 3.5138823986053467, "learning_rate": 3.8948464336271805e-05, "loss": 0.066, "step": 118480 }, { "epoch": 4.005882551810406, "grad_norm": 10.401470184326172, "learning_rate": 3.8940401615898534e-05, "loss": 0.0662, "step": 118490 }, { "epoch": 4.0062206295006595, "grad_norm": 3.548867702484131, "learning_rate": 3.89323391978905e-05, "loss": 0.0586, "step": 118500 }, { "epoch": 4.006558707190912, "grad_norm": 2.8302130699157715, "learning_rate": 3.892427708246818e-05, "loss": 0.0682, "step": 118510 }, { "epoch": 4.006896784881166, "grad_norm": 4.970328330993652, "learning_rate": 3.891621526985195e-05, "loss": 0.0612, "step": 118520 }, { "epoch": 4.007234862571419, "grad_norm": 15.116427421569824, "learning_rate": 3.890815376026224e-05, "loss": 0.0702, "step": 118530 }, { "epoch": 4.007572940261672, "grad_norm": 5.038491249084473, "learning_rate": 3.890009255391944e-05, "loss": 0.0608, "step": 118540 }, { "epoch": 4.007911017951925, "grad_norm": 5.101097106933594, "learning_rate": 3.8892031651043934e-05, "loss": 0.0651, "step": 118550 }, { "epoch": 4.008249095642179, "grad_norm": 6.884790420532227, "learning_rate": 3.888397105185612e-05, "loss": 0.0579, "step": 118560 }, { "epoch": 4.008587173332431, "grad_norm": 16.27268409729004, "learning_rate": 3.8875910756576346e-05, "loss": 0.0622, "step": 118570 }, { "epoch": 4.008925251022685, "grad_norm": 4.140158653259277, "learning_rate": 3.8867850765424986e-05, "loss": 0.0646, "step": 118580 }, { "epoch": 4.0092633287129384, "grad_norm": 3.5025522708892822, "learning_rate": 3.885979107862243e-05, "loss": 0.0635, "step": 118590 }, { "epoch": 4.009601406403191, "grad_norm": 5.748018264770508, "learning_rate": 3.8851731696388975e-05, "loss": 0.0662, "step": 118600 }, { "epoch": 4.009939484093445, "grad_norm": 6.65300989151001, "learning_rate": 3.884367261894502e-05, "loss": 0.0673, "step": 118610 }, { "epoch": 4.010277561783698, "grad_norm": 7.225161075592041, "learning_rate": 3.883561384651083e-05, "loss": 0.0644, "step": 118620 }, { "epoch": 4.010615639473951, "grad_norm": 8.699572563171387, "learning_rate": 3.8827555379306786e-05, "loss": 0.068, "step": 118630 }, { "epoch": 4.010953717164204, "grad_norm": 2.7828116416931152, "learning_rate": 3.881949721755317e-05, "loss": 0.0666, "step": 118640 }, { "epoch": 4.011291794854458, "grad_norm": 173.7318878173828, "learning_rate": 3.881143936147029e-05, "loss": 0.0721, "step": 118650 }, { "epoch": 4.01162987254471, "grad_norm": 4.822621822357178, "learning_rate": 3.880338181127848e-05, "loss": 0.0705, "step": 118660 }, { "epoch": 4.011967950234964, "grad_norm": 3.72015380859375, "learning_rate": 3.879532456719799e-05, "loss": 0.0658, "step": 118670 }, { "epoch": 4.012306027925217, "grad_norm": 2.857835054397583, "learning_rate": 3.878726762944912e-05, "loss": 0.0621, "step": 118680 }, { "epoch": 4.01264410561547, "grad_norm": 3.2446975708007812, "learning_rate": 3.877921099825214e-05, "loss": 0.0622, "step": 118690 }, { "epoch": 4.0129821833057235, "grad_norm": 11.0689115524292, "learning_rate": 3.87711546738273e-05, "loss": 0.0649, "step": 118700 }, { "epoch": 4.013320260995977, "grad_norm": 2.408191442489624, "learning_rate": 3.876309865639489e-05, "loss": 0.0627, "step": 118710 }, { "epoch": 4.01365833868623, "grad_norm": 5.301121711730957, "learning_rate": 3.875504294617514e-05, "loss": 0.065, "step": 118720 }, { "epoch": 4.013996416376483, "grad_norm": 2.6955809593200684, "learning_rate": 3.874698754338829e-05, "loss": 0.0663, "step": 118730 }, { "epoch": 4.014334494066737, "grad_norm": 6.528587818145752, "learning_rate": 3.873893244825457e-05, "loss": 0.0627, "step": 118740 }, { "epoch": 4.014672571756989, "grad_norm": 4.778284549713135, "learning_rate": 3.873087766099422e-05, "loss": 0.0695, "step": 118750 }, { "epoch": 4.015010649447243, "grad_norm": 3.616319179534912, "learning_rate": 3.872282318182743e-05, "loss": 0.0683, "step": 118760 }, { "epoch": 4.015348727137496, "grad_norm": 3.5718977451324463, "learning_rate": 3.871476901097441e-05, "loss": 0.0607, "step": 118770 }, { "epoch": 4.015686804827749, "grad_norm": 4.979985237121582, "learning_rate": 3.870671514865538e-05, "loss": 0.0618, "step": 118780 }, { "epoch": 4.0160248825180025, "grad_norm": 3.5163769721984863, "learning_rate": 3.8698661595090494e-05, "loss": 0.0629, "step": 118790 }, { "epoch": 4.016362960208256, "grad_norm": 12.423202514648438, "learning_rate": 3.869060835049999e-05, "loss": 0.0621, "step": 118800 }, { "epoch": 4.0167010378985095, "grad_norm": 4.641032695770264, "learning_rate": 3.8682555415103964e-05, "loss": 0.0634, "step": 118810 }, { "epoch": 4.017039115588762, "grad_norm": 16.946203231811523, "learning_rate": 3.867450278912266e-05, "loss": 0.0645, "step": 118820 }, { "epoch": 4.017377193279016, "grad_norm": 3.5194878578186035, "learning_rate": 3.866645047277616e-05, "loss": 0.0574, "step": 118830 }, { "epoch": 4.017715270969269, "grad_norm": 5.17702579498291, "learning_rate": 3.865839846628466e-05, "loss": 0.0583, "step": 118840 }, { "epoch": 4.018053348659522, "grad_norm": 5.260003566741943, "learning_rate": 3.86503467698683e-05, "loss": 0.0667, "step": 118850 }, { "epoch": 4.018391426349775, "grad_norm": 5.217527389526367, "learning_rate": 3.864229538374717e-05, "loss": 0.066, "step": 118860 }, { "epoch": 4.018729504040029, "grad_norm": 3.907949447631836, "learning_rate": 3.863424430814144e-05, "loss": 0.0671, "step": 118870 }, { "epoch": 4.0190675817302814, "grad_norm": 9.81484317779541, "learning_rate": 3.862619354327119e-05, "loss": 0.0648, "step": 118880 }, { "epoch": 4.019405659420535, "grad_norm": 6.609201908111572, "learning_rate": 3.861814308935652e-05, "loss": 0.0647, "step": 118890 }, { "epoch": 4.0197437371107885, "grad_norm": 6.144923210144043, "learning_rate": 3.8610092946617584e-05, "loss": 0.063, "step": 118900 }, { "epoch": 4.020081814801041, "grad_norm": 4.847569942474365, "learning_rate": 3.8602043115274386e-05, "loss": 0.0638, "step": 118910 }, { "epoch": 4.020419892491295, "grad_norm": 2.404200553894043, "learning_rate": 3.859399359554707e-05, "loss": 0.0669, "step": 118920 }, { "epoch": 4.020757970181548, "grad_norm": 18.085546493530273, "learning_rate": 3.858594438765567e-05, "loss": 0.0617, "step": 118930 }, { "epoch": 4.021096047871801, "grad_norm": 3.5882058143615723, "learning_rate": 3.857789549182028e-05, "loss": 0.0634, "step": 118940 }, { "epoch": 4.021434125562054, "grad_norm": 8.149802207946777, "learning_rate": 3.856984690826092e-05, "loss": 0.0689, "step": 118950 }, { "epoch": 4.021772203252308, "grad_norm": 5.8586554527282715, "learning_rate": 3.856179863719765e-05, "loss": 0.0705, "step": 118960 }, { "epoch": 4.02211028094256, "grad_norm": 3.491809368133545, "learning_rate": 3.8553750678850524e-05, "loss": 0.057, "step": 118970 }, { "epoch": 4.022448358632814, "grad_norm": 2.4391095638275146, "learning_rate": 3.854570303343955e-05, "loss": 0.0656, "step": 118980 }, { "epoch": 4.022786436323067, "grad_norm": 4.083082675933838, "learning_rate": 3.853765570118475e-05, "loss": 0.0632, "step": 118990 }, { "epoch": 4.02312451401332, "grad_norm": 5.04972505569458, "learning_rate": 3.852960868230612e-05, "loss": 0.0729, "step": 119000 }, { "epoch": 4.023462591703574, "grad_norm": 3.062134027481079, "learning_rate": 3.852156197702371e-05, "loss": 0.0656, "step": 119010 }, { "epoch": 4.023800669393827, "grad_norm": 5.6594624519348145, "learning_rate": 3.851351558555746e-05, "loss": 0.0641, "step": 119020 }, { "epoch": 4.02413874708408, "grad_norm": 7.091756343841553, "learning_rate": 3.850546950812738e-05, "loss": 0.0615, "step": 119030 }, { "epoch": 4.024476824774333, "grad_norm": 4.962924480438232, "learning_rate": 3.8497423744953454e-05, "loss": 0.0658, "step": 119040 }, { "epoch": 4.024814902464587, "grad_norm": 8.077945709228516, "learning_rate": 3.848937829625564e-05, "loss": 0.0682, "step": 119050 }, { "epoch": 4.025152980154839, "grad_norm": 2.528381824493408, "learning_rate": 3.84813331622539e-05, "loss": 0.0618, "step": 119060 }, { "epoch": 4.025491057845093, "grad_norm": 3.2162160873413086, "learning_rate": 3.847328834316818e-05, "loss": 0.065, "step": 119070 }, { "epoch": 4.025829135535346, "grad_norm": 1.4624377489089966, "learning_rate": 3.8465243839218414e-05, "loss": 0.062, "step": 119080 }, { "epoch": 4.026167213225599, "grad_norm": 2.6754825115203857, "learning_rate": 3.845719965062458e-05, "loss": 0.0621, "step": 119090 }, { "epoch": 4.0265052909158525, "grad_norm": 5.231342792510986, "learning_rate": 3.844915577760654e-05, "loss": 0.0608, "step": 119100 }, { "epoch": 4.026843368606106, "grad_norm": 2.182656764984131, "learning_rate": 3.844111222038427e-05, "loss": 0.0645, "step": 119110 }, { "epoch": 4.027181446296359, "grad_norm": 2.4565672874450684, "learning_rate": 3.8433068979177624e-05, "loss": 0.0686, "step": 119120 }, { "epoch": 4.027519523986612, "grad_norm": 5.200992107391357, "learning_rate": 3.842502605420654e-05, "loss": 0.0639, "step": 119130 }, { "epoch": 4.027857601676866, "grad_norm": 2.3295483589172363, "learning_rate": 3.841698344569089e-05, "loss": 0.0733, "step": 119140 }, { "epoch": 4.028195679367118, "grad_norm": 2.2932851314544678, "learning_rate": 3.840894115385056e-05, "loss": 0.0638, "step": 119150 }, { "epoch": 4.028533757057372, "grad_norm": 2.488011360168457, "learning_rate": 3.840089917890544e-05, "loss": 0.0597, "step": 119160 }, { "epoch": 4.028871834747625, "grad_norm": 7.635983467102051, "learning_rate": 3.839285752107536e-05, "loss": 0.0621, "step": 119170 }, { "epoch": 4.029209912437878, "grad_norm": 1.8009809255599976, "learning_rate": 3.838481618058021e-05, "loss": 0.0659, "step": 119180 }, { "epoch": 4.0295479901281315, "grad_norm": 2.414911985397339, "learning_rate": 3.8376775157639804e-05, "loss": 0.061, "step": 119190 }, { "epoch": 4.029886067818385, "grad_norm": 1.794863224029541, "learning_rate": 3.836873445247402e-05, "loss": 0.0653, "step": 119200 }, { "epoch": 4.030224145508638, "grad_norm": 2.4785966873168945, "learning_rate": 3.836069406530264e-05, "loss": 0.0641, "step": 119210 }, { "epoch": 4.030562223198891, "grad_norm": 1.3387227058410645, "learning_rate": 3.835265399634552e-05, "loss": 0.0646, "step": 119220 }, { "epoch": 4.030900300889145, "grad_norm": 7.942816734313965, "learning_rate": 3.834461424582248e-05, "loss": 0.068, "step": 119230 }, { "epoch": 4.031238378579397, "grad_norm": 6.258519172668457, "learning_rate": 3.833657481395329e-05, "loss": 0.0714, "step": 119240 }, { "epoch": 4.031576456269651, "grad_norm": 5.962449073791504, "learning_rate": 3.832853570095778e-05, "loss": 0.0619, "step": 119250 }, { "epoch": 4.031914533959904, "grad_norm": 3.692272186279297, "learning_rate": 3.8320496907055703e-05, "loss": 0.0646, "step": 119260 }, { "epoch": 4.032252611650157, "grad_norm": 5.7900261878967285, "learning_rate": 3.831245843246687e-05, "loss": 0.0651, "step": 119270 }, { "epoch": 4.03259068934041, "grad_norm": 4.276640892028809, "learning_rate": 3.8304420277411015e-05, "loss": 0.0617, "step": 119280 }, { "epoch": 4.032928767030664, "grad_norm": 4.669291973114014, "learning_rate": 3.829638244210791e-05, "loss": 0.0645, "step": 119290 }, { "epoch": 4.033266844720917, "grad_norm": 6.451068878173828, "learning_rate": 3.828834492677734e-05, "loss": 0.0705, "step": 119300 }, { "epoch": 4.03360492241117, "grad_norm": 13.948720932006836, "learning_rate": 3.828030773163899e-05, "loss": 0.064, "step": 119310 }, { "epoch": 4.033943000101424, "grad_norm": 9.42490291595459, "learning_rate": 3.827227085691266e-05, "loss": 0.0642, "step": 119320 }, { "epoch": 4.034281077791676, "grad_norm": 2.531722068786621, "learning_rate": 3.8264234302818004e-05, "loss": 0.0633, "step": 119330 }, { "epoch": 4.03461915548193, "grad_norm": 3.1963860988616943, "learning_rate": 3.825619806957478e-05, "loss": 0.067, "step": 119340 }, { "epoch": 4.034957233172183, "grad_norm": 2.5483264923095703, "learning_rate": 3.82481621574027e-05, "loss": 0.0684, "step": 119350 }, { "epoch": 4.035295310862436, "grad_norm": 3.5443456172943115, "learning_rate": 3.8240126566521444e-05, "loss": 0.0654, "step": 119360 }, { "epoch": 4.035633388552689, "grad_norm": 1.4214801788330078, "learning_rate": 3.8232091297150716e-05, "loss": 0.0683, "step": 119370 }, { "epoch": 4.035971466242943, "grad_norm": 2.231417417526245, "learning_rate": 3.822405634951019e-05, "loss": 0.071, "step": 119380 }, { "epoch": 4.0363095439331955, "grad_norm": 1.8926804065704346, "learning_rate": 3.821602172381954e-05, "loss": 0.0637, "step": 119390 }, { "epoch": 4.036647621623449, "grad_norm": 3.464474678039551, "learning_rate": 3.820798742029842e-05, "loss": 0.0656, "step": 119400 }, { "epoch": 4.0369856993137025, "grad_norm": 2.7269794940948486, "learning_rate": 3.819995343916649e-05, "loss": 0.0706, "step": 119410 }, { "epoch": 4.037323777003955, "grad_norm": 2.259721040725708, "learning_rate": 3.819191978064342e-05, "loss": 0.0712, "step": 119420 }, { "epoch": 4.037661854694209, "grad_norm": 2.5834078788757324, "learning_rate": 3.818388644494881e-05, "loss": 0.0602, "step": 119430 }, { "epoch": 4.037999932384462, "grad_norm": 11.306303024291992, "learning_rate": 3.817585343230232e-05, "loss": 0.0679, "step": 119440 }, { "epoch": 4.038338010074715, "grad_norm": 1.3460572957992554, "learning_rate": 3.8167820742923544e-05, "loss": 0.0701, "step": 119450 }, { "epoch": 4.038676087764968, "grad_norm": 2.992615222930908, "learning_rate": 3.815978837703211e-05, "loss": 0.0666, "step": 119460 }, { "epoch": 4.039014165455222, "grad_norm": 5.515341281890869, "learning_rate": 3.815175633484761e-05, "loss": 0.0642, "step": 119470 }, { "epoch": 4.0393522431454745, "grad_norm": 4.279299736022949, "learning_rate": 3.814372461658964e-05, "loss": 0.0608, "step": 119480 }, { "epoch": 4.039690320835728, "grad_norm": 72.14034271240234, "learning_rate": 3.81356932224778e-05, "loss": 0.0623, "step": 119490 }, { "epoch": 4.0400283985259815, "grad_norm": 3.6030538082122803, "learning_rate": 3.8127662152731615e-05, "loss": 0.0615, "step": 119500 }, { "epoch": 4.040366476216234, "grad_norm": 3.313015937805176, "learning_rate": 3.8119631407570724e-05, "loss": 0.0659, "step": 119510 }, { "epoch": 4.040704553906488, "grad_norm": 2.2041258811950684, "learning_rate": 3.811160098721462e-05, "loss": 0.0598, "step": 119520 }, { "epoch": 4.041042631596741, "grad_norm": 5.558505535125732, "learning_rate": 3.810357089188289e-05, "loss": 0.0693, "step": 119530 }, { "epoch": 4.041380709286994, "grad_norm": 5.899911403656006, "learning_rate": 3.8095541121795064e-05, "loss": 0.0636, "step": 119540 }, { "epoch": 4.041718786977247, "grad_norm": 3.7250401973724365, "learning_rate": 3.808751167717067e-05, "loss": 0.0591, "step": 119550 }, { "epoch": 4.042056864667501, "grad_norm": 11.357765197753906, "learning_rate": 3.8079482558229235e-05, "loss": 0.0706, "step": 119560 }, { "epoch": 4.042394942357753, "grad_norm": 2.9327287673950195, "learning_rate": 3.807145376519026e-05, "loss": 0.0605, "step": 119570 }, { "epoch": 4.042733020048007, "grad_norm": 4.426420211791992, "learning_rate": 3.806342529827327e-05, "loss": 0.0679, "step": 119580 }, { "epoch": 4.0430710977382605, "grad_norm": 4.084841251373291, "learning_rate": 3.8055397157697735e-05, "loss": 0.0671, "step": 119590 }, { "epoch": 4.043409175428513, "grad_norm": 1.7078293561935425, "learning_rate": 3.804736934368315e-05, "loss": 0.0747, "step": 119600 }, { "epoch": 4.043747253118767, "grad_norm": 3.503016471862793, "learning_rate": 3.8039341856449025e-05, "loss": 0.0637, "step": 119610 }, { "epoch": 4.04408533080902, "grad_norm": 2.28024959564209, "learning_rate": 3.803131469621477e-05, "loss": 0.0677, "step": 119620 }, { "epoch": 4.044423408499273, "grad_norm": 7.261044502258301, "learning_rate": 3.802328786319988e-05, "loss": 0.0655, "step": 119630 }, { "epoch": 4.044761486189526, "grad_norm": 5.298566818237305, "learning_rate": 3.801526135762381e-05, "loss": 0.0659, "step": 119640 }, { "epoch": 4.04509956387978, "grad_norm": 4.052220344543457, "learning_rate": 3.8007235179705995e-05, "loss": 0.0597, "step": 119650 }, { "epoch": 4.045437641570032, "grad_norm": 2.2130632400512695, "learning_rate": 3.799920932966586e-05, "loss": 0.0666, "step": 119660 }, { "epoch": 4.045775719260286, "grad_norm": 4.553446292877197, "learning_rate": 3.799118380772282e-05, "loss": 0.0681, "step": 119670 }, { "epoch": 4.046113796950539, "grad_norm": 3.5410091876983643, "learning_rate": 3.798315861409632e-05, "loss": 0.0653, "step": 119680 }, { "epoch": 4.046451874640792, "grad_norm": 3.784515857696533, "learning_rate": 3.797513374900575e-05, "loss": 0.0688, "step": 119690 }, { "epoch": 4.0467899523310455, "grad_norm": 6.722004413604736, "learning_rate": 3.796710921267051e-05, "loss": 0.067, "step": 119700 }, { "epoch": 4.047128030021299, "grad_norm": 5.91940975189209, "learning_rate": 3.795908500530996e-05, "loss": 0.0737, "step": 119710 }, { "epoch": 4.047466107711552, "grad_norm": 5.063194274902344, "learning_rate": 3.7951061127143504e-05, "loss": 0.0729, "step": 119720 }, { "epoch": 4.047804185401805, "grad_norm": 3.7844135761260986, "learning_rate": 3.794303757839053e-05, "loss": 0.0652, "step": 119730 }, { "epoch": 4.048142263092059, "grad_norm": 6.0152974128723145, "learning_rate": 3.793501435927037e-05, "loss": 0.0721, "step": 119740 }, { "epoch": 4.048480340782311, "grad_norm": 8.496665954589844, "learning_rate": 3.7926991470002394e-05, "loss": 0.0667, "step": 119750 }, { "epoch": 4.048818418472565, "grad_norm": 12.168375968933105, "learning_rate": 3.791896891080592e-05, "loss": 0.0689, "step": 119760 }, { "epoch": 4.049156496162818, "grad_norm": 9.048150062561035, "learning_rate": 3.791094668190032e-05, "loss": 0.0694, "step": 119770 }, { "epoch": 4.049494573853071, "grad_norm": 3.9232921600341797, "learning_rate": 3.790292478350487e-05, "loss": 0.0724, "step": 119780 }, { "epoch": 4.0498326515433245, "grad_norm": 3.7409045696258545, "learning_rate": 3.789490321583891e-05, "loss": 0.0737, "step": 119790 }, { "epoch": 4.050170729233578, "grad_norm": 30.46656608581543, "learning_rate": 3.788688197912178e-05, "loss": 0.0681, "step": 119800 }, { "epoch": 4.0505088069238315, "grad_norm": 7.249635219573975, "learning_rate": 3.7878861073572713e-05, "loss": 0.0747, "step": 119810 }, { "epoch": 4.050846884614084, "grad_norm": 21.552509307861328, "learning_rate": 3.7870840499411066e-05, "loss": 0.0768, "step": 119820 }, { "epoch": 4.051184962304338, "grad_norm": 6.010589599609375, "learning_rate": 3.786282025685605e-05, "loss": 0.0701, "step": 119830 }, { "epoch": 4.051523039994591, "grad_norm": 10.40203857421875, "learning_rate": 3.7854800346126985e-05, "loss": 0.0749, "step": 119840 }, { "epoch": 4.051861117684844, "grad_norm": 9.732069969177246, "learning_rate": 3.78467807674431e-05, "loss": 0.0699, "step": 119850 }, { "epoch": 4.052199195375097, "grad_norm": 4.505242347717285, "learning_rate": 3.7838761521023675e-05, "loss": 0.0661, "step": 119860 }, { "epoch": 4.052537273065351, "grad_norm": 3.505113363265991, "learning_rate": 3.7830742607087946e-05, "loss": 0.0759, "step": 119870 }, { "epoch": 4.0528753507556035, "grad_norm": 42.98637390136719, "learning_rate": 3.782272402585513e-05, "loss": 0.0731, "step": 119880 }, { "epoch": 4.053213428445857, "grad_norm": 3.9181859493255615, "learning_rate": 3.781470577754448e-05, "loss": 0.0696, "step": 119890 }, { "epoch": 4.0535515061361105, "grad_norm": 5.933002948760986, "learning_rate": 3.780668786237518e-05, "loss": 0.0774, "step": 119900 }, { "epoch": 4.053889583826363, "grad_norm": 4.168650150299072, "learning_rate": 3.779867028056645e-05, "loss": 0.067, "step": 119910 }, { "epoch": 4.054227661516617, "grad_norm": 23.509376525878906, "learning_rate": 3.779065303233751e-05, "loss": 0.0763, "step": 119920 }, { "epoch": 4.05456573920687, "grad_norm": 9.545059204101562, "learning_rate": 3.778263611790751e-05, "loss": 0.0712, "step": 119930 }, { "epoch": 4.054903816897123, "grad_norm": 16.262466430664062, "learning_rate": 3.777461953749567e-05, "loss": 0.0752, "step": 119940 }, { "epoch": 4.055241894587376, "grad_norm": 11.014688491821289, "learning_rate": 3.7766603291321126e-05, "loss": 0.0746, "step": 119950 }, { "epoch": 4.05557997227763, "grad_norm": 7.622959136962891, "learning_rate": 3.775858737960306e-05, "loss": 0.0736, "step": 119960 }, { "epoch": 4.055918049967882, "grad_norm": 10.921163558959961, "learning_rate": 3.7750571802560605e-05, "loss": 0.0753, "step": 119970 }, { "epoch": 4.056256127658136, "grad_norm": 23.589824676513672, "learning_rate": 3.7742556560412915e-05, "loss": 0.0759, "step": 119980 }, { "epoch": 4.056594205348389, "grad_norm": 17.563621520996094, "learning_rate": 3.773454165337914e-05, "loss": 0.0729, "step": 119990 }, { "epoch": 4.056932283038642, "grad_norm": 9.363299369812012, "learning_rate": 3.772652708167836e-05, "loss": 0.0817, "step": 120000 }, { "epoch": 4.057270360728896, "grad_norm": 17.082035064697266, "learning_rate": 3.7718512845529756e-05, "loss": 0.0728, "step": 120010 }, { "epoch": 4.057608438419149, "grad_norm": 3.6957943439483643, "learning_rate": 3.771049894515236e-05, "loss": 0.0718, "step": 120020 }, { "epoch": 4.057946516109402, "grad_norm": 20.763334274291992, "learning_rate": 3.7702485380765344e-05, "loss": 0.0785, "step": 120030 }, { "epoch": 4.058284593799655, "grad_norm": 6.788820743560791, "learning_rate": 3.769447215258772e-05, "loss": 0.0711, "step": 120040 }, { "epoch": 4.058622671489909, "grad_norm": 4.267161846160889, "learning_rate": 3.768645926083862e-05, "loss": 0.0696, "step": 120050 }, { "epoch": 4.058960749180161, "grad_norm": 6.039779186248779, "learning_rate": 3.7678446705737105e-05, "loss": 0.0651, "step": 120060 }, { "epoch": 4.059298826870415, "grad_norm": 10.751254081726074, "learning_rate": 3.7670434487502214e-05, "loss": 0.0723, "step": 120070 }, { "epoch": 4.059636904560668, "grad_norm": 7.34525728225708, "learning_rate": 3.766242260635303e-05, "loss": 0.0746, "step": 120080 }, { "epoch": 4.059974982250921, "grad_norm": 6.404257774353027, "learning_rate": 3.7654411062508565e-05, "loss": 0.0627, "step": 120090 }, { "epoch": 4.0603130599411745, "grad_norm": 6.031047344207764, "learning_rate": 3.7646399856187845e-05, "loss": 0.0681, "step": 120100 }, { "epoch": 4.060651137631428, "grad_norm": 9.379372596740723, "learning_rate": 3.763838898760995e-05, "loss": 0.061, "step": 120110 }, { "epoch": 4.060989215321681, "grad_norm": 7.706684112548828, "learning_rate": 3.763037845699382e-05, "loss": 0.0705, "step": 120120 }, { "epoch": 4.061327293011934, "grad_norm": 2.2218780517578125, "learning_rate": 3.762236826455853e-05, "loss": 0.0676, "step": 120130 }, { "epoch": 4.061665370702188, "grad_norm": 3.1481943130493164, "learning_rate": 3.761435841052302e-05, "loss": 0.0693, "step": 120140 }, { "epoch": 4.06200344839244, "grad_norm": 4.215840816497803, "learning_rate": 3.760634889510631e-05, "loss": 0.0721, "step": 120150 }, { "epoch": 4.062341526082694, "grad_norm": 3.3755509853363037, "learning_rate": 3.759833971852735e-05, "loss": 0.0708, "step": 120160 }, { "epoch": 4.062679603772947, "grad_norm": 31.72722625732422, "learning_rate": 3.759033088100513e-05, "loss": 0.0649, "step": 120170 }, { "epoch": 4.0630176814632, "grad_norm": 2.1113576889038086, "learning_rate": 3.758232238275862e-05, "loss": 0.0663, "step": 120180 }, { "epoch": 4.0633557591534535, "grad_norm": 5.249423980712891, "learning_rate": 3.757431422400672e-05, "loss": 0.0697, "step": 120190 }, { "epoch": 4.063693836843707, "grad_norm": 2.980125665664673, "learning_rate": 3.756630640496843e-05, "loss": 0.0647, "step": 120200 }, { "epoch": 4.06403191453396, "grad_norm": 9.941596031188965, "learning_rate": 3.755829892586262e-05, "loss": 0.0687, "step": 120210 }, { "epoch": 4.064369992224213, "grad_norm": 10.279169082641602, "learning_rate": 3.755029178690828e-05, "loss": 0.0662, "step": 120220 }, { "epoch": 4.064708069914467, "grad_norm": 4.324321746826172, "learning_rate": 3.754228498832425e-05, "loss": 0.0718, "step": 120230 }, { "epoch": 4.065046147604719, "grad_norm": 14.78425407409668, "learning_rate": 3.753427853032948e-05, "loss": 0.0666, "step": 120240 }, { "epoch": 4.065384225294973, "grad_norm": 2.7926249504089355, "learning_rate": 3.7526272413142865e-05, "loss": 0.0715, "step": 120250 }, { "epoch": 4.065722302985226, "grad_norm": 2.168031692504883, "learning_rate": 3.751826663698326e-05, "loss": 0.0647, "step": 120260 }, { "epoch": 4.066060380675479, "grad_norm": 2.839059591293335, "learning_rate": 3.7510261202069575e-05, "loss": 0.0683, "step": 120270 }, { "epoch": 4.066398458365732, "grad_norm": 12.21830940246582, "learning_rate": 3.750225610862064e-05, "loss": 0.0683, "step": 120280 }, { "epoch": 4.066736536055986, "grad_norm": 3.9460299015045166, "learning_rate": 3.7494251356855317e-05, "loss": 0.0614, "step": 120290 }, { "epoch": 4.067074613746239, "grad_norm": 5.46647834777832, "learning_rate": 3.74862469469925e-05, "loss": 0.0766, "step": 120300 }, { "epoch": 4.067412691436492, "grad_norm": 3.963517665863037, "learning_rate": 3.747824287925095e-05, "loss": 0.0643, "step": 120310 }, { "epoch": 4.067750769126746, "grad_norm": 4.976740837097168, "learning_rate": 3.7470239153849575e-05, "loss": 0.0707, "step": 120320 }, { "epoch": 4.068088846816998, "grad_norm": 2.85428524017334, "learning_rate": 3.7462235771007116e-05, "loss": 0.0686, "step": 120330 }, { "epoch": 4.068426924507252, "grad_norm": 3.6253068447113037, "learning_rate": 3.745423273094245e-05, "loss": 0.0653, "step": 120340 }, { "epoch": 4.068765002197505, "grad_norm": 7.369705677032471, "learning_rate": 3.744623003387433e-05, "loss": 0.0674, "step": 120350 }, { "epoch": 4.069103079887758, "grad_norm": 4.732378005981445, "learning_rate": 3.743822768002157e-05, "loss": 0.0685, "step": 120360 }, { "epoch": 4.069441157578011, "grad_norm": 11.154476165771484, "learning_rate": 3.743022566960295e-05, "loss": 0.0694, "step": 120370 }, { "epoch": 4.069779235268265, "grad_norm": 17.958154678344727, "learning_rate": 3.7422224002837236e-05, "loss": 0.0725, "step": 120380 }, { "epoch": 4.0701173129585175, "grad_norm": 2.259514570236206, "learning_rate": 3.74142226799432e-05, "loss": 0.0655, "step": 120390 }, { "epoch": 4.070455390648771, "grad_norm": 12.127079010009766, "learning_rate": 3.740622170113958e-05, "loss": 0.0626, "step": 120400 }, { "epoch": 4.070793468339025, "grad_norm": 7.753329753875732, "learning_rate": 3.739822106664513e-05, "loss": 0.0725, "step": 120410 }, { "epoch": 4.071131546029277, "grad_norm": 11.327922821044922, "learning_rate": 3.739022077667856e-05, "loss": 0.069, "step": 120420 }, { "epoch": 4.071469623719531, "grad_norm": 18.474353790283203, "learning_rate": 3.738222083145864e-05, "loss": 0.067, "step": 120430 }, { "epoch": 4.071807701409784, "grad_norm": 17.487812042236328, "learning_rate": 3.737422123120406e-05, "loss": 0.0688, "step": 120440 }, { "epoch": 4.072145779100037, "grad_norm": 6.097269058227539, "learning_rate": 3.736622197613352e-05, "loss": 0.0684, "step": 120450 }, { "epoch": 4.07248385679029, "grad_norm": 15.635619163513184, "learning_rate": 3.7358223066465734e-05, "loss": 0.0757, "step": 120460 }, { "epoch": 4.072821934480544, "grad_norm": 4.9031853675842285, "learning_rate": 3.735022450241938e-05, "loss": 0.07, "step": 120470 }, { "epoch": 4.0731600121707965, "grad_norm": 13.569079399108887, "learning_rate": 3.734222628421313e-05, "loss": 0.0624, "step": 120480 }, { "epoch": 4.07349808986105, "grad_norm": 14.319719314575195, "learning_rate": 3.733422841206566e-05, "loss": 0.0639, "step": 120490 }, { "epoch": 4.0738361675513035, "grad_norm": 25.684581756591797, "learning_rate": 3.732623088619562e-05, "loss": 0.0688, "step": 120500 }, { "epoch": 4.074174245241556, "grad_norm": 12.445049285888672, "learning_rate": 3.7318233706821694e-05, "loss": 0.0612, "step": 120510 }, { "epoch": 4.07451232293181, "grad_norm": 8.840681076049805, "learning_rate": 3.731023687416245e-05, "loss": 0.0675, "step": 120520 }, { "epoch": 4.074850400622063, "grad_norm": 11.052570343017578, "learning_rate": 3.730224038843661e-05, "loss": 0.0684, "step": 120530 }, { "epoch": 4.075188478312316, "grad_norm": 15.179387092590332, "learning_rate": 3.729424424986271e-05, "loss": 0.0601, "step": 120540 }, { "epoch": 4.075526556002569, "grad_norm": 2.9875991344451904, "learning_rate": 3.728624845865941e-05, "loss": 0.0692, "step": 120550 }, { "epoch": 4.075864633692823, "grad_norm": 8.041420936584473, "learning_rate": 3.727825301504531e-05, "loss": 0.0649, "step": 120560 }, { "epoch": 4.076202711383075, "grad_norm": 5.467111110687256, "learning_rate": 3.727025791923898e-05, "loss": 0.0705, "step": 120570 }, { "epoch": 4.076540789073329, "grad_norm": 1.195315957069397, "learning_rate": 3.7262263171459034e-05, "loss": 0.0607, "step": 120580 }, { "epoch": 4.0768788667635825, "grad_norm": 17.40740966796875, "learning_rate": 3.7254268771924014e-05, "loss": 0.0706, "step": 120590 }, { "epoch": 4.077216944453835, "grad_norm": 10.843074798583984, "learning_rate": 3.72462747208525e-05, "loss": 0.0671, "step": 120600 }, { "epoch": 4.077555022144089, "grad_norm": 2.8101963996887207, "learning_rate": 3.723828101846304e-05, "loss": 0.0688, "step": 120610 }, { "epoch": 4.077893099834342, "grad_norm": 2.8567166328430176, "learning_rate": 3.723028766497417e-05, "loss": 0.071, "step": 120620 }, { "epoch": 4.078231177524595, "grad_norm": 2.3025999069213867, "learning_rate": 3.7222294660604466e-05, "loss": 0.0618, "step": 120630 }, { "epoch": 4.078569255214848, "grad_norm": 2.736532688140869, "learning_rate": 3.721430200557241e-05, "loss": 0.0737, "step": 120640 }, { "epoch": 4.078907332905102, "grad_norm": 3.70971941947937, "learning_rate": 3.720630970009655e-05, "loss": 0.0628, "step": 120650 }, { "epoch": 4.079245410595354, "grad_norm": 9.616310119628906, "learning_rate": 3.719831774439536e-05, "loss": 0.0738, "step": 120660 }, { "epoch": 4.079583488285608, "grad_norm": 1.4607645273208618, "learning_rate": 3.719032613868736e-05, "loss": 0.0686, "step": 120670 }, { "epoch": 4.079921565975861, "grad_norm": 7.850796222686768, "learning_rate": 3.7182334883191026e-05, "loss": 0.0729, "step": 120680 }, { "epoch": 4.080259643666114, "grad_norm": 2.606698751449585, "learning_rate": 3.717434397812484e-05, "loss": 0.0667, "step": 120690 }, { "epoch": 4.080597721356368, "grad_norm": 7.1872172355651855, "learning_rate": 3.716635342370728e-05, "loss": 0.0723, "step": 120700 }, { "epoch": 4.080935799046621, "grad_norm": 3.3457441329956055, "learning_rate": 3.715836322015678e-05, "loss": 0.0728, "step": 120710 }, { "epoch": 4.081273876736875, "grad_norm": 5.778883934020996, "learning_rate": 3.7150373367691834e-05, "loss": 0.0724, "step": 120720 }, { "epoch": 4.081611954427127, "grad_norm": 68.29132843017578, "learning_rate": 3.714238386653082e-05, "loss": 0.0672, "step": 120730 }, { "epoch": 4.081950032117381, "grad_norm": 30.691219329833984, "learning_rate": 3.7134394716892215e-05, "loss": 0.069, "step": 120740 }, { "epoch": 4.082288109807633, "grad_norm": 6.042230129241943, "learning_rate": 3.7126405918994425e-05, "loss": 0.0691, "step": 120750 }, { "epoch": 4.082626187497887, "grad_norm": 5.385261535644531, "learning_rate": 3.711841747305586e-05, "loss": 0.0684, "step": 120760 }, { "epoch": 4.08296426518814, "grad_norm": 10.04489517211914, "learning_rate": 3.711042937929493e-05, "loss": 0.0763, "step": 120770 }, { "epoch": 4.083302342878394, "grad_norm": 2.9808125495910645, "learning_rate": 3.710244163793001e-05, "loss": 0.0622, "step": 120780 }, { "epoch": 4.0836404205686465, "grad_norm": 3.0722246170043945, "learning_rate": 3.70944542491795e-05, "loss": 0.065, "step": 120790 }, { "epoch": 4.0839784982589, "grad_norm": 11.937106132507324, "learning_rate": 3.708646721326175e-05, "loss": 0.0678, "step": 120800 }, { "epoch": 4.0843165759491535, "grad_norm": 1.9219428300857544, "learning_rate": 3.7078480530395124e-05, "loss": 0.065, "step": 120810 }, { "epoch": 4.084654653639406, "grad_norm": 13.391162872314453, "learning_rate": 3.7070494200798024e-05, "loss": 0.0679, "step": 120820 }, { "epoch": 4.08499273132966, "grad_norm": 3.056389093399048, "learning_rate": 3.706250822468871e-05, "loss": 0.0609, "step": 120830 }, { "epoch": 4.085330809019913, "grad_norm": 3.325279474258423, "learning_rate": 3.705452260228559e-05, "loss": 0.069, "step": 120840 }, { "epoch": 4.085668886710166, "grad_norm": 1.4603667259216309, "learning_rate": 3.704653733380695e-05, "loss": 0.0627, "step": 120850 }, { "epoch": 4.086006964400419, "grad_norm": 7.683762550354004, "learning_rate": 3.703855241947112e-05, "loss": 0.0625, "step": 120860 }, { "epoch": 4.086345042090673, "grad_norm": 4.215174674987793, "learning_rate": 3.703056785949639e-05, "loss": 0.0606, "step": 120870 }, { "epoch": 4.0866831197809255, "grad_norm": 5.469088077545166, "learning_rate": 3.702258365410106e-05, "loss": 0.0625, "step": 120880 }, { "epoch": 4.087021197471179, "grad_norm": 4.3143157958984375, "learning_rate": 3.7014599803503436e-05, "loss": 0.0635, "step": 120890 }, { "epoch": 4.0873592751614325, "grad_norm": 9.456747055053711, "learning_rate": 3.700661630792174e-05, "loss": 0.0656, "step": 120900 }, { "epoch": 4.087697352851685, "grad_norm": 6.253419876098633, "learning_rate": 3.6998633167574314e-05, "loss": 0.0654, "step": 120910 }, { "epoch": 4.088035430541939, "grad_norm": 8.275115966796875, "learning_rate": 3.6990650382679336e-05, "loss": 0.0703, "step": 120920 }, { "epoch": 4.088373508232192, "grad_norm": 7.265529632568359, "learning_rate": 3.6982667953455106e-05, "loss": 0.0656, "step": 120930 }, { "epoch": 4.088711585922445, "grad_norm": 2.6428768634796143, "learning_rate": 3.697468588011985e-05, "loss": 0.0609, "step": 120940 }, { "epoch": 4.089049663612698, "grad_norm": 3.217620611190796, "learning_rate": 3.696670416289177e-05, "loss": 0.0622, "step": 120950 }, { "epoch": 4.089387741302952, "grad_norm": 3.958710193634033, "learning_rate": 3.6958722801989134e-05, "loss": 0.062, "step": 120960 }, { "epoch": 4.089725818993204, "grad_norm": 2.836543083190918, "learning_rate": 3.695074179763009e-05, "loss": 0.067, "step": 120970 }, { "epoch": 4.090063896683458, "grad_norm": 3.942838668823242, "learning_rate": 3.694276115003288e-05, "loss": 0.0637, "step": 120980 }, { "epoch": 4.0904019743737114, "grad_norm": 4.814151287078857, "learning_rate": 3.693478085941565e-05, "loss": 0.0689, "step": 120990 }, { "epoch": 4.090740052063964, "grad_norm": 4.642256259918213, "learning_rate": 3.6926800925996605e-05, "loss": 0.0694, "step": 121000 }, { "epoch": 4.091078129754218, "grad_norm": 3.9944326877593994, "learning_rate": 3.6918821349993946e-05, "loss": 0.0651, "step": 121010 }, { "epoch": 4.091416207444471, "grad_norm": 2.8429312705993652, "learning_rate": 3.691084213162576e-05, "loss": 0.0688, "step": 121020 }, { "epoch": 4.091754285134724, "grad_norm": 5.842804908752441, "learning_rate": 3.690286327111026e-05, "loss": 0.0697, "step": 121030 }, { "epoch": 4.092092362824977, "grad_norm": 3.3241937160491943, "learning_rate": 3.689488476866554e-05, "loss": 0.0692, "step": 121040 }, { "epoch": 4.092430440515231, "grad_norm": 7.336774826049805, "learning_rate": 3.688690662450975e-05, "loss": 0.0679, "step": 121050 }, { "epoch": 4.092768518205483, "grad_norm": 5.3100128173828125, "learning_rate": 3.687892883886102e-05, "loss": 0.065, "step": 121060 }, { "epoch": 4.093106595895737, "grad_norm": 5.8931498527526855, "learning_rate": 3.6870951411937424e-05, "loss": 0.0708, "step": 121070 }, { "epoch": 4.09344467358599, "grad_norm": 2.5368895530700684, "learning_rate": 3.68629743439571e-05, "loss": 0.0718, "step": 121080 }, { "epoch": 4.093782751276243, "grad_norm": 4.266632556915283, "learning_rate": 3.685499763513812e-05, "loss": 0.0645, "step": 121090 }, { "epoch": 4.0941208289664965, "grad_norm": 13.927146911621094, "learning_rate": 3.684702128569857e-05, "loss": 0.0689, "step": 121100 }, { "epoch": 4.09445890665675, "grad_norm": 7.018749237060547, "learning_rate": 3.6839045295856486e-05, "loss": 0.0625, "step": 121110 }, { "epoch": 4.094796984347003, "grad_norm": 56.2640380859375, "learning_rate": 3.683106966582999e-05, "loss": 0.0627, "step": 121120 }, { "epoch": 4.095135062037256, "grad_norm": 194.4099884033203, "learning_rate": 3.68230943958371e-05, "loss": 0.0764, "step": 121130 }, { "epoch": 4.09547313972751, "grad_norm": 9.163331985473633, "learning_rate": 3.681511948609585e-05, "loss": 0.0679, "step": 121140 }, { "epoch": 4.095811217417762, "grad_norm": 10.263666152954102, "learning_rate": 3.680714493682429e-05, "loss": 0.0683, "step": 121150 }, { "epoch": 4.096149295108016, "grad_norm": 2.458411931991577, "learning_rate": 3.679917074824043e-05, "loss": 0.0715, "step": 121160 }, { "epoch": 4.096487372798269, "grad_norm": 9.103837013244629, "learning_rate": 3.6791196920562275e-05, "loss": 0.065, "step": 121170 }, { "epoch": 4.096825450488522, "grad_norm": 3.720478057861328, "learning_rate": 3.678322345400784e-05, "loss": 0.0683, "step": 121180 }, { "epoch": 4.0971635281787755, "grad_norm": 7.015164852142334, "learning_rate": 3.6775250348795086e-05, "loss": 0.0703, "step": 121190 }, { "epoch": 4.097501605869029, "grad_norm": 5.293720245361328, "learning_rate": 3.6767277605142055e-05, "loss": 0.0737, "step": 121200 }, { "epoch": 4.097839683559282, "grad_norm": 10.303234100341797, "learning_rate": 3.675930522326666e-05, "loss": 0.0737, "step": 121210 }, { "epoch": 4.098177761249535, "grad_norm": 4.5174055099487305, "learning_rate": 3.675133320338692e-05, "loss": 0.0713, "step": 121220 }, { "epoch": 4.098515838939789, "grad_norm": 2.967968225479126, "learning_rate": 3.674336154572071e-05, "loss": 0.0623, "step": 121230 }, { "epoch": 4.098853916630041, "grad_norm": 11.341300964355469, "learning_rate": 3.673539025048605e-05, "loss": 0.067, "step": 121240 }, { "epoch": 4.099191994320295, "grad_norm": 5.2723388671875, "learning_rate": 3.6727419317900805e-05, "loss": 0.0655, "step": 121250 }, { "epoch": 4.099530072010548, "grad_norm": 4.974119186401367, "learning_rate": 3.671944874818294e-05, "loss": 0.0689, "step": 121260 }, { "epoch": 4.099868149700801, "grad_norm": 9.477315902709961, "learning_rate": 3.671147854155037e-05, "loss": 0.0727, "step": 121270 }, { "epoch": 4.1002062273910544, "grad_norm": 2.673962354660034, "learning_rate": 3.670350869822097e-05, "loss": 0.0672, "step": 121280 }, { "epoch": 4.100544305081308, "grad_norm": 3.706502914428711, "learning_rate": 3.669553921841266e-05, "loss": 0.0767, "step": 121290 }, { "epoch": 4.100882382771561, "grad_norm": 2.758122682571411, "learning_rate": 3.668757010234329e-05, "loss": 0.0676, "step": 121300 }, { "epoch": 4.101220460461814, "grad_norm": 9.091469764709473, "learning_rate": 3.667960135023075e-05, "loss": 0.071, "step": 121310 }, { "epoch": 4.101558538152068, "grad_norm": 4.49688720703125, "learning_rate": 3.667163296229293e-05, "loss": 0.0787, "step": 121320 }, { "epoch": 4.10189661584232, "grad_norm": 11.134787559509277, "learning_rate": 3.666366493874764e-05, "loss": 0.069, "step": 121330 }, { "epoch": 4.102234693532574, "grad_norm": 5.7735443115234375, "learning_rate": 3.665569727981276e-05, "loss": 0.0704, "step": 121340 }, { "epoch": 4.102572771222827, "grad_norm": 18.4606876373291, "learning_rate": 3.664772998570608e-05, "loss": 0.0723, "step": 121350 }, { "epoch": 4.10291084891308, "grad_norm": 71.52198028564453, "learning_rate": 3.663976305664547e-05, "loss": 0.072, "step": 121360 }, { "epoch": 4.103248926603333, "grad_norm": 7.999277591705322, "learning_rate": 3.66317964928487e-05, "loss": 0.069, "step": 121370 }, { "epoch": 4.103587004293587, "grad_norm": 14.939945220947266, "learning_rate": 3.662383029453359e-05, "loss": 0.0668, "step": 121380 }, { "epoch": 4.1039250819838395, "grad_norm": 7.047635078430176, "learning_rate": 3.661586446191795e-05, "loss": 0.0667, "step": 121390 }, { "epoch": 4.104263159674093, "grad_norm": 4.459183692932129, "learning_rate": 3.6607898995219525e-05, "loss": 0.0694, "step": 121400 }, { "epoch": 4.104601237364347, "grad_norm": 7.719010829925537, "learning_rate": 3.659993389465614e-05, "loss": 0.0705, "step": 121410 }, { "epoch": 4.104939315054599, "grad_norm": 71.9461441040039, "learning_rate": 3.65919691604455e-05, "loss": 0.0628, "step": 121420 }, { "epoch": 4.105277392744853, "grad_norm": 18.062740325927734, "learning_rate": 3.6584004792805415e-05, "loss": 0.0624, "step": 121430 }, { "epoch": 4.105615470435106, "grad_norm": 6.2842841148376465, "learning_rate": 3.657604079195357e-05, "loss": 0.0644, "step": 121440 }, { "epoch": 4.105953548125359, "grad_norm": 7.694125175476074, "learning_rate": 3.656807715810773e-05, "loss": 0.0666, "step": 121450 }, { "epoch": 4.106291625815612, "grad_norm": 6.071134567260742, "learning_rate": 3.6560113891485634e-05, "loss": 0.0686, "step": 121460 }, { "epoch": 4.106629703505866, "grad_norm": 11.491506576538086, "learning_rate": 3.6552150992304954e-05, "loss": 0.0671, "step": 121470 }, { "epoch": 4.1069677811961185, "grad_norm": 5.492159366607666, "learning_rate": 3.654418846078343e-05, "loss": 0.0681, "step": 121480 }, { "epoch": 4.107305858886372, "grad_norm": 5.28079891204834, "learning_rate": 3.653622629713873e-05, "loss": 0.065, "step": 121490 }, { "epoch": 4.1076439365766255, "grad_norm": 3.3262171745300293, "learning_rate": 3.652826450158854e-05, "loss": 0.0602, "step": 121500 }, { "epoch": 4.107982014266878, "grad_norm": 6.153783798217773, "learning_rate": 3.652030307435056e-05, "loss": 0.0634, "step": 121510 }, { "epoch": 4.108320091957132, "grad_norm": 11.542695045471191, "learning_rate": 3.6512342015642396e-05, "loss": 0.0701, "step": 121520 }, { "epoch": 4.108658169647385, "grad_norm": 7.6814165115356445, "learning_rate": 3.650438132568176e-05, "loss": 0.0695, "step": 121530 }, { "epoch": 4.108996247337638, "grad_norm": 8.813420295715332, "learning_rate": 3.6496421004686256e-05, "loss": 0.068, "step": 121540 }, { "epoch": 4.109334325027891, "grad_norm": 38.968624114990234, "learning_rate": 3.648846105287354e-05, "loss": 0.0716, "step": 121550 }, { "epoch": 4.109672402718145, "grad_norm": 7.154969215393066, "learning_rate": 3.6480501470461215e-05, "loss": 0.068, "step": 121560 }, { "epoch": 4.1100104804083974, "grad_norm": 8.350213050842285, "learning_rate": 3.6472542257666896e-05, "loss": 0.0733, "step": 121570 }, { "epoch": 4.110348558098651, "grad_norm": 7.733239650726318, "learning_rate": 3.64645834147082e-05, "loss": 0.0729, "step": 121580 }, { "epoch": 4.1106866357889045, "grad_norm": 3.8557045459747314, "learning_rate": 3.645662494180271e-05, "loss": 0.0685, "step": 121590 }, { "epoch": 4.111024713479157, "grad_norm": 8.029726028442383, "learning_rate": 3.6448666839168e-05, "loss": 0.0702, "step": 121600 }, { "epoch": 4.111362791169411, "grad_norm": 2.5985920429229736, "learning_rate": 3.644070910702164e-05, "loss": 0.067, "step": 121610 }, { "epoch": 4.111700868859664, "grad_norm": 9.764434814453125, "learning_rate": 3.6432751745581216e-05, "loss": 0.064, "step": 121620 }, { "epoch": 4.112038946549917, "grad_norm": 7.946014881134033, "learning_rate": 3.642479475506424e-05, "loss": 0.0652, "step": 121630 }, { "epoch": 4.11237702424017, "grad_norm": 9.460617065429688, "learning_rate": 3.641683813568829e-05, "loss": 0.0646, "step": 121640 }, { "epoch": 4.112715101930424, "grad_norm": 12.464421272277832, "learning_rate": 3.640888188767089e-05, "loss": 0.0647, "step": 121650 }, { "epoch": 4.113053179620676, "grad_norm": 13.635054588317871, "learning_rate": 3.640092601122953e-05, "loss": 0.0693, "step": 121660 }, { "epoch": 4.11339125731093, "grad_norm": 4.142119884490967, "learning_rate": 3.639297050658177e-05, "loss": 0.0709, "step": 121670 }, { "epoch": 4.113729335001183, "grad_norm": 11.205939292907715, "learning_rate": 3.638501537394507e-05, "loss": 0.0642, "step": 121680 }, { "epoch": 4.114067412691437, "grad_norm": 5.886538028717041, "learning_rate": 3.637706061353692e-05, "loss": 0.0727, "step": 121690 }, { "epoch": 4.11440549038169, "grad_norm": 24.272594451904297, "learning_rate": 3.6369106225574856e-05, "loss": 0.0636, "step": 121700 }, { "epoch": 4.114743568071943, "grad_norm": 5.797219276428223, "learning_rate": 3.636115221027627e-05, "loss": 0.0708, "step": 121710 }, { "epoch": 4.115081645762196, "grad_norm": 4.227418422698975, "learning_rate": 3.6353198567858685e-05, "loss": 0.0668, "step": 121720 }, { "epoch": 4.115419723452449, "grad_norm": 5.216978073120117, "learning_rate": 3.63452452985395e-05, "loss": 0.071, "step": 121730 }, { "epoch": 4.115757801142703, "grad_norm": 34.62784957885742, "learning_rate": 3.633729240253619e-05, "loss": 0.0728, "step": 121740 }, { "epoch": 4.116095878832956, "grad_norm": 2.9426047801971436, "learning_rate": 3.632933988006617e-05, "loss": 0.0658, "step": 121750 }, { "epoch": 4.116433956523209, "grad_norm": 21.80902099609375, "learning_rate": 3.6321387731346866e-05, "loss": 0.0654, "step": 121760 }, { "epoch": 4.116772034213462, "grad_norm": 5.317505836486816, "learning_rate": 3.631343595659569e-05, "loss": 0.0711, "step": 121770 }, { "epoch": 4.117110111903716, "grad_norm": 22.69537353515625, "learning_rate": 3.630548455603002e-05, "loss": 0.0679, "step": 121780 }, { "epoch": 4.1174481895939685, "grad_norm": 6.432319641113281, "learning_rate": 3.629753352986727e-05, "loss": 0.0695, "step": 121790 }, { "epoch": 4.117786267284222, "grad_norm": 17.49761199951172, "learning_rate": 3.6289582878324794e-05, "loss": 0.0678, "step": 121800 }, { "epoch": 4.1181243449744755, "grad_norm": 7.483376502990723, "learning_rate": 3.628163260161997e-05, "loss": 0.0663, "step": 121810 }, { "epoch": 4.118462422664728, "grad_norm": 11.5238618850708, "learning_rate": 3.6273682699970154e-05, "loss": 0.0677, "step": 121820 }, { "epoch": 4.118800500354982, "grad_norm": 5.920005798339844, "learning_rate": 3.62657331735927e-05, "loss": 0.0706, "step": 121830 }, { "epoch": 4.119138578045235, "grad_norm": 5.544989585876465, "learning_rate": 3.6257784022704955e-05, "loss": 0.0687, "step": 121840 }, { "epoch": 4.119476655735488, "grad_norm": 5.321504592895508, "learning_rate": 3.624983524752422e-05, "loss": 0.074, "step": 121850 }, { "epoch": 4.119814733425741, "grad_norm": 22.590362548828125, "learning_rate": 3.624188684826783e-05, "loss": 0.0696, "step": 121860 }, { "epoch": 4.120152811115995, "grad_norm": 5.204611778259277, "learning_rate": 3.623393882515308e-05, "loss": 0.065, "step": 121870 }, { "epoch": 4.1204908888062475, "grad_norm": 6.995170593261719, "learning_rate": 3.622599117839728e-05, "loss": 0.0691, "step": 121880 }, { "epoch": 4.120828966496501, "grad_norm": 7.150847434997559, "learning_rate": 3.621804390821769e-05, "loss": 0.0675, "step": 121890 }, { "epoch": 4.1211670441867545, "grad_norm": 5.175670623779297, "learning_rate": 3.6210097014831594e-05, "loss": 0.0687, "step": 121900 }, { "epoch": 4.121505121877007, "grad_norm": 3.6265621185302734, "learning_rate": 3.62021504984563e-05, "loss": 0.0654, "step": 121910 }, { "epoch": 4.121843199567261, "grad_norm": 3.2013936042785645, "learning_rate": 3.6194204359308995e-05, "loss": 0.0606, "step": 121920 }, { "epoch": 4.122181277257514, "grad_norm": 4.021425724029541, "learning_rate": 3.618625859760699e-05, "loss": 0.0623, "step": 121930 }, { "epoch": 4.122519354947767, "grad_norm": 2.49306583404541, "learning_rate": 3.617831321356745e-05, "loss": 0.0645, "step": 121940 }, { "epoch": 4.12285743263802, "grad_norm": 3.8036251068115234, "learning_rate": 3.6170368207407646e-05, "loss": 0.0716, "step": 121950 }, { "epoch": 4.123195510328274, "grad_norm": 3.9274470806121826, "learning_rate": 3.616242357934479e-05, "loss": 0.0658, "step": 121960 }, { "epoch": 4.123533588018526, "grad_norm": 2.981123447418213, "learning_rate": 3.615447932959606e-05, "loss": 0.0645, "step": 121970 }, { "epoch": 4.12387166570878, "grad_norm": 4.1923112869262695, "learning_rate": 3.614653545837868e-05, "loss": 0.0753, "step": 121980 }, { "epoch": 4.1242097433990335, "grad_norm": 7.450328350067139, "learning_rate": 3.613859196590981e-05, "loss": 0.0652, "step": 121990 }, { "epoch": 4.124547821089286, "grad_norm": 2.3434317111968994, "learning_rate": 3.6130648852406627e-05, "loss": 0.071, "step": 122000 }, { "epoch": 4.12488589877954, "grad_norm": 4.515311241149902, "learning_rate": 3.612270611808629e-05, "loss": 0.0662, "step": 122010 }, { "epoch": 4.125223976469793, "grad_norm": 4.773885250091553, "learning_rate": 3.6114763763165946e-05, "loss": 0.0749, "step": 122020 }, { "epoch": 4.125562054160046, "grad_norm": 4.188049793243408, "learning_rate": 3.610682178786276e-05, "loss": 0.0688, "step": 122030 }, { "epoch": 4.125900131850299, "grad_norm": 8.116921424865723, "learning_rate": 3.609888019239384e-05, "loss": 0.0669, "step": 122040 }, { "epoch": 4.126238209540553, "grad_norm": 3.626807928085327, "learning_rate": 3.609093897697632e-05, "loss": 0.071, "step": 122050 }, { "epoch": 4.126576287230805, "grad_norm": 2.650235176086426, "learning_rate": 3.608299814182729e-05, "loss": 0.0771, "step": 122060 }, { "epoch": 4.126914364921059, "grad_norm": 3.5933949947357178, "learning_rate": 3.607505768716388e-05, "loss": 0.0698, "step": 122070 }, { "epoch": 4.127252442611312, "grad_norm": 1.5996057987213135, "learning_rate": 3.606711761320315e-05, "loss": 0.0597, "step": 122080 }, { "epoch": 4.127590520301565, "grad_norm": 2.053677797317505, "learning_rate": 3.605917792016219e-05, "loss": 0.0664, "step": 122090 }, { "epoch": 4.1279285979918185, "grad_norm": 3.1403989791870117, "learning_rate": 3.6051238608258075e-05, "loss": 0.0669, "step": 122100 }, { "epoch": 4.128266675682072, "grad_norm": 4.656165599822998, "learning_rate": 3.604329967770784e-05, "loss": 0.0663, "step": 122110 }, { "epoch": 4.128604753372325, "grad_norm": 3.5678908824920654, "learning_rate": 3.603536112872858e-05, "loss": 0.0687, "step": 122120 }, { "epoch": 4.128942831062578, "grad_norm": 1.8890624046325684, "learning_rate": 3.602742296153726e-05, "loss": 0.0624, "step": 122130 }, { "epoch": 4.129280908752832, "grad_norm": 9.77529239654541, "learning_rate": 3.601948517635097e-05, "loss": 0.0645, "step": 122140 }, { "epoch": 4.129618986443084, "grad_norm": 3.579512119293213, "learning_rate": 3.6011547773386706e-05, "loss": 0.0588, "step": 122150 }, { "epoch": 4.129957064133338, "grad_norm": 2.643087387084961, "learning_rate": 3.600361075286146e-05, "loss": 0.0649, "step": 122160 }, { "epoch": 4.130295141823591, "grad_norm": 5.073329925537109, "learning_rate": 3.599567411499225e-05, "loss": 0.0639, "step": 122170 }, { "epoch": 4.130633219513844, "grad_norm": 6.483582019805908, "learning_rate": 3.5987737859996043e-05, "loss": 0.0665, "step": 122180 }, { "epoch": 4.1309712972040975, "grad_norm": 2.9729552268981934, "learning_rate": 3.5979801988089825e-05, "loss": 0.0661, "step": 122190 }, { "epoch": 4.131309374894351, "grad_norm": 4.548957347869873, "learning_rate": 3.597186649949054e-05, "loss": 0.0679, "step": 122200 }, { "epoch": 4.131647452584604, "grad_norm": 3.6038365364074707, "learning_rate": 3.596393139441515e-05, "loss": 0.0626, "step": 122210 }, { "epoch": 4.131985530274857, "grad_norm": 3.5477242469787598, "learning_rate": 3.5955996673080625e-05, "loss": 0.0752, "step": 122220 }, { "epoch": 4.132323607965111, "grad_norm": 10.288694381713867, "learning_rate": 3.594806233570385e-05, "loss": 0.0641, "step": 122230 }, { "epoch": 4.132661685655363, "grad_norm": 3.5216615200042725, "learning_rate": 3.594012838250179e-05, "loss": 0.0626, "step": 122240 }, { "epoch": 4.132999763345617, "grad_norm": 4.783145904541016, "learning_rate": 3.593219481369134e-05, "loss": 0.0688, "step": 122250 }, { "epoch": 4.13333784103587, "grad_norm": 4.4519500732421875, "learning_rate": 3.592426162948939e-05, "loss": 0.0688, "step": 122260 }, { "epoch": 4.133675918726123, "grad_norm": 7.164364337921143, "learning_rate": 3.591632883011284e-05, "loss": 0.0651, "step": 122270 }, { "epoch": 4.1340139964163765, "grad_norm": 5.9820966720581055, "learning_rate": 3.590839641577857e-05, "loss": 0.0705, "step": 122280 }, { "epoch": 4.13435207410663, "grad_norm": 7.828163146972656, "learning_rate": 3.590046438670345e-05, "loss": 0.0626, "step": 122290 }, { "epoch": 4.134690151796883, "grad_norm": 3.9056129455566406, "learning_rate": 3.589253274310433e-05, "loss": 0.0651, "step": 122300 }, { "epoch": 4.135028229487136, "grad_norm": 4.512481212615967, "learning_rate": 3.588460148519808e-05, "loss": 0.0673, "step": 122310 }, { "epoch": 4.13536630717739, "grad_norm": 3.866457223892212, "learning_rate": 3.587667061320149e-05, "loss": 0.0625, "step": 122320 }, { "epoch": 4.135704384867642, "grad_norm": 10.408917427062988, "learning_rate": 3.586874012733144e-05, "loss": 0.0707, "step": 122330 }, { "epoch": 4.136042462557896, "grad_norm": 1.3931884765625, "learning_rate": 3.586081002780474e-05, "loss": 0.0672, "step": 122340 }, { "epoch": 4.136380540248149, "grad_norm": 4.259265422821045, "learning_rate": 3.585288031483816e-05, "loss": 0.0648, "step": 122350 }, { "epoch": 4.136718617938402, "grad_norm": 3.23447585105896, "learning_rate": 3.584495098864853e-05, "loss": 0.0698, "step": 122360 }, { "epoch": 4.137056695628655, "grad_norm": 1.431882381439209, "learning_rate": 3.583702204945262e-05, "loss": 0.0701, "step": 122370 }, { "epoch": 4.137394773318909, "grad_norm": 2.3986470699310303, "learning_rate": 3.582909349746721e-05, "loss": 0.0607, "step": 122380 }, { "epoch": 4.1377328510091615, "grad_norm": 2.4012832641601562, "learning_rate": 3.5821165332909054e-05, "loss": 0.0628, "step": 122390 }, { "epoch": 4.138070928699415, "grad_norm": 4.070293426513672, "learning_rate": 3.581323755599489e-05, "loss": 0.065, "step": 122400 }, { "epoch": 4.138409006389669, "grad_norm": 3.7110378742218018, "learning_rate": 3.580531016694153e-05, "loss": 0.061, "step": 122410 }, { "epoch": 4.138747084079921, "grad_norm": 11.27017879486084, "learning_rate": 3.5797383165965614e-05, "loss": 0.0672, "step": 122420 }, { "epoch": 4.139085161770175, "grad_norm": 3.0673508644104004, "learning_rate": 3.578945655328394e-05, "loss": 0.0658, "step": 122430 }, { "epoch": 4.139423239460428, "grad_norm": 6.474818706512451, "learning_rate": 3.578153032911315e-05, "loss": 0.0647, "step": 122440 }, { "epoch": 4.139761317150681, "grad_norm": 4.8519792556762695, "learning_rate": 3.577360449367001e-05, "loss": 0.0657, "step": 122450 }, { "epoch": 4.140099394840934, "grad_norm": 3.144373655319214, "learning_rate": 3.576567904717116e-05, "loss": 0.0645, "step": 122460 }, { "epoch": 4.140437472531188, "grad_norm": 3.309048652648926, "learning_rate": 3.575775398983331e-05, "loss": 0.0712, "step": 122470 }, { "epoch": 4.1407755502214405, "grad_norm": 2.535604238510132, "learning_rate": 3.574982932187311e-05, "loss": 0.0666, "step": 122480 }, { "epoch": 4.141113627911694, "grad_norm": 3.325059175491333, "learning_rate": 3.5741905043507215e-05, "loss": 0.0607, "step": 122490 }, { "epoch": 4.1414517056019475, "grad_norm": 3.2749199867248535, "learning_rate": 3.573398115495229e-05, "loss": 0.0666, "step": 122500 }, { "epoch": 4.1417897832922, "grad_norm": 2.637880325317383, "learning_rate": 3.572605765642496e-05, "loss": 0.0656, "step": 122510 }, { "epoch": 4.142127860982454, "grad_norm": 3.5027575492858887, "learning_rate": 3.571813454814183e-05, "loss": 0.0675, "step": 122520 }, { "epoch": 4.142465938672707, "grad_norm": 3.7763783931732178, "learning_rate": 3.571021183031956e-05, "loss": 0.0702, "step": 122530 }, { "epoch": 4.14280401636296, "grad_norm": 3.161378860473633, "learning_rate": 3.570228950317473e-05, "loss": 0.0609, "step": 122540 }, { "epoch": 4.143142094053213, "grad_norm": 2.2153432369232178, "learning_rate": 3.5694367566923934e-05, "loss": 0.0633, "step": 122550 }, { "epoch": 4.143480171743467, "grad_norm": 3.4786529541015625, "learning_rate": 3.568644602178374e-05, "loss": 0.0692, "step": 122560 }, { "epoch": 4.1438182494337195, "grad_norm": 3.402801275253296, "learning_rate": 3.5678524867970754e-05, "loss": 0.0639, "step": 122570 }, { "epoch": 4.144156327123973, "grad_norm": 1.73675537109375, "learning_rate": 3.56706041057015e-05, "loss": 0.0665, "step": 122580 }, { "epoch": 4.1444944048142265, "grad_norm": 2.032819986343384, "learning_rate": 3.5662683735192547e-05, "loss": 0.0669, "step": 122590 }, { "epoch": 4.144832482504479, "grad_norm": 2.795198678970337, "learning_rate": 3.565476375666045e-05, "loss": 0.0587, "step": 122600 }, { "epoch": 4.145170560194733, "grad_norm": 5.9076080322265625, "learning_rate": 3.56468441703217e-05, "loss": 0.0677, "step": 122610 }, { "epoch": 4.145508637884986, "grad_norm": 4.430482387542725, "learning_rate": 3.563892497639287e-05, "loss": 0.0674, "step": 122620 }, { "epoch": 4.145846715575239, "grad_norm": 4.406920433044434, "learning_rate": 3.5631006175090405e-05, "loss": 0.0603, "step": 122630 }, { "epoch": 4.146184793265492, "grad_norm": 3.170257806777954, "learning_rate": 3.562308776663086e-05, "loss": 0.0629, "step": 122640 }, { "epoch": 4.146522870955746, "grad_norm": 4.825927734375, "learning_rate": 3.561516975123067e-05, "loss": 0.0776, "step": 122650 }, { "epoch": 4.146860948645999, "grad_norm": 17.571334838867188, "learning_rate": 3.5607252129106336e-05, "loss": 0.0659, "step": 122660 }, { "epoch": 4.147199026336252, "grad_norm": 5.786108493804932, "learning_rate": 3.559933490047433e-05, "loss": 0.0636, "step": 122670 }, { "epoch": 4.147537104026505, "grad_norm": 4.309851169586182, "learning_rate": 3.55914180655511e-05, "loss": 0.0618, "step": 122680 }, { "epoch": 4.147875181716758, "grad_norm": 17.024839401245117, "learning_rate": 3.5583501624553095e-05, "loss": 0.0712, "step": 122690 }, { "epoch": 4.148213259407012, "grad_norm": 8.036165237426758, "learning_rate": 3.557558557769672e-05, "loss": 0.0661, "step": 122700 }, { "epoch": 4.148551337097265, "grad_norm": 8.682877540588379, "learning_rate": 3.556766992519842e-05, "loss": 0.0629, "step": 122710 }, { "epoch": 4.148889414787519, "grad_norm": 7.859280109405518, "learning_rate": 3.5559754667274625e-05, "loss": 0.0675, "step": 122720 }, { "epoch": 4.149227492477771, "grad_norm": 6.268778324127197, "learning_rate": 3.555183980414167e-05, "loss": 0.0641, "step": 122730 }, { "epoch": 4.149565570168025, "grad_norm": 5.360494136810303, "learning_rate": 3.5543925336016024e-05, "loss": 0.0668, "step": 122740 }, { "epoch": 4.149903647858278, "grad_norm": 12.41577434539795, "learning_rate": 3.553601126311402e-05, "loss": 0.0659, "step": 122750 }, { "epoch": 4.150241725548531, "grad_norm": 10.674849510192871, "learning_rate": 3.552809758565203e-05, "loss": 0.065, "step": 122760 }, { "epoch": 4.150579803238784, "grad_norm": 6.468790054321289, "learning_rate": 3.5520184303846414e-05, "loss": 0.0648, "step": 122770 }, { "epoch": 4.150917880929038, "grad_norm": 3.9257938861846924, "learning_rate": 3.551227141791352e-05, "loss": 0.0644, "step": 122780 }, { "epoch": 4.1512559586192905, "grad_norm": 21.51272964477539, "learning_rate": 3.550435892806969e-05, "loss": 0.063, "step": 122790 }, { "epoch": 4.151594036309544, "grad_norm": 5.400784015655518, "learning_rate": 3.5496446834531236e-05, "loss": 0.0555, "step": 122800 }, { "epoch": 4.151932113999798, "grad_norm": 7.210638046264648, "learning_rate": 3.548853513751448e-05, "loss": 0.071, "step": 122810 }, { "epoch": 4.15227019169005, "grad_norm": 3.5442054271698, "learning_rate": 3.548062383723571e-05, "loss": 0.0634, "step": 122820 }, { "epoch": 4.152608269380304, "grad_norm": 8.454010963439941, "learning_rate": 3.547271293391124e-05, "loss": 0.064, "step": 122830 }, { "epoch": 4.152946347070557, "grad_norm": 9.26883316040039, "learning_rate": 3.546480242775733e-05, "loss": 0.0676, "step": 122840 }, { "epoch": 4.15328442476081, "grad_norm": 3.5750207901000977, "learning_rate": 3.5456892318990264e-05, "loss": 0.0652, "step": 122850 }, { "epoch": 4.153622502451063, "grad_norm": 14.132333755493164, "learning_rate": 3.544898260782631e-05, "loss": 0.0744, "step": 122860 }, { "epoch": 4.153960580141317, "grad_norm": 3.7149319648742676, "learning_rate": 3.54410732944817e-05, "loss": 0.0604, "step": 122870 }, { "epoch": 4.1542986578315695, "grad_norm": 4.137649059295654, "learning_rate": 3.543316437917268e-05, "loss": 0.0687, "step": 122880 }, { "epoch": 4.154636735521823, "grad_norm": 3.9652600288391113, "learning_rate": 3.5425255862115465e-05, "loss": 0.0596, "step": 122890 }, { "epoch": 4.1549748132120765, "grad_norm": 4.314509391784668, "learning_rate": 3.541734774352627e-05, "loss": 0.0641, "step": 122900 }, { "epoch": 4.155312890902329, "grad_norm": 5.439142227172852, "learning_rate": 3.540944002362134e-05, "loss": 0.062, "step": 122910 }, { "epoch": 4.155650968592583, "grad_norm": 2.256227731704712, "learning_rate": 3.54015327026168e-05, "loss": 0.0744, "step": 122920 }, { "epoch": 4.155989046282836, "grad_norm": 1.75209379196167, "learning_rate": 3.539362578072891e-05, "loss": 0.0611, "step": 122930 }, { "epoch": 4.156327123973089, "grad_norm": 3.926325798034668, "learning_rate": 3.538571925817378e-05, "loss": 0.0637, "step": 122940 }, { "epoch": 4.156665201663342, "grad_norm": 4.211982727050781, "learning_rate": 3.53778131351676e-05, "loss": 0.072, "step": 122950 }, { "epoch": 4.157003279353596, "grad_norm": 4.17979097366333, "learning_rate": 3.5369907411926516e-05, "loss": 0.0625, "step": 122960 }, { "epoch": 4.157341357043848, "grad_norm": 3.556828260421753, "learning_rate": 3.536200208866666e-05, "loss": 0.0679, "step": 122970 }, { "epoch": 4.157679434734102, "grad_norm": 4.4578423500061035, "learning_rate": 3.535409716560418e-05, "loss": 0.0654, "step": 122980 }, { "epoch": 4.1580175124243555, "grad_norm": 5.860812187194824, "learning_rate": 3.534619264295518e-05, "loss": 0.0575, "step": 122990 }, { "epoch": 4.158355590114608, "grad_norm": 72.4918212890625, "learning_rate": 3.5338288520935766e-05, "loss": 0.063, "step": 123000 }, { "epoch": 4.158693667804862, "grad_norm": 9.643309593200684, "learning_rate": 3.533038479976203e-05, "loss": 0.0726, "step": 123010 }, { "epoch": 4.159031745495115, "grad_norm": 3.4660484790802, "learning_rate": 3.532248147965007e-05, "loss": 0.0735, "step": 123020 }, { "epoch": 4.159369823185368, "grad_norm": 4.701113224029541, "learning_rate": 3.5314578560815936e-05, "loss": 0.0684, "step": 123030 }, { "epoch": 4.159707900875621, "grad_norm": 2.9442615509033203, "learning_rate": 3.5306676043475714e-05, "loss": 0.0671, "step": 123040 }, { "epoch": 4.160045978565875, "grad_norm": 2.0644960403442383, "learning_rate": 3.529877392784546e-05, "loss": 0.0595, "step": 123050 }, { "epoch": 4.160384056256127, "grad_norm": 3.2091524600982666, "learning_rate": 3.52908722141412e-05, "loss": 0.0705, "step": 123060 }, { "epoch": 4.160722133946381, "grad_norm": 4.54995059967041, "learning_rate": 3.528297090257897e-05, "loss": 0.0632, "step": 123070 }, { "epoch": 4.161060211636634, "grad_norm": 2.5825538635253906, "learning_rate": 3.527506999337479e-05, "loss": 0.0607, "step": 123080 }, { "epoch": 4.161398289326887, "grad_norm": 1.7636291980743408, "learning_rate": 3.526716948674465e-05, "loss": 0.064, "step": 123090 }, { "epoch": 4.161736367017141, "grad_norm": 4.082272052764893, "learning_rate": 3.525926938290457e-05, "loss": 0.0606, "step": 123100 }, { "epoch": 4.162074444707394, "grad_norm": 2.70182204246521, "learning_rate": 3.525136968207051e-05, "loss": 0.0662, "step": 123110 }, { "epoch": 4.162412522397647, "grad_norm": 3.47428822517395, "learning_rate": 3.52434703844585e-05, "loss": 0.0717, "step": 123120 }, { "epoch": 4.1627506000879, "grad_norm": 11.109970092773438, "learning_rate": 3.523557149028443e-05, "loss": 0.0709, "step": 123130 }, { "epoch": 4.163088677778154, "grad_norm": 8.452210426330566, "learning_rate": 3.5227672999764316e-05, "loss": 0.0635, "step": 123140 }, { "epoch": 4.163426755468406, "grad_norm": 1.2725841999053955, "learning_rate": 3.5219774913114046e-05, "loss": 0.059, "step": 123150 }, { "epoch": 4.16376483315866, "grad_norm": 7.945720195770264, "learning_rate": 3.5211877230549586e-05, "loss": 0.0621, "step": 123160 }, { "epoch": 4.164102910848913, "grad_norm": 96.47887420654297, "learning_rate": 3.520397995228686e-05, "loss": 0.0676, "step": 123170 }, { "epoch": 4.164440988539166, "grad_norm": 2.606004238128662, "learning_rate": 3.5196083078541754e-05, "loss": 0.0612, "step": 123180 }, { "epoch": 4.1647790662294195, "grad_norm": 5.034487724304199, "learning_rate": 3.5188186609530186e-05, "loss": 0.0667, "step": 123190 }, { "epoch": 4.165117143919673, "grad_norm": 10.083173751831055, "learning_rate": 3.5180290545468014e-05, "loss": 0.0575, "step": 123200 }, { "epoch": 4.165455221609926, "grad_norm": 6.707938194274902, "learning_rate": 3.517239488657115e-05, "loss": 0.0601, "step": 123210 }, { "epoch": 4.165793299300179, "grad_norm": 2.0577595233917236, "learning_rate": 3.516449963305541e-05, "loss": 0.0616, "step": 123220 }, { "epoch": 4.166131376990433, "grad_norm": 2.004951238632202, "learning_rate": 3.5156604785136685e-05, "loss": 0.0622, "step": 123230 }, { "epoch": 4.166469454680685, "grad_norm": 47.889400482177734, "learning_rate": 3.5148710343030825e-05, "loss": 0.0685, "step": 123240 }, { "epoch": 4.166807532370939, "grad_norm": 3.0329430103302, "learning_rate": 3.514081630695364e-05, "loss": 0.061, "step": 123250 }, { "epoch": 4.167145610061192, "grad_norm": 3.304661273956299, "learning_rate": 3.513292267712096e-05, "loss": 0.0607, "step": 123260 }, { "epoch": 4.167483687751445, "grad_norm": 1.715553641319275, "learning_rate": 3.5125029453748576e-05, "loss": 0.0718, "step": 123270 }, { "epoch": 4.1678217654416985, "grad_norm": 3.4028759002685547, "learning_rate": 3.511713663705231e-05, "loss": 0.0634, "step": 123280 }, { "epoch": 4.168159843131952, "grad_norm": 4.816648960113525, "learning_rate": 3.510924422724793e-05, "loss": 0.0631, "step": 123290 }, { "epoch": 4.168497920822205, "grad_norm": 2.581334352493286, "learning_rate": 3.510135222455121e-05, "loss": 0.0669, "step": 123300 }, { "epoch": 4.168835998512458, "grad_norm": 2.5891342163085938, "learning_rate": 3.5093460629177946e-05, "loss": 0.0696, "step": 123310 }, { "epoch": 4.169174076202712, "grad_norm": 2.5754244327545166, "learning_rate": 3.508556944134384e-05, "loss": 0.0576, "step": 123320 }, { "epoch": 4.169512153892964, "grad_norm": 2.9470138549804688, "learning_rate": 3.507767866126469e-05, "loss": 0.0807, "step": 123330 }, { "epoch": 4.169850231583218, "grad_norm": 3.1335277557373047, "learning_rate": 3.506978828915617e-05, "loss": 0.0679, "step": 123340 }, { "epoch": 4.170188309273471, "grad_norm": 4.359880447387695, "learning_rate": 3.5061898325234046e-05, "loss": 0.0619, "step": 123350 }, { "epoch": 4.170526386963724, "grad_norm": 2.4281699657440186, "learning_rate": 3.505400876971402e-05, "loss": 0.0675, "step": 123360 }, { "epoch": 4.170864464653977, "grad_norm": 8.298009872436523, "learning_rate": 3.5046119622811754e-05, "loss": 0.062, "step": 123370 }, { "epoch": 4.171202542344231, "grad_norm": 2.2317192554473877, "learning_rate": 3.503823088474299e-05, "loss": 0.0672, "step": 123380 }, { "epoch": 4.171540620034484, "grad_norm": 4.110592842102051, "learning_rate": 3.503034255572335e-05, "loss": 0.0669, "step": 123390 }, { "epoch": 4.171878697724737, "grad_norm": 9.680811882019043, "learning_rate": 3.502245463596854e-05, "loss": 0.0638, "step": 123400 }, { "epoch": 4.172216775414991, "grad_norm": 3.399864673614502, "learning_rate": 3.501456712569418e-05, "loss": 0.0631, "step": 123410 }, { "epoch": 4.172554853105243, "grad_norm": 3.262510061264038, "learning_rate": 3.5006680025115906e-05, "loss": 0.0638, "step": 123420 }, { "epoch": 4.172892930795497, "grad_norm": 4.184964656829834, "learning_rate": 3.499879333444941e-05, "loss": 0.0661, "step": 123430 }, { "epoch": 4.17323100848575, "grad_norm": 7.373373985290527, "learning_rate": 3.499090705391023e-05, "loss": 0.0637, "step": 123440 }, { "epoch": 4.173569086176003, "grad_norm": 4.616551876068115, "learning_rate": 3.498302118371403e-05, "loss": 0.066, "step": 123450 }, { "epoch": 4.173907163866256, "grad_norm": 5.893165111541748, "learning_rate": 3.497513572407639e-05, "loss": 0.0645, "step": 123460 }, { "epoch": 4.17424524155651, "grad_norm": 6.429777145385742, "learning_rate": 3.49672506752129e-05, "loss": 0.0609, "step": 123470 }, { "epoch": 4.1745833192467625, "grad_norm": 4.381947040557861, "learning_rate": 3.495936603733911e-05, "loss": 0.0667, "step": 123480 }, { "epoch": 4.174921396937016, "grad_norm": 7.830214023590088, "learning_rate": 3.4951481810670616e-05, "loss": 0.0621, "step": 123490 }, { "epoch": 4.1752594746272695, "grad_norm": 5.3676958084106445, "learning_rate": 3.494359799542296e-05, "loss": 0.0654, "step": 123500 }, { "epoch": 4.175597552317522, "grad_norm": 7.353477954864502, "learning_rate": 3.493571459181167e-05, "loss": 0.0676, "step": 123510 }, { "epoch": 4.175935630007776, "grad_norm": 4.630681037902832, "learning_rate": 3.492783160005229e-05, "loss": 0.07, "step": 123520 }, { "epoch": 4.176273707698029, "grad_norm": 2.1878559589385986, "learning_rate": 3.491994902036031e-05, "loss": 0.0616, "step": 123530 }, { "epoch": 4.176611785388282, "grad_norm": 8.670875549316406, "learning_rate": 3.4912066852951276e-05, "loss": 0.0635, "step": 123540 }, { "epoch": 4.176949863078535, "grad_norm": 2.7878212928771973, "learning_rate": 3.490418509804068e-05, "loss": 0.0619, "step": 123550 }, { "epoch": 4.177287940768789, "grad_norm": 4.078543186187744, "learning_rate": 3.489630375584397e-05, "loss": 0.0629, "step": 123560 }, { "epoch": 4.1776260184590415, "grad_norm": 2.160358428955078, "learning_rate": 3.488842282657665e-05, "loss": 0.0641, "step": 123570 }, { "epoch": 4.177964096149295, "grad_norm": 3.0736072063446045, "learning_rate": 3.4880542310454164e-05, "loss": 0.0627, "step": 123580 }, { "epoch": 4.1783021738395485, "grad_norm": 2.1627390384674072, "learning_rate": 3.487266220769199e-05, "loss": 0.0617, "step": 123590 }, { "epoch": 4.178640251529801, "grad_norm": 7.504239559173584, "learning_rate": 3.4864782518505536e-05, "loss": 0.0655, "step": 123600 }, { "epoch": 4.178978329220055, "grad_norm": 3.892869472503662, "learning_rate": 3.485690324311023e-05, "loss": 0.0637, "step": 123610 }, { "epoch": 4.179316406910308, "grad_norm": 6.444902420043945, "learning_rate": 3.484902438172153e-05, "loss": 0.0645, "step": 123620 }, { "epoch": 4.179654484600562, "grad_norm": 3.1083967685699463, "learning_rate": 3.484114593455478e-05, "loss": 0.0666, "step": 123630 }, { "epoch": 4.179992562290814, "grad_norm": 1.9602530002593994, "learning_rate": 3.483326790182544e-05, "loss": 0.0609, "step": 123640 }, { "epoch": 4.180330639981068, "grad_norm": 2.8241541385650635, "learning_rate": 3.482539028374882e-05, "loss": 0.0717, "step": 123650 }, { "epoch": 4.18066871767132, "grad_norm": 11.372855186462402, "learning_rate": 3.481751308054036e-05, "loss": 0.0634, "step": 123660 }, { "epoch": 4.181006795361574, "grad_norm": 4.486420631408691, "learning_rate": 3.480963629241537e-05, "loss": 0.0628, "step": 123670 }, { "epoch": 4.1813448730518274, "grad_norm": 4.806526184082031, "learning_rate": 3.480175991958922e-05, "loss": 0.0664, "step": 123680 }, { "epoch": 4.181682950742081, "grad_norm": 2.393019914627075, "learning_rate": 3.479388396227725e-05, "loss": 0.0665, "step": 123690 }, { "epoch": 4.182021028432334, "grad_norm": 58.80308151245117, "learning_rate": 3.4786008420694774e-05, "loss": 0.0639, "step": 123700 }, { "epoch": 4.182359106122587, "grad_norm": 5.092795372009277, "learning_rate": 3.477813329505711e-05, "loss": 0.0651, "step": 123710 }, { "epoch": 4.182697183812841, "grad_norm": 3.7337024211883545, "learning_rate": 3.477025858557957e-05, "loss": 0.0664, "step": 123720 }, { "epoch": 4.183035261503093, "grad_norm": 4.94437313079834, "learning_rate": 3.476238429247742e-05, "loss": 0.0674, "step": 123730 }, { "epoch": 4.183373339193347, "grad_norm": 12.61115837097168, "learning_rate": 3.475451041596598e-05, "loss": 0.069, "step": 123740 }, { "epoch": 4.1837114168836, "grad_norm": 13.476776123046875, "learning_rate": 3.474663695626049e-05, "loss": 0.0625, "step": 123750 }, { "epoch": 4.184049494573853, "grad_norm": 3.1291885375976562, "learning_rate": 3.473876391357623e-05, "loss": 0.0641, "step": 123760 }, { "epoch": 4.184387572264106, "grad_norm": 5.468232154846191, "learning_rate": 3.473089128812841e-05, "loss": 0.0587, "step": 123770 }, { "epoch": 4.18472564995436, "grad_norm": 3.669523000717163, "learning_rate": 3.47230190801323e-05, "loss": 0.0607, "step": 123780 }, { "epoch": 4.1850637276446125, "grad_norm": 5.238617897033691, "learning_rate": 3.4715147289803105e-05, "loss": 0.0655, "step": 123790 }, { "epoch": 4.185401805334866, "grad_norm": 4.252004623413086, "learning_rate": 3.470727591735604e-05, "loss": 0.0652, "step": 123800 }, { "epoch": 4.18573988302512, "grad_norm": 3.468120813369751, "learning_rate": 3.469940496300631e-05, "loss": 0.0627, "step": 123810 }, { "epoch": 4.186077960715372, "grad_norm": 3.1499156951904297, "learning_rate": 3.469153442696908e-05, "loss": 0.0677, "step": 123820 }, { "epoch": 4.186416038405626, "grad_norm": 4.696857452392578, "learning_rate": 3.468366430945959e-05, "loss": 0.0686, "step": 123830 }, { "epoch": 4.186754116095879, "grad_norm": 2.016944646835327, "learning_rate": 3.4675794610692914e-05, "loss": 0.0647, "step": 123840 }, { "epoch": 4.187092193786132, "grad_norm": 2.3450417518615723, "learning_rate": 3.4667925330884296e-05, "loss": 0.0679, "step": 123850 }, { "epoch": 4.187430271476385, "grad_norm": 1.4535560607910156, "learning_rate": 3.4660056470248816e-05, "loss": 0.0694, "step": 123860 }, { "epoch": 4.187768349166639, "grad_norm": 3.004603147506714, "learning_rate": 3.465218802900163e-05, "loss": 0.07, "step": 123870 }, { "epoch": 4.1881064268568915, "grad_norm": 3.454340696334839, "learning_rate": 3.464432000735787e-05, "loss": 0.0598, "step": 123880 }, { "epoch": 4.188444504547145, "grad_norm": 5.048361301422119, "learning_rate": 3.4636452405532624e-05, "loss": 0.0671, "step": 123890 }, { "epoch": 4.1887825822373985, "grad_norm": 2.645996570587158, "learning_rate": 3.4628585223741e-05, "loss": 0.0628, "step": 123900 }, { "epoch": 4.189120659927651, "grad_norm": 2.6349329948425293, "learning_rate": 3.4620718462198084e-05, "loss": 0.0591, "step": 123910 }, { "epoch": 4.189458737617905, "grad_norm": 2.3438379764556885, "learning_rate": 3.4612852121118926e-05, "loss": 0.0695, "step": 123920 }, { "epoch": 4.189796815308158, "grad_norm": 3.9579665660858154, "learning_rate": 3.4604986200718646e-05, "loss": 0.0664, "step": 123930 }, { "epoch": 4.190134892998411, "grad_norm": 2.924797773361206, "learning_rate": 3.459712070121223e-05, "loss": 0.0682, "step": 123940 }, { "epoch": 4.190472970688664, "grad_norm": 3.28920316696167, "learning_rate": 3.458925562281476e-05, "loss": 0.0671, "step": 123950 }, { "epoch": 4.190811048378918, "grad_norm": 4.8334150314331055, "learning_rate": 3.4581390965741254e-05, "loss": 0.0683, "step": 123960 }, { "epoch": 4.1911491260691704, "grad_norm": 3.152031421661377, "learning_rate": 3.457352673020673e-05, "loss": 0.0668, "step": 123970 }, { "epoch": 4.191487203759424, "grad_norm": 2.596306562423706, "learning_rate": 3.456566291642618e-05, "loss": 0.0662, "step": 123980 }, { "epoch": 4.1918252814496775, "grad_norm": 3.9898171424865723, "learning_rate": 3.4557799524614606e-05, "loss": 0.0668, "step": 123990 }, { "epoch": 4.19216335913993, "grad_norm": 2.2245919704437256, "learning_rate": 3.4549936554987e-05, "loss": 0.0647, "step": 124000 }, { "epoch": 4.192501436830184, "grad_norm": 1.5641281604766846, "learning_rate": 3.454207400775832e-05, "loss": 0.0705, "step": 124010 }, { "epoch": 4.192839514520437, "grad_norm": 4.193557262420654, "learning_rate": 3.453421188314354e-05, "loss": 0.0647, "step": 124020 }, { "epoch": 4.19317759221069, "grad_norm": 1.8342176675796509, "learning_rate": 3.452635018135757e-05, "loss": 0.0607, "step": 124030 }, { "epoch": 4.193515669900943, "grad_norm": 5.506082057952881, "learning_rate": 3.451848890261541e-05, "loss": 0.064, "step": 124040 }, { "epoch": 4.193853747591197, "grad_norm": 2.795487403869629, "learning_rate": 3.451062804713191e-05, "loss": 0.0674, "step": 124050 }, { "epoch": 4.194191825281449, "grad_norm": 6.398024082183838, "learning_rate": 3.4502767615122036e-05, "loss": 0.0673, "step": 124060 }, { "epoch": 4.194529902971703, "grad_norm": 10.639300346374512, "learning_rate": 3.449490760680068e-05, "loss": 0.0687, "step": 124070 }, { "epoch": 4.194867980661956, "grad_norm": 4.048648357391357, "learning_rate": 3.448704802238272e-05, "loss": 0.0651, "step": 124080 }, { "epoch": 4.195206058352209, "grad_norm": 3.701655626296997, "learning_rate": 3.447918886208304e-05, "loss": 0.0713, "step": 124090 }, { "epoch": 4.195544136042463, "grad_norm": 2.6915934085845947, "learning_rate": 3.44713301261165e-05, "loss": 0.0683, "step": 124100 }, { "epoch": 4.195882213732716, "grad_norm": 3.62479305267334, "learning_rate": 3.446347181469796e-05, "loss": 0.0627, "step": 124110 }, { "epoch": 4.196220291422969, "grad_norm": 106.46134948730469, "learning_rate": 3.445561392804228e-05, "loss": 0.0634, "step": 124120 }, { "epoch": 4.196558369113222, "grad_norm": 9.358619689941406, "learning_rate": 3.444775646636425e-05, "loss": 0.0657, "step": 124130 }, { "epoch": 4.196896446803476, "grad_norm": 9.62603759765625, "learning_rate": 3.443989942987875e-05, "loss": 0.0621, "step": 124140 }, { "epoch": 4.197234524493728, "grad_norm": 12.84558391571045, "learning_rate": 3.443204281880051e-05, "loss": 0.0684, "step": 124150 }, { "epoch": 4.197572602183982, "grad_norm": 45.56466293334961, "learning_rate": 3.442418663334441e-05, "loss": 0.0607, "step": 124160 }, { "epoch": 4.197910679874235, "grad_norm": 15.936711311340332, "learning_rate": 3.441633087372518e-05, "loss": 0.0579, "step": 124170 }, { "epoch": 4.198248757564488, "grad_norm": 2.9167375564575195, "learning_rate": 3.4408475540157606e-05, "loss": 0.0608, "step": 124180 }, { "epoch": 4.1985868352547415, "grad_norm": 4.8826751708984375, "learning_rate": 3.4400620632856464e-05, "loss": 0.0628, "step": 124190 }, { "epoch": 4.198924912944995, "grad_norm": 4.950031280517578, "learning_rate": 3.439276615203648e-05, "loss": 0.0643, "step": 124200 }, { "epoch": 4.199262990635248, "grad_norm": 5.93264627456665, "learning_rate": 3.438491209791242e-05, "loss": 0.0676, "step": 124210 }, { "epoch": 4.199601068325501, "grad_norm": 13.321993827819824, "learning_rate": 3.4377058470698985e-05, "loss": 0.0583, "step": 124220 }, { "epoch": 4.199939146015755, "grad_norm": 10.88984489440918, "learning_rate": 3.436920527061091e-05, "loss": 0.0576, "step": 124230 }, { "epoch": 4.200277223706007, "grad_norm": 3.466015338897705, "learning_rate": 3.4361352497862874e-05, "loss": 0.0697, "step": 124240 }, { "epoch": 4.200615301396261, "grad_norm": 12.589377403259277, "learning_rate": 3.4353500152669593e-05, "loss": 0.0686, "step": 124250 }, { "epoch": 4.200953379086514, "grad_norm": 4.132269382476807, "learning_rate": 3.434564823524575e-05, "loss": 0.0683, "step": 124260 }, { "epoch": 4.201291456776767, "grad_norm": 7.785521507263184, "learning_rate": 3.4337796745806e-05, "loss": 0.059, "step": 124270 }, { "epoch": 4.2016295344670205, "grad_norm": 4.374282360076904, "learning_rate": 3.432994568456501e-05, "loss": 0.0597, "step": 124280 }, { "epoch": 4.201967612157274, "grad_norm": 14.02731990814209, "learning_rate": 3.4322095051737404e-05, "loss": 0.069, "step": 124290 }, { "epoch": 4.202305689847527, "grad_norm": 2.959494113922119, "learning_rate": 3.431424484753784e-05, "loss": 0.0708, "step": 124300 }, { "epoch": 4.20264376753778, "grad_norm": 5.229691028594971, "learning_rate": 3.4306395072180916e-05, "loss": 0.0722, "step": 124310 }, { "epoch": 4.202981845228034, "grad_norm": 4.92301082611084, "learning_rate": 3.4298545725881256e-05, "loss": 0.0631, "step": 124320 }, { "epoch": 4.203319922918286, "grad_norm": 2.287644624710083, "learning_rate": 3.429069680885348e-05, "loss": 0.069, "step": 124330 }, { "epoch": 4.20365800060854, "grad_norm": 3.836034059524536, "learning_rate": 3.4282848321312124e-05, "loss": 0.0647, "step": 124340 }, { "epoch": 4.203996078298793, "grad_norm": 7.970068454742432, "learning_rate": 3.4275000263471814e-05, "loss": 0.0706, "step": 124350 }, { "epoch": 4.204334155989046, "grad_norm": 3.718967914581299, "learning_rate": 3.426715263554706e-05, "loss": 0.0679, "step": 124360 }, { "epoch": 4.204672233679299, "grad_norm": 2.0474705696105957, "learning_rate": 3.425930543775246e-05, "loss": 0.0642, "step": 124370 }, { "epoch": 4.205010311369553, "grad_norm": 7.238474369049072, "learning_rate": 3.425145867030254e-05, "loss": 0.0696, "step": 124380 }, { "epoch": 4.205348389059806, "grad_norm": 4.9649152755737305, "learning_rate": 3.424361233341182e-05, "loss": 0.0671, "step": 124390 }, { "epoch": 4.205686466750059, "grad_norm": 1.5788995027542114, "learning_rate": 3.4235766427294825e-05, "loss": 0.0585, "step": 124400 }, { "epoch": 4.206024544440313, "grad_norm": 4.333350658416748, "learning_rate": 3.422792095216604e-05, "loss": 0.0686, "step": 124410 }, { "epoch": 4.206362622130565, "grad_norm": 3.1918201446533203, "learning_rate": 3.422007590823999e-05, "loss": 0.0651, "step": 124420 }, { "epoch": 4.206700699820819, "grad_norm": 21.752626419067383, "learning_rate": 3.4212231295731114e-05, "loss": 0.0598, "step": 124430 }, { "epoch": 4.207038777511072, "grad_norm": 3.4684653282165527, "learning_rate": 3.42043871148539e-05, "loss": 0.0698, "step": 124440 }, { "epoch": 4.207376855201325, "grad_norm": 3.937594175338745, "learning_rate": 3.419654336582283e-05, "loss": 0.0641, "step": 124450 }, { "epoch": 4.207714932891578, "grad_norm": 3.8791205883026123, "learning_rate": 3.4188700048852314e-05, "loss": 0.0665, "step": 124460 }, { "epoch": 4.208053010581832, "grad_norm": 2.8333961963653564, "learning_rate": 3.418085716415681e-05, "loss": 0.0693, "step": 124470 }, { "epoch": 4.2083910882720845, "grad_norm": 4.524318218231201, "learning_rate": 3.417301471195071e-05, "loss": 0.0648, "step": 124480 }, { "epoch": 4.208729165962338, "grad_norm": 18.292909622192383, "learning_rate": 3.4165172692448465e-05, "loss": 0.071, "step": 124490 }, { "epoch": 4.2090672436525916, "grad_norm": 6.083944797515869, "learning_rate": 3.415733110586443e-05, "loss": 0.0695, "step": 124500 }, { "epoch": 4.209405321342844, "grad_norm": 5.462231159210205, "learning_rate": 3.414948995241301e-05, "loss": 0.0686, "step": 124510 }, { "epoch": 4.209743399033098, "grad_norm": 6.983648777008057, "learning_rate": 3.4141649232308604e-05, "loss": 0.0638, "step": 124520 }, { "epoch": 4.210081476723351, "grad_norm": 6.037243843078613, "learning_rate": 3.4133808945765524e-05, "loss": 0.0702, "step": 124530 }, { "epoch": 4.210419554413604, "grad_norm": 4.836452960968018, "learning_rate": 3.412596909299817e-05, "loss": 0.0656, "step": 124540 }, { "epoch": 4.210757632103857, "grad_norm": 2.756222724914551, "learning_rate": 3.411812967422083e-05, "loss": 0.0664, "step": 124550 }, { "epoch": 4.211095709794111, "grad_norm": 3.9875712394714355, "learning_rate": 3.411029068964787e-05, "loss": 0.0686, "step": 124560 }, { "epoch": 4.2114337874843635, "grad_norm": 6.882508277893066, "learning_rate": 3.410245213949361e-05, "loss": 0.0659, "step": 124570 }, { "epoch": 4.211771865174617, "grad_norm": 3.8369646072387695, "learning_rate": 3.4094614023972315e-05, "loss": 0.0694, "step": 124580 }, { "epoch": 4.2121099428648705, "grad_norm": 3.4486124515533447, "learning_rate": 3.408677634329831e-05, "loss": 0.0607, "step": 124590 }, { "epoch": 4.212448020555124, "grad_norm": 3.2054264545440674, "learning_rate": 3.407893909768586e-05, "loss": 0.0635, "step": 124600 }, { "epoch": 4.212786098245377, "grad_norm": 7.264907360076904, "learning_rate": 3.407110228734923e-05, "loss": 0.0642, "step": 124610 }, { "epoch": 4.21312417593563, "grad_norm": 3.478196859359741, "learning_rate": 3.406326591250267e-05, "loss": 0.0653, "step": 124620 }, { "epoch": 4.213462253625883, "grad_norm": 5.413522720336914, "learning_rate": 3.405542997336043e-05, "loss": 0.0623, "step": 124630 }, { "epoch": 4.213800331316136, "grad_norm": 22.4851016998291, "learning_rate": 3.404759447013678e-05, "loss": 0.0649, "step": 124640 }, { "epoch": 4.21413840900639, "grad_norm": 10.053766250610352, "learning_rate": 3.403975940304586e-05, "loss": 0.0668, "step": 124650 }, { "epoch": 4.214476486696643, "grad_norm": 20.28791618347168, "learning_rate": 3.4031924772301955e-05, "loss": 0.062, "step": 124660 }, { "epoch": 4.214814564386896, "grad_norm": 6.410284519195557, "learning_rate": 3.402409057811921e-05, "loss": 0.0686, "step": 124670 }, { "epoch": 4.2151526420771495, "grad_norm": 9.528641700744629, "learning_rate": 3.401625682071184e-05, "loss": 0.0588, "step": 124680 }, { "epoch": 4.215490719767403, "grad_norm": 43.55938720703125, "learning_rate": 3.400842350029399e-05, "loss": 0.0594, "step": 124690 }, { "epoch": 4.215828797457656, "grad_norm": 12.453950881958008, "learning_rate": 3.400059061707983e-05, "loss": 0.0697, "step": 124700 }, { "epoch": 4.216166875147909, "grad_norm": 5.856115818023682, "learning_rate": 3.3992758171283536e-05, "loss": 0.0692, "step": 124710 }, { "epoch": 4.216504952838163, "grad_norm": 6.027766227722168, "learning_rate": 3.3984926163119204e-05, "loss": 0.0633, "step": 124720 }, { "epoch": 4.216843030528415, "grad_norm": 4.082973480224609, "learning_rate": 3.3977094592800976e-05, "loss": 0.0609, "step": 124730 }, { "epoch": 4.217181108218669, "grad_norm": 12.978191375732422, "learning_rate": 3.396926346054295e-05, "loss": 0.0641, "step": 124740 }, { "epoch": 4.217519185908922, "grad_norm": 19.700674057006836, "learning_rate": 3.3961432766559244e-05, "loss": 0.0611, "step": 124750 }, { "epoch": 4.217857263599175, "grad_norm": 5.407935619354248, "learning_rate": 3.3953602511063954e-05, "loss": 0.0586, "step": 124760 }, { "epoch": 4.218195341289428, "grad_norm": 9.10937786102295, "learning_rate": 3.394577269427113e-05, "loss": 0.0686, "step": 124770 }, { "epoch": 4.218533418979682, "grad_norm": 5.448326110839844, "learning_rate": 3.393794331639486e-05, "loss": 0.0616, "step": 124780 }, { "epoch": 4.2188714966699346, "grad_norm": 168.93923950195312, "learning_rate": 3.393011437764918e-05, "loss": 0.0599, "step": 124790 }, { "epoch": 4.219209574360188, "grad_norm": 54.236480712890625, "learning_rate": 3.392228587824814e-05, "loss": 0.065, "step": 124800 }, { "epoch": 4.219547652050442, "grad_norm": 3.5404326915740967, "learning_rate": 3.3914457818405766e-05, "loss": 0.0652, "step": 124810 }, { "epoch": 4.219885729740694, "grad_norm": 3.207796573638916, "learning_rate": 3.3906630198336045e-05, "loss": 0.064, "step": 124820 }, { "epoch": 4.220223807430948, "grad_norm": 6.63956356048584, "learning_rate": 3.389880301825305e-05, "loss": 0.0698, "step": 124830 }, { "epoch": 4.220561885121201, "grad_norm": 18.86829948425293, "learning_rate": 3.38909762783707e-05, "loss": 0.0643, "step": 124840 }, { "epoch": 4.220899962811454, "grad_norm": 17.152267456054688, "learning_rate": 3.3883149978903036e-05, "loss": 0.0674, "step": 124850 }, { "epoch": 4.221238040501707, "grad_norm": 4.225857734680176, "learning_rate": 3.387532412006396e-05, "loss": 0.0593, "step": 124860 }, { "epoch": 4.221576118191961, "grad_norm": 9.394107818603516, "learning_rate": 3.386749870206749e-05, "loss": 0.0641, "step": 124870 }, { "epoch": 4.2219141958822135, "grad_norm": 4.591907501220703, "learning_rate": 3.385967372512754e-05, "loss": 0.0576, "step": 124880 }, { "epoch": 4.222252273572467, "grad_norm": 5.959919452667236, "learning_rate": 3.385184918945804e-05, "loss": 0.0638, "step": 124890 }, { "epoch": 4.2225903512627205, "grad_norm": 8.829001426696777, "learning_rate": 3.384402509527292e-05, "loss": 0.0693, "step": 124900 }, { "epoch": 4.222928428952973, "grad_norm": 11.439581871032715, "learning_rate": 3.383620144278609e-05, "loss": 0.0714, "step": 124910 }, { "epoch": 4.223266506643227, "grad_norm": 12.838629722595215, "learning_rate": 3.382837823221144e-05, "loss": 0.0714, "step": 124920 }, { "epoch": 4.22360458433348, "grad_norm": 17.95448875427246, "learning_rate": 3.382055546376285e-05, "loss": 0.0682, "step": 124930 }, { "epoch": 4.223942662023733, "grad_norm": 5.68960428237915, "learning_rate": 3.3812733137654176e-05, "loss": 0.0677, "step": 124940 }, { "epoch": 4.224280739713986, "grad_norm": 8.88604736328125, "learning_rate": 3.380491125409932e-05, "loss": 0.0691, "step": 124950 }, { "epoch": 4.22461881740424, "grad_norm": 3.7446513175964355, "learning_rate": 3.3797089813312103e-05, "loss": 0.0694, "step": 124960 }, { "epoch": 4.2249568950944925, "grad_norm": 4.031774044036865, "learning_rate": 3.378926881550637e-05, "loss": 0.0611, "step": 124970 }, { "epoch": 4.225294972784746, "grad_norm": 2.143540859222412, "learning_rate": 3.3781448260895925e-05, "loss": 0.0595, "step": 124980 }, { "epoch": 4.2256330504749995, "grad_norm": 13.294830322265625, "learning_rate": 3.37736281496946e-05, "loss": 0.0634, "step": 124990 }, { "epoch": 4.225971128165252, "grad_norm": 3.843163251876831, "learning_rate": 3.376580848211618e-05, "loss": 0.0676, "step": 125000 }, { "epoch": 4.226309205855506, "grad_norm": 4.667377471923828, "learning_rate": 3.375798925837446e-05, "loss": 0.0684, "step": 125010 }, { "epoch": 4.226647283545759, "grad_norm": 2.773723602294922, "learning_rate": 3.375017047868322e-05, "loss": 0.0626, "step": 125020 }, { "epoch": 4.226985361236012, "grad_norm": 3.913008451461792, "learning_rate": 3.374235214325618e-05, "loss": 0.0637, "step": 125030 }, { "epoch": 4.227323438926265, "grad_norm": 4.696812152862549, "learning_rate": 3.373453425230717e-05, "loss": 0.0624, "step": 125040 }, { "epoch": 4.227661516616519, "grad_norm": 3.9785659313201904, "learning_rate": 3.3726716806049855e-05, "loss": 0.0661, "step": 125050 }, { "epoch": 4.227999594306771, "grad_norm": 3.9464566707611084, "learning_rate": 3.3718899804698014e-05, "loss": 0.0727, "step": 125060 }, { "epoch": 4.228337671997025, "grad_norm": 2.884615421295166, "learning_rate": 3.37110832484653e-05, "loss": 0.064, "step": 125070 }, { "epoch": 4.228675749687278, "grad_norm": 4.292041778564453, "learning_rate": 3.3703267137565465e-05, "loss": 0.0688, "step": 125080 }, { "epoch": 4.229013827377531, "grad_norm": 2.8222908973693848, "learning_rate": 3.36954514722122e-05, "loss": 0.0715, "step": 125090 }, { "epoch": 4.229351905067785, "grad_norm": 3.5745787620544434, "learning_rate": 3.368763625261915e-05, "loss": 0.0653, "step": 125100 }, { "epoch": 4.229689982758038, "grad_norm": 1.9534369707107544, "learning_rate": 3.3679821479000005e-05, "loss": 0.0601, "step": 125110 }, { "epoch": 4.230028060448291, "grad_norm": 3.571617364883423, "learning_rate": 3.36720071515684e-05, "loss": 0.0686, "step": 125120 }, { "epoch": 4.230366138138544, "grad_norm": 7.8875837326049805, "learning_rate": 3.366419327053798e-05, "loss": 0.0698, "step": 125130 }, { "epoch": 4.230704215828798, "grad_norm": 4.0288543701171875, "learning_rate": 3.365637983612241e-05, "loss": 0.0674, "step": 125140 }, { "epoch": 4.23104229351905, "grad_norm": 16.696630477905273, "learning_rate": 3.364856684853523e-05, "loss": 0.0654, "step": 125150 }, { "epoch": 4.231380371209304, "grad_norm": 5.922128677368164, "learning_rate": 3.364075430799013e-05, "loss": 0.0639, "step": 125160 }, { "epoch": 4.231718448899557, "grad_norm": 2.6782615184783936, "learning_rate": 3.363294221470064e-05, "loss": 0.0677, "step": 125170 }, { "epoch": 4.23205652658981, "grad_norm": 2.699779748916626, "learning_rate": 3.362513056888038e-05, "loss": 0.0622, "step": 125180 }, { "epoch": 4.2323946042800635, "grad_norm": 3.796637535095215, "learning_rate": 3.361731937074288e-05, "loss": 0.0623, "step": 125190 }, { "epoch": 4.232732681970317, "grad_norm": 4.785669803619385, "learning_rate": 3.3609508620501716e-05, "loss": 0.0705, "step": 125200 }, { "epoch": 4.23307075966057, "grad_norm": 4.44460916519165, "learning_rate": 3.3601698318370433e-05, "loss": 0.0671, "step": 125210 }, { "epoch": 4.233408837350823, "grad_norm": 8.493370056152344, "learning_rate": 3.3593888464562556e-05, "loss": 0.0713, "step": 125220 }, { "epoch": 4.233746915041077, "grad_norm": 10.885411262512207, "learning_rate": 3.358607905929161e-05, "loss": 0.0594, "step": 125230 }, { "epoch": 4.234084992731329, "grad_norm": 19.158735275268555, "learning_rate": 3.357827010277108e-05, "loss": 0.0646, "step": 125240 }, { "epoch": 4.234423070421583, "grad_norm": 47.254722595214844, "learning_rate": 3.3570461595214504e-05, "loss": 0.0707, "step": 125250 }, { "epoch": 4.234761148111836, "grad_norm": 24.827003479003906, "learning_rate": 3.35626535368353e-05, "loss": 0.0641, "step": 125260 }, { "epoch": 4.235099225802089, "grad_norm": 5.323698043823242, "learning_rate": 3.355484592784699e-05, "loss": 0.0656, "step": 125270 }, { "epoch": 4.2354373034923425, "grad_norm": 3.873889923095703, "learning_rate": 3.3547038768463025e-05, "loss": 0.0702, "step": 125280 }, { "epoch": 4.235775381182596, "grad_norm": 2.82195782661438, "learning_rate": 3.353923205889683e-05, "loss": 0.0698, "step": 125290 }, { "epoch": 4.236113458872849, "grad_norm": 6.040563583374023, "learning_rate": 3.353142579936185e-05, "loss": 0.0665, "step": 125300 }, { "epoch": 4.236451536563102, "grad_norm": 4.365820407867432, "learning_rate": 3.35236199900715e-05, "loss": 0.0716, "step": 125310 }, { "epoch": 4.236789614253356, "grad_norm": 11.048956871032715, "learning_rate": 3.3515814631239175e-05, "loss": 0.0637, "step": 125320 }, { "epoch": 4.237127691943608, "grad_norm": 1.7861847877502441, "learning_rate": 3.350800972307831e-05, "loss": 0.0689, "step": 125330 }, { "epoch": 4.237465769633862, "grad_norm": 4.692168712615967, "learning_rate": 3.3500205265802234e-05, "loss": 0.0683, "step": 125340 }, { "epoch": 4.237803847324115, "grad_norm": 9.703969955444336, "learning_rate": 3.349240125962438e-05, "loss": 0.0643, "step": 125350 }, { "epoch": 4.238141925014368, "grad_norm": 4.217187404632568, "learning_rate": 3.348459770475805e-05, "loss": 0.0654, "step": 125360 }, { "epoch": 4.238480002704621, "grad_norm": 3.682755947113037, "learning_rate": 3.347679460141663e-05, "loss": 0.0644, "step": 125370 }, { "epoch": 4.238818080394875, "grad_norm": 7.112089157104492, "learning_rate": 3.3468991949813426e-05, "loss": 0.0699, "step": 125380 }, { "epoch": 4.239156158085128, "grad_norm": 3.1539762020111084, "learning_rate": 3.3461189750161774e-05, "loss": 0.0641, "step": 125390 }, { "epoch": 4.239494235775381, "grad_norm": 6.014593601226807, "learning_rate": 3.345338800267499e-05, "loss": 0.0641, "step": 125400 }, { "epoch": 4.239832313465635, "grad_norm": 2.847867012023926, "learning_rate": 3.3445586707566356e-05, "loss": 0.0657, "step": 125410 }, { "epoch": 4.240170391155887, "grad_norm": 2.2088303565979004, "learning_rate": 3.343778586504917e-05, "loss": 0.0634, "step": 125420 }, { "epoch": 4.240508468846141, "grad_norm": 1.4793657064437866, "learning_rate": 3.342998547533668e-05, "loss": 0.0651, "step": 125430 }, { "epoch": 4.240846546536394, "grad_norm": 2.943948268890381, "learning_rate": 3.3422185538642175e-05, "loss": 0.0649, "step": 125440 }, { "epoch": 4.241184624226647, "grad_norm": 7.585553169250488, "learning_rate": 3.341438605517887e-05, "loss": 0.0668, "step": 125450 }, { "epoch": 4.2415227019169, "grad_norm": 4.522591590881348, "learning_rate": 3.340658702516003e-05, "loss": 0.0649, "step": 125460 }, { "epoch": 4.241860779607154, "grad_norm": 2.8666582107543945, "learning_rate": 3.339878844879888e-05, "loss": 0.0626, "step": 125470 }, { "epoch": 4.2421988572974065, "grad_norm": 7.886550426483154, "learning_rate": 3.33909903263086e-05, "loss": 0.0675, "step": 125480 }, { "epoch": 4.24253693498766, "grad_norm": 2.786870241165161, "learning_rate": 3.3383192657902424e-05, "loss": 0.0588, "step": 125490 }, { "epoch": 4.242875012677914, "grad_norm": 2.4276537895202637, "learning_rate": 3.33753954437935e-05, "loss": 0.0637, "step": 125500 }, { "epoch": 4.243213090368166, "grad_norm": 3.2318060398101807, "learning_rate": 3.3367598684195034e-05, "loss": 0.0619, "step": 125510 }, { "epoch": 4.24355116805842, "grad_norm": 1.7607145309448242, "learning_rate": 3.335980237932016e-05, "loss": 0.0638, "step": 125520 }, { "epoch": 4.243889245748673, "grad_norm": 1.5643129348754883, "learning_rate": 3.335200652938203e-05, "loss": 0.0572, "step": 125530 }, { "epoch": 4.244227323438926, "grad_norm": 2.4928932189941406, "learning_rate": 3.334421113459381e-05, "loss": 0.0626, "step": 125540 }, { "epoch": 4.244565401129179, "grad_norm": 1.2935130596160889, "learning_rate": 3.333641619516857e-05, "loss": 0.0659, "step": 125550 }, { "epoch": 4.244903478819433, "grad_norm": 3.375326156616211, "learning_rate": 3.3328621711319494e-05, "loss": 0.0588, "step": 125560 }, { "epoch": 4.245241556509686, "grad_norm": 6.918107509613037, "learning_rate": 3.332082768325959e-05, "loss": 0.0651, "step": 125570 }, { "epoch": 4.245579634199939, "grad_norm": 7.895354747772217, "learning_rate": 3.331303411120201e-05, "loss": 0.0662, "step": 125580 }, { "epoch": 4.2459177118901925, "grad_norm": 2.687863826751709, "learning_rate": 3.330524099535982e-05, "loss": 0.0635, "step": 125590 }, { "epoch": 4.246255789580445, "grad_norm": 6.01027774810791, "learning_rate": 3.3297448335946056e-05, "loss": 0.0621, "step": 125600 }, { "epoch": 4.246593867270699, "grad_norm": 9.746926307678223, "learning_rate": 3.328965613317379e-05, "loss": 0.0565, "step": 125610 }, { "epoch": 4.246931944960952, "grad_norm": 6.308382511138916, "learning_rate": 3.328186438725604e-05, "loss": 0.0664, "step": 125620 }, { "epoch": 4.247270022651206, "grad_norm": 2.1707398891448975, "learning_rate": 3.327407309840585e-05, "loss": 0.0645, "step": 125630 }, { "epoch": 4.247608100341458, "grad_norm": 2.8283448219299316, "learning_rate": 3.32662822668362e-05, "loss": 0.0678, "step": 125640 }, { "epoch": 4.247946178031712, "grad_norm": 8.523335456848145, "learning_rate": 3.3258491892760104e-05, "loss": 0.0617, "step": 125650 }, { "epoch": 4.248284255721965, "grad_norm": 1.9966133832931519, "learning_rate": 3.3250701976390565e-05, "loss": 0.0682, "step": 125660 }, { "epoch": 4.248622333412218, "grad_norm": 3.383854389190674, "learning_rate": 3.324291251794053e-05, "loss": 0.0625, "step": 125670 }, { "epoch": 4.2489604111024715, "grad_norm": 2.0290627479553223, "learning_rate": 3.3235123517622985e-05, "loss": 0.0617, "step": 125680 }, { "epoch": 4.249298488792725, "grad_norm": 12.63961124420166, "learning_rate": 3.3227334975650854e-05, "loss": 0.0634, "step": 125690 }, { "epoch": 4.249636566482978, "grad_norm": 2.403399705886841, "learning_rate": 3.32195468922371e-05, "loss": 0.0643, "step": 125700 }, { "epoch": 4.249974644173231, "grad_norm": 5.985455513000488, "learning_rate": 3.321175926759461e-05, "loss": 0.0619, "step": 125710 }, { "epoch": 4.250312721863485, "grad_norm": 3.153402805328369, "learning_rate": 3.320397210193631e-05, "loss": 0.0592, "step": 125720 }, { "epoch": 4.250650799553737, "grad_norm": 3.443840265274048, "learning_rate": 3.319618539547512e-05, "loss": 0.0597, "step": 125730 }, { "epoch": 4.250988877243991, "grad_norm": 2.387148141860962, "learning_rate": 3.318839914842389e-05, "loss": 0.0615, "step": 125740 }, { "epoch": 4.251326954934244, "grad_norm": 2.9942715167999268, "learning_rate": 3.318061336099554e-05, "loss": 0.063, "step": 125750 }, { "epoch": 4.251665032624497, "grad_norm": 2.308135747909546, "learning_rate": 3.317282803340288e-05, "loss": 0.0714, "step": 125760 }, { "epoch": 4.25200311031475, "grad_norm": 1.8897732496261597, "learning_rate": 3.316504316585878e-05, "loss": 0.0655, "step": 125770 }, { "epoch": 4.252341188005004, "grad_norm": 1.7640036344528198, "learning_rate": 3.3157258758576084e-05, "loss": 0.0642, "step": 125780 }, { "epoch": 4.252679265695257, "grad_norm": 3.1971347332000732, "learning_rate": 3.31494748117676e-05, "loss": 0.06, "step": 125790 }, { "epoch": 4.25301734338551, "grad_norm": 31.619068145751953, "learning_rate": 3.314169132564615e-05, "loss": 0.0659, "step": 125800 }, { "epoch": 4.253355421075764, "grad_norm": 2.4920146465301514, "learning_rate": 3.313390830042452e-05, "loss": 0.0583, "step": 125810 }, { "epoch": 4.253693498766016, "grad_norm": 3.983344078063965, "learning_rate": 3.312612573631551e-05, "loss": 0.0658, "step": 125820 }, { "epoch": 4.25403157645627, "grad_norm": 20.31688117980957, "learning_rate": 3.311834363353187e-05, "loss": 0.0645, "step": 125830 }, { "epoch": 4.254369654146523, "grad_norm": 6.989761829376221, "learning_rate": 3.311056199228637e-05, "loss": 0.0631, "step": 125840 }, { "epoch": 4.254707731836776, "grad_norm": 4.526477336883545, "learning_rate": 3.310278081279179e-05, "loss": 0.0684, "step": 125850 }, { "epoch": 4.255045809527029, "grad_norm": 3.336740732192993, "learning_rate": 3.3095000095260795e-05, "loss": 0.069, "step": 125860 }, { "epoch": 4.255383887217283, "grad_norm": 5.77692985534668, "learning_rate": 3.308721983990618e-05, "loss": 0.07, "step": 125870 }, { "epoch": 4.2557219649075355, "grad_norm": 2.6482114791870117, "learning_rate": 3.30794400469406e-05, "loss": 0.0652, "step": 125880 }, { "epoch": 4.256060042597789, "grad_norm": 3.976583242416382, "learning_rate": 3.307166071657678e-05, "loss": 0.0602, "step": 125890 }, { "epoch": 4.2563981202880425, "grad_norm": 3.471837043762207, "learning_rate": 3.306388184902739e-05, "loss": 0.0568, "step": 125900 }, { "epoch": 4.256736197978295, "grad_norm": 3.4562928676605225, "learning_rate": 3.3056103444505115e-05, "loss": 0.0643, "step": 125910 }, { "epoch": 4.257074275668549, "grad_norm": 2.3046109676361084, "learning_rate": 3.304832550322261e-05, "loss": 0.062, "step": 125920 }, { "epoch": 4.257412353358802, "grad_norm": 1.4765653610229492, "learning_rate": 3.3040548025392504e-05, "loss": 0.0618, "step": 125930 }, { "epoch": 4.257750431049055, "grad_norm": 2.5210959911346436, "learning_rate": 3.3032771011227456e-05, "loss": 0.07, "step": 125940 }, { "epoch": 4.258088508739308, "grad_norm": 2.6893911361694336, "learning_rate": 3.302499446094006e-05, "loss": 0.0543, "step": 125950 }, { "epoch": 4.258426586429562, "grad_norm": 3.4914591312408447, "learning_rate": 3.301721837474294e-05, "loss": 0.0652, "step": 125960 }, { "epoch": 4.2587646641198145, "grad_norm": 3.032801389694214, "learning_rate": 3.300944275284871e-05, "loss": 0.0631, "step": 125970 }, { "epoch": 4.259102741810068, "grad_norm": 3.8996775150299072, "learning_rate": 3.300166759546991e-05, "loss": 0.0639, "step": 125980 }, { "epoch": 4.2594408195003215, "grad_norm": 2.0200603008270264, "learning_rate": 3.2993892902819155e-05, "loss": 0.0589, "step": 125990 }, { "epoch": 4.259778897190574, "grad_norm": 2.6111538410186768, "learning_rate": 3.298611867510896e-05, "loss": 0.0638, "step": 126000 }, { "epoch": 4.260116974880828, "grad_norm": 2.1195778846740723, "learning_rate": 3.29783449125519e-05, "loss": 0.0642, "step": 126010 }, { "epoch": 4.260455052571081, "grad_norm": 2.533590793609619, "learning_rate": 3.297057161536048e-05, "loss": 0.0637, "step": 126020 }, { "epoch": 4.260793130261334, "grad_norm": 1.3900651931762695, "learning_rate": 3.296279878374723e-05, "loss": 0.0673, "step": 126030 }, { "epoch": 4.261131207951587, "grad_norm": 1.5238896608352661, "learning_rate": 3.29550264179247e-05, "loss": 0.0556, "step": 126040 }, { "epoch": 4.261469285641841, "grad_norm": 2.7239174842834473, "learning_rate": 3.2947254518105296e-05, "loss": 0.0609, "step": 126050 }, { "epoch": 4.261807363332093, "grad_norm": 2.1988155841827393, "learning_rate": 3.2939483084501584e-05, "loss": 0.0661, "step": 126060 }, { "epoch": 4.262145441022347, "grad_norm": 4.463598251342773, "learning_rate": 3.293171211732598e-05, "loss": 0.0658, "step": 126070 }, { "epoch": 4.2624835187126004, "grad_norm": 1.4496444463729858, "learning_rate": 3.292394161679095e-05, "loss": 0.0539, "step": 126080 }, { "epoch": 4.262821596402853, "grad_norm": 2.335946798324585, "learning_rate": 3.291617158310895e-05, "loss": 0.0644, "step": 126090 }, { "epoch": 4.263159674093107, "grad_norm": 3.4189682006835938, "learning_rate": 3.290840201649239e-05, "loss": 0.0652, "step": 126100 }, { "epoch": 4.26349775178336, "grad_norm": 3.0228781700134277, "learning_rate": 3.2900632917153705e-05, "loss": 0.0654, "step": 126110 }, { "epoch": 4.263835829473613, "grad_norm": 2.574462413787842, "learning_rate": 3.2892864285305294e-05, "loss": 0.0624, "step": 126120 }, { "epoch": 4.264173907163866, "grad_norm": 9.604905128479004, "learning_rate": 3.2885096121159556e-05, "loss": 0.0638, "step": 126130 }, { "epoch": 4.26451198485412, "grad_norm": 2.8002007007598877, "learning_rate": 3.287732842492884e-05, "loss": 0.0564, "step": 126140 }, { "epoch": 4.264850062544372, "grad_norm": 3.1759109497070312, "learning_rate": 3.286956119682553e-05, "loss": 0.0599, "step": 126150 }, { "epoch": 4.265188140234626, "grad_norm": 2.696178436279297, "learning_rate": 3.2861794437061995e-05, "loss": 0.0583, "step": 126160 }, { "epoch": 4.265526217924879, "grad_norm": 1.4668159484863281, "learning_rate": 3.285402814585056e-05, "loss": 0.0544, "step": 126170 }, { "epoch": 4.265864295615132, "grad_norm": 0.9401497840881348, "learning_rate": 3.2846262323403556e-05, "loss": 0.0589, "step": 126180 }, { "epoch": 4.2662023733053855, "grad_norm": 2.801222562789917, "learning_rate": 3.2838496969933284e-05, "loss": 0.061, "step": 126190 }, { "epoch": 4.266540450995639, "grad_norm": 1.9098389148712158, "learning_rate": 3.283073208565207e-05, "loss": 0.0634, "step": 126200 }, { "epoch": 4.266878528685892, "grad_norm": 5.271843433380127, "learning_rate": 3.2822967670772185e-05, "loss": 0.0621, "step": 126210 }, { "epoch": 4.267216606376145, "grad_norm": 2.368391275405884, "learning_rate": 3.281520372550592e-05, "loss": 0.0716, "step": 126220 }, { "epoch": 4.267554684066399, "grad_norm": 1.3249387741088867, "learning_rate": 3.280744025006552e-05, "loss": 0.0591, "step": 126230 }, { "epoch": 4.267892761756651, "grad_norm": 1.6924853324890137, "learning_rate": 3.279967724466323e-05, "loss": 0.0589, "step": 126240 }, { "epoch": 4.268230839446905, "grad_norm": 5.639789581298828, "learning_rate": 3.279191470951134e-05, "loss": 0.059, "step": 126250 }, { "epoch": 4.268568917137158, "grad_norm": 3.798671007156372, "learning_rate": 3.2784152644822e-05, "loss": 0.0574, "step": 126260 }, { "epoch": 4.268906994827411, "grad_norm": 30.254684448242188, "learning_rate": 3.277639105080749e-05, "loss": 0.0679, "step": 126270 }, { "epoch": 4.2692450725176645, "grad_norm": 2.2913615703582764, "learning_rate": 3.276862992767995e-05, "loss": 0.0608, "step": 126280 }, { "epoch": 4.269583150207918, "grad_norm": 2.4231722354888916, "learning_rate": 3.276086927565161e-05, "loss": 0.0684, "step": 126290 }, { "epoch": 4.269921227898171, "grad_norm": 3.1985464096069336, "learning_rate": 3.275310909493462e-05, "loss": 0.0616, "step": 126300 }, { "epoch": 4.270259305588424, "grad_norm": 1.0893710851669312, "learning_rate": 3.2745349385741156e-05, "loss": 0.058, "step": 126310 }, { "epoch": 4.270597383278678, "grad_norm": 1.8117080926895142, "learning_rate": 3.273759014828336e-05, "loss": 0.063, "step": 126320 }, { "epoch": 4.27093546096893, "grad_norm": 2.8036651611328125, "learning_rate": 3.272983138277335e-05, "loss": 0.0675, "step": 126330 }, { "epoch": 4.271273538659184, "grad_norm": 1.491942048072815, "learning_rate": 3.2722073089423256e-05, "loss": 0.0597, "step": 126340 }, { "epoch": 4.271611616349437, "grad_norm": 1.4789129495620728, "learning_rate": 3.2714315268445226e-05, "loss": 0.0631, "step": 126350 }, { "epoch": 4.27194969403969, "grad_norm": 3.4626309871673584, "learning_rate": 3.270655792005129e-05, "loss": 0.0592, "step": 126360 }, { "epoch": 4.2722877717299435, "grad_norm": 5.946259021759033, "learning_rate": 3.269880104445358e-05, "loss": 0.0579, "step": 126370 }, { "epoch": 4.272625849420197, "grad_norm": 1.89745032787323, "learning_rate": 3.269104464186414e-05, "loss": 0.066, "step": 126380 }, { "epoch": 4.27296392711045, "grad_norm": 2.3411705493927, "learning_rate": 3.268328871249505e-05, "loss": 0.0577, "step": 126390 }, { "epoch": 4.273302004800703, "grad_norm": 2.0467329025268555, "learning_rate": 3.267553325655833e-05, "loss": 0.0582, "step": 126400 }, { "epoch": 4.273640082490957, "grad_norm": 1.6733150482177734, "learning_rate": 3.266777827426603e-05, "loss": 0.0665, "step": 126410 }, { "epoch": 4.273978160181209, "grad_norm": 2.1073355674743652, "learning_rate": 3.266002376583016e-05, "loss": 0.0603, "step": 126420 }, { "epoch": 4.274316237871463, "grad_norm": 2.7478644847869873, "learning_rate": 3.265226973146272e-05, "loss": 0.0541, "step": 126430 }, { "epoch": 4.274654315561716, "grad_norm": 2.8023622035980225, "learning_rate": 3.264451617137572e-05, "loss": 0.0613, "step": 126440 }, { "epoch": 4.274992393251969, "grad_norm": 1.7254565954208374, "learning_rate": 3.2636763085781105e-05, "loss": 0.0607, "step": 126450 }, { "epoch": 4.275330470942222, "grad_norm": 2.1315019130706787, "learning_rate": 3.2629010474890895e-05, "loss": 0.0568, "step": 126460 }, { "epoch": 4.275668548632476, "grad_norm": 6.657186031341553, "learning_rate": 3.262125833891698e-05, "loss": 0.0682, "step": 126470 }, { "epoch": 4.276006626322729, "grad_norm": 5.724409580230713, "learning_rate": 3.261350667807135e-05, "loss": 0.0702, "step": 126480 }, { "epoch": 4.276344704012982, "grad_norm": 3.56337571144104, "learning_rate": 3.260575549256592e-05, "loss": 0.0613, "step": 126490 }, { "epoch": 4.276682781703236, "grad_norm": 2.180640697479248, "learning_rate": 3.259800478261259e-05, "loss": 0.0648, "step": 126500 }, { "epoch": 4.277020859393488, "grad_norm": 1.7323062419891357, "learning_rate": 3.2590254548423285e-05, "loss": 0.0639, "step": 126510 }, { "epoch": 4.277358937083742, "grad_norm": 4.964233875274658, "learning_rate": 3.258250479020987e-05, "loss": 0.0597, "step": 126520 }, { "epoch": 4.277697014773995, "grad_norm": 4.2802653312683105, "learning_rate": 3.2574755508184215e-05, "loss": 0.0694, "step": 126530 }, { "epoch": 4.278035092464249, "grad_norm": 1.9415490627288818, "learning_rate": 3.2567006702558235e-05, "loss": 0.058, "step": 126540 }, { "epoch": 4.278373170154501, "grad_norm": 2.688138484954834, "learning_rate": 3.25592583735437e-05, "loss": 0.0585, "step": 126550 }, { "epoch": 4.278711247844755, "grad_norm": 3.861377477645874, "learning_rate": 3.2551510521352535e-05, "loss": 0.0581, "step": 126560 }, { "epoch": 4.2790493255350075, "grad_norm": 4.572569370269775, "learning_rate": 3.254376314619648e-05, "loss": 0.0627, "step": 126570 }, { "epoch": 4.279387403225261, "grad_norm": 2.208994150161743, "learning_rate": 3.2536016248287405e-05, "loss": 0.0654, "step": 126580 }, { "epoch": 4.2797254809155145, "grad_norm": 8.768376350402832, "learning_rate": 3.252826982783707e-05, "loss": 0.067, "step": 126590 }, { "epoch": 4.280063558605768, "grad_norm": 2.5122854709625244, "learning_rate": 3.252052388505728e-05, "loss": 0.0668, "step": 126600 }, { "epoch": 4.280401636296021, "grad_norm": 3.130986213684082, "learning_rate": 3.251277842015981e-05, "loss": 0.0643, "step": 126610 }, { "epoch": 4.280739713986274, "grad_norm": 3.9150326251983643, "learning_rate": 3.250503343335639e-05, "loss": 0.0674, "step": 126620 }, { "epoch": 4.281077791676527, "grad_norm": 2.0964534282684326, "learning_rate": 3.24972889248588e-05, "loss": 0.068, "step": 126630 }, { "epoch": 4.28141586936678, "grad_norm": 3.616518974304199, "learning_rate": 3.248954489487874e-05, "loss": 0.0639, "step": 126640 }, { "epoch": 4.281753947057034, "grad_norm": 4.92343807220459, "learning_rate": 3.248180134362796e-05, "loss": 0.0622, "step": 126650 }, { "epoch": 4.282092024747287, "grad_norm": 28.402584075927734, "learning_rate": 3.247405827131813e-05, "loss": 0.0654, "step": 126660 }, { "epoch": 4.28243010243754, "grad_norm": 2.0217978954315186, "learning_rate": 3.246631567816097e-05, "loss": 0.0638, "step": 126670 }, { "epoch": 4.2827681801277935, "grad_norm": 8.85047721862793, "learning_rate": 3.245857356436817e-05, "loss": 0.0619, "step": 126680 }, { "epoch": 4.283106257818047, "grad_norm": 2.296213388442993, "learning_rate": 3.245083193015135e-05, "loss": 0.0586, "step": 126690 }, { "epoch": 4.2834443355083, "grad_norm": 6.462766647338867, "learning_rate": 3.244309077572222e-05, "loss": 0.0648, "step": 126700 }, { "epoch": 4.283782413198553, "grad_norm": 2.105166435241699, "learning_rate": 3.2435350101292375e-05, "loss": 0.0643, "step": 126710 }, { "epoch": 4.284120490888807, "grad_norm": 2.6413846015930176, "learning_rate": 3.2427609907073454e-05, "loss": 0.059, "step": 126720 }, { "epoch": 4.284458568579059, "grad_norm": 1.606965184211731, "learning_rate": 3.2419870193277105e-05, "loss": 0.0629, "step": 126730 }, { "epoch": 4.284796646269313, "grad_norm": 23.4086856842041, "learning_rate": 3.2412130960114864e-05, "loss": 0.0595, "step": 126740 }, { "epoch": 4.285134723959566, "grad_norm": 5.456451892852783, "learning_rate": 3.2404392207798394e-05, "loss": 0.0605, "step": 126750 }, { "epoch": 4.285472801649819, "grad_norm": 6.580088138580322, "learning_rate": 3.239665393653919e-05, "loss": 0.0731, "step": 126760 }, { "epoch": 4.285810879340072, "grad_norm": 2.7908849716186523, "learning_rate": 3.23889161465489e-05, "loss": 0.0632, "step": 126770 }, { "epoch": 4.286148957030326, "grad_norm": 4.250744342803955, "learning_rate": 3.238117883803898e-05, "loss": 0.066, "step": 126780 }, { "epoch": 4.286487034720579, "grad_norm": 3.3871147632598877, "learning_rate": 3.237344201122103e-05, "loss": 0.066, "step": 126790 }, { "epoch": 4.286825112410832, "grad_norm": 2.418571949005127, "learning_rate": 3.236570566630656e-05, "loss": 0.0641, "step": 126800 }, { "epoch": 4.287163190101086, "grad_norm": 3.8870251178741455, "learning_rate": 3.235796980350707e-05, "loss": 0.0623, "step": 126810 }, { "epoch": 4.287501267791338, "grad_norm": 1.7108536958694458, "learning_rate": 3.235023442303406e-05, "loss": 0.0587, "step": 126820 }, { "epoch": 4.287839345481592, "grad_norm": 4.9415106773376465, "learning_rate": 3.2342499525099e-05, "loss": 0.0677, "step": 126830 }, { "epoch": 4.288177423171845, "grad_norm": 3.6939218044281006, "learning_rate": 3.233476510991338e-05, "loss": 0.06, "step": 126840 }, { "epoch": 4.288515500862098, "grad_norm": 1.8688143491744995, "learning_rate": 3.2327031177688636e-05, "loss": 0.0649, "step": 126850 }, { "epoch": 4.288853578552351, "grad_norm": 5.7348504066467285, "learning_rate": 3.231929772863621e-05, "loss": 0.0643, "step": 126860 }, { "epoch": 4.289191656242605, "grad_norm": 4.917889595031738, "learning_rate": 3.231156476296757e-05, "loss": 0.0578, "step": 126870 }, { "epoch": 4.2895297339328575, "grad_norm": 2.9191486835479736, "learning_rate": 3.2303832280894084e-05, "loss": 0.0555, "step": 126880 }, { "epoch": 4.289867811623111, "grad_norm": 8.434808731079102, "learning_rate": 3.2296100282627184e-05, "loss": 0.0582, "step": 126890 }, { "epoch": 4.2902058893133646, "grad_norm": 4.071129322052002, "learning_rate": 3.2288368768378254e-05, "loss": 0.0622, "step": 126900 }, { "epoch": 4.290543967003617, "grad_norm": 10.039997100830078, "learning_rate": 3.2280637738358666e-05, "loss": 0.0647, "step": 126910 }, { "epoch": 4.290882044693871, "grad_norm": 1.969456672668457, "learning_rate": 3.227290719277978e-05, "loss": 0.0581, "step": 126920 }, { "epoch": 4.291220122384124, "grad_norm": 8.48049259185791, "learning_rate": 3.226517713185294e-05, "loss": 0.058, "step": 126930 }, { "epoch": 4.291558200074377, "grad_norm": 2.259380578994751, "learning_rate": 3.225744755578952e-05, "loss": 0.0614, "step": 126940 }, { "epoch": 4.29189627776463, "grad_norm": 6.276857852935791, "learning_rate": 3.224971846480079e-05, "loss": 0.063, "step": 126950 }, { "epoch": 4.292234355454884, "grad_norm": 1.9792624711990356, "learning_rate": 3.224198985909813e-05, "loss": 0.0626, "step": 126960 }, { "epoch": 4.2925724331451365, "grad_norm": 3.727356433868408, "learning_rate": 3.2234261738892756e-05, "loss": 0.065, "step": 126970 }, { "epoch": 4.29291051083539, "grad_norm": 2.520401954650879, "learning_rate": 3.2226534104396e-05, "loss": 0.0656, "step": 126980 }, { "epoch": 4.2932485885256435, "grad_norm": 6.2982378005981445, "learning_rate": 3.2218806955819144e-05, "loss": 0.0578, "step": 126990 }, { "epoch": 4.293586666215896, "grad_norm": 3.0027103424072266, "learning_rate": 3.221108029337341e-05, "loss": 0.0683, "step": 127000 }, { "epoch": 4.29392474390615, "grad_norm": 4.053765296936035, "learning_rate": 3.220335411727007e-05, "loss": 0.0629, "step": 127010 }, { "epoch": 4.294262821596403, "grad_norm": 2.40116286277771, "learning_rate": 3.219562842772033e-05, "loss": 0.059, "step": 127020 }, { "epoch": 4.294600899286656, "grad_norm": 1.7565239667892456, "learning_rate": 3.2187903224935426e-05, "loss": 0.0622, "step": 127030 }, { "epoch": 4.294938976976909, "grad_norm": 8.261090278625488, "learning_rate": 3.218017850912655e-05, "loss": 0.0684, "step": 127040 }, { "epoch": 4.295277054667163, "grad_norm": 3.4145781993865967, "learning_rate": 3.2172454280504886e-05, "loss": 0.0594, "step": 127050 }, { "epoch": 4.295615132357415, "grad_norm": 4.131454944610596, "learning_rate": 3.216473053928166e-05, "loss": 0.0583, "step": 127060 }, { "epoch": 4.295953210047669, "grad_norm": 11.676361083984375, "learning_rate": 3.215700728566797e-05, "loss": 0.069, "step": 127070 }, { "epoch": 4.2962912877379225, "grad_norm": 2.970506191253662, "learning_rate": 3.214928451987501e-05, "loss": 0.064, "step": 127080 }, { "epoch": 4.296629365428175, "grad_norm": 5.000486850738525, "learning_rate": 3.2141562242113906e-05, "loss": 0.0595, "step": 127090 }, { "epoch": 4.296967443118429, "grad_norm": 3.4636363983154297, "learning_rate": 3.213384045259579e-05, "loss": 0.0623, "step": 127100 }, { "epoch": 4.297305520808682, "grad_norm": 2.4348957538604736, "learning_rate": 3.2126119151531744e-05, "loss": 0.0677, "step": 127110 }, { "epoch": 4.297643598498935, "grad_norm": 9.794581413269043, "learning_rate": 3.211839833913289e-05, "loss": 0.0596, "step": 127120 }, { "epoch": 4.297981676189188, "grad_norm": 2.8055975437164307, "learning_rate": 3.211067801561032e-05, "loss": 0.067, "step": 127130 }, { "epoch": 4.298319753879442, "grad_norm": 1.978790283203125, "learning_rate": 3.210295818117507e-05, "loss": 0.0618, "step": 127140 }, { "epoch": 4.298657831569694, "grad_norm": 3.0540058612823486, "learning_rate": 3.209523883603826e-05, "loss": 0.0618, "step": 127150 }, { "epoch": 4.298995909259948, "grad_norm": 3.726816415786743, "learning_rate": 3.2087519980410854e-05, "loss": 0.0596, "step": 127160 }, { "epoch": 4.299333986950201, "grad_norm": 4.3930792808532715, "learning_rate": 3.2079801614503924e-05, "loss": 0.0612, "step": 127170 }, { "epoch": 4.299672064640454, "grad_norm": 9.140077590942383, "learning_rate": 3.207208373852851e-05, "loss": 0.0627, "step": 127180 }, { "epoch": 4.3000101423307076, "grad_norm": 4.270317554473877, "learning_rate": 3.206436635269557e-05, "loss": 0.065, "step": 127190 }, { "epoch": 4.300348220020961, "grad_norm": 1.8337258100509644, "learning_rate": 3.2056649457216135e-05, "loss": 0.0574, "step": 127200 }, { "epoch": 4.300686297711214, "grad_norm": 2.919293165206909, "learning_rate": 3.204893305230115e-05, "loss": 0.0589, "step": 127210 }, { "epoch": 4.301024375401467, "grad_norm": 2.7934608459472656, "learning_rate": 3.2041217138161605e-05, "loss": 0.0574, "step": 127220 }, { "epoch": 4.301362453091721, "grad_norm": 3.0581417083740234, "learning_rate": 3.2033501715008416e-05, "loss": 0.0663, "step": 127230 }, { "epoch": 4.301700530781973, "grad_norm": 5.664783954620361, "learning_rate": 3.202578678305254e-05, "loss": 0.0591, "step": 127240 }, { "epoch": 4.302038608472227, "grad_norm": 2.8012008666992188, "learning_rate": 3.201807234250493e-05, "loss": 0.0633, "step": 127250 }, { "epoch": 4.30237668616248, "grad_norm": 14.297454833984375, "learning_rate": 3.201035839357643e-05, "loss": 0.0595, "step": 127260 }, { "epoch": 4.302714763852733, "grad_norm": 3.477717638015747, "learning_rate": 3.2002644936478e-05, "loss": 0.0598, "step": 127270 }, { "epoch": 4.3030528415429865, "grad_norm": 5.107539653778076, "learning_rate": 3.199493197142048e-05, "loss": 0.0659, "step": 127280 }, { "epoch": 4.30339091923324, "grad_norm": 1.7298572063446045, "learning_rate": 3.198721949861477e-05, "loss": 0.0604, "step": 127290 }, { "epoch": 4.303728996923493, "grad_norm": 1.5671703815460205, "learning_rate": 3.1979507518271705e-05, "loss": 0.0596, "step": 127300 }, { "epoch": 4.304067074613746, "grad_norm": 3.8034403324127197, "learning_rate": 3.197179603060213e-05, "loss": 0.0578, "step": 127310 }, { "epoch": 4.304405152304, "grad_norm": 1.7329787015914917, "learning_rate": 3.196408503581689e-05, "loss": 0.0579, "step": 127320 }, { "epoch": 4.304743229994252, "grad_norm": 3.9660868644714355, "learning_rate": 3.195637453412678e-05, "loss": 0.0636, "step": 127330 }, { "epoch": 4.305081307684506, "grad_norm": 4.583698749542236, "learning_rate": 3.194866452574262e-05, "loss": 0.0628, "step": 127340 }, { "epoch": 4.305419385374759, "grad_norm": 4.078701496124268, "learning_rate": 3.194095501087517e-05, "loss": 0.0595, "step": 127350 }, { "epoch": 4.305757463065012, "grad_norm": 1.3986566066741943, "learning_rate": 3.193324598973524e-05, "loss": 0.062, "step": 127360 }, { "epoch": 4.3060955407552655, "grad_norm": 2.3245537281036377, "learning_rate": 3.192553746253358e-05, "loss": 0.0735, "step": 127370 }, { "epoch": 4.306433618445519, "grad_norm": 54.10919952392578, "learning_rate": 3.191782942948093e-05, "loss": 0.0696, "step": 127380 }, { "epoch": 4.306771696135772, "grad_norm": 4.187588691711426, "learning_rate": 3.191012189078804e-05, "loss": 0.068, "step": 127390 }, { "epoch": 4.307109773826025, "grad_norm": 5.079721927642822, "learning_rate": 3.190241484666561e-05, "loss": 0.0617, "step": 127400 }, { "epoch": 4.307447851516279, "grad_norm": 3.3747270107269287, "learning_rate": 3.1894708297324373e-05, "loss": 0.0558, "step": 127410 }, { "epoch": 4.307785929206531, "grad_norm": 1.8893285989761353, "learning_rate": 3.188700224297499e-05, "loss": 0.0627, "step": 127420 }, { "epoch": 4.308124006896785, "grad_norm": 2.0561420917510986, "learning_rate": 3.1879296683828164e-05, "loss": 0.0642, "step": 127430 }, { "epoch": 4.308462084587038, "grad_norm": 4.404224872589111, "learning_rate": 3.187159162009458e-05, "loss": 0.0545, "step": 127440 }, { "epoch": 4.308800162277292, "grad_norm": 5.0960469245910645, "learning_rate": 3.186388705198484e-05, "loss": 0.0659, "step": 127450 }, { "epoch": 4.309138239967544, "grad_norm": 4.298086643218994, "learning_rate": 3.1856182979709646e-05, "loss": 0.0705, "step": 127460 }, { "epoch": 4.309476317657798, "grad_norm": 3.053753137588501, "learning_rate": 3.184847940347957e-05, "loss": 0.0626, "step": 127470 }, { "epoch": 4.3098143953480506, "grad_norm": 2.7055084705352783, "learning_rate": 3.1840776323505264e-05, "loss": 0.0642, "step": 127480 }, { "epoch": 4.310152473038304, "grad_norm": 2.8735368251800537, "learning_rate": 3.18330737399973e-05, "loss": 0.0572, "step": 127490 }, { "epoch": 4.310490550728558, "grad_norm": 7.736539363861084, "learning_rate": 3.182537165316627e-05, "loss": 0.0648, "step": 127500 }, { "epoch": 4.310828628418811, "grad_norm": 2.792689561843872, "learning_rate": 3.181767006322277e-05, "loss": 0.058, "step": 127510 }, { "epoch": 4.311166706109064, "grad_norm": 2.5181596279144287, "learning_rate": 3.1809968970377324e-05, "loss": 0.065, "step": 127520 }, { "epoch": 4.311504783799317, "grad_norm": 3.211451530456543, "learning_rate": 3.1802268374840503e-05, "loss": 0.0603, "step": 127530 }, { "epoch": 4.31184286148957, "grad_norm": 6.030186653137207, "learning_rate": 3.179456827682282e-05, "loss": 0.0644, "step": 127540 }, { "epoch": 4.312180939179823, "grad_norm": 5.58892297744751, "learning_rate": 3.178686867653479e-05, "loss": 0.0631, "step": 127550 }, { "epoch": 4.312519016870077, "grad_norm": 3.664511203765869, "learning_rate": 3.177916957418694e-05, "loss": 0.0607, "step": 127560 }, { "epoch": 4.31285709456033, "grad_norm": 1.2103819847106934, "learning_rate": 3.177147096998975e-05, "loss": 0.0603, "step": 127570 }, { "epoch": 4.313195172250583, "grad_norm": 3.172823667526245, "learning_rate": 3.1763772864153705e-05, "loss": 0.0634, "step": 127580 }, { "epoch": 4.3135332499408365, "grad_norm": 2.420517683029175, "learning_rate": 3.175607525688925e-05, "loss": 0.0652, "step": 127590 }, { "epoch": 4.313871327631089, "grad_norm": 3.127838611602783, "learning_rate": 3.174837814840684e-05, "loss": 0.0664, "step": 127600 }, { "epoch": 4.314209405321343, "grad_norm": 1.6917195320129395, "learning_rate": 3.1740681538916914e-05, "loss": 0.0591, "step": 127610 }, { "epoch": 4.314547483011596, "grad_norm": 4.5725555419921875, "learning_rate": 3.1732985428629905e-05, "loss": 0.0674, "step": 127620 }, { "epoch": 4.31488556070185, "grad_norm": 2.2368364334106445, "learning_rate": 3.172528981775621e-05, "loss": 0.0608, "step": 127630 }, { "epoch": 4.315223638392102, "grad_norm": 2.6153764724731445, "learning_rate": 3.17175947065062e-05, "loss": 0.0611, "step": 127640 }, { "epoch": 4.315561716082356, "grad_norm": 3.6867120265960693, "learning_rate": 3.170990009509033e-05, "loss": 0.0674, "step": 127650 }, { "epoch": 4.315899793772609, "grad_norm": 3.887521505355835, "learning_rate": 3.1702205983718874e-05, "loss": 0.0651, "step": 127660 }, { "epoch": 4.316237871462862, "grad_norm": 3.27738881111145, "learning_rate": 3.169451237260227e-05, "loss": 0.0659, "step": 127670 }, { "epoch": 4.3165759491531155, "grad_norm": 1.4679709672927856, "learning_rate": 3.168681926195079e-05, "loss": 0.0642, "step": 127680 }, { "epoch": 4.316914026843369, "grad_norm": 2.5028679370880127, "learning_rate": 3.16791266519748e-05, "loss": 0.0599, "step": 127690 }, { "epoch": 4.317252104533622, "grad_norm": 1.436031699180603, "learning_rate": 3.1671434542884615e-05, "loss": 0.064, "step": 127700 }, { "epoch": 4.317590182223875, "grad_norm": 2.173978090286255, "learning_rate": 3.166374293489052e-05, "loss": 0.0679, "step": 127710 }, { "epoch": 4.317928259914129, "grad_norm": 3.1755754947662354, "learning_rate": 3.1656051828202804e-05, "loss": 0.0607, "step": 127720 }, { "epoch": 4.318266337604381, "grad_norm": 1.7959483861923218, "learning_rate": 3.164836122303173e-05, "loss": 0.063, "step": 127730 }, { "epoch": 4.318604415294635, "grad_norm": 6.784398555755615, "learning_rate": 3.1640671119587565e-05, "loss": 0.0676, "step": 127740 }, { "epoch": 4.318942492984888, "grad_norm": 3.203616142272949, "learning_rate": 3.1632981518080575e-05, "loss": 0.0658, "step": 127750 }, { "epoch": 4.319280570675141, "grad_norm": 5.043331623077393, "learning_rate": 3.162529241872094e-05, "loss": 0.0678, "step": 127760 }, { "epoch": 4.319618648365394, "grad_norm": 2.733161449432373, "learning_rate": 3.1617603821718936e-05, "loss": 0.068, "step": 127770 }, { "epoch": 4.319956726055648, "grad_norm": 5.485974311828613, "learning_rate": 3.160991572728472e-05, "loss": 0.0639, "step": 127780 }, { "epoch": 4.320294803745901, "grad_norm": 1.4796937704086304, "learning_rate": 3.160222813562852e-05, "loss": 0.0549, "step": 127790 }, { "epoch": 4.320632881436154, "grad_norm": 2.3008358478546143, "learning_rate": 3.159454104696047e-05, "loss": 0.0715, "step": 127800 }, { "epoch": 4.320970959126408, "grad_norm": 4.425876140594482, "learning_rate": 3.158685446149076e-05, "loss": 0.0701, "step": 127810 }, { "epoch": 4.32130903681666, "grad_norm": 5.006903171539307, "learning_rate": 3.157916837942954e-05, "loss": 0.0612, "step": 127820 }, { "epoch": 4.321647114506914, "grad_norm": 4.532048225402832, "learning_rate": 3.157148280098692e-05, "loss": 0.0666, "step": 127830 }, { "epoch": 4.321985192197167, "grad_norm": 4.158028602600098, "learning_rate": 3.156379772637305e-05, "loss": 0.0646, "step": 127840 }, { "epoch": 4.32232326988742, "grad_norm": 3.518234968185425, "learning_rate": 3.1556113155798006e-05, "loss": 0.0657, "step": 127850 }, { "epoch": 4.322661347577673, "grad_norm": 2.510113477706909, "learning_rate": 3.1548429089471925e-05, "loss": 0.0627, "step": 127860 }, { "epoch": 4.322999425267927, "grad_norm": 2.2422149181365967, "learning_rate": 3.154074552760484e-05, "loss": 0.0599, "step": 127870 }, { "epoch": 4.3233375029581795, "grad_norm": 9.142190933227539, "learning_rate": 3.1533062470406836e-05, "loss": 0.0679, "step": 127880 }, { "epoch": 4.323675580648433, "grad_norm": 6.6137189865112305, "learning_rate": 3.1525379918087984e-05, "loss": 0.0566, "step": 127890 }, { "epoch": 4.324013658338687, "grad_norm": 5.794495105743408, "learning_rate": 3.151769787085829e-05, "loss": 0.0634, "step": 127900 }, { "epoch": 4.324351736028939, "grad_norm": 2.059530258178711, "learning_rate": 3.151001632892781e-05, "loss": 0.0598, "step": 127910 }, { "epoch": 4.324689813719193, "grad_norm": 2.0671281814575195, "learning_rate": 3.150233529250652e-05, "loss": 0.0654, "step": 127920 }, { "epoch": 4.325027891409446, "grad_norm": 3.077897310256958, "learning_rate": 3.149465476180442e-05, "loss": 0.0679, "step": 127930 }, { "epoch": 4.325365969099699, "grad_norm": 3.187565803527832, "learning_rate": 3.148697473703155e-05, "loss": 0.0711, "step": 127940 }, { "epoch": 4.325704046789952, "grad_norm": 2.4980435371398926, "learning_rate": 3.14792952183978e-05, "loss": 0.0585, "step": 127950 }, { "epoch": 4.326042124480206, "grad_norm": 2.353686571121216, "learning_rate": 3.1471616206113185e-05, "loss": 0.0631, "step": 127960 }, { "epoch": 4.3263802021704585, "grad_norm": 3.2727131843566895, "learning_rate": 3.14639377003876e-05, "loss": 0.0572, "step": 127970 }, { "epoch": 4.326718279860712, "grad_norm": 2.1881234645843506, "learning_rate": 3.1456259701431006e-05, "loss": 0.0596, "step": 127980 }, { "epoch": 4.3270563575509655, "grad_norm": 1.9073847532272339, "learning_rate": 3.1448582209453304e-05, "loss": 0.068, "step": 127990 }, { "epoch": 4.327394435241218, "grad_norm": 1.481218695640564, "learning_rate": 3.144090522466438e-05, "loss": 0.0621, "step": 128000 }, { "epoch": 4.327732512931472, "grad_norm": 18.726465225219727, "learning_rate": 3.143322874727417e-05, "loss": 0.0592, "step": 128010 }, { "epoch": 4.328070590621725, "grad_norm": 2.9135053157806396, "learning_rate": 3.1425552777492486e-05, "loss": 0.0634, "step": 128020 }, { "epoch": 4.328408668311978, "grad_norm": 2.2346384525299072, "learning_rate": 3.141787731552922e-05, "loss": 0.0641, "step": 128030 }, { "epoch": 4.328746746002231, "grad_norm": 2.6640632152557373, "learning_rate": 3.1410202361594196e-05, "loss": 0.0571, "step": 128040 }, { "epoch": 4.329084823692485, "grad_norm": 3.347663164138794, "learning_rate": 3.140252791589727e-05, "loss": 0.067, "step": 128050 }, { "epoch": 4.329422901382737, "grad_norm": 1.3158515691757202, "learning_rate": 3.139485397864822e-05, "loss": 0.0631, "step": 128060 }, { "epoch": 4.329760979072991, "grad_norm": 4.897714138031006, "learning_rate": 3.138718055005689e-05, "loss": 0.0677, "step": 128070 }, { "epoch": 4.3300990567632445, "grad_norm": 2.2576663494110107, "learning_rate": 3.137950763033305e-05, "loss": 0.0643, "step": 128080 }, { "epoch": 4.330437134453497, "grad_norm": 3.2788033485412598, "learning_rate": 3.137183521968647e-05, "loss": 0.0584, "step": 128090 }, { "epoch": 4.330775212143751, "grad_norm": 1.6085779666900635, "learning_rate": 3.1364163318326934e-05, "loss": 0.07, "step": 128100 }, { "epoch": 4.331113289834004, "grad_norm": 3.1899211406707764, "learning_rate": 3.135649192646416e-05, "loss": 0.063, "step": 128110 }, { "epoch": 4.331451367524257, "grad_norm": 3.3528969287872314, "learning_rate": 3.13488210443079e-05, "loss": 0.0563, "step": 128120 }, { "epoch": 4.33178944521451, "grad_norm": 2.8677053451538086, "learning_rate": 3.134115067206786e-05, "loss": 0.0603, "step": 128130 }, { "epoch": 4.332127522904764, "grad_norm": 3.354743719100952, "learning_rate": 3.133348080995373e-05, "loss": 0.0614, "step": 128140 }, { "epoch": 4.332465600595016, "grad_norm": 4.6579437255859375, "learning_rate": 3.132581145817526e-05, "loss": 0.0565, "step": 128150 }, { "epoch": 4.33280367828527, "grad_norm": 2.2384026050567627, "learning_rate": 3.131814261694206e-05, "loss": 0.0635, "step": 128160 }, { "epoch": 4.333141755975523, "grad_norm": 7.008220195770264, "learning_rate": 3.131047428646384e-05, "loss": 0.0498, "step": 128170 }, { "epoch": 4.333479833665776, "grad_norm": 1.9605876207351685, "learning_rate": 3.130280646695022e-05, "loss": 0.0555, "step": 128180 }, { "epoch": 4.33381791135603, "grad_norm": 1.2019966840744019, "learning_rate": 3.129513915861084e-05, "loss": 0.0549, "step": 128190 }, { "epoch": 4.334155989046283, "grad_norm": 2.209913730621338, "learning_rate": 3.128747236165535e-05, "loss": 0.0674, "step": 128200 }, { "epoch": 4.334494066736536, "grad_norm": 2.775224447250366, "learning_rate": 3.127980607629333e-05, "loss": 0.057, "step": 128210 }, { "epoch": 4.334832144426789, "grad_norm": 2.404534339904785, "learning_rate": 3.127214030273438e-05, "loss": 0.0653, "step": 128220 }, { "epoch": 4.335170222117043, "grad_norm": 1.18602454662323, "learning_rate": 3.126447504118807e-05, "loss": 0.0573, "step": 128230 }, { "epoch": 4.335508299807295, "grad_norm": 5.480309963226318, "learning_rate": 3.125681029186398e-05, "loss": 0.0602, "step": 128240 }, { "epoch": 4.335846377497549, "grad_norm": 1.9854310750961304, "learning_rate": 3.124914605497164e-05, "loss": 0.0637, "step": 128250 }, { "epoch": 4.336184455187802, "grad_norm": 2.300708055496216, "learning_rate": 3.1241482330720606e-05, "loss": 0.0662, "step": 128260 }, { "epoch": 4.336522532878055, "grad_norm": 6.376862049102783, "learning_rate": 3.123381911932042e-05, "loss": 0.0567, "step": 128270 }, { "epoch": 4.3368606105683085, "grad_norm": 1.7560515403747559, "learning_rate": 3.122615642098055e-05, "loss": 0.0692, "step": 128280 }, { "epoch": 4.337198688258562, "grad_norm": 1.940779209136963, "learning_rate": 3.1218494235910524e-05, "loss": 0.0666, "step": 128290 }, { "epoch": 4.337536765948815, "grad_norm": 37.02724838256836, "learning_rate": 3.1210832564319794e-05, "loss": 0.0621, "step": 128300 }, { "epoch": 4.337874843639068, "grad_norm": 1.6029068231582642, "learning_rate": 3.120317140641786e-05, "loss": 0.0597, "step": 128310 }, { "epoch": 4.338212921329322, "grad_norm": 19.456336975097656, "learning_rate": 3.119551076241415e-05, "loss": 0.0706, "step": 128320 }, { "epoch": 4.338550999019574, "grad_norm": 2.0918776988983154, "learning_rate": 3.11878506325181e-05, "loss": 0.0713, "step": 128330 }, { "epoch": 4.338889076709828, "grad_norm": 1.7033723592758179, "learning_rate": 3.118019101693917e-05, "loss": 0.0594, "step": 128340 }, { "epoch": 4.339227154400081, "grad_norm": 1.6247899532318115, "learning_rate": 3.1172531915886716e-05, "loss": 0.0611, "step": 128350 }, { "epoch": 4.339565232090334, "grad_norm": 31.778093338012695, "learning_rate": 3.1164873329570196e-05, "loss": 0.0629, "step": 128360 }, { "epoch": 4.3399033097805875, "grad_norm": 2.6632497310638428, "learning_rate": 3.115721525819894e-05, "loss": 0.0664, "step": 128370 }, { "epoch": 4.340241387470841, "grad_norm": 2.8935158252716064, "learning_rate": 3.114955770198234e-05, "loss": 0.0646, "step": 128380 }, { "epoch": 4.340579465161094, "grad_norm": 5.160099029541016, "learning_rate": 3.1141900661129775e-05, "loss": 0.0629, "step": 128390 }, { "epoch": 4.340917542851347, "grad_norm": 8.066012382507324, "learning_rate": 3.1134244135850544e-05, "loss": 0.055, "step": 128400 }, { "epoch": 4.341255620541601, "grad_norm": 1.73219895362854, "learning_rate": 3.112658812635401e-05, "loss": 0.0607, "step": 128410 }, { "epoch": 4.341593698231854, "grad_norm": 1.7200490236282349, "learning_rate": 3.111893263284944e-05, "loss": 0.0667, "step": 128420 }, { "epoch": 4.341931775922107, "grad_norm": 2.2201359272003174, "learning_rate": 3.111127765554618e-05, "loss": 0.0645, "step": 128430 }, { "epoch": 4.34226985361236, "grad_norm": 2.9680798053741455, "learning_rate": 3.110362319465348e-05, "loss": 0.059, "step": 128440 }, { "epoch": 4.342607931302613, "grad_norm": 2.710545539855957, "learning_rate": 3.1095969250380605e-05, "loss": 0.0651, "step": 128450 }, { "epoch": 4.342946008992866, "grad_norm": 1.6151394844055176, "learning_rate": 3.1088315822936865e-05, "loss": 0.0623, "step": 128460 }, { "epoch": 4.34328408668312, "grad_norm": 3.072028875350952, "learning_rate": 3.108066291253144e-05, "loss": 0.0617, "step": 128470 }, { "epoch": 4.3436221643733735, "grad_norm": 3.190732955932617, "learning_rate": 3.10730105193736e-05, "loss": 0.064, "step": 128480 }, { "epoch": 4.343960242063626, "grad_norm": 4.84998083114624, "learning_rate": 3.106535864367253e-05, "loss": 0.0626, "step": 128490 }, { "epoch": 4.34429831975388, "grad_norm": 3.5804855823516846, "learning_rate": 3.105770728563745e-05, "loss": 0.0654, "step": 128500 }, { "epoch": 4.344636397444132, "grad_norm": 7.928800106048584, "learning_rate": 3.1050056445477526e-05, "loss": 0.0636, "step": 128510 }, { "epoch": 4.344974475134386, "grad_norm": 1.1301673650741577, "learning_rate": 3.104240612340195e-05, "loss": 0.0671, "step": 128520 }, { "epoch": 4.345312552824639, "grad_norm": 1.9943287372589111, "learning_rate": 3.103475631961987e-05, "loss": 0.0668, "step": 128530 }, { "epoch": 4.345650630514893, "grad_norm": 4.816030025482178, "learning_rate": 3.102710703434042e-05, "loss": 0.0595, "step": 128540 }, { "epoch": 4.345988708205145, "grad_norm": 1.7016187906265259, "learning_rate": 3.101945826777274e-05, "loss": 0.0624, "step": 128550 }, { "epoch": 4.346326785895399, "grad_norm": 1.4278695583343506, "learning_rate": 3.1011810020125924e-05, "loss": 0.0642, "step": 128560 }, { "epoch": 4.3466648635856515, "grad_norm": 4.527178764343262, "learning_rate": 3.10041622916091e-05, "loss": 0.0614, "step": 128570 }, { "epoch": 4.347002941275905, "grad_norm": 1.639775037765503, "learning_rate": 3.099651508243136e-05, "loss": 0.0704, "step": 128580 }, { "epoch": 4.3473410189661585, "grad_norm": 1.8754279613494873, "learning_rate": 3.098886839280173e-05, "loss": 0.0652, "step": 128590 }, { "epoch": 4.347679096656412, "grad_norm": 1.4170129299163818, "learning_rate": 3.0981222222929315e-05, "loss": 0.061, "step": 128600 }, { "epoch": 4.348017174346665, "grad_norm": 2.8751559257507324, "learning_rate": 3.0973576573023134e-05, "loss": 0.0654, "step": 128610 }, { "epoch": 4.348355252036918, "grad_norm": 1.0038686990737915, "learning_rate": 3.0965931443292226e-05, "loss": 0.0603, "step": 128620 }, { "epoch": 4.348693329727172, "grad_norm": 2.4829649925231934, "learning_rate": 3.095828683394559e-05, "loss": 0.0579, "step": 128630 }, { "epoch": 4.349031407417424, "grad_norm": 2.312267303466797, "learning_rate": 3.0950642745192234e-05, "loss": 0.066, "step": 128640 }, { "epoch": 4.349369485107678, "grad_norm": 1.5570248365402222, "learning_rate": 3.0942999177241186e-05, "loss": 0.0635, "step": 128650 }, { "epoch": 4.349707562797931, "grad_norm": 4.466812610626221, "learning_rate": 3.0935356130301336e-05, "loss": 0.0632, "step": 128660 }, { "epoch": 4.350045640488184, "grad_norm": 2.537135124206543, "learning_rate": 3.092771360458172e-05, "loss": 0.0556, "step": 128670 }, { "epoch": 4.3503837181784375, "grad_norm": 1.4612805843353271, "learning_rate": 3.092007160029122e-05, "loss": 0.0615, "step": 128680 }, { "epoch": 4.350721795868691, "grad_norm": 1.2909510135650635, "learning_rate": 3.0912430117638824e-05, "loss": 0.062, "step": 128690 }, { "epoch": 4.351059873558944, "grad_norm": 16.19412612915039, "learning_rate": 3.0904789156833394e-05, "loss": 0.0581, "step": 128700 }, { "epoch": 4.351397951249197, "grad_norm": 4.514779567718506, "learning_rate": 3.089714871808387e-05, "loss": 0.0593, "step": 128710 }, { "epoch": 4.351736028939451, "grad_norm": 3.564624786376953, "learning_rate": 3.088950880159912e-05, "loss": 0.0643, "step": 128720 }, { "epoch": 4.352074106629703, "grad_norm": 1.5759261846542358, "learning_rate": 3.0881869407588014e-05, "loss": 0.0584, "step": 128730 }, { "epoch": 4.352412184319957, "grad_norm": 1.0610662698745728, "learning_rate": 3.087423053625943e-05, "loss": 0.0586, "step": 128740 }, { "epoch": 4.35275026201021, "grad_norm": 2.1450417041778564, "learning_rate": 3.086659218782218e-05, "loss": 0.0609, "step": 128750 }, { "epoch": 4.353088339700463, "grad_norm": 1.318472146987915, "learning_rate": 3.0858954362485105e-05, "loss": 0.0568, "step": 128760 }, { "epoch": 4.3534264173907165, "grad_norm": 1.7495224475860596, "learning_rate": 3.0851317060457045e-05, "loss": 0.0727, "step": 128770 }, { "epoch": 4.35376449508097, "grad_norm": 139.12168884277344, "learning_rate": 3.084368028194677e-05, "loss": 0.0589, "step": 128780 }, { "epoch": 4.354102572771223, "grad_norm": 1.9769268035888672, "learning_rate": 3.08360440271631e-05, "loss": 0.0638, "step": 128790 }, { "epoch": 4.354440650461476, "grad_norm": 2.034059524536133, "learning_rate": 3.082840829631477e-05, "loss": 0.0686, "step": 128800 }, { "epoch": 4.35477872815173, "grad_norm": 1.7159090042114258, "learning_rate": 3.0820773089610564e-05, "loss": 0.0622, "step": 128810 }, { "epoch": 4.355116805841982, "grad_norm": 2.7827274799346924, "learning_rate": 3.081313840725921e-05, "loss": 0.0631, "step": 128820 }, { "epoch": 4.355454883532236, "grad_norm": 1.6010736227035522, "learning_rate": 3.0805504249469444e-05, "loss": 0.0559, "step": 128830 }, { "epoch": 4.355792961222489, "grad_norm": 1.7837481498718262, "learning_rate": 3.079787061644999e-05, "loss": 0.066, "step": 128840 }, { "epoch": 4.356131038912742, "grad_norm": 3.34452486038208, "learning_rate": 3.079023750840952e-05, "loss": 0.0636, "step": 128850 }, { "epoch": 4.356469116602995, "grad_norm": 3.776167869567871, "learning_rate": 3.078260492555677e-05, "loss": 0.0636, "step": 128860 }, { "epoch": 4.356807194293249, "grad_norm": 2.255206823348999, "learning_rate": 3.077497286810035e-05, "loss": 0.0625, "step": 128870 }, { "epoch": 4.3571452719835015, "grad_norm": 1.1761629581451416, "learning_rate": 3.076734133624898e-05, "loss": 0.0644, "step": 128880 }, { "epoch": 4.357483349673755, "grad_norm": 1.672555923461914, "learning_rate": 3.075971033021124e-05, "loss": 0.0567, "step": 128890 }, { "epoch": 4.357821427364009, "grad_norm": 1.7320526838302612, "learning_rate": 3.0752079850195805e-05, "loss": 0.0534, "step": 128900 }, { "epoch": 4.358159505054261, "grad_norm": 3.844299793243408, "learning_rate": 3.074444989641128e-05, "loss": 0.0604, "step": 128910 }, { "epoch": 4.358497582744515, "grad_norm": 9.835800170898438, "learning_rate": 3.0736820469066255e-05, "loss": 0.0638, "step": 128920 }, { "epoch": 4.358835660434768, "grad_norm": 2.2817273139953613, "learning_rate": 3.072919156836933e-05, "loss": 0.0557, "step": 128930 }, { "epoch": 4.359173738125021, "grad_norm": 3.4220199584960938, "learning_rate": 3.072156319452906e-05, "loss": 0.061, "step": 128940 }, { "epoch": 4.359511815815274, "grad_norm": 2.0658721923828125, "learning_rate": 3.0713935347754e-05, "loss": 0.0646, "step": 128950 }, { "epoch": 4.359849893505528, "grad_norm": 3.0793955326080322, "learning_rate": 3.0706308028252726e-05, "loss": 0.0663, "step": 128960 }, { "epoch": 4.3601879711957805, "grad_norm": 2.3638150691986084, "learning_rate": 3.0698681236233714e-05, "loss": 0.0579, "step": 128970 }, { "epoch": 4.360526048886034, "grad_norm": 1.6367303133010864, "learning_rate": 3.0691054971905535e-05, "loss": 0.0592, "step": 128980 }, { "epoch": 4.3608641265762875, "grad_norm": 3.765115976333618, "learning_rate": 3.068342923547665e-05, "loss": 0.0665, "step": 128990 }, { "epoch": 4.36120220426654, "grad_norm": 2.1022164821624756, "learning_rate": 3.067580402715555e-05, "loss": 0.0532, "step": 129000 }, { "epoch": 4.361540281956794, "grad_norm": 1.7020796537399292, "learning_rate": 3.06681793471507e-05, "loss": 0.0668, "step": 129010 }, { "epoch": 4.361878359647047, "grad_norm": 2.3965835571289062, "learning_rate": 3.066055519567057e-05, "loss": 0.0633, "step": 129020 }, { "epoch": 4.3622164373373, "grad_norm": 1.7669674158096313, "learning_rate": 3.06529315729236e-05, "loss": 0.0656, "step": 129030 }, { "epoch": 4.362554515027553, "grad_norm": 1.0977030992507935, "learning_rate": 3.064530847911821e-05, "loss": 0.0595, "step": 129040 }, { "epoch": 4.362892592717807, "grad_norm": 1.7805938720703125, "learning_rate": 3.063768591446282e-05, "loss": 0.0629, "step": 129050 }, { "epoch": 4.3632306704080595, "grad_norm": 1.1180098056793213, "learning_rate": 3.06300638791658e-05, "loss": 0.0608, "step": 129060 }, { "epoch": 4.363568748098313, "grad_norm": 1.3987436294555664, "learning_rate": 3.06224423734356e-05, "loss": 0.0569, "step": 129070 }, { "epoch": 4.3639068257885665, "grad_norm": 2.1009953022003174, "learning_rate": 3.06148213974805e-05, "loss": 0.0653, "step": 129080 }, { "epoch": 4.364244903478819, "grad_norm": 2.982910394668579, "learning_rate": 3.060720095150892e-05, "loss": 0.0633, "step": 129090 }, { "epoch": 4.364582981169073, "grad_norm": 0.8815516233444214, "learning_rate": 3.0599581035729184e-05, "loss": 0.0616, "step": 129100 }, { "epoch": 4.364921058859326, "grad_norm": 1.6460812091827393, "learning_rate": 3.059196165034961e-05, "loss": 0.0624, "step": 129110 }, { "epoch": 4.365259136549579, "grad_norm": 4.769246578216553, "learning_rate": 3.058434279557852e-05, "loss": 0.0624, "step": 129120 }, { "epoch": 4.365597214239832, "grad_norm": 1.5039658546447754, "learning_rate": 3.057672447162419e-05, "loss": 0.0574, "step": 129130 }, { "epoch": 4.365935291930086, "grad_norm": 1.775173544883728, "learning_rate": 3.0569106678694915e-05, "loss": 0.0633, "step": 129140 }, { "epoch": 4.366273369620338, "grad_norm": 2.0028157234191895, "learning_rate": 3.056148941699899e-05, "loss": 0.0619, "step": 129150 }, { "epoch": 4.366611447310592, "grad_norm": 1.9378856420516968, "learning_rate": 3.055387268674461e-05, "loss": 0.066, "step": 129160 }, { "epoch": 4.366949525000845, "grad_norm": 2.79299259185791, "learning_rate": 3.054625648814007e-05, "loss": 0.0617, "step": 129170 }, { "epoch": 4.367287602691098, "grad_norm": 1.5657604932785034, "learning_rate": 3.053864082139355e-05, "loss": 0.0599, "step": 129180 }, { "epoch": 4.367625680381352, "grad_norm": 3.9209070205688477, "learning_rate": 3.053102568671329e-05, "loss": 0.0589, "step": 129190 }, { "epoch": 4.367963758071605, "grad_norm": 1.7890056371688843, "learning_rate": 3.052341108430746e-05, "loss": 0.0661, "step": 129200 }, { "epoch": 4.368301835761858, "grad_norm": 1.8553036451339722, "learning_rate": 3.051579701438426e-05, "loss": 0.0643, "step": 129210 }, { "epoch": 4.368639913452111, "grad_norm": 2.608973503112793, "learning_rate": 3.0508183477151864e-05, "loss": 0.0606, "step": 129220 }, { "epoch": 4.368977991142365, "grad_norm": 199.91065979003906, "learning_rate": 3.0500570472818398e-05, "loss": 0.066, "step": 129230 }, { "epoch": 4.369316068832617, "grad_norm": 2.8373160362243652, "learning_rate": 3.049295800159202e-05, "loss": 0.058, "step": 129240 }, { "epoch": 4.369654146522871, "grad_norm": 1.498258352279663, "learning_rate": 3.0485346063680825e-05, "loss": 0.0618, "step": 129250 }, { "epoch": 4.369992224213124, "grad_norm": 1.5613305568695068, "learning_rate": 3.0477734659292957e-05, "loss": 0.0631, "step": 129260 }, { "epoch": 4.370330301903377, "grad_norm": 1.5217324495315552, "learning_rate": 3.0470123788636457e-05, "loss": 0.0627, "step": 129270 }, { "epoch": 4.3706683795936305, "grad_norm": 1.2658653259277344, "learning_rate": 3.046251345191945e-05, "loss": 0.058, "step": 129280 }, { "epoch": 4.371006457283884, "grad_norm": 1.58878755569458, "learning_rate": 3.0454903649349996e-05, "loss": 0.0535, "step": 129290 }, { "epoch": 4.371344534974137, "grad_norm": 2.802992820739746, "learning_rate": 3.0447294381136128e-05, "loss": 0.0576, "step": 129300 }, { "epoch": 4.37168261266439, "grad_norm": 1.5270887613296509, "learning_rate": 3.0439685647485895e-05, "loss": 0.0671, "step": 129310 }, { "epoch": 4.372020690354644, "grad_norm": 2.6274006366729736, "learning_rate": 3.0432077448607294e-05, "loss": 0.0583, "step": 129320 }, { "epoch": 4.372358768044896, "grad_norm": 1.1916592121124268, "learning_rate": 3.0424469784708353e-05, "loss": 0.0587, "step": 129330 }, { "epoch": 4.37269684573515, "grad_norm": 1.3727558851242065, "learning_rate": 3.0416862655997046e-05, "loss": 0.0604, "step": 129340 }, { "epoch": 4.373034923425403, "grad_norm": 1.354479193687439, "learning_rate": 3.0409256062681353e-05, "loss": 0.062, "step": 129350 }, { "epoch": 4.373373001115656, "grad_norm": 1.7475156784057617, "learning_rate": 3.040165000496926e-05, "loss": 0.0589, "step": 129360 }, { "epoch": 4.3737110788059095, "grad_norm": 2.0150251388549805, "learning_rate": 3.0394044483068674e-05, "loss": 0.055, "step": 129370 }, { "epoch": 4.374049156496163, "grad_norm": 4.464443683624268, "learning_rate": 3.038643949718757e-05, "loss": 0.0609, "step": 129380 }, { "epoch": 4.3743872341864165, "grad_norm": 1.2003253698349, "learning_rate": 3.0378835047533805e-05, "loss": 0.0658, "step": 129390 }, { "epoch": 4.374725311876669, "grad_norm": 1.9527827501296997, "learning_rate": 3.0371231134315335e-05, "loss": 0.0593, "step": 129400 }, { "epoch": 4.375063389566923, "grad_norm": 2.0335047245025635, "learning_rate": 3.0363627757740044e-05, "loss": 0.06, "step": 129410 }, { "epoch": 4.375401467257175, "grad_norm": 2.1189188957214355, "learning_rate": 3.035602491801578e-05, "loss": 0.0609, "step": 129420 }, { "epoch": 4.375739544947429, "grad_norm": 1.650436282157898, "learning_rate": 3.0348422615350425e-05, "loss": 0.0585, "step": 129430 }, { "epoch": 4.376077622637682, "grad_norm": 2.229982614517212, "learning_rate": 3.03408208499518e-05, "loss": 0.0595, "step": 129440 }, { "epoch": 4.376415700327936, "grad_norm": 0.6897432208061218, "learning_rate": 3.0333219622027764e-05, "loss": 0.0538, "step": 129450 }, { "epoch": 4.376753778018188, "grad_norm": 6.036888122558594, "learning_rate": 3.0325618931786104e-05, "loss": 0.0557, "step": 129460 }, { "epoch": 4.377091855708442, "grad_norm": 3.1264827251434326, "learning_rate": 3.0318018779434615e-05, "loss": 0.06, "step": 129470 }, { "epoch": 4.377429933398695, "grad_norm": 2.910421133041382, "learning_rate": 3.0310419165181124e-05, "loss": 0.059, "step": 129480 }, { "epoch": 4.377768011088948, "grad_norm": 2.09312105178833, "learning_rate": 3.0302820089233364e-05, "loss": 0.0587, "step": 129490 }, { "epoch": 4.378106088779202, "grad_norm": 2.8329224586486816, "learning_rate": 3.0295221551799114e-05, "loss": 0.0646, "step": 129500 }, { "epoch": 4.378444166469455, "grad_norm": 2.2472784519195557, "learning_rate": 3.0287623553086088e-05, "loss": 0.0573, "step": 129510 }, { "epoch": 4.378782244159708, "grad_norm": 1.4907691478729248, "learning_rate": 3.0280026093302044e-05, "loss": 0.0594, "step": 129520 }, { "epoch": 4.379120321849961, "grad_norm": 2.4875388145446777, "learning_rate": 3.027242917265467e-05, "loss": 0.057, "step": 129530 }, { "epoch": 4.379458399540214, "grad_norm": 2.0165607929229736, "learning_rate": 3.0264832791351667e-05, "loss": 0.0649, "step": 129540 }, { "epoch": 4.379796477230467, "grad_norm": 1.562097191810608, "learning_rate": 3.025723694960073e-05, "loss": 0.0576, "step": 129550 }, { "epoch": 4.380134554920721, "grad_norm": 1.180854320526123, "learning_rate": 3.0249641647609507e-05, "loss": 0.055, "step": 129560 }, { "epoch": 4.380472632610974, "grad_norm": 4.140839099884033, "learning_rate": 3.0242046885585684e-05, "loss": 0.0603, "step": 129570 }, { "epoch": 4.380810710301227, "grad_norm": 4.1622796058654785, "learning_rate": 3.0234452663736846e-05, "loss": 0.0605, "step": 129580 }, { "epoch": 4.3811487879914806, "grad_norm": 2.75252628326416, "learning_rate": 3.0226858982270657e-05, "loss": 0.0617, "step": 129590 }, { "epoch": 4.381486865681734, "grad_norm": 3.7689456939697266, "learning_rate": 3.0219265841394728e-05, "loss": 0.0652, "step": 129600 }, { "epoch": 4.381824943371987, "grad_norm": 2.6772820949554443, "learning_rate": 3.021167324131663e-05, "loss": 0.0609, "step": 129610 }, { "epoch": 4.38216302106224, "grad_norm": 1.743660569190979, "learning_rate": 3.020408118224396e-05, "loss": 0.0646, "step": 129620 }, { "epoch": 4.382501098752494, "grad_norm": 2.2081592082977295, "learning_rate": 3.019648966438426e-05, "loss": 0.0583, "step": 129630 }, { "epoch": 4.382839176442746, "grad_norm": 3.112281322479248, "learning_rate": 3.0188898687945106e-05, "loss": 0.059, "step": 129640 }, { "epoch": 4.383177254133, "grad_norm": 1.2787853479385376, "learning_rate": 3.018130825313401e-05, "loss": 0.0565, "step": 129650 }, { "epoch": 4.383515331823253, "grad_norm": 8.655645370483398, "learning_rate": 3.017371836015849e-05, "loss": 0.0597, "step": 129660 }, { "epoch": 4.383853409513506, "grad_norm": 1.9917584657669067, "learning_rate": 3.0166129009226085e-05, "loss": 0.0591, "step": 129670 }, { "epoch": 4.3841914872037595, "grad_norm": 0.9996249675750732, "learning_rate": 3.0158540200544238e-05, "loss": 0.0631, "step": 129680 }, { "epoch": 4.384529564894013, "grad_norm": 1.354905128479004, "learning_rate": 3.015095193432047e-05, "loss": 0.058, "step": 129690 }, { "epoch": 4.384867642584266, "grad_norm": 1.7153912782669067, "learning_rate": 3.0143364210762202e-05, "loss": 0.0651, "step": 129700 }, { "epoch": 4.385205720274519, "grad_norm": 1.1835259199142456, "learning_rate": 3.0135777030076918e-05, "loss": 0.0599, "step": 129710 }, { "epoch": 4.385543797964773, "grad_norm": 2.41151762008667, "learning_rate": 3.012819039247201e-05, "loss": 0.0607, "step": 129720 }, { "epoch": 4.385881875655025, "grad_norm": 1.8331822156906128, "learning_rate": 3.0120604298154908e-05, "loss": 0.0572, "step": 129730 }, { "epoch": 4.386219953345279, "grad_norm": 2.513364553451538, "learning_rate": 3.0113018747333033e-05, "loss": 0.0624, "step": 129740 }, { "epoch": 4.386558031035532, "grad_norm": 3.558570384979248, "learning_rate": 3.010543374021374e-05, "loss": 0.0615, "step": 129750 }, { "epoch": 4.386896108725785, "grad_norm": 2.7642412185668945, "learning_rate": 3.009784927700443e-05, "loss": 0.0621, "step": 129760 }, { "epoch": 4.3872341864160385, "grad_norm": 1.8281365633010864, "learning_rate": 3.0090265357912416e-05, "loss": 0.0624, "step": 129770 }, { "epoch": 4.387572264106292, "grad_norm": 3.352837085723877, "learning_rate": 3.0082681983145084e-05, "loss": 0.0585, "step": 129780 }, { "epoch": 4.387910341796545, "grad_norm": 2.6248679161071777, "learning_rate": 3.007509915290975e-05, "loss": 0.0556, "step": 129790 }, { "epoch": 4.388248419486798, "grad_norm": 3.405498504638672, "learning_rate": 3.0067516867413715e-05, "loss": 0.0623, "step": 129800 }, { "epoch": 4.388586497177052, "grad_norm": 4.0201215744018555, "learning_rate": 3.005993512686428e-05, "loss": 0.0618, "step": 129810 }, { "epoch": 4.388924574867304, "grad_norm": 2.070521593093872, "learning_rate": 3.0052353931468724e-05, "loss": 0.0577, "step": 129820 }, { "epoch": 4.389262652557558, "grad_norm": 2.7898683547973633, "learning_rate": 3.004477328143433e-05, "loss": 0.0583, "step": 129830 }, { "epoch": 4.389600730247811, "grad_norm": 1.5520423650741577, "learning_rate": 3.0037193176968326e-05, "loss": 0.0624, "step": 129840 }, { "epoch": 4.389938807938064, "grad_norm": 1.228999376296997, "learning_rate": 3.002961361827795e-05, "loss": 0.0588, "step": 129850 }, { "epoch": 4.390276885628317, "grad_norm": 2.46246075630188, "learning_rate": 3.0022034605570466e-05, "loss": 0.0569, "step": 129860 }, { "epoch": 4.390614963318571, "grad_norm": 4.832151889801025, "learning_rate": 3.0014456139053016e-05, "loss": 0.061, "step": 129870 }, { "epoch": 4.3909530410088236, "grad_norm": 1.3493643999099731, "learning_rate": 3.0006878218932866e-05, "loss": 0.0545, "step": 129880 }, { "epoch": 4.391291118699077, "grad_norm": 2.6004068851470947, "learning_rate": 2.9999300845417116e-05, "loss": 0.0574, "step": 129890 }, { "epoch": 4.391629196389331, "grad_norm": 1.3744087219238281, "learning_rate": 2.9991724018712993e-05, "loss": 0.0571, "step": 129900 }, { "epoch": 4.391967274079583, "grad_norm": 5.6896772384643555, "learning_rate": 2.99841477390276e-05, "loss": 0.065, "step": 129910 }, { "epoch": 4.392305351769837, "grad_norm": 2.374258041381836, "learning_rate": 2.9976572006568094e-05, "loss": 0.0566, "step": 129920 }, { "epoch": 4.39264342946009, "grad_norm": 2.4703094959259033, "learning_rate": 2.9968996821541602e-05, "loss": 0.0645, "step": 129930 }, { "epoch": 4.392981507150343, "grad_norm": 2.335062026977539, "learning_rate": 2.9961422184155196e-05, "loss": 0.0587, "step": 129940 }, { "epoch": 4.393319584840596, "grad_norm": 1.9322201013565063, "learning_rate": 2.9953848094615988e-05, "loss": 0.0566, "step": 129950 }, { "epoch": 4.39365766253085, "grad_norm": 2.8954551219940186, "learning_rate": 2.9946274553131038e-05, "loss": 0.0582, "step": 129960 }, { "epoch": 4.3939957402211025, "grad_norm": 5.738729476928711, "learning_rate": 2.993870155990739e-05, "loss": 0.0606, "step": 129970 }, { "epoch": 4.394333817911356, "grad_norm": 2.9273123741149902, "learning_rate": 2.9931129115152128e-05, "loss": 0.0627, "step": 129980 }, { "epoch": 4.3946718956016095, "grad_norm": 2.922438621520996, "learning_rate": 2.9923557219072246e-05, "loss": 0.0639, "step": 129990 }, { "epoch": 4.395009973291862, "grad_norm": 10.555357933044434, "learning_rate": 2.991598587187478e-05, "loss": 0.0568, "step": 130000 }, { "epoch": 4.395348050982116, "grad_norm": 4.282050609588623, "learning_rate": 2.9908415073766704e-05, "loss": 0.0563, "step": 130010 }, { "epoch": 4.395686128672369, "grad_norm": 16.026052474975586, "learning_rate": 2.990084482495502e-05, "loss": 0.0603, "step": 130020 }, { "epoch": 4.396024206362622, "grad_norm": 2.6609303951263428, "learning_rate": 2.989327512564668e-05, "loss": 0.0563, "step": 130030 }, { "epoch": 4.396362284052875, "grad_norm": 2.6230757236480713, "learning_rate": 2.988570597604864e-05, "loss": 0.0651, "step": 130040 }, { "epoch": 4.396700361743129, "grad_norm": 3.232327938079834, "learning_rate": 2.9878137376367854e-05, "loss": 0.0626, "step": 130050 }, { "epoch": 4.3970384394333815, "grad_norm": 2.7037599086761475, "learning_rate": 2.9870569326811217e-05, "loss": 0.0627, "step": 130060 }, { "epoch": 4.397376517123635, "grad_norm": 2.528032064437866, "learning_rate": 2.986300182758568e-05, "loss": 0.0612, "step": 130070 }, { "epoch": 4.3977145948138885, "grad_norm": 14.688286781311035, "learning_rate": 2.985543487889807e-05, "loss": 0.0563, "step": 130080 }, { "epoch": 4.398052672504141, "grad_norm": 1.8971072435379028, "learning_rate": 2.9847868480955345e-05, "loss": 0.0646, "step": 130090 }, { "epoch": 4.398390750194395, "grad_norm": 3.3670167922973633, "learning_rate": 2.984030263396429e-05, "loss": 0.0595, "step": 130100 }, { "epoch": 4.398728827884648, "grad_norm": 1.5208845138549805, "learning_rate": 2.983273733813179e-05, "loss": 0.0655, "step": 130110 }, { "epoch": 4.399066905574901, "grad_norm": 1.915351390838623, "learning_rate": 2.9825172593664695e-05, "loss": 0.0625, "step": 130120 }, { "epoch": 4.399404983265154, "grad_norm": 5.687646389007568, "learning_rate": 2.9817608400769786e-05, "loss": 0.0664, "step": 130130 }, { "epoch": 4.399743060955408, "grad_norm": 253.65017700195312, "learning_rate": 2.9810044759653894e-05, "loss": 0.0582, "step": 130140 }, { "epoch": 4.40008113864566, "grad_norm": 2.8071584701538086, "learning_rate": 2.9802481670523786e-05, "loss": 0.0662, "step": 130150 }, { "epoch": 4.400419216335914, "grad_norm": 1.2333749532699585, "learning_rate": 2.9794919133586224e-05, "loss": 0.0658, "step": 130160 }, { "epoch": 4.400757294026167, "grad_norm": 15.231711387634277, "learning_rate": 2.9787357149048018e-05, "loss": 0.0615, "step": 130170 }, { "epoch": 4.40109537171642, "grad_norm": 1.6713509559631348, "learning_rate": 2.9779795717115842e-05, "loss": 0.0608, "step": 130180 }, { "epoch": 4.401433449406674, "grad_norm": 3.1041831970214844, "learning_rate": 2.977223483799647e-05, "loss": 0.0615, "step": 130190 }, { "epoch": 4.401771527096927, "grad_norm": 2.0479421615600586, "learning_rate": 2.9764674511896595e-05, "loss": 0.0582, "step": 130200 }, { "epoch": 4.40210960478718, "grad_norm": 17.164588928222656, "learning_rate": 2.975711473902293e-05, "loss": 0.058, "step": 130210 }, { "epoch": 4.402447682477433, "grad_norm": 5.916240692138672, "learning_rate": 2.974955551958213e-05, "loss": 0.0649, "step": 130220 }, { "epoch": 4.402785760167687, "grad_norm": 2.490407705307007, "learning_rate": 2.974199685378087e-05, "loss": 0.0683, "step": 130230 }, { "epoch": 4.403123837857939, "grad_norm": 1.6628981828689575, "learning_rate": 2.9734438741825816e-05, "loss": 0.0593, "step": 130240 }, { "epoch": 4.403461915548193, "grad_norm": 3.4058732986450195, "learning_rate": 2.9726881183923582e-05, "loss": 0.0588, "step": 130250 }, { "epoch": 4.403799993238446, "grad_norm": 3.4062626361846924, "learning_rate": 2.9719324180280816e-05, "loss": 0.0553, "step": 130260 }, { "epoch": 4.404138070928699, "grad_norm": 3.04353666305542, "learning_rate": 2.9711767731104083e-05, "loss": 0.063, "step": 130270 }, { "epoch": 4.4044761486189525, "grad_norm": 1.998482584953308, "learning_rate": 2.970421183660003e-05, "loss": 0.0632, "step": 130280 }, { "epoch": 4.404814226309206, "grad_norm": 1.4551950693130493, "learning_rate": 2.969665649697517e-05, "loss": 0.0569, "step": 130290 }, { "epoch": 4.405152303999459, "grad_norm": 3.716677188873291, "learning_rate": 2.9689101712436108e-05, "loss": 0.0675, "step": 130300 }, { "epoch": 4.405490381689712, "grad_norm": 2.7588934898376465, "learning_rate": 2.968154748318938e-05, "loss": 0.0638, "step": 130310 }, { "epoch": 4.405828459379966, "grad_norm": 2.29596209526062, "learning_rate": 2.96739938094415e-05, "loss": 0.0631, "step": 130320 }, { "epoch": 4.406166537070218, "grad_norm": 2.290700912475586, "learning_rate": 2.9666440691399012e-05, "loss": 0.057, "step": 130330 }, { "epoch": 4.406504614760472, "grad_norm": 5.3985395431518555, "learning_rate": 2.9658888129268373e-05, "loss": 0.064, "step": 130340 }, { "epoch": 4.406842692450725, "grad_norm": 2.1196155548095703, "learning_rate": 2.965133612325609e-05, "loss": 0.0664, "step": 130350 }, { "epoch": 4.407180770140979, "grad_norm": 2.592373847961426, "learning_rate": 2.9643784673568665e-05, "loss": 0.0597, "step": 130360 }, { "epoch": 4.4075188478312315, "grad_norm": 5.408497333526611, "learning_rate": 2.963623378041248e-05, "loss": 0.0628, "step": 130370 }, { "epoch": 4.407856925521485, "grad_norm": 3.756464958190918, "learning_rate": 2.9628683443994047e-05, "loss": 0.0662, "step": 130380 }, { "epoch": 4.408195003211738, "grad_norm": 3.48476505279541, "learning_rate": 2.962113366451973e-05, "loss": 0.0579, "step": 130390 }, { "epoch": 4.408533080901991, "grad_norm": 3.5055339336395264, "learning_rate": 2.9613584442195973e-05, "loss": 0.0554, "step": 130400 }, { "epoch": 4.408871158592245, "grad_norm": 2.3988382816314697, "learning_rate": 2.960603577722916e-05, "loss": 0.0602, "step": 130410 }, { "epoch": 4.409209236282498, "grad_norm": 1.4465886354446411, "learning_rate": 2.9598487669825658e-05, "loss": 0.0603, "step": 130420 }, { "epoch": 4.409547313972751, "grad_norm": 2.847817897796631, "learning_rate": 2.9590940120191855e-05, "loss": 0.0633, "step": 130430 }, { "epoch": 4.409885391663004, "grad_norm": 4.5763044357299805, "learning_rate": 2.958339312853407e-05, "loss": 0.0624, "step": 130440 }, { "epoch": 4.410223469353257, "grad_norm": 1.7855960130691528, "learning_rate": 2.9575846695058657e-05, "loss": 0.0595, "step": 130450 }, { "epoch": 4.4105615470435104, "grad_norm": 2.875638246536255, "learning_rate": 2.9568300819971904e-05, "loss": 0.0637, "step": 130460 }, { "epoch": 4.410899624733764, "grad_norm": 12.027176856994629, "learning_rate": 2.956075550348015e-05, "loss": 0.0568, "step": 130470 }, { "epoch": 4.4112377024240175, "grad_norm": 10.23793888092041, "learning_rate": 2.9553210745789638e-05, "loss": 0.0642, "step": 130480 }, { "epoch": 4.41157578011427, "grad_norm": 4.506106376647949, "learning_rate": 2.9545666547106676e-05, "loss": 0.0607, "step": 130490 }, { "epoch": 4.411913857804524, "grad_norm": 3.9415385723114014, "learning_rate": 2.9538122907637522e-05, "loss": 0.0604, "step": 130500 }, { "epoch": 4.412251935494776, "grad_norm": 8.403304100036621, "learning_rate": 2.9530579827588378e-05, "loss": 0.0588, "step": 130510 }, { "epoch": 4.41259001318503, "grad_norm": 2.016908645629883, "learning_rate": 2.9523037307165507e-05, "loss": 0.059, "step": 130520 }, { "epoch": 4.412928090875283, "grad_norm": 1.6132574081420898, "learning_rate": 2.9515495346575096e-05, "loss": 0.0634, "step": 130530 }, { "epoch": 4.413266168565537, "grad_norm": 3.4010226726531982, "learning_rate": 2.9507953946023354e-05, "loss": 0.0594, "step": 130540 }, { "epoch": 4.413604246255789, "grad_norm": 6.638599395751953, "learning_rate": 2.950041310571644e-05, "loss": 0.061, "step": 130550 }, { "epoch": 4.413942323946043, "grad_norm": 2.924365997314453, "learning_rate": 2.9492872825860522e-05, "loss": 0.0574, "step": 130560 }, { "epoch": 4.414280401636296, "grad_norm": 3.120478391647339, "learning_rate": 2.9485333106661788e-05, "loss": 0.0586, "step": 130570 }, { "epoch": 4.414618479326549, "grad_norm": 2.980494976043701, "learning_rate": 2.947779394832631e-05, "loss": 0.064, "step": 130580 }, { "epoch": 4.414956557016803, "grad_norm": 3.548131227493286, "learning_rate": 2.947025535106026e-05, "loss": 0.0587, "step": 130590 }, { "epoch": 4.415294634707056, "grad_norm": 7.915720462799072, "learning_rate": 2.94627173150697e-05, "loss": 0.0619, "step": 130600 }, { "epoch": 4.415632712397309, "grad_norm": 7.673823356628418, "learning_rate": 2.9455179840560733e-05, "loss": 0.0648, "step": 130610 }, { "epoch": 4.415970790087562, "grad_norm": 1.3112510442733765, "learning_rate": 2.944764292773944e-05, "loss": 0.0623, "step": 130620 }, { "epoch": 4.416308867777816, "grad_norm": 4.654965400695801, "learning_rate": 2.9440106576811855e-05, "loss": 0.0609, "step": 130630 }, { "epoch": 4.416646945468068, "grad_norm": 1.954119324684143, "learning_rate": 2.943257078798405e-05, "loss": 0.0635, "step": 130640 }, { "epoch": 4.416985023158322, "grad_norm": 1.8479934930801392, "learning_rate": 2.9425035561462023e-05, "loss": 0.0643, "step": 130650 }, { "epoch": 4.417323100848575, "grad_norm": 2.4343645572662354, "learning_rate": 2.9417500897451798e-05, "loss": 0.0587, "step": 130660 }, { "epoch": 4.417661178538828, "grad_norm": 4.810944080352783, "learning_rate": 2.940996679615936e-05, "loss": 0.0614, "step": 130670 }, { "epoch": 4.4179992562290815, "grad_norm": 2.447622060775757, "learning_rate": 2.9402433257790685e-05, "loss": 0.0646, "step": 130680 }, { "epoch": 4.418337333919335, "grad_norm": 4.928944110870361, "learning_rate": 2.9394900282551762e-05, "loss": 0.06, "step": 130690 }, { "epoch": 4.418675411609588, "grad_norm": 3.8257930278778076, "learning_rate": 2.938736787064852e-05, "loss": 0.0622, "step": 130700 }, { "epoch": 4.419013489299841, "grad_norm": 2.0765650272369385, "learning_rate": 2.93798360222869e-05, "loss": 0.059, "step": 130710 }, { "epoch": 4.419351566990095, "grad_norm": 1.5043082237243652, "learning_rate": 2.9372304737672812e-05, "loss": 0.0624, "step": 130720 }, { "epoch": 4.419689644680347, "grad_norm": 2.786884307861328, "learning_rate": 2.9364774017012175e-05, "loss": 0.0564, "step": 130730 }, { "epoch": 4.420027722370601, "grad_norm": 1.367050051689148, "learning_rate": 2.9357243860510853e-05, "loss": 0.0606, "step": 130740 }, { "epoch": 4.420365800060854, "grad_norm": 17.992042541503906, "learning_rate": 2.934971426837472e-05, "loss": 0.0677, "step": 130750 }, { "epoch": 4.420703877751107, "grad_norm": 5.640562057495117, "learning_rate": 2.9342185240809662e-05, "loss": 0.0555, "step": 130760 }, { "epoch": 4.4210419554413605, "grad_norm": 1.7865688800811768, "learning_rate": 2.9334656778021475e-05, "loss": 0.0591, "step": 130770 }, { "epoch": 4.421380033131614, "grad_norm": 2.7332799434661865, "learning_rate": 2.9327128880216038e-05, "loss": 0.0635, "step": 130780 }, { "epoch": 4.421718110821867, "grad_norm": 2.7302145957946777, "learning_rate": 2.9319601547599095e-05, "loss": 0.0589, "step": 130790 }, { "epoch": 4.42205618851212, "grad_norm": 17.43799591064453, "learning_rate": 2.931207478037649e-05, "loss": 0.0679, "step": 130800 }, { "epoch": 4.422394266202374, "grad_norm": 2.138336658477783, "learning_rate": 2.9304548578754e-05, "loss": 0.062, "step": 130810 }, { "epoch": 4.422732343892626, "grad_norm": 1.4795678853988647, "learning_rate": 2.929702294293737e-05, "loss": 0.055, "step": 130820 }, { "epoch": 4.42307042158288, "grad_norm": 1.463821530342102, "learning_rate": 2.928949787313236e-05, "loss": 0.0494, "step": 130830 }, { "epoch": 4.423408499273133, "grad_norm": 2.0152947902679443, "learning_rate": 2.928197336954469e-05, "loss": 0.0575, "step": 130840 }, { "epoch": 4.423746576963386, "grad_norm": 1.3769571781158447, "learning_rate": 2.927444943238009e-05, "loss": 0.0605, "step": 130850 }, { "epoch": 4.424084654653639, "grad_norm": 4.250340461730957, "learning_rate": 2.9266926061844258e-05, "loss": 0.0581, "step": 130860 }, { "epoch": 4.424422732343893, "grad_norm": 1.918073058128357, "learning_rate": 2.925940325814286e-05, "loss": 0.0565, "step": 130870 }, { "epoch": 4.424760810034146, "grad_norm": 3.753911256790161, "learning_rate": 2.9251881021481614e-05, "loss": 0.0584, "step": 130880 }, { "epoch": 4.425098887724399, "grad_norm": 2.4609217643737793, "learning_rate": 2.924435935206612e-05, "loss": 0.0646, "step": 130890 }, { "epoch": 4.425436965414653, "grad_norm": 2.2652761936187744, "learning_rate": 2.923683825010206e-05, "loss": 0.0579, "step": 130900 }, { "epoch": 4.425775043104905, "grad_norm": 4.445440292358398, "learning_rate": 2.9229317715795036e-05, "loss": 0.0584, "step": 130910 }, { "epoch": 4.426113120795159, "grad_norm": 2.760119676589966, "learning_rate": 2.9221797749350665e-05, "loss": 0.0599, "step": 130920 }, { "epoch": 4.426451198485412, "grad_norm": 2.006681203842163, "learning_rate": 2.9214278350974535e-05, "loss": 0.0534, "step": 130930 }, { "epoch": 4.426789276175665, "grad_norm": 6.95731782913208, "learning_rate": 2.9206759520872222e-05, "loss": 0.0645, "step": 130940 }, { "epoch": 4.427127353865918, "grad_norm": 5.7148518562316895, "learning_rate": 2.919924125924931e-05, "loss": 0.0642, "step": 130950 }, { "epoch": 4.427465431556172, "grad_norm": 1.391814947128296, "learning_rate": 2.9191723566311307e-05, "loss": 0.0574, "step": 130960 }, { "epoch": 4.4278035092464245, "grad_norm": 5.783514022827148, "learning_rate": 2.918420644226378e-05, "loss": 0.0562, "step": 130970 }, { "epoch": 4.428141586936678, "grad_norm": 7.818664073944092, "learning_rate": 2.9176689887312204e-05, "loss": 0.0629, "step": 130980 }, { "epoch": 4.4284796646269315, "grad_norm": 3.1713998317718506, "learning_rate": 2.916917390166213e-05, "loss": 0.0625, "step": 130990 }, { "epoch": 4.428817742317184, "grad_norm": 5.031571388244629, "learning_rate": 2.9161658485519012e-05, "loss": 0.064, "step": 131000 }, { "epoch": 4.429155820007438, "grad_norm": 1.9251978397369385, "learning_rate": 2.9154143639088304e-05, "loss": 0.0655, "step": 131010 }, { "epoch": 4.429493897697691, "grad_norm": 7.3441972732543945, "learning_rate": 2.9146629362575495e-05, "loss": 0.0604, "step": 131020 }, { "epoch": 4.429831975387944, "grad_norm": 4.269627571105957, "learning_rate": 2.913911565618599e-05, "loss": 0.0701, "step": 131030 }, { "epoch": 4.430170053078197, "grad_norm": 1.7966169118881226, "learning_rate": 2.9131602520125256e-05, "loss": 0.0618, "step": 131040 }, { "epoch": 4.430508130768451, "grad_norm": 1.2080060243606567, "learning_rate": 2.9124089954598645e-05, "loss": 0.0667, "step": 131050 }, { "epoch": 4.4308462084587035, "grad_norm": 10.803662300109863, "learning_rate": 2.9116577959811558e-05, "loss": 0.0617, "step": 131060 }, { "epoch": 4.431184286148957, "grad_norm": 3.57836651802063, "learning_rate": 2.910906653596941e-05, "loss": 0.0613, "step": 131070 }, { "epoch": 4.4315223638392105, "grad_norm": 5.024834156036377, "learning_rate": 2.910155568327753e-05, "loss": 0.0594, "step": 131080 }, { "epoch": 4.431860441529463, "grad_norm": 3.4828951358795166, "learning_rate": 2.9094045401941273e-05, "loss": 0.0574, "step": 131090 }, { "epoch": 4.432198519219717, "grad_norm": 1.9177781343460083, "learning_rate": 2.908653569216594e-05, "loss": 0.0639, "step": 131100 }, { "epoch": 4.43253659690997, "grad_norm": 1.2752689123153687, "learning_rate": 2.9079026554156874e-05, "loss": 0.0589, "step": 131110 }, { "epoch": 4.432874674600223, "grad_norm": 2.852108955383301, "learning_rate": 2.9071517988119374e-05, "loss": 0.0547, "step": 131120 }, { "epoch": 4.433212752290476, "grad_norm": 1.2536816596984863, "learning_rate": 2.9064009994258684e-05, "loss": 0.072, "step": 131130 }, { "epoch": 4.43355082998073, "grad_norm": 1.0709068775177002, "learning_rate": 2.905650257278012e-05, "loss": 0.0635, "step": 131140 }, { "epoch": 4.433888907670982, "grad_norm": 1.609447956085205, "learning_rate": 2.9048995723888884e-05, "loss": 0.062, "step": 131150 }, { "epoch": 4.434226985361236, "grad_norm": 1.3577756881713867, "learning_rate": 2.9041489447790256e-05, "loss": 0.059, "step": 131160 }, { "epoch": 4.4345650630514895, "grad_norm": 3.41227126121521, "learning_rate": 2.903398374468943e-05, "loss": 0.0608, "step": 131170 }, { "epoch": 4.434903140741742, "grad_norm": 1.7468341588974, "learning_rate": 2.9026478614791598e-05, "loss": 0.0645, "step": 131180 }, { "epoch": 4.435241218431996, "grad_norm": 2.5945823192596436, "learning_rate": 2.9018974058301972e-05, "loss": 0.055, "step": 131190 }, { "epoch": 4.435579296122249, "grad_norm": 2.3044917583465576, "learning_rate": 2.901147007542572e-05, "loss": 0.0625, "step": 131200 }, { "epoch": 4.435917373812502, "grad_norm": 2.849147319793701, "learning_rate": 2.9003966666367988e-05, "loss": 0.0594, "step": 131210 }, { "epoch": 4.436255451502755, "grad_norm": 1.2166063785552979, "learning_rate": 2.8996463831333898e-05, "loss": 0.0621, "step": 131220 }, { "epoch": 4.436593529193009, "grad_norm": 2.72135066986084, "learning_rate": 2.8988961570528616e-05, "loss": 0.0593, "step": 131230 }, { "epoch": 4.436931606883261, "grad_norm": 4.614528656005859, "learning_rate": 2.8981459884157214e-05, "loss": 0.0651, "step": 131240 }, { "epoch": 4.437269684573515, "grad_norm": 2.1426916122436523, "learning_rate": 2.8973958772424824e-05, "loss": 0.0594, "step": 131250 }, { "epoch": 4.437607762263768, "grad_norm": 4.687005996704102, "learning_rate": 2.8966458235536498e-05, "loss": 0.0576, "step": 131260 }, { "epoch": 4.437945839954021, "grad_norm": 1.8415459394454956, "learning_rate": 2.895895827369729e-05, "loss": 0.0558, "step": 131270 }, { "epoch": 4.4382839176442745, "grad_norm": 2.1257071495056152, "learning_rate": 2.895145888711227e-05, "loss": 0.0668, "step": 131280 }, { "epoch": 4.438621995334528, "grad_norm": 2.4471824169158936, "learning_rate": 2.8943960075986458e-05, "loss": 0.0627, "step": 131290 }, { "epoch": 4.438960073024781, "grad_norm": 3.2103710174560547, "learning_rate": 2.8936461840524864e-05, "loss": 0.0604, "step": 131300 }, { "epoch": 4.439298150715034, "grad_norm": 1.9324289560317993, "learning_rate": 2.8928964180932477e-05, "loss": 0.058, "step": 131310 }, { "epoch": 4.439636228405288, "grad_norm": 1.5573266744613647, "learning_rate": 2.892146709741429e-05, "loss": 0.0632, "step": 131320 }, { "epoch": 4.439974306095541, "grad_norm": 2.261155128479004, "learning_rate": 2.8913970590175315e-05, "loss": 0.0606, "step": 131330 }, { "epoch": 4.440312383785794, "grad_norm": 3.184318780899048, "learning_rate": 2.8906474659420425e-05, "loss": 0.0575, "step": 131340 }, { "epoch": 4.440650461476047, "grad_norm": 2.436875581741333, "learning_rate": 2.8898979305354607e-05, "loss": 0.0617, "step": 131350 }, { "epoch": 4.4409885391663, "grad_norm": 5.122133255004883, "learning_rate": 2.8891484528182748e-05, "loss": 0.0652, "step": 131360 }, { "epoch": 4.4413266168565535, "grad_norm": 2.3062973022460938, "learning_rate": 2.8883990328109795e-05, "loss": 0.0621, "step": 131370 }, { "epoch": 4.441664694546807, "grad_norm": 1.7260915040969849, "learning_rate": 2.8876496705340618e-05, "loss": 0.0576, "step": 131380 }, { "epoch": 4.4420027722370605, "grad_norm": 2.0511724948883057, "learning_rate": 2.8869003660080052e-05, "loss": 0.0605, "step": 131390 }, { "epoch": 4.442340849927313, "grad_norm": 8.32862663269043, "learning_rate": 2.886151119253302e-05, "loss": 0.0615, "step": 131400 }, { "epoch": 4.442678927617567, "grad_norm": 1.2545239925384521, "learning_rate": 2.8854019302904322e-05, "loss": 0.0629, "step": 131410 }, { "epoch": 4.443017005307819, "grad_norm": 3.2392289638519287, "learning_rate": 2.8846527991398796e-05, "loss": 0.0655, "step": 131420 }, { "epoch": 4.443355082998073, "grad_norm": 2.249918222427368, "learning_rate": 2.883903725822123e-05, "loss": 0.0563, "step": 131430 }, { "epoch": 4.443693160688326, "grad_norm": 6.484011173248291, "learning_rate": 2.8831547103576427e-05, "loss": 0.0653, "step": 131440 }, { "epoch": 4.44403123837858, "grad_norm": 6.459186553955078, "learning_rate": 2.8824057527669203e-05, "loss": 0.0528, "step": 131450 }, { "epoch": 4.4443693160688325, "grad_norm": 9.551736831665039, "learning_rate": 2.8816568530704292e-05, "loss": 0.0662, "step": 131460 }, { "epoch": 4.444707393759086, "grad_norm": 1.849258542060852, "learning_rate": 2.8809080112886433e-05, "loss": 0.069, "step": 131470 }, { "epoch": 4.445045471449339, "grad_norm": 2.8169238567352295, "learning_rate": 2.8801592274420342e-05, "loss": 0.056, "step": 131480 }, { "epoch": 4.445383549139592, "grad_norm": 3.331664562225342, "learning_rate": 2.8794105015510776e-05, "loss": 0.0601, "step": 131490 }, { "epoch": 4.445721626829846, "grad_norm": 1.9910635948181152, "learning_rate": 2.8786618336362413e-05, "loss": 0.0596, "step": 131500 }, { "epoch": 4.446059704520099, "grad_norm": 8.575541496276855, "learning_rate": 2.877913223717992e-05, "loss": 0.0618, "step": 131510 }, { "epoch": 4.446397782210352, "grad_norm": 1.8821659088134766, "learning_rate": 2.8771646718167987e-05, "loss": 0.0607, "step": 131520 }, { "epoch": 4.446735859900605, "grad_norm": 2.9534225463867188, "learning_rate": 2.876416177953125e-05, "loss": 0.0562, "step": 131530 }, { "epoch": 4.447073937590859, "grad_norm": 2.2195167541503906, "learning_rate": 2.8756677421474388e-05, "loss": 0.0654, "step": 131540 }, { "epoch": 4.447412015281111, "grad_norm": 1.4640018939971924, "learning_rate": 2.8749193644201943e-05, "loss": 0.064, "step": 131550 }, { "epoch": 4.447750092971365, "grad_norm": 3.7595901489257812, "learning_rate": 2.8741710447918558e-05, "loss": 0.0688, "step": 131560 }, { "epoch": 4.448088170661618, "grad_norm": 2.9353044033050537, "learning_rate": 2.8734227832828846e-05, "loss": 0.0665, "step": 131570 }, { "epoch": 4.448426248351871, "grad_norm": 7.493204593658447, "learning_rate": 2.8726745799137354e-05, "loss": 0.0644, "step": 131580 }, { "epoch": 4.448764326042125, "grad_norm": 1.512353539466858, "learning_rate": 2.8719264347048646e-05, "loss": 0.0631, "step": 131590 }, { "epoch": 4.449102403732378, "grad_norm": 4.098930835723877, "learning_rate": 2.8711783476767228e-05, "loss": 0.0661, "step": 131600 }, { "epoch": 4.449440481422631, "grad_norm": 2.1562247276306152, "learning_rate": 2.870430318849768e-05, "loss": 0.0634, "step": 131610 }, { "epoch": 4.449778559112884, "grad_norm": 5.949795246124268, "learning_rate": 2.869682348244448e-05, "loss": 0.0571, "step": 131620 }, { "epoch": 4.450116636803138, "grad_norm": 3.728623390197754, "learning_rate": 2.8689344358812098e-05, "loss": 0.0569, "step": 131630 }, { "epoch": 4.45045471449339, "grad_norm": 2.794555187225342, "learning_rate": 2.8681865817805054e-05, "loss": 0.0597, "step": 131640 }, { "epoch": 4.450792792183644, "grad_norm": 1.4319734573364258, "learning_rate": 2.8674387859627777e-05, "loss": 0.0581, "step": 131650 }, { "epoch": 4.451130869873897, "grad_norm": 4.777217388153076, "learning_rate": 2.8666910484484742e-05, "loss": 0.0614, "step": 131660 }, { "epoch": 4.45146894756415, "grad_norm": 1.815445899963379, "learning_rate": 2.865943369258036e-05, "loss": 0.0534, "step": 131670 }, { "epoch": 4.4518070252544035, "grad_norm": 2.130035161972046, "learning_rate": 2.865195748411904e-05, "loss": 0.0608, "step": 131680 }, { "epoch": 4.452145102944657, "grad_norm": 2.3153982162475586, "learning_rate": 2.8644481859305162e-05, "loss": 0.0656, "step": 131690 }, { "epoch": 4.45248318063491, "grad_norm": 0.9827023148536682, "learning_rate": 2.8637006818343155e-05, "loss": 0.0615, "step": 131700 }, { "epoch": 4.452821258325163, "grad_norm": 2.4988372325897217, "learning_rate": 2.862953236143735e-05, "loss": 0.059, "step": 131710 }, { "epoch": 4.453159336015417, "grad_norm": 1.528972864151001, "learning_rate": 2.8622058488792082e-05, "loss": 0.0633, "step": 131720 }, { "epoch": 4.453497413705669, "grad_norm": 1.4127339124679565, "learning_rate": 2.861458520061172e-05, "loss": 0.0604, "step": 131730 }, { "epoch": 4.453835491395923, "grad_norm": 2.1038098335266113, "learning_rate": 2.8607112497100552e-05, "loss": 0.0577, "step": 131740 }, { "epoch": 4.454173569086176, "grad_norm": 1.3431264162063599, "learning_rate": 2.8599640378462912e-05, "loss": 0.0575, "step": 131750 }, { "epoch": 4.454511646776429, "grad_norm": 4.531091213226318, "learning_rate": 2.8592168844903055e-05, "loss": 0.0614, "step": 131760 }, { "epoch": 4.4548497244666825, "grad_norm": 1.2967942953109741, "learning_rate": 2.8584697896625245e-05, "loss": 0.0536, "step": 131770 }, { "epoch": 4.455187802156936, "grad_norm": 4.981910705566406, "learning_rate": 2.8577227533833765e-05, "loss": 0.0559, "step": 131780 }, { "epoch": 4.455525879847189, "grad_norm": 2.2812142372131348, "learning_rate": 2.856975775673284e-05, "loss": 0.0564, "step": 131790 }, { "epoch": 4.455863957537442, "grad_norm": 1.2280337810516357, "learning_rate": 2.8562288565526683e-05, "loss": 0.0601, "step": 131800 }, { "epoch": 4.456202035227696, "grad_norm": 1.689337134361267, "learning_rate": 2.8554819960419493e-05, "loss": 0.0536, "step": 131810 }, { "epoch": 4.456540112917948, "grad_norm": 2.764521360397339, "learning_rate": 2.8547351941615465e-05, "loss": 0.0652, "step": 131820 }, { "epoch": 4.456878190608202, "grad_norm": 1.430167555809021, "learning_rate": 2.853988450931881e-05, "loss": 0.0583, "step": 131830 }, { "epoch": 4.457216268298455, "grad_norm": 1.5623219013214111, "learning_rate": 2.8532417663733613e-05, "loss": 0.0604, "step": 131840 }, { "epoch": 4.457554345988708, "grad_norm": 5.246077537536621, "learning_rate": 2.8524951405064072e-05, "loss": 0.0595, "step": 131850 }, { "epoch": 4.457892423678961, "grad_norm": 3.143159866333008, "learning_rate": 2.851748573351427e-05, "loss": 0.0624, "step": 131860 }, { "epoch": 4.458230501369215, "grad_norm": 3.2380685806274414, "learning_rate": 2.851002064928836e-05, "loss": 0.0666, "step": 131870 }, { "epoch": 4.458568579059468, "grad_norm": 0.9500526785850525, "learning_rate": 2.850255615259041e-05, "loss": 0.0607, "step": 131880 }, { "epoch": 4.458906656749721, "grad_norm": 1.915840744972229, "learning_rate": 2.8495092243624476e-05, "loss": 0.0643, "step": 131890 }, { "epoch": 4.459244734439975, "grad_norm": 3.1276795864105225, "learning_rate": 2.8487628922594667e-05, "loss": 0.0636, "step": 131900 }, { "epoch": 4.459582812130227, "grad_norm": 1.668232798576355, "learning_rate": 2.8480166189705005e-05, "loss": 0.0616, "step": 131910 }, { "epoch": 4.459920889820481, "grad_norm": 1.911577820777893, "learning_rate": 2.8472704045159504e-05, "loss": 0.0623, "step": 131920 }, { "epoch": 4.460258967510734, "grad_norm": 3.7826435565948486, "learning_rate": 2.8465242489162176e-05, "loss": 0.0605, "step": 131930 }, { "epoch": 4.460597045200987, "grad_norm": 1.7275190353393555, "learning_rate": 2.8457781521917042e-05, "loss": 0.065, "step": 131940 }, { "epoch": 4.46093512289124, "grad_norm": 1.480808973312378, "learning_rate": 2.845032114362805e-05, "loss": 0.0574, "step": 131950 }, { "epoch": 4.461273200581494, "grad_norm": 11.819090843200684, "learning_rate": 2.8442861354499208e-05, "loss": 0.0637, "step": 131960 }, { "epoch": 4.4616112782717465, "grad_norm": 1.3941508531570435, "learning_rate": 2.8435402154734425e-05, "loss": 0.0557, "step": 131970 }, { "epoch": 4.461949355962, "grad_norm": 5.117754936218262, "learning_rate": 2.8427943544537638e-05, "loss": 0.0591, "step": 131980 }, { "epoch": 4.4622874336522536, "grad_norm": 3.6543047428131104, "learning_rate": 2.8420485524112782e-05, "loss": 0.0604, "step": 131990 }, { "epoch": 4.462625511342506, "grad_norm": 3.2444827556610107, "learning_rate": 2.8413028093663752e-05, "loss": 0.0601, "step": 132000 }, { "epoch": 4.46296358903276, "grad_norm": 1.5228374004364014, "learning_rate": 2.8405571253394397e-05, "loss": 0.0602, "step": 132010 }, { "epoch": 4.463301666723013, "grad_norm": 1.1505972146987915, "learning_rate": 2.8398115003508633e-05, "loss": 0.0552, "step": 132020 }, { "epoch": 4.463639744413266, "grad_norm": 2.3417067527770996, "learning_rate": 2.839065934421027e-05, "loss": 0.0599, "step": 132030 }, { "epoch": 4.463977822103519, "grad_norm": 2.6455392837524414, "learning_rate": 2.83832042757032e-05, "loss": 0.0574, "step": 132040 }, { "epoch": 4.464315899793773, "grad_norm": 1.582334041595459, "learning_rate": 2.8375749798191175e-05, "loss": 0.0579, "step": 132050 }, { "epoch": 4.4646539774840255, "grad_norm": 1.0891779661178589, "learning_rate": 2.8368295911878036e-05, "loss": 0.0588, "step": 132060 }, { "epoch": 4.464992055174279, "grad_norm": 2.9255878925323486, "learning_rate": 2.8360842616967543e-05, "loss": 0.0569, "step": 132070 }, { "epoch": 4.4653301328645325, "grad_norm": 1.8502622842788696, "learning_rate": 2.83533899136635e-05, "loss": 0.0627, "step": 132080 }, { "epoch": 4.465668210554785, "grad_norm": 1.673378825187683, "learning_rate": 2.8345937802169654e-05, "loss": 0.0592, "step": 132090 }, { "epoch": 4.466006288245039, "grad_norm": 2.5486090183258057, "learning_rate": 2.8338486282689702e-05, "loss": 0.0584, "step": 132100 }, { "epoch": 4.466344365935292, "grad_norm": 1.5888519287109375, "learning_rate": 2.8331035355427428e-05, "loss": 0.0588, "step": 132110 }, { "epoch": 4.466682443625545, "grad_norm": 4.317534446716309, "learning_rate": 2.8323585020586495e-05, "loss": 0.0675, "step": 132120 }, { "epoch": 4.467020521315798, "grad_norm": 1.956990122795105, "learning_rate": 2.8316135278370616e-05, "loss": 0.0569, "step": 132130 }, { "epoch": 4.467358599006052, "grad_norm": 2.652399778366089, "learning_rate": 2.830868612898342e-05, "loss": 0.0611, "step": 132140 }, { "epoch": 4.467696676696304, "grad_norm": 1.8553266525268555, "learning_rate": 2.83012375726286e-05, "loss": 0.0623, "step": 132150 }, { "epoch": 4.468034754386558, "grad_norm": 1.077699065208435, "learning_rate": 2.8293789609509825e-05, "loss": 0.0594, "step": 132160 }, { "epoch": 4.4683728320768115, "grad_norm": 2.107391357421875, "learning_rate": 2.8286342239830676e-05, "loss": 0.0575, "step": 132170 }, { "epoch": 4.468710909767064, "grad_norm": 1.3869692087173462, "learning_rate": 2.8278895463794775e-05, "loss": 0.059, "step": 132180 }, { "epoch": 4.469048987457318, "grad_norm": 1.537182331085205, "learning_rate": 2.8271449281605687e-05, "loss": 0.0596, "step": 132190 }, { "epoch": 4.469387065147571, "grad_norm": 1.5261046886444092, "learning_rate": 2.826400369346704e-05, "loss": 0.0578, "step": 132200 }, { "epoch": 4.469725142837824, "grad_norm": 1.7854385375976562, "learning_rate": 2.8256558699582358e-05, "loss": 0.06, "step": 132210 }, { "epoch": 4.470063220528077, "grad_norm": 8.11050033569336, "learning_rate": 2.8249114300155177e-05, "loss": 0.0588, "step": 132220 }, { "epoch": 4.470401298218331, "grad_norm": 4.072573184967041, "learning_rate": 2.824167049538906e-05, "loss": 0.0596, "step": 132230 }, { "epoch": 4.470739375908583, "grad_norm": 3.962231159210205, "learning_rate": 2.823422728548748e-05, "loss": 0.0609, "step": 132240 }, { "epoch": 4.471077453598837, "grad_norm": 1.1233105659484863, "learning_rate": 2.8226784670653988e-05, "loss": 0.0635, "step": 132250 }, { "epoch": 4.47141553128909, "grad_norm": 1.8179508447647095, "learning_rate": 2.8219342651091984e-05, "loss": 0.0609, "step": 132260 }, { "epoch": 4.471753608979343, "grad_norm": 2.2318198680877686, "learning_rate": 2.8211901227004965e-05, "loss": 0.0597, "step": 132270 }, { "epoch": 4.4720916866695966, "grad_norm": 11.657684326171875, "learning_rate": 2.8204460398596407e-05, "loss": 0.0576, "step": 132280 }, { "epoch": 4.47242976435985, "grad_norm": 1.0700626373291016, "learning_rate": 2.819702016606971e-05, "loss": 0.0537, "step": 132290 }, { "epoch": 4.472767842050104, "grad_norm": 1.1708014011383057, "learning_rate": 2.818958052962829e-05, "loss": 0.064, "step": 132300 }, { "epoch": 4.473105919740356, "grad_norm": 2.0234434604644775, "learning_rate": 2.8182141489475523e-05, "loss": 0.0601, "step": 132310 }, { "epoch": 4.47344399743061, "grad_norm": 1.650113582611084, "learning_rate": 2.817470304581483e-05, "loss": 0.059, "step": 132320 }, { "epoch": 4.473782075120862, "grad_norm": 1.5319862365722656, "learning_rate": 2.816726519884956e-05, "loss": 0.0515, "step": 132330 }, { "epoch": 4.474120152811116, "grad_norm": 1.337246298789978, "learning_rate": 2.815982794878304e-05, "loss": 0.0593, "step": 132340 }, { "epoch": 4.474458230501369, "grad_norm": 0.9234004616737366, "learning_rate": 2.8152391295818637e-05, "loss": 0.06, "step": 132350 }, { "epoch": 4.474796308191623, "grad_norm": 1.664420247077942, "learning_rate": 2.8144955240159627e-05, "loss": 0.0573, "step": 132360 }, { "epoch": 4.4751343858818755, "grad_norm": 0.9087253212928772, "learning_rate": 2.8137519782009347e-05, "loss": 0.0507, "step": 132370 }, { "epoch": 4.475472463572129, "grad_norm": 2.133213996887207, "learning_rate": 2.8130084921571064e-05, "loss": 0.0596, "step": 132380 }, { "epoch": 4.475810541262382, "grad_norm": 5.02879524230957, "learning_rate": 2.812265065904803e-05, "loss": 0.0586, "step": 132390 }, { "epoch": 4.476148618952635, "grad_norm": 0.7353662848472595, "learning_rate": 2.8115216994643524e-05, "loss": 0.0571, "step": 132400 }, { "epoch": 4.476486696642889, "grad_norm": 4.099560260772705, "learning_rate": 2.8107783928560762e-05, "loss": 0.062, "step": 132410 }, { "epoch": 4.476824774333142, "grad_norm": 1.4094518423080444, "learning_rate": 2.810035146100296e-05, "loss": 0.061, "step": 132420 }, { "epoch": 4.477162852023395, "grad_norm": 0.8518404364585876, "learning_rate": 2.8092919592173305e-05, "loss": 0.0548, "step": 132430 }, { "epoch": 4.477500929713648, "grad_norm": 3.1364781856536865, "learning_rate": 2.808548832227502e-05, "loss": 0.0618, "step": 132440 }, { "epoch": 4.477839007403901, "grad_norm": 1.7685554027557373, "learning_rate": 2.807805765151123e-05, "loss": 0.0585, "step": 132450 }, { "epoch": 4.4781770850941545, "grad_norm": 1.6602702140808105, "learning_rate": 2.807062758008513e-05, "loss": 0.0574, "step": 132460 }, { "epoch": 4.478515162784408, "grad_norm": 3.720179319381714, "learning_rate": 2.806319810819983e-05, "loss": 0.0594, "step": 132470 }, { "epoch": 4.4788532404746615, "grad_norm": 0.6579815149307251, "learning_rate": 2.805576923605843e-05, "loss": 0.0589, "step": 132480 }, { "epoch": 4.479191318164914, "grad_norm": 2.560014486312866, "learning_rate": 2.8048340963864083e-05, "loss": 0.0609, "step": 132490 }, { "epoch": 4.479529395855168, "grad_norm": 1.5580432415008545, "learning_rate": 2.8040913291819842e-05, "loss": 0.063, "step": 132500 }, { "epoch": 4.479867473545421, "grad_norm": 3.3685543537139893, "learning_rate": 2.803348622012878e-05, "loss": 0.0592, "step": 132510 }, { "epoch": 4.480205551235674, "grad_norm": 3.5505614280700684, "learning_rate": 2.802605974899394e-05, "loss": 0.0643, "step": 132520 }, { "epoch": 4.480543628925927, "grad_norm": 1.2014578580856323, "learning_rate": 2.8018633878618365e-05, "loss": 0.0597, "step": 132530 }, { "epoch": 4.480881706616181, "grad_norm": 1.921950340270996, "learning_rate": 2.8011208609205126e-05, "loss": 0.0665, "step": 132540 }, { "epoch": 4.481219784306433, "grad_norm": 2.5683884620666504, "learning_rate": 2.800378394095714e-05, "loss": 0.0588, "step": 132550 }, { "epoch": 4.481557861996687, "grad_norm": 1.8042280673980713, "learning_rate": 2.7996359874077455e-05, "loss": 0.071, "step": 132560 }, { "epoch": 4.4818959396869404, "grad_norm": 1.3258061408996582, "learning_rate": 2.7988936408769007e-05, "loss": 0.0646, "step": 132570 }, { "epoch": 4.482234017377193, "grad_norm": 1.5596948862075806, "learning_rate": 2.7981513545234788e-05, "loss": 0.059, "step": 132580 }, { "epoch": 4.482572095067447, "grad_norm": 1.3642675876617432, "learning_rate": 2.7974091283677716e-05, "loss": 0.0609, "step": 132590 }, { "epoch": 4.4829101727577, "grad_norm": 1.067521572113037, "learning_rate": 2.79666696243007e-05, "loss": 0.0658, "step": 132600 }, { "epoch": 4.483248250447953, "grad_norm": 1.6665693521499634, "learning_rate": 2.7959248567306672e-05, "loss": 0.069, "step": 132610 }, { "epoch": 4.483586328138206, "grad_norm": 3.2045724391937256, "learning_rate": 2.795182811289851e-05, "loss": 0.0603, "step": 132620 }, { "epoch": 4.48392440582846, "grad_norm": 2.137916088104248, "learning_rate": 2.7944408261279088e-05, "loss": 0.0631, "step": 132630 }, { "epoch": 4.484262483518712, "grad_norm": 1.1704866886138916, "learning_rate": 2.7936989012651227e-05, "loss": 0.0681, "step": 132640 }, { "epoch": 4.484600561208966, "grad_norm": 3.506012439727783, "learning_rate": 2.7929570367217804e-05, "loss": 0.0589, "step": 132650 }, { "epoch": 4.484938638899219, "grad_norm": 2.0019583702087402, "learning_rate": 2.792215232518166e-05, "loss": 0.0637, "step": 132660 }, { "epoch": 4.485276716589472, "grad_norm": 3.195699691772461, "learning_rate": 2.7914734886745574e-05, "loss": 0.0535, "step": 132670 }, { "epoch": 4.4856147942797255, "grad_norm": 2.432878017425537, "learning_rate": 2.7907318052112342e-05, "loss": 0.0565, "step": 132680 }, { "epoch": 4.485952871969979, "grad_norm": 2.9422287940979004, "learning_rate": 2.789990182148471e-05, "loss": 0.0564, "step": 132690 }, { "epoch": 4.486290949660232, "grad_norm": 4.999020099639893, "learning_rate": 2.789248619506548e-05, "loss": 0.0551, "step": 132700 }, { "epoch": 4.486629027350485, "grad_norm": 1.7201915979385376, "learning_rate": 2.7885071173057375e-05, "loss": 0.0564, "step": 132710 }, { "epoch": 4.486967105040739, "grad_norm": 1.4267774820327759, "learning_rate": 2.7877656755663095e-05, "loss": 0.064, "step": 132720 }, { "epoch": 4.487305182730991, "grad_norm": 1.2989219427108765, "learning_rate": 2.787024294308539e-05, "loss": 0.062, "step": 132730 }, { "epoch": 4.487643260421245, "grad_norm": 1.190753698348999, "learning_rate": 2.78628297355269e-05, "loss": 0.055, "step": 132740 }, { "epoch": 4.487981338111498, "grad_norm": 2.403522491455078, "learning_rate": 2.7855417133190375e-05, "loss": 0.0556, "step": 132750 }, { "epoch": 4.488319415801751, "grad_norm": 2.3869051933288574, "learning_rate": 2.7848005136278388e-05, "loss": 0.0652, "step": 132760 }, { "epoch": 4.4886574934920045, "grad_norm": 1.8519079685211182, "learning_rate": 2.7840593744993615e-05, "loss": 0.0574, "step": 132770 }, { "epoch": 4.488995571182258, "grad_norm": 3.4175965785980225, "learning_rate": 2.7833182959538706e-05, "loss": 0.063, "step": 132780 }, { "epoch": 4.489333648872511, "grad_norm": 1.3924006223678589, "learning_rate": 2.782577278011625e-05, "loss": 0.0571, "step": 132790 }, { "epoch": 4.489671726562764, "grad_norm": 10.184234619140625, "learning_rate": 2.781836320692883e-05, "loss": 0.0648, "step": 132800 }, { "epoch": 4.490009804253018, "grad_norm": 1.58555006980896, "learning_rate": 2.781095424017901e-05, "loss": 0.0633, "step": 132810 }, { "epoch": 4.49034788194327, "grad_norm": 0.8494608402252197, "learning_rate": 2.7803545880069386e-05, "loss": 0.0594, "step": 132820 }, { "epoch": 4.490685959633524, "grad_norm": 3.2743895053863525, "learning_rate": 2.779613812680248e-05, "loss": 0.0543, "step": 132830 }, { "epoch": 4.491024037323777, "grad_norm": 1.0490790605545044, "learning_rate": 2.778873098058079e-05, "loss": 0.0551, "step": 132840 }, { "epoch": 4.49136211501403, "grad_norm": 1.5825284719467163, "learning_rate": 2.7781324441606875e-05, "loss": 0.0609, "step": 132850 }, { "epoch": 4.4917001927042834, "grad_norm": 0.9609864354133606, "learning_rate": 2.7773918510083184e-05, "loss": 0.0566, "step": 132860 }, { "epoch": 4.492038270394537, "grad_norm": 2.383127212524414, "learning_rate": 2.776651318621223e-05, "loss": 0.0605, "step": 132870 }, { "epoch": 4.49237634808479, "grad_norm": 3.9811160564422607, "learning_rate": 2.7759108470196448e-05, "loss": 0.0629, "step": 132880 }, { "epoch": 4.492714425775043, "grad_norm": 1.449899673461914, "learning_rate": 2.7751704362238283e-05, "loss": 0.061, "step": 132890 }, { "epoch": 4.493052503465297, "grad_norm": 1.6717082262039185, "learning_rate": 2.774430086254015e-05, "loss": 0.0614, "step": 132900 }, { "epoch": 4.493390581155549, "grad_norm": 0.931752622127533, "learning_rate": 2.773689797130449e-05, "loss": 0.0639, "step": 132910 }, { "epoch": 4.493728658845803, "grad_norm": 2.8448967933654785, "learning_rate": 2.7729495688733676e-05, "loss": 0.0617, "step": 132920 }, { "epoch": 4.494066736536056, "grad_norm": 1.2353605031967163, "learning_rate": 2.7722094015030058e-05, "loss": 0.0644, "step": 132930 }, { "epoch": 4.494404814226309, "grad_norm": 1.4561455249786377, "learning_rate": 2.7714692950396048e-05, "loss": 0.0575, "step": 132940 }, { "epoch": 4.494742891916562, "grad_norm": 2.341707468032837, "learning_rate": 2.770729249503393e-05, "loss": 0.062, "step": 132950 }, { "epoch": 4.495080969606816, "grad_norm": 7.272103786468506, "learning_rate": 2.7699892649146085e-05, "loss": 0.0684, "step": 132960 }, { "epoch": 4.4954190472970685, "grad_norm": 5.661157608032227, "learning_rate": 2.76924934129348e-05, "loss": 0.0524, "step": 132970 }, { "epoch": 4.495757124987322, "grad_norm": 1.7047135829925537, "learning_rate": 2.7685094786602344e-05, "loss": 0.0624, "step": 132980 }, { "epoch": 4.496095202677576, "grad_norm": 1.938265085220337, "learning_rate": 2.767769677035103e-05, "loss": 0.0673, "step": 132990 }, { "epoch": 4.496433280367828, "grad_norm": 10.182535171508789, "learning_rate": 2.76702993643831e-05, "loss": 0.0592, "step": 133000 }, { "epoch": 4.496771358058082, "grad_norm": 1.3399009704589844, "learning_rate": 2.7662902568900807e-05, "loss": 0.0524, "step": 133010 }, { "epoch": 4.497109435748335, "grad_norm": 3.3198204040527344, "learning_rate": 2.7655506384106334e-05, "loss": 0.0593, "step": 133020 }, { "epoch": 4.497447513438588, "grad_norm": 2.4438636302948, "learning_rate": 2.7648110810201923e-05, "loss": 0.0557, "step": 133030 }, { "epoch": 4.497785591128841, "grad_norm": 10.251264572143555, "learning_rate": 2.7640715847389808e-05, "loss": 0.0609, "step": 133040 }, { "epoch": 4.498123668819095, "grad_norm": 1.020188570022583, "learning_rate": 2.7633321495872084e-05, "loss": 0.0564, "step": 133050 }, { "epoch": 4.4984617465093475, "grad_norm": 1.4800502061843872, "learning_rate": 2.7625927755850965e-05, "loss": 0.0548, "step": 133060 }, { "epoch": 4.498799824199601, "grad_norm": 6.047771453857422, "learning_rate": 2.761853462752856e-05, "loss": 0.0594, "step": 133070 }, { "epoch": 4.4991379018898545, "grad_norm": 0.9262950420379639, "learning_rate": 2.7611142111107018e-05, "loss": 0.0556, "step": 133080 }, { "epoch": 4.499475979580107, "grad_norm": 0.9503784775733948, "learning_rate": 2.760375020678845e-05, "loss": 0.0529, "step": 133090 }, { "epoch": 4.499814057270361, "grad_norm": 0.7052894830703735, "learning_rate": 2.7596358914774913e-05, "loss": 0.0635, "step": 133100 }, { "epoch": 4.500152134960614, "grad_norm": 15.248459815979004, "learning_rate": 2.758896823526853e-05, "loss": 0.0539, "step": 133110 }, { "epoch": 4.500490212650867, "grad_norm": 1.2342925071716309, "learning_rate": 2.7581578168471333e-05, "loss": 0.0627, "step": 133120 }, { "epoch": 4.50082829034112, "grad_norm": 0.6636614799499512, "learning_rate": 2.7574188714585368e-05, "loss": 0.0577, "step": 133130 }, { "epoch": 4.501166368031374, "grad_norm": 0.9851417541503906, "learning_rate": 2.756679987381264e-05, "loss": 0.0607, "step": 133140 }, { "epoch": 4.5015044457216264, "grad_norm": 1.0113182067871094, "learning_rate": 2.7559411646355195e-05, "loss": 0.0601, "step": 133150 }, { "epoch": 4.50184252341188, "grad_norm": 0.7045333385467529, "learning_rate": 2.755202403241498e-05, "loss": 0.0611, "step": 133160 }, { "epoch": 4.5021806011021335, "grad_norm": 2.501746416091919, "learning_rate": 2.754463703219402e-05, "loss": 0.0587, "step": 133170 }, { "epoch": 4.502518678792386, "grad_norm": 3.391277313232422, "learning_rate": 2.7537250645894252e-05, "loss": 0.0592, "step": 133180 }, { "epoch": 4.50285675648264, "grad_norm": 3.2578444480895996, "learning_rate": 2.7529864873717582e-05, "loss": 0.0644, "step": 133190 }, { "epoch": 4.503194834172893, "grad_norm": 1.8961513042449951, "learning_rate": 2.752247971586599e-05, "loss": 0.067, "step": 133200 }, { "epoch": 4.503532911863147, "grad_norm": 2.4251291751861572, "learning_rate": 2.751509517254136e-05, "loss": 0.0619, "step": 133210 }, { "epoch": 4.503870989553399, "grad_norm": 1.6720187664031982, "learning_rate": 2.7507711243945554e-05, "loss": 0.0533, "step": 133220 }, { "epoch": 4.504209067243653, "grad_norm": 2.174292802810669, "learning_rate": 2.75003279302805e-05, "loss": 0.066, "step": 133230 }, { "epoch": 4.504547144933905, "grad_norm": 6.178433895111084, "learning_rate": 2.7492945231748003e-05, "loss": 0.0645, "step": 133240 }, { "epoch": 4.504885222624159, "grad_norm": 2.1190972328186035, "learning_rate": 2.7485563148549974e-05, "loss": 0.0634, "step": 133250 }, { "epoch": 4.505223300314412, "grad_norm": 0.783755362033844, "learning_rate": 2.7478181680888145e-05, "loss": 0.0582, "step": 133260 }, { "epoch": 4.505561378004666, "grad_norm": 7.964619159698486, "learning_rate": 2.7470800828964387e-05, "loss": 0.0613, "step": 133270 }, { "epoch": 4.505899455694919, "grad_norm": 1.241595983505249, "learning_rate": 2.7463420592980455e-05, "loss": 0.0642, "step": 133280 }, { "epoch": 4.506237533385172, "grad_norm": 1.5432876348495483, "learning_rate": 2.7456040973138154e-05, "loss": 0.063, "step": 133290 }, { "epoch": 4.506575611075425, "grad_norm": 1.8144985437393188, "learning_rate": 2.7448661969639222e-05, "loss": 0.0589, "step": 133300 }, { "epoch": 4.506913688765678, "grad_norm": 1.890823483467102, "learning_rate": 2.7441283582685384e-05, "loss": 0.0615, "step": 133310 }, { "epoch": 4.507251766455932, "grad_norm": 2.2585959434509277, "learning_rate": 2.7433905812478393e-05, "loss": 0.0583, "step": 133320 }, { "epoch": 4.507589844146185, "grad_norm": 1.1758428812026978, "learning_rate": 2.742652865921994e-05, "loss": 0.0553, "step": 133330 }, { "epoch": 4.507927921836438, "grad_norm": 2.081629514694214, "learning_rate": 2.7419152123111718e-05, "loss": 0.0652, "step": 133340 }, { "epoch": 4.508265999526691, "grad_norm": 2.0431413650512695, "learning_rate": 2.741177620435537e-05, "loss": 0.0571, "step": 133350 }, { "epoch": 4.508604077216944, "grad_norm": 5.5213775634765625, "learning_rate": 2.7404400903152578e-05, "loss": 0.0624, "step": 133360 }, { "epoch": 4.5089421549071975, "grad_norm": 0.9804826378822327, "learning_rate": 2.739702621970499e-05, "loss": 0.0644, "step": 133370 }, { "epoch": 4.509280232597451, "grad_norm": 1.2903963327407837, "learning_rate": 2.738965215421422e-05, "loss": 0.0621, "step": 133380 }, { "epoch": 4.5096183102877045, "grad_norm": 1.231966257095337, "learning_rate": 2.738227870688186e-05, "loss": 0.0538, "step": 133390 }, { "epoch": 4.509956387977957, "grad_norm": 1.0110409259796143, "learning_rate": 2.7374905877909492e-05, "loss": 0.0569, "step": 133400 }, { "epoch": 4.510294465668211, "grad_norm": 10.947840690612793, "learning_rate": 2.7367533667498712e-05, "loss": 0.06, "step": 133410 }, { "epoch": 4.510632543358463, "grad_norm": 1.1094176769256592, "learning_rate": 2.736016207585106e-05, "loss": 0.0535, "step": 133420 }, { "epoch": 4.510970621048717, "grad_norm": 2.760159730911255, "learning_rate": 2.7352791103168046e-05, "loss": 0.0597, "step": 133430 }, { "epoch": 4.51130869873897, "grad_norm": 0.8313745856285095, "learning_rate": 2.7345420749651242e-05, "loss": 0.0602, "step": 133440 }, { "epoch": 4.511646776429224, "grad_norm": 0.8439033031463623, "learning_rate": 2.7338051015502096e-05, "loss": 0.0585, "step": 133450 }, { "epoch": 4.5119848541194765, "grad_norm": 1.7179691791534424, "learning_rate": 2.7330681900922166e-05, "loss": 0.0571, "step": 133460 }, { "epoch": 4.51232293180973, "grad_norm": 1.0997226238250732, "learning_rate": 2.732331340611284e-05, "loss": 0.0622, "step": 133470 }, { "epoch": 4.512661009499983, "grad_norm": 0.9400103688240051, "learning_rate": 2.73159455312756e-05, "loss": 0.0651, "step": 133480 }, { "epoch": 4.512999087190236, "grad_norm": 2.112426519393921, "learning_rate": 2.7308578276611912e-05, "loss": 0.0616, "step": 133490 }, { "epoch": 4.51333716488049, "grad_norm": 2.0658371448516846, "learning_rate": 2.7301211642323165e-05, "loss": 0.0562, "step": 133500 }, { "epoch": 4.513675242570743, "grad_norm": 1.2909531593322754, "learning_rate": 2.729384562861077e-05, "loss": 0.0592, "step": 133510 }, { "epoch": 4.514013320260996, "grad_norm": 1.4541161060333252, "learning_rate": 2.728648023567608e-05, "loss": 0.0535, "step": 133520 }, { "epoch": 4.514351397951249, "grad_norm": 1.3087947368621826, "learning_rate": 2.7279115463720505e-05, "loss": 0.0651, "step": 133530 }, { "epoch": 4.514689475641503, "grad_norm": 1.7951314449310303, "learning_rate": 2.7271751312945387e-05, "loss": 0.0696, "step": 133540 }, { "epoch": 4.515027553331755, "grad_norm": 3.467609167098999, "learning_rate": 2.7264387783552026e-05, "loss": 0.0673, "step": 133550 }, { "epoch": 4.515365631022009, "grad_norm": 0.991723895072937, "learning_rate": 2.7257024875741776e-05, "loss": 0.0619, "step": 133560 }, { "epoch": 4.5157037087122625, "grad_norm": 1.0667316913604736, "learning_rate": 2.724966258971591e-05, "loss": 0.063, "step": 133570 }, { "epoch": 4.516041786402515, "grad_norm": 1.643800973892212, "learning_rate": 2.7242300925675733e-05, "loss": 0.0575, "step": 133580 }, { "epoch": 4.516379864092769, "grad_norm": 1.3371717929840088, "learning_rate": 2.7234939883822508e-05, "loss": 0.0609, "step": 133590 }, { "epoch": 4.516717941783022, "grad_norm": 1.0214601755142212, "learning_rate": 2.7227579464357444e-05, "loss": 0.0588, "step": 133600 }, { "epoch": 4.517056019473275, "grad_norm": 1.1194621324539185, "learning_rate": 2.7220219667481826e-05, "loss": 0.061, "step": 133610 }, { "epoch": 4.517394097163528, "grad_norm": 1.1941540241241455, "learning_rate": 2.7212860493396853e-05, "loss": 0.0631, "step": 133620 }, { "epoch": 4.517732174853782, "grad_norm": 0.9131146669387817, "learning_rate": 2.72055019423037e-05, "loss": 0.0616, "step": 133630 }, { "epoch": 4.518070252544034, "grad_norm": 1.0042067766189575, "learning_rate": 2.7198144014403548e-05, "loss": 0.0596, "step": 133640 }, { "epoch": 4.518408330234288, "grad_norm": 1.3967753648757935, "learning_rate": 2.7190786709897594e-05, "loss": 0.061, "step": 133650 }, { "epoch": 4.518746407924541, "grad_norm": 1.4759994745254517, "learning_rate": 2.7183430028986935e-05, "loss": 0.0638, "step": 133660 }, { "epoch": 4.519084485614794, "grad_norm": 1.5674099922180176, "learning_rate": 2.717607397187275e-05, "loss": 0.0616, "step": 133670 }, { "epoch": 4.5194225633050475, "grad_norm": 1.4657433032989502, "learning_rate": 2.716871853875613e-05, "loss": 0.0602, "step": 133680 }, { "epoch": 4.519760640995301, "grad_norm": 4.846495151519775, "learning_rate": 2.7161363729838152e-05, "loss": 0.0577, "step": 133690 }, { "epoch": 4.520098718685554, "grad_norm": 1.2455925941467285, "learning_rate": 2.7154009545319926e-05, "loss": 0.0565, "step": 133700 }, { "epoch": 4.520436796375807, "grad_norm": 3.284134864807129, "learning_rate": 2.714665598540249e-05, "loss": 0.0597, "step": 133710 }, { "epoch": 4.520774874066061, "grad_norm": 4.483253479003906, "learning_rate": 2.71393030502869e-05, "loss": 0.0583, "step": 133720 }, { "epoch": 4.521112951756313, "grad_norm": 1.4178916215896606, "learning_rate": 2.713195074017416e-05, "loss": 0.058, "step": 133730 }, { "epoch": 4.521451029446567, "grad_norm": 3.6039953231811523, "learning_rate": 2.7124599055265286e-05, "loss": 0.0581, "step": 133740 }, { "epoch": 4.52178910713682, "grad_norm": 0.9904290437698364, "learning_rate": 2.7117247995761332e-05, "loss": 0.0609, "step": 133750 }, { "epoch": 4.522127184827073, "grad_norm": 2.3022756576538086, "learning_rate": 2.7109897561863173e-05, "loss": 0.052, "step": 133760 }, { "epoch": 4.5224652625173265, "grad_norm": 1.1345936059951782, "learning_rate": 2.7102547753771833e-05, "loss": 0.0631, "step": 133770 }, { "epoch": 4.52280334020758, "grad_norm": 0.8680142164230347, "learning_rate": 2.7095198571688217e-05, "loss": 0.0604, "step": 133780 }, { "epoch": 4.523141417897833, "grad_norm": 1.15824556350708, "learning_rate": 2.7087850015813287e-05, "loss": 0.0551, "step": 133790 }, { "epoch": 4.523479495588086, "grad_norm": 2.7719974517822266, "learning_rate": 2.7080502086347936e-05, "loss": 0.0595, "step": 133800 }, { "epoch": 4.52381757327834, "grad_norm": 0.9737204313278198, "learning_rate": 2.7073154783493023e-05, "loss": 0.0584, "step": 133810 }, { "epoch": 4.524155650968592, "grad_norm": 1.3258060216903687, "learning_rate": 2.7065808107449463e-05, "loss": 0.0589, "step": 133820 }, { "epoch": 4.524493728658846, "grad_norm": 2.567368745803833, "learning_rate": 2.70584620584181e-05, "loss": 0.0562, "step": 133830 }, { "epoch": 4.524831806349099, "grad_norm": 1.458799123764038, "learning_rate": 2.705111663659976e-05, "loss": 0.053, "step": 133840 }, { "epoch": 4.525169884039352, "grad_norm": 1.4886293411254883, "learning_rate": 2.7043771842195248e-05, "loss": 0.065, "step": 133850 }, { "epoch": 4.5255079617296055, "grad_norm": 6.2095513343811035, "learning_rate": 2.7036427675405395e-05, "loss": 0.055, "step": 133860 }, { "epoch": 4.525846039419859, "grad_norm": 1.6464309692382812, "learning_rate": 2.7029084136430992e-05, "loss": 0.0619, "step": 133870 }, { "epoch": 4.526184117110112, "grad_norm": 3.260122299194336, "learning_rate": 2.7021741225472806e-05, "loss": 0.0521, "step": 133880 }, { "epoch": 4.526522194800365, "grad_norm": 4.777489185333252, "learning_rate": 2.7014398942731588e-05, "loss": 0.0602, "step": 133890 }, { "epoch": 4.526860272490619, "grad_norm": 1.65574049949646, "learning_rate": 2.7007057288408044e-05, "loss": 0.057, "step": 133900 }, { "epoch": 4.527198350180871, "grad_norm": 1.8822448253631592, "learning_rate": 2.6999716262702934e-05, "loss": 0.0553, "step": 133910 }, { "epoch": 4.527536427871125, "grad_norm": 2.492225170135498, "learning_rate": 2.6992375865816944e-05, "loss": 0.0615, "step": 133920 }, { "epoch": 4.527874505561378, "grad_norm": 5.102876663208008, "learning_rate": 2.698503609795073e-05, "loss": 0.0637, "step": 133930 }, { "epoch": 4.528212583251631, "grad_norm": 1.2235766649246216, "learning_rate": 2.6977696959305e-05, "loss": 0.063, "step": 133940 }, { "epoch": 4.528550660941884, "grad_norm": 1.8791075944900513, "learning_rate": 2.697035845008037e-05, "loss": 0.062, "step": 133950 }, { "epoch": 4.528888738632138, "grad_norm": 1.168080449104309, "learning_rate": 2.6963020570477527e-05, "loss": 0.0582, "step": 133960 }, { "epoch": 4.5292268163223905, "grad_norm": 2.1384286880493164, "learning_rate": 2.695568332069701e-05, "loss": 0.0583, "step": 133970 }, { "epoch": 4.529564894012644, "grad_norm": 2.7744944095611572, "learning_rate": 2.6948346700939452e-05, "loss": 0.0578, "step": 133980 }, { "epoch": 4.529902971702898, "grad_norm": 1.694847822189331, "learning_rate": 2.6941010711405458e-05, "loss": 0.0576, "step": 133990 }, { "epoch": 4.53024104939315, "grad_norm": 6.021186828613281, "learning_rate": 2.6933675352295563e-05, "loss": 0.0573, "step": 134000 }, { "epoch": 4.530579127083404, "grad_norm": 8.098760604858398, "learning_rate": 2.692634062381032e-05, "loss": 0.057, "step": 134010 }, { "epoch": 4.530917204773657, "grad_norm": 0.9072909355163574, "learning_rate": 2.6919006526150237e-05, "loss": 0.0532, "step": 134020 }, { "epoch": 4.53125528246391, "grad_norm": 2.065021276473999, "learning_rate": 2.6911673059515863e-05, "loss": 0.0598, "step": 134030 }, { "epoch": 4.531593360154163, "grad_norm": 1.299538493156433, "learning_rate": 2.690434022410767e-05, "loss": 0.0582, "step": 134040 }, { "epoch": 4.531931437844417, "grad_norm": 1.8951616287231445, "learning_rate": 2.6897008020126125e-05, "loss": 0.0559, "step": 134050 }, { "epoch": 4.5322695155346695, "grad_norm": 4.459753036499023, "learning_rate": 2.6889676447771718e-05, "loss": 0.0539, "step": 134060 }, { "epoch": 4.532607593224923, "grad_norm": 1.6277265548706055, "learning_rate": 2.6882345507244855e-05, "loss": 0.0621, "step": 134070 }, { "epoch": 4.5329456709151765, "grad_norm": 2.9623005390167236, "learning_rate": 2.6875015198746002e-05, "loss": 0.0559, "step": 134080 }, { "epoch": 4.533283748605429, "grad_norm": 2.787815809249878, "learning_rate": 2.686768552247555e-05, "loss": 0.061, "step": 134090 }, { "epoch": 4.533621826295683, "grad_norm": 1.4020401239395142, "learning_rate": 2.686035647863388e-05, "loss": 0.0558, "step": 134100 }, { "epoch": 4.533959903985936, "grad_norm": 2.8676304817199707, "learning_rate": 2.6853028067421348e-05, "loss": 0.0606, "step": 134110 }, { "epoch": 4.534297981676189, "grad_norm": 1.22694730758667, "learning_rate": 2.684570028903836e-05, "loss": 0.0598, "step": 134120 }, { "epoch": 4.534636059366442, "grad_norm": 2.167489767074585, "learning_rate": 2.6838373143685224e-05, "loss": 0.0631, "step": 134130 }, { "epoch": 4.534974137056696, "grad_norm": 11.005229949951172, "learning_rate": 2.6831046631562247e-05, "loss": 0.0632, "step": 134140 }, { "epoch": 4.5353122147469485, "grad_norm": 0.9976478815078735, "learning_rate": 2.682372075286977e-05, "loss": 0.0659, "step": 134150 }, { "epoch": 4.535650292437202, "grad_norm": 1.3593907356262207, "learning_rate": 2.6816395507808035e-05, "loss": 0.0595, "step": 134160 }, { "epoch": 4.5359883701274555, "grad_norm": 1.076362133026123, "learning_rate": 2.6809070896577366e-05, "loss": 0.0618, "step": 134170 }, { "epoch": 4.536326447817709, "grad_norm": 1.6110217571258545, "learning_rate": 2.6801746919377977e-05, "loss": 0.0614, "step": 134180 }, { "epoch": 4.536664525507962, "grad_norm": 0.5932989120483398, "learning_rate": 2.67944235764101e-05, "loss": 0.0574, "step": 134190 }, { "epoch": 4.537002603198215, "grad_norm": 1.0332462787628174, "learning_rate": 2.6787100867873983e-05, "loss": 0.0581, "step": 134200 }, { "epoch": 4.537340680888468, "grad_norm": 4.217190265655518, "learning_rate": 2.67797787939698e-05, "loss": 0.0646, "step": 134210 }, { "epoch": 4.537678758578721, "grad_norm": 1.1929041147232056, "learning_rate": 2.6772457354897752e-05, "loss": 0.0605, "step": 134220 }, { "epoch": 4.538016836268975, "grad_norm": 1.4896669387817383, "learning_rate": 2.676513655085797e-05, "loss": 0.0597, "step": 134230 }, { "epoch": 4.538354913959228, "grad_norm": 1.2971572875976562, "learning_rate": 2.6757816382050627e-05, "loss": 0.0632, "step": 134240 }, { "epoch": 4.538692991649481, "grad_norm": 2.0910913944244385, "learning_rate": 2.675049684867589e-05, "loss": 0.0628, "step": 134250 }, { "epoch": 4.539031069339734, "grad_norm": 1.973419427871704, "learning_rate": 2.6743177950933795e-05, "loss": 0.0631, "step": 134260 }, { "epoch": 4.539369147029987, "grad_norm": 1.9776257276535034, "learning_rate": 2.6735859689024494e-05, "loss": 0.0648, "step": 134270 }, { "epoch": 4.539707224720241, "grad_norm": 2.7657852172851562, "learning_rate": 2.6728542063148032e-05, "loss": 0.0606, "step": 134280 }, { "epoch": 4.540045302410494, "grad_norm": 1.2164658308029175, "learning_rate": 2.672122507350451e-05, "loss": 0.0577, "step": 134290 }, { "epoch": 4.540383380100748, "grad_norm": 1.2942209243774414, "learning_rate": 2.6713908720293945e-05, "loss": 0.058, "step": 134300 }, { "epoch": 4.540721457791, "grad_norm": 1.1720223426818848, "learning_rate": 2.670659300371635e-05, "loss": 0.0582, "step": 134310 }, { "epoch": 4.541059535481254, "grad_norm": 1.2476145029067993, "learning_rate": 2.6699277923971776e-05, "loss": 0.0578, "step": 134320 }, { "epoch": 4.541397613171506, "grad_norm": 1.6003330945968628, "learning_rate": 2.6691963481260184e-05, "loss": 0.0599, "step": 134330 }, { "epoch": 4.54173569086176, "grad_norm": 1.1853318214416504, "learning_rate": 2.6684649675781553e-05, "loss": 0.0575, "step": 134340 }, { "epoch": 4.542073768552013, "grad_norm": 0.7148733139038086, "learning_rate": 2.6677336507735823e-05, "loss": 0.0507, "step": 134350 }, { "epoch": 4.542411846242267, "grad_norm": 1.4511843919754028, "learning_rate": 2.6670023977322972e-05, "loss": 0.056, "step": 134360 }, { "epoch": 4.5427499239325195, "grad_norm": 26.145326614379883, "learning_rate": 2.666271208474288e-05, "loss": 0.0588, "step": 134370 }, { "epoch": 4.543088001622773, "grad_norm": 1.297700047492981, "learning_rate": 2.6655400830195486e-05, "loss": 0.0652, "step": 134380 }, { "epoch": 4.543426079313026, "grad_norm": 1.3976738452911377, "learning_rate": 2.6648090213880674e-05, "loss": 0.0602, "step": 134390 }, { "epoch": 4.543764157003279, "grad_norm": 2.106340169906616, "learning_rate": 2.6640780235998276e-05, "loss": 0.0554, "step": 134400 }, { "epoch": 4.544102234693533, "grad_norm": 2.5840909481048584, "learning_rate": 2.663347089674819e-05, "loss": 0.0554, "step": 134410 }, { "epoch": 4.544440312383786, "grad_norm": 2.925921678543091, "learning_rate": 2.6626162196330235e-05, "loss": 0.0638, "step": 134420 }, { "epoch": 4.544778390074039, "grad_norm": 9.566205978393555, "learning_rate": 2.6618854134944205e-05, "loss": 0.0647, "step": 134430 }, { "epoch": 4.545116467764292, "grad_norm": 3.087867021560669, "learning_rate": 2.6611546712789938e-05, "loss": 0.066, "step": 134440 }, { "epoch": 4.545454545454545, "grad_norm": 3.980938196182251, "learning_rate": 2.6604239930067176e-05, "loss": 0.0585, "step": 134450 }, { "epoch": 4.5457926231447985, "grad_norm": 2.7499189376831055, "learning_rate": 2.6596933786975743e-05, "loss": 0.0587, "step": 134460 }, { "epoch": 4.546130700835052, "grad_norm": 2.7734155654907227, "learning_rate": 2.6589628283715305e-05, "loss": 0.0664, "step": 134470 }, { "epoch": 4.5464687785253055, "grad_norm": 1.237011194229126, "learning_rate": 2.6582323420485665e-05, "loss": 0.0624, "step": 134480 }, { "epoch": 4.546806856215558, "grad_norm": 3.2041683197021484, "learning_rate": 2.6575019197486478e-05, "loss": 0.0608, "step": 134490 }, { "epoch": 4.547144933905812, "grad_norm": 1.9279626607894897, "learning_rate": 2.6567715614917487e-05, "loss": 0.059, "step": 134500 }, { "epoch": 4.547483011596065, "grad_norm": 1.8504621982574463, "learning_rate": 2.6560412672978353e-05, "loss": 0.0625, "step": 134510 }, { "epoch": 4.547821089286318, "grad_norm": 4.513645648956299, "learning_rate": 2.6553110371868706e-05, "loss": 0.063, "step": 134520 }, { "epoch": 4.548159166976571, "grad_norm": 1.897128701210022, "learning_rate": 2.6545808711788234e-05, "loss": 0.0621, "step": 134530 }, { "epoch": 4.548497244666825, "grad_norm": 0.901578426361084, "learning_rate": 2.653850769293654e-05, "loss": 0.0621, "step": 134540 }, { "epoch": 4.548835322357077, "grad_norm": 3.1229372024536133, "learning_rate": 2.653120731551323e-05, "loss": 0.0606, "step": 134550 }, { "epoch": 4.549173400047331, "grad_norm": 2.500288724899292, "learning_rate": 2.6523907579717878e-05, "loss": 0.058, "step": 134560 }, { "epoch": 4.5495114777375845, "grad_norm": 1.7863969802856445, "learning_rate": 2.6516608485750073e-05, "loss": 0.0617, "step": 134570 }, { "epoch": 4.549849555427837, "grad_norm": 3.4490926265716553, "learning_rate": 2.6509310033809386e-05, "loss": 0.0608, "step": 134580 }, { "epoch": 4.550187633118091, "grad_norm": 1.697717308998108, "learning_rate": 2.650201222409534e-05, "loss": 0.054, "step": 134590 }, { "epoch": 4.550525710808344, "grad_norm": 4.060670852661133, "learning_rate": 2.649471505680745e-05, "loss": 0.0593, "step": 134600 }, { "epoch": 4.550863788498597, "grad_norm": 2.6070473194122314, "learning_rate": 2.6487418532145202e-05, "loss": 0.062, "step": 134610 }, { "epoch": 4.55120186618885, "grad_norm": 1.3447619676589966, "learning_rate": 2.648012265030812e-05, "loss": 0.0601, "step": 134620 }, { "epoch": 4.551539943879104, "grad_norm": 2.176332473754883, "learning_rate": 2.647282741149565e-05, "loss": 0.068, "step": 134630 }, { "epoch": 4.551878021569356, "grad_norm": 2.5496914386749268, "learning_rate": 2.6465532815907214e-05, "loss": 0.0596, "step": 134640 }, { "epoch": 4.55221609925961, "grad_norm": 3.0059568881988525, "learning_rate": 2.645823886374229e-05, "loss": 0.0575, "step": 134650 }, { "epoch": 4.552554176949863, "grad_norm": 5.578258991241455, "learning_rate": 2.645094555520026e-05, "loss": 0.0602, "step": 134660 }, { "epoch": 4.552892254640116, "grad_norm": 3.1031410694122314, "learning_rate": 2.6443652890480558e-05, "loss": 0.057, "step": 134670 }, { "epoch": 4.55323033233037, "grad_norm": 2.604891061782837, "learning_rate": 2.6436360869782504e-05, "loss": 0.0655, "step": 134680 }, { "epoch": 4.553568410020623, "grad_norm": 1.6625826358795166, "learning_rate": 2.6429069493305493e-05, "loss": 0.0604, "step": 134690 }, { "epoch": 4.553906487710876, "grad_norm": 1.7676324844360352, "learning_rate": 2.6421778761248884e-05, "loss": 0.0585, "step": 134700 }, { "epoch": 4.554244565401129, "grad_norm": 1.7445838451385498, "learning_rate": 2.641448867381199e-05, "loss": 0.0642, "step": 134710 }, { "epoch": 4.554582643091383, "grad_norm": 1.025159478187561, "learning_rate": 2.6407199231194103e-05, "loss": 0.0553, "step": 134720 }, { "epoch": 4.554920720781635, "grad_norm": 1.3796666860580444, "learning_rate": 2.6399910433594514e-05, "loss": 0.0535, "step": 134730 }, { "epoch": 4.555258798471889, "grad_norm": 1.1175470352172852, "learning_rate": 2.6392622281212532e-05, "loss": 0.0614, "step": 134740 }, { "epoch": 4.555596876162142, "grad_norm": 1.1900103092193604, "learning_rate": 2.638533477424738e-05, "loss": 0.0577, "step": 134750 }, { "epoch": 4.555934953852395, "grad_norm": 3.1641008853912354, "learning_rate": 2.6378047912898286e-05, "loss": 0.0604, "step": 134760 }, { "epoch": 4.5562730315426485, "grad_norm": 1.4253557920455933, "learning_rate": 2.637076169736451e-05, "loss": 0.0541, "step": 134770 }, { "epoch": 4.556611109232902, "grad_norm": 1.5592957735061646, "learning_rate": 2.636347612784521e-05, "loss": 0.0554, "step": 134780 }, { "epoch": 4.556949186923155, "grad_norm": 3.07714581489563, "learning_rate": 2.6356191204539614e-05, "loss": 0.0599, "step": 134790 }, { "epoch": 4.557287264613408, "grad_norm": 2.5444746017456055, "learning_rate": 2.6348906927646866e-05, "loss": 0.0521, "step": 134800 }, { "epoch": 4.557625342303662, "grad_norm": 2.389941930770874, "learning_rate": 2.6341623297366103e-05, "loss": 0.0621, "step": 134810 }, { "epoch": 4.557963419993914, "grad_norm": 1.4804197549819946, "learning_rate": 2.6334340313896484e-05, "loss": 0.0586, "step": 134820 }, { "epoch": 4.558301497684168, "grad_norm": 3.996507167816162, "learning_rate": 2.6327057977437108e-05, "loss": 0.0575, "step": 134830 }, { "epoch": 4.558639575374421, "grad_norm": 0.849563479423523, "learning_rate": 2.6319776288187083e-05, "loss": 0.0584, "step": 134840 }, { "epoch": 4.558977653064674, "grad_norm": 0.9309192895889282, "learning_rate": 2.6312495246345454e-05, "loss": 0.0612, "step": 134850 }, { "epoch": 4.5593157307549275, "grad_norm": 1.6514819860458374, "learning_rate": 2.6305214852111325e-05, "loss": 0.0612, "step": 134860 }, { "epoch": 4.559653808445181, "grad_norm": 2.335685968399048, "learning_rate": 2.62979351056837e-05, "loss": 0.0689, "step": 134870 }, { "epoch": 4.559991886135434, "grad_norm": 0.8530099391937256, "learning_rate": 2.6290656007261642e-05, "loss": 0.0686, "step": 134880 }, { "epoch": 4.560329963825687, "grad_norm": 3.5049777030944824, "learning_rate": 2.6283377557044144e-05, "loss": 0.0573, "step": 134890 }, { "epoch": 4.560668041515941, "grad_norm": 1.279998540878296, "learning_rate": 2.6276099755230176e-05, "loss": 0.0591, "step": 134900 }, { "epoch": 4.561006119206193, "grad_norm": 1.293962836265564, "learning_rate": 2.626882260201875e-05, "loss": 0.0564, "step": 134910 }, { "epoch": 4.561344196896447, "grad_norm": 0.7933605909347534, "learning_rate": 2.6261546097608792e-05, "loss": 0.062, "step": 134920 }, { "epoch": 4.5616822745867, "grad_norm": 3.3728835582733154, "learning_rate": 2.6254270242199247e-05, "loss": 0.0601, "step": 134930 }, { "epoch": 4.562020352276953, "grad_norm": 1.5441769361495972, "learning_rate": 2.6246995035989018e-05, "loss": 0.0597, "step": 134940 }, { "epoch": 4.562358429967206, "grad_norm": 1.5144963264465332, "learning_rate": 2.6239720479177015e-05, "loss": 0.0623, "step": 134950 }, { "epoch": 4.56269650765746, "grad_norm": 0.7146522998809814, "learning_rate": 2.623244657196217e-05, "loss": 0.0499, "step": 134960 }, { "epoch": 4.563034585347713, "grad_norm": 1.164465308189392, "learning_rate": 2.6225173314543266e-05, "loss": 0.0565, "step": 134970 }, { "epoch": 4.563372663037966, "grad_norm": 1.196141242980957, "learning_rate": 2.6217900707119213e-05, "loss": 0.0546, "step": 134980 }, { "epoch": 4.56371074072822, "grad_norm": 1.1302911043167114, "learning_rate": 2.6210628749888798e-05, "loss": 0.0589, "step": 134990 }, { "epoch": 4.564048818418472, "grad_norm": 1.9384121894836426, "learning_rate": 2.620335744305088e-05, "loss": 0.0633, "step": 135000 }, { "epoch": 4.564386896108726, "grad_norm": 1.9414516687393188, "learning_rate": 2.619608678680423e-05, "loss": 0.0564, "step": 135010 }, { "epoch": 4.564724973798979, "grad_norm": 1.5753730535507202, "learning_rate": 2.61888167813476e-05, "loss": 0.0542, "step": 135020 }, { "epoch": 4.565063051489232, "grad_norm": 8.249584197998047, "learning_rate": 2.6181547426879793e-05, "loss": 0.0586, "step": 135030 }, { "epoch": 4.565401129179485, "grad_norm": 7.060030937194824, "learning_rate": 2.6174278723599533e-05, "loss": 0.054, "step": 135040 }, { "epoch": 4.565739206869739, "grad_norm": 1.6707117557525635, "learning_rate": 2.6167010671705543e-05, "loss": 0.0599, "step": 135050 }, { "epoch": 4.5660772845599915, "grad_norm": 1.4618830680847168, "learning_rate": 2.615974327139651e-05, "loss": 0.0621, "step": 135060 }, { "epoch": 4.566415362250245, "grad_norm": 2.512784242630005, "learning_rate": 2.615247652287114e-05, "loss": 0.0641, "step": 135070 }, { "epoch": 4.5667534399404985, "grad_norm": 2.264869213104248, "learning_rate": 2.6145210426328114e-05, "loss": 0.0646, "step": 135080 }, { "epoch": 4.567091517630751, "grad_norm": 1.5118441581726074, "learning_rate": 2.6137944981966078e-05, "loss": 0.0553, "step": 135090 }, { "epoch": 4.567429595321005, "grad_norm": 1.0879939794540405, "learning_rate": 2.6130680189983665e-05, "loss": 0.0547, "step": 135100 }, { "epoch": 4.567767673011258, "grad_norm": 1.7756240367889404, "learning_rate": 2.6123416050579463e-05, "loss": 0.0569, "step": 135110 }, { "epoch": 4.568105750701511, "grad_norm": 0.8247563242912292, "learning_rate": 2.6116152563952118e-05, "loss": 0.0589, "step": 135120 }, { "epoch": 4.568443828391764, "grad_norm": 1.6727288961410522, "learning_rate": 2.610888973030019e-05, "loss": 0.0585, "step": 135130 }, { "epoch": 4.568781906082018, "grad_norm": 1.127225399017334, "learning_rate": 2.6101627549822223e-05, "loss": 0.0628, "step": 135140 }, { "epoch": 4.569119983772271, "grad_norm": 2.991887331008911, "learning_rate": 2.60943660227168e-05, "loss": 0.0582, "step": 135150 }, { "epoch": 4.569458061462524, "grad_norm": 1.7622554302215576, "learning_rate": 2.6087105149182407e-05, "loss": 0.0582, "step": 135160 }, { "epoch": 4.5697961391527775, "grad_norm": 0.9312082529067993, "learning_rate": 2.6079844929417612e-05, "loss": 0.0587, "step": 135170 }, { "epoch": 4.57013421684303, "grad_norm": 0.9736689329147339, "learning_rate": 2.607258536362084e-05, "loss": 0.0559, "step": 135180 }, { "epoch": 4.570472294533284, "grad_norm": 2.2705936431884766, "learning_rate": 2.6065326451990584e-05, "loss": 0.0584, "step": 135190 }, { "epoch": 4.570810372223537, "grad_norm": 1.8489701747894287, "learning_rate": 2.605806819472534e-05, "loss": 0.0593, "step": 135200 }, { "epoch": 4.571148449913791, "grad_norm": 1.4464051723480225, "learning_rate": 2.6050810592023513e-05, "loss": 0.06, "step": 135210 }, { "epoch": 4.571486527604043, "grad_norm": 1.5325493812561035, "learning_rate": 2.604355364408353e-05, "loss": 0.0598, "step": 135220 }, { "epoch": 4.571824605294297, "grad_norm": 2.054497718811035, "learning_rate": 2.6036297351103768e-05, "loss": 0.058, "step": 135230 }, { "epoch": 4.572162682984549, "grad_norm": 0.8287267684936523, "learning_rate": 2.6029041713282653e-05, "loss": 0.0562, "step": 135240 }, { "epoch": 4.572500760674803, "grad_norm": 1.2093420028686523, "learning_rate": 2.602178673081853e-05, "loss": 0.0557, "step": 135250 }, { "epoch": 4.5728388383650564, "grad_norm": 1.1810723543167114, "learning_rate": 2.601453240390973e-05, "loss": 0.0597, "step": 135260 }, { "epoch": 4.57317691605531, "grad_norm": 1.4000262022018433, "learning_rate": 2.6007278732754624e-05, "loss": 0.0556, "step": 135270 }, { "epoch": 4.573514993745563, "grad_norm": 1.3302357196807861, "learning_rate": 2.600002571755149e-05, "loss": 0.0582, "step": 135280 }, { "epoch": 4.573853071435816, "grad_norm": 1.5931565761566162, "learning_rate": 2.599277335849865e-05, "loss": 0.0587, "step": 135290 }, { "epoch": 4.574191149126069, "grad_norm": 1.1669107675552368, "learning_rate": 2.5985521655794366e-05, "loss": 0.0614, "step": 135300 }, { "epoch": 4.574529226816322, "grad_norm": 2.376505136489868, "learning_rate": 2.59782706096369e-05, "loss": 0.0592, "step": 135310 }, { "epoch": 4.574867304506576, "grad_norm": 4.039593696594238, "learning_rate": 2.5971020220224474e-05, "loss": 0.0576, "step": 135320 }, { "epoch": 4.575205382196829, "grad_norm": 1.1458936929702759, "learning_rate": 2.5963770487755346e-05, "loss": 0.0614, "step": 135330 }, { "epoch": 4.575543459887082, "grad_norm": 3.3318216800689697, "learning_rate": 2.5956521412427703e-05, "loss": 0.0579, "step": 135340 }, { "epoch": 4.575881537577335, "grad_norm": 52.612335205078125, "learning_rate": 2.5949272994439712e-05, "loss": 0.0555, "step": 135350 }, { "epoch": 4.576219615267588, "grad_norm": 74.13028717041016, "learning_rate": 2.5942025233989587e-05, "loss": 0.0564, "step": 135360 }, { "epoch": 4.5765576929578415, "grad_norm": 3.8695790767669678, "learning_rate": 2.5934778131275427e-05, "loss": 0.061, "step": 135370 }, { "epoch": 4.576895770648095, "grad_norm": 1.1454102993011475, "learning_rate": 2.592753168649541e-05, "loss": 0.0533, "step": 135380 }, { "epoch": 4.577233848338349, "grad_norm": 1.2298839092254639, "learning_rate": 2.592028589984764e-05, "loss": 0.0584, "step": 135390 }, { "epoch": 4.577571926028601, "grad_norm": 7.875279903411865, "learning_rate": 2.5913040771530182e-05, "loss": 0.0581, "step": 135400 }, { "epoch": 4.577910003718855, "grad_norm": 6.510344505310059, "learning_rate": 2.590579630174116e-05, "loss": 0.0599, "step": 135410 }, { "epoch": 4.578248081409107, "grad_norm": 1.1562556028366089, "learning_rate": 2.589855249067862e-05, "loss": 0.0548, "step": 135420 }, { "epoch": 4.578586159099361, "grad_norm": 1.5022450685501099, "learning_rate": 2.58913093385406e-05, "loss": 0.0558, "step": 135430 }, { "epoch": 4.578924236789614, "grad_norm": 1.5712809562683105, "learning_rate": 2.5884066845525102e-05, "loss": 0.0543, "step": 135440 }, { "epoch": 4.579262314479868, "grad_norm": 1.0730289220809937, "learning_rate": 2.5876825011830158e-05, "loss": 0.0574, "step": 135450 }, { "epoch": 4.5796003921701205, "grad_norm": 2.2140252590179443, "learning_rate": 2.5869583837653793e-05, "loss": 0.0543, "step": 135460 }, { "epoch": 4.579938469860374, "grad_norm": 1.5106412172317505, "learning_rate": 2.5862343323193904e-05, "loss": 0.0729, "step": 135470 }, { "epoch": 4.5802765475506275, "grad_norm": 2.394566774368286, "learning_rate": 2.58551034686485e-05, "loss": 0.0577, "step": 135480 }, { "epoch": 4.58061462524088, "grad_norm": 2.2916452884674072, "learning_rate": 2.5847864274215473e-05, "loss": 0.0601, "step": 135490 }, { "epoch": 4.580952702931134, "grad_norm": 0.7555733919143677, "learning_rate": 2.5840625740092788e-05, "loss": 0.0599, "step": 135500 }, { "epoch": 4.581290780621387, "grad_norm": 13.142096519470215, "learning_rate": 2.5833387866478313e-05, "loss": 0.0547, "step": 135510 }, { "epoch": 4.58162885831164, "grad_norm": 0.8981982469558716, "learning_rate": 2.5826150653569924e-05, "loss": 0.0638, "step": 135520 }, { "epoch": 4.581966936001893, "grad_norm": 5.030513763427734, "learning_rate": 2.5818914101565504e-05, "loss": 0.0633, "step": 135530 }, { "epoch": 4.582305013692147, "grad_norm": 1.6471900939941406, "learning_rate": 2.5811678210662893e-05, "loss": 0.0609, "step": 135540 }, { "epoch": 4.5826430913823994, "grad_norm": 1.1845555305480957, "learning_rate": 2.5804442981059917e-05, "loss": 0.0642, "step": 135550 }, { "epoch": 4.582981169072653, "grad_norm": 1.6362073421478271, "learning_rate": 2.5797208412954355e-05, "loss": 0.0545, "step": 135560 }, { "epoch": 4.5833192467629065, "grad_norm": 1.3045696020126343, "learning_rate": 2.5789974506544047e-05, "loss": 0.0587, "step": 135570 }, { "epoch": 4.583657324453159, "grad_norm": 1.453399658203125, "learning_rate": 2.578274126202671e-05, "loss": 0.0545, "step": 135580 }, { "epoch": 4.583995402143413, "grad_norm": 5.692244529724121, "learning_rate": 2.5775508679600157e-05, "loss": 0.0547, "step": 135590 }, { "epoch": 4.584333479833666, "grad_norm": 1.3832523822784424, "learning_rate": 2.5768276759462097e-05, "loss": 0.0595, "step": 135600 }, { "epoch": 4.584671557523919, "grad_norm": 6.3861212730407715, "learning_rate": 2.5761045501810222e-05, "loss": 0.0569, "step": 135610 }, { "epoch": 4.585009635214172, "grad_norm": 2.77095627784729, "learning_rate": 2.5753814906842278e-05, "loss": 0.0634, "step": 135620 }, { "epoch": 4.585347712904426, "grad_norm": 3.177673101425171, "learning_rate": 2.574658497475593e-05, "loss": 0.0559, "step": 135630 }, { "epoch": 4.585685790594678, "grad_norm": 1.1115849018096924, "learning_rate": 2.5739355705748814e-05, "loss": 0.0588, "step": 135640 }, { "epoch": 4.586023868284932, "grad_norm": 1.1277527809143066, "learning_rate": 2.573212710001862e-05, "loss": 0.059, "step": 135650 }, { "epoch": 4.586361945975185, "grad_norm": 1.4992051124572754, "learning_rate": 2.572489915776293e-05, "loss": 0.0565, "step": 135660 }, { "epoch": 4.586700023665438, "grad_norm": 0.7758736610412598, "learning_rate": 2.5717671879179423e-05, "loss": 0.0597, "step": 135670 }, { "epoch": 4.587038101355692, "grad_norm": 1.1181867122650146, "learning_rate": 2.57104452644656e-05, "loss": 0.0613, "step": 135680 }, { "epoch": 4.587376179045945, "grad_norm": 1.6662636995315552, "learning_rate": 2.5703219313819104e-05, "loss": 0.0563, "step": 135690 }, { "epoch": 4.587714256736198, "grad_norm": 1.1735761165618896, "learning_rate": 2.5695994027437443e-05, "loss": 0.0606, "step": 135700 }, { "epoch": 4.588052334426451, "grad_norm": 1.8740240335464478, "learning_rate": 2.5688769405518188e-05, "loss": 0.0595, "step": 135710 }, { "epoch": 4.588390412116705, "grad_norm": 3.4824347496032715, "learning_rate": 2.568154544825885e-05, "loss": 0.0642, "step": 135720 }, { "epoch": 4.588728489806957, "grad_norm": 1.6245871782302856, "learning_rate": 2.56743221558569e-05, "loss": 0.0576, "step": 135730 }, { "epoch": 4.589066567497211, "grad_norm": 2.222670078277588, "learning_rate": 2.5667099528509864e-05, "loss": 0.0618, "step": 135740 }, { "epoch": 4.589404645187464, "grad_norm": 1.1652255058288574, "learning_rate": 2.5659877566415186e-05, "loss": 0.0563, "step": 135750 }, { "epoch": 4.589742722877717, "grad_norm": 1.737668752670288, "learning_rate": 2.5652656269770313e-05, "loss": 0.058, "step": 135760 }, { "epoch": 4.5900808005679705, "grad_norm": 1.6266053915023804, "learning_rate": 2.5645435638772647e-05, "loss": 0.0578, "step": 135770 }, { "epoch": 4.590418878258224, "grad_norm": 1.4846034049987793, "learning_rate": 2.5638215673619627e-05, "loss": 0.0608, "step": 135780 }, { "epoch": 4.590756955948477, "grad_norm": 1.584498643875122, "learning_rate": 2.5630996374508653e-05, "loss": 0.0563, "step": 135790 }, { "epoch": 4.59109503363873, "grad_norm": 0.929975152015686, "learning_rate": 2.5623777741637085e-05, "loss": 0.0599, "step": 135800 }, { "epoch": 4.591433111328984, "grad_norm": 3.1729464530944824, "learning_rate": 2.5616559775202276e-05, "loss": 0.0602, "step": 135810 }, { "epoch": 4.591771189019236, "grad_norm": 3.2440624237060547, "learning_rate": 2.560934247540154e-05, "loss": 0.0621, "step": 135820 }, { "epoch": 4.59210926670949, "grad_norm": 3.8506057262420654, "learning_rate": 2.5602125842432222e-05, "loss": 0.0646, "step": 135830 }, { "epoch": 4.592447344399743, "grad_norm": 7.4036478996276855, "learning_rate": 2.559490987649165e-05, "loss": 0.0514, "step": 135840 }, { "epoch": 4.592785422089996, "grad_norm": 2.5239806175231934, "learning_rate": 2.5587694577777043e-05, "loss": 0.0612, "step": 135850 }, { "epoch": 4.5931234997802495, "grad_norm": 1.1234873533248901, "learning_rate": 2.5580479946485707e-05, "loss": 0.0561, "step": 135860 }, { "epoch": 4.593461577470503, "grad_norm": 12.477019309997559, "learning_rate": 2.557326598281486e-05, "loss": 0.0599, "step": 135870 }, { "epoch": 4.593799655160756, "grad_norm": 1.3474671840667725, "learning_rate": 2.5566052686961783e-05, "loss": 0.0611, "step": 135880 }, { "epoch": 4.594137732851009, "grad_norm": 1.1630207300186157, "learning_rate": 2.55588400591236e-05, "loss": 0.058, "step": 135890 }, { "epoch": 4.594475810541263, "grad_norm": 3.8574533462524414, "learning_rate": 2.555162809949756e-05, "loss": 0.0558, "step": 135900 }, { "epoch": 4.594813888231515, "grad_norm": 1.155577540397644, "learning_rate": 2.5544416808280837e-05, "loss": 0.0654, "step": 135910 }, { "epoch": 4.595151965921769, "grad_norm": 1.7565010786056519, "learning_rate": 2.553720618567057e-05, "loss": 0.0606, "step": 135920 }, { "epoch": 4.595490043612022, "grad_norm": 0.9082628488540649, "learning_rate": 2.5529996231863896e-05, "loss": 0.0652, "step": 135930 }, { "epoch": 4.595828121302275, "grad_norm": 1.2161593437194824, "learning_rate": 2.552278694705791e-05, "loss": 0.0623, "step": 135940 }, { "epoch": 4.596166198992528, "grad_norm": 1.537314534187317, "learning_rate": 2.551557833144976e-05, "loss": 0.0588, "step": 135950 }, { "epoch": 4.596504276682782, "grad_norm": 1.7375277280807495, "learning_rate": 2.5508370385236476e-05, "loss": 0.0573, "step": 135960 }, { "epoch": 4.596842354373035, "grad_norm": 1.1495509147644043, "learning_rate": 2.550116310861516e-05, "loss": 0.0564, "step": 135970 }, { "epoch": 4.597180432063288, "grad_norm": 2.9863438606262207, "learning_rate": 2.5493956501782846e-05, "loss": 0.0563, "step": 135980 }, { "epoch": 4.597518509753542, "grad_norm": 3.357692241668701, "learning_rate": 2.5486750564936533e-05, "loss": 0.0546, "step": 135990 }, { "epoch": 4.597856587443794, "grad_norm": 2.480271100997925, "learning_rate": 2.5479545298273273e-05, "loss": 0.0613, "step": 136000 }, { "epoch": 4.598194665134048, "grad_norm": 1.9066122770309448, "learning_rate": 2.5472340701990027e-05, "loss": 0.0634, "step": 136010 }, { "epoch": 4.598532742824301, "grad_norm": 1.7670384645462036, "learning_rate": 2.546513677628376e-05, "loss": 0.0551, "step": 136020 }, { "epoch": 4.598870820514554, "grad_norm": 77.12948608398438, "learning_rate": 2.545793352135145e-05, "loss": 0.0616, "step": 136030 }, { "epoch": 4.599208898204807, "grad_norm": 2.0494868755340576, "learning_rate": 2.545073093739e-05, "loss": 0.0587, "step": 136040 }, { "epoch": 4.599546975895061, "grad_norm": 1.1189182996749878, "learning_rate": 2.544352902459638e-05, "loss": 0.0559, "step": 136050 }, { "epoch": 4.5998850535853135, "grad_norm": 1.5043870210647583, "learning_rate": 2.543632778316741e-05, "loss": 0.0611, "step": 136060 }, { "epoch": 4.600223131275567, "grad_norm": 1.8386647701263428, "learning_rate": 2.542912721330004e-05, "loss": 0.0627, "step": 136070 }, { "epoch": 4.6005612089658205, "grad_norm": 0.7481908202171326, "learning_rate": 2.5421927315191075e-05, "loss": 0.0598, "step": 136080 }, { "epoch": 4.600899286656073, "grad_norm": 5.357445240020752, "learning_rate": 2.54147280890374e-05, "loss": 0.0564, "step": 136090 }, { "epoch": 4.601237364346327, "grad_norm": 0.6773801445960999, "learning_rate": 2.5407529535035828e-05, "loss": 0.0566, "step": 136100 }, { "epoch": 4.60157544203658, "grad_norm": 1.7354323863983154, "learning_rate": 2.5400331653383136e-05, "loss": 0.0703, "step": 136110 }, { "epoch": 4.601913519726834, "grad_norm": 2.0411365032196045, "learning_rate": 2.5393134444276156e-05, "loss": 0.0622, "step": 136120 }, { "epoch": 4.602251597417086, "grad_norm": 5.596672058105469, "learning_rate": 2.538593790791164e-05, "loss": 0.064, "step": 136130 }, { "epoch": 4.60258967510734, "grad_norm": 3.2472665309906006, "learning_rate": 2.5378742044486333e-05, "loss": 0.0595, "step": 136140 }, { "epoch": 4.6029277527975925, "grad_norm": 1.8483372926712036, "learning_rate": 2.5371546854196947e-05, "loss": 0.0634, "step": 136150 }, { "epoch": 4.603265830487846, "grad_norm": 0.9121569395065308, "learning_rate": 2.5364352337240214e-05, "loss": 0.0525, "step": 136160 }, { "epoch": 4.6036039081780995, "grad_norm": 2.4491145610809326, "learning_rate": 2.535715849381286e-05, "loss": 0.0547, "step": 136170 }, { "epoch": 4.603941985868353, "grad_norm": 1.477278232574463, "learning_rate": 2.534996532411153e-05, "loss": 0.057, "step": 136180 }, { "epoch": 4.604280063558606, "grad_norm": 3.579596757888794, "learning_rate": 2.534277282833289e-05, "loss": 0.0584, "step": 136190 }, { "epoch": 4.604618141248859, "grad_norm": 2.22847580909729, "learning_rate": 2.533558100667356e-05, "loss": 0.0641, "step": 136200 }, { "epoch": 4.604956218939112, "grad_norm": 1.2420629262924194, "learning_rate": 2.532838985933019e-05, "loss": 0.0561, "step": 136210 }, { "epoch": 4.605294296629365, "grad_norm": 0.7174719572067261, "learning_rate": 2.5321199386499377e-05, "loss": 0.0569, "step": 136220 }, { "epoch": 4.605632374319619, "grad_norm": 1.4445133209228516, "learning_rate": 2.5314009588377684e-05, "loss": 0.0577, "step": 136230 }, { "epoch": 4.605970452009872, "grad_norm": 1.4122687578201294, "learning_rate": 2.5306820465161714e-05, "loss": 0.0614, "step": 136240 }, { "epoch": 4.606308529700125, "grad_norm": 1.2518854141235352, "learning_rate": 2.5299632017047976e-05, "loss": 0.0649, "step": 136250 }, { "epoch": 4.6066466073903785, "grad_norm": 1.0168859958648682, "learning_rate": 2.529244424423306e-05, "loss": 0.0605, "step": 136260 }, { "epoch": 4.606984685080631, "grad_norm": 1.770700216293335, "learning_rate": 2.5285257146913388e-05, "loss": 0.0507, "step": 136270 }, { "epoch": 4.607322762770885, "grad_norm": 1.8042289018630981, "learning_rate": 2.527807072528551e-05, "loss": 0.0605, "step": 136280 }, { "epoch": 4.607660840461138, "grad_norm": 1.0971137285232544, "learning_rate": 2.5270884979545907e-05, "loss": 0.0597, "step": 136290 }, { "epoch": 4.607998918151392, "grad_norm": 3.78950572013855, "learning_rate": 2.526369990989102e-05, "loss": 0.0636, "step": 136300 }, { "epoch": 4.608336995841644, "grad_norm": 1.3965425491333008, "learning_rate": 2.525651551651729e-05, "loss": 0.0511, "step": 136310 }, { "epoch": 4.608675073531898, "grad_norm": 5.261653900146484, "learning_rate": 2.5249331799621102e-05, "loss": 0.0526, "step": 136320 }, { "epoch": 4.60901315122215, "grad_norm": 0.9051552414894104, "learning_rate": 2.52421487593989e-05, "loss": 0.0525, "step": 136330 }, { "epoch": 4.609351228912404, "grad_norm": 2.4925262928009033, "learning_rate": 2.5234966396047055e-05, "loss": 0.0564, "step": 136340 }, { "epoch": 4.609689306602657, "grad_norm": 1.2261348962783813, "learning_rate": 2.522778470976191e-05, "loss": 0.0606, "step": 136350 }, { "epoch": 4.610027384292911, "grad_norm": 0.8000897169113159, "learning_rate": 2.5220603700739842e-05, "loss": 0.0563, "step": 136360 }, { "epoch": 4.6103654619831635, "grad_norm": 1.2456493377685547, "learning_rate": 2.5213423369177136e-05, "loss": 0.0609, "step": 136370 }, { "epoch": 4.610703539673417, "grad_norm": 1.245578408241272, "learning_rate": 2.5206243715270146e-05, "loss": 0.0555, "step": 136380 }, { "epoch": 4.61104161736367, "grad_norm": 1.1715117692947388, "learning_rate": 2.5199064739215132e-05, "loss": 0.0554, "step": 136390 }, { "epoch": 4.611379695053923, "grad_norm": 2.1017611026763916, "learning_rate": 2.5191886441208356e-05, "loss": 0.0593, "step": 136400 }, { "epoch": 4.611717772744177, "grad_norm": 0.9869874715805054, "learning_rate": 2.5184708821446106e-05, "loss": 0.0639, "step": 136410 }, { "epoch": 4.61205585043443, "grad_norm": 1.6437034606933594, "learning_rate": 2.5177531880124596e-05, "loss": 0.0572, "step": 136420 }, { "epoch": 4.612393928124683, "grad_norm": 1.1320500373840332, "learning_rate": 2.5170355617440034e-05, "loss": 0.0565, "step": 136430 }, { "epoch": 4.612732005814936, "grad_norm": 22.38646125793457, "learning_rate": 2.5163180033588606e-05, "loss": 0.056, "step": 136440 }, { "epoch": 4.61307008350519, "grad_norm": 3.1223034858703613, "learning_rate": 2.5156005128766526e-05, "loss": 0.0606, "step": 136450 }, { "epoch": 4.6134081611954425, "grad_norm": 2.0787482261657715, "learning_rate": 2.5148830903169908e-05, "loss": 0.0577, "step": 136460 }, { "epoch": 4.613746238885696, "grad_norm": 0.8509978652000427, "learning_rate": 2.514165735699493e-05, "loss": 0.0589, "step": 136470 }, { "epoch": 4.6140843165759495, "grad_norm": 1.8148469924926758, "learning_rate": 2.5134484490437715e-05, "loss": 0.0549, "step": 136480 }, { "epoch": 4.614422394266202, "grad_norm": 1.1675443649291992, "learning_rate": 2.5127312303694324e-05, "loss": 0.0632, "step": 136490 }, { "epoch": 4.614760471956456, "grad_norm": 1.4126025438308716, "learning_rate": 2.512014079696089e-05, "loss": 0.0536, "step": 136500 }, { "epoch": 4.615098549646709, "grad_norm": 7.763212203979492, "learning_rate": 2.5112969970433464e-05, "loss": 0.0561, "step": 136510 }, { "epoch": 4.615436627336962, "grad_norm": 1.5667600631713867, "learning_rate": 2.5105799824308083e-05, "loss": 0.0556, "step": 136520 }, { "epoch": 4.615774705027215, "grad_norm": 0.8523601293563843, "learning_rate": 2.5098630358780774e-05, "loss": 0.0595, "step": 136530 }, { "epoch": 4.616112782717469, "grad_norm": 2.3183021545410156, "learning_rate": 2.509146157404755e-05, "loss": 0.0576, "step": 136540 }, { "epoch": 4.6164508604077215, "grad_norm": 1.0136996507644653, "learning_rate": 2.5084293470304443e-05, "loss": 0.0555, "step": 136550 }, { "epoch": 4.616788938097975, "grad_norm": 4.706289768218994, "learning_rate": 2.5077126047747358e-05, "loss": 0.0597, "step": 136560 }, { "epoch": 4.6171270157882285, "grad_norm": 1.798675775527954, "learning_rate": 2.5069959306572306e-05, "loss": 0.0616, "step": 136570 }, { "epoch": 4.617465093478481, "grad_norm": 0.6789514422416687, "learning_rate": 2.506279324697518e-05, "loss": 0.0556, "step": 136580 }, { "epoch": 4.617803171168735, "grad_norm": 0.7975255846977234, "learning_rate": 2.5055627869151934e-05, "loss": 0.0613, "step": 136590 }, { "epoch": 4.618141248858988, "grad_norm": 1.8799172639846802, "learning_rate": 2.5048463173298453e-05, "loss": 0.0593, "step": 136600 }, { "epoch": 4.618479326549241, "grad_norm": 1.0302356481552124, "learning_rate": 2.5041299159610593e-05, "loss": 0.0556, "step": 136610 }, { "epoch": 4.618817404239494, "grad_norm": 1.079514980316162, "learning_rate": 2.5034135828284256e-05, "loss": 0.06, "step": 136620 }, { "epoch": 4.619155481929748, "grad_norm": 1.1210212707519531, "learning_rate": 2.5026973179515273e-05, "loss": 0.0582, "step": 136630 }, { "epoch": 4.61949355962, "grad_norm": 1.2898852825164795, "learning_rate": 2.5019811213499456e-05, "loss": 0.0621, "step": 136640 }, { "epoch": 4.619831637310254, "grad_norm": 1.2295259237289429, "learning_rate": 2.5012649930432596e-05, "loss": 0.0614, "step": 136650 }, { "epoch": 4.620169715000507, "grad_norm": 1.279533863067627, "learning_rate": 2.50054893305105e-05, "loss": 0.0593, "step": 136660 }, { "epoch": 4.62050779269076, "grad_norm": 0.7390416264533997, "learning_rate": 2.4998329413928962e-05, "loss": 0.0551, "step": 136670 }, { "epoch": 4.620845870381014, "grad_norm": 2.2558398246765137, "learning_rate": 2.49911701808837e-05, "loss": 0.0614, "step": 136680 }, { "epoch": 4.621183948071267, "grad_norm": 2.744084596633911, "learning_rate": 2.498401163157045e-05, "loss": 0.0604, "step": 136690 }, { "epoch": 4.62152202576152, "grad_norm": 0.671920657157898, "learning_rate": 2.4976853766184905e-05, "loss": 0.0566, "step": 136700 }, { "epoch": 4.621860103451773, "grad_norm": 0.8006687164306641, "learning_rate": 2.496969658492279e-05, "loss": 0.0609, "step": 136710 }, { "epoch": 4.622198181142027, "grad_norm": 0.9795480966567993, "learning_rate": 2.4962540087979774e-05, "loss": 0.0603, "step": 136720 }, { "epoch": 4.622536258832279, "grad_norm": 1.03180730342865, "learning_rate": 2.4955384275551484e-05, "loss": 0.0601, "step": 136730 }, { "epoch": 4.622874336522533, "grad_norm": 0.9797732830047607, "learning_rate": 2.49482291478336e-05, "loss": 0.0569, "step": 136740 }, { "epoch": 4.623212414212786, "grad_norm": 1.0683186054229736, "learning_rate": 2.4941074705021707e-05, "loss": 0.0596, "step": 136750 }, { "epoch": 4.623550491903039, "grad_norm": 1.0695667266845703, "learning_rate": 2.4933920947311447e-05, "loss": 0.0576, "step": 136760 }, { "epoch": 4.6238885695932925, "grad_norm": 1.1918977499008179, "learning_rate": 2.4926767874898332e-05, "loss": 0.0509, "step": 136770 }, { "epoch": 4.624226647283546, "grad_norm": 1.0062514543533325, "learning_rate": 2.4919615487977987e-05, "loss": 0.0576, "step": 136780 }, { "epoch": 4.624564724973799, "grad_norm": 0.7397075891494751, "learning_rate": 2.491246378674591e-05, "loss": 0.0551, "step": 136790 }, { "epoch": 4.624902802664052, "grad_norm": 1.3083127737045288, "learning_rate": 2.4905312771397672e-05, "loss": 0.0593, "step": 136800 }, { "epoch": 4.625240880354306, "grad_norm": 3.3149049282073975, "learning_rate": 2.4898162442128753e-05, "loss": 0.0581, "step": 136810 }, { "epoch": 4.625578958044558, "grad_norm": 1.5436475276947021, "learning_rate": 2.4891012799134622e-05, "loss": 0.0605, "step": 136820 }, { "epoch": 4.625917035734812, "grad_norm": 3.2792458534240723, "learning_rate": 2.4883863842610788e-05, "loss": 0.0567, "step": 136830 }, { "epoch": 4.626255113425065, "grad_norm": 1.3204935789108276, "learning_rate": 2.487671557275268e-05, "loss": 0.0521, "step": 136840 }, { "epoch": 4.626593191115318, "grad_norm": 1.6763834953308105, "learning_rate": 2.486956798975571e-05, "loss": 0.0558, "step": 136850 }, { "epoch": 4.6269312688055715, "grad_norm": 1.2038273811340332, "learning_rate": 2.486242109381533e-05, "loss": 0.0556, "step": 136860 }, { "epoch": 4.627269346495825, "grad_norm": 1.9411481618881226, "learning_rate": 2.4855274885126893e-05, "loss": 0.0531, "step": 136870 }, { "epoch": 4.627607424186078, "grad_norm": 2.4968013763427734, "learning_rate": 2.4848129363885818e-05, "loss": 0.063, "step": 136880 }, { "epoch": 4.627945501876331, "grad_norm": 0.8290150761604309, "learning_rate": 2.4840984530287438e-05, "loss": 0.0544, "step": 136890 }, { "epoch": 4.628283579566585, "grad_norm": 1.2837263345718384, "learning_rate": 2.4833840384527085e-05, "loss": 0.0627, "step": 136900 }, { "epoch": 4.628621657256837, "grad_norm": 1.2463164329528809, "learning_rate": 2.482669692680006e-05, "loss": 0.0622, "step": 136910 }, { "epoch": 4.628959734947091, "grad_norm": 1.4779973030090332, "learning_rate": 2.4819554157301705e-05, "loss": 0.0538, "step": 136920 }, { "epoch": 4.629297812637344, "grad_norm": 1.1425176858901978, "learning_rate": 2.4812412076227283e-05, "loss": 0.0561, "step": 136930 }, { "epoch": 4.629635890327597, "grad_norm": 1.9700731039047241, "learning_rate": 2.480527068377203e-05, "loss": 0.0596, "step": 136940 }, { "epoch": 4.62997396801785, "grad_norm": 1.4131242036819458, "learning_rate": 2.479812998013123e-05, "loss": 0.0579, "step": 136950 }, { "epoch": 4.630312045708104, "grad_norm": 1.1443902254104614, "learning_rate": 2.4790989965500073e-05, "loss": 0.0591, "step": 136960 }, { "epoch": 4.630650123398357, "grad_norm": 0.7123228907585144, "learning_rate": 2.4783850640073814e-05, "loss": 0.0626, "step": 136970 }, { "epoch": 4.63098820108861, "grad_norm": 0.9521005749702454, "learning_rate": 2.477671200404757e-05, "loss": 0.0529, "step": 136980 }, { "epoch": 4.631326278778864, "grad_norm": 1.2756390571594238, "learning_rate": 2.476957405761654e-05, "loss": 0.0652, "step": 136990 }, { "epoch": 4.631664356469116, "grad_norm": 1.6463512182235718, "learning_rate": 2.4762436800975897e-05, "loss": 0.0567, "step": 137000 }, { "epoch": 4.63200243415937, "grad_norm": 1.8255163431167603, "learning_rate": 2.4755300234320748e-05, "loss": 0.0601, "step": 137010 }, { "epoch": 4.632340511849623, "grad_norm": 2.8936030864715576, "learning_rate": 2.4748164357846215e-05, "loss": 0.0601, "step": 137020 }, { "epoch": 4.632678589539876, "grad_norm": 1.4901673793792725, "learning_rate": 2.4741029171747353e-05, "loss": 0.0501, "step": 137030 }, { "epoch": 4.633016667230129, "grad_norm": 0.9466025233268738, "learning_rate": 2.473389467621927e-05, "loss": 0.0638, "step": 137040 }, { "epoch": 4.633354744920383, "grad_norm": 1.8250256776809692, "learning_rate": 2.4726760871457057e-05, "loss": 0.0587, "step": 137050 }, { "epoch": 4.6336928226106355, "grad_norm": 1.565938949584961, "learning_rate": 2.471962775765566e-05, "loss": 0.0566, "step": 137060 }, { "epoch": 4.634030900300889, "grad_norm": 1.1350041627883911, "learning_rate": 2.4712495335010168e-05, "loss": 0.0559, "step": 137070 }, { "epoch": 4.634368977991143, "grad_norm": 5.025607109069824, "learning_rate": 2.470536360371553e-05, "loss": 0.0582, "step": 137080 }, { "epoch": 4.634707055681396, "grad_norm": 1.7430375814437866, "learning_rate": 2.4698232563966765e-05, "loss": 0.0581, "step": 137090 }, { "epoch": 4.635045133371649, "grad_norm": 8.364096641540527, "learning_rate": 2.4691102215958823e-05, "loss": 0.0585, "step": 137100 }, { "epoch": 4.635383211061902, "grad_norm": 1.1295305490493774, "learning_rate": 2.4683972559886616e-05, "loss": 0.064, "step": 137110 }, { "epoch": 4.635721288752155, "grad_norm": 1.0163179636001587, "learning_rate": 2.467684359594511e-05, "loss": 0.0561, "step": 137120 }, { "epoch": 4.636059366442408, "grad_norm": 0.8588621616363525, "learning_rate": 2.4669715324329186e-05, "loss": 0.0593, "step": 137130 }, { "epoch": 4.636397444132662, "grad_norm": 5.695322036743164, "learning_rate": 2.4662587745233723e-05, "loss": 0.0585, "step": 137140 }, { "epoch": 4.636735521822915, "grad_norm": 2.617252826690674, "learning_rate": 2.465546085885358e-05, "loss": 0.0533, "step": 137150 }, { "epoch": 4.637073599513168, "grad_norm": 0.8900040984153748, "learning_rate": 2.4648334665383634e-05, "loss": 0.0666, "step": 137160 }, { "epoch": 4.6374116772034215, "grad_norm": 0.8908359408378601, "learning_rate": 2.464120916501867e-05, "loss": 0.0569, "step": 137170 }, { "epoch": 4.637749754893674, "grad_norm": 1.0601487159729004, "learning_rate": 2.4634084357953547e-05, "loss": 0.057, "step": 137180 }, { "epoch": 4.638087832583928, "grad_norm": 1.0511122941970825, "learning_rate": 2.4626960244383022e-05, "loss": 0.058, "step": 137190 }, { "epoch": 4.638425910274181, "grad_norm": 3.6615006923675537, "learning_rate": 2.4619836824501853e-05, "loss": 0.0604, "step": 137200 }, { "epoch": 4.638763987964435, "grad_norm": 2.0258336067199707, "learning_rate": 2.461271409850483e-05, "loss": 0.0594, "step": 137210 }, { "epoch": 4.639102065654687, "grad_norm": 0.8542396426200867, "learning_rate": 2.4605592066586668e-05, "loss": 0.058, "step": 137220 }, { "epoch": 4.639440143344941, "grad_norm": 1.1067384481430054, "learning_rate": 2.459847072894206e-05, "loss": 0.0561, "step": 137230 }, { "epoch": 4.639778221035193, "grad_norm": 1.4162336587905884, "learning_rate": 2.4591350085765735e-05, "loss": 0.0574, "step": 137240 }, { "epoch": 4.640116298725447, "grad_norm": 1.4085711240768433, "learning_rate": 2.458423013725234e-05, "loss": 0.0562, "step": 137250 }, { "epoch": 4.6404543764157005, "grad_norm": 0.7399296164512634, "learning_rate": 2.4577110883596577e-05, "loss": 0.0562, "step": 137260 }, { "epoch": 4.640792454105954, "grad_norm": 0.7550387382507324, "learning_rate": 2.4569992324993015e-05, "loss": 0.0567, "step": 137270 }, { "epoch": 4.641130531796207, "grad_norm": 0.5098269581794739, "learning_rate": 2.456287446163633e-05, "loss": 0.0563, "step": 137280 }, { "epoch": 4.64146860948646, "grad_norm": 0.7699083685874939, "learning_rate": 2.4555757293721084e-05, "loss": 0.0591, "step": 137290 }, { "epoch": 4.641806687176713, "grad_norm": 1.3371570110321045, "learning_rate": 2.4548640821441887e-05, "loss": 0.0585, "step": 137300 }, { "epoch": 4.642144764866966, "grad_norm": 2.5617547035217285, "learning_rate": 2.454152504499329e-05, "loss": 0.0674, "step": 137310 }, { "epoch": 4.64248284255722, "grad_norm": 1.327582836151123, "learning_rate": 2.4534409964569815e-05, "loss": 0.0578, "step": 137320 }, { "epoch": 4.642820920247473, "grad_norm": 0.9394257664680481, "learning_rate": 2.4527295580366028e-05, "loss": 0.0607, "step": 137330 }, { "epoch": 4.643158997937726, "grad_norm": 1.1089006662368774, "learning_rate": 2.4520181892576405e-05, "loss": 0.0543, "step": 137340 }, { "epoch": 4.643497075627979, "grad_norm": 1.4274702072143555, "learning_rate": 2.451306890139544e-05, "loss": 0.0571, "step": 137350 }, { "epoch": 4.643835153318232, "grad_norm": 0.8010379076004028, "learning_rate": 2.4505956607017578e-05, "loss": 0.0559, "step": 137360 }, { "epoch": 4.644173231008486, "grad_norm": 0.8310776948928833, "learning_rate": 2.4498845009637283e-05, "loss": 0.0565, "step": 137370 }, { "epoch": 4.644511308698739, "grad_norm": 1.8486058712005615, "learning_rate": 2.4491734109449004e-05, "loss": 0.0601, "step": 137380 }, { "epoch": 4.644849386388993, "grad_norm": 2.538238763809204, "learning_rate": 2.4484623906647132e-05, "loss": 0.06, "step": 137390 }, { "epoch": 4.645187464079245, "grad_norm": 2.024138927459717, "learning_rate": 2.4477514401426053e-05, "loss": 0.0612, "step": 137400 }, { "epoch": 4.645525541769499, "grad_norm": 0.758311927318573, "learning_rate": 2.447040559398013e-05, "loss": 0.0615, "step": 137410 }, { "epoch": 4.645863619459752, "grad_norm": 1.3505421876907349, "learning_rate": 2.446329748450374e-05, "loss": 0.0565, "step": 137420 }, { "epoch": 4.646201697150005, "grad_norm": 0.8515104651451111, "learning_rate": 2.4456190073191203e-05, "loss": 0.0583, "step": 137430 }, { "epoch": 4.646539774840258, "grad_norm": 1.2885550260543823, "learning_rate": 2.4449083360236817e-05, "loss": 0.0529, "step": 137440 }, { "epoch": 4.646877852530512, "grad_norm": 2.654581069946289, "learning_rate": 2.4441977345834914e-05, "loss": 0.0589, "step": 137450 }, { "epoch": 4.6472159302207645, "grad_norm": 1.8147015571594238, "learning_rate": 2.443487203017973e-05, "loss": 0.0603, "step": 137460 }, { "epoch": 4.647554007911018, "grad_norm": 1.3492467403411865, "learning_rate": 2.442776741346557e-05, "loss": 0.057, "step": 137470 }, { "epoch": 4.6478920856012715, "grad_norm": 1.3274569511413574, "learning_rate": 2.442066349588661e-05, "loss": 0.0567, "step": 137480 }, { "epoch": 4.648230163291524, "grad_norm": 1.7618087530136108, "learning_rate": 2.44135602776371e-05, "loss": 0.0537, "step": 137490 }, { "epoch": 4.648568240981778, "grad_norm": 1.1184457540512085, "learning_rate": 2.4406457758911254e-05, "loss": 0.0521, "step": 137500 }, { "epoch": 4.648906318672031, "grad_norm": 1.4944453239440918, "learning_rate": 2.4399355939903245e-05, "loss": 0.056, "step": 137510 }, { "epoch": 4.649244396362284, "grad_norm": 11.11139965057373, "learning_rate": 2.4392254820807226e-05, "loss": 0.0547, "step": 137520 }, { "epoch": 4.649582474052537, "grad_norm": 0.6654959917068481, "learning_rate": 2.4385154401817318e-05, "loss": 0.0602, "step": 137530 }, { "epoch": 4.649920551742791, "grad_norm": 0.9402692317962646, "learning_rate": 2.437805468312768e-05, "loss": 0.0525, "step": 137540 }, { "epoch": 4.6502586294330435, "grad_norm": 4.075307846069336, "learning_rate": 2.437095566493241e-05, "loss": 0.0577, "step": 137550 }, { "epoch": 4.650596707123297, "grad_norm": 1.5133421421051025, "learning_rate": 2.436385734742556e-05, "loss": 0.0633, "step": 137560 }, { "epoch": 4.6509347848135505, "grad_norm": 9.443784713745117, "learning_rate": 2.4356759730801245e-05, "loss": 0.0589, "step": 137570 }, { "epoch": 4.651272862503803, "grad_norm": 1.975181221961975, "learning_rate": 2.4349662815253455e-05, "loss": 0.0585, "step": 137580 }, { "epoch": 4.651610940194057, "grad_norm": 0.9218162298202515, "learning_rate": 2.4342566600976273e-05, "loss": 0.063, "step": 137590 }, { "epoch": 4.65194901788431, "grad_norm": 1.3637157678604126, "learning_rate": 2.433547108816368e-05, "loss": 0.0577, "step": 137600 }, { "epoch": 4.652287095574563, "grad_norm": 6.352208137512207, "learning_rate": 2.4328376277009647e-05, "loss": 0.0539, "step": 137610 }, { "epoch": 4.652625173264816, "grad_norm": 13.951313018798828, "learning_rate": 2.432128216770818e-05, "loss": 0.0631, "step": 137620 }, { "epoch": 4.65296325095507, "grad_norm": 0.5879385471343994, "learning_rate": 2.431418876045321e-05, "loss": 0.0533, "step": 137630 }, { "epoch": 4.653301328645322, "grad_norm": 2.060046911239624, "learning_rate": 2.430709605543867e-05, "loss": 0.059, "step": 137640 }, { "epoch": 4.653639406335576, "grad_norm": 1.2471435070037842, "learning_rate": 2.430000405285846e-05, "loss": 0.0526, "step": 137650 }, { "epoch": 4.6539774840258294, "grad_norm": 1.8223330974578857, "learning_rate": 2.4292912752906494e-05, "loss": 0.0589, "step": 137660 }, { "epoch": 4.654315561716082, "grad_norm": 1.8450353145599365, "learning_rate": 2.428582215577662e-05, "loss": 0.0678, "step": 137670 }, { "epoch": 4.654653639406336, "grad_norm": 2.281888723373413, "learning_rate": 2.4278732261662727e-05, "loss": 0.0588, "step": 137680 }, { "epoch": 4.654991717096589, "grad_norm": 0.9627587199211121, "learning_rate": 2.4271643070758633e-05, "loss": 0.0545, "step": 137690 }, { "epoch": 4.655329794786842, "grad_norm": 0.9619908928871155, "learning_rate": 2.426455458325813e-05, "loss": 0.0587, "step": 137700 }, { "epoch": 4.655667872477095, "grad_norm": 1.4865199327468872, "learning_rate": 2.4257466799355055e-05, "loss": 0.0553, "step": 137710 }, { "epoch": 4.656005950167349, "grad_norm": 2.83803653717041, "learning_rate": 2.4250379719243167e-05, "loss": 0.0597, "step": 137720 }, { "epoch": 4.656344027857601, "grad_norm": 0.6727590560913086, "learning_rate": 2.4243293343116224e-05, "loss": 0.0571, "step": 137730 }, { "epoch": 4.656682105547855, "grad_norm": 5.902261734008789, "learning_rate": 2.4236207671167948e-05, "loss": 0.0591, "step": 137740 }, { "epoch": 4.657020183238108, "grad_norm": 1.772660255432129, "learning_rate": 2.4229122703592073e-05, "loss": 0.0551, "step": 137750 }, { "epoch": 4.657358260928361, "grad_norm": 9.422469139099121, "learning_rate": 2.422203844058234e-05, "loss": 0.06, "step": 137760 }, { "epoch": 4.6576963386186145, "grad_norm": 0.5986063480377197, "learning_rate": 2.4214954882332353e-05, "loss": 0.0582, "step": 137770 }, { "epoch": 4.658034416308868, "grad_norm": 1.1674896478652954, "learning_rate": 2.4207872029035835e-05, "loss": 0.0601, "step": 137780 }, { "epoch": 4.658372493999121, "grad_norm": 2.3669958114624023, "learning_rate": 2.420078988088638e-05, "loss": 0.0564, "step": 137790 }, { "epoch": 4.658710571689374, "grad_norm": 1.5541839599609375, "learning_rate": 2.419370843807766e-05, "loss": 0.0519, "step": 137800 }, { "epoch": 4.659048649379628, "grad_norm": 1.1092798709869385, "learning_rate": 2.418662770080326e-05, "loss": 0.0553, "step": 137810 }, { "epoch": 4.65938672706988, "grad_norm": 1.3723962306976318, "learning_rate": 2.417954766925674e-05, "loss": 0.0589, "step": 137820 }, { "epoch": 4.659724804760134, "grad_norm": 2.4406678676605225, "learning_rate": 2.4172468343631704e-05, "loss": 0.0622, "step": 137830 }, { "epoch": 4.660062882450387, "grad_norm": 1.0202971696853638, "learning_rate": 2.4165389724121677e-05, "loss": 0.0647, "step": 137840 }, { "epoch": 4.66040096014064, "grad_norm": 0.8928374648094177, "learning_rate": 2.4158311810920198e-05, "loss": 0.0555, "step": 137850 }, { "epoch": 4.6607390378308935, "grad_norm": 0.7359710931777954, "learning_rate": 2.415123460422074e-05, "loss": 0.0598, "step": 137860 }, { "epoch": 4.661077115521147, "grad_norm": 1.843060851097107, "learning_rate": 2.414415810421682e-05, "loss": 0.0555, "step": 137870 }, { "epoch": 4.6614151932114, "grad_norm": 0.9441421031951904, "learning_rate": 2.413708231110192e-05, "loss": 0.0522, "step": 137880 }, { "epoch": 4.661753270901653, "grad_norm": 3.1500399112701416, "learning_rate": 2.413000722506948e-05, "loss": 0.0619, "step": 137890 }, { "epoch": 4.662091348591907, "grad_norm": 3.0390713214874268, "learning_rate": 2.412293284631292e-05, "loss": 0.0578, "step": 137900 }, { "epoch": 4.662429426282159, "grad_norm": 1.5869220495224, "learning_rate": 2.4115859175025628e-05, "loss": 0.0585, "step": 137910 }, { "epoch": 4.662767503972413, "grad_norm": 1.3210036754608154, "learning_rate": 2.4108786211401047e-05, "loss": 0.0579, "step": 137920 }, { "epoch": 4.663105581662666, "grad_norm": 1.3469382524490356, "learning_rate": 2.4101713955632528e-05, "loss": 0.0547, "step": 137930 }, { "epoch": 4.663443659352919, "grad_norm": 0.9754753112792969, "learning_rate": 2.4094642407913398e-05, "loss": 0.06, "step": 137940 }, { "epoch": 4.6637817370431724, "grad_norm": 1.4080173969268799, "learning_rate": 2.4087571568437028e-05, "loss": 0.0626, "step": 137950 }, { "epoch": 4.664119814733426, "grad_norm": 2.143479824066162, "learning_rate": 2.40805014373967e-05, "loss": 0.0603, "step": 137960 }, { "epoch": 4.664457892423679, "grad_norm": 1.4493799209594727, "learning_rate": 2.4073432014985764e-05, "loss": 0.0563, "step": 137970 }, { "epoch": 4.664795970113932, "grad_norm": 0.9986018538475037, "learning_rate": 2.4066363301397415e-05, "loss": 0.0596, "step": 137980 }, { "epoch": 4.665134047804186, "grad_norm": 0.7834295630455017, "learning_rate": 2.4059295296824958e-05, "loss": 0.0587, "step": 137990 }, { "epoch": 4.665472125494438, "grad_norm": 1.3902173042297363, "learning_rate": 2.4052228001461636e-05, "loss": 0.0559, "step": 138000 }, { "epoch": 4.665810203184692, "grad_norm": 1.3711268901824951, "learning_rate": 2.404516141550065e-05, "loss": 0.0573, "step": 138010 }, { "epoch": 4.666148280874945, "grad_norm": 0.5002287030220032, "learning_rate": 2.4038095539135203e-05, "loss": 0.0586, "step": 138020 }, { "epoch": 4.666486358565198, "grad_norm": 1.5135267972946167, "learning_rate": 2.403103037255845e-05, "loss": 0.0565, "step": 138030 }, { "epoch": 4.666824436255451, "grad_norm": 0.7177193760871887, "learning_rate": 2.402396591596359e-05, "loss": 0.0579, "step": 138040 }, { "epoch": 4.667162513945705, "grad_norm": 0.8529400825500488, "learning_rate": 2.4016902169543747e-05, "loss": 0.0537, "step": 138050 }, { "epoch": 4.667500591635958, "grad_norm": 1.8523188829421997, "learning_rate": 2.4009839133492018e-05, "loss": 0.0597, "step": 138060 }, { "epoch": 4.667838669326211, "grad_norm": 0.9400379657745361, "learning_rate": 2.4002776808001538e-05, "loss": 0.0603, "step": 138070 }, { "epoch": 4.668176747016465, "grad_norm": 0.6929036974906921, "learning_rate": 2.3995715193265355e-05, "loss": 0.0583, "step": 138080 }, { "epoch": 4.668514824706717, "grad_norm": 1.3914748430252075, "learning_rate": 2.398865428947657e-05, "loss": 0.0595, "step": 138090 }, { "epoch": 4.668852902396971, "grad_norm": 1.0358762741088867, "learning_rate": 2.3981594096828203e-05, "loss": 0.0605, "step": 138100 }, { "epoch": 4.669190980087224, "grad_norm": 0.9227786660194397, "learning_rate": 2.3974534615513278e-05, "loss": 0.0607, "step": 138110 }, { "epoch": 4.669529057777478, "grad_norm": 1.373060941696167, "learning_rate": 2.3967475845724785e-05, "loss": 0.0533, "step": 138120 }, { "epoch": 4.66986713546773, "grad_norm": 1.5336346626281738, "learning_rate": 2.3960417787655736e-05, "loss": 0.0645, "step": 138130 }, { "epoch": 4.670205213157984, "grad_norm": 2.9867360591888428, "learning_rate": 2.3953360441499085e-05, "loss": 0.0584, "step": 138140 }, { "epoch": 4.6705432908482365, "grad_norm": 15.131095886230469, "learning_rate": 2.394630380744775e-05, "loss": 0.0574, "step": 138150 }, { "epoch": 4.67088136853849, "grad_norm": 0.9281431436538696, "learning_rate": 2.393924788569471e-05, "loss": 0.0551, "step": 138160 }, { "epoch": 4.6712194462287435, "grad_norm": 1.422518014907837, "learning_rate": 2.3932192676432813e-05, "loss": 0.0549, "step": 138170 }, { "epoch": 4.671557523918997, "grad_norm": 2.3803186416625977, "learning_rate": 2.3925138179855022e-05, "loss": 0.0631, "step": 138180 }, { "epoch": 4.67189560160925, "grad_norm": 2.148075819015503, "learning_rate": 2.391808439615411e-05, "loss": 0.0617, "step": 138190 }, { "epoch": 4.672233679299503, "grad_norm": 0.7948472499847412, "learning_rate": 2.391103132552298e-05, "loss": 0.0558, "step": 138200 }, { "epoch": 4.672571756989756, "grad_norm": 1.0799407958984375, "learning_rate": 2.3903978968154468e-05, "loss": 0.0585, "step": 138210 }, { "epoch": 4.672909834680009, "grad_norm": 1.0434296131134033, "learning_rate": 2.3896927324241368e-05, "loss": 0.0646, "step": 138220 }, { "epoch": 4.673247912370263, "grad_norm": 7.097239017486572, "learning_rate": 2.3889876393976468e-05, "loss": 0.0614, "step": 138230 }, { "epoch": 4.673585990060516, "grad_norm": 1.5507932901382446, "learning_rate": 2.388282617755252e-05, "loss": 0.0593, "step": 138240 }, { "epoch": 4.673924067750769, "grad_norm": 1.7909339666366577, "learning_rate": 2.3875776675162292e-05, "loss": 0.0567, "step": 138250 }, { "epoch": 4.6742621454410225, "grad_norm": 1.3765720129013062, "learning_rate": 2.3868727886998555e-05, "loss": 0.056, "step": 138260 }, { "epoch": 4.674600223131275, "grad_norm": 1.1153112649917603, "learning_rate": 2.386167981325394e-05, "loss": 0.0587, "step": 138270 }, { "epoch": 4.674938300821529, "grad_norm": 1.928034782409668, "learning_rate": 2.38546324541212e-05, "loss": 0.0625, "step": 138280 }, { "epoch": 4.675276378511782, "grad_norm": 4.089865207672119, "learning_rate": 2.3847585809792965e-05, "loss": 0.0554, "step": 138290 }, { "epoch": 4.675614456202036, "grad_norm": 2.408531665802002, "learning_rate": 2.3840539880461926e-05, "loss": 0.0567, "step": 138300 }, { "epoch": 4.675952533892288, "grad_norm": 1.9679447412490845, "learning_rate": 2.3833494666320706e-05, "loss": 0.065, "step": 138310 }, { "epoch": 4.676290611582542, "grad_norm": 2.6065666675567627, "learning_rate": 2.3826450167561887e-05, "loss": 0.0652, "step": 138320 }, { "epoch": 4.676628689272794, "grad_norm": 1.963936686515808, "learning_rate": 2.3819406384378107e-05, "loss": 0.0609, "step": 138330 }, { "epoch": 4.676966766963048, "grad_norm": 1.1300876140594482, "learning_rate": 2.3812363316961922e-05, "loss": 0.0539, "step": 138340 }, { "epoch": 4.677304844653301, "grad_norm": 0.8445937037467957, "learning_rate": 2.380532096550589e-05, "loss": 0.0635, "step": 138350 }, { "epoch": 4.677642922343555, "grad_norm": 3.304083824157715, "learning_rate": 2.3798279330202518e-05, "loss": 0.0567, "step": 138360 }, { "epoch": 4.677981000033808, "grad_norm": 1.4955599308013916, "learning_rate": 2.3791238411244366e-05, "loss": 0.05, "step": 138370 }, { "epoch": 4.678319077724061, "grad_norm": 0.8333058953285217, "learning_rate": 2.3784198208823887e-05, "loss": 0.0568, "step": 138380 }, { "epoch": 4.678657155414315, "grad_norm": 1.709395408630371, "learning_rate": 2.37771587231336e-05, "loss": 0.06, "step": 138390 }, { "epoch": 4.678995233104567, "grad_norm": 1.223320484161377, "learning_rate": 2.377011995436595e-05, "loss": 0.0531, "step": 138400 }, { "epoch": 4.679333310794821, "grad_norm": 0.9655226469039917, "learning_rate": 2.376308190271334e-05, "loss": 0.0552, "step": 138410 }, { "epoch": 4.679671388485074, "grad_norm": 1.1099117994308472, "learning_rate": 2.3756044568368237e-05, "loss": 0.0577, "step": 138420 }, { "epoch": 4.680009466175327, "grad_norm": 1.2308125495910645, "learning_rate": 2.374900795152301e-05, "loss": 0.0611, "step": 138430 }, { "epoch": 4.68034754386558, "grad_norm": 1.5696016550064087, "learning_rate": 2.3741972052370025e-05, "loss": 0.0618, "step": 138440 }, { "epoch": 4.680685621555834, "grad_norm": 1.8621914386749268, "learning_rate": 2.3734936871101687e-05, "loss": 0.0569, "step": 138450 }, { "epoch": 4.6810236992460865, "grad_norm": 1.0228766202926636, "learning_rate": 2.372790240791028e-05, "loss": 0.0641, "step": 138460 }, { "epoch": 4.68136177693634, "grad_norm": 1.0372499227523804, "learning_rate": 2.3720868662988193e-05, "loss": 0.0541, "step": 138470 }, { "epoch": 4.6816998546265935, "grad_norm": 1.7243908643722534, "learning_rate": 2.3713835636527647e-05, "loss": 0.0568, "step": 138480 }, { "epoch": 4.682037932316846, "grad_norm": 1.0866363048553467, "learning_rate": 2.3706803328720977e-05, "loss": 0.0601, "step": 138490 }, { "epoch": 4.6823760100071, "grad_norm": 1.6570477485656738, "learning_rate": 2.3699771739760406e-05, "loss": 0.0615, "step": 138500 }, { "epoch": 4.682714087697353, "grad_norm": 1.9450033903121948, "learning_rate": 2.3692740869838215e-05, "loss": 0.0605, "step": 138510 }, { "epoch": 4.683052165387606, "grad_norm": 1.3301421403884888, "learning_rate": 2.368571071914661e-05, "loss": 0.0558, "step": 138520 }, { "epoch": 4.683390243077859, "grad_norm": 1.7483329772949219, "learning_rate": 2.3678681287877763e-05, "loss": 0.057, "step": 138530 }, { "epoch": 4.683728320768113, "grad_norm": 1.0240825414657593, "learning_rate": 2.36716525762239e-05, "loss": 0.0523, "step": 138540 }, { "epoch": 4.6840663984583655, "grad_norm": 3.7854137420654297, "learning_rate": 2.3664624584377164e-05, "loss": 0.0628, "step": 138550 }, { "epoch": 4.684404476148619, "grad_norm": 0.7681785821914673, "learning_rate": 2.3657597312529706e-05, "loss": 0.0586, "step": 138560 }, { "epoch": 4.6847425538388725, "grad_norm": 1.3929470777511597, "learning_rate": 2.3650570760873615e-05, "loss": 0.054, "step": 138570 }, { "epoch": 4.685080631529125, "grad_norm": 2.3352222442626953, "learning_rate": 2.364354492960102e-05, "loss": 0.0591, "step": 138580 }, { "epoch": 4.685418709219379, "grad_norm": 1.4686272144317627, "learning_rate": 2.363651981890403e-05, "loss": 0.052, "step": 138590 }, { "epoch": 4.685756786909632, "grad_norm": 0.5806270241737366, "learning_rate": 2.3629495428974674e-05, "loss": 0.0549, "step": 138600 }, { "epoch": 4.686094864599885, "grad_norm": 3.0779850482940674, "learning_rate": 2.3622471760005015e-05, "loss": 0.0586, "step": 138610 }, { "epoch": 4.686432942290138, "grad_norm": 2.163224697113037, "learning_rate": 2.3615448812187047e-05, "loss": 0.0633, "step": 138620 }, { "epoch": 4.686771019980392, "grad_norm": 1.5996968746185303, "learning_rate": 2.360842658571281e-05, "loss": 0.0615, "step": 138630 }, { "epoch": 4.687109097670644, "grad_norm": 0.9304125308990479, "learning_rate": 2.3601405080774286e-05, "loss": 0.0636, "step": 138640 }, { "epoch": 4.687447175360898, "grad_norm": 0.9657942056655884, "learning_rate": 2.3594384297563405e-05, "loss": 0.0585, "step": 138650 }, { "epoch": 4.6877852530511515, "grad_norm": 1.987786889076233, "learning_rate": 2.358736423627215e-05, "loss": 0.0594, "step": 138660 }, { "epoch": 4.688123330741404, "grad_norm": 0.7815678715705872, "learning_rate": 2.358034489709242e-05, "loss": 0.0554, "step": 138670 }, { "epoch": 4.688461408431658, "grad_norm": 0.895014226436615, "learning_rate": 2.357332628021618e-05, "loss": 0.0538, "step": 138680 }, { "epoch": 4.688799486121911, "grad_norm": 0.7246031165122986, "learning_rate": 2.3566308385835224e-05, "loss": 0.0623, "step": 138690 }, { "epoch": 4.689137563812164, "grad_norm": 1.1631683111190796, "learning_rate": 2.355929121414147e-05, "loss": 0.0517, "step": 138700 }, { "epoch": 4.689475641502417, "grad_norm": 0.6173862814903259, "learning_rate": 2.3552274765326777e-05, "loss": 0.0584, "step": 138710 }, { "epoch": 4.689813719192671, "grad_norm": 4.537398338317871, "learning_rate": 2.3545259039582962e-05, "loss": 0.0589, "step": 138720 }, { "epoch": 4.690151796882923, "grad_norm": 1.1477869749069214, "learning_rate": 2.353824403710182e-05, "loss": 0.0615, "step": 138730 }, { "epoch": 4.690489874573177, "grad_norm": 1.8433420658111572, "learning_rate": 2.353122975807513e-05, "loss": 0.055, "step": 138740 }, { "epoch": 4.69082795226343, "grad_norm": 2.1337339878082275, "learning_rate": 2.3524216202694687e-05, "loss": 0.0552, "step": 138750 }, { "epoch": 4.691166029953683, "grad_norm": 1.7926994562149048, "learning_rate": 2.3517203371152237e-05, "loss": 0.0591, "step": 138760 }, { "epoch": 4.6915041076439365, "grad_norm": 1.788108229637146, "learning_rate": 2.351019126363947e-05, "loss": 0.053, "step": 138770 }, { "epoch": 4.69184218533419, "grad_norm": 0.8164689540863037, "learning_rate": 2.3503179880348147e-05, "loss": 0.0605, "step": 138780 }, { "epoch": 4.692180263024443, "grad_norm": 0.9001650810241699, "learning_rate": 2.349616922146991e-05, "loss": 0.0555, "step": 138790 }, { "epoch": 4.692518340714696, "grad_norm": 6.600106716156006, "learning_rate": 2.3489159287196465e-05, "loss": 0.0627, "step": 138800 }, { "epoch": 4.69285641840495, "grad_norm": 0.850636899471283, "learning_rate": 2.348215007771945e-05, "loss": 0.059, "step": 138810 }, { "epoch": 4.693194496095202, "grad_norm": 1.3818601369857788, "learning_rate": 2.3475141593230472e-05, "loss": 0.0535, "step": 138820 }, { "epoch": 4.693532573785456, "grad_norm": 0.9158030152320862, "learning_rate": 2.346813383392117e-05, "loss": 0.0614, "step": 138830 }, { "epoch": 4.693870651475709, "grad_norm": 0.9881512522697449, "learning_rate": 2.3461126799983126e-05, "loss": 0.0582, "step": 138840 }, { "epoch": 4.694208729165962, "grad_norm": 0.7333078384399414, "learning_rate": 2.3454120491607905e-05, "loss": 0.0597, "step": 138850 }, { "epoch": 4.6945468068562155, "grad_norm": 0.7085418701171875, "learning_rate": 2.344711490898704e-05, "loss": 0.0583, "step": 138860 }, { "epoch": 4.694884884546469, "grad_norm": 1.1982600688934326, "learning_rate": 2.344011005231209e-05, "loss": 0.0601, "step": 138870 }, { "epoch": 4.695222962236722, "grad_norm": 1.8246629238128662, "learning_rate": 2.343310592177454e-05, "loss": 0.0595, "step": 138880 }, { "epoch": 4.695561039926975, "grad_norm": 1.2796591520309448, "learning_rate": 2.342610251756591e-05, "loss": 0.0591, "step": 138890 }, { "epoch": 4.695899117617229, "grad_norm": 3.000670909881592, "learning_rate": 2.3419099839877655e-05, "loss": 0.0541, "step": 138900 }, { "epoch": 4.696237195307481, "grad_norm": 2.4133801460266113, "learning_rate": 2.3412097888901207e-05, "loss": 0.0631, "step": 138910 }, { "epoch": 4.696575272997735, "grad_norm": 2.143648624420166, "learning_rate": 2.340509666482803e-05, "loss": 0.0582, "step": 138920 }, { "epoch": 4.696913350687988, "grad_norm": 1.2287439107894897, "learning_rate": 2.339809616784952e-05, "loss": 0.0546, "step": 138930 }, { "epoch": 4.697251428378241, "grad_norm": 0.9895800352096558, "learning_rate": 2.339109639815707e-05, "loss": 0.0604, "step": 138940 }, { "epoch": 4.6975895060684945, "grad_norm": 0.7595320343971252, "learning_rate": 2.3384097355942025e-05, "loss": 0.0589, "step": 138950 }, { "epoch": 4.697927583758748, "grad_norm": 0.5971038937568665, "learning_rate": 2.337709904139576e-05, "loss": 0.0595, "step": 138960 }, { "epoch": 4.698265661449001, "grad_norm": 1.1828405857086182, "learning_rate": 2.3370101454709642e-05, "loss": 0.0595, "step": 138970 }, { "epoch": 4.698603739139254, "grad_norm": 1.4589868783950806, "learning_rate": 2.3363104596074914e-05, "loss": 0.0615, "step": 138980 }, { "epoch": 4.698941816829508, "grad_norm": 0.9727356433868408, "learning_rate": 2.3356108465682912e-05, "loss": 0.0587, "step": 138990 }, { "epoch": 4.69927989451976, "grad_norm": 0.8269190192222595, "learning_rate": 2.3349113063724876e-05, "loss": 0.058, "step": 139000 }, { "epoch": 4.699617972210014, "grad_norm": 1.7564570903778076, "learning_rate": 2.33421183903921e-05, "loss": 0.0488, "step": 139010 }, { "epoch": 4.699956049900267, "grad_norm": 1.2844356298446655, "learning_rate": 2.333512444587579e-05, "loss": 0.0597, "step": 139020 }, { "epoch": 4.700294127590521, "grad_norm": 0.9407668113708496, "learning_rate": 2.332813123036714e-05, "loss": 0.054, "step": 139030 }, { "epoch": 4.700632205280773, "grad_norm": 0.9310268759727478, "learning_rate": 2.332113874405738e-05, "loss": 0.0627, "step": 139040 }, { "epoch": 4.700970282971027, "grad_norm": 0.9728630781173706, "learning_rate": 2.331414698713767e-05, "loss": 0.0619, "step": 139050 }, { "epoch": 4.7013083606612796, "grad_norm": 0.7940311431884766, "learning_rate": 2.3307155959799148e-05, "loss": 0.0546, "step": 139060 }, { "epoch": 4.701646438351533, "grad_norm": 5.617149353027344, "learning_rate": 2.3300165662232942e-05, "loss": 0.0523, "step": 139070 }, { "epoch": 4.701984516041787, "grad_norm": 1.4102649688720703, "learning_rate": 2.329317609463017e-05, "loss": 0.0575, "step": 139080 }, { "epoch": 4.70232259373204, "grad_norm": 1.7739295959472656, "learning_rate": 2.3286187257181953e-05, "loss": 0.0566, "step": 139090 }, { "epoch": 4.702660671422293, "grad_norm": 2.0660769939422607, "learning_rate": 2.327919915007934e-05, "loss": 0.0608, "step": 139100 }, { "epoch": 4.702998749112546, "grad_norm": 1.1458812952041626, "learning_rate": 2.3272211773513385e-05, "loss": 0.057, "step": 139110 }, { "epoch": 4.703336826802799, "grad_norm": 1.6507166624069214, "learning_rate": 2.32652251276751e-05, "loss": 0.0542, "step": 139120 }, { "epoch": 4.703674904493052, "grad_norm": 1.8811355829238892, "learning_rate": 2.3258239212755527e-05, "loss": 0.0546, "step": 139130 }, { "epoch": 4.704012982183306, "grad_norm": 1.4073095321655273, "learning_rate": 2.3251254028945657e-05, "loss": 0.0564, "step": 139140 }, { "epoch": 4.704351059873559, "grad_norm": 92.38789367675781, "learning_rate": 2.3244269576436433e-05, "loss": 0.0552, "step": 139150 }, { "epoch": 4.704689137563812, "grad_norm": 2.5632450580596924, "learning_rate": 2.3237285855418845e-05, "loss": 0.0534, "step": 139160 }, { "epoch": 4.7050272152540655, "grad_norm": 1.422435998916626, "learning_rate": 2.3230302866083788e-05, "loss": 0.0566, "step": 139170 }, { "epoch": 4.705365292944318, "grad_norm": 1.3650814294815063, "learning_rate": 2.322332060862223e-05, "loss": 0.0533, "step": 139180 }, { "epoch": 4.705703370634572, "grad_norm": 0.5569343566894531, "learning_rate": 2.3216339083224985e-05, "loss": 0.0515, "step": 139190 }, { "epoch": 4.706041448324825, "grad_norm": 0.9168305397033691, "learning_rate": 2.3209358290082977e-05, "loss": 0.0604, "step": 139200 }, { "epoch": 4.706379526015079, "grad_norm": 2.2629151344299316, "learning_rate": 2.3202378229387062e-05, "loss": 0.0539, "step": 139210 }, { "epoch": 4.706717603705331, "grad_norm": 0.9011463522911072, "learning_rate": 2.3195398901328057e-05, "loss": 0.0629, "step": 139220 }, { "epoch": 4.707055681395585, "grad_norm": 0.6843764185905457, "learning_rate": 2.318842030609678e-05, "loss": 0.0536, "step": 139230 }, { "epoch": 4.7073937590858375, "grad_norm": 1.209521770477295, "learning_rate": 2.3181442443884e-05, "loss": 0.0547, "step": 139240 }, { "epoch": 4.707731836776091, "grad_norm": 0.6680712103843689, "learning_rate": 2.3174465314880528e-05, "loss": 0.0539, "step": 139250 }, { "epoch": 4.7080699144663445, "grad_norm": 1.0899547338485718, "learning_rate": 2.316748891927709e-05, "loss": 0.0509, "step": 139260 }, { "epoch": 4.708407992156598, "grad_norm": 1.490858793258667, "learning_rate": 2.3160513257264416e-05, "loss": 0.0558, "step": 139270 }, { "epoch": 4.708746069846851, "grad_norm": 1.7224034070968628, "learning_rate": 2.3153538329033237e-05, "loss": 0.0611, "step": 139280 }, { "epoch": 4.709084147537104, "grad_norm": 1.1080732345581055, "learning_rate": 2.314656413477422e-05, "loss": 0.0643, "step": 139290 }, { "epoch": 4.709422225227357, "grad_norm": 1.062528371810913, "learning_rate": 2.313959067467807e-05, "loss": 0.0533, "step": 139300 }, { "epoch": 4.70976030291761, "grad_norm": 1.0653882026672363, "learning_rate": 2.3132617948935415e-05, "loss": 0.0559, "step": 139310 }, { "epoch": 4.710098380607864, "grad_norm": 3.036609649658203, "learning_rate": 2.3125645957736898e-05, "loss": 0.061, "step": 139320 }, { "epoch": 4.710436458298117, "grad_norm": 0.9672738909721375, "learning_rate": 2.3118674701273104e-05, "loss": 0.0524, "step": 139330 }, { "epoch": 4.71077453598837, "grad_norm": 1.5548536777496338, "learning_rate": 2.3111704179734657e-05, "loss": 0.0619, "step": 139340 }, { "epoch": 4.711112613678623, "grad_norm": 0.6399564743041992, "learning_rate": 2.310473439331212e-05, "loss": 0.0553, "step": 139350 }, { "epoch": 4.711450691368877, "grad_norm": 0.7836256623268127, "learning_rate": 2.3097765342196027e-05, "loss": 0.0533, "step": 139360 }, { "epoch": 4.71178876905913, "grad_norm": 1.1476070880889893, "learning_rate": 2.3090797026576937e-05, "loss": 0.0566, "step": 139370 }, { "epoch": 4.712126846749383, "grad_norm": 1.3116449117660522, "learning_rate": 2.308382944664533e-05, "loss": 0.0636, "step": 139380 }, { "epoch": 4.712464924439637, "grad_norm": 1.2109462022781372, "learning_rate": 2.3076862602591746e-05, "loss": 0.0617, "step": 139390 }, { "epoch": 4.712803002129889, "grad_norm": 1.1276363134384155, "learning_rate": 2.306989649460659e-05, "loss": 0.0571, "step": 139400 }, { "epoch": 4.713141079820143, "grad_norm": 1.1708316802978516, "learning_rate": 2.3062931122880348e-05, "loss": 0.0541, "step": 139410 }, { "epoch": 4.713479157510396, "grad_norm": 1.0967776775360107, "learning_rate": 2.305596648760347e-05, "loss": 0.0581, "step": 139420 }, { "epoch": 4.713817235200649, "grad_norm": 1.3559616804122925, "learning_rate": 2.3049002588966346e-05, "loss": 0.056, "step": 139430 }, { "epoch": 4.714155312890902, "grad_norm": 0.8802688121795654, "learning_rate": 2.3042039427159363e-05, "loss": 0.054, "step": 139440 }, { "epoch": 4.714493390581156, "grad_norm": 2.5642998218536377, "learning_rate": 2.3035077002372874e-05, "loss": 0.056, "step": 139450 }, { "epoch": 4.7148314682714085, "grad_norm": 21.318500518798828, "learning_rate": 2.302811531479725e-05, "loss": 0.0651, "step": 139460 }, { "epoch": 4.715169545961662, "grad_norm": 4.2622599601745605, "learning_rate": 2.302115436462286e-05, "loss": 0.0609, "step": 139470 }, { "epoch": 4.715507623651916, "grad_norm": 1.9110958576202393, "learning_rate": 2.3014194152039935e-05, "loss": 0.0547, "step": 139480 }, { "epoch": 4.715845701342168, "grad_norm": 0.8599556684494019, "learning_rate": 2.3007234677238825e-05, "loss": 0.0612, "step": 139490 }, { "epoch": 4.716183779032422, "grad_norm": 0.9024580717086792, "learning_rate": 2.300027594040976e-05, "loss": 0.0554, "step": 139500 }, { "epoch": 4.716521856722675, "grad_norm": 0.8063489198684692, "learning_rate": 2.2993317941743027e-05, "loss": 0.0551, "step": 139510 }, { "epoch": 4.716859934412928, "grad_norm": 0.5893672704696655, "learning_rate": 2.2986360681428832e-05, "loss": 0.0579, "step": 139520 }, { "epoch": 4.717198012103181, "grad_norm": 2.058610677719116, "learning_rate": 2.297940415965737e-05, "loss": 0.0542, "step": 139530 }, { "epoch": 4.717536089793435, "grad_norm": 2.195347785949707, "learning_rate": 2.297244837661886e-05, "loss": 0.0517, "step": 139540 }, { "epoch": 4.7178741674836875, "grad_norm": 1.4246346950531006, "learning_rate": 2.2965493332503468e-05, "loss": 0.0599, "step": 139550 }, { "epoch": 4.718212245173941, "grad_norm": 1.7574446201324463, "learning_rate": 2.2958539027501324e-05, "loss": 0.0585, "step": 139560 }, { "epoch": 4.7185503228641945, "grad_norm": 1.4261054992675781, "learning_rate": 2.295158546180255e-05, "loss": 0.0563, "step": 139570 }, { "epoch": 4.718888400554447, "grad_norm": 1.5542281866073608, "learning_rate": 2.2944632635597286e-05, "loss": 0.0618, "step": 139580 }, { "epoch": 4.719226478244701, "grad_norm": 0.9044556021690369, "learning_rate": 2.293768054907558e-05, "loss": 0.0562, "step": 139590 }, { "epoch": 4.719564555934954, "grad_norm": 1.6366115808486938, "learning_rate": 2.2930729202427542e-05, "loss": 0.0524, "step": 139600 }, { "epoch": 4.719902633625207, "grad_norm": 2.7596867084503174, "learning_rate": 2.2923778595843198e-05, "loss": 0.0514, "step": 139610 }, { "epoch": 4.72024071131546, "grad_norm": 1.7409764528274536, "learning_rate": 2.291682872951255e-05, "loss": 0.0531, "step": 139620 }, { "epoch": 4.720578789005714, "grad_norm": 2.044843912124634, "learning_rate": 2.290987960362566e-05, "loss": 0.0579, "step": 139630 }, { "epoch": 4.720916866695966, "grad_norm": 1.2697983980178833, "learning_rate": 2.2902931218372475e-05, "loss": 0.0617, "step": 139640 }, { "epoch": 4.72125494438622, "grad_norm": 1.3005290031433105, "learning_rate": 2.2895983573942958e-05, "loss": 0.0519, "step": 139650 }, { "epoch": 4.7215930220764735, "grad_norm": 2.1708016395568848, "learning_rate": 2.2889036670527086e-05, "loss": 0.052, "step": 139660 }, { "epoch": 4.721931099766726, "grad_norm": 1.2535275220870972, "learning_rate": 2.2882090508314746e-05, "loss": 0.0589, "step": 139670 }, { "epoch": 4.72226917745698, "grad_norm": 1.5106743574142456, "learning_rate": 2.28751450874959e-05, "loss": 0.0561, "step": 139680 }, { "epoch": 4.722607255147233, "grad_norm": 11.35407829284668, "learning_rate": 2.286820040826036e-05, "loss": 0.0571, "step": 139690 }, { "epoch": 4.722945332837486, "grad_norm": 1.0953503847122192, "learning_rate": 2.286125647079805e-05, "loss": 0.0562, "step": 139700 }, { "epoch": 4.723283410527739, "grad_norm": 1.887446641921997, "learning_rate": 2.2854313275298767e-05, "loss": 0.0668, "step": 139710 }, { "epoch": 4.723621488217993, "grad_norm": 1.268044114112854, "learning_rate": 2.2847370821952385e-05, "loss": 0.059, "step": 139720 }, { "epoch": 4.723959565908245, "grad_norm": 2.51472544670105, "learning_rate": 2.284042911094868e-05, "loss": 0.0541, "step": 139730 }, { "epoch": 4.724297643598499, "grad_norm": 1.8901373147964478, "learning_rate": 2.283348814247742e-05, "loss": 0.0549, "step": 139740 }, { "epoch": 4.724635721288752, "grad_norm": 1.2059674263000488, "learning_rate": 2.2826547916728415e-05, "loss": 0.0554, "step": 139750 }, { "epoch": 4.724973798979005, "grad_norm": 0.8408617973327637, "learning_rate": 2.2819608433891376e-05, "loss": 0.055, "step": 139760 }, { "epoch": 4.725311876669259, "grad_norm": 1.8949743509292603, "learning_rate": 2.281266969415603e-05, "loss": 0.0537, "step": 139770 }, { "epoch": 4.725649954359512, "grad_norm": 1.189696192741394, "learning_rate": 2.2805731697712063e-05, "loss": 0.0606, "step": 139780 }, { "epoch": 4.725988032049765, "grad_norm": 0.6846686005592346, "learning_rate": 2.2798794444749177e-05, "loss": 0.0571, "step": 139790 }, { "epoch": 4.726326109740018, "grad_norm": 1.0642696619033813, "learning_rate": 2.279185793545705e-05, "loss": 0.0547, "step": 139800 }, { "epoch": 4.726664187430272, "grad_norm": 2.299992799758911, "learning_rate": 2.278492217002531e-05, "loss": 0.0587, "step": 139810 }, { "epoch": 4.727002265120524, "grad_norm": 8.976980209350586, "learning_rate": 2.2777987148643564e-05, "loss": 0.0602, "step": 139820 }, { "epoch": 4.727340342810778, "grad_norm": 1.4563498497009277, "learning_rate": 2.277105287150141e-05, "loss": 0.0583, "step": 139830 }, { "epoch": 4.727678420501031, "grad_norm": 1.5057309865951538, "learning_rate": 2.276411933878846e-05, "loss": 0.0544, "step": 139840 }, { "epoch": 4.728016498191284, "grad_norm": 1.2582106590270996, "learning_rate": 2.2757186550694247e-05, "loss": 0.0515, "step": 139850 }, { "epoch": 4.7283545758815375, "grad_norm": 1.7960294485092163, "learning_rate": 2.275025450740831e-05, "loss": 0.0554, "step": 139860 }, { "epoch": 4.728692653571791, "grad_norm": 3.2839150428771973, "learning_rate": 2.2743323209120193e-05, "loss": 0.0529, "step": 139870 }, { "epoch": 4.729030731262044, "grad_norm": 2.2013113498687744, "learning_rate": 2.2736392656019363e-05, "loss": 0.0607, "step": 139880 }, { "epoch": 4.729368808952297, "grad_norm": 1.1690559387207031, "learning_rate": 2.2729462848295347e-05, "loss": 0.0582, "step": 139890 }, { "epoch": 4.729706886642551, "grad_norm": 1.629504680633545, "learning_rate": 2.272253378613754e-05, "loss": 0.0568, "step": 139900 }, { "epoch": 4.730044964332803, "grad_norm": 1.347910761833191, "learning_rate": 2.2715605469735406e-05, "loss": 0.0506, "step": 139910 }, { "epoch": 4.730383042023057, "grad_norm": 1.1165932416915894, "learning_rate": 2.2708677899278392e-05, "loss": 0.0644, "step": 139920 }, { "epoch": 4.73072111971331, "grad_norm": 1.0780842304229736, "learning_rate": 2.270175107495587e-05, "loss": 0.0558, "step": 139930 }, { "epoch": 4.731059197403563, "grad_norm": 2.2132601737976074, "learning_rate": 2.269482499695722e-05, "loss": 0.0587, "step": 139940 }, { "epoch": 4.7313972750938165, "grad_norm": 1.6148602962493896, "learning_rate": 2.2687899665471775e-05, "loss": 0.059, "step": 139950 }, { "epoch": 4.73173535278407, "grad_norm": 1.4265483617782593, "learning_rate": 2.268097508068891e-05, "loss": 0.0522, "step": 139960 }, { "epoch": 4.732073430474323, "grad_norm": 0.7728821039199829, "learning_rate": 2.267405124279792e-05, "loss": 0.0603, "step": 139970 }, { "epoch": 4.732411508164576, "grad_norm": 1.1928365230560303, "learning_rate": 2.2667128151988094e-05, "loss": 0.0557, "step": 139980 }, { "epoch": 4.73274958585483, "grad_norm": 0.5914612412452698, "learning_rate": 2.266020580844873e-05, "loss": 0.0562, "step": 139990 }, { "epoch": 4.733087663545083, "grad_norm": 2.568779945373535, "learning_rate": 2.2653284212369048e-05, "loss": 0.0602, "step": 140000 }, { "epoch": 4.733425741235336, "grad_norm": 1.7630386352539062, "learning_rate": 2.2646363363938324e-05, "loss": 0.0564, "step": 140010 }, { "epoch": 4.733763818925589, "grad_norm": 3.424133777618408, "learning_rate": 2.2639443263345745e-05, "loss": 0.0566, "step": 140020 }, { "epoch": 4.734101896615842, "grad_norm": 8.94363021850586, "learning_rate": 2.263252391078049e-05, "loss": 0.0575, "step": 140030 }, { "epoch": 4.734439974306095, "grad_norm": 1.815407395362854, "learning_rate": 2.2625605306431773e-05, "loss": 0.056, "step": 140040 }, { "epoch": 4.734778051996349, "grad_norm": 0.9181845784187317, "learning_rate": 2.261868745048872e-05, "loss": 0.0537, "step": 140050 }, { "epoch": 4.7351161296866024, "grad_norm": 0.821004331111908, "learning_rate": 2.2611770343140464e-05, "loss": 0.0517, "step": 140060 }, { "epoch": 4.735454207376855, "grad_norm": 1.0125842094421387, "learning_rate": 2.26048539845761e-05, "loss": 0.0559, "step": 140070 }, { "epoch": 4.735792285067109, "grad_norm": 1.9291132688522339, "learning_rate": 2.2597938374984755e-05, "loss": 0.0522, "step": 140080 }, { "epoch": 4.736130362757361, "grad_norm": 1.299849033355713, "learning_rate": 2.2591023514555464e-05, "loss": 0.0576, "step": 140090 }, { "epoch": 4.736468440447615, "grad_norm": 0.869774341583252, "learning_rate": 2.258410940347731e-05, "loss": 0.055, "step": 140100 }, { "epoch": 4.736806518137868, "grad_norm": 0.8634256720542908, "learning_rate": 2.2577196041939302e-05, "loss": 0.0591, "step": 140110 }, { "epoch": 4.737144595828122, "grad_norm": 2.2504043579101562, "learning_rate": 2.257028343013044e-05, "loss": 0.0571, "step": 140120 }, { "epoch": 4.737482673518374, "grad_norm": 1.2942273616790771, "learning_rate": 2.256337156823974e-05, "loss": 0.0542, "step": 140130 }, { "epoch": 4.737820751208628, "grad_norm": 1.1376234292984009, "learning_rate": 2.2556460456456152e-05, "loss": 0.0605, "step": 140140 }, { "epoch": 4.7381588288988805, "grad_norm": 0.7259756326675415, "learning_rate": 2.2549550094968624e-05, "loss": 0.0534, "step": 140150 }, { "epoch": 4.738496906589134, "grad_norm": 2.860811710357666, "learning_rate": 2.2542640483966067e-05, "loss": 0.057, "step": 140160 }, { "epoch": 4.7388349842793875, "grad_norm": 1.8236234188079834, "learning_rate": 2.2535731623637403e-05, "loss": 0.0595, "step": 140170 }, { "epoch": 4.739173061969641, "grad_norm": 1.7722543478012085, "learning_rate": 2.252882351417156e-05, "loss": 0.0624, "step": 140180 }, { "epoch": 4.739511139659894, "grad_norm": 0.9092447757720947, "learning_rate": 2.2521916155757317e-05, "loss": 0.0561, "step": 140190 }, { "epoch": 4.739849217350147, "grad_norm": 1.9396753311157227, "learning_rate": 2.2515009548583587e-05, "loss": 0.0586, "step": 140200 }, { "epoch": 4.7401872950404, "grad_norm": 1.4766063690185547, "learning_rate": 2.250810369283915e-05, "loss": 0.0606, "step": 140210 }, { "epoch": 4.740525372730653, "grad_norm": 1.2963508367538452, "learning_rate": 2.250119858871285e-05, "loss": 0.0631, "step": 140220 }, { "epoch": 4.740863450420907, "grad_norm": 0.8163883686065674, "learning_rate": 2.2494294236393447e-05, "loss": 0.049, "step": 140230 }, { "epoch": 4.74120152811116, "grad_norm": 1.650563359260559, "learning_rate": 2.2487390636069695e-05, "loss": 0.0614, "step": 140240 }, { "epoch": 4.741539605801413, "grad_norm": 1.2236255407333374, "learning_rate": 2.248048778793036e-05, "loss": 0.0569, "step": 140250 }, { "epoch": 4.7418776834916665, "grad_norm": 1.2656742334365845, "learning_rate": 2.2473585692164144e-05, "loss": 0.053, "step": 140260 }, { "epoch": 4.742215761181919, "grad_norm": 1.3380773067474365, "learning_rate": 2.2466684348959765e-05, "loss": 0.0674, "step": 140270 }, { "epoch": 4.742553838872173, "grad_norm": 1.0318635702133179, "learning_rate": 2.245978375850587e-05, "loss": 0.0559, "step": 140280 }, { "epoch": 4.742891916562426, "grad_norm": 2.5411458015441895, "learning_rate": 2.2452883920991134e-05, "loss": 0.0598, "step": 140290 }, { "epoch": 4.74322999425268, "grad_norm": 0.9003878235816956, "learning_rate": 2.2445984836604228e-05, "loss": 0.0615, "step": 140300 }, { "epoch": 4.743568071942932, "grad_norm": 2.6518664360046387, "learning_rate": 2.2439086505533747e-05, "loss": 0.0613, "step": 140310 }, { "epoch": 4.743906149633186, "grad_norm": 0.7268229722976685, "learning_rate": 2.2432188927968283e-05, "loss": 0.0554, "step": 140320 }, { "epoch": 4.744244227323439, "grad_norm": 0.6996737122535706, "learning_rate": 2.2425292104096397e-05, "loss": 0.0619, "step": 140330 }, { "epoch": 4.744582305013692, "grad_norm": 1.6640262603759766, "learning_rate": 2.2418396034106687e-05, "loss": 0.059, "step": 140340 }, { "epoch": 4.7449203827039454, "grad_norm": 1.1374156475067139, "learning_rate": 2.2411500718187666e-05, "loss": 0.0605, "step": 140350 }, { "epoch": 4.745258460394199, "grad_norm": 1.346150517463684, "learning_rate": 2.2404606156527835e-05, "loss": 0.0635, "step": 140360 }, { "epoch": 4.745596538084452, "grad_norm": 1.4074310064315796, "learning_rate": 2.239771234931572e-05, "loss": 0.057, "step": 140370 }, { "epoch": 4.745934615774705, "grad_norm": 0.9606496095657349, "learning_rate": 2.2390819296739763e-05, "loss": 0.0592, "step": 140380 }, { "epoch": 4.746272693464959, "grad_norm": 0.8627146482467651, "learning_rate": 2.2383926998988468e-05, "loss": 0.0545, "step": 140390 }, { "epoch": 4.746610771155211, "grad_norm": 0.9321141839027405, "learning_rate": 2.23770354562502e-05, "loss": 0.057, "step": 140400 }, { "epoch": 4.746948848845465, "grad_norm": 1.0934436321258545, "learning_rate": 2.2370144668713404e-05, "loss": 0.0559, "step": 140410 }, { "epoch": 4.747286926535718, "grad_norm": 0.7247810959815979, "learning_rate": 2.2363254636566494e-05, "loss": 0.0618, "step": 140420 }, { "epoch": 4.747625004225971, "grad_norm": 0.6222270727157593, "learning_rate": 2.2356365359997818e-05, "loss": 0.0595, "step": 140430 }, { "epoch": 4.747963081916224, "grad_norm": 1.081223726272583, "learning_rate": 2.2349476839195725e-05, "loss": 0.0554, "step": 140440 }, { "epoch": 4.748301159606478, "grad_norm": 1.5016531944274902, "learning_rate": 2.234258907434853e-05, "loss": 0.0523, "step": 140450 }, { "epoch": 4.7486392372967305, "grad_norm": 0.9358691573143005, "learning_rate": 2.233570206564458e-05, "loss": 0.061, "step": 140460 }, { "epoch": 4.748977314986984, "grad_norm": 7.000713348388672, "learning_rate": 2.232881581327214e-05, "loss": 0.0533, "step": 140470 }, { "epoch": 4.749315392677238, "grad_norm": 0.8665565252304077, "learning_rate": 2.2321930317419464e-05, "loss": 0.0568, "step": 140480 }, { "epoch": 4.74965347036749, "grad_norm": 0.547561526298523, "learning_rate": 2.231504557827483e-05, "loss": 0.0619, "step": 140490 }, { "epoch": 4.749991548057744, "grad_norm": 0.9194198250770569, "learning_rate": 2.2308161596026434e-05, "loss": 0.0576, "step": 140500 }, { "epoch": 4.750329625747997, "grad_norm": 1.6881494522094727, "learning_rate": 2.2301278370862516e-05, "loss": 0.0594, "step": 140510 }, { "epoch": 4.75066770343825, "grad_norm": 1.6866490840911865, "learning_rate": 2.2294395902971238e-05, "loss": 0.0574, "step": 140520 }, { "epoch": 4.751005781128503, "grad_norm": 1.7253450155258179, "learning_rate": 2.228751419254077e-05, "loss": 0.0566, "step": 140530 }, { "epoch": 4.751343858818757, "grad_norm": 0.7539060115814209, "learning_rate": 2.228063323975923e-05, "loss": 0.0543, "step": 140540 }, { "epoch": 4.7516819365090095, "grad_norm": 0.6866176128387451, "learning_rate": 2.2273753044814787e-05, "loss": 0.057, "step": 140550 }, { "epoch": 4.752020014199263, "grad_norm": 6.727355480194092, "learning_rate": 2.2266873607895517e-05, "loss": 0.0573, "step": 140560 }, { "epoch": 4.7523580918895165, "grad_norm": 1.0987756252288818, "learning_rate": 2.2259994929189486e-05, "loss": 0.061, "step": 140570 }, { "epoch": 4.752696169579769, "grad_norm": 4.316006660461426, "learning_rate": 2.2253117008884795e-05, "loss": 0.0605, "step": 140580 }, { "epoch": 4.753034247270023, "grad_norm": 1.4890520572662354, "learning_rate": 2.2246239847169435e-05, "loss": 0.0516, "step": 140590 }, { "epoch": 4.753372324960276, "grad_norm": 1.7476764917373657, "learning_rate": 2.2239363444231492e-05, "loss": 0.0563, "step": 140600 }, { "epoch": 4.753710402650529, "grad_norm": 0.6872177720069885, "learning_rate": 2.2232487800258882e-05, "loss": 0.056, "step": 140610 }, { "epoch": 4.754048480340782, "grad_norm": 1.007149338722229, "learning_rate": 2.2225612915439632e-05, "loss": 0.0543, "step": 140620 }, { "epoch": 4.754386558031036, "grad_norm": 1.3823134899139404, "learning_rate": 2.2218738789961702e-05, "loss": 0.0651, "step": 140630 }, { "epoch": 4.7547246357212885, "grad_norm": 2.769927740097046, "learning_rate": 2.221186542401302e-05, "loss": 0.0606, "step": 140640 }, { "epoch": 4.755062713411542, "grad_norm": 1.160771369934082, "learning_rate": 2.2204992817781494e-05, "loss": 0.0612, "step": 140650 }, { "epoch": 4.7554007911017955, "grad_norm": 0.7466312050819397, "learning_rate": 2.2198120971455004e-05, "loss": 0.0549, "step": 140660 }, { "epoch": 4.755738868792048, "grad_norm": 0.9118491411209106, "learning_rate": 2.219124988522145e-05, "loss": 0.0567, "step": 140670 }, { "epoch": 4.756076946482302, "grad_norm": 1.306219220161438, "learning_rate": 2.218437955926871e-05, "loss": 0.0552, "step": 140680 }, { "epoch": 4.756415024172555, "grad_norm": 1.560558795928955, "learning_rate": 2.217750999378454e-05, "loss": 0.065, "step": 140690 }, { "epoch": 4.756753101862808, "grad_norm": 0.6842731237411499, "learning_rate": 2.217064118895682e-05, "loss": 0.0619, "step": 140700 }, { "epoch": 4.757091179553061, "grad_norm": 1.1018285751342773, "learning_rate": 2.2163773144973293e-05, "loss": 0.059, "step": 140710 }, { "epoch": 4.757429257243315, "grad_norm": 0.6225124001502991, "learning_rate": 2.2156905862021766e-05, "loss": 0.0543, "step": 140720 }, { "epoch": 4.757767334933567, "grad_norm": 2.6565630435943604, "learning_rate": 2.215003934028998e-05, "loss": 0.0571, "step": 140730 }, { "epoch": 4.758105412623821, "grad_norm": 0.9299263954162598, "learning_rate": 2.2143173579965632e-05, "loss": 0.0591, "step": 140740 }, { "epoch": 4.758443490314074, "grad_norm": 0.5836097002029419, "learning_rate": 2.213630858123647e-05, "loss": 0.0614, "step": 140750 }, { "epoch": 4.758781568004327, "grad_norm": 0.782623291015625, "learning_rate": 2.2129444344290167e-05, "loss": 0.0586, "step": 140760 }, { "epoch": 4.759119645694581, "grad_norm": 2.158674478530884, "learning_rate": 2.2122580869314385e-05, "loss": 0.0589, "step": 140770 }, { "epoch": 4.759457723384834, "grad_norm": 0.9706925749778748, "learning_rate": 2.2115718156496752e-05, "loss": 0.0613, "step": 140780 }, { "epoch": 4.759795801075087, "grad_norm": 0.6674181818962097, "learning_rate": 2.2108856206024926e-05, "loss": 0.0562, "step": 140790 }, { "epoch": 4.76013387876534, "grad_norm": 2.3184754848480225, "learning_rate": 2.2101995018086473e-05, "loss": 0.0546, "step": 140800 }, { "epoch": 4.760471956455594, "grad_norm": 1.6785753965377808, "learning_rate": 2.2095134592869017e-05, "loss": 0.0604, "step": 140810 }, { "epoch": 4.760810034145846, "grad_norm": 4.069780349731445, "learning_rate": 2.2088274930560095e-05, "loss": 0.0535, "step": 140820 }, { "epoch": 4.7611481118361, "grad_norm": 1.0099680423736572, "learning_rate": 2.208141603134723e-05, "loss": 0.0546, "step": 140830 }, { "epoch": 4.761486189526353, "grad_norm": 2.598931312561035, "learning_rate": 2.2074557895417986e-05, "loss": 0.0518, "step": 140840 }, { "epoch": 4.761824267216606, "grad_norm": 1.1534734964370728, "learning_rate": 2.2067700522959834e-05, "loss": 0.061, "step": 140850 }, { "epoch": 4.7621623449068595, "grad_norm": 1.5524804592132568, "learning_rate": 2.206084391416024e-05, "loss": 0.0534, "step": 140860 }, { "epoch": 4.762500422597113, "grad_norm": 0.528096079826355, "learning_rate": 2.2053988069206695e-05, "loss": 0.0542, "step": 140870 }, { "epoch": 4.762838500287366, "grad_norm": 2.4609732627868652, "learning_rate": 2.2047132988286596e-05, "loss": 0.0573, "step": 140880 }, { "epoch": 4.763176577977619, "grad_norm": 1.4245699644088745, "learning_rate": 2.2040278671587422e-05, "loss": 0.064, "step": 140890 }, { "epoch": 4.763514655667873, "grad_norm": 1.3794604539871216, "learning_rate": 2.203342511929648e-05, "loss": 0.0529, "step": 140900 }, { "epoch": 4.763852733358125, "grad_norm": 5.788247585296631, "learning_rate": 2.2026572331601215e-05, "loss": 0.054, "step": 140910 }, { "epoch": 4.764190811048379, "grad_norm": 1.9327763319015503, "learning_rate": 2.2019720308688933e-05, "loss": 0.0516, "step": 140920 }, { "epoch": 4.764528888738632, "grad_norm": 1.194419026374817, "learning_rate": 2.2012869050747003e-05, "loss": 0.0548, "step": 140930 }, { "epoch": 4.764866966428885, "grad_norm": 0.7822787761688232, "learning_rate": 2.2006018557962722e-05, "loss": 0.0561, "step": 140940 }, { "epoch": 4.7652050441191385, "grad_norm": 0.8766574263572693, "learning_rate": 2.1999168830523354e-05, "loss": 0.0562, "step": 140950 }, { "epoch": 4.765543121809392, "grad_norm": 0.924314558506012, "learning_rate": 2.1992319868616212e-05, "loss": 0.0545, "step": 140960 }, { "epoch": 4.7658811994996455, "grad_norm": 1.1429909467697144, "learning_rate": 2.1985471672428522e-05, "loss": 0.0514, "step": 140970 }, { "epoch": 4.766219277189898, "grad_norm": 1.211456537246704, "learning_rate": 2.197862424214751e-05, "loss": 0.0568, "step": 140980 }, { "epoch": 4.766557354880152, "grad_norm": 0.8689630627632141, "learning_rate": 2.1971777577960367e-05, "loss": 0.0512, "step": 140990 }, { "epoch": 4.766895432570404, "grad_norm": 1.4539412260055542, "learning_rate": 2.1964931680054295e-05, "loss": 0.0607, "step": 141000 }, { "epoch": 4.767233510260658, "grad_norm": 0.7940934300422668, "learning_rate": 2.195808654861648e-05, "loss": 0.0511, "step": 141010 }, { "epoch": 4.767571587950911, "grad_norm": 0.9630980491638184, "learning_rate": 2.1951242183834043e-05, "loss": 0.0561, "step": 141020 }, { "epoch": 4.767909665641165, "grad_norm": 2.1299638748168945, "learning_rate": 2.1944398585894103e-05, "loss": 0.0586, "step": 141030 }, { "epoch": 4.768247743331417, "grad_norm": 1.6219943761825562, "learning_rate": 2.193755575498375e-05, "loss": 0.0626, "step": 141040 }, { "epoch": 4.768585821021671, "grad_norm": 0.9091047644615173, "learning_rate": 2.19307136912901e-05, "loss": 0.053, "step": 141050 }, { "epoch": 4.768923898711924, "grad_norm": 1.0233110189437866, "learning_rate": 2.1923872395000193e-05, "loss": 0.0552, "step": 141060 }, { "epoch": 4.769261976402177, "grad_norm": 0.9406636357307434, "learning_rate": 2.1917031866301048e-05, "loss": 0.0597, "step": 141070 }, { "epoch": 4.769600054092431, "grad_norm": 1.4157884120941162, "learning_rate": 2.191019210537972e-05, "loss": 0.0567, "step": 141080 }, { "epoch": 4.769938131782684, "grad_norm": 1.0109716653823853, "learning_rate": 2.1903353112423165e-05, "loss": 0.055, "step": 141090 }, { "epoch": 4.770276209472937, "grad_norm": 0.9109995365142822, "learning_rate": 2.189651488761842e-05, "loss": 0.0539, "step": 141100 }, { "epoch": 4.77061428716319, "grad_norm": 1.2002015113830566, "learning_rate": 2.1889677431152362e-05, "loss": 0.0509, "step": 141110 }, { "epoch": 4.770952364853443, "grad_norm": 1.4712625741958618, "learning_rate": 2.1882840743211962e-05, "loss": 0.0594, "step": 141120 }, { "epoch": 4.771290442543696, "grad_norm": 1.0878078937530518, "learning_rate": 2.1876004823984154e-05, "loss": 0.0609, "step": 141130 }, { "epoch": 4.77162852023395, "grad_norm": 4.018407344818115, "learning_rate": 2.186916967365581e-05, "loss": 0.0608, "step": 141140 }, { "epoch": 4.771966597924203, "grad_norm": 0.9707043170928955, "learning_rate": 2.1862335292413793e-05, "loss": 0.0553, "step": 141150 }, { "epoch": 4.772304675614456, "grad_norm": 0.9499292373657227, "learning_rate": 2.1855501680444945e-05, "loss": 0.0608, "step": 141160 }, { "epoch": 4.7726427533047096, "grad_norm": 1.7761069536209106, "learning_rate": 2.1848668837936127e-05, "loss": 0.0572, "step": 141170 }, { "epoch": 4.772980830994962, "grad_norm": 1.8126369714736938, "learning_rate": 2.184183676507412e-05, "loss": 0.0542, "step": 141180 }, { "epoch": 4.773318908685216, "grad_norm": 1.8189611434936523, "learning_rate": 2.1835005462045706e-05, "loss": 0.0583, "step": 141190 }, { "epoch": 4.773656986375469, "grad_norm": 0.7227108478546143, "learning_rate": 2.1828174929037677e-05, "loss": 0.0601, "step": 141200 }, { "epoch": 4.773995064065723, "grad_norm": 1.1892848014831543, "learning_rate": 2.182134516623674e-05, "loss": 0.0532, "step": 141210 }, { "epoch": 4.774333141755975, "grad_norm": 1.1265590190887451, "learning_rate": 2.1814516173829663e-05, "loss": 0.0564, "step": 141220 }, { "epoch": 4.774671219446229, "grad_norm": 2.3573267459869385, "learning_rate": 2.1807687952003124e-05, "loss": 0.0546, "step": 141230 }, { "epoch": 4.7750092971364815, "grad_norm": 0.7318335175514221, "learning_rate": 2.1800860500943786e-05, "loss": 0.0559, "step": 141240 }, { "epoch": 4.775347374826735, "grad_norm": 0.6387603282928467, "learning_rate": 2.1794033820838344e-05, "loss": 0.051, "step": 141250 }, { "epoch": 4.7756854525169885, "grad_norm": 1.3226664066314697, "learning_rate": 2.1787207911873424e-05, "loss": 0.0561, "step": 141260 }, { "epoch": 4.776023530207242, "grad_norm": 1.6344796419143677, "learning_rate": 2.1780382774235642e-05, "loss": 0.0558, "step": 141270 }, { "epoch": 4.776361607897495, "grad_norm": 1.3306190967559814, "learning_rate": 2.1773558408111573e-05, "loss": 0.0616, "step": 141280 }, { "epoch": 4.776699685587748, "grad_norm": 0.5119927525520325, "learning_rate": 2.1766734813687827e-05, "loss": 0.0568, "step": 141290 }, { "epoch": 4.777037763278002, "grad_norm": 3.4029958248138428, "learning_rate": 2.1759911991150933e-05, "loss": 0.0594, "step": 141300 }, { "epoch": 4.777375840968254, "grad_norm": 4.673276901245117, "learning_rate": 2.1753089940687456e-05, "loss": 0.0589, "step": 141310 }, { "epoch": 4.777713918658508, "grad_norm": 1.199865460395813, "learning_rate": 2.1746268662483883e-05, "loss": 0.0595, "step": 141320 }, { "epoch": 4.778051996348761, "grad_norm": 2.073610305786133, "learning_rate": 2.1739448156726688e-05, "loss": 0.0578, "step": 141330 }, { "epoch": 4.778390074039014, "grad_norm": 0.8976482152938843, "learning_rate": 2.1732628423602386e-05, "loss": 0.0576, "step": 141340 }, { "epoch": 4.7787281517292675, "grad_norm": 1.1094739437103271, "learning_rate": 2.172580946329741e-05, "loss": 0.0551, "step": 141350 }, { "epoch": 4.779066229419521, "grad_norm": 1.0856794118881226, "learning_rate": 2.1718991275998173e-05, "loss": 0.0584, "step": 141360 }, { "epoch": 4.779404307109774, "grad_norm": 1.6294492483139038, "learning_rate": 2.1712173861891072e-05, "loss": 0.054, "step": 141370 }, { "epoch": 4.779742384800027, "grad_norm": 2.074777364730835, "learning_rate": 2.1705357221162515e-05, "loss": 0.0566, "step": 141380 }, { "epoch": 4.780080462490281, "grad_norm": 2.334567070007324, "learning_rate": 2.1698541353998892e-05, "loss": 0.0634, "step": 141390 }, { "epoch": 4.780418540180533, "grad_norm": 0.9986705183982849, "learning_rate": 2.169172626058648e-05, "loss": 0.0613, "step": 141400 }, { "epoch": 4.780756617870787, "grad_norm": 1.1791044473648071, "learning_rate": 2.1684911941111663e-05, "loss": 0.0566, "step": 141410 }, { "epoch": 4.78109469556104, "grad_norm": 1.1587225198745728, "learning_rate": 2.1678098395760694e-05, "loss": 0.0559, "step": 141420 }, { "epoch": 4.781432773251293, "grad_norm": 0.6796258091926575, "learning_rate": 2.167128562471989e-05, "loss": 0.0597, "step": 141430 }, { "epoch": 4.781770850941546, "grad_norm": 7.700169563293457, "learning_rate": 2.1664473628175495e-05, "loss": 0.059, "step": 141440 }, { "epoch": 4.7821089286318, "grad_norm": 1.4156683683395386, "learning_rate": 2.165766240631373e-05, "loss": 0.0517, "step": 141450 }, { "epoch": 4.7824470063220526, "grad_norm": 2.0361528396606445, "learning_rate": 2.1650851959320845e-05, "loss": 0.0588, "step": 141460 }, { "epoch": 4.782785084012306, "grad_norm": 1.0679689645767212, "learning_rate": 2.1644042287383015e-05, "loss": 0.0596, "step": 141470 }, { "epoch": 4.78312316170256, "grad_norm": 3.908698081970215, "learning_rate": 2.163723339068642e-05, "loss": 0.0624, "step": 141480 }, { "epoch": 4.783461239392812, "grad_norm": 1.5571184158325195, "learning_rate": 2.1630425269417186e-05, "loss": 0.0532, "step": 141490 }, { "epoch": 4.783799317083066, "grad_norm": 2.6023664474487305, "learning_rate": 2.1623617923761468e-05, "loss": 0.0615, "step": 141500 }, { "epoch": 4.784137394773319, "grad_norm": 1.1671468019485474, "learning_rate": 2.1616811353905398e-05, "loss": 0.0513, "step": 141510 }, { "epoch": 4.784475472463572, "grad_norm": 1.090164303779602, "learning_rate": 2.1610005560035035e-05, "loss": 0.056, "step": 141520 }, { "epoch": 4.784813550153825, "grad_norm": 0.8921260833740234, "learning_rate": 2.160320054233646e-05, "loss": 0.0528, "step": 141530 }, { "epoch": 4.785151627844079, "grad_norm": 0.5502805113792419, "learning_rate": 2.1596396300995697e-05, "loss": 0.0554, "step": 141540 }, { "epoch": 4.7854897055343315, "grad_norm": 0.8923749923706055, "learning_rate": 2.1589592836198802e-05, "loss": 0.0545, "step": 141550 }, { "epoch": 4.785827783224585, "grad_norm": 2.4618563652038574, "learning_rate": 2.1582790148131767e-05, "loss": 0.0553, "step": 141560 }, { "epoch": 4.7861658609148385, "grad_norm": 1.4261537790298462, "learning_rate": 2.1575988236980556e-05, "loss": 0.0626, "step": 141570 }, { "epoch": 4.786503938605091, "grad_norm": 3.4674577713012695, "learning_rate": 2.1569187102931164e-05, "loss": 0.0572, "step": 141580 }, { "epoch": 4.786842016295345, "grad_norm": 0.9910939335823059, "learning_rate": 2.15623867461695e-05, "loss": 0.0579, "step": 141590 }, { "epoch": 4.787180093985598, "grad_norm": 2.6081063747406006, "learning_rate": 2.1555587166881535e-05, "loss": 0.0591, "step": 141600 }, { "epoch": 4.787518171675851, "grad_norm": 2.269138813018799, "learning_rate": 2.15487883652531e-05, "loss": 0.0493, "step": 141610 }, { "epoch": 4.787856249366104, "grad_norm": 1.3724764585494995, "learning_rate": 2.154199034147009e-05, "loss": 0.0483, "step": 141620 }, { "epoch": 4.788194327056358, "grad_norm": 0.8368911743164062, "learning_rate": 2.1535193095718403e-05, "loss": 0.0553, "step": 141630 }, { "epoch": 4.7885324047466105, "grad_norm": 1.1612167358398438, "learning_rate": 2.152839662818384e-05, "loss": 0.0508, "step": 141640 }, { "epoch": 4.788870482436864, "grad_norm": 1.2269200086593628, "learning_rate": 2.1521600939052212e-05, "loss": 0.056, "step": 141650 }, { "epoch": 4.7892085601271175, "grad_norm": 1.1430959701538086, "learning_rate": 2.15148060285093e-05, "loss": 0.0503, "step": 141660 }, { "epoch": 4.78954663781737, "grad_norm": 0.9808061122894287, "learning_rate": 2.1508011896740908e-05, "loss": 0.0556, "step": 141670 }, { "epoch": 4.789884715507624, "grad_norm": 0.7107192873954773, "learning_rate": 2.1501218543932767e-05, "loss": 0.0581, "step": 141680 }, { "epoch": 4.790222793197877, "grad_norm": 0.669911801815033, "learning_rate": 2.1494425970270587e-05, "loss": 0.0521, "step": 141690 }, { "epoch": 4.79056087088813, "grad_norm": 2.8554351329803467, "learning_rate": 2.1487634175940107e-05, "loss": 0.0561, "step": 141700 }, { "epoch": 4.790898948578383, "grad_norm": 8.053348541259766, "learning_rate": 2.1480843161126985e-05, "loss": 0.053, "step": 141710 }, { "epoch": 4.791237026268637, "grad_norm": 0.9845897555351257, "learning_rate": 2.1474052926016906e-05, "loss": 0.0521, "step": 141720 }, { "epoch": 4.791575103958889, "grad_norm": 1.6722707748413086, "learning_rate": 2.146726347079551e-05, "loss": 0.0546, "step": 141730 }, { "epoch": 4.791913181649143, "grad_norm": 2.345289468765259, "learning_rate": 2.146047479564841e-05, "loss": 0.0531, "step": 141740 }, { "epoch": 4.792251259339396, "grad_norm": 1.700573444366455, "learning_rate": 2.1453686900761188e-05, "loss": 0.0521, "step": 141750 }, { "epoch": 4.792589337029649, "grad_norm": 1.2561132907867432, "learning_rate": 2.1446899786319462e-05, "loss": 0.0638, "step": 141760 }, { "epoch": 4.792927414719903, "grad_norm": 1.9327553510665894, "learning_rate": 2.144011345250877e-05, "loss": 0.0574, "step": 141770 }, { "epoch": 4.793265492410156, "grad_norm": 4.062613010406494, "learning_rate": 2.143332789951463e-05, "loss": 0.0592, "step": 141780 }, { "epoch": 4.793603570100409, "grad_norm": 0.8247944116592407, "learning_rate": 2.1426543127522585e-05, "loss": 0.06, "step": 141790 }, { "epoch": 4.793941647790662, "grad_norm": 4.194366931915283, "learning_rate": 2.14197591367181e-05, "loss": 0.0633, "step": 141800 }, { "epoch": 4.794279725480916, "grad_norm": 1.6654484272003174, "learning_rate": 2.1412975927286698e-05, "loss": 0.0575, "step": 141810 }, { "epoch": 4.794617803171168, "grad_norm": 0.9089292883872986, "learning_rate": 2.1406193499413762e-05, "loss": 0.0585, "step": 141820 }, { "epoch": 4.794955880861422, "grad_norm": 6.536923408508301, "learning_rate": 2.1399411853284756e-05, "loss": 0.0531, "step": 141830 }, { "epoch": 4.795293958551675, "grad_norm": 2.337787628173828, "learning_rate": 2.13926309890851e-05, "loss": 0.0586, "step": 141840 }, { "epoch": 4.795632036241928, "grad_norm": 0.6470574140548706, "learning_rate": 2.1385850907000164e-05, "loss": 0.057, "step": 141850 }, { "epoch": 4.7959701139321815, "grad_norm": 1.5226256847381592, "learning_rate": 2.137907160721532e-05, "loss": 0.055, "step": 141860 }, { "epoch": 4.796308191622435, "grad_norm": 0.7710916996002197, "learning_rate": 2.1372293089915877e-05, "loss": 0.0652, "step": 141870 }, { "epoch": 4.796646269312688, "grad_norm": 0.7874986529350281, "learning_rate": 2.1365515355287192e-05, "loss": 0.061, "step": 141880 }, { "epoch": 4.796984347002941, "grad_norm": 1.1185259819030762, "learning_rate": 2.13587384035146e-05, "loss": 0.0552, "step": 141890 }, { "epoch": 4.797322424693195, "grad_norm": 0.8373257517814636, "learning_rate": 2.1351962234783297e-05, "loss": 0.0612, "step": 141900 }, { "epoch": 4.797660502383447, "grad_norm": 5.462630748748779, "learning_rate": 2.1345186849278603e-05, "loss": 0.0573, "step": 141910 }, { "epoch": 4.797998580073701, "grad_norm": 0.6663561463356018, "learning_rate": 2.1338412247185724e-05, "loss": 0.0628, "step": 141920 }, { "epoch": 4.798336657763954, "grad_norm": 1.1356043815612793, "learning_rate": 2.1331638428689898e-05, "loss": 0.0554, "step": 141930 }, { "epoch": 4.798674735454208, "grad_norm": 0.6255533695220947, "learning_rate": 2.1324865393976305e-05, "loss": 0.0594, "step": 141940 }, { "epoch": 4.7990128131444605, "grad_norm": 0.8370229601860046, "learning_rate": 2.131809314323011e-05, "loss": 0.0573, "step": 141950 }, { "epoch": 4.799350890834714, "grad_norm": 1.4585515260696411, "learning_rate": 2.1311321676636486e-05, "loss": 0.0633, "step": 141960 }, { "epoch": 4.799688968524967, "grad_norm": 0.6161798238754272, "learning_rate": 2.130455099438055e-05, "loss": 0.0588, "step": 141970 }, { "epoch": 4.80002704621522, "grad_norm": 1.392695426940918, "learning_rate": 2.129778109664741e-05, "loss": 0.06, "step": 141980 }, { "epoch": 4.800365123905474, "grad_norm": 0.8530558943748474, "learning_rate": 2.1291011983622133e-05, "loss": 0.0637, "step": 141990 }, { "epoch": 4.800703201595727, "grad_norm": 0.6850916743278503, "learning_rate": 2.1284243655489822e-05, "loss": 0.0547, "step": 142000 }, { "epoch": 4.80104127928598, "grad_norm": 1.042775273323059, "learning_rate": 2.1277476112435484e-05, "loss": 0.0571, "step": 142010 }, { "epoch": 4.801379356976233, "grad_norm": 2.179074764251709, "learning_rate": 2.1270709354644175e-05, "loss": 0.0588, "step": 142020 }, { "epoch": 4.801717434666486, "grad_norm": 1.0977171659469604, "learning_rate": 2.1263943382300876e-05, "loss": 0.0592, "step": 142030 }, { "epoch": 4.802055512356739, "grad_norm": 0.6946473121643066, "learning_rate": 2.1257178195590553e-05, "loss": 0.0536, "step": 142040 }, { "epoch": 4.802393590046993, "grad_norm": 9.12421703338623, "learning_rate": 2.1250413794698205e-05, "loss": 0.054, "step": 142050 }, { "epoch": 4.8027316677372465, "grad_norm": 2.0984692573547363, "learning_rate": 2.1243650179808738e-05, "loss": 0.0573, "step": 142060 }, { "epoch": 4.803069745427499, "grad_norm": 1.5619421005249023, "learning_rate": 2.1236887351107053e-05, "loss": 0.0561, "step": 142070 }, { "epoch": 4.803407823117753, "grad_norm": 2.2870028018951416, "learning_rate": 2.1230125308778082e-05, "loss": 0.0554, "step": 142080 }, { "epoch": 4.803745900808005, "grad_norm": 1.8143681287765503, "learning_rate": 2.1223364053006657e-05, "loss": 0.0611, "step": 142090 }, { "epoch": 4.804083978498259, "grad_norm": 2.4644455909729004, "learning_rate": 2.1216603583977685e-05, "loss": 0.0604, "step": 142100 }, { "epoch": 4.804422056188512, "grad_norm": 1.1243605613708496, "learning_rate": 2.120984390187592e-05, "loss": 0.0521, "step": 142110 }, { "epoch": 4.804760133878766, "grad_norm": 1.33639657497406, "learning_rate": 2.1203085006886225e-05, "loss": 0.0542, "step": 142120 }, { "epoch": 4.805098211569018, "grad_norm": 4.69370174407959, "learning_rate": 2.119632689919335e-05, "loss": 0.0588, "step": 142130 }, { "epoch": 4.805436289259272, "grad_norm": 1.0150693655014038, "learning_rate": 2.1189569578982093e-05, "loss": 0.0559, "step": 142140 }, { "epoch": 4.8057743669495245, "grad_norm": 1.0554416179656982, "learning_rate": 2.1182813046437184e-05, "loss": 0.0593, "step": 142150 }, { "epoch": 4.806112444639778, "grad_norm": 2.5016818046569824, "learning_rate": 2.117605730174332e-05, "loss": 0.0586, "step": 142160 }, { "epoch": 4.806450522330032, "grad_norm": 1.4183036088943481, "learning_rate": 2.116930234508524e-05, "loss": 0.0582, "step": 142170 }, { "epoch": 4.806788600020285, "grad_norm": 2.2166335582733154, "learning_rate": 2.11625481766476e-05, "loss": 0.0597, "step": 142180 }, { "epoch": 4.807126677710538, "grad_norm": 1.4360451698303223, "learning_rate": 2.1155794796615058e-05, "loss": 0.0581, "step": 142190 }, { "epoch": 4.807464755400791, "grad_norm": 1.452776551246643, "learning_rate": 2.1149042205172236e-05, "loss": 0.0598, "step": 142200 }, { "epoch": 4.807802833091044, "grad_norm": 1.3375943899154663, "learning_rate": 2.1142290402503763e-05, "loss": 0.0559, "step": 142210 }, { "epoch": 4.808140910781297, "grad_norm": 6.0562286376953125, "learning_rate": 2.113553938879424e-05, "loss": 0.063, "step": 142220 }, { "epoch": 4.808478988471551, "grad_norm": 1.2829017639160156, "learning_rate": 2.112878916422823e-05, "loss": 0.0574, "step": 142230 }, { "epoch": 4.808817066161804, "grad_norm": 3.5581204891204834, "learning_rate": 2.112203972899028e-05, "loss": 0.0588, "step": 142240 }, { "epoch": 4.809155143852057, "grad_norm": 1.6499253511428833, "learning_rate": 2.1115291083264893e-05, "loss": 0.057, "step": 142250 }, { "epoch": 4.8094932215423105, "grad_norm": 1.6931830644607544, "learning_rate": 2.1108543227236616e-05, "loss": 0.0525, "step": 142260 }, { "epoch": 4.809831299232564, "grad_norm": 1.1050678491592407, "learning_rate": 2.1101796161089916e-05, "loss": 0.0569, "step": 142270 }, { "epoch": 4.810169376922817, "grad_norm": 2.195465564727783, "learning_rate": 2.1095049885009234e-05, "loss": 0.0572, "step": 142280 }, { "epoch": 4.81050745461307, "grad_norm": 7.01752233505249, "learning_rate": 2.1088304399179042e-05, "loss": 0.0531, "step": 142290 }, { "epoch": 4.810845532303324, "grad_norm": 1.4712021350860596, "learning_rate": 2.108155970378373e-05, "loss": 0.0522, "step": 142300 }, { "epoch": 4.811183609993576, "grad_norm": 2.6580426692962646, "learning_rate": 2.1074815799007753e-05, "loss": 0.0539, "step": 142310 }, { "epoch": 4.81152168768383, "grad_norm": 1.356094479560852, "learning_rate": 2.10680726850354e-05, "loss": 0.0538, "step": 142320 }, { "epoch": 4.811859765374083, "grad_norm": 1.366996169090271, "learning_rate": 2.106133036205108e-05, "loss": 0.0615, "step": 142330 }, { "epoch": 4.812197843064336, "grad_norm": 1.2582969665527344, "learning_rate": 2.1054588830239124e-05, "loss": 0.0547, "step": 142340 }, { "epoch": 4.8125359207545895, "grad_norm": 0.9022339582443237, "learning_rate": 2.1047848089783846e-05, "loss": 0.0602, "step": 142350 }, { "epoch": 4.812873998444843, "grad_norm": 2.133354663848877, "learning_rate": 2.1041108140869514e-05, "loss": 0.0627, "step": 142360 }, { "epoch": 4.813212076135096, "grad_norm": 2.363340139389038, "learning_rate": 2.1034368983680397e-05, "loss": 0.0592, "step": 142370 }, { "epoch": 4.813550153825349, "grad_norm": 0.8359466791152954, "learning_rate": 2.102763061840076e-05, "loss": 0.0534, "step": 142380 }, { "epoch": 4.813888231515603, "grad_norm": 1.4847419261932373, "learning_rate": 2.1020893045214823e-05, "loss": 0.0552, "step": 142390 }, { "epoch": 4.814226309205855, "grad_norm": 2.8565123081207275, "learning_rate": 2.1014156264306767e-05, "loss": 0.0591, "step": 142400 }, { "epoch": 4.814564386896109, "grad_norm": 1.73942232131958, "learning_rate": 2.1007420275860802e-05, "loss": 0.0504, "step": 142410 }, { "epoch": 4.814902464586362, "grad_norm": 0.9020158648490906, "learning_rate": 2.1000685080061062e-05, "loss": 0.0546, "step": 142420 }, { "epoch": 4.815240542276615, "grad_norm": 0.6110806465148926, "learning_rate": 2.0993950677091713e-05, "loss": 0.05, "step": 142430 }, { "epoch": 4.815578619966868, "grad_norm": 1.0095129013061523, "learning_rate": 2.098721706713685e-05, "loss": 0.0509, "step": 142440 }, { "epoch": 4.815916697657122, "grad_norm": 1.520017385482788, "learning_rate": 2.098048425038056e-05, "loss": 0.0539, "step": 142450 }, { "epoch": 4.816254775347375, "grad_norm": 3.471313238143921, "learning_rate": 2.097375222700695e-05, "loss": 0.0641, "step": 142460 }, { "epoch": 4.816592853037628, "grad_norm": 2.0403945446014404, "learning_rate": 2.0967020997200043e-05, "loss": 0.0551, "step": 142470 }, { "epoch": 4.816930930727882, "grad_norm": 2.7898051738739014, "learning_rate": 2.0960290561143873e-05, "loss": 0.0569, "step": 142480 }, { "epoch": 4.817269008418134, "grad_norm": 0.8476406335830688, "learning_rate": 2.0953560919022436e-05, "loss": 0.0566, "step": 142490 }, { "epoch": 4.817607086108388, "grad_norm": 1.1465574502944946, "learning_rate": 2.094683207101975e-05, "loss": 0.0549, "step": 142500 }, { "epoch": 4.817945163798641, "grad_norm": 1.4129443168640137, "learning_rate": 2.094010401731974e-05, "loss": 0.056, "step": 142510 }, { "epoch": 4.818283241488894, "grad_norm": 1.8231266736984253, "learning_rate": 2.0933376758106386e-05, "loss": 0.0508, "step": 142520 }, { "epoch": 4.818621319179147, "grad_norm": 1.820440649986267, "learning_rate": 2.09266502935636e-05, "loss": 0.053, "step": 142530 }, { "epoch": 4.818959396869401, "grad_norm": 1.2756084203720093, "learning_rate": 2.0919924623875248e-05, "loss": 0.0597, "step": 142540 }, { "epoch": 4.8192974745596535, "grad_norm": 1.497436761856079, "learning_rate": 2.091319974922525e-05, "loss": 0.0592, "step": 142550 }, { "epoch": 4.819635552249907, "grad_norm": 1.811413049697876, "learning_rate": 2.0906475669797444e-05, "loss": 0.0602, "step": 142560 }, { "epoch": 4.8199736299401605, "grad_norm": 1.4914758205413818, "learning_rate": 2.0899752385775662e-05, "loss": 0.0602, "step": 142570 }, { "epoch": 4.820311707630413, "grad_norm": 1.0713976621627808, "learning_rate": 2.0893029897343696e-05, "loss": 0.0561, "step": 142580 }, { "epoch": 4.820649785320667, "grad_norm": 3.5284533500671387, "learning_rate": 2.088630820468536e-05, "loss": 0.0602, "step": 142590 }, { "epoch": 4.82098786301092, "grad_norm": 1.356444239616394, "learning_rate": 2.087958730798446e-05, "loss": 0.0601, "step": 142600 }, { "epoch": 4.821325940701173, "grad_norm": 1.591773271560669, "learning_rate": 2.0872867207424668e-05, "loss": 0.058, "step": 142610 }, { "epoch": 4.821664018391426, "grad_norm": 0.8750753998756409, "learning_rate": 2.086614790318976e-05, "loss": 0.0548, "step": 142620 }, { "epoch": 4.82200209608168, "grad_norm": 0.779781699180603, "learning_rate": 2.0859429395463405e-05, "loss": 0.064, "step": 142630 }, { "epoch": 4.8223401737719325, "grad_norm": 1.874503254890442, "learning_rate": 2.0852711684429314e-05, "loss": 0.056, "step": 142640 }, { "epoch": 4.822678251462186, "grad_norm": 1.0068292617797852, "learning_rate": 2.0845994770271142e-05, "loss": 0.0564, "step": 142650 }, { "epoch": 4.8230163291524395, "grad_norm": 1.4362539052963257, "learning_rate": 2.0839278653172505e-05, "loss": 0.053, "step": 142660 }, { "epoch": 4.823354406842692, "grad_norm": 2.799887180328369, "learning_rate": 2.0832563333317046e-05, "loss": 0.0524, "step": 142670 }, { "epoch": 4.823692484532946, "grad_norm": 2.7538280487060547, "learning_rate": 2.0825848810888355e-05, "loss": 0.0554, "step": 142680 }, { "epoch": 4.824030562223199, "grad_norm": 2.4268901348114014, "learning_rate": 2.081913508606999e-05, "loss": 0.0566, "step": 142690 }, { "epoch": 4.824368639913452, "grad_norm": 1.447513222694397, "learning_rate": 2.0812422159045497e-05, "loss": 0.0617, "step": 142700 }, { "epoch": 4.824706717603705, "grad_norm": 2.614575147628784, "learning_rate": 2.080571002999841e-05, "loss": 0.0542, "step": 142710 }, { "epoch": 4.825044795293959, "grad_norm": 0.9156649112701416, "learning_rate": 2.0798998699112266e-05, "loss": 0.0544, "step": 142720 }, { "epoch": 4.825382872984211, "grad_norm": 1.2485499382019043, "learning_rate": 2.0792288166570528e-05, "loss": 0.0559, "step": 142730 }, { "epoch": 4.825720950674465, "grad_norm": 1.4588452577590942, "learning_rate": 2.0785578432556656e-05, "loss": 0.0581, "step": 142740 }, { "epoch": 4.8260590283647185, "grad_norm": 1.0416038036346436, "learning_rate": 2.0778869497254077e-05, "loss": 0.0549, "step": 142750 }, { "epoch": 4.826397106054971, "grad_norm": 0.922433614730835, "learning_rate": 2.0772161360846248e-05, "loss": 0.0523, "step": 142760 }, { "epoch": 4.826735183745225, "grad_norm": 1.6813398599624634, "learning_rate": 2.0765454023516546e-05, "loss": 0.0599, "step": 142770 }, { "epoch": 4.827073261435478, "grad_norm": 1.95017671585083, "learning_rate": 2.0758747485448328e-05, "loss": 0.0536, "step": 142780 }, { "epoch": 4.827411339125731, "grad_norm": 0.9720635414123535, "learning_rate": 2.0752041746824985e-05, "loss": 0.0516, "step": 142790 }, { "epoch": 4.827749416815984, "grad_norm": 1.1408754587173462, "learning_rate": 2.0745336807829817e-05, "loss": 0.0552, "step": 142800 }, { "epoch": 4.828087494506238, "grad_norm": 8.219165802001953, "learning_rate": 2.073863266864619e-05, "loss": 0.0645, "step": 142810 }, { "epoch": 4.82842557219649, "grad_norm": 17.719148635864258, "learning_rate": 2.073192932945731e-05, "loss": 0.0607, "step": 142820 }, { "epoch": 4.828763649886744, "grad_norm": 1.5897347927093506, "learning_rate": 2.0725226790446495e-05, "loss": 0.0547, "step": 142830 }, { "epoch": 4.829101727576997, "grad_norm": 2.211437702178955, "learning_rate": 2.0718525051796993e-05, "loss": 0.0582, "step": 142840 }, { "epoch": 4.82943980526725, "grad_norm": 2.0061817169189453, "learning_rate": 2.0711824113692018e-05, "loss": 0.0592, "step": 142850 }, { "epoch": 4.8297778829575035, "grad_norm": 1.2050267457962036, "learning_rate": 2.0705123976314765e-05, "loss": 0.0566, "step": 142860 }, { "epoch": 4.830115960647757, "grad_norm": 1.1925091743469238, "learning_rate": 2.06984246398484e-05, "loss": 0.058, "step": 142870 }, { "epoch": 4.83045403833801, "grad_norm": 1.638687252998352, "learning_rate": 2.0691726104476122e-05, "loss": 0.0572, "step": 142880 }, { "epoch": 4.830792116028263, "grad_norm": 0.7916667461395264, "learning_rate": 2.068502837038104e-05, "loss": 0.0583, "step": 142890 }, { "epoch": 4.831130193718517, "grad_norm": 0.6492868661880493, "learning_rate": 2.0678331437746244e-05, "loss": 0.0557, "step": 142900 }, { "epoch": 4.83146827140877, "grad_norm": 2.039184331893921, "learning_rate": 2.0671635306754872e-05, "loss": 0.0629, "step": 142910 }, { "epoch": 4.831806349099023, "grad_norm": 1.272766351699829, "learning_rate": 2.0664939977589953e-05, "loss": 0.0639, "step": 142920 }, { "epoch": 4.832144426789276, "grad_norm": 1.9840620756149292, "learning_rate": 2.0658245450434576e-05, "loss": 0.0535, "step": 142930 }, { "epoch": 4.832482504479529, "grad_norm": 0.6176023483276367, "learning_rate": 2.0651551725471746e-05, "loss": 0.0581, "step": 142940 }, { "epoch": 4.8328205821697825, "grad_norm": 0.7883182168006897, "learning_rate": 2.0644858802884463e-05, "loss": 0.0619, "step": 142950 }, { "epoch": 4.833158659860036, "grad_norm": 1.3604885339736938, "learning_rate": 2.0638166682855697e-05, "loss": 0.0519, "step": 142960 }, { "epoch": 4.8334967375502895, "grad_norm": 1.3904238939285278, "learning_rate": 2.0631475365568438e-05, "loss": 0.057, "step": 142970 }, { "epoch": 4.833834815240542, "grad_norm": 1.264306664466858, "learning_rate": 2.062478485120561e-05, "loss": 0.0554, "step": 142980 }, { "epoch": 4.834172892930796, "grad_norm": 1.1037311553955078, "learning_rate": 2.0618095139950112e-05, "loss": 0.0613, "step": 142990 }, { "epoch": 4.834510970621048, "grad_norm": 1.0425466299057007, "learning_rate": 2.061140623198487e-05, "loss": 0.0557, "step": 143000 }, { "epoch": 4.834849048311302, "grad_norm": 1.415763258934021, "learning_rate": 2.0604718127492728e-05, "loss": 0.0559, "step": 143010 }, { "epoch": 4.835187126001555, "grad_norm": 5.095673084259033, "learning_rate": 2.0598030826656577e-05, "loss": 0.0579, "step": 143020 }, { "epoch": 4.835525203691809, "grad_norm": 1.0254759788513184, "learning_rate": 2.059134432965919e-05, "loss": 0.0602, "step": 143030 }, { "epoch": 4.8358632813820615, "grad_norm": 0.8319883346557617, "learning_rate": 2.0584658636683395e-05, "loss": 0.0494, "step": 143040 }, { "epoch": 4.836201359072315, "grad_norm": 1.8858954906463623, "learning_rate": 2.0577973747911998e-05, "loss": 0.0566, "step": 143050 }, { "epoch": 4.836539436762568, "grad_norm": 0.6307166218757629, "learning_rate": 2.0571289663527753e-05, "loss": 0.0541, "step": 143060 }, { "epoch": 4.836877514452821, "grad_norm": 0.7805625796318054, "learning_rate": 2.0564606383713392e-05, "loss": 0.0589, "step": 143070 }, { "epoch": 4.837215592143075, "grad_norm": 1.2756564617156982, "learning_rate": 2.055792390865161e-05, "loss": 0.0588, "step": 143080 }, { "epoch": 4.837553669833328, "grad_norm": 1.9575090408325195, "learning_rate": 2.055124223852513e-05, "loss": 0.0563, "step": 143090 }, { "epoch": 4.837891747523581, "grad_norm": 0.7692888379096985, "learning_rate": 2.0544561373516664e-05, "loss": 0.0556, "step": 143100 }, { "epoch": 4.838229825213834, "grad_norm": 1.5816634893417358, "learning_rate": 2.0537881313808783e-05, "loss": 0.055, "step": 143110 }, { "epoch": 4.838567902904087, "grad_norm": 2.6471593379974365, "learning_rate": 2.0531202059584177e-05, "loss": 0.0586, "step": 143120 }, { "epoch": 4.83890598059434, "grad_norm": 1.8267879486083984, "learning_rate": 2.052452361102542e-05, "loss": 0.0539, "step": 143130 }, { "epoch": 4.839244058284594, "grad_norm": 1.0242457389831543, "learning_rate": 2.051784596831513e-05, "loss": 0.0562, "step": 143140 }, { "epoch": 4.839582135974847, "grad_norm": 1.25618314743042, "learning_rate": 2.0511169131635855e-05, "loss": 0.0572, "step": 143150 }, { "epoch": 4.8399202136651, "grad_norm": 1.0345802307128906, "learning_rate": 2.0504493101170117e-05, "loss": 0.057, "step": 143160 }, { "epoch": 4.840258291355354, "grad_norm": 2.3703715801239014, "learning_rate": 2.0497817877100484e-05, "loss": 0.0584, "step": 143170 }, { "epoch": 4.840596369045606, "grad_norm": 1.3491524457931519, "learning_rate": 2.049114345960942e-05, "loss": 0.0581, "step": 143180 }, { "epoch": 4.84093444673586, "grad_norm": 1.329222321510315, "learning_rate": 2.0484469848879407e-05, "loss": 0.0598, "step": 143190 }, { "epoch": 4.841272524426113, "grad_norm": 1.2428227663040161, "learning_rate": 2.0477797045092884e-05, "loss": 0.0547, "step": 143200 }, { "epoch": 4.841610602116367, "grad_norm": 1.199743390083313, "learning_rate": 2.0471125048432317e-05, "loss": 0.0549, "step": 143210 }, { "epoch": 4.841948679806619, "grad_norm": 2.3536336421966553, "learning_rate": 2.046445385908008e-05, "loss": 0.0598, "step": 143220 }, { "epoch": 4.842286757496873, "grad_norm": 1.4846582412719727, "learning_rate": 2.0457783477218596e-05, "loss": 0.0565, "step": 143230 }, { "epoch": 4.842624835187126, "grad_norm": 1.0983500480651855, "learning_rate": 2.045111390303021e-05, "loss": 0.0548, "step": 143240 }, { "epoch": 4.842962912877379, "grad_norm": 1.4294581413269043, "learning_rate": 2.0444445136697254e-05, "loss": 0.0603, "step": 143250 }, { "epoch": 4.8433009905676325, "grad_norm": 0.8662699460983276, "learning_rate": 2.0437777178402086e-05, "loss": 0.0513, "step": 143260 }, { "epoch": 4.843639068257886, "grad_norm": 1.2632209062576294, "learning_rate": 2.0431110028326976e-05, "loss": 0.0573, "step": 143270 }, { "epoch": 4.843977145948139, "grad_norm": 1.4632023572921753, "learning_rate": 2.04244436866542e-05, "loss": 0.0563, "step": 143280 }, { "epoch": 4.844315223638392, "grad_norm": 1.8300211429595947, "learning_rate": 2.041777815356603e-05, "loss": 0.0533, "step": 143290 }, { "epoch": 4.844653301328646, "grad_norm": 1.8427258729934692, "learning_rate": 2.0411113429244684e-05, "loss": 0.054, "step": 143300 }, { "epoch": 4.844991379018898, "grad_norm": 0.9189862608909607, "learning_rate": 2.040444951387241e-05, "loss": 0.0592, "step": 143310 }, { "epoch": 4.845329456709152, "grad_norm": 8.944393157958984, "learning_rate": 2.0397786407631338e-05, "loss": 0.0541, "step": 143320 }, { "epoch": 4.845667534399405, "grad_norm": 0.6441840529441833, "learning_rate": 2.0391124110703673e-05, "loss": 0.0547, "step": 143330 }, { "epoch": 4.846005612089658, "grad_norm": 1.0838027000427246, "learning_rate": 2.0384462623271543e-05, "loss": 0.0553, "step": 143340 }, { "epoch": 4.8463436897799115, "grad_norm": 2.219411849975586, "learning_rate": 2.0377801945517095e-05, "loss": 0.056, "step": 143350 }, { "epoch": 4.846681767470165, "grad_norm": 4.332850456237793, "learning_rate": 2.0371142077622408e-05, "loss": 0.0631, "step": 143360 }, { "epoch": 4.847019845160418, "grad_norm": 1.5091285705566406, "learning_rate": 2.0364483019769554e-05, "loss": 0.0574, "step": 143370 }, { "epoch": 4.847357922850671, "grad_norm": 1.0217183828353882, "learning_rate": 2.0357824772140614e-05, "loss": 0.059, "step": 143380 }, { "epoch": 4.847696000540925, "grad_norm": 1.0469423532485962, "learning_rate": 2.035116733491761e-05, "loss": 0.0626, "step": 143390 }, { "epoch": 4.848034078231177, "grad_norm": 0.8953725695610046, "learning_rate": 2.0344510708282556e-05, "loss": 0.061, "step": 143400 }, { "epoch": 4.848372155921431, "grad_norm": 1.1956604719161987, "learning_rate": 2.0337854892417414e-05, "loss": 0.0586, "step": 143410 }, { "epoch": 4.848710233611684, "grad_norm": 3.3213565349578857, "learning_rate": 2.033119988750418e-05, "loss": 0.0591, "step": 143420 }, { "epoch": 4.849048311301937, "grad_norm": 2.007581949234009, "learning_rate": 2.032454569372481e-05, "loss": 0.0566, "step": 143430 }, { "epoch": 4.84938638899219, "grad_norm": 1.4274684190750122, "learning_rate": 2.0317892311261215e-05, "loss": 0.0543, "step": 143440 }, { "epoch": 4.849724466682444, "grad_norm": 0.971023440361023, "learning_rate": 2.031123974029529e-05, "loss": 0.0534, "step": 143450 }, { "epoch": 4.850062544372697, "grad_norm": 1.8129961490631104, "learning_rate": 2.0304587981008904e-05, "loss": 0.0573, "step": 143460 }, { "epoch": 4.85040062206295, "grad_norm": 1.5557576417922974, "learning_rate": 2.029793703358394e-05, "loss": 0.0571, "step": 143470 }, { "epoch": 4.850738699753204, "grad_norm": 4.084322452545166, "learning_rate": 2.0291286898202223e-05, "loss": 0.0563, "step": 143480 }, { "epoch": 4.851076777443456, "grad_norm": 1.313073754310608, "learning_rate": 2.028463757504554e-05, "loss": 0.0564, "step": 143490 }, { "epoch": 4.85141485513371, "grad_norm": 1.1936190128326416, "learning_rate": 2.027798906429572e-05, "loss": 0.058, "step": 143500 }, { "epoch": 4.851752932823963, "grad_norm": 1.0369120836257935, "learning_rate": 2.0271341366134493e-05, "loss": 0.0568, "step": 143510 }, { "epoch": 4.852091010514216, "grad_norm": 1.9283260107040405, "learning_rate": 2.026469448074367e-05, "loss": 0.058, "step": 143520 }, { "epoch": 4.852429088204469, "grad_norm": 2.4399702548980713, "learning_rate": 2.025804840830488e-05, "loss": 0.0538, "step": 143530 }, { "epoch": 4.852767165894723, "grad_norm": 3.055398464202881, "learning_rate": 2.025140314899988e-05, "loss": 0.055, "step": 143540 }, { "epoch": 4.8531052435849755, "grad_norm": 1.6599526405334473, "learning_rate": 2.0244758703010358e-05, "loss": 0.0561, "step": 143550 }, { "epoch": 4.853443321275229, "grad_norm": 7.557242393493652, "learning_rate": 2.023811507051796e-05, "loss": 0.0576, "step": 143560 }, { "epoch": 4.8537813989654826, "grad_norm": 1.2451646327972412, "learning_rate": 2.023147225170431e-05, "loss": 0.0551, "step": 143570 }, { "epoch": 4.854119476655735, "grad_norm": 0.8086908459663391, "learning_rate": 2.022483024675101e-05, "loss": 0.0499, "step": 143580 }, { "epoch": 4.854457554345989, "grad_norm": 1.2907999753952026, "learning_rate": 2.0218189055839677e-05, "loss": 0.0565, "step": 143590 }, { "epoch": 4.854795632036242, "grad_norm": 2.987086772918701, "learning_rate": 2.021154867915187e-05, "loss": 0.0536, "step": 143600 }, { "epoch": 4.855133709726495, "grad_norm": 2.43204402923584, "learning_rate": 2.0204909116869115e-05, "loss": 0.0625, "step": 143610 }, { "epoch": 4.855471787416748, "grad_norm": 4.182828903198242, "learning_rate": 2.0198270369172968e-05, "loss": 0.0582, "step": 143620 }, { "epoch": 4.855809865107002, "grad_norm": 5.379994869232178, "learning_rate": 2.0191632436244894e-05, "loss": 0.0585, "step": 143630 }, { "epoch": 4.8561479427972545, "grad_norm": 1.2747043371200562, "learning_rate": 2.018499531826641e-05, "loss": 0.0546, "step": 143640 }, { "epoch": 4.856486020487508, "grad_norm": 2.694348096847534, "learning_rate": 2.017835901541895e-05, "loss": 0.0558, "step": 143650 }, { "epoch": 4.8568240981777615, "grad_norm": 1.6975743770599365, "learning_rate": 2.017172352788394e-05, "loss": 0.0589, "step": 143660 }, { "epoch": 4.857162175868014, "grad_norm": 1.207773208618164, "learning_rate": 2.0165088855842813e-05, "loss": 0.0571, "step": 143670 }, { "epoch": 4.857500253558268, "grad_norm": 1.6258686780929565, "learning_rate": 2.015845499947695e-05, "loss": 0.0628, "step": 143680 }, { "epoch": 4.857838331248521, "grad_norm": 4.172394275665283, "learning_rate": 2.015182195896772e-05, "loss": 0.0579, "step": 143690 }, { "epoch": 4.858176408938774, "grad_norm": 0.9361477494239807, "learning_rate": 2.0145189734496446e-05, "loss": 0.0584, "step": 143700 }, { "epoch": 4.858514486629027, "grad_norm": 1.4371168613433838, "learning_rate": 2.0138558326244483e-05, "loss": 0.0564, "step": 143710 }, { "epoch": 4.858852564319281, "grad_norm": 1.6542928218841553, "learning_rate": 2.0131927734393103e-05, "loss": 0.0565, "step": 143720 }, { "epoch": 4.859190642009533, "grad_norm": 2.4430065155029297, "learning_rate": 2.0125297959123613e-05, "loss": 0.058, "step": 143730 }, { "epoch": 4.859528719699787, "grad_norm": 1.1241040229797363, "learning_rate": 2.0118669000617253e-05, "loss": 0.062, "step": 143740 }, { "epoch": 4.8598667973900405, "grad_norm": 0.9812861680984497, "learning_rate": 2.011204085905524e-05, "loss": 0.0586, "step": 143750 }, { "epoch": 4.860204875080293, "grad_norm": 0.7339780926704407, "learning_rate": 2.0105413534618822e-05, "loss": 0.054, "step": 143760 }, { "epoch": 4.860542952770547, "grad_norm": 1.0815215110778809, "learning_rate": 2.0098787027489163e-05, "loss": 0.0537, "step": 143770 }, { "epoch": 4.8608810304608, "grad_norm": 2.6487605571746826, "learning_rate": 2.0092161337847433e-05, "loss": 0.0567, "step": 143780 }, { "epoch": 4.861219108151053, "grad_norm": 2.3930881023406982, "learning_rate": 2.008553646587476e-05, "loss": 0.0545, "step": 143790 }, { "epoch": 4.861557185841306, "grad_norm": 7.7058610916137695, "learning_rate": 2.007891241175228e-05, "loss": 0.0497, "step": 143800 }, { "epoch": 4.86189526353156, "grad_norm": 2.6854472160339355, "learning_rate": 2.0072289175661136e-05, "loss": 0.0539, "step": 143810 }, { "epoch": 4.862233341221812, "grad_norm": 5.5882039070129395, "learning_rate": 2.006566675778232e-05, "loss": 0.0587, "step": 143820 }, { "epoch": 4.862571418912066, "grad_norm": 8.770378112792969, "learning_rate": 2.0059045158296948e-05, "loss": 0.0561, "step": 143830 }, { "epoch": 4.862909496602319, "grad_norm": 1.0929776430130005, "learning_rate": 2.0052424377386013e-05, "loss": 0.0601, "step": 143840 }, { "epoch": 4.863247574292572, "grad_norm": 1.282503604888916, "learning_rate": 2.004580441523057e-05, "loss": 0.0558, "step": 143850 }, { "epoch": 4.8635856519828256, "grad_norm": 2.613649606704712, "learning_rate": 2.0039185272011575e-05, "loss": 0.0631, "step": 143860 }, { "epoch": 4.863923729673079, "grad_norm": 4.256990432739258, "learning_rate": 2.0032566947909986e-05, "loss": 0.0562, "step": 143870 }, { "epoch": 4.864261807363333, "grad_norm": 0.7453588247299194, "learning_rate": 2.0025949443106772e-05, "loss": 0.0536, "step": 143880 }, { "epoch": 4.864599885053585, "grad_norm": 1.9365004301071167, "learning_rate": 2.0019332757782838e-05, "loss": 0.0578, "step": 143890 }, { "epoch": 4.864937962743839, "grad_norm": 1.8639204502105713, "learning_rate": 2.0012716892119087e-05, "loss": 0.0601, "step": 143900 }, { "epoch": 4.865276040434091, "grad_norm": 5.547634601593018, "learning_rate": 2.000610184629637e-05, "loss": 0.0608, "step": 143910 }, { "epoch": 4.865614118124345, "grad_norm": 1.819685459136963, "learning_rate": 1.9999487620495566e-05, "loss": 0.0566, "step": 143920 }, { "epoch": 4.865952195814598, "grad_norm": 2.4835257530212402, "learning_rate": 1.9992874214897512e-05, "loss": 0.0643, "step": 143930 }, { "epoch": 4.866290273504852, "grad_norm": 1.5711935758590698, "learning_rate": 1.9986261629683008e-05, "loss": 0.0609, "step": 143940 }, { "epoch": 4.8666283511951045, "grad_norm": 5.214644908905029, "learning_rate": 1.997964986503283e-05, "loss": 0.0533, "step": 143950 }, { "epoch": 4.866966428885358, "grad_norm": 5.392519950866699, "learning_rate": 1.9973038921127728e-05, "loss": 0.0587, "step": 143960 }, { "epoch": 4.867304506575611, "grad_norm": 0.9681711792945862, "learning_rate": 1.9966428798148483e-05, "loss": 0.0573, "step": 143970 }, { "epoch": 4.867642584265864, "grad_norm": 1.7077912092208862, "learning_rate": 1.9959819496275794e-05, "loss": 0.0551, "step": 143980 }, { "epoch": 4.867980661956118, "grad_norm": 3.104732036590576, "learning_rate": 1.995321101569033e-05, "loss": 0.0516, "step": 143990 }, { "epoch": 4.868318739646371, "grad_norm": 1.8052856922149658, "learning_rate": 1.994660335657281e-05, "loss": 0.054, "step": 144000 }, { "epoch": 4.868656817336624, "grad_norm": 6.489900588989258, "learning_rate": 1.9939996519103848e-05, "loss": 0.0586, "step": 144010 }, { "epoch": 4.868994895026877, "grad_norm": 1.569758415222168, "learning_rate": 1.9933390503464123e-05, "loss": 0.0596, "step": 144020 }, { "epoch": 4.86933297271713, "grad_norm": 2.6495118141174316, "learning_rate": 1.992678530983417e-05, "loss": 0.0523, "step": 144030 }, { "epoch": 4.8696710504073835, "grad_norm": 1.1463299989700317, "learning_rate": 1.9920180938394607e-05, "loss": 0.0519, "step": 144040 }, { "epoch": 4.870009128097637, "grad_norm": 3.4661824703216553, "learning_rate": 1.9913577389326015e-05, "loss": 0.0625, "step": 144050 }, { "epoch": 4.8703472057878905, "grad_norm": 1.075652837753296, "learning_rate": 1.9906974662808912e-05, "loss": 0.061, "step": 144060 }, { "epoch": 4.870685283478143, "grad_norm": 2.480005979537964, "learning_rate": 1.9900372759023822e-05, "loss": 0.0605, "step": 144070 }, { "epoch": 4.871023361168397, "grad_norm": 2.4594085216522217, "learning_rate": 1.9893771678151213e-05, "loss": 0.0554, "step": 144080 }, { "epoch": 4.871361438858649, "grad_norm": 1.3843625783920288, "learning_rate": 1.988717142037159e-05, "loss": 0.0587, "step": 144090 }, { "epoch": 4.871699516548903, "grad_norm": 2.8374924659729004, "learning_rate": 1.9880571985865393e-05, "loss": 0.0509, "step": 144100 }, { "epoch": 4.872037594239156, "grad_norm": 1.7539478540420532, "learning_rate": 1.987397337481302e-05, "loss": 0.0571, "step": 144110 }, { "epoch": 4.87237567192941, "grad_norm": 20.19286346435547, "learning_rate": 1.986737558739492e-05, "loss": 0.0551, "step": 144120 }, { "epoch": 4.872713749619662, "grad_norm": 1.2371366024017334, "learning_rate": 1.9860778623791432e-05, "loss": 0.0565, "step": 144130 }, { "epoch": 4.873051827309916, "grad_norm": 1.7094142436981201, "learning_rate": 1.985418248418295e-05, "loss": 0.0554, "step": 144140 }, { "epoch": 4.8733899050001686, "grad_norm": 2.6982216835021973, "learning_rate": 1.9847587168749794e-05, "loss": 0.0559, "step": 144150 }, { "epoch": 4.873727982690422, "grad_norm": 1.373329758644104, "learning_rate": 1.984099267767227e-05, "loss": 0.0554, "step": 144160 }, { "epoch": 4.874066060380676, "grad_norm": 1.3832143545150757, "learning_rate": 1.9834399011130665e-05, "loss": 0.0556, "step": 144170 }, { "epoch": 4.874404138070929, "grad_norm": 1.9576356410980225, "learning_rate": 1.982780616930528e-05, "loss": 0.0596, "step": 144180 }, { "epoch": 4.874742215761182, "grad_norm": 2.955887794494629, "learning_rate": 1.9821214152376332e-05, "loss": 0.0563, "step": 144190 }, { "epoch": 4.875080293451435, "grad_norm": 8.738587379455566, "learning_rate": 1.9814622960524038e-05, "loss": 0.055, "step": 144200 }, { "epoch": 4.875418371141689, "grad_norm": 38.325096130371094, "learning_rate": 1.980803259392863e-05, "loss": 0.0599, "step": 144210 }, { "epoch": 4.875756448831941, "grad_norm": 2.8108885288238525, "learning_rate": 1.9801443052770245e-05, "loss": 0.0652, "step": 144220 }, { "epoch": 4.876094526522195, "grad_norm": 3.2991039752960205, "learning_rate": 1.9794854337229085e-05, "loss": 0.0596, "step": 144230 }, { "epoch": 4.876432604212448, "grad_norm": 6.444465637207031, "learning_rate": 1.9788266447485258e-05, "loss": 0.0523, "step": 144240 }, { "epoch": 4.876770681902701, "grad_norm": 1.1280288696289062, "learning_rate": 1.978167938371886e-05, "loss": 0.0531, "step": 144250 }, { "epoch": 4.8771087595929545, "grad_norm": 16.413543701171875, "learning_rate": 1.9775093146110013e-05, "loss": 0.0517, "step": 144260 }, { "epoch": 4.877446837283208, "grad_norm": 1.7453097105026245, "learning_rate": 1.9768507734838766e-05, "loss": 0.054, "step": 144270 }, { "epoch": 4.877784914973461, "grad_norm": 1.7930494546890259, "learning_rate": 1.9761923150085156e-05, "loss": 0.0602, "step": 144280 }, { "epoch": 4.878122992663714, "grad_norm": 1.4509984254837036, "learning_rate": 1.9755339392029186e-05, "loss": 0.06, "step": 144290 }, { "epoch": 4.878461070353968, "grad_norm": 1.2557398080825806, "learning_rate": 1.974875646085088e-05, "loss": 0.0603, "step": 144300 }, { "epoch": 4.87879914804422, "grad_norm": 3.736323595046997, "learning_rate": 1.9742174356730248e-05, "loss": 0.0573, "step": 144310 }, { "epoch": 4.879137225734474, "grad_norm": 2.1381537914276123, "learning_rate": 1.973559307984716e-05, "loss": 0.0565, "step": 144320 }, { "epoch": 4.879475303424727, "grad_norm": 4.035816669464111, "learning_rate": 1.9729012630381604e-05, "loss": 0.0546, "step": 144330 }, { "epoch": 4.87981338111498, "grad_norm": 1.970908761024475, "learning_rate": 1.9722433008513457e-05, "loss": 0.056, "step": 144340 }, { "epoch": 4.8801514588052335, "grad_norm": 3.01643443107605, "learning_rate": 1.971585421442263e-05, "loss": 0.06, "step": 144350 }, { "epoch": 4.880489536495487, "grad_norm": 1.5370680093765259, "learning_rate": 1.9709276248288978e-05, "loss": 0.054, "step": 144360 }, { "epoch": 4.88082761418574, "grad_norm": 1.2152405977249146, "learning_rate": 1.9702699110292307e-05, "loss": 0.0569, "step": 144370 }, { "epoch": 4.881165691875993, "grad_norm": 1.922504186630249, "learning_rate": 1.9696122800612482e-05, "loss": 0.0566, "step": 144380 }, { "epoch": 4.881503769566247, "grad_norm": 3.647596836090088, "learning_rate": 1.9689547319429274e-05, "loss": 0.0572, "step": 144390 }, { "epoch": 4.881841847256499, "grad_norm": 1.8214526176452637, "learning_rate": 1.968297266692246e-05, "loss": 0.0585, "step": 144400 }, { "epoch": 4.882179924946753, "grad_norm": 1.206263780593872, "learning_rate": 1.9676398843271767e-05, "loss": 0.0548, "step": 144410 }, { "epoch": 4.882518002637006, "grad_norm": 1.4874769449234009, "learning_rate": 1.9669825848656948e-05, "loss": 0.0578, "step": 144420 }, { "epoch": 4.882856080327259, "grad_norm": 1.8295494318008423, "learning_rate": 1.9663253683257684e-05, "loss": 0.0545, "step": 144430 }, { "epoch": 4.883194158017512, "grad_norm": 6.492847919464111, "learning_rate": 1.965668234725368e-05, "loss": 0.0581, "step": 144440 }, { "epoch": 4.883532235707766, "grad_norm": 0.9524174332618713, "learning_rate": 1.9650111840824582e-05, "loss": 0.0595, "step": 144450 }, { "epoch": 4.883870313398019, "grad_norm": 1.9252289533615112, "learning_rate": 1.964354216415001e-05, "loss": 0.0621, "step": 144460 }, { "epoch": 4.884208391088272, "grad_norm": 0.9804880619049072, "learning_rate": 1.9636973317409603e-05, "loss": 0.0555, "step": 144470 }, { "epoch": 4.884546468778526, "grad_norm": 1.6923476457595825, "learning_rate": 1.9630405300782946e-05, "loss": 0.0582, "step": 144480 }, { "epoch": 4.884884546468778, "grad_norm": 1.2931370735168457, "learning_rate": 1.962383811444957e-05, "loss": 0.0555, "step": 144490 }, { "epoch": 4.885222624159032, "grad_norm": 0.9816908240318298, "learning_rate": 1.9617271758589077e-05, "loss": 0.0555, "step": 144500 }, { "epoch": 4.885560701849285, "grad_norm": 1.5417087078094482, "learning_rate": 1.9610706233380936e-05, "loss": 0.0538, "step": 144510 }, { "epoch": 4.885898779539538, "grad_norm": 2.077317953109741, "learning_rate": 1.960414153900471e-05, "loss": 0.0683, "step": 144520 }, { "epoch": 4.886236857229791, "grad_norm": 5.062514305114746, "learning_rate": 1.9597577675639795e-05, "loss": 0.0581, "step": 144530 }, { "epoch": 4.886574934920045, "grad_norm": 1.7119402885437012, "learning_rate": 1.9591014643465706e-05, "loss": 0.0563, "step": 144540 }, { "epoch": 4.8869130126102975, "grad_norm": 5.95920991897583, "learning_rate": 1.958445244266183e-05, "loss": 0.0548, "step": 144550 }, { "epoch": 4.887251090300551, "grad_norm": 4.553107738494873, "learning_rate": 1.9577891073407622e-05, "loss": 0.0527, "step": 144560 }, { "epoch": 4.887589167990805, "grad_norm": 1.0940114259719849, "learning_rate": 1.9571330535882437e-05, "loss": 0.0547, "step": 144570 }, { "epoch": 4.887927245681057, "grad_norm": 1.7535896301269531, "learning_rate": 1.9564770830265633e-05, "loss": 0.0554, "step": 144580 }, { "epoch": 4.888265323371311, "grad_norm": 1.5903323888778687, "learning_rate": 1.955821195673657e-05, "loss": 0.0596, "step": 144590 }, { "epoch": 4.888603401061564, "grad_norm": 0.9592286944389343, "learning_rate": 1.955165391547456e-05, "loss": 0.0637, "step": 144600 }, { "epoch": 4.888941478751817, "grad_norm": 1.5688517093658447, "learning_rate": 1.9545096706658893e-05, "loss": 0.0626, "step": 144610 }, { "epoch": 4.88927955644207, "grad_norm": 1.8677788972854614, "learning_rate": 1.9538540330468826e-05, "loss": 0.0539, "step": 144620 }, { "epoch": 4.889617634132324, "grad_norm": 1.5184991359710693, "learning_rate": 1.953198478708362e-05, "loss": 0.0531, "step": 144630 }, { "epoch": 4.8899557118225765, "grad_norm": 1.9498268365859985, "learning_rate": 1.9525430076682516e-05, "loss": 0.0549, "step": 144640 }, { "epoch": 4.89029378951283, "grad_norm": 2.0617594718933105, "learning_rate": 1.9518876199444707e-05, "loss": 0.0546, "step": 144650 }, { "epoch": 4.8906318672030835, "grad_norm": 1.5117982625961304, "learning_rate": 1.9512323155549377e-05, "loss": 0.0574, "step": 144660 }, { "epoch": 4.890969944893336, "grad_norm": 4.1074934005737305, "learning_rate": 1.9505770945175656e-05, "loss": 0.056, "step": 144670 }, { "epoch": 4.89130802258359, "grad_norm": 2.161064386367798, "learning_rate": 1.949921956850272e-05, "loss": 0.0526, "step": 144680 }, { "epoch": 4.891646100273843, "grad_norm": 2.553476572036743, "learning_rate": 1.9492669025709665e-05, "loss": 0.0597, "step": 144690 }, { "epoch": 4.891984177964096, "grad_norm": 2.0330469608306885, "learning_rate": 1.9486119316975553e-05, "loss": 0.0526, "step": 144700 }, { "epoch": 4.892322255654349, "grad_norm": 0.7940866351127625, "learning_rate": 1.94795704424795e-05, "loss": 0.0544, "step": 144710 }, { "epoch": 4.892660333344603, "grad_norm": 2.2399401664733887, "learning_rate": 1.9473022402400503e-05, "loss": 0.0525, "step": 144720 }, { "epoch": 4.892998411034855, "grad_norm": 1.5814237594604492, "learning_rate": 1.9466475196917643e-05, "loss": 0.0513, "step": 144730 }, { "epoch": 4.893336488725109, "grad_norm": 1.376792073249817, "learning_rate": 1.9459928826209845e-05, "loss": 0.0511, "step": 144740 }, { "epoch": 4.8936745664153625, "grad_norm": 1.798573613166809, "learning_rate": 1.945338329045612e-05, "loss": 0.0579, "step": 144750 }, { "epoch": 4.894012644105615, "grad_norm": 1.4244738817214966, "learning_rate": 1.9446838589835437e-05, "loss": 0.0549, "step": 144760 }, { "epoch": 4.894350721795869, "grad_norm": 2.783906936645508, "learning_rate": 1.944029472452671e-05, "loss": 0.0631, "step": 144770 }, { "epoch": 4.894688799486122, "grad_norm": 2.7959959506988525, "learning_rate": 1.943375169470884e-05, "loss": 0.0575, "step": 144780 }, { "epoch": 4.895026877176375, "grad_norm": 1.263317346572876, "learning_rate": 1.94272095005607e-05, "loss": 0.059, "step": 144790 }, { "epoch": 4.895364954866628, "grad_norm": 1.4874396324157715, "learning_rate": 1.942066814226119e-05, "loss": 0.0552, "step": 144800 }, { "epoch": 4.895703032556882, "grad_norm": 1.294701337814331, "learning_rate": 1.9414127619989115e-05, "loss": 0.0519, "step": 144810 }, { "epoch": 4.896041110247134, "grad_norm": 3.099613666534424, "learning_rate": 1.940758793392329e-05, "loss": 0.0555, "step": 144820 }, { "epoch": 4.896379187937388, "grad_norm": 1.5058344602584839, "learning_rate": 1.9401049084242533e-05, "loss": 0.0528, "step": 144830 }, { "epoch": 4.896717265627641, "grad_norm": 1.393373966217041, "learning_rate": 1.9394511071125582e-05, "loss": 0.0586, "step": 144840 }, { "epoch": 4.897055343317895, "grad_norm": 5.36557674407959, "learning_rate": 1.9387973894751217e-05, "loss": 0.0587, "step": 144850 }, { "epoch": 4.897393421008148, "grad_norm": 1.775654673576355, "learning_rate": 1.9381437555298142e-05, "loss": 0.0609, "step": 144860 }, { "epoch": 4.897731498698401, "grad_norm": 1.9348700046539307, "learning_rate": 1.937490205294505e-05, "loss": 0.0518, "step": 144870 }, { "epoch": 4.898069576388654, "grad_norm": 1.6901577711105347, "learning_rate": 1.936836738787064e-05, "loss": 0.0569, "step": 144880 }, { "epoch": 4.898407654078907, "grad_norm": 3.944443464279175, "learning_rate": 1.9361833560253567e-05, "loss": 0.0608, "step": 144890 }, { "epoch": 4.898745731769161, "grad_norm": 2.2452824115753174, "learning_rate": 1.935530057027245e-05, "loss": 0.061, "step": 144900 }, { "epoch": 4.899083809459414, "grad_norm": 2.1321773529052734, "learning_rate": 1.9348768418105884e-05, "loss": 0.0557, "step": 144910 }, { "epoch": 4.899421887149667, "grad_norm": 1.5714465379714966, "learning_rate": 1.934223710393249e-05, "loss": 0.0562, "step": 144920 }, { "epoch": 4.89975996483992, "grad_norm": 2.3654234409332275, "learning_rate": 1.9335706627930804e-05, "loss": 0.0586, "step": 144930 }, { "epoch": 4.900098042530173, "grad_norm": 4.3381242752075195, "learning_rate": 1.932917699027939e-05, "loss": 0.0584, "step": 144940 }, { "epoch": 4.9004361202204265, "grad_norm": 1.4844268560409546, "learning_rate": 1.932264819115676e-05, "loss": 0.0563, "step": 144950 }, { "epoch": 4.90077419791068, "grad_norm": 8.470353126525879, "learning_rate": 1.9316120230741385e-05, "loss": 0.0585, "step": 144960 }, { "epoch": 4.9011122756009335, "grad_norm": 1.7124607563018799, "learning_rate": 1.9309593109211773e-05, "loss": 0.0612, "step": 144970 }, { "epoch": 4.901450353291186, "grad_norm": 3.689936399459839, "learning_rate": 1.9303066826746358e-05, "loss": 0.058, "step": 144980 }, { "epoch": 4.90178843098144, "grad_norm": 2.9233922958374023, "learning_rate": 1.929654138352356e-05, "loss": 0.0574, "step": 144990 }, { "epoch": 4.902126508671692, "grad_norm": 1.5331237316131592, "learning_rate": 1.9290016779721763e-05, "loss": 0.0535, "step": 145000 }, { "epoch": 4.902464586361946, "grad_norm": 4.0178680419921875, "learning_rate": 1.928349301551938e-05, "loss": 0.065, "step": 145010 }, { "epoch": 4.902802664052199, "grad_norm": 1.1943289041519165, "learning_rate": 1.927697009109479e-05, "loss": 0.0597, "step": 145020 }, { "epoch": 4.903140741742453, "grad_norm": 2.6975386142730713, "learning_rate": 1.9270448006626256e-05, "loss": 0.0582, "step": 145030 }, { "epoch": 4.9034788194327055, "grad_norm": 7.5003862380981445, "learning_rate": 1.926392676229215e-05, "loss": 0.0512, "step": 145040 }, { "epoch": 4.903816897122959, "grad_norm": 3.8957979679107666, "learning_rate": 1.9257406358270724e-05, "loss": 0.0576, "step": 145050 }, { "epoch": 4.904154974813212, "grad_norm": 4.89808464050293, "learning_rate": 1.9250886794740275e-05, "loss": 0.0509, "step": 145060 }, { "epoch": 4.904493052503465, "grad_norm": 2.4431161880493164, "learning_rate": 1.9244368071879027e-05, "loss": 0.0581, "step": 145070 }, { "epoch": 4.904831130193719, "grad_norm": 1.893417477607727, "learning_rate": 1.9237850189865187e-05, "loss": 0.0607, "step": 145080 }, { "epoch": 4.905169207883972, "grad_norm": 2.0492007732391357, "learning_rate": 1.9231333148876984e-05, "loss": 0.0578, "step": 145090 }, { "epoch": 4.905507285574225, "grad_norm": 2.5881097316741943, "learning_rate": 1.9224816949092574e-05, "loss": 0.0587, "step": 145100 }, { "epoch": 4.905845363264478, "grad_norm": 2.116960287094116, "learning_rate": 1.9218301590690103e-05, "loss": 0.0586, "step": 145110 }, { "epoch": 4.906183440954731, "grad_norm": 11.518045425415039, "learning_rate": 1.9211787073847682e-05, "loss": 0.0555, "step": 145120 }, { "epoch": 4.906521518644984, "grad_norm": 3.0563600063323975, "learning_rate": 1.920527339874344e-05, "loss": 0.0606, "step": 145130 }, { "epoch": 4.906859596335238, "grad_norm": 1.9135963916778564, "learning_rate": 1.9198760565555473e-05, "loss": 0.053, "step": 145140 }, { "epoch": 4.9071976740254915, "grad_norm": 2.6623787879943848, "learning_rate": 1.9192248574461818e-05, "loss": 0.0583, "step": 145150 }, { "epoch": 4.907535751715744, "grad_norm": 9.947526931762695, "learning_rate": 1.9185737425640517e-05, "loss": 0.059, "step": 145160 }, { "epoch": 4.907873829405998, "grad_norm": 4.441251754760742, "learning_rate": 1.9179227119269555e-05, "loss": 0.0607, "step": 145170 }, { "epoch": 4.908211907096251, "grad_norm": 2.520185708999634, "learning_rate": 1.9172717655526967e-05, "loss": 0.0608, "step": 145180 }, { "epoch": 4.908549984786504, "grad_norm": 3.1079163551330566, "learning_rate": 1.91662090345907e-05, "loss": 0.0585, "step": 145190 }, { "epoch": 4.908888062476757, "grad_norm": 4.131626129150391, "learning_rate": 1.915970125663867e-05, "loss": 0.0527, "step": 145200 }, { "epoch": 4.909226140167011, "grad_norm": 1.7599427700042725, "learning_rate": 1.915319432184884e-05, "loss": 0.0602, "step": 145210 }, { "epoch": 4.909564217857263, "grad_norm": 10.475113868713379, "learning_rate": 1.9146688230399075e-05, "loss": 0.0502, "step": 145220 }, { "epoch": 4.909902295547517, "grad_norm": 2.6995668411254883, "learning_rate": 1.9140182982467292e-05, "loss": 0.0576, "step": 145230 }, { "epoch": 4.91024037323777, "grad_norm": 1.8762348890304565, "learning_rate": 1.9133678578231286e-05, "loss": 0.0566, "step": 145240 }, { "epoch": 4.910578450928023, "grad_norm": 2.848806858062744, "learning_rate": 1.9127175017868902e-05, "loss": 0.0584, "step": 145250 }, { "epoch": 4.9109165286182765, "grad_norm": 4.012544631958008, "learning_rate": 1.9120672301557978e-05, "loss": 0.0654, "step": 145260 }, { "epoch": 4.91125460630853, "grad_norm": 4.16076135635376, "learning_rate": 1.911417042947627e-05, "loss": 0.0609, "step": 145270 }, { "epoch": 4.911592683998783, "grad_norm": 2.377847909927368, "learning_rate": 1.9107669401801543e-05, "loss": 0.0634, "step": 145280 }, { "epoch": 4.911930761689036, "grad_norm": 4.97927713394165, "learning_rate": 1.9101169218711508e-05, "loss": 0.0554, "step": 145290 }, { "epoch": 4.91226883937929, "grad_norm": 1.8018802404403687, "learning_rate": 1.9094669880383915e-05, "loss": 0.0584, "step": 145300 }, { "epoch": 4.912606917069542, "grad_norm": 9.240036964416504, "learning_rate": 1.908817138699644e-05, "loss": 0.067, "step": 145310 }, { "epoch": 4.912944994759796, "grad_norm": 2.778226137161255, "learning_rate": 1.908167373872673e-05, "loss": 0.0519, "step": 145320 }, { "epoch": 4.913283072450049, "grad_norm": 8.596787452697754, "learning_rate": 1.907517693575246e-05, "loss": 0.0518, "step": 145330 }, { "epoch": 4.913621150140302, "grad_norm": 2.1140129566192627, "learning_rate": 1.9068680978251213e-05, "loss": 0.0526, "step": 145340 }, { "epoch": 4.9139592278305555, "grad_norm": 7.930289268493652, "learning_rate": 1.9062185866400628e-05, "loss": 0.0563, "step": 145350 }, { "epoch": 4.914297305520809, "grad_norm": 2.4727797508239746, "learning_rate": 1.905569160037826e-05, "loss": 0.0537, "step": 145360 }, { "epoch": 4.914635383211062, "grad_norm": 1.664157748222351, "learning_rate": 1.904919818036166e-05, "loss": 0.0526, "step": 145370 }, { "epoch": 4.914973460901315, "grad_norm": 1.7187919616699219, "learning_rate": 1.9042705606528333e-05, "loss": 0.0641, "step": 145380 }, { "epoch": 4.915311538591569, "grad_norm": 1.4955557584762573, "learning_rate": 1.9036213879055825e-05, "loss": 0.0589, "step": 145390 }, { "epoch": 4.915649616281821, "grad_norm": 2.5189692974090576, "learning_rate": 1.902972299812159e-05, "loss": 0.0573, "step": 145400 }, { "epoch": 4.915987693972075, "grad_norm": 2.6720919609069824, "learning_rate": 1.9023232963903086e-05, "loss": 0.0582, "step": 145410 }, { "epoch": 4.916325771662328, "grad_norm": 2.8996591567993164, "learning_rate": 1.9016743776577768e-05, "loss": 0.0543, "step": 145420 }, { "epoch": 4.916663849352581, "grad_norm": 1.891143560409546, "learning_rate": 1.9010255436323016e-05, "loss": 0.057, "step": 145430 }, { "epoch": 4.9170019270428345, "grad_norm": 4.5384039878845215, "learning_rate": 1.900376794331626e-05, "loss": 0.0586, "step": 145440 }, { "epoch": 4.917340004733088, "grad_norm": 4.415981769561768, "learning_rate": 1.8997281297734835e-05, "loss": 0.0588, "step": 145450 }, { "epoch": 4.917678082423341, "grad_norm": 3.282421112060547, "learning_rate": 1.8990795499756087e-05, "loss": 0.0565, "step": 145460 }, { "epoch": 4.918016160113594, "grad_norm": 1.5971934795379639, "learning_rate": 1.898431054955735e-05, "loss": 0.0627, "step": 145470 }, { "epoch": 4.918354237803848, "grad_norm": 3.6159932613372803, "learning_rate": 1.897782644731591e-05, "loss": 0.0554, "step": 145480 }, { "epoch": 4.9186923154941, "grad_norm": 1.9782854318618774, "learning_rate": 1.8971343193209045e-05, "loss": 0.0611, "step": 145490 }, { "epoch": 4.919030393184354, "grad_norm": 2.3046858310699463, "learning_rate": 1.896486078741398e-05, "loss": 0.0565, "step": 145500 }, { "epoch": 4.919368470874607, "grad_norm": 3.043426752090454, "learning_rate": 1.895837923010796e-05, "loss": 0.0554, "step": 145510 }, { "epoch": 4.91970654856486, "grad_norm": 2.1866729259490967, "learning_rate": 1.895189852146823e-05, "loss": 0.0593, "step": 145520 }, { "epoch": 4.920044626255113, "grad_norm": 2.3520944118499756, "learning_rate": 1.894541866167189e-05, "loss": 0.0584, "step": 145530 }, { "epoch": 4.920382703945367, "grad_norm": 4.697669506072998, "learning_rate": 1.893893965089616e-05, "loss": 0.0566, "step": 145540 }, { "epoch": 4.9207207816356195, "grad_norm": 3.0703983306884766, "learning_rate": 1.8932461489318125e-05, "loss": 0.0546, "step": 145550 }, { "epoch": 4.921058859325873, "grad_norm": 2.267871618270874, "learning_rate": 1.8925984177114946e-05, "loss": 0.0545, "step": 145560 }, { "epoch": 4.921396937016127, "grad_norm": 6.323413372039795, "learning_rate": 1.8919507714463685e-05, "loss": 0.0543, "step": 145570 }, { "epoch": 4.921735014706379, "grad_norm": 1.3127671480178833, "learning_rate": 1.891303210154139e-05, "loss": 0.0572, "step": 145580 }, { "epoch": 4.922073092396633, "grad_norm": 1.3206770420074463, "learning_rate": 1.8906557338525132e-05, "loss": 0.0563, "step": 145590 }, { "epoch": 4.922411170086886, "grad_norm": 2.3414218425750732, "learning_rate": 1.8900083425591913e-05, "loss": 0.0594, "step": 145600 }, { "epoch": 4.922749247777139, "grad_norm": 1.3877923488616943, "learning_rate": 1.8893610362918736e-05, "loss": 0.0562, "step": 145610 }, { "epoch": 4.923087325467392, "grad_norm": 3.081429958343506, "learning_rate": 1.8887138150682548e-05, "loss": 0.0587, "step": 145620 }, { "epoch": 4.923425403157646, "grad_norm": 2.6028614044189453, "learning_rate": 1.8880666789060326e-05, "loss": 0.0557, "step": 145630 }, { "epoch": 4.9237634808478985, "grad_norm": 1.8281971216201782, "learning_rate": 1.8874196278228973e-05, "loss": 0.0532, "step": 145640 }, { "epoch": 4.924101558538152, "grad_norm": 2.2373127937316895, "learning_rate": 1.8867726618365412e-05, "loss": 0.0593, "step": 145650 }, { "epoch": 4.9244396362284055, "grad_norm": 2.6317176818847656, "learning_rate": 1.886125780964651e-05, "loss": 0.0655, "step": 145660 }, { "epoch": 4.924777713918658, "grad_norm": 2.4761946201324463, "learning_rate": 1.8854789852249106e-05, "loss": 0.0546, "step": 145670 }, { "epoch": 4.925115791608912, "grad_norm": 2.472341537475586, "learning_rate": 1.884832274635006e-05, "loss": 0.0565, "step": 145680 }, { "epoch": 4.925453869299165, "grad_norm": 2.095691442489624, "learning_rate": 1.884185649212617e-05, "loss": 0.056, "step": 145690 }, { "epoch": 4.925791946989418, "grad_norm": 1.8023015260696411, "learning_rate": 1.88353910897542e-05, "loss": 0.0567, "step": 145700 }, { "epoch": 4.926130024679671, "grad_norm": 2.3389759063720703, "learning_rate": 1.882892653941095e-05, "loss": 0.0507, "step": 145710 }, { "epoch": 4.926468102369925, "grad_norm": 2.0827407836914062, "learning_rate": 1.882246284127312e-05, "loss": 0.0591, "step": 145720 }, { "epoch": 4.9268061800601775, "grad_norm": 4.392285346984863, "learning_rate": 1.8815999995517475e-05, "loss": 0.0611, "step": 145730 }, { "epoch": 4.927144257750431, "grad_norm": 1.8750255107879639, "learning_rate": 1.8809538002320644e-05, "loss": 0.0597, "step": 145740 }, { "epoch": 4.9274823354406845, "grad_norm": 1.3110260963439941, "learning_rate": 1.8803076861859338e-05, "loss": 0.0612, "step": 145750 }, { "epoch": 4.927820413130937, "grad_norm": 2.439847230911255, "learning_rate": 1.879661657431017e-05, "loss": 0.0573, "step": 145760 }, { "epoch": 4.928158490821191, "grad_norm": 1.0739012956619263, "learning_rate": 1.8790157139849813e-05, "loss": 0.0568, "step": 145770 }, { "epoch": 4.928496568511444, "grad_norm": 4.092965126037598, "learning_rate": 1.8783698558654823e-05, "loss": 0.0596, "step": 145780 }, { "epoch": 4.928834646201697, "grad_norm": 5.780481338500977, "learning_rate": 1.8777240830901777e-05, "loss": 0.0541, "step": 145790 }, { "epoch": 4.92917272389195, "grad_norm": 0.7574576735496521, "learning_rate": 1.877078395676725e-05, "loss": 0.0511, "step": 145800 }, { "epoch": 4.929510801582204, "grad_norm": 2.4464423656463623, "learning_rate": 1.8764327936427757e-05, "loss": 0.0572, "step": 145810 }, { "epoch": 4.929848879272457, "grad_norm": 4.214411735534668, "learning_rate": 1.8757872770059803e-05, "loss": 0.0548, "step": 145820 }, { "epoch": 4.93018695696271, "grad_norm": 2.862257957458496, "learning_rate": 1.875141845783986e-05, "loss": 0.0553, "step": 145830 }, { "epoch": 4.930525034652963, "grad_norm": 2.710301637649536, "learning_rate": 1.8744964999944392e-05, "loss": 0.0601, "step": 145840 }, { "epoch": 4.930863112343216, "grad_norm": 1.8309859037399292, "learning_rate": 1.873851239654985e-05, "loss": 0.0514, "step": 145850 }, { "epoch": 4.93120119003347, "grad_norm": 0.673391580581665, "learning_rate": 1.8732060647832645e-05, "loss": 0.0625, "step": 145860 }, { "epoch": 4.931539267723723, "grad_norm": 1.4196094274520874, "learning_rate": 1.8725609753969146e-05, "loss": 0.0526, "step": 145870 }, { "epoch": 4.931877345413977, "grad_norm": 1.979047179222107, "learning_rate": 1.871915971513572e-05, "loss": 0.0518, "step": 145880 }, { "epoch": 4.932215423104229, "grad_norm": 1.7233775854110718, "learning_rate": 1.8712710531508724e-05, "loss": 0.058, "step": 145890 }, { "epoch": 4.932553500794483, "grad_norm": 0.9806755185127258, "learning_rate": 1.8706262203264475e-05, "loss": 0.0608, "step": 145900 }, { "epoch": 4.932891578484735, "grad_norm": 1.4507540464401245, "learning_rate": 1.869981473057924e-05, "loss": 0.0564, "step": 145910 }, { "epoch": 4.933229656174989, "grad_norm": 2.1838736534118652, "learning_rate": 1.869336811362933e-05, "loss": 0.052, "step": 145920 }, { "epoch": 4.933567733865242, "grad_norm": 1.2853896617889404, "learning_rate": 1.868692235259095e-05, "loss": 0.0612, "step": 145930 }, { "epoch": 4.933905811555496, "grad_norm": 1.4636306762695312, "learning_rate": 1.8680477447640394e-05, "loss": 0.0617, "step": 145940 }, { "epoch": 4.9342438892457485, "grad_norm": 1.419885277748108, "learning_rate": 1.8674033398953777e-05, "loss": 0.055, "step": 145950 }, { "epoch": 4.934581966936002, "grad_norm": 1.3837573528289795, "learning_rate": 1.8667590206707318e-05, "loss": 0.0604, "step": 145960 }, { "epoch": 4.934920044626255, "grad_norm": 1.1356523036956787, "learning_rate": 1.8661147871077196e-05, "loss": 0.0582, "step": 145970 }, { "epoch": 4.935258122316508, "grad_norm": 2.070366144180298, "learning_rate": 1.8654706392239514e-05, "loss": 0.0514, "step": 145980 }, { "epoch": 4.935596200006762, "grad_norm": 1.4211781024932861, "learning_rate": 1.8648265770370384e-05, "loss": 0.0606, "step": 145990 }, { "epoch": 4.935934277697015, "grad_norm": 1.2177287340164185, "learning_rate": 1.8641826005645867e-05, "loss": 0.0539, "step": 146000 }, { "epoch": 4.936272355387268, "grad_norm": 0.8882477879524231, "learning_rate": 1.863538709824207e-05, "loss": 0.0635, "step": 146010 }, { "epoch": 4.936610433077521, "grad_norm": 1.3478795289993286, "learning_rate": 1.8628949048335005e-05, "loss": 0.0572, "step": 146020 }, { "epoch": 4.936948510767774, "grad_norm": 2.03311824798584, "learning_rate": 1.8622511856100665e-05, "loss": 0.0539, "step": 146030 }, { "epoch": 4.9372865884580275, "grad_norm": 2.158392906188965, "learning_rate": 1.8616075521715088e-05, "loss": 0.057, "step": 146040 }, { "epoch": 4.937624666148281, "grad_norm": 2.827777624130249, "learning_rate": 1.8609640045354194e-05, "loss": 0.0565, "step": 146050 }, { "epoch": 4.9379627438385345, "grad_norm": 1.8286153078079224, "learning_rate": 1.8603205427193964e-05, "loss": 0.0565, "step": 146060 }, { "epoch": 4.938300821528787, "grad_norm": 3.443614959716797, "learning_rate": 1.8596771667410306e-05, "loss": 0.0551, "step": 146070 }, { "epoch": 4.938638899219041, "grad_norm": 5.780159950256348, "learning_rate": 1.8590338766179093e-05, "loss": 0.0583, "step": 146080 }, { "epoch": 4.938976976909293, "grad_norm": 1.2708673477172852, "learning_rate": 1.8583906723676242e-05, "loss": 0.0567, "step": 146090 }, { "epoch": 4.939315054599547, "grad_norm": 1.423203468322754, "learning_rate": 1.857747554007757e-05, "loss": 0.0581, "step": 146100 }, { "epoch": 4.9396531322898, "grad_norm": 2.1805953979492188, "learning_rate": 1.857104521555892e-05, "loss": 0.0571, "step": 146110 }, { "epoch": 4.939991209980054, "grad_norm": 1.6558165550231934, "learning_rate": 1.856461575029606e-05, "loss": 0.0599, "step": 146120 }, { "epoch": 4.940329287670306, "grad_norm": 1.1335656642913818, "learning_rate": 1.8558187144464822e-05, "loss": 0.0523, "step": 146130 }, { "epoch": 4.94066736536056, "grad_norm": 3.5298659801483154, "learning_rate": 1.8551759398240914e-05, "loss": 0.0534, "step": 146140 }, { "epoch": 4.9410054430508135, "grad_norm": 1.5825015306472778, "learning_rate": 1.8545332511800106e-05, "loss": 0.0567, "step": 146150 }, { "epoch": 4.941343520741066, "grad_norm": 1.305723786354065, "learning_rate": 1.853890648531809e-05, "loss": 0.0664, "step": 146160 }, { "epoch": 4.94168159843132, "grad_norm": 0.9600107669830322, "learning_rate": 1.8532481318970535e-05, "loss": 0.0561, "step": 146170 }, { "epoch": 4.942019676121573, "grad_norm": 2.6463749408721924, "learning_rate": 1.8526057012933134e-05, "loss": 0.0549, "step": 146180 }, { "epoch": 4.942357753811826, "grad_norm": 1.582468867301941, "learning_rate": 1.851963356738151e-05, "loss": 0.057, "step": 146190 }, { "epoch": 4.942695831502079, "grad_norm": 2.711737632751465, "learning_rate": 1.8513210982491284e-05, "loss": 0.0624, "step": 146200 }, { "epoch": 4.943033909192333, "grad_norm": 4.153191566467285, "learning_rate": 1.8506789258438016e-05, "loss": 0.062, "step": 146210 }, { "epoch": 4.943371986882585, "grad_norm": 1.3303406238555908, "learning_rate": 1.8500368395397306e-05, "loss": 0.051, "step": 146220 }, { "epoch": 4.943710064572839, "grad_norm": 2.2935292720794678, "learning_rate": 1.8493948393544718e-05, "loss": 0.0572, "step": 146230 }, { "epoch": 4.944048142263092, "grad_norm": 1.2372500896453857, "learning_rate": 1.8487529253055712e-05, "loss": 0.0624, "step": 146240 }, { "epoch": 4.944386219953345, "grad_norm": 0.7429509162902832, "learning_rate": 1.8481110974105835e-05, "loss": 0.0558, "step": 146250 }, { "epoch": 4.9447242976435986, "grad_norm": 1.2279119491577148, "learning_rate": 1.8474693556870527e-05, "loss": 0.0549, "step": 146260 }, { "epoch": 4.945062375333852, "grad_norm": 2.030634880065918, "learning_rate": 1.846827700152527e-05, "loss": 0.0542, "step": 146270 }, { "epoch": 4.945400453024105, "grad_norm": 1.1888673305511475, "learning_rate": 1.846186130824547e-05, "loss": 0.0607, "step": 146280 }, { "epoch": 4.945738530714358, "grad_norm": 2.0571160316467285, "learning_rate": 1.8455446477206523e-05, "loss": 0.0573, "step": 146290 }, { "epoch": 4.946076608404612, "grad_norm": 1.8561238050460815, "learning_rate": 1.844903250858383e-05, "loss": 0.053, "step": 146300 }, { "epoch": 4.946414686094864, "grad_norm": 2.2092432975769043, "learning_rate": 1.844261940255274e-05, "loss": 0.0557, "step": 146310 }, { "epoch": 4.946752763785118, "grad_norm": 0.9590107202529907, "learning_rate": 1.8436207159288578e-05, "loss": 0.0576, "step": 146320 }, { "epoch": 4.947090841475371, "grad_norm": 0.9211825728416443, "learning_rate": 1.8429795778966634e-05, "loss": 0.055, "step": 146330 }, { "epoch": 4.947428919165624, "grad_norm": 1.4490821361541748, "learning_rate": 1.8423385261762216e-05, "loss": 0.053, "step": 146340 }, { "epoch": 4.9477669968558775, "grad_norm": 7.931228160858154, "learning_rate": 1.84169756078506e-05, "loss": 0.051, "step": 146350 }, { "epoch": 4.948105074546131, "grad_norm": 0.8999221324920654, "learning_rate": 1.8410566817407004e-05, "loss": 0.0546, "step": 146360 }, { "epoch": 4.948443152236384, "grad_norm": 1.6102198362350464, "learning_rate": 1.8404158890606644e-05, "loss": 0.0687, "step": 146370 }, { "epoch": 4.948781229926637, "grad_norm": 3.062453031539917, "learning_rate": 1.8397751827624693e-05, "loss": 0.0555, "step": 146380 }, { "epoch": 4.949119307616891, "grad_norm": 1.1030241250991821, "learning_rate": 1.839134562863635e-05, "loss": 0.056, "step": 146390 }, { "epoch": 4.949457385307143, "grad_norm": 1.6278711557388306, "learning_rate": 1.838494029381675e-05, "loss": 0.0585, "step": 146400 }, { "epoch": 4.949795462997397, "grad_norm": 1.613722801208496, "learning_rate": 1.837853582334098e-05, "loss": 0.0546, "step": 146410 }, { "epoch": 4.95013354068765, "grad_norm": 1.390149712562561, "learning_rate": 1.837213221738418e-05, "loss": 0.0585, "step": 146420 }, { "epoch": 4.950471618377903, "grad_norm": 1.1706050634384155, "learning_rate": 1.8365729476121386e-05, "loss": 0.0543, "step": 146430 }, { "epoch": 4.9508096960681565, "grad_norm": 2.7374980449676514, "learning_rate": 1.8359327599727698e-05, "loss": 0.0542, "step": 146440 }, { "epoch": 4.95114777375841, "grad_norm": 1.5477063655853271, "learning_rate": 1.8352926588378074e-05, "loss": 0.0481, "step": 146450 }, { "epoch": 4.951485851448663, "grad_norm": 1.527674913406372, "learning_rate": 1.8346526442247546e-05, "loss": 0.055, "step": 146460 }, { "epoch": 4.951823929138916, "grad_norm": 0.8318172097206116, "learning_rate": 1.8340127161511112e-05, "loss": 0.0578, "step": 146470 }, { "epoch": 4.95216200682917, "grad_norm": 1.5790148973464966, "learning_rate": 1.833372874634371e-05, "loss": 0.062, "step": 146480 }, { "epoch": 4.952500084519422, "grad_norm": 0.9659658670425415, "learning_rate": 1.832733119692026e-05, "loss": 0.056, "step": 146490 }, { "epoch": 4.952838162209676, "grad_norm": 3.478381633758545, "learning_rate": 1.8320934513415667e-05, "loss": 0.0533, "step": 146500 }, { "epoch": 4.953176239899929, "grad_norm": 1.0403341054916382, "learning_rate": 1.8314538696004834e-05, "loss": 0.0554, "step": 146510 }, { "epoch": 4.953514317590182, "grad_norm": 4.909489154815674, "learning_rate": 1.830814374486261e-05, "loss": 0.0603, "step": 146520 }, { "epoch": 4.953852395280435, "grad_norm": 7.264016628265381, "learning_rate": 1.830174966016382e-05, "loss": 0.0536, "step": 146530 }, { "epoch": 4.954190472970689, "grad_norm": 2.307742118835449, "learning_rate": 1.8295356442083294e-05, "loss": 0.06, "step": 146540 }, { "epoch": 4.9545285506609416, "grad_norm": 2.6164841651916504, "learning_rate": 1.82889640907958e-05, "loss": 0.061, "step": 146550 }, { "epoch": 4.954866628351195, "grad_norm": 2.876692056655884, "learning_rate": 1.828257260647613e-05, "loss": 0.0542, "step": 146560 }, { "epoch": 4.955204706041449, "grad_norm": 1.1647701263427734, "learning_rate": 1.8276181989299006e-05, "loss": 0.0555, "step": 146570 }, { "epoch": 4.955542783731701, "grad_norm": 2.1602351665496826, "learning_rate": 1.8269792239439156e-05, "loss": 0.0543, "step": 146580 }, { "epoch": 4.955880861421955, "grad_norm": 1.355591058731079, "learning_rate": 1.8263403357071242e-05, "loss": 0.0519, "step": 146590 }, { "epoch": 4.956218939112208, "grad_norm": 1.1697372198104858, "learning_rate": 1.8257015342369978e-05, "loss": 0.0566, "step": 146600 }, { "epoch": 4.956557016802461, "grad_norm": 1.2429524660110474, "learning_rate": 1.8250628195509988e-05, "loss": 0.0507, "step": 146610 }, { "epoch": 4.956895094492714, "grad_norm": 1.609033465385437, "learning_rate": 1.824424191666588e-05, "loss": 0.0551, "step": 146620 }, { "epoch": 4.957233172182968, "grad_norm": 1.0930156707763672, "learning_rate": 1.823785650601228e-05, "loss": 0.0564, "step": 146630 }, { "epoch": 4.9575712498732205, "grad_norm": 0.8808678388595581, "learning_rate": 1.823147196372374e-05, "loss": 0.0575, "step": 146640 }, { "epoch": 4.957909327563474, "grad_norm": 0.9625693559646606, "learning_rate": 1.8225088289974835e-05, "loss": 0.057, "step": 146650 }, { "epoch": 4.9582474052537275, "grad_norm": 1.2226295471191406, "learning_rate": 1.821870548494008e-05, "loss": 0.0534, "step": 146660 }, { "epoch": 4.95858548294398, "grad_norm": 1.1030579805374146, "learning_rate": 1.821232354879396e-05, "loss": 0.0598, "step": 146670 }, { "epoch": 4.958923560634234, "grad_norm": 2.8689231872558594, "learning_rate": 1.8205942481710987e-05, "loss": 0.0538, "step": 146680 }, { "epoch": 4.959261638324487, "grad_norm": 1.2364579439163208, "learning_rate": 1.8199562283865607e-05, "loss": 0.0569, "step": 146690 }, { "epoch": 4.95959971601474, "grad_norm": 3.533081293106079, "learning_rate": 1.819318295543224e-05, "loss": 0.0567, "step": 146700 }, { "epoch": 4.959937793704993, "grad_norm": 1.0650714635849, "learning_rate": 1.818680449658529e-05, "loss": 0.064, "step": 146710 }, { "epoch": 4.960275871395247, "grad_norm": 0.5600181221961975, "learning_rate": 1.8180426907499154e-05, "loss": 0.0517, "step": 146720 }, { "epoch": 4.9606139490854995, "grad_norm": 1.4802075624465942, "learning_rate": 1.8174050188348224e-05, "loss": 0.0529, "step": 146730 }, { "epoch": 4.960952026775753, "grad_norm": 1.7554092407226562, "learning_rate": 1.8167674339306773e-05, "loss": 0.0543, "step": 146740 }, { "epoch": 4.9612901044660065, "grad_norm": 1.105726718902588, "learning_rate": 1.816129936054916e-05, "loss": 0.0603, "step": 146750 }, { "epoch": 4.961628182156259, "grad_norm": 1.8008980751037598, "learning_rate": 1.815492525224965e-05, "loss": 0.0605, "step": 146760 }, { "epoch": 4.961966259846513, "grad_norm": 1.2884323596954346, "learning_rate": 1.8148552014582532e-05, "loss": 0.0593, "step": 146770 }, { "epoch": 4.962304337536766, "grad_norm": 1.3396785259246826, "learning_rate": 1.8142179647722042e-05, "loss": 0.0558, "step": 146780 }, { "epoch": 4.96264241522702, "grad_norm": 0.9442320466041565, "learning_rate": 1.8135808151842364e-05, "loss": 0.0559, "step": 146790 }, { "epoch": 4.962980492917272, "grad_norm": 1.1232693195343018, "learning_rate": 1.812943752711775e-05, "loss": 0.0536, "step": 146800 }, { "epoch": 4.963318570607526, "grad_norm": 1.366355299949646, "learning_rate": 1.8123067773722336e-05, "loss": 0.0596, "step": 146810 }, { "epoch": 4.963656648297778, "grad_norm": 1.5362412929534912, "learning_rate": 1.811669889183027e-05, "loss": 0.0604, "step": 146820 }, { "epoch": 4.963994725988032, "grad_norm": 0.8571761846542358, "learning_rate": 1.811033088161566e-05, "loss": 0.0533, "step": 146830 }, { "epoch": 4.964332803678285, "grad_norm": 1.2937510013580322, "learning_rate": 1.810396374325265e-05, "loss": 0.0674, "step": 146840 }, { "epoch": 4.964670881368539, "grad_norm": 3.3847968578338623, "learning_rate": 1.8097597476915263e-05, "loss": 0.0569, "step": 146850 }, { "epoch": 4.965008959058792, "grad_norm": 0.8486058712005615, "learning_rate": 1.8091232082777597e-05, "loss": 0.0596, "step": 146860 }, { "epoch": 4.965347036749045, "grad_norm": 2.003005266189575, "learning_rate": 1.8084867561013658e-05, "loss": 0.0643, "step": 146870 }, { "epoch": 4.965685114439298, "grad_norm": 1.2631844282150269, "learning_rate": 1.807850391179743e-05, "loss": 0.0521, "step": 146880 }, { "epoch": 4.966023192129551, "grad_norm": 1.6257630586624146, "learning_rate": 1.8072141135302927e-05, "loss": 0.0549, "step": 146890 }, { "epoch": 4.966361269819805, "grad_norm": 0.9532197713851929, "learning_rate": 1.80657792317041e-05, "loss": 0.0589, "step": 146900 }, { "epoch": 4.966699347510058, "grad_norm": 2.075819730758667, "learning_rate": 1.805941820117485e-05, "loss": 0.0568, "step": 146910 }, { "epoch": 4.967037425200311, "grad_norm": 1.1557062864303589, "learning_rate": 1.805305804388912e-05, "loss": 0.0538, "step": 146920 }, { "epoch": 4.967375502890564, "grad_norm": 0.8236413598060608, "learning_rate": 1.804669876002077e-05, "loss": 0.0604, "step": 146930 }, { "epoch": 4.967713580580817, "grad_norm": 1.0387771129608154, "learning_rate": 1.8040340349743702e-05, "loss": 0.0558, "step": 146940 }, { "epoch": 4.9680516582710705, "grad_norm": 1.4286656379699707, "learning_rate": 1.8033982813231685e-05, "loss": 0.062, "step": 146950 }, { "epoch": 4.968389735961324, "grad_norm": 1.532679796218872, "learning_rate": 1.8027626150658594e-05, "loss": 0.0506, "step": 146960 }, { "epoch": 4.968727813651578, "grad_norm": 3.292067766189575, "learning_rate": 1.8021270362198174e-05, "loss": 0.0555, "step": 146970 }, { "epoch": 4.96906589134183, "grad_norm": 1.059385895729065, "learning_rate": 1.8014915448024224e-05, "loss": 0.0559, "step": 146980 }, { "epoch": 4.969403969032084, "grad_norm": 0.9336453080177307, "learning_rate": 1.8008561408310466e-05, "loss": 0.0579, "step": 146990 }, { "epoch": 4.969742046722336, "grad_norm": 0.8123382925987244, "learning_rate": 1.8002208243230607e-05, "loss": 0.0509, "step": 147000 }, { "epoch": 4.97008012441259, "grad_norm": 1.197586178779602, "learning_rate": 1.799585595295837e-05, "loss": 0.056, "step": 147010 }, { "epoch": 4.970418202102843, "grad_norm": 3.061798334121704, "learning_rate": 1.7989504537667406e-05, "loss": 0.0566, "step": 147020 }, { "epoch": 4.970756279793097, "grad_norm": 0.7378565669059753, "learning_rate": 1.7983153997531365e-05, "loss": 0.0544, "step": 147030 }, { "epoch": 4.9710943574833495, "grad_norm": 1.1966841220855713, "learning_rate": 1.7976804332723847e-05, "loss": 0.0517, "step": 147040 }, { "epoch": 4.971432435173603, "grad_norm": 1.6921075582504272, "learning_rate": 1.7970455543418467e-05, "loss": 0.0532, "step": 147050 }, { "epoch": 4.971770512863856, "grad_norm": 1.8736145496368408, "learning_rate": 1.796410762978882e-05, "loss": 0.0568, "step": 147060 }, { "epoch": 4.972108590554109, "grad_norm": 1.253579020500183, "learning_rate": 1.7957760592008434e-05, "loss": 0.0522, "step": 147070 }, { "epoch": 4.972446668244363, "grad_norm": 2.639291524887085, "learning_rate": 1.7951414430250833e-05, "loss": 0.0564, "step": 147080 }, { "epoch": 4.972784745934616, "grad_norm": 0.9463704824447632, "learning_rate": 1.79450691446895e-05, "loss": 0.054, "step": 147090 }, { "epoch": 4.973122823624869, "grad_norm": 1.0834461450576782, "learning_rate": 1.7938724735497963e-05, "loss": 0.0563, "step": 147100 }, { "epoch": 4.973460901315122, "grad_norm": 1.164165735244751, "learning_rate": 1.793238120284964e-05, "loss": 0.0528, "step": 147110 }, { "epoch": 4.973798979005376, "grad_norm": 1.7046345472335815, "learning_rate": 1.792603854691795e-05, "loss": 0.056, "step": 147120 }, { "epoch": 4.9741370566956284, "grad_norm": 0.7583950757980347, "learning_rate": 1.7919696767876336e-05, "loss": 0.0544, "step": 147130 }, { "epoch": 4.974475134385882, "grad_norm": 1.0411866903305054, "learning_rate": 1.791335586589814e-05, "loss": 0.053, "step": 147140 }, { "epoch": 4.9748132120761355, "grad_norm": 0.8518784046173096, "learning_rate": 1.7907015841156782e-05, "loss": 0.0586, "step": 147150 }, { "epoch": 4.975151289766388, "grad_norm": 0.7199668288230896, "learning_rate": 1.7900676693825515e-05, "loss": 0.0607, "step": 147160 }, { "epoch": 4.975489367456642, "grad_norm": 0.6498321890830994, "learning_rate": 1.7894338424077683e-05, "loss": 0.0512, "step": 147170 }, { "epoch": 4.975827445146895, "grad_norm": 1.4584691524505615, "learning_rate": 1.78880010320866e-05, "loss": 0.057, "step": 147180 }, { "epoch": 4.976165522837148, "grad_norm": 1.4022812843322754, "learning_rate": 1.7881664518025503e-05, "loss": 0.0526, "step": 147190 }, { "epoch": 4.976503600527401, "grad_norm": 0.5819947123527527, "learning_rate": 1.7875328882067634e-05, "loss": 0.0544, "step": 147200 }, { "epoch": 4.976841678217655, "grad_norm": 1.2263147830963135, "learning_rate": 1.7868994124386183e-05, "loss": 0.0647, "step": 147210 }, { "epoch": 4.977179755907907, "grad_norm": 2.890615463256836, "learning_rate": 1.7862660245154383e-05, "loss": 0.0594, "step": 147220 }, { "epoch": 4.977517833598161, "grad_norm": 2.29248046875, "learning_rate": 1.785632724454538e-05, "loss": 0.0566, "step": 147230 }, { "epoch": 4.977855911288414, "grad_norm": 1.871482014656067, "learning_rate": 1.7849995122732292e-05, "loss": 0.0562, "step": 147240 }, { "epoch": 4.978193988978667, "grad_norm": 2.728611707687378, "learning_rate": 1.784366387988828e-05, "loss": 0.0554, "step": 147250 }, { "epoch": 4.978532066668921, "grad_norm": 0.9457591772079468, "learning_rate": 1.78373335161864e-05, "loss": 0.0552, "step": 147260 }, { "epoch": 4.978870144359174, "grad_norm": 0.8495649695396423, "learning_rate": 1.783100403179975e-05, "loss": 0.0529, "step": 147270 }, { "epoch": 4.979208222049427, "grad_norm": 0.8858650326728821, "learning_rate": 1.782467542690137e-05, "loss": 0.0545, "step": 147280 }, { "epoch": 4.97954629973968, "grad_norm": 1.137728214263916, "learning_rate": 1.7818347701664262e-05, "loss": 0.0618, "step": 147290 }, { "epoch": 4.979884377429934, "grad_norm": 0.5589549541473389, "learning_rate": 1.7812020856261445e-05, "loss": 0.061, "step": 147300 }, { "epoch": 4.980222455120186, "grad_norm": 0.9561123251914978, "learning_rate": 1.7805694890865894e-05, "loss": 0.0593, "step": 147310 }, { "epoch": 4.98056053281044, "grad_norm": 1.8133399486541748, "learning_rate": 1.7799369805650546e-05, "loss": 0.0521, "step": 147320 }, { "epoch": 4.980898610500693, "grad_norm": 1.0414122343063354, "learning_rate": 1.7793045600788317e-05, "loss": 0.0583, "step": 147330 }, { "epoch": 4.981236688190946, "grad_norm": 1.2870948314666748, "learning_rate": 1.7786722276452134e-05, "loss": 0.0578, "step": 147340 }, { "epoch": 4.9815747658811995, "grad_norm": 1.1352571249008179, "learning_rate": 1.778039983281485e-05, "loss": 0.0543, "step": 147350 }, { "epoch": 4.981912843571453, "grad_norm": 0.9253603219985962, "learning_rate": 1.777407827004934e-05, "loss": 0.0538, "step": 147360 }, { "epoch": 4.982250921261706, "grad_norm": 2.4285507202148438, "learning_rate": 1.7767757588328426e-05, "loss": 0.0558, "step": 147370 }, { "epoch": 4.982588998951959, "grad_norm": 2.6370880603790283, "learning_rate": 1.7761437787824897e-05, "loss": 0.0616, "step": 147380 }, { "epoch": 4.982927076642213, "grad_norm": 1.4020835161209106, "learning_rate": 1.775511886871156e-05, "loss": 0.0562, "step": 147390 }, { "epoch": 4.983265154332465, "grad_norm": 0.9898300766944885, "learning_rate": 1.774880083116116e-05, "loss": 0.0555, "step": 147400 }, { "epoch": 4.983603232022719, "grad_norm": 3.276685953140259, "learning_rate": 1.774248367534643e-05, "loss": 0.0595, "step": 147410 }, { "epoch": 4.983941309712972, "grad_norm": 0.999132513999939, "learning_rate": 1.773616740144007e-05, "loss": 0.0571, "step": 147420 }, { "epoch": 4.984279387403225, "grad_norm": 3.06105637550354, "learning_rate": 1.7729852009614762e-05, "loss": 0.0642, "step": 147430 }, { "epoch": 4.9846174650934785, "grad_norm": 2.9465601444244385, "learning_rate": 1.772353750004321e-05, "loss": 0.0533, "step": 147440 }, { "epoch": 4.984955542783732, "grad_norm": 0.9507155418395996, "learning_rate": 1.7717223872897986e-05, "loss": 0.0509, "step": 147450 }, { "epoch": 4.985293620473985, "grad_norm": 3.665579080581665, "learning_rate": 1.771091112835175e-05, "loss": 0.0591, "step": 147460 }, { "epoch": 4.985631698164238, "grad_norm": 0.6640101671218872, "learning_rate": 1.7704599266577054e-05, "loss": 0.0528, "step": 147470 }, { "epoch": 4.985969775854492, "grad_norm": 2.3113081455230713, "learning_rate": 1.76982882877465e-05, "loss": 0.0534, "step": 147480 }, { "epoch": 4.986307853544744, "grad_norm": 1.9997419118881226, "learning_rate": 1.769197819203261e-05, "loss": 0.0536, "step": 147490 }, { "epoch": 4.986645931234998, "grad_norm": 3.0153496265411377, "learning_rate": 1.768566897960788e-05, "loss": 0.0604, "step": 147500 }, { "epoch": 4.986984008925251, "grad_norm": 1.5717346668243408, "learning_rate": 1.7679360650644838e-05, "loss": 0.0506, "step": 147510 }, { "epoch": 4.987322086615504, "grad_norm": 0.5814973711967468, "learning_rate": 1.767305320531593e-05, "loss": 0.0536, "step": 147520 }, { "epoch": 4.987660164305757, "grad_norm": 4.00648307800293, "learning_rate": 1.766674664379361e-05, "loss": 0.0522, "step": 147530 }, { "epoch": 4.987998241996011, "grad_norm": 2.9259092807769775, "learning_rate": 1.766044096625027e-05, "loss": 0.0538, "step": 147540 }, { "epoch": 4.988336319686264, "grad_norm": 1.3377474546432495, "learning_rate": 1.7654136172858323e-05, "loss": 0.0512, "step": 147550 }, { "epoch": 4.988674397376517, "grad_norm": 2.9191997051239014, "learning_rate": 1.7647832263790158e-05, "loss": 0.057, "step": 147560 }, { "epoch": 4.989012475066771, "grad_norm": 10.856626510620117, "learning_rate": 1.7641529239218106e-05, "loss": 0.0606, "step": 147570 }, { "epoch": 4.989350552757023, "grad_norm": 13.221273422241211, "learning_rate": 1.763522709931449e-05, "loss": 0.0558, "step": 147580 }, { "epoch": 4.989688630447277, "grad_norm": 1.4982151985168457, "learning_rate": 1.762892584425158e-05, "loss": 0.0567, "step": 147590 }, { "epoch": 4.99002670813753, "grad_norm": 1.0621399879455566, "learning_rate": 1.7622625474201703e-05, "loss": 0.0574, "step": 147600 }, { "epoch": 4.990364785827783, "grad_norm": 3.3442158699035645, "learning_rate": 1.761632598933708e-05, "loss": 0.059, "step": 147610 }, { "epoch": 4.990702863518036, "grad_norm": 0.8127245306968689, "learning_rate": 1.761002738982992e-05, "loss": 0.0538, "step": 147620 }, { "epoch": 4.99104094120829, "grad_norm": 6.032869815826416, "learning_rate": 1.760372967585246e-05, "loss": 0.0517, "step": 147630 }, { "epoch": 4.9913790188985425, "grad_norm": 1.3289401531219482, "learning_rate": 1.7597432847576844e-05, "loss": 0.0568, "step": 147640 }, { "epoch": 4.991717096588796, "grad_norm": 5.570511341094971, "learning_rate": 1.7591136905175278e-05, "loss": 0.0576, "step": 147650 }, { "epoch": 4.9920551742790495, "grad_norm": 1.149009346961975, "learning_rate": 1.758484184881981e-05, "loss": 0.056, "step": 147660 }, { "epoch": 4.992393251969302, "grad_norm": 0.8209218382835388, "learning_rate": 1.7578547678682594e-05, "loss": 0.0557, "step": 147670 }, { "epoch": 4.992731329659556, "grad_norm": 1.425983190536499, "learning_rate": 1.757225439493572e-05, "loss": 0.0544, "step": 147680 }, { "epoch": 4.993069407349809, "grad_norm": 1.5071336030960083, "learning_rate": 1.756596199775123e-05, "loss": 0.0586, "step": 147690 }, { "epoch": 4.993407485040062, "grad_norm": 3.2090933322906494, "learning_rate": 1.7559670487301157e-05, "loss": 0.0497, "step": 147700 }, { "epoch": 4.993745562730315, "grad_norm": 4.5493950843811035, "learning_rate": 1.7553379863757487e-05, "loss": 0.0583, "step": 147710 }, { "epoch": 4.994083640420569, "grad_norm": 1.810114860534668, "learning_rate": 1.7547090127292233e-05, "loss": 0.0559, "step": 147720 }, { "epoch": 4.9944217181108215, "grad_norm": 1.8153077363967896, "learning_rate": 1.7540801278077345e-05, "loss": 0.0518, "step": 147730 }, { "epoch": 4.994759795801075, "grad_norm": 1.009922742843628, "learning_rate": 1.753451331628474e-05, "loss": 0.0608, "step": 147740 }, { "epoch": 4.9950978734913285, "grad_norm": 1.1406335830688477, "learning_rate": 1.7528226242086364e-05, "loss": 0.0568, "step": 147750 }, { "epoch": 4.995435951181582, "grad_norm": 2.875593423843384, "learning_rate": 1.752194005565406e-05, "loss": 0.0566, "step": 147760 }, { "epoch": 4.995774028871835, "grad_norm": 1.1221063137054443, "learning_rate": 1.751565475715974e-05, "loss": 0.0609, "step": 147770 }, { "epoch": 4.996112106562088, "grad_norm": 3.729917049407959, "learning_rate": 1.7509370346775207e-05, "loss": 0.0521, "step": 147780 }, { "epoch": 4.996450184252341, "grad_norm": 1.562198281288147, "learning_rate": 1.7503086824672287e-05, "loss": 0.0532, "step": 147790 }, { "epoch": 4.996788261942594, "grad_norm": 1.5327262878417969, "learning_rate": 1.7496804191022747e-05, "loss": 0.0611, "step": 147800 }, { "epoch": 4.997126339632848, "grad_norm": 0.8901724219322205, "learning_rate": 1.7490522445998363e-05, "loss": 0.0558, "step": 147810 }, { "epoch": 4.997464417323101, "grad_norm": 1.0873987674713135, "learning_rate": 1.7484241589770928e-05, "loss": 0.0534, "step": 147820 }, { "epoch": 4.997802495013354, "grad_norm": 3.673447370529175, "learning_rate": 1.7477961622512066e-05, "loss": 0.0608, "step": 147830 }, { "epoch": 4.9981405727036075, "grad_norm": 2.04459285736084, "learning_rate": 1.7471682544393542e-05, "loss": 0.056, "step": 147840 }, { "epoch": 4.99847865039386, "grad_norm": 1.2926775217056274, "learning_rate": 1.7465404355586974e-05, "loss": 0.0553, "step": 147850 }, { "epoch": 4.998816728084114, "grad_norm": 0.9814385175704956, "learning_rate": 1.7459127056264046e-05, "loss": 0.0522, "step": 147860 }, { "epoch": 4.999154805774367, "grad_norm": 1.776982069015503, "learning_rate": 1.7452850646596363e-05, "loss": 0.052, "step": 147870 }, { "epoch": 4.999492883464621, "grad_norm": 4.470863342285156, "learning_rate": 1.7446575126755497e-05, "loss": 0.0529, "step": 147880 }, { "epoch": 4.999830961154873, "grad_norm": 1.662373423576355, "learning_rate": 1.7440300496913054e-05, "loss": 0.0574, "step": 147890 }, { "epoch": 5.000169038845127, "grad_norm": 2.9186275005340576, "learning_rate": 1.7434026757240564e-05, "loss": 0.058, "step": 147900 }, { "epoch": 5.00050711653538, "grad_norm": 56.20478820800781, "learning_rate": 1.7427753907909545e-05, "loss": 0.0543, "step": 147910 }, { "epoch": 5.000845194225633, "grad_norm": 1.5234259366989136, "learning_rate": 1.7421481949091483e-05, "loss": 0.0518, "step": 147920 }, { "epoch": 5.001183271915886, "grad_norm": 2.0028295516967773, "learning_rate": 1.7415210880957867e-05, "loss": 0.0541, "step": 147930 }, { "epoch": 5.00152134960614, "grad_norm": 1.00912344455719, "learning_rate": 1.7408940703680154e-05, "loss": 0.0581, "step": 147940 }, { "epoch": 5.0018594272963925, "grad_norm": 1.5730777978897095, "learning_rate": 1.7402671417429765e-05, "loss": 0.0532, "step": 147950 }, { "epoch": 5.002197504986646, "grad_norm": 1.6069891452789307, "learning_rate": 1.7396403022378095e-05, "loss": 0.0568, "step": 147960 }, { "epoch": 5.0025355826769, "grad_norm": 1.499285340309143, "learning_rate": 1.7390135518696492e-05, "loss": 0.0554, "step": 147970 }, { "epoch": 5.002873660367152, "grad_norm": 1.543206810951233, "learning_rate": 1.7383868906556355e-05, "loss": 0.0527, "step": 147980 }, { "epoch": 5.003211738057406, "grad_norm": 1.8861908912658691, "learning_rate": 1.7377603186128988e-05, "loss": 0.0575, "step": 147990 }, { "epoch": 5.003549815747659, "grad_norm": 6.6229329109191895, "learning_rate": 1.737133835758568e-05, "loss": 0.0609, "step": 148000 }, { "epoch": 5.003887893437912, "grad_norm": 0.870597243309021, "learning_rate": 1.7365074421097737e-05, "loss": 0.0442, "step": 148010 }, { "epoch": 5.004225971128165, "grad_norm": 0.8494949340820312, "learning_rate": 1.7358811376836387e-05, "loss": 0.054, "step": 148020 }, { "epoch": 5.004564048818419, "grad_norm": 2.1966819763183594, "learning_rate": 1.7352549224972897e-05, "loss": 0.0546, "step": 148030 }, { "epoch": 5.0049021265086715, "grad_norm": 1.0610116720199585, "learning_rate": 1.734628796567842e-05, "loss": 0.0472, "step": 148040 }, { "epoch": 5.005240204198925, "grad_norm": 12.291837692260742, "learning_rate": 1.7340027599124175e-05, "loss": 0.0505, "step": 148050 }, { "epoch": 5.0055782818891785, "grad_norm": 1.342314600944519, "learning_rate": 1.7333768125481293e-05, "loss": 0.0525, "step": 148060 }, { "epoch": 5.005916359579431, "grad_norm": 3.721796989440918, "learning_rate": 1.7327509544920928e-05, "loss": 0.0563, "step": 148070 }, { "epoch": 5.006254437269685, "grad_norm": 5.573915958404541, "learning_rate": 1.7321251857614183e-05, "loss": 0.0526, "step": 148080 }, { "epoch": 5.006592514959938, "grad_norm": 1.4944769144058228, "learning_rate": 1.7314995063732113e-05, "loss": 0.0556, "step": 148090 }, { "epoch": 5.006930592650191, "grad_norm": 1.7705368995666504, "learning_rate": 1.7308739163445825e-05, "loss": 0.062, "step": 148100 }, { "epoch": 5.007268670340444, "grad_norm": 1.3228733539581299, "learning_rate": 1.730248415692632e-05, "loss": 0.0527, "step": 148110 }, { "epoch": 5.007606748030698, "grad_norm": 6.141819953918457, "learning_rate": 1.7296230044344598e-05, "loss": 0.0587, "step": 148120 }, { "epoch": 5.0079448257209505, "grad_norm": 1.8899974822998047, "learning_rate": 1.728997682587168e-05, "loss": 0.0556, "step": 148130 }, { "epoch": 5.008282903411204, "grad_norm": 0.940069854259491, "learning_rate": 1.728372450167849e-05, "loss": 0.0547, "step": 148140 }, { "epoch": 5.0086209811014575, "grad_norm": 1.7737746238708496, "learning_rate": 1.7277473071935996e-05, "loss": 0.0563, "step": 148150 }, { "epoch": 5.00895905879171, "grad_norm": 2.3123342990875244, "learning_rate": 1.727122253681509e-05, "loss": 0.0545, "step": 148160 }, { "epoch": 5.009297136481964, "grad_norm": 1.1579118967056274, "learning_rate": 1.726497289648667e-05, "loss": 0.0545, "step": 148170 }, { "epoch": 5.009635214172217, "grad_norm": 3.5666861534118652, "learning_rate": 1.725872415112158e-05, "loss": 0.0558, "step": 148180 }, { "epoch": 5.00997329186247, "grad_norm": 1.5764034986495972, "learning_rate": 1.725247630089068e-05, "loss": 0.0557, "step": 148190 }, { "epoch": 5.010311369552723, "grad_norm": 4.205766677856445, "learning_rate": 1.724622934596478e-05, "loss": 0.0561, "step": 148200 }, { "epoch": 5.010649447242977, "grad_norm": 6.1412272453308105, "learning_rate": 1.723998328651465e-05, "loss": 0.0599, "step": 148210 }, { "epoch": 5.010987524933229, "grad_norm": 1.6811915636062622, "learning_rate": 1.723373812271108e-05, "loss": 0.0479, "step": 148220 }, { "epoch": 5.011325602623483, "grad_norm": 1.0054007768630981, "learning_rate": 1.7227493854724784e-05, "loss": 0.0546, "step": 148230 }, { "epoch": 5.011663680313736, "grad_norm": 1.6800742149353027, "learning_rate": 1.722125048272653e-05, "loss": 0.0589, "step": 148240 }, { "epoch": 5.012001758003989, "grad_norm": 0.9811999201774597, "learning_rate": 1.7215008006886946e-05, "loss": 0.0557, "step": 148250 }, { "epoch": 5.012339835694243, "grad_norm": 4.557526588439941, "learning_rate": 1.7208766427376715e-05, "loss": 0.0582, "step": 148260 }, { "epoch": 5.012677913384496, "grad_norm": 2.7051289081573486, "learning_rate": 1.7202525744366515e-05, "loss": 0.0603, "step": 148270 }, { "epoch": 5.013015991074749, "grad_norm": 1.5353885889053345, "learning_rate": 1.7196285958026937e-05, "loss": 0.0505, "step": 148280 }, { "epoch": 5.013354068765002, "grad_norm": 1.541609287261963, "learning_rate": 1.7190047068528582e-05, "loss": 0.0566, "step": 148290 }, { "epoch": 5.013692146455256, "grad_norm": 3.9376204013824463, "learning_rate": 1.718380907604199e-05, "loss": 0.0578, "step": 148300 }, { "epoch": 5.014030224145508, "grad_norm": 1.4772729873657227, "learning_rate": 1.7177571980737732e-05, "loss": 0.0519, "step": 148310 }, { "epoch": 5.014368301835762, "grad_norm": 1.235614538192749, "learning_rate": 1.7171335782786357e-05, "loss": 0.0532, "step": 148320 }, { "epoch": 5.014706379526015, "grad_norm": 2.2254345417022705, "learning_rate": 1.71651004823583e-05, "loss": 0.0587, "step": 148330 }, { "epoch": 5.015044457216268, "grad_norm": 2.2888152599334717, "learning_rate": 1.7158866079624075e-05, "loss": 0.0515, "step": 148340 }, { "epoch": 5.0153825349065215, "grad_norm": 2.018404245376587, "learning_rate": 1.7152632574754094e-05, "loss": 0.0607, "step": 148350 }, { "epoch": 5.015720612596775, "grad_norm": 2.3193657398223877, "learning_rate": 1.714639996791882e-05, "loss": 0.0582, "step": 148360 }, { "epoch": 5.016058690287028, "grad_norm": 0.5995416641235352, "learning_rate": 1.7140168259288625e-05, "loss": 0.0524, "step": 148370 }, { "epoch": 5.016396767977281, "grad_norm": 1.3676129579544067, "learning_rate": 1.7133937449033866e-05, "loss": 0.0581, "step": 148380 }, { "epoch": 5.016734845667535, "grad_norm": 1.275436282157898, "learning_rate": 1.7127707537324926e-05, "loss": 0.052, "step": 148390 }, { "epoch": 5.017072923357787, "grad_norm": 2.7229020595550537, "learning_rate": 1.712147852433212e-05, "loss": 0.0547, "step": 148400 }, { "epoch": 5.017411001048041, "grad_norm": 1.4065146446228027, "learning_rate": 1.711525041022573e-05, "loss": 0.0513, "step": 148410 }, { "epoch": 5.017749078738294, "grad_norm": 1.1051232814788818, "learning_rate": 1.7109023195176033e-05, "loss": 0.0515, "step": 148420 }, { "epoch": 5.018087156428547, "grad_norm": 1.8561971187591553, "learning_rate": 1.7102796879353296e-05, "loss": 0.0548, "step": 148430 }, { "epoch": 5.0184252341188005, "grad_norm": 1.3375059366226196, "learning_rate": 1.7096571462927713e-05, "loss": 0.06, "step": 148440 }, { "epoch": 5.018763311809054, "grad_norm": 1.1882597208023071, "learning_rate": 1.709034694606953e-05, "loss": 0.0556, "step": 148450 }, { "epoch": 5.019101389499307, "grad_norm": 2.0836596488952637, "learning_rate": 1.708412332894889e-05, "loss": 0.053, "step": 148460 }, { "epoch": 5.01943946718956, "grad_norm": 4.299455165863037, "learning_rate": 1.707790061173594e-05, "loss": 0.0601, "step": 148470 }, { "epoch": 5.019777544879814, "grad_norm": 1.80311918258667, "learning_rate": 1.7071678794600836e-05, "loss": 0.0561, "step": 148480 }, { "epoch": 5.020115622570066, "grad_norm": 3.545386552810669, "learning_rate": 1.7065457877713663e-05, "loss": 0.0549, "step": 148490 }, { "epoch": 5.02045370026032, "grad_norm": 2.1218292713165283, "learning_rate": 1.7059237861244482e-05, "loss": 0.0573, "step": 148500 }, { "epoch": 5.020791777950573, "grad_norm": 1.986930251121521, "learning_rate": 1.705301874536338e-05, "loss": 0.0553, "step": 148510 }, { "epoch": 5.021129855640826, "grad_norm": 2.0830888748168945, "learning_rate": 1.704680053024035e-05, "loss": 0.0541, "step": 148520 }, { "epoch": 5.021467933331079, "grad_norm": 0.7605462074279785, "learning_rate": 1.7040583216045448e-05, "loss": 0.0534, "step": 148530 }, { "epoch": 5.021806011021333, "grad_norm": 1.4882643222808838, "learning_rate": 1.703436680294858e-05, "loss": 0.0526, "step": 148540 }, { "epoch": 5.022144088711586, "grad_norm": 1.1249927282333374, "learning_rate": 1.702815129111976e-05, "loss": 0.0523, "step": 148550 }, { "epoch": 5.022482166401839, "grad_norm": 0.8323595523834229, "learning_rate": 1.702193668072888e-05, "loss": 0.0545, "step": 148560 }, { "epoch": 5.022820244092093, "grad_norm": 1.0815917253494263, "learning_rate": 1.701572297194588e-05, "loss": 0.0593, "step": 148570 }, { "epoch": 5.023158321782345, "grad_norm": 1.0058631896972656, "learning_rate": 1.7009510164940616e-05, "loss": 0.0616, "step": 148580 }, { "epoch": 5.023496399472599, "grad_norm": 3.6679210662841797, "learning_rate": 1.7003298259882944e-05, "loss": 0.0624, "step": 148590 }, { "epoch": 5.023834477162852, "grad_norm": 2.7038097381591797, "learning_rate": 1.6997087256942706e-05, "loss": 0.0542, "step": 148600 }, { "epoch": 5.024172554853105, "grad_norm": 1.6303244829177856, "learning_rate": 1.6990877156289712e-05, "loss": 0.0519, "step": 148610 }, { "epoch": 5.024510632543358, "grad_norm": 1.7733445167541504, "learning_rate": 1.6984667958093735e-05, "loss": 0.0563, "step": 148620 }, { "epoch": 5.024848710233612, "grad_norm": 1.2803657054901123, "learning_rate": 1.6978459662524514e-05, "loss": 0.0509, "step": 148630 }, { "epoch": 5.0251867879238645, "grad_norm": 0.9887422323226929, "learning_rate": 1.69722522697518e-05, "loss": 0.0587, "step": 148640 }, { "epoch": 5.025524865614118, "grad_norm": 0.9091978073120117, "learning_rate": 1.6966045779945315e-05, "loss": 0.0597, "step": 148650 }, { "epoch": 5.0258629433043716, "grad_norm": 1.905074119567871, "learning_rate": 1.6959840193274734e-05, "loss": 0.0621, "step": 148660 }, { "epoch": 5.026201020994624, "grad_norm": 2.585204601287842, "learning_rate": 1.695363550990971e-05, "loss": 0.0502, "step": 148670 }, { "epoch": 5.026539098684878, "grad_norm": 1.7922219038009644, "learning_rate": 1.6947431730019853e-05, "loss": 0.0647, "step": 148680 }, { "epoch": 5.026877176375131, "grad_norm": 2.2362380027770996, "learning_rate": 1.6941228853774814e-05, "loss": 0.0536, "step": 148690 }, { "epoch": 5.027215254065384, "grad_norm": 1.5935004949569702, "learning_rate": 1.6935026881344162e-05, "loss": 0.0565, "step": 148700 }, { "epoch": 5.027553331755637, "grad_norm": 7.972198963165283, "learning_rate": 1.6928825812897436e-05, "loss": 0.0543, "step": 148710 }, { "epoch": 5.027891409445891, "grad_norm": 1.7155064344406128, "learning_rate": 1.69226256486042e-05, "loss": 0.0591, "step": 148720 }, { "epoch": 5.0282294871361435, "grad_norm": 0.7581393122673035, "learning_rate": 1.6916426388633943e-05, "loss": 0.063, "step": 148730 }, { "epoch": 5.028567564826397, "grad_norm": 2.1795458793640137, "learning_rate": 1.691022803315619e-05, "loss": 0.0552, "step": 148740 }, { "epoch": 5.0289056425166505, "grad_norm": 2.469700574874878, "learning_rate": 1.6904030582340337e-05, "loss": 0.0588, "step": 148750 }, { "epoch": 5.029243720206903, "grad_norm": 0.6432977318763733, "learning_rate": 1.6897834036355846e-05, "loss": 0.0564, "step": 148760 }, { "epoch": 5.029581797897157, "grad_norm": 2.0281035900115967, "learning_rate": 1.6891638395372167e-05, "loss": 0.0588, "step": 148770 }, { "epoch": 5.02991987558741, "grad_norm": 6.884765148162842, "learning_rate": 1.6885443659558647e-05, "loss": 0.0543, "step": 148780 }, { "epoch": 5.030257953277663, "grad_norm": 1.0394668579101562, "learning_rate": 1.6879249829084658e-05, "loss": 0.0607, "step": 148790 }, { "epoch": 5.030596030967916, "grad_norm": 2.161729574203491, "learning_rate": 1.687305690411952e-05, "loss": 0.0576, "step": 148800 }, { "epoch": 5.03093410865817, "grad_norm": 1.4738472700119019, "learning_rate": 1.6866864884832583e-05, "loss": 0.0593, "step": 148810 }, { "epoch": 5.031272186348422, "grad_norm": 1.2576942443847656, "learning_rate": 1.686067377139311e-05, "loss": 0.0607, "step": 148820 }, { "epoch": 5.031610264038676, "grad_norm": 1.91702139377594, "learning_rate": 1.6854483563970348e-05, "loss": 0.0594, "step": 148830 }, { "epoch": 5.0319483417289295, "grad_norm": 3.4042959213256836, "learning_rate": 1.6848294262733567e-05, "loss": 0.0518, "step": 148840 }, { "epoch": 5.032286419419182, "grad_norm": 2.1377322673797607, "learning_rate": 1.6842105867851953e-05, "loss": 0.055, "step": 148850 }, { "epoch": 5.032624497109436, "grad_norm": 1.194929838180542, "learning_rate": 1.6835918379494726e-05, "loss": 0.0552, "step": 148860 }, { "epoch": 5.032962574799689, "grad_norm": 3.1303420066833496, "learning_rate": 1.6829731797831027e-05, "loss": 0.0683, "step": 148870 }, { "epoch": 5.033300652489943, "grad_norm": 5.711938858032227, "learning_rate": 1.6823546123029982e-05, "loss": 0.0574, "step": 148880 }, { "epoch": 5.033638730180195, "grad_norm": 1.295685887336731, "learning_rate": 1.6817361355260742e-05, "loss": 0.0557, "step": 148890 }, { "epoch": 5.033976807870449, "grad_norm": 1.0176191329956055, "learning_rate": 1.6811177494692377e-05, "loss": 0.0548, "step": 148900 }, { "epoch": 5.034314885560702, "grad_norm": 8.799394607543945, "learning_rate": 1.6804994541493953e-05, "loss": 0.0558, "step": 148910 }, { "epoch": 5.034652963250955, "grad_norm": 1.4291050434112549, "learning_rate": 1.6798812495834482e-05, "loss": 0.0554, "step": 148920 }, { "epoch": 5.034991040941208, "grad_norm": 2.2538201808929443, "learning_rate": 1.6792631357883025e-05, "loss": 0.0615, "step": 148930 }, { "epoch": 5.035329118631462, "grad_norm": 13.188941955566406, "learning_rate": 1.678645112780854e-05, "loss": 0.0487, "step": 148940 }, { "epoch": 5.035667196321715, "grad_norm": 2.589240550994873, "learning_rate": 1.6780271805780013e-05, "loss": 0.0569, "step": 148950 }, { "epoch": 5.036005274011968, "grad_norm": 0.7264499664306641, "learning_rate": 1.6774093391966377e-05, "loss": 0.0575, "step": 148960 }, { "epoch": 5.036343351702222, "grad_norm": 22.0365047454834, "learning_rate": 1.6767915886536535e-05, "loss": 0.0566, "step": 148970 }, { "epoch": 5.036681429392474, "grad_norm": 6.93467903137207, "learning_rate": 1.6761739289659395e-05, "loss": 0.0578, "step": 148980 }, { "epoch": 5.037019507082728, "grad_norm": 0.8877121806144714, "learning_rate": 1.675556360150382e-05, "loss": 0.0597, "step": 148990 }, { "epoch": 5.037357584772981, "grad_norm": 5.424368381500244, "learning_rate": 1.674938882223865e-05, "loss": 0.059, "step": 149000 }, { "epoch": 5.037695662463234, "grad_norm": 2.0282046794891357, "learning_rate": 1.6743214952032677e-05, "loss": 0.0576, "step": 149010 }, { "epoch": 5.038033740153487, "grad_norm": 6.293201923370361, "learning_rate": 1.6737041991054713e-05, "loss": 0.0559, "step": 149020 }, { "epoch": 5.038371817843741, "grad_norm": 2.3081729412078857, "learning_rate": 1.673086993947356e-05, "loss": 0.0517, "step": 149030 }, { "epoch": 5.0387098955339935, "grad_norm": 1.2971216440200806, "learning_rate": 1.6724698797457888e-05, "loss": 0.0533, "step": 149040 }, { "epoch": 5.039047973224247, "grad_norm": 3.9409048557281494, "learning_rate": 1.6718528565176465e-05, "loss": 0.0557, "step": 149050 }, { "epoch": 5.0393860509145005, "grad_norm": 1.2787593603134155, "learning_rate": 1.671235924279796e-05, "loss": 0.0584, "step": 149060 }, { "epoch": 5.039724128604753, "grad_norm": 2.9838974475860596, "learning_rate": 1.6706190830491053e-05, "loss": 0.0496, "step": 149070 }, { "epoch": 5.040062206295007, "grad_norm": 7.972280025482178, "learning_rate": 1.670002332842439e-05, "loss": 0.0555, "step": 149080 }, { "epoch": 5.04040028398526, "grad_norm": 4.573690414428711, "learning_rate": 1.669385673676656e-05, "loss": 0.0547, "step": 149090 }, { "epoch": 5.040738361675513, "grad_norm": 1.011527180671692, "learning_rate": 1.6687691055686187e-05, "loss": 0.0583, "step": 149100 }, { "epoch": 5.041076439365766, "grad_norm": 5.112907886505127, "learning_rate": 1.6681526285351828e-05, "loss": 0.0539, "step": 149110 }, { "epoch": 5.04141451705602, "grad_norm": 1.4658639430999756, "learning_rate": 1.6675362425932028e-05, "loss": 0.0613, "step": 149120 }, { "epoch": 5.0417525947462725, "grad_norm": 1.2307811975479126, "learning_rate": 1.6669199477595275e-05, "loss": 0.0592, "step": 149130 }, { "epoch": 5.042090672436526, "grad_norm": 0.942038893699646, "learning_rate": 1.6663037440510093e-05, "loss": 0.0535, "step": 149140 }, { "epoch": 5.0424287501267795, "grad_norm": 1.8383392095565796, "learning_rate": 1.6656876314844956e-05, "loss": 0.0569, "step": 149150 }, { "epoch": 5.042766827817032, "grad_norm": 1.9900140762329102, "learning_rate": 1.6650716100768298e-05, "loss": 0.0585, "step": 149160 }, { "epoch": 5.043104905507286, "grad_norm": 2.151512861251831, "learning_rate": 1.664455679844853e-05, "loss": 0.0514, "step": 149170 }, { "epoch": 5.043442983197539, "grad_norm": 2.259124994277954, "learning_rate": 1.663839840805403e-05, "loss": 0.0568, "step": 149180 }, { "epoch": 5.043781060887792, "grad_norm": 6.950267791748047, "learning_rate": 1.66322409297532e-05, "loss": 0.0602, "step": 149190 }, { "epoch": 5.044119138578045, "grad_norm": 5.314477920532227, "learning_rate": 1.662608436371437e-05, "loss": 0.0515, "step": 149200 }, { "epoch": 5.044457216268299, "grad_norm": 1.575122594833374, "learning_rate": 1.6619928710105843e-05, "loss": 0.0562, "step": 149210 }, { "epoch": 5.044795293958551, "grad_norm": 0.8457520008087158, "learning_rate": 1.6613773969095937e-05, "loss": 0.0538, "step": 149220 }, { "epoch": 5.045133371648805, "grad_norm": 1.587524652481079, "learning_rate": 1.6607620140852893e-05, "loss": 0.0558, "step": 149230 }, { "epoch": 5.0454714493390584, "grad_norm": 3.3375179767608643, "learning_rate": 1.6601467225545e-05, "loss": 0.0533, "step": 149240 }, { "epoch": 5.045809527029311, "grad_norm": 3.925893783569336, "learning_rate": 1.6595315223340414e-05, "loss": 0.0554, "step": 149250 }, { "epoch": 5.046147604719565, "grad_norm": 3.703202724456787, "learning_rate": 1.6589164134407376e-05, "loss": 0.0552, "step": 149260 }, { "epoch": 5.046485682409818, "grad_norm": 4.142205238342285, "learning_rate": 1.6583013958914022e-05, "loss": 0.0611, "step": 149270 }, { "epoch": 5.046823760100071, "grad_norm": 2.815671920776367, "learning_rate": 1.6576864697028534e-05, "loss": 0.0571, "step": 149280 }, { "epoch": 5.047161837790324, "grad_norm": 2.9164466857910156, "learning_rate": 1.6570716348919e-05, "loss": 0.0563, "step": 149290 }, { "epoch": 5.047499915480578, "grad_norm": 1.596847414970398, "learning_rate": 1.6564568914753516e-05, "loss": 0.0564, "step": 149300 }, { "epoch": 5.04783799317083, "grad_norm": 5.822439670562744, "learning_rate": 1.6558422394700163e-05, "loss": 0.0496, "step": 149310 }, { "epoch": 5.048176070861084, "grad_norm": 1.1161353588104248, "learning_rate": 1.6552276788926984e-05, "loss": 0.0567, "step": 149320 }, { "epoch": 5.048514148551337, "grad_norm": 0.9981541633605957, "learning_rate": 1.6546132097601975e-05, "loss": 0.0576, "step": 149330 }, { "epoch": 5.04885222624159, "grad_norm": 0.9244236946105957, "learning_rate": 1.6539988320893158e-05, "loss": 0.0564, "step": 149340 }, { "epoch": 5.0491903039318435, "grad_norm": 1.178499698638916, "learning_rate": 1.653384545896848e-05, "loss": 0.0506, "step": 149350 }, { "epoch": 5.049528381622097, "grad_norm": 1.2656574249267578, "learning_rate": 1.6527703511995908e-05, "loss": 0.0541, "step": 149360 }, { "epoch": 5.04986645931235, "grad_norm": 1.1083083152770996, "learning_rate": 1.6521562480143347e-05, "loss": 0.0589, "step": 149370 }, { "epoch": 5.050204537002603, "grad_norm": 1.0852125883102417, "learning_rate": 1.6515422363578692e-05, "loss": 0.0535, "step": 149380 }, { "epoch": 5.050542614692857, "grad_norm": 3.7971837520599365, "learning_rate": 1.650928316246979e-05, "loss": 0.0583, "step": 149390 }, { "epoch": 5.050880692383109, "grad_norm": 5.130583763122559, "learning_rate": 1.6503144876984522e-05, "loss": 0.055, "step": 149400 }, { "epoch": 5.051218770073363, "grad_norm": 2.1986169815063477, "learning_rate": 1.649700750729069e-05, "loss": 0.0589, "step": 149410 }, { "epoch": 5.051556847763616, "grad_norm": 1.4677767753601074, "learning_rate": 1.6490871053556073e-05, "loss": 0.0552, "step": 149420 }, { "epoch": 5.051894925453869, "grad_norm": 6.009707927703857, "learning_rate": 1.6484735515948467e-05, "loss": 0.0542, "step": 149430 }, { "epoch": 5.0522330031441225, "grad_norm": 2.4209535121917725, "learning_rate": 1.647860089463558e-05, "loss": 0.0538, "step": 149440 }, { "epoch": 5.052571080834376, "grad_norm": 1.6262189149856567, "learning_rate": 1.647246718978519e-05, "loss": 0.0533, "step": 149450 }, { "epoch": 5.052909158524629, "grad_norm": 2.6671574115753174, "learning_rate": 1.6466334401564915e-05, "loss": 0.0611, "step": 149460 }, { "epoch": 5.053247236214882, "grad_norm": 7.095553874969482, "learning_rate": 1.646020253014246e-05, "loss": 0.059, "step": 149470 }, { "epoch": 5.053585313905136, "grad_norm": 1.7044925689697266, "learning_rate": 1.6454071575685488e-05, "loss": 0.0483, "step": 149480 }, { "epoch": 5.053923391595388, "grad_norm": 2.9545090198516846, "learning_rate": 1.6447941538361593e-05, "loss": 0.0515, "step": 149490 }, { "epoch": 5.054261469285642, "grad_norm": 2.4626779556274414, "learning_rate": 1.644181241833837e-05, "loss": 0.0533, "step": 149500 }, { "epoch": 5.054599546975895, "grad_norm": 1.664520263671875, "learning_rate": 1.6435684215783377e-05, "loss": 0.058, "step": 149510 }, { "epoch": 5.054937624666148, "grad_norm": 1.7895560264587402, "learning_rate": 1.642955693086416e-05, "loss": 0.058, "step": 149520 }, { "epoch": 5.0552757023564014, "grad_norm": 1.6588422060012817, "learning_rate": 1.6423430563748282e-05, "loss": 0.0592, "step": 149530 }, { "epoch": 5.055613780046655, "grad_norm": 1.4797883033752441, "learning_rate": 1.6417305114603165e-05, "loss": 0.058, "step": 149540 }, { "epoch": 5.055951857736908, "grad_norm": 1.4703712463378906, "learning_rate": 1.6411180583596324e-05, "loss": 0.0512, "step": 149550 }, { "epoch": 5.056289935427161, "grad_norm": 1.3184268474578857, "learning_rate": 1.6405056970895178e-05, "loss": 0.0591, "step": 149560 }, { "epoch": 5.056628013117415, "grad_norm": 3.7600977420806885, "learning_rate": 1.639893427666716e-05, "loss": 0.0618, "step": 149570 }, { "epoch": 5.056966090807667, "grad_norm": 4.273800849914551, "learning_rate": 1.6392812501079663e-05, "loss": 0.0545, "step": 149580 }, { "epoch": 5.057304168497921, "grad_norm": 1.4255657196044922, "learning_rate": 1.638669164430003e-05, "loss": 0.0566, "step": 149590 }, { "epoch": 5.057642246188174, "grad_norm": 2.6238527297973633, "learning_rate": 1.638057170649563e-05, "loss": 0.0559, "step": 149600 }, { "epoch": 5.057980323878427, "grad_norm": 3.847079038619995, "learning_rate": 1.6374452687833775e-05, "loss": 0.0537, "step": 149610 }, { "epoch": 5.05831840156868, "grad_norm": 6.967441558837891, "learning_rate": 1.6368334588481748e-05, "loss": 0.0498, "step": 149620 }, { "epoch": 5.058656479258934, "grad_norm": 1.58884859085083, "learning_rate": 1.63622174086068e-05, "loss": 0.0543, "step": 149630 }, { "epoch": 5.0589945569491865, "grad_norm": 6.577290058135986, "learning_rate": 1.6356101148376212e-05, "loss": 0.0574, "step": 149640 }, { "epoch": 5.05933263463944, "grad_norm": 4.249659538269043, "learning_rate": 1.6349985807957163e-05, "loss": 0.0596, "step": 149650 }, { "epoch": 5.059670712329694, "grad_norm": 4.582082271575928, "learning_rate": 1.6343871387516874e-05, "loss": 0.0601, "step": 149660 }, { "epoch": 5.060008790019946, "grad_norm": 1.3998690843582153, "learning_rate": 1.63377578872225e-05, "loss": 0.0587, "step": 149670 }, { "epoch": 5.0603468677102, "grad_norm": 32.50126647949219, "learning_rate": 1.6331645307241155e-05, "loss": 0.061, "step": 149680 }, { "epoch": 5.060684945400453, "grad_norm": 6.117778778076172, "learning_rate": 1.6325533647740005e-05, "loss": 0.0589, "step": 149690 }, { "epoch": 5.061023023090706, "grad_norm": 2.9238107204437256, "learning_rate": 1.6319422908886107e-05, "loss": 0.0574, "step": 149700 }, { "epoch": 5.061361100780959, "grad_norm": 1.7750908136367798, "learning_rate": 1.631331309084652e-05, "loss": 0.0587, "step": 149710 }, { "epoch": 5.061699178471213, "grad_norm": 5.176365852355957, "learning_rate": 1.6307204193788317e-05, "loss": 0.0604, "step": 149720 }, { "epoch": 5.0620372561614655, "grad_norm": 1.291732907295227, "learning_rate": 1.630109621787847e-05, "loss": 0.0534, "step": 149730 }, { "epoch": 5.062375333851719, "grad_norm": 1.4818058013916016, "learning_rate": 1.629498916328403e-05, "loss": 0.0545, "step": 149740 }, { "epoch": 5.0627134115419725, "grad_norm": 1.388647437095642, "learning_rate": 1.6288883030171892e-05, "loss": 0.0524, "step": 149750 }, { "epoch": 5.063051489232225, "grad_norm": 1.506593942642212, "learning_rate": 1.628277781870904e-05, "loss": 0.0524, "step": 149760 }, { "epoch": 5.063389566922479, "grad_norm": 1.2139590978622437, "learning_rate": 1.6276673529062365e-05, "loss": 0.0547, "step": 149770 }, { "epoch": 5.063727644612732, "grad_norm": 2.317668914794922, "learning_rate": 1.6270570161398775e-05, "loss": 0.0563, "step": 149780 }, { "epoch": 5.064065722302985, "grad_norm": 1.4565446376800537, "learning_rate": 1.6264467715885128e-05, "loss": 0.0559, "step": 149790 }, { "epoch": 5.064403799993238, "grad_norm": 1.8385621309280396, "learning_rate": 1.6258366192688246e-05, "loss": 0.051, "step": 149800 }, { "epoch": 5.064741877683492, "grad_norm": 1.0322140455245972, "learning_rate": 1.625226559197497e-05, "loss": 0.0557, "step": 149810 }, { "epoch": 5.065079955373745, "grad_norm": 1.4340087175369263, "learning_rate": 1.6246165913912075e-05, "loss": 0.0576, "step": 149820 }, { "epoch": 5.065418033063998, "grad_norm": 2.8839850425720215, "learning_rate": 1.6240067158666324e-05, "loss": 0.0581, "step": 149830 }, { "epoch": 5.0657561107542515, "grad_norm": 1.572183609008789, "learning_rate": 1.6233969326404437e-05, "loss": 0.0552, "step": 149840 }, { "epoch": 5.066094188444504, "grad_norm": 1.2521711587905884, "learning_rate": 1.6227872417293144e-05, "loss": 0.0528, "step": 149850 }, { "epoch": 5.066432266134758, "grad_norm": 1.8186907768249512, "learning_rate": 1.6221776431499148e-05, "loss": 0.0581, "step": 149860 }, { "epoch": 5.066770343825011, "grad_norm": 5.183618068695068, "learning_rate": 1.6215681369189094e-05, "loss": 0.054, "step": 149870 }, { "epoch": 5.067108421515265, "grad_norm": 2.4475257396698, "learning_rate": 1.6209587230529614e-05, "loss": 0.0555, "step": 149880 }, { "epoch": 5.067446499205517, "grad_norm": 4.584283828735352, "learning_rate": 1.6203494015687316e-05, "loss": 0.0568, "step": 149890 }, { "epoch": 5.067784576895771, "grad_norm": 1.9584107398986816, "learning_rate": 1.6197401724828808e-05, "loss": 0.0534, "step": 149900 }, { "epoch": 5.068122654586024, "grad_norm": 2.660073757171631, "learning_rate": 1.6191310358120637e-05, "loss": 0.0564, "step": 149910 }, { "epoch": 5.068460732276277, "grad_norm": 3.209461212158203, "learning_rate": 1.6185219915729326e-05, "loss": 0.0541, "step": 149920 }, { "epoch": 5.06879880996653, "grad_norm": 2.2596187591552734, "learning_rate": 1.6179130397821412e-05, "loss": 0.054, "step": 149930 }, { "epoch": 5.069136887656784, "grad_norm": 1.9661006927490234, "learning_rate": 1.6173041804563354e-05, "loss": 0.0521, "step": 149940 }, { "epoch": 5.069474965347037, "grad_norm": 0.6623036861419678, "learning_rate": 1.6166954136121654e-05, "loss": 0.0468, "step": 149950 }, { "epoch": 5.06981304303729, "grad_norm": 4.264469623565674, "learning_rate": 1.6160867392662688e-05, "loss": 0.0591, "step": 149960 }, { "epoch": 5.070151120727544, "grad_norm": 1.3066065311431885, "learning_rate": 1.6154781574352894e-05, "loss": 0.0573, "step": 149970 }, { "epoch": 5.070489198417796, "grad_norm": 1.6764148473739624, "learning_rate": 1.6148696681358672e-05, "loss": 0.0581, "step": 149980 }, { "epoch": 5.07082727610805, "grad_norm": 5.873584747314453, "learning_rate": 1.6142612713846367e-05, "loss": 0.0565, "step": 149990 }, { "epoch": 5.071165353798303, "grad_norm": 2.6291041374206543, "learning_rate": 1.613652967198231e-05, "loss": 0.0499, "step": 150000 }, { "epoch": 5.071503431488556, "grad_norm": 2.4121897220611572, "learning_rate": 1.6130447555932788e-05, "loss": 0.0546, "step": 150010 }, { "epoch": 5.071841509178809, "grad_norm": 4.449934005737305, "learning_rate": 1.612436636586413e-05, "loss": 0.0587, "step": 150020 }, { "epoch": 5.072179586869063, "grad_norm": 0.8295425176620483, "learning_rate": 1.6118286101942566e-05, "loss": 0.0564, "step": 150030 }, { "epoch": 5.0725176645593155, "grad_norm": 0.9520647525787354, "learning_rate": 1.6112206764334313e-05, "loss": 0.0557, "step": 150040 }, { "epoch": 5.072855742249569, "grad_norm": 17.090347290039062, "learning_rate": 1.6106128353205612e-05, "loss": 0.0525, "step": 150050 }, { "epoch": 5.0731938199398225, "grad_norm": 1.1476147174835205, "learning_rate": 1.6100050868722617e-05, "loss": 0.0448, "step": 150060 }, { "epoch": 5.073531897630075, "grad_norm": 5.343655109405518, "learning_rate": 1.609397431105151e-05, "loss": 0.0552, "step": 150070 }, { "epoch": 5.073869975320329, "grad_norm": 9.975224494934082, "learning_rate": 1.6087898680358405e-05, "loss": 0.0564, "step": 150080 }, { "epoch": 5.074208053010582, "grad_norm": 1.6911470890045166, "learning_rate": 1.6081823976809395e-05, "loss": 0.0591, "step": 150090 }, { "epoch": 5.074546130700835, "grad_norm": 7.647157669067383, "learning_rate": 1.6075750200570595e-05, "loss": 0.0512, "step": 150100 }, { "epoch": 5.074884208391088, "grad_norm": 2.5534486770629883, "learning_rate": 1.6069677351808033e-05, "loss": 0.0527, "step": 150110 }, { "epoch": 5.075222286081342, "grad_norm": 1.1588799953460693, "learning_rate": 1.606360543068775e-05, "loss": 0.0546, "step": 150120 }, { "epoch": 5.0755603637715945, "grad_norm": 2.0330207347869873, "learning_rate": 1.6057534437375728e-05, "loss": 0.0515, "step": 150130 }, { "epoch": 5.075898441461848, "grad_norm": 1.8184229135513306, "learning_rate": 1.6051464372037977e-05, "loss": 0.0618, "step": 150140 }, { "epoch": 5.0762365191521015, "grad_norm": 9.104740142822266, "learning_rate": 1.604539523484042e-05, "loss": 0.0551, "step": 150150 }, { "epoch": 5.076574596842354, "grad_norm": 1.8677144050598145, "learning_rate": 1.6039327025949018e-05, "loss": 0.0543, "step": 150160 }, { "epoch": 5.076912674532608, "grad_norm": 1.5167896747589111, "learning_rate": 1.603325974552966e-05, "loss": 0.0573, "step": 150170 }, { "epoch": 5.077250752222861, "grad_norm": 11.578763008117676, "learning_rate": 1.60271933937482e-05, "loss": 0.0577, "step": 150180 }, { "epoch": 5.077588829913114, "grad_norm": 2.168396234512329, "learning_rate": 1.6021127970770527e-05, "loss": 0.0585, "step": 150190 }, { "epoch": 5.077926907603367, "grad_norm": 3.8266146183013916, "learning_rate": 1.601506347676245e-05, "loss": 0.06, "step": 150200 }, { "epoch": 5.078264985293621, "grad_norm": 1.160468578338623, "learning_rate": 1.600899991188977e-05, "loss": 0.0541, "step": 150210 }, { "epoch": 5.078603062983873, "grad_norm": 3.191155433654785, "learning_rate": 1.6002937276318248e-05, "loss": 0.0545, "step": 150220 }, { "epoch": 5.078941140674127, "grad_norm": 0.8374627828598022, "learning_rate": 1.599687557021365e-05, "loss": 0.0581, "step": 150230 }, { "epoch": 5.0792792183643805, "grad_norm": 2.2655367851257324, "learning_rate": 1.5990814793741725e-05, "loss": 0.0526, "step": 150240 }, { "epoch": 5.079617296054633, "grad_norm": 1.0908393859863281, "learning_rate": 1.5984754947068122e-05, "loss": 0.0546, "step": 150250 }, { "epoch": 5.079955373744887, "grad_norm": 2.099364995956421, "learning_rate": 1.5978696030358548e-05, "loss": 0.0551, "step": 150260 }, { "epoch": 5.08029345143514, "grad_norm": 1.535664439201355, "learning_rate": 1.597263804377863e-05, "loss": 0.0584, "step": 150270 }, { "epoch": 5.080631529125393, "grad_norm": 2.5248100757598877, "learning_rate": 1.5966580987494025e-05, "loss": 0.053, "step": 150280 }, { "epoch": 5.080969606815646, "grad_norm": 1.5446604490280151, "learning_rate": 1.5960524861670306e-05, "loss": 0.0598, "step": 150290 }, { "epoch": 5.0813076845059, "grad_norm": 2.34511137008667, "learning_rate": 1.5954469666473038e-05, "loss": 0.0564, "step": 150300 }, { "epoch": 5.081645762196152, "grad_norm": 0.8718681335449219, "learning_rate": 1.594841540206779e-05, "loss": 0.0542, "step": 150310 }, { "epoch": 5.081983839886406, "grad_norm": 0.8599544763565063, "learning_rate": 1.5942362068620075e-05, "loss": 0.0538, "step": 150320 }, { "epoch": 5.082321917576659, "grad_norm": 1.1367912292480469, "learning_rate": 1.5936309666295384e-05, "loss": 0.0527, "step": 150330 }, { "epoch": 5.082659995266912, "grad_norm": 1.4913270473480225, "learning_rate": 1.5930258195259174e-05, "loss": 0.0577, "step": 150340 }, { "epoch": 5.0829980729571655, "grad_norm": 4.697709560394287, "learning_rate": 1.592420765567691e-05, "loss": 0.0548, "step": 150350 }, { "epoch": 5.083336150647419, "grad_norm": 1.5320460796356201, "learning_rate": 1.5918158047714014e-05, "loss": 0.0591, "step": 150360 }, { "epoch": 5.083674228337672, "grad_norm": 1.3948806524276733, "learning_rate": 1.5912109371535882e-05, "loss": 0.0523, "step": 150370 }, { "epoch": 5.084012306027925, "grad_norm": 1.1940526962280273, "learning_rate": 1.5906061627307868e-05, "loss": 0.0492, "step": 150380 }, { "epoch": 5.084350383718179, "grad_norm": 3.7094669342041016, "learning_rate": 1.5900014815195308e-05, "loss": 0.0527, "step": 150390 }, { "epoch": 5.084688461408431, "grad_norm": 1.8569655418395996, "learning_rate": 1.5893968935363545e-05, "loss": 0.0555, "step": 150400 }, { "epoch": 5.085026539098685, "grad_norm": 0.7218925356864929, "learning_rate": 1.588792398797786e-05, "loss": 0.056, "step": 150410 }, { "epoch": 5.085364616788938, "grad_norm": 3.443225622177124, "learning_rate": 1.58818799732035e-05, "loss": 0.0545, "step": 150420 }, { "epoch": 5.085702694479191, "grad_norm": 7.240528106689453, "learning_rate": 1.5875836891205742e-05, "loss": 0.059, "step": 150430 }, { "epoch": 5.0860407721694445, "grad_norm": 1.8347675800323486, "learning_rate": 1.586979474214977e-05, "loss": 0.0576, "step": 150440 }, { "epoch": 5.086378849859698, "grad_norm": 4.2890448570251465, "learning_rate": 1.5863753526200814e-05, "loss": 0.0597, "step": 150450 }, { "epoch": 5.086716927549951, "grad_norm": 1.6198198795318604, "learning_rate": 1.5857713243523986e-05, "loss": 0.0583, "step": 150460 }, { "epoch": 5.087055005240204, "grad_norm": 7.649960517883301, "learning_rate": 1.5851673894284445e-05, "loss": 0.0614, "step": 150470 }, { "epoch": 5.087393082930458, "grad_norm": 1.9390895366668701, "learning_rate": 1.584563547864733e-05, "loss": 0.0573, "step": 150480 }, { "epoch": 5.08773116062071, "grad_norm": 12.782936096191406, "learning_rate": 1.583959799677771e-05, "loss": 0.0503, "step": 150490 }, { "epoch": 5.088069238310964, "grad_norm": 2.0150961875915527, "learning_rate": 1.583356144884065e-05, "loss": 0.0535, "step": 150500 }, { "epoch": 5.088407316001217, "grad_norm": 1.4628170728683472, "learning_rate": 1.582752583500116e-05, "loss": 0.0586, "step": 150510 }, { "epoch": 5.08874539369147, "grad_norm": 1.0081602334976196, "learning_rate": 1.58214911554243e-05, "loss": 0.0588, "step": 150520 }, { "epoch": 5.0890834713817235, "grad_norm": 1.9543519020080566, "learning_rate": 1.581545741027503e-05, "loss": 0.0557, "step": 150530 }, { "epoch": 5.089421549071977, "grad_norm": 2.030202627182007, "learning_rate": 1.580942459971829e-05, "loss": 0.0562, "step": 150540 }, { "epoch": 5.08975962676223, "grad_norm": 1.5186967849731445, "learning_rate": 1.5803392723919063e-05, "loss": 0.0605, "step": 150550 }, { "epoch": 5.090097704452483, "grad_norm": 2.5736899375915527, "learning_rate": 1.579736178304221e-05, "loss": 0.0605, "step": 150560 }, { "epoch": 5.090435782142737, "grad_norm": 1.480314016342163, "learning_rate": 1.579133177725265e-05, "loss": 0.0514, "step": 150570 }, { "epoch": 5.090773859832989, "grad_norm": 1.9670417308807373, "learning_rate": 1.5785302706715233e-05, "loss": 0.0533, "step": 150580 }, { "epoch": 5.091111937523243, "grad_norm": 1.5196268558502197, "learning_rate": 1.5779274571594782e-05, "loss": 0.0494, "step": 150590 }, { "epoch": 5.091450015213496, "grad_norm": 1.1619234085083008, "learning_rate": 1.5773247372056093e-05, "loss": 0.0536, "step": 150600 }, { "epoch": 5.091788092903749, "grad_norm": 5.760233402252197, "learning_rate": 1.5767221108263978e-05, "loss": 0.0528, "step": 150610 }, { "epoch": 5.092126170594002, "grad_norm": 1.3513975143432617, "learning_rate": 1.5761195780383174e-05, "loss": 0.0586, "step": 150620 }, { "epoch": 5.092464248284256, "grad_norm": 3.323361396789551, "learning_rate": 1.5755171388578406e-05, "loss": 0.0547, "step": 150630 }, { "epoch": 5.0928023259745085, "grad_norm": 36.70823669433594, "learning_rate": 1.57491479330144e-05, "loss": 0.0618, "step": 150640 }, { "epoch": 5.093140403664762, "grad_norm": 2.597668170928955, "learning_rate": 1.574312541385581e-05, "loss": 0.0631, "step": 150650 }, { "epoch": 5.093478481355016, "grad_norm": 1.314437747001648, "learning_rate": 1.5737103831267323e-05, "loss": 0.0567, "step": 150660 }, { "epoch": 5.093816559045268, "grad_norm": 1.1265634298324585, "learning_rate": 1.573108318541352e-05, "loss": 0.0604, "step": 150670 }, { "epoch": 5.094154636735522, "grad_norm": 0.7743209600448608, "learning_rate": 1.572506347645903e-05, "loss": 0.0603, "step": 150680 }, { "epoch": 5.094492714425775, "grad_norm": 1.3541077375411987, "learning_rate": 1.5719044704568443e-05, "loss": 0.0577, "step": 150690 }, { "epoch": 5.094830792116028, "grad_norm": 6.824957847595215, "learning_rate": 1.5713026869906294e-05, "loss": 0.0612, "step": 150700 }, { "epoch": 5.095168869806281, "grad_norm": 2.1541695594787598, "learning_rate": 1.570700997263711e-05, "loss": 0.058, "step": 150710 }, { "epoch": 5.095506947496535, "grad_norm": 4.132026672363281, "learning_rate": 1.5700994012925368e-05, "loss": 0.0606, "step": 150720 }, { "epoch": 5.0958450251867875, "grad_norm": 1.7694646120071411, "learning_rate": 1.5694978990935572e-05, "loss": 0.059, "step": 150730 }, { "epoch": 5.096183102877041, "grad_norm": 1.8298251628875732, "learning_rate": 1.5688964906832193e-05, "loss": 0.0541, "step": 150740 }, { "epoch": 5.0965211805672945, "grad_norm": 1.8782113790512085, "learning_rate": 1.568295176077959e-05, "loss": 0.0578, "step": 150750 }, { "epoch": 5.096859258257547, "grad_norm": 3.5695841312408447, "learning_rate": 1.567693955294221e-05, "loss": 0.0595, "step": 150760 }, { "epoch": 5.097197335947801, "grad_norm": 1.5402913093566895, "learning_rate": 1.5670928283484393e-05, "loss": 0.0659, "step": 150770 }, { "epoch": 5.097535413638054, "grad_norm": 1.3452867269515991, "learning_rate": 1.566491795257051e-05, "loss": 0.0594, "step": 150780 }, { "epoch": 5.097873491328308, "grad_norm": 5.2280778884887695, "learning_rate": 1.5658908560364884e-05, "loss": 0.0599, "step": 150790 }, { "epoch": 5.09821156901856, "grad_norm": 1.2104240655899048, "learning_rate": 1.565290010703177e-05, "loss": 0.0557, "step": 150800 }, { "epoch": 5.098549646708814, "grad_norm": 1.3031492233276367, "learning_rate": 1.5646892592735478e-05, "loss": 0.0584, "step": 150810 }, { "epoch": 5.0988877243990665, "grad_norm": 1.7293165922164917, "learning_rate": 1.5640886017640243e-05, "loss": 0.0555, "step": 150820 }, { "epoch": 5.09922580208932, "grad_norm": 2.4743001461029053, "learning_rate": 1.5634880381910273e-05, "loss": 0.0583, "step": 150830 }, { "epoch": 5.0995638797795735, "grad_norm": 1.7738690376281738, "learning_rate": 1.562887568570975e-05, "loss": 0.0528, "step": 150840 }, { "epoch": 5.099901957469827, "grad_norm": 2.3735508918762207, "learning_rate": 1.5622871929202864e-05, "loss": 0.0563, "step": 150850 }, { "epoch": 5.10024003516008, "grad_norm": 1.6786998510360718, "learning_rate": 1.5616869112553733e-05, "loss": 0.0568, "step": 150860 }, { "epoch": 5.100578112850333, "grad_norm": 5.86676025390625, "learning_rate": 1.5610867235926502e-05, "loss": 0.0509, "step": 150870 }, { "epoch": 5.100916190540587, "grad_norm": 0.9708353877067566, "learning_rate": 1.5604866299485233e-05, "loss": 0.0507, "step": 150880 }, { "epoch": 5.101254268230839, "grad_norm": 2.0867111682891846, "learning_rate": 1.559886630339399e-05, "loss": 0.0652, "step": 150890 }, { "epoch": 5.101592345921093, "grad_norm": 1.9514981508255005, "learning_rate": 1.5592867247816827e-05, "loss": 0.057, "step": 150900 }, { "epoch": 5.101930423611346, "grad_norm": 3.9839489459991455, "learning_rate": 1.5586869132917752e-05, "loss": 0.0555, "step": 150910 }, { "epoch": 5.102268501301599, "grad_norm": 2.942196846008301, "learning_rate": 1.5580871958860726e-05, "loss": 0.0564, "step": 150920 }, { "epoch": 5.102606578991852, "grad_norm": 1.193907380104065, "learning_rate": 1.557487572580975e-05, "loss": 0.0512, "step": 150930 }, { "epoch": 5.102944656682106, "grad_norm": 2.2237191200256348, "learning_rate": 1.5568880433928722e-05, "loss": 0.0569, "step": 150940 }, { "epoch": 5.103282734372359, "grad_norm": 3.0470759868621826, "learning_rate": 1.55628860833816e-05, "loss": 0.0562, "step": 150950 }, { "epoch": 5.103620812062612, "grad_norm": 3.9070017337799072, "learning_rate": 1.55568926743322e-05, "loss": 0.0521, "step": 150960 }, { "epoch": 5.103958889752866, "grad_norm": 1.688449740409851, "learning_rate": 1.5550900206944434e-05, "loss": 0.052, "step": 150970 }, { "epoch": 5.104296967443118, "grad_norm": 7.215394496917725, "learning_rate": 1.5544908681382105e-05, "loss": 0.0575, "step": 150980 }, { "epoch": 5.104635045133372, "grad_norm": 2.4785330295562744, "learning_rate": 1.5538918097809036e-05, "loss": 0.0524, "step": 150990 }, { "epoch": 5.104973122823625, "grad_norm": 2.6051998138427734, "learning_rate": 1.5532928456389e-05, "loss": 0.0552, "step": 151000 }, { "epoch": 5.105311200513878, "grad_norm": 1.3413078784942627, "learning_rate": 1.5526939757285736e-05, "loss": 0.0588, "step": 151010 }, { "epoch": 5.105649278204131, "grad_norm": 1.2660388946533203, "learning_rate": 1.5520952000663e-05, "loss": 0.0539, "step": 151020 }, { "epoch": 5.105987355894385, "grad_norm": 5.351236343383789, "learning_rate": 1.5514965186684487e-05, "loss": 0.0547, "step": 151030 }, { "epoch": 5.1063254335846375, "grad_norm": 2.2425262928009033, "learning_rate": 1.5508979315513877e-05, "loss": 0.0563, "step": 151040 }, { "epoch": 5.106663511274891, "grad_norm": 1.6341965198516846, "learning_rate": 1.550299438731479e-05, "loss": 0.0622, "step": 151050 }, { "epoch": 5.107001588965145, "grad_norm": 6.73335075378418, "learning_rate": 1.5497010402250876e-05, "loss": 0.0539, "step": 151060 }, { "epoch": 5.107339666655397, "grad_norm": 2.884810209274292, "learning_rate": 1.549102736048576e-05, "loss": 0.054, "step": 151070 }, { "epoch": 5.107677744345651, "grad_norm": 3.0866379737854004, "learning_rate": 1.5485045262182984e-05, "loss": 0.0543, "step": 151080 }, { "epoch": 5.108015822035904, "grad_norm": 1.9406158924102783, "learning_rate": 1.547906410750611e-05, "loss": 0.0553, "step": 151090 }, { "epoch": 5.108353899726157, "grad_norm": 0.9488702416419983, "learning_rate": 1.5473083896618628e-05, "loss": 0.0604, "step": 151100 }, { "epoch": 5.10869197741641, "grad_norm": 0.6298236846923828, "learning_rate": 1.5467104629684087e-05, "loss": 0.0531, "step": 151110 }, { "epoch": 5.109030055106664, "grad_norm": 1.176436424255371, "learning_rate": 1.546112630686593e-05, "loss": 0.0602, "step": 151120 }, { "epoch": 5.1093681327969165, "grad_norm": 2.2019524574279785, "learning_rate": 1.5455148928327583e-05, "loss": 0.057, "step": 151130 }, { "epoch": 5.10970621048717, "grad_norm": 1.168046474456787, "learning_rate": 1.544917249423251e-05, "loss": 0.0532, "step": 151140 }, { "epoch": 5.1100442881774235, "grad_norm": 1.4022406339645386, "learning_rate": 1.544319700474406e-05, "loss": 0.0541, "step": 151150 }, { "epoch": 5.110382365867676, "grad_norm": 0.7158960103988647, "learning_rate": 1.5437222460025652e-05, "loss": 0.0616, "step": 151160 }, { "epoch": 5.11072044355793, "grad_norm": 1.9812254905700684, "learning_rate": 1.5431248860240573e-05, "loss": 0.0499, "step": 151170 }, { "epoch": 5.111058521248183, "grad_norm": 2.9945921897888184, "learning_rate": 1.5425276205552152e-05, "loss": 0.0563, "step": 151180 }, { "epoch": 5.111396598938436, "grad_norm": 0.7051414251327515, "learning_rate": 1.5419304496123715e-05, "loss": 0.0569, "step": 151190 }, { "epoch": 5.111734676628689, "grad_norm": 1.444376826286316, "learning_rate": 1.5413333732118502e-05, "loss": 0.0567, "step": 151200 }, { "epoch": 5.112072754318943, "grad_norm": 0.8196407556533813, "learning_rate": 1.5407363913699746e-05, "loss": 0.0549, "step": 151210 }, { "epoch": 5.112410832009195, "grad_norm": 1.5122430324554443, "learning_rate": 1.5401395041030653e-05, "loss": 0.0502, "step": 151220 }, { "epoch": 5.112748909699449, "grad_norm": 1.098992109298706, "learning_rate": 1.5395427114274437e-05, "loss": 0.0582, "step": 151230 }, { "epoch": 5.1130869873897025, "grad_norm": 0.6121402978897095, "learning_rate": 1.5389460133594243e-05, "loss": 0.0562, "step": 151240 }, { "epoch": 5.113425065079955, "grad_norm": 1.1049365997314453, "learning_rate": 1.5383494099153185e-05, "loss": 0.0593, "step": 151250 }, { "epoch": 5.113763142770209, "grad_norm": 4.087777614593506, "learning_rate": 1.5377529011114416e-05, "loss": 0.0561, "step": 151260 }, { "epoch": 5.114101220460462, "grad_norm": 1.8978430032730103, "learning_rate": 1.5371564869640977e-05, "loss": 0.0525, "step": 151270 }, { "epoch": 5.114439298150715, "grad_norm": 1.0974479913711548, "learning_rate": 1.5365601674895964e-05, "loss": 0.0601, "step": 151280 }, { "epoch": 5.114777375840968, "grad_norm": 1.955474615097046, "learning_rate": 1.535963942704239e-05, "loss": 0.0554, "step": 151290 }, { "epoch": 5.115115453531222, "grad_norm": 2.104170083999634, "learning_rate": 1.5353678126243242e-05, "loss": 0.0549, "step": 151300 }, { "epoch": 5.115453531221474, "grad_norm": 5.265700340270996, "learning_rate": 1.5347717772661535e-05, "loss": 0.0632, "step": 151310 }, { "epoch": 5.115791608911728, "grad_norm": 5.461947917938232, "learning_rate": 1.53417583664602e-05, "loss": 0.0488, "step": 151320 }, { "epoch": 5.116129686601981, "grad_norm": 0.6703312397003174, "learning_rate": 1.5335799907802178e-05, "loss": 0.0524, "step": 151330 }, { "epoch": 5.116467764292234, "grad_norm": 0.9768701195716858, "learning_rate": 1.5329842396850347e-05, "loss": 0.0559, "step": 151340 }, { "epoch": 5.116805841982488, "grad_norm": 1.0955696105957031, "learning_rate": 1.5323885833767615e-05, "loss": 0.0517, "step": 151350 }, { "epoch": 5.117143919672741, "grad_norm": 2.1818125247955322, "learning_rate": 1.5317930218716802e-05, "loss": 0.0613, "step": 151360 }, { "epoch": 5.117481997362994, "grad_norm": 0.8344327211380005, "learning_rate": 1.5311975551860763e-05, "loss": 0.0559, "step": 151370 }, { "epoch": 5.117820075053247, "grad_norm": 1.9477368593215942, "learning_rate": 1.5306021833362282e-05, "loss": 0.0494, "step": 151380 }, { "epoch": 5.118158152743501, "grad_norm": 1.278668761253357, "learning_rate": 1.530006906338411e-05, "loss": 0.0521, "step": 151390 }, { "epoch": 5.118496230433753, "grad_norm": 0.9477869868278503, "learning_rate": 1.5294117242089033e-05, "loss": 0.0564, "step": 151400 }, { "epoch": 5.118834308124007, "grad_norm": 1.832993745803833, "learning_rate": 1.528816636963975e-05, "loss": 0.056, "step": 151410 }, { "epoch": 5.11917238581426, "grad_norm": 2.4503514766693115, "learning_rate": 1.528221644619896e-05, "loss": 0.0541, "step": 151420 }, { "epoch": 5.119510463504513, "grad_norm": 2.4758167266845703, "learning_rate": 1.527626747192931e-05, "loss": 0.0586, "step": 151430 }, { "epoch": 5.1198485411947665, "grad_norm": 0.4666004180908203, "learning_rate": 1.527031944699347e-05, "loss": 0.054, "step": 151440 }, { "epoch": 5.12018661888502, "grad_norm": 0.6202061772346497, "learning_rate": 1.5264372371554074e-05, "loss": 0.055, "step": 151450 }, { "epoch": 5.120524696575273, "grad_norm": 1.0082128047943115, "learning_rate": 1.5258426245773661e-05, "loss": 0.0599, "step": 151460 }, { "epoch": 5.120862774265526, "grad_norm": 1.3867967128753662, "learning_rate": 1.5252481069814834e-05, "loss": 0.0604, "step": 151470 }, { "epoch": 5.12120085195578, "grad_norm": 0.6677021384239197, "learning_rate": 1.5246536843840103e-05, "loss": 0.0566, "step": 151480 }, { "epoch": 5.121538929646032, "grad_norm": 0.9720016717910767, "learning_rate": 1.5240593568012017e-05, "loss": 0.0535, "step": 151490 }, { "epoch": 5.121877007336286, "grad_norm": 1.219748854637146, "learning_rate": 1.5234651242493047e-05, "loss": 0.0557, "step": 151500 }, { "epoch": 5.122215085026539, "grad_norm": 1.8671079874038696, "learning_rate": 1.5228709867445635e-05, "loss": 0.056, "step": 151510 }, { "epoch": 5.122553162716792, "grad_norm": 1.1342698335647583, "learning_rate": 1.5222769443032243e-05, "loss": 0.0567, "step": 151520 }, { "epoch": 5.1228912404070455, "grad_norm": 2.6433544158935547, "learning_rate": 1.521682996941527e-05, "loss": 0.0569, "step": 151530 }, { "epoch": 5.123229318097299, "grad_norm": 1.8397693634033203, "learning_rate": 1.5210891446757103e-05, "loss": 0.0502, "step": 151540 }, { "epoch": 5.123567395787552, "grad_norm": 1.0656116008758545, "learning_rate": 1.5204953875220073e-05, "loss": 0.0509, "step": 151550 }, { "epoch": 5.123905473477805, "grad_norm": 0.6641805171966553, "learning_rate": 1.5199017254966536e-05, "loss": 0.0582, "step": 151560 }, { "epoch": 5.124243551168059, "grad_norm": 1.272286057472229, "learning_rate": 1.5193081586158809e-05, "loss": 0.0529, "step": 151570 }, { "epoch": 5.124581628858311, "grad_norm": 3.437065601348877, "learning_rate": 1.5187146868959157e-05, "loss": 0.054, "step": 151580 }, { "epoch": 5.124919706548565, "grad_norm": 1.9265884160995483, "learning_rate": 1.5181213103529824e-05, "loss": 0.057, "step": 151590 }, { "epoch": 5.125257784238818, "grad_norm": 2.291425943374634, "learning_rate": 1.5175280290033033e-05, "loss": 0.0579, "step": 151600 }, { "epoch": 5.125595861929071, "grad_norm": 1.5589556694030762, "learning_rate": 1.5169348428631015e-05, "loss": 0.0616, "step": 151610 }, { "epoch": 5.125933939619324, "grad_norm": 2.0735151767730713, "learning_rate": 1.5163417519485917e-05, "loss": 0.0492, "step": 151620 }, { "epoch": 5.126272017309578, "grad_norm": 1.5563808679580688, "learning_rate": 1.5157487562759887e-05, "loss": 0.0518, "step": 151630 }, { "epoch": 5.126610094999831, "grad_norm": 2.1499507427215576, "learning_rate": 1.515155855861507e-05, "loss": 0.0551, "step": 151640 }, { "epoch": 5.126948172690084, "grad_norm": 1.4419529438018799, "learning_rate": 1.5145630507213537e-05, "loss": 0.0607, "step": 151650 }, { "epoch": 5.127286250380338, "grad_norm": 1.6173590421676636, "learning_rate": 1.5139703408717399e-05, "loss": 0.0573, "step": 151660 }, { "epoch": 5.12762432807059, "grad_norm": 0.8555018305778503, "learning_rate": 1.5133777263288646e-05, "loss": 0.0556, "step": 151670 }, { "epoch": 5.127962405760844, "grad_norm": 3.6655642986297607, "learning_rate": 1.512785207108932e-05, "loss": 0.0525, "step": 151680 }, { "epoch": 5.128300483451097, "grad_norm": 4.429382801055908, "learning_rate": 1.512192783228144e-05, "loss": 0.0585, "step": 151690 }, { "epoch": 5.12863856114135, "grad_norm": 0.9639086723327637, "learning_rate": 1.5116004547026947e-05, "loss": 0.065, "step": 151700 }, { "epoch": 5.128976638831603, "grad_norm": 0.8606576919555664, "learning_rate": 1.511008221548778e-05, "loss": 0.06, "step": 151710 }, { "epoch": 5.129314716521857, "grad_norm": 2.0032079219818115, "learning_rate": 1.5104160837825849e-05, "loss": 0.0602, "step": 151720 }, { "epoch": 5.1296527942121095, "grad_norm": 0.9993412494659424, "learning_rate": 1.5098240414203063e-05, "loss": 0.059, "step": 151730 }, { "epoch": 5.129990871902363, "grad_norm": 1.0227482318878174, "learning_rate": 1.5092320944781269e-05, "loss": 0.0537, "step": 151740 }, { "epoch": 5.1303289495926165, "grad_norm": 1.1820452213287354, "learning_rate": 1.5086402429722297e-05, "loss": 0.0573, "step": 151750 }, { "epoch": 5.13066702728287, "grad_norm": 3.900435209274292, "learning_rate": 1.5080484869187977e-05, "loss": 0.0588, "step": 151760 }, { "epoch": 5.131005104973123, "grad_norm": 9.840682983398438, "learning_rate": 1.5074568263340067e-05, "loss": 0.0552, "step": 151770 }, { "epoch": 5.131343182663376, "grad_norm": 2.728797435760498, "learning_rate": 1.5068652612340355e-05, "loss": 0.0543, "step": 151780 }, { "epoch": 5.131681260353629, "grad_norm": 2.258681297302246, "learning_rate": 1.5062737916350561e-05, "loss": 0.0594, "step": 151790 }, { "epoch": 5.132019338043882, "grad_norm": 0.9303366541862488, "learning_rate": 1.5056824175532386e-05, "loss": 0.0532, "step": 151800 }, { "epoch": 5.132357415734136, "grad_norm": 1.711774468421936, "learning_rate": 1.505091139004749e-05, "loss": 0.0595, "step": 151810 }, { "epoch": 5.132695493424389, "grad_norm": 5.28030252456665, "learning_rate": 1.5044999560057566e-05, "loss": 0.0509, "step": 151820 }, { "epoch": 5.133033571114642, "grad_norm": 1.6226352453231812, "learning_rate": 1.5039088685724223e-05, "loss": 0.0574, "step": 151830 }, { "epoch": 5.1333716488048955, "grad_norm": 1.5695921182632446, "learning_rate": 1.5033178767209044e-05, "loss": 0.0556, "step": 151840 }, { "epoch": 5.133709726495149, "grad_norm": 1.6235721111297607, "learning_rate": 1.5027269804673639e-05, "loss": 0.0605, "step": 151850 }, { "epoch": 5.134047804185402, "grad_norm": 1.2864725589752197, "learning_rate": 1.5021361798279521e-05, "loss": 0.0576, "step": 151860 }, { "epoch": 5.134385881875655, "grad_norm": 1.260827660560608, "learning_rate": 1.5015454748188263e-05, "loss": 0.0521, "step": 151870 }, { "epoch": 5.134723959565909, "grad_norm": 1.101770043373108, "learning_rate": 1.5009548654561295e-05, "loss": 0.0574, "step": 151880 }, { "epoch": 5.135062037256161, "grad_norm": 0.9723201394081116, "learning_rate": 1.5003643517560123e-05, "loss": 0.0489, "step": 151890 }, { "epoch": 5.135400114946415, "grad_norm": 8.350037574768066, "learning_rate": 1.4997739337346211e-05, "loss": 0.0598, "step": 151900 }, { "epoch": 5.135738192636668, "grad_norm": 31.221403121948242, "learning_rate": 1.4991836114080948e-05, "loss": 0.0516, "step": 151910 }, { "epoch": 5.136076270326921, "grad_norm": 0.9739503264427185, "learning_rate": 1.4985933847925737e-05, "loss": 0.0508, "step": 151920 }, { "epoch": 5.1364143480171744, "grad_norm": 1.446077585220337, "learning_rate": 1.4980032539041927e-05, "loss": 0.056, "step": 151930 }, { "epoch": 5.136752425707428, "grad_norm": 1.892605185508728, "learning_rate": 1.4974132187590868e-05, "loss": 0.0604, "step": 151940 }, { "epoch": 5.137090503397681, "grad_norm": 1.3035573959350586, "learning_rate": 1.4968232793733905e-05, "loss": 0.058, "step": 151950 }, { "epoch": 5.137428581087934, "grad_norm": 1.1909542083740234, "learning_rate": 1.4962334357632268e-05, "loss": 0.056, "step": 151960 }, { "epoch": 5.137766658778188, "grad_norm": 0.8331212401390076, "learning_rate": 1.4956436879447267e-05, "loss": 0.0504, "step": 151970 }, { "epoch": 5.13810473646844, "grad_norm": 3.7622485160827637, "learning_rate": 1.4950540359340092e-05, "loss": 0.0531, "step": 151980 }, { "epoch": 5.138442814158694, "grad_norm": 1.0648155212402344, "learning_rate": 1.4944644797471995e-05, "loss": 0.052, "step": 151990 }, { "epoch": 5.138780891848947, "grad_norm": 1.6298044919967651, "learning_rate": 1.4938750194004142e-05, "loss": 0.0569, "step": 152000 }, { "epoch": 5.1391189695392, "grad_norm": 1.653601884841919, "learning_rate": 1.4932856549097674e-05, "loss": 0.0527, "step": 152010 }, { "epoch": 5.139457047229453, "grad_norm": 1.5071297883987427, "learning_rate": 1.4926963862913746e-05, "loss": 0.056, "step": 152020 }, { "epoch": 5.139795124919707, "grad_norm": 1.6459310054779053, "learning_rate": 1.4921072135613445e-05, "loss": 0.0469, "step": 152030 }, { "epoch": 5.1401332026099595, "grad_norm": 1.8048205375671387, "learning_rate": 1.4915181367357856e-05, "loss": 0.0577, "step": 152040 }, { "epoch": 5.140471280300213, "grad_norm": 2.20229172706604, "learning_rate": 1.4909291558308013e-05, "loss": 0.0593, "step": 152050 }, { "epoch": 5.140809357990467, "grad_norm": 3.9168295860290527, "learning_rate": 1.4903402708624974e-05, "loss": 0.0542, "step": 152060 }, { "epoch": 5.141147435680719, "grad_norm": 1.6187039613723755, "learning_rate": 1.4897514818469705e-05, "loss": 0.0529, "step": 152070 }, { "epoch": 5.141485513370973, "grad_norm": 0.974872350692749, "learning_rate": 1.4891627888003201e-05, "loss": 0.0545, "step": 152080 }, { "epoch": 5.141823591061226, "grad_norm": 7.6622209548950195, "learning_rate": 1.4885741917386409e-05, "loss": 0.0536, "step": 152090 }, { "epoch": 5.142161668751479, "grad_norm": 0.8140506744384766, "learning_rate": 1.4879856906780221e-05, "loss": 0.0498, "step": 152100 }, { "epoch": 5.142499746441732, "grad_norm": 1.2565895318984985, "learning_rate": 1.487397285634557e-05, "loss": 0.0551, "step": 152110 }, { "epoch": 5.142837824131986, "grad_norm": 1.6944199800491333, "learning_rate": 1.4868089766243305e-05, "loss": 0.0548, "step": 152120 }, { "epoch": 5.1431759018222385, "grad_norm": 1.5723834037780762, "learning_rate": 1.4862207636634246e-05, "loss": 0.0551, "step": 152130 }, { "epoch": 5.143513979512492, "grad_norm": 2.4190258979797363, "learning_rate": 1.4856326467679248e-05, "loss": 0.057, "step": 152140 }, { "epoch": 5.1438520572027455, "grad_norm": 1.8528738021850586, "learning_rate": 1.4850446259539069e-05, "loss": 0.0495, "step": 152150 }, { "epoch": 5.144190134892998, "grad_norm": 3.251801013946533, "learning_rate": 1.4844567012374511e-05, "loss": 0.0554, "step": 152160 }, { "epoch": 5.144528212583252, "grad_norm": 1.655106544494629, "learning_rate": 1.4838688726346256e-05, "loss": 0.056, "step": 152170 }, { "epoch": 5.144866290273505, "grad_norm": 2.5226807594299316, "learning_rate": 1.4832811401615054e-05, "loss": 0.0513, "step": 152180 }, { "epoch": 5.145204367963758, "grad_norm": 6.5063090324401855, "learning_rate": 1.4826935038341566e-05, "loss": 0.0592, "step": 152190 }, { "epoch": 5.145542445654011, "grad_norm": 3.6225972175598145, "learning_rate": 1.4821059636686474e-05, "loss": 0.0519, "step": 152200 }, { "epoch": 5.145880523344265, "grad_norm": 2.370021104812622, "learning_rate": 1.4815185196810395e-05, "loss": 0.0562, "step": 152210 }, { "epoch": 5.1462186010345174, "grad_norm": 4.458953380584717, "learning_rate": 1.4809311718873914e-05, "loss": 0.0601, "step": 152220 }, { "epoch": 5.146556678724771, "grad_norm": 1.0831849575042725, "learning_rate": 1.480343920303765e-05, "loss": 0.0529, "step": 152230 }, { "epoch": 5.1468947564150245, "grad_norm": 1.2630380392074585, "learning_rate": 1.4797567649462135e-05, "loss": 0.0499, "step": 152240 }, { "epoch": 5.147232834105277, "grad_norm": 1.5234864950180054, "learning_rate": 1.4791697058307896e-05, "loss": 0.0501, "step": 152250 }, { "epoch": 5.147570911795531, "grad_norm": 1.5690726041793823, "learning_rate": 1.4785827429735415e-05, "loss": 0.0564, "step": 152260 }, { "epoch": 5.147908989485784, "grad_norm": 1.9100220203399658, "learning_rate": 1.4779958763905189e-05, "loss": 0.0601, "step": 152270 }, { "epoch": 5.148247067176037, "grad_norm": 1.95477294921875, "learning_rate": 1.4774091060977668e-05, "loss": 0.0545, "step": 152280 }, { "epoch": 5.14858514486629, "grad_norm": 2.2018020153045654, "learning_rate": 1.4768224321113266e-05, "loss": 0.0568, "step": 152290 }, { "epoch": 5.148923222556544, "grad_norm": 1.3986438512802124, "learning_rate": 1.4762358544472372e-05, "loss": 0.0523, "step": 152300 }, { "epoch": 5.149261300246796, "grad_norm": 0.9582329988479614, "learning_rate": 1.4756493731215343e-05, "loss": 0.0498, "step": 152310 }, { "epoch": 5.14959937793705, "grad_norm": 1.1358698606491089, "learning_rate": 1.4750629881502553e-05, "loss": 0.05, "step": 152320 }, { "epoch": 5.149937455627303, "grad_norm": 0.7460318207740784, "learning_rate": 1.4744766995494298e-05, "loss": 0.0545, "step": 152330 }, { "epoch": 5.150275533317556, "grad_norm": 1.017162799835205, "learning_rate": 1.473890507335085e-05, "loss": 0.0593, "step": 152340 }, { "epoch": 5.15061361100781, "grad_norm": 6.922269344329834, "learning_rate": 1.4733044115232513e-05, "loss": 0.0533, "step": 152350 }, { "epoch": 5.150951688698063, "grad_norm": 1.375982403755188, "learning_rate": 1.4727184121299482e-05, "loss": 0.0589, "step": 152360 }, { "epoch": 5.151289766388316, "grad_norm": 1.412628412246704, "learning_rate": 1.4721325091712012e-05, "loss": 0.0569, "step": 152370 }, { "epoch": 5.151627844078569, "grad_norm": 1.3204370737075806, "learning_rate": 1.4715467026630236e-05, "loss": 0.0541, "step": 152380 }, { "epoch": 5.151965921768823, "grad_norm": 1.6002106666564941, "learning_rate": 1.470960992621433e-05, "loss": 0.057, "step": 152390 }, { "epoch": 5.152303999459075, "grad_norm": 4.753597736358643, "learning_rate": 1.4703753790624452e-05, "loss": 0.0597, "step": 152400 }, { "epoch": 5.152642077149329, "grad_norm": 11.969529151916504, "learning_rate": 1.4697898620020679e-05, "loss": 0.0624, "step": 152410 }, { "epoch": 5.152980154839582, "grad_norm": 2.107203245162964, "learning_rate": 1.46920444145631e-05, "loss": 0.0492, "step": 152420 }, { "epoch": 5.153318232529835, "grad_norm": 4.037163734436035, "learning_rate": 1.4686191174411745e-05, "loss": 0.0548, "step": 152430 }, { "epoch": 5.1536563102200885, "grad_norm": 3.4374887943267822, "learning_rate": 1.468033889972667e-05, "loss": 0.0568, "step": 152440 }, { "epoch": 5.153994387910342, "grad_norm": 4.008044719696045, "learning_rate": 1.4674487590667863e-05, "loss": 0.0558, "step": 152450 }, { "epoch": 5.154332465600595, "grad_norm": 1.4816608428955078, "learning_rate": 1.4668637247395278e-05, "loss": 0.0577, "step": 152460 }, { "epoch": 5.154670543290848, "grad_norm": 1.019754409790039, "learning_rate": 1.4662787870068895e-05, "loss": 0.0556, "step": 152470 }, { "epoch": 5.155008620981102, "grad_norm": 2.382852554321289, "learning_rate": 1.4656939458848602e-05, "loss": 0.0537, "step": 152480 }, { "epoch": 5.155346698671354, "grad_norm": 0.5005947351455688, "learning_rate": 1.4651092013894313e-05, "loss": 0.056, "step": 152490 }, { "epoch": 5.155684776361608, "grad_norm": 0.9319718480110168, "learning_rate": 1.4645245535365898e-05, "loss": 0.0549, "step": 152500 }, { "epoch": 5.156022854051861, "grad_norm": 0.9893869161605835, "learning_rate": 1.4639400023423167e-05, "loss": 0.0511, "step": 152510 }, { "epoch": 5.156360931742114, "grad_norm": 0.7212774753570557, "learning_rate": 1.463355547822597e-05, "loss": 0.0541, "step": 152520 }, { "epoch": 5.1566990094323675, "grad_norm": 1.0234376192092896, "learning_rate": 1.4627711899934082e-05, "loss": 0.0579, "step": 152530 }, { "epoch": 5.157037087122621, "grad_norm": 3.808866024017334, "learning_rate": 1.4621869288707263e-05, "loss": 0.0557, "step": 152540 }, { "epoch": 5.157375164812874, "grad_norm": 0.8539414405822754, "learning_rate": 1.4616027644705227e-05, "loss": 0.0513, "step": 152550 }, { "epoch": 5.157713242503127, "grad_norm": 2.0224804878234863, "learning_rate": 1.461018696808772e-05, "loss": 0.0508, "step": 152560 }, { "epoch": 5.158051320193381, "grad_norm": 2.8269200325012207, "learning_rate": 1.4604347259014395e-05, "loss": 0.0484, "step": 152570 }, { "epoch": 5.158389397883633, "grad_norm": 2.4957072734832764, "learning_rate": 1.4598508517644927e-05, "loss": 0.0551, "step": 152580 }, { "epoch": 5.158727475573887, "grad_norm": 1.4962177276611328, "learning_rate": 1.4592670744138937e-05, "loss": 0.0581, "step": 152590 }, { "epoch": 5.15906555326414, "grad_norm": 1.4297956228256226, "learning_rate": 1.4586833938656014e-05, "loss": 0.0582, "step": 152600 }, { "epoch": 5.159403630954393, "grad_norm": 15.283293724060059, "learning_rate": 1.4580998101355764e-05, "loss": 0.0524, "step": 152610 }, { "epoch": 5.159741708644646, "grad_norm": 0.7967422008514404, "learning_rate": 1.4575163232397715e-05, "loss": 0.0553, "step": 152620 }, { "epoch": 5.1600797863349, "grad_norm": 1.259534478187561, "learning_rate": 1.4569329331941401e-05, "loss": 0.0478, "step": 152630 }, { "epoch": 5.160417864025153, "grad_norm": 0.7933604717254639, "learning_rate": 1.4563496400146292e-05, "loss": 0.0521, "step": 152640 }, { "epoch": 5.160755941715406, "grad_norm": 1.188995599746704, "learning_rate": 1.455766443717188e-05, "loss": 0.0553, "step": 152650 }, { "epoch": 5.16109401940566, "grad_norm": 0.9975546598434448, "learning_rate": 1.4551833443177637e-05, "loss": 0.058, "step": 152660 }, { "epoch": 5.161432097095912, "grad_norm": 2.3292930126190186, "learning_rate": 1.4546003418322917e-05, "loss": 0.0583, "step": 152670 }, { "epoch": 5.161770174786166, "grad_norm": 3.6985421180725098, "learning_rate": 1.4540174362767166e-05, "loss": 0.0519, "step": 152680 }, { "epoch": 5.162108252476419, "grad_norm": 2.0082015991210938, "learning_rate": 1.4534346276669709e-05, "loss": 0.0595, "step": 152690 }, { "epoch": 5.162446330166672, "grad_norm": 1.4002457857131958, "learning_rate": 1.4528519160189912e-05, "loss": 0.0545, "step": 152700 }, { "epoch": 5.162784407856925, "grad_norm": 1.022592544555664, "learning_rate": 1.4522693013487077e-05, "loss": 0.0554, "step": 152710 }, { "epoch": 5.163122485547179, "grad_norm": 8.680479049682617, "learning_rate": 1.4516867836720477e-05, "loss": 0.0567, "step": 152720 }, { "epoch": 5.163460563237432, "grad_norm": 1.2389806509017944, "learning_rate": 1.4511043630049388e-05, "loss": 0.056, "step": 152730 }, { "epoch": 5.163798640927685, "grad_norm": 1.1548000574111938, "learning_rate": 1.4505220393633034e-05, "loss": 0.0597, "step": 152740 }, { "epoch": 5.1641367186179385, "grad_norm": 1.29010808467865, "learning_rate": 1.4499398127630626e-05, "loss": 0.0519, "step": 152750 }, { "epoch": 5.164474796308191, "grad_norm": 1.161380410194397, "learning_rate": 1.4493576832201316e-05, "loss": 0.0481, "step": 152760 }, { "epoch": 5.164812873998445, "grad_norm": 0.7832587361335754, "learning_rate": 1.448775650750428e-05, "loss": 0.0615, "step": 152770 }, { "epoch": 5.165150951688698, "grad_norm": 1.935197353363037, "learning_rate": 1.4481937153698649e-05, "loss": 0.0564, "step": 152780 }, { "epoch": 5.165489029378952, "grad_norm": 0.9185633063316345, "learning_rate": 1.447611877094352e-05, "loss": 0.0574, "step": 152790 }, { "epoch": 5.165827107069204, "grad_norm": 2.772942543029785, "learning_rate": 1.4470301359397954e-05, "loss": 0.0532, "step": 152800 }, { "epoch": 5.166165184759458, "grad_norm": 3.4332902431488037, "learning_rate": 1.4464484919220983e-05, "loss": 0.0571, "step": 152810 }, { "epoch": 5.166503262449711, "grad_norm": 1.082136869430542, "learning_rate": 1.4458669450571661e-05, "loss": 0.0614, "step": 152820 }, { "epoch": 5.166841340139964, "grad_norm": 1.8444387912750244, "learning_rate": 1.4452854953608968e-05, "loss": 0.0589, "step": 152830 }, { "epoch": 5.1671794178302175, "grad_norm": 1.6513434648513794, "learning_rate": 1.4447041428491848e-05, "loss": 0.0536, "step": 152840 }, { "epoch": 5.167517495520471, "grad_norm": 0.8349406719207764, "learning_rate": 1.4441228875379276e-05, "loss": 0.0537, "step": 152850 }, { "epoch": 5.167855573210724, "grad_norm": 2.148259162902832, "learning_rate": 1.4435417294430125e-05, "loss": 0.0539, "step": 152860 }, { "epoch": 5.168193650900977, "grad_norm": 4.52598762512207, "learning_rate": 1.4429606685803342e-05, "loss": 0.0596, "step": 152870 }, { "epoch": 5.168531728591231, "grad_norm": 1.105053424835205, "learning_rate": 1.4423797049657717e-05, "loss": 0.0545, "step": 152880 }, { "epoch": 5.168869806281483, "grad_norm": 0.7819306254386902, "learning_rate": 1.4417988386152114e-05, "loss": 0.0596, "step": 152890 }, { "epoch": 5.169207883971737, "grad_norm": 2.2332816123962402, "learning_rate": 1.441218069544536e-05, "loss": 0.0558, "step": 152900 }, { "epoch": 5.16954596166199, "grad_norm": 1.6882028579711914, "learning_rate": 1.4406373977696213e-05, "loss": 0.0458, "step": 152910 }, { "epoch": 5.169884039352243, "grad_norm": 1.309547781944275, "learning_rate": 1.4400568233063433e-05, "loss": 0.054, "step": 152920 }, { "epoch": 5.1702221170424965, "grad_norm": 3.017831325531006, "learning_rate": 1.4394763461705723e-05, "loss": 0.0491, "step": 152930 }, { "epoch": 5.17056019473275, "grad_norm": 3.9961743354797363, "learning_rate": 1.438895966378182e-05, "loss": 0.054, "step": 152940 }, { "epoch": 5.170898272423003, "grad_norm": 1.901611089706421, "learning_rate": 1.4383156839450384e-05, "loss": 0.0529, "step": 152950 }, { "epoch": 5.171236350113256, "grad_norm": 1.0296525955200195, "learning_rate": 1.437735498887004e-05, "loss": 0.0558, "step": 152960 }, { "epoch": 5.17157442780351, "grad_norm": 3.0908126831054688, "learning_rate": 1.4371554112199448e-05, "loss": 0.0624, "step": 152970 }, { "epoch": 5.171912505493762, "grad_norm": 0.9975261092185974, "learning_rate": 1.4365754209597166e-05, "loss": 0.0563, "step": 152980 }, { "epoch": 5.172250583184016, "grad_norm": 1.8556572198867798, "learning_rate": 1.4359955281221788e-05, "loss": 0.0533, "step": 152990 }, { "epoch": 5.172588660874269, "grad_norm": 1.1609143018722534, "learning_rate": 1.4354157327231843e-05, "loss": 0.06, "step": 153000 }, { "epoch": 5.172926738564522, "grad_norm": 1.8398855924606323, "learning_rate": 1.434836034778585e-05, "loss": 0.0575, "step": 153010 }, { "epoch": 5.173264816254775, "grad_norm": 1.1222294569015503, "learning_rate": 1.4342564343042276e-05, "loss": 0.0595, "step": 153020 }, { "epoch": 5.173602893945029, "grad_norm": 1.611200213432312, "learning_rate": 1.4336769313159609e-05, "loss": 0.0548, "step": 153030 }, { "epoch": 5.1739409716352815, "grad_norm": 0.8441150784492493, "learning_rate": 1.433097525829627e-05, "loss": 0.0513, "step": 153040 }, { "epoch": 5.174279049325535, "grad_norm": 1.5980846881866455, "learning_rate": 1.4325182178610651e-05, "loss": 0.0593, "step": 153050 }, { "epoch": 5.174617127015789, "grad_norm": 1.0759851932525635, "learning_rate": 1.4319390074261169e-05, "loss": 0.0591, "step": 153060 }, { "epoch": 5.174955204706041, "grad_norm": 1.2698625326156616, "learning_rate": 1.4313598945406143e-05, "loss": 0.0585, "step": 153070 }, { "epoch": 5.175293282396295, "grad_norm": 1.2348906993865967, "learning_rate": 1.4307808792203941e-05, "loss": 0.052, "step": 153080 }, { "epoch": 5.175631360086548, "grad_norm": 2.3062446117401123, "learning_rate": 1.4302019614812807e-05, "loss": 0.0529, "step": 153090 }, { "epoch": 5.175969437776801, "grad_norm": 3.3022055625915527, "learning_rate": 1.4296231413391049e-05, "loss": 0.0626, "step": 153100 }, { "epoch": 5.176307515467054, "grad_norm": 2.8699402809143066, "learning_rate": 1.4290444188096924e-05, "loss": 0.0536, "step": 153110 }, { "epoch": 5.176645593157308, "grad_norm": 1.3786991834640503, "learning_rate": 1.4284657939088642e-05, "loss": 0.0558, "step": 153120 }, { "epoch": 5.1769836708475605, "grad_norm": 1.8186895847320557, "learning_rate": 1.427887266652439e-05, "loss": 0.055, "step": 153130 }, { "epoch": 5.177321748537814, "grad_norm": 2.015214204788208, "learning_rate": 1.4273088370562326e-05, "loss": 0.057, "step": 153140 }, { "epoch": 5.1776598262280675, "grad_norm": 27.097232818603516, "learning_rate": 1.4267305051360607e-05, "loss": 0.061, "step": 153150 }, { "epoch": 5.17799790391832, "grad_norm": 3.0679807662963867, "learning_rate": 1.4261522709077375e-05, "loss": 0.0559, "step": 153160 }, { "epoch": 5.178335981608574, "grad_norm": 1.7906371355056763, "learning_rate": 1.4255741343870649e-05, "loss": 0.0518, "step": 153170 }, { "epoch": 5.178674059298827, "grad_norm": 4.824442386627197, "learning_rate": 1.4249960955898544e-05, "loss": 0.053, "step": 153180 }, { "epoch": 5.17901213698908, "grad_norm": 2.4149281978607178, "learning_rate": 1.4244181545319062e-05, "loss": 0.06, "step": 153190 }, { "epoch": 5.179350214679333, "grad_norm": 0.9542326927185059, "learning_rate": 1.4238403112290239e-05, "loss": 0.0563, "step": 153200 }, { "epoch": 5.179688292369587, "grad_norm": 1.1117057800292969, "learning_rate": 1.423262565697004e-05, "loss": 0.056, "step": 153210 }, { "epoch": 5.1800263700598395, "grad_norm": 3.4236228466033936, "learning_rate": 1.42268491795164e-05, "loss": 0.0601, "step": 153220 }, { "epoch": 5.180364447750093, "grad_norm": 1.2763513326644897, "learning_rate": 1.4221073680087283e-05, "loss": 0.0635, "step": 153230 }, { "epoch": 5.1807025254403465, "grad_norm": 1.5264390707015991, "learning_rate": 1.4215299158840566e-05, "loss": 0.0588, "step": 153240 }, { "epoch": 5.181040603130599, "grad_norm": 2.5844430923461914, "learning_rate": 1.4209525615934133e-05, "loss": 0.0526, "step": 153250 }, { "epoch": 5.181378680820853, "grad_norm": 1.7362326383590698, "learning_rate": 1.4203753051525803e-05, "loss": 0.0655, "step": 153260 }, { "epoch": 5.181716758511106, "grad_norm": 2.268726110458374, "learning_rate": 1.4197981465773436e-05, "loss": 0.0543, "step": 153270 }, { "epoch": 5.182054836201359, "grad_norm": 2.0953359603881836, "learning_rate": 1.4192210858834786e-05, "loss": 0.0529, "step": 153280 }, { "epoch": 5.182392913891612, "grad_norm": 1.975332498550415, "learning_rate": 1.4186441230867653e-05, "loss": 0.0539, "step": 153290 }, { "epoch": 5.182730991581866, "grad_norm": 3.4764461517333984, "learning_rate": 1.4180672582029768e-05, "loss": 0.0623, "step": 153300 }, { "epoch": 5.183069069272118, "grad_norm": 0.8642148375511169, "learning_rate": 1.417490491247882e-05, "loss": 0.0592, "step": 153310 }, { "epoch": 5.183407146962372, "grad_norm": 1.9307019710540771, "learning_rate": 1.4169138222372524e-05, "loss": 0.0568, "step": 153320 }, { "epoch": 5.183745224652625, "grad_norm": 2.486889600753784, "learning_rate": 1.4163372511868533e-05, "loss": 0.055, "step": 153330 }, { "epoch": 5.184083302342878, "grad_norm": 3.8227038383483887, "learning_rate": 1.415760778112446e-05, "loss": 0.0527, "step": 153340 }, { "epoch": 5.184421380033132, "grad_norm": 10.744603157043457, "learning_rate": 1.4151844030297934e-05, "loss": 0.0551, "step": 153350 }, { "epoch": 5.184759457723385, "grad_norm": 0.9236378073692322, "learning_rate": 1.4146081259546517e-05, "loss": 0.0526, "step": 153360 }, { "epoch": 5.185097535413638, "grad_norm": 2.096304416656494, "learning_rate": 1.4140319469027796e-05, "loss": 0.0554, "step": 153370 }, { "epoch": 5.185435613103891, "grad_norm": 1.5679285526275635, "learning_rate": 1.4134558658899232e-05, "loss": 0.0591, "step": 153380 }, { "epoch": 5.185773690794145, "grad_norm": 2.4804556369781494, "learning_rate": 1.4128798829318385e-05, "loss": 0.0587, "step": 153390 }, { "epoch": 5.186111768484397, "grad_norm": 1.752263069152832, "learning_rate": 1.4123039980442681e-05, "loss": 0.0578, "step": 153400 }, { "epoch": 5.186449846174651, "grad_norm": 0.9944906830787659, "learning_rate": 1.4117282112429602e-05, "loss": 0.0598, "step": 153410 }, { "epoch": 5.186787923864904, "grad_norm": 1.690108060836792, "learning_rate": 1.411152522543655e-05, "loss": 0.0525, "step": 153420 }, { "epoch": 5.187126001555157, "grad_norm": 2.083526849746704, "learning_rate": 1.4105769319620905e-05, "loss": 0.0555, "step": 153430 }, { "epoch": 5.1874640792454105, "grad_norm": 2.472588300704956, "learning_rate": 1.4100014395140049e-05, "loss": 0.0595, "step": 153440 }, { "epoch": 5.187802156935664, "grad_norm": 1.131553053855896, "learning_rate": 1.409426045215132e-05, "loss": 0.0508, "step": 153450 }, { "epoch": 5.188140234625917, "grad_norm": 1.8941891193389893, "learning_rate": 1.4088507490812015e-05, "loss": 0.0499, "step": 153460 }, { "epoch": 5.18847831231617, "grad_norm": 5.265225887298584, "learning_rate": 1.4082755511279411e-05, "loss": 0.06, "step": 153470 }, { "epoch": 5.188816390006424, "grad_norm": 0.7694581151008606, "learning_rate": 1.4077004513710778e-05, "loss": 0.0507, "step": 153480 }, { "epoch": 5.189154467696676, "grad_norm": 1.5721607208251953, "learning_rate": 1.4071254498263365e-05, "loss": 0.0623, "step": 153490 }, { "epoch": 5.18949254538693, "grad_norm": 2.3599064350128174, "learning_rate": 1.4065505465094348e-05, "loss": 0.0591, "step": 153500 }, { "epoch": 5.189830623077183, "grad_norm": 0.8712780475616455, "learning_rate": 1.405975741436092e-05, "loss": 0.0612, "step": 153510 }, { "epoch": 5.190168700767436, "grad_norm": 0.7923301458358765, "learning_rate": 1.4054010346220198e-05, "loss": 0.0544, "step": 153520 }, { "epoch": 5.1905067784576895, "grad_norm": 1.483587622642517, "learning_rate": 1.4048264260829353e-05, "loss": 0.0532, "step": 153530 }, { "epoch": 5.190844856147943, "grad_norm": 2.371459722518921, "learning_rate": 1.4042519158345452e-05, "loss": 0.0601, "step": 153540 }, { "epoch": 5.191182933838196, "grad_norm": 1.357189655303955, "learning_rate": 1.4036775038925554e-05, "loss": 0.0543, "step": 153550 }, { "epoch": 5.191521011528449, "grad_norm": 5.258639335632324, "learning_rate": 1.4031031902726727e-05, "loss": 0.0565, "step": 153560 }, { "epoch": 5.191859089218703, "grad_norm": 1.795247197151184, "learning_rate": 1.4025289749905967e-05, "loss": 0.06, "step": 153570 }, { "epoch": 5.192197166908955, "grad_norm": 1.2248095273971558, "learning_rate": 1.4019548580620296e-05, "loss": 0.0498, "step": 153580 }, { "epoch": 5.192535244599209, "grad_norm": 2.8444082736968994, "learning_rate": 1.4013808395026617e-05, "loss": 0.0587, "step": 153590 }, { "epoch": 5.192873322289462, "grad_norm": 2.0535147190093994, "learning_rate": 1.4008069193281903e-05, "loss": 0.0472, "step": 153600 }, { "epoch": 5.193211399979715, "grad_norm": 2.0363011360168457, "learning_rate": 1.4002330975543072e-05, "loss": 0.0528, "step": 153610 }, { "epoch": 5.193549477669968, "grad_norm": 1.7938152551651, "learning_rate": 1.3996593741966979e-05, "loss": 0.05, "step": 153620 }, { "epoch": 5.193887555360222, "grad_norm": 2.3323445320129395, "learning_rate": 1.3990857492710496e-05, "loss": 0.0554, "step": 153630 }, { "epoch": 5.194225633050475, "grad_norm": 2.9749233722686768, "learning_rate": 1.398512222793042e-05, "loss": 0.054, "step": 153640 }, { "epoch": 5.194563710740728, "grad_norm": 1.383242130279541, "learning_rate": 1.397938794778359e-05, "loss": 0.0539, "step": 153650 }, { "epoch": 5.194901788430982, "grad_norm": 1.7015022039413452, "learning_rate": 1.3973654652426766e-05, "loss": 0.0517, "step": 153660 }, { "epoch": 5.195239866121234, "grad_norm": 1.5102678537368774, "learning_rate": 1.3967922342016665e-05, "loss": 0.0581, "step": 153670 }, { "epoch": 5.195577943811488, "grad_norm": 1.3428667783737183, "learning_rate": 1.3962191016710053e-05, "loss": 0.0521, "step": 153680 }, { "epoch": 5.195916021501741, "grad_norm": 5.075372695922852, "learning_rate": 1.395646067666358e-05, "loss": 0.0488, "step": 153690 }, { "epoch": 5.196254099191995, "grad_norm": 4.390892028808594, "learning_rate": 1.3950731322033949e-05, "loss": 0.0561, "step": 153700 }, { "epoch": 5.196592176882247, "grad_norm": 0.5436379313468933, "learning_rate": 1.3945002952977782e-05, "loss": 0.0525, "step": 153710 }, { "epoch": 5.196930254572501, "grad_norm": 8.975680351257324, "learning_rate": 1.3939275569651672e-05, "loss": 0.0587, "step": 153720 }, { "epoch": 5.1972683322627535, "grad_norm": 1.0651112794876099, "learning_rate": 1.3933549172212234e-05, "loss": 0.0619, "step": 153730 }, { "epoch": 5.197606409953007, "grad_norm": 2.415602207183838, "learning_rate": 1.3927823760816012e-05, "loss": 0.0545, "step": 153740 }, { "epoch": 5.197944487643261, "grad_norm": 0.8738201856613159, "learning_rate": 1.3922099335619537e-05, "loss": 0.0582, "step": 153750 }, { "epoch": 5.198282565333514, "grad_norm": 1.8080474138259888, "learning_rate": 1.3916375896779298e-05, "loss": 0.0568, "step": 153760 }, { "epoch": 5.198620643023767, "grad_norm": 0.9909391403198242, "learning_rate": 1.3910653444451805e-05, "loss": 0.0576, "step": 153770 }, { "epoch": 5.19895872071402, "grad_norm": 0.9455398917198181, "learning_rate": 1.3904931978793467e-05, "loss": 0.0605, "step": 153780 }, { "epoch": 5.199296798404274, "grad_norm": 1.0386186838150024, "learning_rate": 1.389921149996075e-05, "loss": 0.0615, "step": 153790 }, { "epoch": 5.199634876094526, "grad_norm": 1.5883396863937378, "learning_rate": 1.3893492008110027e-05, "loss": 0.0579, "step": 153800 }, { "epoch": 5.19997295378478, "grad_norm": 80.4474868774414, "learning_rate": 1.388777350339765e-05, "loss": 0.0514, "step": 153810 }, { "epoch": 5.200311031475033, "grad_norm": 1.2000072002410889, "learning_rate": 1.3882055985979992e-05, "loss": 0.054, "step": 153820 }, { "epoch": 5.200649109165286, "grad_norm": 1.3412936925888062, "learning_rate": 1.3876339456013355e-05, "loss": 0.0519, "step": 153830 }, { "epoch": 5.2009871868555395, "grad_norm": 1.3549323081970215, "learning_rate": 1.387062391365403e-05, "loss": 0.0545, "step": 153840 }, { "epoch": 5.201325264545793, "grad_norm": 2.3879973888397217, "learning_rate": 1.3864909359058259e-05, "loss": 0.0591, "step": 153850 }, { "epoch": 5.201663342236046, "grad_norm": 3.0141146183013916, "learning_rate": 1.3859195792382284e-05, "loss": 0.0588, "step": 153860 }, { "epoch": 5.202001419926299, "grad_norm": 3.281985282897949, "learning_rate": 1.3853483213782358e-05, "loss": 0.0576, "step": 153870 }, { "epoch": 5.202339497616553, "grad_norm": 2.0580666065216064, "learning_rate": 1.3847771623414584e-05, "loss": 0.056, "step": 153880 }, { "epoch": 5.202677575306805, "grad_norm": 2.3727917671203613, "learning_rate": 1.384206102143517e-05, "loss": 0.0576, "step": 153890 }, { "epoch": 5.203015652997059, "grad_norm": 0.879542350769043, "learning_rate": 1.3836351408000208e-05, "loss": 0.054, "step": 153900 }, { "epoch": 5.203353730687312, "grad_norm": 1.7097785472869873, "learning_rate": 1.383064278326583e-05, "loss": 0.0526, "step": 153910 }, { "epoch": 5.203691808377565, "grad_norm": 1.731665015220642, "learning_rate": 1.3824935147388084e-05, "loss": 0.0564, "step": 153920 }, { "epoch": 5.2040298860678185, "grad_norm": 1.8095202445983887, "learning_rate": 1.3819228500523013e-05, "loss": 0.0627, "step": 153930 }, { "epoch": 5.204367963758072, "grad_norm": 1.841101884841919, "learning_rate": 1.3813522842826654e-05, "loss": 0.0617, "step": 153940 }, { "epoch": 5.204706041448325, "grad_norm": 1.0797985792160034, "learning_rate": 1.380781817445499e-05, "loss": 0.051, "step": 153950 }, { "epoch": 5.205044119138578, "grad_norm": 1.4961659908294678, "learning_rate": 1.3802114495563972e-05, "loss": 0.057, "step": 153960 }, { "epoch": 5.205382196828832, "grad_norm": 5.585787296295166, "learning_rate": 1.3796411806309533e-05, "loss": 0.0545, "step": 153970 }, { "epoch": 5.205720274519084, "grad_norm": 1.5035104751586914, "learning_rate": 1.3790710106847598e-05, "loss": 0.0569, "step": 153980 }, { "epoch": 5.206058352209338, "grad_norm": 1.7635854482650757, "learning_rate": 1.3785009397334059e-05, "loss": 0.0571, "step": 153990 }, { "epoch": 5.206396429899591, "grad_norm": 2.528285503387451, "learning_rate": 1.3779309677924757e-05, "loss": 0.0562, "step": 154000 }, { "epoch": 5.206734507589844, "grad_norm": 1.2768986225128174, "learning_rate": 1.3773610948775518e-05, "loss": 0.0564, "step": 154010 }, { "epoch": 5.207072585280097, "grad_norm": 4.204798221588135, "learning_rate": 1.3767913210042132e-05, "loss": 0.0541, "step": 154020 }, { "epoch": 5.207410662970351, "grad_norm": 0.8111608028411865, "learning_rate": 1.3762216461880406e-05, "loss": 0.055, "step": 154030 }, { "epoch": 5.207748740660604, "grad_norm": 2.058224678039551, "learning_rate": 1.3756520704446063e-05, "loss": 0.0524, "step": 154040 }, { "epoch": 5.208086818350857, "grad_norm": 0.8137797117233276, "learning_rate": 1.3750825937894818e-05, "loss": 0.0539, "step": 154050 }, { "epoch": 5.208424896041111, "grad_norm": 2.033372402191162, "learning_rate": 1.3745132162382385e-05, "loss": 0.0571, "step": 154060 }, { "epoch": 5.208762973731363, "grad_norm": 2.4678773880004883, "learning_rate": 1.3739439378064405e-05, "loss": 0.05, "step": 154070 }, { "epoch": 5.209101051421617, "grad_norm": 2.332206964492798, "learning_rate": 1.3733747585096562e-05, "loss": 0.0591, "step": 154080 }, { "epoch": 5.20943912911187, "grad_norm": 11.256566047668457, "learning_rate": 1.3728056783634403e-05, "loss": 0.0542, "step": 154090 }, { "epoch": 5.209777206802123, "grad_norm": 1.4687976837158203, "learning_rate": 1.3722366973833545e-05, "loss": 0.0556, "step": 154100 }, { "epoch": 5.210115284492376, "grad_norm": 2.828995943069458, "learning_rate": 1.3716678155849555e-05, "loss": 0.0573, "step": 154110 }, { "epoch": 5.21045336218263, "grad_norm": 2.0668141841888428, "learning_rate": 1.3710990329837957e-05, "loss": 0.0571, "step": 154120 }, { "epoch": 5.2107914398728825, "grad_norm": 0.8497850894927979, "learning_rate": 1.3705303495954252e-05, "loss": 0.0558, "step": 154130 }, { "epoch": 5.211129517563136, "grad_norm": 1.0853149890899658, "learning_rate": 1.3699617654353902e-05, "loss": 0.0546, "step": 154140 }, { "epoch": 5.2114675952533895, "grad_norm": 1.1245770454406738, "learning_rate": 1.3693932805192378e-05, "loss": 0.0578, "step": 154150 }, { "epoch": 5.211805672943642, "grad_norm": 0.8823725581169128, "learning_rate": 1.3688248948625093e-05, "loss": 0.0623, "step": 154160 }, { "epoch": 5.212143750633896, "grad_norm": 1.045266032218933, "learning_rate": 1.3682566084807424e-05, "loss": 0.0576, "step": 154170 }, { "epoch": 5.212481828324149, "grad_norm": 1.4119296073913574, "learning_rate": 1.367688421389477e-05, "loss": 0.064, "step": 154180 }, { "epoch": 5.212819906014402, "grad_norm": 1.3838295936584473, "learning_rate": 1.3671203336042443e-05, "loss": 0.053, "step": 154190 }, { "epoch": 5.213157983704655, "grad_norm": 3.1870508193969727, "learning_rate": 1.3665523451405782e-05, "loss": 0.0558, "step": 154200 }, { "epoch": 5.213496061394909, "grad_norm": 1.1189414262771606, "learning_rate": 1.3659844560140057e-05, "loss": 0.0573, "step": 154210 }, { "epoch": 5.2138341390851615, "grad_norm": 1.6454236507415771, "learning_rate": 1.365416666240053e-05, "loss": 0.0581, "step": 154220 }, { "epoch": 5.214172216775415, "grad_norm": 4.898404121398926, "learning_rate": 1.364848975834242e-05, "loss": 0.0582, "step": 154230 }, { "epoch": 5.2145102944656685, "grad_norm": 2.4381680488586426, "learning_rate": 1.3642813848120956e-05, "loss": 0.059, "step": 154240 }, { "epoch": 5.214848372155921, "grad_norm": 1.2839789390563965, "learning_rate": 1.3637138931891307e-05, "loss": 0.056, "step": 154250 }, { "epoch": 5.215186449846175, "grad_norm": 4.397793292999268, "learning_rate": 1.3631465009808597e-05, "loss": 0.0513, "step": 154260 }, { "epoch": 5.215524527536428, "grad_norm": 1.803473711013794, "learning_rate": 1.3625792082027988e-05, "loss": 0.0543, "step": 154270 }, { "epoch": 5.215862605226681, "grad_norm": 0.9031655788421631, "learning_rate": 1.3620120148704547e-05, "loss": 0.0545, "step": 154280 }, { "epoch": 5.216200682916934, "grad_norm": 3.9888007640838623, "learning_rate": 1.361444920999338e-05, "loss": 0.0591, "step": 154290 }, { "epoch": 5.216538760607188, "grad_norm": 1.7354404926300049, "learning_rate": 1.3608779266049476e-05, "loss": 0.0508, "step": 154300 }, { "epoch": 5.21687683829744, "grad_norm": 5.722426891326904, "learning_rate": 1.3603110317027874e-05, "loss": 0.0575, "step": 154310 }, { "epoch": 5.217214915987694, "grad_norm": 3.4453327655792236, "learning_rate": 1.3597442363083574e-05, "loss": 0.0513, "step": 154320 }, { "epoch": 5.2175529936779474, "grad_norm": 3.3956046104431152, "learning_rate": 1.3591775404371526e-05, "loss": 0.0551, "step": 154330 }, { "epoch": 5.2178910713682, "grad_norm": 1.7502864599227905, "learning_rate": 1.3586109441046657e-05, "loss": 0.0537, "step": 154340 }, { "epoch": 5.218229149058454, "grad_norm": 1.8962386846542358, "learning_rate": 1.3580444473263865e-05, "loss": 0.0566, "step": 154350 }, { "epoch": 5.218567226748707, "grad_norm": 2.805600166320801, "learning_rate": 1.3574780501178036e-05, "loss": 0.0478, "step": 154360 }, { "epoch": 5.21890530443896, "grad_norm": 1.2865920066833496, "learning_rate": 1.3569117524944058e-05, "loss": 0.0526, "step": 154370 }, { "epoch": 5.219243382129213, "grad_norm": 4.967658519744873, "learning_rate": 1.3563455544716692e-05, "loss": 0.0572, "step": 154380 }, { "epoch": 5.219581459819467, "grad_norm": 1.6362026929855347, "learning_rate": 1.3557794560650772e-05, "loss": 0.055, "step": 154390 }, { "epoch": 5.219919537509719, "grad_norm": 2.2882297039031982, "learning_rate": 1.3552134572901042e-05, "loss": 0.0537, "step": 154400 }, { "epoch": 5.220257615199973, "grad_norm": 4.2070722579956055, "learning_rate": 1.3546475581622281e-05, "loss": 0.0625, "step": 154410 }, { "epoch": 5.220595692890226, "grad_norm": 2.641094923019409, "learning_rate": 1.354081758696918e-05, "loss": 0.0583, "step": 154420 }, { "epoch": 5.220933770580479, "grad_norm": 8.202475547790527, "learning_rate": 1.3535160589096413e-05, "loss": 0.0553, "step": 154430 }, { "epoch": 5.2212718482707325, "grad_norm": 4.435732364654541, "learning_rate": 1.352950458815867e-05, "loss": 0.0518, "step": 154440 }, { "epoch": 5.221609925960986, "grad_norm": 6.166194915771484, "learning_rate": 1.3523849584310571e-05, "loss": 0.0571, "step": 154450 }, { "epoch": 5.221948003651239, "grad_norm": 0.8464879989624023, "learning_rate": 1.3518195577706728e-05, "loss": 0.0519, "step": 154460 }, { "epoch": 5.222286081341492, "grad_norm": 1.495042085647583, "learning_rate": 1.351254256850169e-05, "loss": 0.0604, "step": 154470 }, { "epoch": 5.222624159031746, "grad_norm": 2.066664934158325, "learning_rate": 1.3506890556850044e-05, "loss": 0.0527, "step": 154480 }, { "epoch": 5.222962236721998, "grad_norm": 4.103691101074219, "learning_rate": 1.3501239542906291e-05, "loss": 0.0616, "step": 154490 }, { "epoch": 5.223300314412252, "grad_norm": 1.4205124378204346, "learning_rate": 1.3495589526824947e-05, "loss": 0.0604, "step": 154500 }, { "epoch": 5.223638392102505, "grad_norm": 1.5516835451126099, "learning_rate": 1.3489940508760474e-05, "loss": 0.0561, "step": 154510 }, { "epoch": 5.223976469792758, "grad_norm": 1.4348617792129517, "learning_rate": 1.3484292488867295e-05, "loss": 0.0543, "step": 154520 }, { "epoch": 5.2243145474830115, "grad_norm": 1.853437900543213, "learning_rate": 1.347864546729986e-05, "loss": 0.0578, "step": 154530 }, { "epoch": 5.224652625173265, "grad_norm": 1.6103320121765137, "learning_rate": 1.3472999444212531e-05, "loss": 0.0549, "step": 154540 }, { "epoch": 5.224990702863518, "grad_norm": 1.9148564338684082, "learning_rate": 1.3467354419759664e-05, "loss": 0.0566, "step": 154550 }, { "epoch": 5.225328780553771, "grad_norm": 8.538545608520508, "learning_rate": 1.3461710394095617e-05, "loss": 0.0643, "step": 154560 }, { "epoch": 5.225666858244025, "grad_norm": 2.4144515991210938, "learning_rate": 1.3456067367374664e-05, "loss": 0.0627, "step": 154570 }, { "epoch": 5.226004935934277, "grad_norm": 1.26972496509552, "learning_rate": 1.3450425339751133e-05, "loss": 0.0598, "step": 154580 }, { "epoch": 5.226343013624531, "grad_norm": 2.0439374446868896, "learning_rate": 1.344478431137921e-05, "loss": 0.0569, "step": 154590 }, { "epoch": 5.226681091314784, "grad_norm": 2.7787973880767822, "learning_rate": 1.3439144282413168e-05, "loss": 0.0559, "step": 154600 }, { "epoch": 5.227019169005037, "grad_norm": 2.5548269748687744, "learning_rate": 1.3433505253007172e-05, "loss": 0.0588, "step": 154610 }, { "epoch": 5.2273572466952904, "grad_norm": 1.4681247472763062, "learning_rate": 1.3427867223315416e-05, "loss": 0.0589, "step": 154620 }, { "epoch": 5.227695324385544, "grad_norm": 1.6304441690444946, "learning_rate": 1.3422230193492031e-05, "loss": 0.0544, "step": 154630 }, { "epoch": 5.228033402075797, "grad_norm": 1.0005953311920166, "learning_rate": 1.3416594163691121e-05, "loss": 0.0522, "step": 154640 }, { "epoch": 5.22837147976605, "grad_norm": 1.2816640138626099, "learning_rate": 1.3410959134066791e-05, "loss": 0.0539, "step": 154650 }, { "epoch": 5.228709557456304, "grad_norm": 1.322757601737976, "learning_rate": 1.3405325104773097e-05, "loss": 0.0588, "step": 154660 }, { "epoch": 5.229047635146557, "grad_norm": 1.2839224338531494, "learning_rate": 1.3399692075964066e-05, "loss": 0.0551, "step": 154670 }, { "epoch": 5.22938571283681, "grad_norm": 0.9377827048301697, "learning_rate": 1.3394060047793688e-05, "loss": 0.0584, "step": 154680 }, { "epoch": 5.229723790527063, "grad_norm": 3.1915676593780518, "learning_rate": 1.3388429020415955e-05, "loss": 0.0555, "step": 154690 }, { "epoch": 5.230061868217316, "grad_norm": 1.6066333055496216, "learning_rate": 1.3382798993984835e-05, "loss": 0.0567, "step": 154700 }, { "epoch": 5.230399945907569, "grad_norm": 7.338314056396484, "learning_rate": 1.337716996865423e-05, "loss": 0.0559, "step": 154710 }, { "epoch": 5.230738023597823, "grad_norm": 1.3257558345794678, "learning_rate": 1.3371541944578042e-05, "loss": 0.0521, "step": 154720 }, { "epoch": 5.231076101288076, "grad_norm": 4.260495185852051, "learning_rate": 1.3365914921910121e-05, "loss": 0.0559, "step": 154730 }, { "epoch": 5.231414178978329, "grad_norm": 1.4340574741363525, "learning_rate": 1.336028890080434e-05, "loss": 0.053, "step": 154740 }, { "epoch": 5.231752256668583, "grad_norm": 3.0930800437927246, "learning_rate": 1.3354663881414497e-05, "loss": 0.0567, "step": 154750 }, { "epoch": 5.232090334358836, "grad_norm": 0.9418935775756836, "learning_rate": 1.3349039863894358e-05, "loss": 0.0679, "step": 154760 }, { "epoch": 5.232428412049089, "grad_norm": 1.1262882947921753, "learning_rate": 1.3343416848397716e-05, "loss": 0.0572, "step": 154770 }, { "epoch": 5.232766489739342, "grad_norm": 1.1734075546264648, "learning_rate": 1.3337794835078271e-05, "loss": 0.0539, "step": 154780 }, { "epoch": 5.233104567429596, "grad_norm": 1.4369789361953735, "learning_rate": 1.3332173824089778e-05, "loss": 0.0564, "step": 154790 }, { "epoch": 5.233442645119848, "grad_norm": 3.1348612308502197, "learning_rate": 1.3326553815585841e-05, "loss": 0.0518, "step": 154800 }, { "epoch": 5.233780722810102, "grad_norm": 1.3673131465911865, "learning_rate": 1.3320934809720143e-05, "loss": 0.0519, "step": 154810 }, { "epoch": 5.234118800500355, "grad_norm": 4.26170539855957, "learning_rate": 1.3315316806646328e-05, "loss": 0.0538, "step": 154820 }, { "epoch": 5.234456878190608, "grad_norm": 1.12274169921875, "learning_rate": 1.3309699806517967e-05, "loss": 0.0543, "step": 154830 }, { "epoch": 5.2347949558808615, "grad_norm": 8.299077033996582, "learning_rate": 1.3304083809488627e-05, "loss": 0.0577, "step": 154840 }, { "epoch": 5.235133033571115, "grad_norm": 1.7222886085510254, "learning_rate": 1.329846881571184e-05, "loss": 0.0477, "step": 154850 }, { "epoch": 5.235471111261368, "grad_norm": 0.9144846200942993, "learning_rate": 1.3292854825341138e-05, "loss": 0.052, "step": 154860 }, { "epoch": 5.235809188951621, "grad_norm": 1.0314984321594238, "learning_rate": 1.3287241838529996e-05, "loss": 0.0543, "step": 154870 }, { "epoch": 5.236147266641875, "grad_norm": 0.8665504455566406, "learning_rate": 1.3281629855431854e-05, "loss": 0.0552, "step": 154880 }, { "epoch": 5.236485344332127, "grad_norm": 2.321354627609253, "learning_rate": 1.3276018876200175e-05, "loss": 0.0582, "step": 154890 }, { "epoch": 5.236823422022381, "grad_norm": 2.4757533073425293, "learning_rate": 1.3270408900988324e-05, "loss": 0.0555, "step": 154900 }, { "epoch": 5.237161499712634, "grad_norm": 2.806579113006592, "learning_rate": 1.3264799929949712e-05, "loss": 0.0506, "step": 154910 }, { "epoch": 5.237499577402887, "grad_norm": 0.9350165128707886, "learning_rate": 1.3259191963237672e-05, "loss": 0.0456, "step": 154920 }, { "epoch": 5.2378376550931405, "grad_norm": 3.2813446521759033, "learning_rate": 1.3253585001005503e-05, "loss": 0.0527, "step": 154930 }, { "epoch": 5.238175732783394, "grad_norm": 1.9652036428451538, "learning_rate": 1.3247979043406534e-05, "loss": 0.0568, "step": 154940 }, { "epoch": 5.238513810473647, "grad_norm": 1.2175300121307373, "learning_rate": 1.3242374090594007e-05, "loss": 0.0558, "step": 154950 }, { "epoch": 5.2388518881639, "grad_norm": 0.9582188725471497, "learning_rate": 1.3236770142721162e-05, "loss": 0.0631, "step": 154960 }, { "epoch": 5.239189965854154, "grad_norm": 0.818858802318573, "learning_rate": 1.3231167199941197e-05, "loss": 0.0544, "step": 154970 }, { "epoch": 5.239528043544406, "grad_norm": 3.496182918548584, "learning_rate": 1.3225565262407324e-05, "loss": 0.0575, "step": 154980 }, { "epoch": 5.23986612123466, "grad_norm": 1.6921576261520386, "learning_rate": 1.3219964330272661e-05, "loss": 0.0509, "step": 154990 }, { "epoch": 5.240204198924913, "grad_norm": 3.2299766540527344, "learning_rate": 1.3214364403690377e-05, "loss": 0.052, "step": 155000 }, { "epoch": 5.240542276615166, "grad_norm": 2.0855190753936768, "learning_rate": 1.3208765482813546e-05, "loss": 0.0584, "step": 155010 }, { "epoch": 5.240880354305419, "grad_norm": 0.4822808504104614, "learning_rate": 1.3203167567795233e-05, "loss": 0.0568, "step": 155020 }, { "epoch": 5.241218431995673, "grad_norm": 1.5964053869247437, "learning_rate": 1.3197570658788504e-05, "loss": 0.0542, "step": 155030 }, { "epoch": 5.241556509685926, "grad_norm": 1.359748363494873, "learning_rate": 1.3191974755946374e-05, "loss": 0.0597, "step": 155040 }, { "epoch": 5.241894587376179, "grad_norm": 0.9851025938987732, "learning_rate": 1.3186379859421821e-05, "loss": 0.0626, "step": 155050 }, { "epoch": 5.242232665066433, "grad_norm": 0.9171826243400574, "learning_rate": 1.3180785969367803e-05, "loss": 0.0514, "step": 155060 }, { "epoch": 5.242570742756685, "grad_norm": 2.247685670852661, "learning_rate": 1.317519308593726e-05, "loss": 0.054, "step": 155070 }, { "epoch": 5.242908820446939, "grad_norm": 2.0400025844573975, "learning_rate": 1.3169601209283133e-05, "loss": 0.056, "step": 155080 }, { "epoch": 5.243246898137192, "grad_norm": 1.0896661281585693, "learning_rate": 1.3164010339558246e-05, "loss": 0.0505, "step": 155090 }, { "epoch": 5.243584975827445, "grad_norm": 0.7646083831787109, "learning_rate": 1.315842047691549e-05, "loss": 0.0496, "step": 155100 }, { "epoch": 5.243923053517698, "grad_norm": 0.5699598789215088, "learning_rate": 1.3152831621507666e-05, "loss": 0.0457, "step": 155110 }, { "epoch": 5.244261131207952, "grad_norm": 1.6792545318603516, "learning_rate": 1.3147243773487595e-05, "loss": 0.0523, "step": 155120 }, { "epoch": 5.2445992088982045, "grad_norm": 1.5496327877044678, "learning_rate": 1.3141656933008034e-05, "loss": 0.0554, "step": 155130 }, { "epoch": 5.244937286588458, "grad_norm": 0.9642779231071472, "learning_rate": 1.313607110022171e-05, "loss": 0.0625, "step": 155140 }, { "epoch": 5.2452753642787115, "grad_norm": 1.2373228073120117, "learning_rate": 1.3130486275281367e-05, "loss": 0.0541, "step": 155150 }, { "epoch": 5.245613441968964, "grad_norm": 1.584525227546692, "learning_rate": 1.3124902458339678e-05, "loss": 0.0529, "step": 155160 }, { "epoch": 5.245951519659218, "grad_norm": 1.8487682342529297, "learning_rate": 1.3119319649549305e-05, "loss": 0.0578, "step": 155170 }, { "epoch": 5.246289597349471, "grad_norm": 1.0967731475830078, "learning_rate": 1.3113737849062862e-05, "loss": 0.0536, "step": 155180 }, { "epoch": 5.246627675039724, "grad_norm": 1.5810184478759766, "learning_rate": 1.3108157057032966e-05, "loss": 0.0621, "step": 155190 }, { "epoch": 5.246965752729977, "grad_norm": 1.2063523530960083, "learning_rate": 1.3102577273612216e-05, "loss": 0.0499, "step": 155200 }, { "epoch": 5.247303830420231, "grad_norm": 2.069453001022339, "learning_rate": 1.309699849895314e-05, "loss": 0.0567, "step": 155210 }, { "epoch": 5.2476419081104835, "grad_norm": 5.237931728363037, "learning_rate": 1.309142073320826e-05, "loss": 0.0578, "step": 155220 }, { "epoch": 5.247979985800737, "grad_norm": 7.120741367340088, "learning_rate": 1.3085843976530055e-05, "loss": 0.0591, "step": 155230 }, { "epoch": 5.2483180634909905, "grad_norm": 2.2587289810180664, "learning_rate": 1.3080268229071024e-05, "loss": 0.0598, "step": 155240 }, { "epoch": 5.248656141181243, "grad_norm": 1.0974024534225464, "learning_rate": 1.3074693490983591e-05, "loss": 0.0598, "step": 155250 }, { "epoch": 5.248994218871497, "grad_norm": 2.2121798992156982, "learning_rate": 1.3069119762420146e-05, "loss": 0.0601, "step": 155260 }, { "epoch": 5.24933229656175, "grad_norm": 0.8239262104034424, "learning_rate": 1.3063547043533114e-05, "loss": 0.0574, "step": 155270 }, { "epoch": 5.249670374252003, "grad_norm": 1.0920226573944092, "learning_rate": 1.305797533447481e-05, "loss": 0.055, "step": 155280 }, { "epoch": 5.250008451942256, "grad_norm": 2.4147844314575195, "learning_rate": 1.3052404635397613e-05, "loss": 0.0543, "step": 155290 }, { "epoch": 5.25034652963251, "grad_norm": 1.3589636087417603, "learning_rate": 1.3046834946453763e-05, "loss": 0.0538, "step": 155300 }, { "epoch": 5.250684607322762, "grad_norm": 1.925589919090271, "learning_rate": 1.3041266267795565e-05, "loss": 0.0533, "step": 155310 }, { "epoch": 5.251022685013016, "grad_norm": 2.223843574523926, "learning_rate": 1.3035698599575274e-05, "loss": 0.0632, "step": 155320 }, { "epoch": 5.2513607627032695, "grad_norm": 2.507404088973999, "learning_rate": 1.3030131941945095e-05, "loss": 0.0534, "step": 155330 }, { "epoch": 5.251698840393522, "grad_norm": 2.2184524536132812, "learning_rate": 1.3024566295057223e-05, "loss": 0.056, "step": 155340 }, { "epoch": 5.252036918083776, "grad_norm": 2.1096458435058594, "learning_rate": 1.3019001659063801e-05, "loss": 0.0496, "step": 155350 }, { "epoch": 5.252374995774029, "grad_norm": 1.1493942737579346, "learning_rate": 1.3013438034116998e-05, "loss": 0.05, "step": 155360 }, { "epoch": 5.252713073464282, "grad_norm": 2.122880220413208, "learning_rate": 1.3007875420368904e-05, "loss": 0.0571, "step": 155370 }, { "epoch": 5.253051151154535, "grad_norm": 2.0755414962768555, "learning_rate": 1.3002313817971584e-05, "loss": 0.0557, "step": 155380 }, { "epoch": 5.253389228844789, "grad_norm": 2.61114239692688, "learning_rate": 1.2996753227077113e-05, "loss": 0.0563, "step": 155390 }, { "epoch": 5.253727306535041, "grad_norm": 3.790133476257324, "learning_rate": 1.2991193647837501e-05, "loss": 0.054, "step": 155400 }, { "epoch": 5.254065384225295, "grad_norm": 7.828193664550781, "learning_rate": 1.2985635080404762e-05, "loss": 0.0548, "step": 155410 }, { "epoch": 5.254403461915548, "grad_norm": 2.2960822582244873, "learning_rate": 1.2980077524930856e-05, "loss": 0.0562, "step": 155420 }, { "epoch": 5.254741539605801, "grad_norm": 1.0010627508163452, "learning_rate": 1.2974520981567718e-05, "loss": 0.0587, "step": 155430 }, { "epoch": 5.2550796172960546, "grad_norm": 1.6696701049804688, "learning_rate": 1.2968965450467258e-05, "loss": 0.0565, "step": 155440 }, { "epoch": 5.255417694986308, "grad_norm": 1.574523687362671, "learning_rate": 1.2963410931781378e-05, "loss": 0.0552, "step": 155450 }, { "epoch": 5.255755772676561, "grad_norm": 1.661720633506775, "learning_rate": 1.2957857425661934e-05, "loss": 0.0537, "step": 155460 }, { "epoch": 5.256093850366814, "grad_norm": 1.3997784852981567, "learning_rate": 1.2952304932260739e-05, "loss": 0.0579, "step": 155470 }, { "epoch": 5.256431928057068, "grad_norm": 4.514571189880371, "learning_rate": 1.294675345172962e-05, "loss": 0.0517, "step": 155480 }, { "epoch": 5.25677000574732, "grad_norm": 1.6321905851364136, "learning_rate": 1.2941202984220329e-05, "loss": 0.0565, "step": 155490 }, { "epoch": 5.257108083437574, "grad_norm": 2.7038862705230713, "learning_rate": 1.2935653529884639e-05, "loss": 0.0565, "step": 155500 }, { "epoch": 5.257446161127827, "grad_norm": 1.2014529705047607, "learning_rate": 1.2930105088874262e-05, "loss": 0.0574, "step": 155510 }, { "epoch": 5.25778423881808, "grad_norm": 2.324488878250122, "learning_rate": 1.2924557661340864e-05, "loss": 0.0575, "step": 155520 }, { "epoch": 5.2581223165083335, "grad_norm": 0.5541766285896301, "learning_rate": 1.2919011247436147e-05, "loss": 0.0598, "step": 155530 }, { "epoch": 5.258460394198587, "grad_norm": 0.6885697245597839, "learning_rate": 1.2913465847311735e-05, "loss": 0.0608, "step": 155540 }, { "epoch": 5.25879847188884, "grad_norm": 3.7130067348480225, "learning_rate": 1.2907921461119232e-05, "loss": 0.0562, "step": 155550 }, { "epoch": 5.259136549579093, "grad_norm": 2.2955191135406494, "learning_rate": 1.2902378089010208e-05, "loss": 0.0527, "step": 155560 }, { "epoch": 5.259474627269347, "grad_norm": 0.547805905342102, "learning_rate": 1.2896835731136225e-05, "loss": 0.0543, "step": 155570 }, { "epoch": 5.2598127049596, "grad_norm": 0.7208532094955444, "learning_rate": 1.289129438764885e-05, "loss": 0.0517, "step": 155580 }, { "epoch": 5.260150782649853, "grad_norm": 1.088046908378601, "learning_rate": 1.2885754058699512e-05, "loss": 0.0591, "step": 155590 }, { "epoch": 5.260488860340106, "grad_norm": 0.5144289135932922, "learning_rate": 1.2880214744439733e-05, "loss": 0.0501, "step": 155600 }, { "epoch": 5.260826938030359, "grad_norm": 2.6207187175750732, "learning_rate": 1.2874676445020922e-05, "loss": 0.0496, "step": 155610 }, { "epoch": 5.2611650157206125, "grad_norm": 2.7035863399505615, "learning_rate": 1.2869139160594523e-05, "loss": 0.0541, "step": 155620 }, { "epoch": 5.261503093410866, "grad_norm": 0.7351069450378418, "learning_rate": 1.286360289131191e-05, "loss": 0.0562, "step": 155630 }, { "epoch": 5.2618411711011195, "grad_norm": 1.0032035112380981, "learning_rate": 1.2858067637324428e-05, "loss": 0.0564, "step": 155640 }, { "epoch": 5.262179248791372, "grad_norm": 1.1836395263671875, "learning_rate": 1.2852533398783435e-05, "loss": 0.0462, "step": 155650 }, { "epoch": 5.262517326481626, "grad_norm": 0.8070502281188965, "learning_rate": 1.2847000175840229e-05, "loss": 0.046, "step": 155660 }, { "epoch": 5.262855404171878, "grad_norm": 1.6161452531814575, "learning_rate": 1.2841467968646076e-05, "loss": 0.0512, "step": 155670 }, { "epoch": 5.263193481862132, "grad_norm": 1.6572469472885132, "learning_rate": 1.2835936777352214e-05, "loss": 0.0555, "step": 155680 }, { "epoch": 5.263531559552385, "grad_norm": 1.5916668176651, "learning_rate": 1.2830406602109896e-05, "loss": 0.0554, "step": 155690 }, { "epoch": 5.263869637242639, "grad_norm": 1.1665518283843994, "learning_rate": 1.2824877443070282e-05, "loss": 0.0499, "step": 155700 }, { "epoch": 5.264207714932891, "grad_norm": 2.838524103164673, "learning_rate": 1.2819349300384565e-05, "loss": 0.0605, "step": 155710 }, { "epoch": 5.264545792623145, "grad_norm": 0.8648155927658081, "learning_rate": 1.2813822174203877e-05, "loss": 0.0516, "step": 155720 }, { "epoch": 5.2648838703133976, "grad_norm": 1.616317629814148, "learning_rate": 1.2808296064679298e-05, "loss": 0.055, "step": 155730 }, { "epoch": 5.265221948003651, "grad_norm": 0.9290036559104919, "learning_rate": 1.2802770971961952e-05, "loss": 0.0507, "step": 155740 }, { "epoch": 5.265560025693905, "grad_norm": 0.7284842729568481, "learning_rate": 1.2797246896202869e-05, "loss": 0.055, "step": 155750 }, { "epoch": 5.265898103384158, "grad_norm": 1.1634589433670044, "learning_rate": 1.2791723837553066e-05, "loss": 0.0478, "step": 155760 }, { "epoch": 5.266236181074411, "grad_norm": 1.4773774147033691, "learning_rate": 1.2786201796163571e-05, "loss": 0.0585, "step": 155770 }, { "epoch": 5.266574258764664, "grad_norm": 3.758727550506592, "learning_rate": 1.2780680772185328e-05, "loss": 0.0573, "step": 155780 }, { "epoch": 5.266912336454918, "grad_norm": 3.027157783508301, "learning_rate": 1.2775160765769313e-05, "loss": 0.0574, "step": 155790 }, { "epoch": 5.26725041414517, "grad_norm": 1.7439626455307007, "learning_rate": 1.2769641777066389e-05, "loss": 0.0506, "step": 155800 }, { "epoch": 5.267588491835424, "grad_norm": 1.6148903369903564, "learning_rate": 1.2764123806227486e-05, "loss": 0.0572, "step": 155810 }, { "epoch": 5.267926569525677, "grad_norm": 1.6486804485321045, "learning_rate": 1.2758606853403438e-05, "loss": 0.0497, "step": 155820 }, { "epoch": 5.26826464721593, "grad_norm": 1.950381875038147, "learning_rate": 1.2753090918745104e-05, "loss": 0.0605, "step": 155830 }, { "epoch": 5.2686027249061835, "grad_norm": 1.3843241930007935, "learning_rate": 1.2747576002403267e-05, "loss": 0.0545, "step": 155840 }, { "epoch": 5.268940802596437, "grad_norm": 0.9496616125106812, "learning_rate": 1.2742062104528691e-05, "loss": 0.054, "step": 155850 }, { "epoch": 5.26927888028669, "grad_norm": 1.733742594718933, "learning_rate": 1.273654922527216e-05, "loss": 0.0514, "step": 155860 }, { "epoch": 5.269616957976943, "grad_norm": 0.6351923942565918, "learning_rate": 1.2731037364784371e-05, "loss": 0.0572, "step": 155870 }, { "epoch": 5.269955035667197, "grad_norm": 2.5965981483459473, "learning_rate": 1.2725526523216019e-05, "loss": 0.0493, "step": 155880 }, { "epoch": 5.270293113357449, "grad_norm": 1.4486266374588013, "learning_rate": 1.2720016700717752e-05, "loss": 0.0512, "step": 155890 }, { "epoch": 5.270631191047703, "grad_norm": 2.9726099967956543, "learning_rate": 1.2714507897440226e-05, "loss": 0.0533, "step": 155900 }, { "epoch": 5.270969268737956, "grad_norm": 2.8440442085266113, "learning_rate": 1.2709000113534058e-05, "loss": 0.0498, "step": 155910 }, { "epoch": 5.271307346428209, "grad_norm": 2.579615592956543, "learning_rate": 1.270349334914982e-05, "loss": 0.0555, "step": 155920 }, { "epoch": 5.2716454241184625, "grad_norm": 0.6369153261184692, "learning_rate": 1.2697987604438061e-05, "loss": 0.0545, "step": 155930 }, { "epoch": 5.271983501808716, "grad_norm": 0.8339663743972778, "learning_rate": 1.2692482879549295e-05, "loss": 0.0537, "step": 155940 }, { "epoch": 5.272321579498969, "grad_norm": 16.66429328918457, "learning_rate": 1.2686979174634045e-05, "loss": 0.051, "step": 155950 }, { "epoch": 5.272659657189222, "grad_norm": 2.5782630443573, "learning_rate": 1.2681476489842763e-05, "loss": 0.0564, "step": 155960 }, { "epoch": 5.272997734879476, "grad_norm": 1.2065068483352661, "learning_rate": 1.2675974825325882e-05, "loss": 0.0571, "step": 155970 }, { "epoch": 5.273335812569728, "grad_norm": 0.839428722858429, "learning_rate": 1.2670474181233843e-05, "loss": 0.0509, "step": 155980 }, { "epoch": 5.273673890259982, "grad_norm": 0.7082859873771667, "learning_rate": 1.2664974557716996e-05, "loss": 0.0511, "step": 155990 }, { "epoch": 5.274011967950235, "grad_norm": 2.0181500911712646, "learning_rate": 1.2659475954925754e-05, "loss": 0.0576, "step": 156000 }, { "epoch": 5.274350045640488, "grad_norm": 0.823239266872406, "learning_rate": 1.2653978373010383e-05, "loss": 0.0592, "step": 156010 }, { "epoch": 5.274688123330741, "grad_norm": 0.6506600975990295, "learning_rate": 1.2648481812121204e-05, "loss": 0.0497, "step": 156020 }, { "epoch": 5.275026201020995, "grad_norm": 1.0935379266738892, "learning_rate": 1.2642986272408524e-05, "loss": 0.0516, "step": 156030 }, { "epoch": 5.275364278711248, "grad_norm": 0.9955401420593262, "learning_rate": 1.2637491754022562e-05, "loss": 0.0505, "step": 156040 }, { "epoch": 5.275702356401501, "grad_norm": 1.938071846961975, "learning_rate": 1.2631998257113536e-05, "loss": 0.0563, "step": 156050 }, { "epoch": 5.276040434091755, "grad_norm": 1.332034707069397, "learning_rate": 1.2626505781831626e-05, "loss": 0.0553, "step": 156060 }, { "epoch": 5.276378511782007, "grad_norm": 1.15915846824646, "learning_rate": 1.2621014328327024e-05, "loss": 0.0592, "step": 156070 }, { "epoch": 5.276716589472261, "grad_norm": 2.3706793785095215, "learning_rate": 1.2615523896749849e-05, "loss": 0.0564, "step": 156080 }, { "epoch": 5.277054667162514, "grad_norm": 22.798677444458008, "learning_rate": 1.2610034487250188e-05, "loss": 0.0543, "step": 156090 }, { "epoch": 5.277392744852767, "grad_norm": 2.6588168144226074, "learning_rate": 1.2604546099978153e-05, "loss": 0.0612, "step": 156100 }, { "epoch": 5.27773082254302, "grad_norm": 1.2805315256118774, "learning_rate": 1.2599058735083763e-05, "loss": 0.0529, "step": 156110 }, { "epoch": 5.278068900233274, "grad_norm": 1.496004581451416, "learning_rate": 1.2593572392717074e-05, "loss": 0.0572, "step": 156120 }, { "epoch": 5.2784069779235265, "grad_norm": 1.262248158454895, "learning_rate": 1.258808707302806e-05, "loss": 0.0513, "step": 156130 }, { "epoch": 5.27874505561378, "grad_norm": 2.598463296890259, "learning_rate": 1.2582602776166674e-05, "loss": 0.0499, "step": 156140 }, { "epoch": 5.279083133304034, "grad_norm": 1.487646460533142, "learning_rate": 1.2577119502282886e-05, "loss": 0.0592, "step": 156150 }, { "epoch": 5.279421210994286, "grad_norm": 1.4520410299301147, "learning_rate": 1.2571637251526598e-05, "loss": 0.056, "step": 156160 }, { "epoch": 5.27975928868454, "grad_norm": 3.7994561195373535, "learning_rate": 1.2566156024047682e-05, "loss": 0.0549, "step": 156170 }, { "epoch": 5.280097366374793, "grad_norm": 2.4967291355133057, "learning_rate": 1.2560675819995982e-05, "loss": 0.0488, "step": 156180 }, { "epoch": 5.280435444065046, "grad_norm": 0.8707157373428345, "learning_rate": 1.2555196639521355e-05, "loss": 0.0533, "step": 156190 }, { "epoch": 5.280773521755299, "grad_norm": 1.7061681747436523, "learning_rate": 1.2549718482773571e-05, "loss": 0.0573, "step": 156200 }, { "epoch": 5.281111599445553, "grad_norm": 0.9770539402961731, "learning_rate": 1.2544241349902424e-05, "loss": 0.0565, "step": 156210 }, { "epoch": 5.2814496771358055, "grad_norm": 0.531427800655365, "learning_rate": 1.2538765241057654e-05, "loss": 0.0555, "step": 156220 }, { "epoch": 5.281787754826059, "grad_norm": 1.2527260780334473, "learning_rate": 1.253329015638895e-05, "loss": 0.0518, "step": 156230 }, { "epoch": 5.2821258325163125, "grad_norm": 1.0200095176696777, "learning_rate": 1.252781609604603e-05, "loss": 0.0571, "step": 156240 }, { "epoch": 5.282463910206565, "grad_norm": 4.388890743255615, "learning_rate": 1.2522343060178543e-05, "loss": 0.0561, "step": 156250 }, { "epoch": 5.282801987896819, "grad_norm": 1.4063422679901123, "learning_rate": 1.2516871048936118e-05, "loss": 0.0566, "step": 156260 }, { "epoch": 5.283140065587072, "grad_norm": 2.196969509124756, "learning_rate": 1.2511400062468343e-05, "loss": 0.0511, "step": 156270 }, { "epoch": 5.283478143277325, "grad_norm": 1.3209962844848633, "learning_rate": 1.2505930100924806e-05, "loss": 0.0545, "step": 156280 }, { "epoch": 5.283816220967578, "grad_norm": 0.8759955167770386, "learning_rate": 1.2500461164455079e-05, "loss": 0.0567, "step": 156290 }, { "epoch": 5.284154298657832, "grad_norm": 8.32209587097168, "learning_rate": 1.2494993253208632e-05, "loss": 0.0545, "step": 156300 }, { "epoch": 5.284492376348084, "grad_norm": 0.6960728764533997, "learning_rate": 1.2489526367334991e-05, "loss": 0.0504, "step": 156310 }, { "epoch": 5.284830454038338, "grad_norm": 1.0050297975540161, "learning_rate": 1.2484060506983597e-05, "loss": 0.0516, "step": 156320 }, { "epoch": 5.2851685317285915, "grad_norm": 0.5470283031463623, "learning_rate": 1.2478595672303905e-05, "loss": 0.0581, "step": 156330 }, { "epoch": 5.285506609418844, "grad_norm": 0.6762330532073975, "learning_rate": 1.2473131863445309e-05, "loss": 0.0499, "step": 156340 }, { "epoch": 5.285844687109098, "grad_norm": 0.948797881603241, "learning_rate": 1.2467669080557181e-05, "loss": 0.0561, "step": 156350 }, { "epoch": 5.286182764799351, "grad_norm": 1.5308748483657837, "learning_rate": 1.2462207323788893e-05, "loss": 0.0551, "step": 156360 }, { "epoch": 5.286520842489604, "grad_norm": 0.8406094312667847, "learning_rate": 1.2456746593289753e-05, "loss": 0.0489, "step": 156370 }, { "epoch": 5.286858920179857, "grad_norm": 0.6921539902687073, "learning_rate": 1.2451286889209058e-05, "loss": 0.0575, "step": 156380 }, { "epoch": 5.287196997870111, "grad_norm": 1.0575028657913208, "learning_rate": 1.2445828211696053e-05, "loss": 0.0524, "step": 156390 }, { "epoch": 5.287535075560363, "grad_norm": 2.6288390159606934, "learning_rate": 1.2440370560900005e-05, "loss": 0.0533, "step": 156400 }, { "epoch": 5.287873153250617, "grad_norm": 1.5672025680541992, "learning_rate": 1.2434913936970122e-05, "loss": 0.0548, "step": 156410 }, { "epoch": 5.28821123094087, "grad_norm": 1.1058783531188965, "learning_rate": 1.2429458340055583e-05, "loss": 0.057, "step": 156420 }, { "epoch": 5.288549308631123, "grad_norm": 2.2781970500946045, "learning_rate": 1.2424003770305542e-05, "loss": 0.0545, "step": 156430 }, { "epoch": 5.288887386321377, "grad_norm": 1.6973472833633423, "learning_rate": 1.2418550227869103e-05, "loss": 0.0514, "step": 156440 }, { "epoch": 5.28922546401163, "grad_norm": 0.856371283531189, "learning_rate": 1.2413097712895393e-05, "loss": 0.0578, "step": 156450 }, { "epoch": 5.289563541701883, "grad_norm": 0.9866112470626831, "learning_rate": 1.2407646225533476e-05, "loss": 0.0543, "step": 156460 }, { "epoch": 5.289901619392136, "grad_norm": 1.403830885887146, "learning_rate": 1.2402195765932373e-05, "loss": 0.063, "step": 156470 }, { "epoch": 5.29023969708239, "grad_norm": 1.8966916799545288, "learning_rate": 1.2396746334241127e-05, "loss": 0.0547, "step": 156480 }, { "epoch": 5.290577774772642, "grad_norm": 1.619748592376709, "learning_rate": 1.2391297930608697e-05, "loss": 0.0491, "step": 156490 }, { "epoch": 5.290915852462896, "grad_norm": 2.3566505908966064, "learning_rate": 1.238585055518408e-05, "loss": 0.0589, "step": 156500 }, { "epoch": 5.291253930153149, "grad_norm": 1.1721590757369995, "learning_rate": 1.2380404208116148e-05, "loss": 0.0566, "step": 156510 }, { "epoch": 5.291592007843402, "grad_norm": 2.349924087524414, "learning_rate": 1.2374958889553828e-05, "loss": 0.0567, "step": 156520 }, { "epoch": 5.2919300855336555, "grad_norm": 0.83970707654953, "learning_rate": 1.236951459964602e-05, "loss": 0.0554, "step": 156530 }, { "epoch": 5.292268163223909, "grad_norm": 2.319551467895508, "learning_rate": 1.2364071338541539e-05, "loss": 0.0544, "step": 156540 }, { "epoch": 5.2926062409141625, "grad_norm": 1.693235993385315, "learning_rate": 1.2358629106389213e-05, "loss": 0.0563, "step": 156550 }, { "epoch": 5.292944318604415, "grad_norm": 0.8878909349441528, "learning_rate": 1.2353187903337809e-05, "loss": 0.0534, "step": 156560 }, { "epoch": 5.293282396294669, "grad_norm": 2.27337384223938, "learning_rate": 1.2347747729536119e-05, "loss": 0.0576, "step": 156570 }, { "epoch": 5.293620473984921, "grad_norm": 1.655716896057129, "learning_rate": 1.2342308585132861e-05, "loss": 0.0526, "step": 156580 }, { "epoch": 5.293958551675175, "grad_norm": 1.4476600885391235, "learning_rate": 1.2336870470276724e-05, "loss": 0.064, "step": 156590 }, { "epoch": 5.294296629365428, "grad_norm": 1.4560438394546509, "learning_rate": 1.2331433385116415e-05, "loss": 0.0547, "step": 156600 }, { "epoch": 5.294634707055682, "grad_norm": 0.8085801005363464, "learning_rate": 1.2325997329800549e-05, "loss": 0.0533, "step": 156610 }, { "epoch": 5.2949727847459345, "grad_norm": 1.7666780948638916, "learning_rate": 1.2320562304477778e-05, "loss": 0.0521, "step": 156620 }, { "epoch": 5.295310862436188, "grad_norm": 1.233705997467041, "learning_rate": 1.2315128309296681e-05, "loss": 0.0537, "step": 156630 }, { "epoch": 5.295648940126441, "grad_norm": 1.0145059823989868, "learning_rate": 1.2309695344405814e-05, "loss": 0.0537, "step": 156640 }, { "epoch": 5.295987017816694, "grad_norm": 1.251590609550476, "learning_rate": 1.2304263409953703e-05, "loss": 0.0537, "step": 156650 }, { "epoch": 5.296325095506948, "grad_norm": 2.714597702026367, "learning_rate": 1.2298832506088887e-05, "loss": 0.0553, "step": 156660 }, { "epoch": 5.296663173197201, "grad_norm": 1.1496124267578125, "learning_rate": 1.2293402632959822e-05, "loss": 0.055, "step": 156670 }, { "epoch": 5.297001250887454, "grad_norm": 2.0578525066375732, "learning_rate": 1.2287973790714952e-05, "loss": 0.0526, "step": 156680 }, { "epoch": 5.297339328577707, "grad_norm": 0.9067716002464294, "learning_rate": 1.2282545979502725e-05, "loss": 0.0573, "step": 156690 }, { "epoch": 5.29767740626796, "grad_norm": 2.0047521591186523, "learning_rate": 1.2277119199471504e-05, "loss": 0.0531, "step": 156700 }, { "epoch": 5.298015483958213, "grad_norm": 2.782819986343384, "learning_rate": 1.2271693450769688e-05, "loss": 0.0605, "step": 156710 }, { "epoch": 5.298353561648467, "grad_norm": 1.5981003046035767, "learning_rate": 1.22662687335456e-05, "loss": 0.0544, "step": 156720 }, { "epoch": 5.2986916393387204, "grad_norm": 1.3761190176010132, "learning_rate": 1.2260845047947533e-05, "loss": 0.0469, "step": 156730 }, { "epoch": 5.299029717028973, "grad_norm": 1.2230629920959473, "learning_rate": 1.22554223941238e-05, "loss": 0.0538, "step": 156740 }, { "epoch": 5.299367794719227, "grad_norm": 2.238948345184326, "learning_rate": 1.2250000772222641e-05, "loss": 0.0571, "step": 156750 }, { "epoch": 5.29970587240948, "grad_norm": 2.5200209617614746, "learning_rate": 1.2244580182392274e-05, "loss": 0.0529, "step": 156760 }, { "epoch": 5.300043950099733, "grad_norm": 1.5823508501052856, "learning_rate": 1.2239160624780894e-05, "loss": 0.0544, "step": 156770 }, { "epoch": 5.300382027789986, "grad_norm": 1.281950831413269, "learning_rate": 1.2233742099536665e-05, "loss": 0.0567, "step": 156780 }, { "epoch": 5.30072010548024, "grad_norm": 2.634007692337036, "learning_rate": 1.2228324606807778e-05, "loss": 0.0615, "step": 156790 }, { "epoch": 5.301058183170492, "grad_norm": 3.3597571849823, "learning_rate": 1.2222908146742273e-05, "loss": 0.0512, "step": 156800 }, { "epoch": 5.301396260860746, "grad_norm": 1.7674179077148438, "learning_rate": 1.2217492719488282e-05, "loss": 0.0554, "step": 156810 }, { "epoch": 5.301734338550999, "grad_norm": 0.9300185441970825, "learning_rate": 1.2212078325193832e-05, "loss": 0.0559, "step": 156820 }, { "epoch": 5.302072416241252, "grad_norm": 3.3526971340179443, "learning_rate": 1.2206664964006975e-05, "loss": 0.0536, "step": 156830 }, { "epoch": 5.3024104939315055, "grad_norm": 0.9612993001937866, "learning_rate": 1.2201252636075705e-05, "loss": 0.0588, "step": 156840 }, { "epoch": 5.302748571621759, "grad_norm": 0.8292630910873413, "learning_rate": 1.219584134154797e-05, "loss": 0.052, "step": 156850 }, { "epoch": 5.303086649312012, "grad_norm": 2.557426929473877, "learning_rate": 1.2190431080571747e-05, "loss": 0.0529, "step": 156860 }, { "epoch": 5.303424727002265, "grad_norm": 1.1430994272232056, "learning_rate": 1.2185021853294932e-05, "loss": 0.0525, "step": 156870 }, { "epoch": 5.303762804692519, "grad_norm": 11.529040336608887, "learning_rate": 1.2179613659865413e-05, "loss": 0.0569, "step": 156880 }, { "epoch": 5.304100882382771, "grad_norm": 13.27650260925293, "learning_rate": 1.2174206500431035e-05, "loss": 0.0619, "step": 156890 }, { "epoch": 5.304438960073025, "grad_norm": 2.841343879699707, "learning_rate": 1.2168800375139655e-05, "loss": 0.0534, "step": 156900 }, { "epoch": 5.304777037763278, "grad_norm": 1.1318246126174927, "learning_rate": 1.2163395284139046e-05, "loss": 0.0511, "step": 156910 }, { "epoch": 5.305115115453531, "grad_norm": 0.571349561214447, "learning_rate": 1.2157991227577009e-05, "loss": 0.0564, "step": 156920 }, { "epoch": 5.3054531931437845, "grad_norm": 1.409796118736267, "learning_rate": 1.2152588205601273e-05, "loss": 0.0608, "step": 156930 }, { "epoch": 5.305791270834038, "grad_norm": 1.1058340072631836, "learning_rate": 1.2147186218359547e-05, "loss": 0.0517, "step": 156940 }, { "epoch": 5.306129348524291, "grad_norm": 0.9310762286186218, "learning_rate": 1.2141785265999538e-05, "loss": 0.0557, "step": 156950 }, { "epoch": 5.306467426214544, "grad_norm": 0.7837740182876587, "learning_rate": 1.2136385348668899e-05, "loss": 0.0511, "step": 156960 }, { "epoch": 5.306805503904798, "grad_norm": 1.331998586654663, "learning_rate": 1.2130986466515249e-05, "loss": 0.0593, "step": 156970 }, { "epoch": 5.30714358159505, "grad_norm": 1.3838237524032593, "learning_rate": 1.2125588619686213e-05, "loss": 0.062, "step": 156980 }, { "epoch": 5.307481659285304, "grad_norm": 2.5206377506256104, "learning_rate": 1.2120191808329345e-05, "loss": 0.0579, "step": 156990 }, { "epoch": 5.307819736975557, "grad_norm": 1.7177801132202148, "learning_rate": 1.2114796032592229e-05, "loss": 0.0553, "step": 157000 }, { "epoch": 5.30815781466581, "grad_norm": 1.0694481134414673, "learning_rate": 1.2109401292622325e-05, "loss": 0.0566, "step": 157010 }, { "epoch": 5.3084958923560635, "grad_norm": 1.2039928436279297, "learning_rate": 1.2104007588567173e-05, "loss": 0.0529, "step": 157020 }, { "epoch": 5.308833970046317, "grad_norm": 1.0410809516906738, "learning_rate": 1.2098614920574202e-05, "loss": 0.052, "step": 157030 }, { "epoch": 5.30917204773657, "grad_norm": 0.9464603662490845, "learning_rate": 1.2093223288790873e-05, "loss": 0.0605, "step": 157040 }, { "epoch": 5.309510125426823, "grad_norm": 0.981722354888916, "learning_rate": 1.2087832693364576e-05, "loss": 0.0501, "step": 157050 }, { "epoch": 5.309848203117077, "grad_norm": 3.534052610397339, "learning_rate": 1.2082443134442678e-05, "loss": 0.0545, "step": 157060 }, { "epoch": 5.310186280807329, "grad_norm": 0.8831528425216675, "learning_rate": 1.2077054612172555e-05, "loss": 0.0538, "step": 157070 }, { "epoch": 5.310524358497583, "grad_norm": 1.3057676553726196, "learning_rate": 1.2071667126701514e-05, "loss": 0.0619, "step": 157080 }, { "epoch": 5.310862436187836, "grad_norm": 0.854574978351593, "learning_rate": 1.2066280678176845e-05, "loss": 0.0565, "step": 157090 }, { "epoch": 5.311200513878089, "grad_norm": 1.4878188371658325, "learning_rate": 1.2060895266745793e-05, "loss": 0.0549, "step": 157100 }, { "epoch": 5.311538591568342, "grad_norm": 1.0225590467453003, "learning_rate": 1.205551089255561e-05, "loss": 0.0532, "step": 157110 }, { "epoch": 5.311876669258596, "grad_norm": 1.379388451576233, "learning_rate": 1.2050127555753522e-05, "loss": 0.0552, "step": 157120 }, { "epoch": 5.3122147469488485, "grad_norm": 6.402449607849121, "learning_rate": 1.2044745256486689e-05, "loss": 0.0508, "step": 157130 }, { "epoch": 5.312552824639102, "grad_norm": 3.308112859725952, "learning_rate": 1.2039363994902264e-05, "loss": 0.0551, "step": 157140 }, { "epoch": 5.312890902329356, "grad_norm": 0.9459111094474792, "learning_rate": 1.2033983771147344e-05, "loss": 0.0517, "step": 157150 }, { "epoch": 5.313228980019608, "grad_norm": 0.9584141969680786, "learning_rate": 1.2028604585369069e-05, "loss": 0.0494, "step": 157160 }, { "epoch": 5.313567057709862, "grad_norm": 0.4982077181339264, "learning_rate": 1.2023226437714474e-05, "loss": 0.0539, "step": 157170 }, { "epoch": 5.313905135400115, "grad_norm": 1.1462368965148926, "learning_rate": 1.2017849328330583e-05, "loss": 0.0545, "step": 157180 }, { "epoch": 5.314243213090368, "grad_norm": 1.3019423484802246, "learning_rate": 1.2012473257364442e-05, "loss": 0.05, "step": 157190 }, { "epoch": 5.314581290780621, "grad_norm": 0.7637320756912231, "learning_rate": 1.2007098224962993e-05, "loss": 0.0482, "step": 157200 }, { "epoch": 5.314919368470875, "grad_norm": 0.9089078903198242, "learning_rate": 1.2001724231273237e-05, "loss": 0.0595, "step": 157210 }, { "epoch": 5.3152574461611275, "grad_norm": 3.961552858352661, "learning_rate": 1.1996351276442037e-05, "loss": 0.0571, "step": 157220 }, { "epoch": 5.315595523851381, "grad_norm": 0.7636926770210266, "learning_rate": 1.1990979360616317e-05, "loss": 0.0518, "step": 157230 }, { "epoch": 5.3159336015416345, "grad_norm": 1.1375058889389038, "learning_rate": 1.198560848394295e-05, "loss": 0.0564, "step": 157240 }, { "epoch": 5.316271679231887, "grad_norm": 0.8606793880462646, "learning_rate": 1.198023864656877e-05, "loss": 0.0536, "step": 157250 }, { "epoch": 5.316609756922141, "grad_norm": 1.4462734460830688, "learning_rate": 1.197486984864058e-05, "loss": 0.0525, "step": 157260 }, { "epoch": 5.316947834612394, "grad_norm": 1.6057381629943848, "learning_rate": 1.1969502090305152e-05, "loss": 0.0581, "step": 157270 }, { "epoch": 5.317285912302647, "grad_norm": 0.8296048641204834, "learning_rate": 1.1964135371709262e-05, "loss": 0.0548, "step": 157280 }, { "epoch": 5.3176239899929, "grad_norm": 1.715634822845459, "learning_rate": 1.1958769692999621e-05, "loss": 0.0565, "step": 157290 }, { "epoch": 5.317962067683154, "grad_norm": 3.3927621841430664, "learning_rate": 1.1953405054322913e-05, "loss": 0.0539, "step": 157300 }, { "epoch": 5.3183001453734065, "grad_norm": 4.275928497314453, "learning_rate": 1.1948041455825826e-05, "loss": 0.0481, "step": 157310 }, { "epoch": 5.31863822306366, "grad_norm": 0.5217121839523315, "learning_rate": 1.194267889765498e-05, "loss": 0.0548, "step": 157320 }, { "epoch": 5.3189763007539135, "grad_norm": 1.353476881980896, "learning_rate": 1.1937317379957013e-05, "loss": 0.0517, "step": 157330 }, { "epoch": 5.319314378444166, "grad_norm": 0.8022008538246155, "learning_rate": 1.1931956902878483e-05, "loss": 0.0552, "step": 157340 }, { "epoch": 5.31965245613442, "grad_norm": 29.408864974975586, "learning_rate": 1.1926597466565936e-05, "loss": 0.0604, "step": 157350 }, { "epoch": 5.319990533824673, "grad_norm": 0.8912310600280762, "learning_rate": 1.192123907116593e-05, "loss": 0.0542, "step": 157360 }, { "epoch": 5.320328611514926, "grad_norm": 3.3421695232391357, "learning_rate": 1.1915881716824939e-05, "loss": 0.0532, "step": 157370 }, { "epoch": 5.320666689205179, "grad_norm": 2.5883724689483643, "learning_rate": 1.1910525403689437e-05, "loss": 0.0531, "step": 157380 }, { "epoch": 5.321004766895433, "grad_norm": 4.510624408721924, "learning_rate": 1.1905170131905846e-05, "loss": 0.0612, "step": 157390 }, { "epoch": 5.321342844585685, "grad_norm": 3.794806957244873, "learning_rate": 1.1899815901620604e-05, "loss": 0.0574, "step": 157400 }, { "epoch": 5.321680922275939, "grad_norm": 4.955822467803955, "learning_rate": 1.1894462712980065e-05, "loss": 0.0585, "step": 157410 }, { "epoch": 5.322018999966192, "grad_norm": 1.471339225769043, "learning_rate": 1.1889110566130618e-05, "loss": 0.0603, "step": 157420 }, { "epoch": 5.322357077656445, "grad_norm": 0.8078528642654419, "learning_rate": 1.1883759461218568e-05, "loss": 0.0537, "step": 157430 }, { "epoch": 5.322695155346699, "grad_norm": 1.0791022777557373, "learning_rate": 1.1878409398390206e-05, "loss": 0.0547, "step": 157440 }, { "epoch": 5.323033233036952, "grad_norm": 1.5016155242919922, "learning_rate": 1.1873060377791816e-05, "loss": 0.0584, "step": 157450 }, { "epoch": 5.323371310727205, "grad_norm": 0.7655881643295288, "learning_rate": 1.1867712399569642e-05, "loss": 0.0523, "step": 157460 }, { "epoch": 5.323709388417458, "grad_norm": 1.2248377799987793, "learning_rate": 1.1862365463869878e-05, "loss": 0.0514, "step": 157470 }, { "epoch": 5.324047466107712, "grad_norm": 1.3597112894058228, "learning_rate": 1.18570195708387e-05, "loss": 0.0609, "step": 157480 }, { "epoch": 5.324385543797964, "grad_norm": 1.4644008874893188, "learning_rate": 1.185167472062228e-05, "loss": 0.0512, "step": 157490 }, { "epoch": 5.324723621488218, "grad_norm": 1.5847293138504028, "learning_rate": 1.1846330913366766e-05, "loss": 0.0531, "step": 157500 }, { "epoch": 5.325061699178471, "grad_norm": 2.8707170486450195, "learning_rate": 1.1840988149218207e-05, "loss": 0.0576, "step": 157510 }, { "epoch": 5.325399776868725, "grad_norm": 0.6864296197891235, "learning_rate": 1.1835646428322711e-05, "loss": 0.055, "step": 157520 }, { "epoch": 5.3257378545589775, "grad_norm": 0.9095708727836609, "learning_rate": 1.1830305750826287e-05, "loss": 0.0534, "step": 157530 }, { "epoch": 5.326075932249231, "grad_norm": 1.699470043182373, "learning_rate": 1.1824966116874975e-05, "loss": 0.0552, "step": 157540 }, { "epoch": 5.326414009939484, "grad_norm": 0.7981312274932861, "learning_rate": 1.1819627526614752e-05, "loss": 0.0623, "step": 157550 }, { "epoch": 5.326752087629737, "grad_norm": 0.928968608379364, "learning_rate": 1.1814289980191556e-05, "loss": 0.0507, "step": 157560 }, { "epoch": 5.327090165319991, "grad_norm": 0.9331954121589661, "learning_rate": 1.1808953477751339e-05, "loss": 0.0543, "step": 157570 }, { "epoch": 5.327428243010244, "grad_norm": 0.8628295063972473, "learning_rate": 1.1803618019439983e-05, "loss": 0.0571, "step": 157580 }, { "epoch": 5.327766320700497, "grad_norm": 2.60916805267334, "learning_rate": 1.1798283605403365e-05, "loss": 0.0546, "step": 157590 }, { "epoch": 5.32810439839075, "grad_norm": 2.0645251274108887, "learning_rate": 1.1792950235787303e-05, "loss": 0.0562, "step": 157600 }, { "epoch": 5.328442476081003, "grad_norm": 1.0275768041610718, "learning_rate": 1.1787617910737636e-05, "loss": 0.048, "step": 157610 }, { "epoch": 5.3287805537712565, "grad_norm": 0.9384077787399292, "learning_rate": 1.1782286630400147e-05, "loss": 0.0557, "step": 157620 }, { "epoch": 5.32911863146151, "grad_norm": 1.4149683713912964, "learning_rate": 1.1776956394920586e-05, "loss": 0.0491, "step": 157630 }, { "epoch": 5.3294567091517635, "grad_norm": 1.4943934679031372, "learning_rate": 1.1771627204444675e-05, "loss": 0.0527, "step": 157640 }, { "epoch": 5.329794786842016, "grad_norm": 1.1632375717163086, "learning_rate": 1.1766299059118102e-05, "loss": 0.0518, "step": 157650 }, { "epoch": 5.33013286453227, "grad_norm": 2.7712411880493164, "learning_rate": 1.1760971959086564e-05, "loss": 0.0508, "step": 157660 }, { "epoch": 5.330470942222522, "grad_norm": 2.354149103164673, "learning_rate": 1.1755645904495682e-05, "loss": 0.0494, "step": 157670 }, { "epoch": 5.330809019912776, "grad_norm": 1.406001329421997, "learning_rate": 1.1750320895491068e-05, "loss": 0.0565, "step": 157680 }, { "epoch": 5.331147097603029, "grad_norm": 0.8773280382156372, "learning_rate": 1.1744996932218316e-05, "loss": 0.0576, "step": 157690 }, { "epoch": 5.331485175293283, "grad_norm": 1.0135753154754639, "learning_rate": 1.1739674014822966e-05, "loss": 0.0511, "step": 157700 }, { "epoch": 5.331823252983535, "grad_norm": 2.549739122390747, "learning_rate": 1.1734352143450584e-05, "loss": 0.0538, "step": 157710 }, { "epoch": 5.332161330673789, "grad_norm": 0.8652122616767883, "learning_rate": 1.1729031318246608e-05, "loss": 0.0532, "step": 157720 }, { "epoch": 5.3324994083640425, "grad_norm": 3.2654314041137695, "learning_rate": 1.1723711539356541e-05, "loss": 0.0505, "step": 157730 }, { "epoch": 5.332837486054295, "grad_norm": 1.5460611581802368, "learning_rate": 1.1718392806925831e-05, "loss": 0.0568, "step": 157740 }, { "epoch": 5.333175563744549, "grad_norm": 1.043407917022705, "learning_rate": 1.1713075121099887e-05, "loss": 0.065, "step": 157750 }, { "epoch": 5.333513641434802, "grad_norm": 3.5139899253845215, "learning_rate": 1.1707758482024079e-05, "loss": 0.0609, "step": 157760 }, { "epoch": 5.333851719125055, "grad_norm": 1.5287926197052002, "learning_rate": 1.1702442889843756e-05, "loss": 0.0568, "step": 157770 }, { "epoch": 5.334189796815308, "grad_norm": 1.3570871353149414, "learning_rate": 1.1697128344704273e-05, "loss": 0.0515, "step": 157780 }, { "epoch": 5.334527874505562, "grad_norm": 0.5913051962852478, "learning_rate": 1.169181484675091e-05, "loss": 0.0478, "step": 157790 }, { "epoch": 5.334865952195814, "grad_norm": 0.9713449478149414, "learning_rate": 1.1686502396128918e-05, "loss": 0.0602, "step": 157800 }, { "epoch": 5.335204029886068, "grad_norm": 2.6581578254699707, "learning_rate": 1.1681190992983577e-05, "loss": 0.0571, "step": 157810 }, { "epoch": 5.335542107576321, "grad_norm": 0.6039271354675293, "learning_rate": 1.1675880637460062e-05, "loss": 0.0525, "step": 157820 }, { "epoch": 5.335880185266574, "grad_norm": 5.794922351837158, "learning_rate": 1.1670571329703589e-05, "loss": 0.0525, "step": 157830 }, { "epoch": 5.3362182629568276, "grad_norm": 1.3171229362487793, "learning_rate": 1.1665263069859295e-05, "loss": 0.0617, "step": 157840 }, { "epoch": 5.336556340647081, "grad_norm": 1.4654514789581299, "learning_rate": 1.1659955858072308e-05, "loss": 0.0561, "step": 157850 }, { "epoch": 5.336894418337334, "grad_norm": 0.8605064749717712, "learning_rate": 1.1654649694487707e-05, "loss": 0.0566, "step": 157860 }, { "epoch": 5.337232496027587, "grad_norm": 4.245710849761963, "learning_rate": 1.1649344579250593e-05, "loss": 0.0603, "step": 157870 }, { "epoch": 5.337570573717841, "grad_norm": 1.3084386587142944, "learning_rate": 1.1644040512505994e-05, "loss": 0.0498, "step": 157880 }, { "epoch": 5.337908651408093, "grad_norm": 7.151374816894531, "learning_rate": 1.16387374943989e-05, "loss": 0.0535, "step": 157890 }, { "epoch": 5.338246729098347, "grad_norm": 1.106492280960083, "learning_rate": 1.1633435525074327e-05, "loss": 0.0563, "step": 157900 }, { "epoch": 5.3385848067886, "grad_norm": 0.8883970379829407, "learning_rate": 1.1628134604677193e-05, "loss": 0.0561, "step": 157910 }, { "epoch": 5.338922884478853, "grad_norm": 0.7616534233093262, "learning_rate": 1.1622834733352467e-05, "loss": 0.056, "step": 157920 }, { "epoch": 5.3392609621691065, "grad_norm": 2.63767671585083, "learning_rate": 1.1617535911245015e-05, "loss": 0.0573, "step": 157930 }, { "epoch": 5.33959903985936, "grad_norm": 1.1784378290176392, "learning_rate": 1.1612238138499698e-05, "loss": 0.0551, "step": 157940 }, { "epoch": 5.339937117549613, "grad_norm": 1.7946699857711792, "learning_rate": 1.1606941415261385e-05, "loss": 0.0557, "step": 157950 }, { "epoch": 5.340275195239866, "grad_norm": 4.456087112426758, "learning_rate": 1.1601645741674872e-05, "loss": 0.0508, "step": 157960 }, { "epoch": 5.34061327293012, "grad_norm": 0.6939095258712769, "learning_rate": 1.1596351117884935e-05, "loss": 0.0526, "step": 157970 }, { "epoch": 5.340951350620372, "grad_norm": 1.0752546787261963, "learning_rate": 1.1591057544036315e-05, "loss": 0.0576, "step": 157980 }, { "epoch": 5.341289428310626, "grad_norm": 1.040043592453003, "learning_rate": 1.1585765020273758e-05, "loss": 0.0576, "step": 157990 }, { "epoch": 5.341627506000879, "grad_norm": 1.1587786674499512, "learning_rate": 1.1580473546741983e-05, "loss": 0.0595, "step": 158000 }, { "epoch": 5.341965583691132, "grad_norm": 4.635684967041016, "learning_rate": 1.1575183123585592e-05, "loss": 0.0637, "step": 158010 }, { "epoch": 5.3423036613813855, "grad_norm": 0.7798436880111694, "learning_rate": 1.1569893750949279e-05, "loss": 0.0556, "step": 158020 }, { "epoch": 5.342641739071639, "grad_norm": 1.1801519393920898, "learning_rate": 1.1564605428977615e-05, "loss": 0.0546, "step": 158030 }, { "epoch": 5.342979816761892, "grad_norm": 1.8363641500473022, "learning_rate": 1.1559318157815213e-05, "loss": 0.0637, "step": 158040 }, { "epoch": 5.343317894452145, "grad_norm": 1.2629770040512085, "learning_rate": 1.1554031937606614e-05, "loss": 0.0551, "step": 158050 }, { "epoch": 5.343655972142399, "grad_norm": 2.619649887084961, "learning_rate": 1.1548746768496326e-05, "loss": 0.0574, "step": 158060 }, { "epoch": 5.343994049832651, "grad_norm": 0.9517643451690674, "learning_rate": 1.1543462650628866e-05, "loss": 0.0551, "step": 158070 }, { "epoch": 5.344332127522905, "grad_norm": 2.0851213932037354, "learning_rate": 1.153817958414869e-05, "loss": 0.052, "step": 158080 }, { "epoch": 5.344670205213158, "grad_norm": 1.3858379125595093, "learning_rate": 1.1532897569200234e-05, "loss": 0.0578, "step": 158090 }, { "epoch": 5.345008282903411, "grad_norm": 1.4630876779556274, "learning_rate": 1.1527616605927899e-05, "loss": 0.0535, "step": 158100 }, { "epoch": 5.345346360593664, "grad_norm": 1.3132355213165283, "learning_rate": 1.1522336694476083e-05, "loss": 0.055, "step": 158110 }, { "epoch": 5.345684438283918, "grad_norm": 1.468520164489746, "learning_rate": 1.1517057834989109e-05, "loss": 0.0556, "step": 158120 }, { "epoch": 5.3460225159741706, "grad_norm": 0.8769325613975525, "learning_rate": 1.1511780027611335e-05, "loss": 0.0558, "step": 158130 }, { "epoch": 5.346360593664424, "grad_norm": 1.2927898168563843, "learning_rate": 1.1506503272487035e-05, "loss": 0.0545, "step": 158140 }, { "epoch": 5.346698671354678, "grad_norm": 1.9188839197158813, "learning_rate": 1.1501227569760459e-05, "loss": 0.0562, "step": 158150 }, { "epoch": 5.34703674904493, "grad_norm": 1.9192253351211548, "learning_rate": 1.149595291957587e-05, "loss": 0.0505, "step": 158160 }, { "epoch": 5.347374826735184, "grad_norm": 0.9941936731338501, "learning_rate": 1.1490679322077464e-05, "loss": 0.0549, "step": 158170 }, { "epoch": 5.347712904425437, "grad_norm": 2.673953056335449, "learning_rate": 1.1485406777409407e-05, "loss": 0.0536, "step": 158180 }, { "epoch": 5.34805098211569, "grad_norm": 0.7416162490844727, "learning_rate": 1.1480135285715871e-05, "loss": 0.0522, "step": 158190 }, { "epoch": 5.348389059805943, "grad_norm": 2.033343553543091, "learning_rate": 1.1474864847140953e-05, "loss": 0.0533, "step": 158200 }, { "epoch": 5.348727137496197, "grad_norm": 4.404703140258789, "learning_rate": 1.1469595461828781e-05, "loss": 0.0587, "step": 158210 }, { "epoch": 5.3490652151864495, "grad_norm": 1.697493553161621, "learning_rate": 1.1464327129923369e-05, "loss": 0.0608, "step": 158220 }, { "epoch": 5.349403292876703, "grad_norm": 1.1931304931640625, "learning_rate": 1.1459059851568788e-05, "loss": 0.0508, "step": 158230 }, { "epoch": 5.3497413705669565, "grad_norm": 0.9245196580886841, "learning_rate": 1.1453793626909021e-05, "loss": 0.058, "step": 158240 }, { "epoch": 5.350079448257209, "grad_norm": 5.603155612945557, "learning_rate": 1.1448528456088064e-05, "loss": 0.0563, "step": 158250 }, { "epoch": 5.350417525947463, "grad_norm": 1.480354905128479, "learning_rate": 1.144326433924986e-05, "loss": 0.0471, "step": 158260 }, { "epoch": 5.350755603637716, "grad_norm": 1.3855193853378296, "learning_rate": 1.1438001276538308e-05, "loss": 0.0577, "step": 158270 }, { "epoch": 5.351093681327969, "grad_norm": 0.8040406703948975, "learning_rate": 1.143273926809733e-05, "loss": 0.056, "step": 158280 }, { "epoch": 5.351431759018222, "grad_norm": 5.14314603805542, "learning_rate": 1.1427478314070772e-05, "loss": 0.0508, "step": 158290 }, { "epoch": 5.351769836708476, "grad_norm": 1.3316552639007568, "learning_rate": 1.1422218414602465e-05, "loss": 0.0564, "step": 158300 }, { "epoch": 5.3521079143987285, "grad_norm": 1.162514090538025, "learning_rate": 1.1416959569836195e-05, "loss": 0.0545, "step": 158310 }, { "epoch": 5.352445992088982, "grad_norm": 2.9236299991607666, "learning_rate": 1.1411701779915761e-05, "loss": 0.0555, "step": 158320 }, { "epoch": 5.3527840697792355, "grad_norm": 2.1487648487091064, "learning_rate": 1.1406445044984915e-05, "loss": 0.0512, "step": 158330 }, { "epoch": 5.353122147469488, "grad_norm": 1.30890953540802, "learning_rate": 1.1401189365187364e-05, "loss": 0.0543, "step": 158340 }, { "epoch": 5.353460225159742, "grad_norm": 1.0268100500106812, "learning_rate": 1.1395934740666792e-05, "loss": 0.0609, "step": 158350 }, { "epoch": 5.353798302849995, "grad_norm": 0.7560364007949829, "learning_rate": 1.1390681171566848e-05, "loss": 0.0546, "step": 158360 }, { "epoch": 5.354136380540248, "grad_norm": 2.8104023933410645, "learning_rate": 1.1385428658031194e-05, "loss": 0.0618, "step": 158370 }, { "epoch": 5.354474458230501, "grad_norm": 1.3925899267196655, "learning_rate": 1.1380177200203407e-05, "loss": 0.0586, "step": 158380 }, { "epoch": 5.354812535920755, "grad_norm": 5.450962066650391, "learning_rate": 1.137492679822706e-05, "loss": 0.0505, "step": 158390 }, { "epoch": 5.355150613611007, "grad_norm": 0.7649700045585632, "learning_rate": 1.1369677452245708e-05, "loss": 0.0535, "step": 158400 }, { "epoch": 5.355488691301261, "grad_norm": 3.3618364334106445, "learning_rate": 1.1364429162402857e-05, "loss": 0.0595, "step": 158410 }, { "epoch": 5.355826768991514, "grad_norm": 1.9849817752838135, "learning_rate": 1.1359181928842023e-05, "loss": 0.0547, "step": 158420 }, { "epoch": 5.356164846681767, "grad_norm": 1.089793086051941, "learning_rate": 1.1353935751706606e-05, "loss": 0.0527, "step": 158430 }, { "epoch": 5.356502924372021, "grad_norm": 0.7620612382888794, "learning_rate": 1.1348690631140075e-05, "loss": 0.0573, "step": 158440 }, { "epoch": 5.356841002062274, "grad_norm": 1.523444652557373, "learning_rate": 1.1343446567285832e-05, "loss": 0.0587, "step": 158450 }, { "epoch": 5.357179079752527, "grad_norm": 2.239962100982666, "learning_rate": 1.1338203560287237e-05, "loss": 0.0545, "step": 158460 }, { "epoch": 5.35751715744278, "grad_norm": 0.9301265478134155, "learning_rate": 1.133296161028763e-05, "loss": 0.0564, "step": 158470 }, { "epoch": 5.357855235133034, "grad_norm": 1.6209940910339355, "learning_rate": 1.1327720717430313e-05, "loss": 0.0591, "step": 158480 }, { "epoch": 5.358193312823287, "grad_norm": 1.1250464916229248, "learning_rate": 1.13224808818586e-05, "loss": 0.0564, "step": 158490 }, { "epoch": 5.35853139051354, "grad_norm": 1.3238266706466675, "learning_rate": 1.1317242103715726e-05, "loss": 0.0522, "step": 158500 }, { "epoch": 5.358869468203793, "grad_norm": 1.1572147607803345, "learning_rate": 1.1312004383144903e-05, "loss": 0.0543, "step": 158510 }, { "epoch": 5.359207545894046, "grad_norm": 1.6704659461975098, "learning_rate": 1.1306767720289369e-05, "loss": 0.0546, "step": 158520 }, { "epoch": 5.3595456235842995, "grad_norm": 2.2167770862579346, "learning_rate": 1.1301532115292246e-05, "loss": 0.051, "step": 158530 }, { "epoch": 5.359883701274553, "grad_norm": 1.0873990058898926, "learning_rate": 1.1296297568296716e-05, "loss": 0.0537, "step": 158540 }, { "epoch": 5.360221778964807, "grad_norm": 1.3744666576385498, "learning_rate": 1.1291064079445867e-05, "loss": 0.0522, "step": 158550 }, { "epoch": 5.360559856655059, "grad_norm": 2.784262180328369, "learning_rate": 1.1285831648882772e-05, "loss": 0.0602, "step": 158560 }, { "epoch": 5.360897934345313, "grad_norm": 4.328278064727783, "learning_rate": 1.1280600276750514e-05, "loss": 0.0516, "step": 158570 }, { "epoch": 5.361236012035565, "grad_norm": 0.9183080196380615, "learning_rate": 1.127536996319209e-05, "loss": 0.0507, "step": 158580 }, { "epoch": 5.361574089725819, "grad_norm": 0.5643103122711182, "learning_rate": 1.1270140708350508e-05, "loss": 0.0583, "step": 158590 }, { "epoch": 5.361912167416072, "grad_norm": 1.50651216506958, "learning_rate": 1.1264912512368714e-05, "loss": 0.0538, "step": 158600 }, { "epoch": 5.362250245106326, "grad_norm": 1.0079530477523804, "learning_rate": 1.1259685375389673e-05, "loss": 0.0545, "step": 158610 }, { "epoch": 5.3625883227965785, "grad_norm": 3.275040626525879, "learning_rate": 1.1254459297556264e-05, "loss": 0.0526, "step": 158620 }, { "epoch": 5.362926400486832, "grad_norm": 0.9151995778083801, "learning_rate": 1.12492342790114e-05, "loss": 0.0476, "step": 158630 }, { "epoch": 5.363264478177085, "grad_norm": 0.7471290826797485, "learning_rate": 1.1244010319897913e-05, "loss": 0.0539, "step": 158640 }, { "epoch": 5.363602555867338, "grad_norm": 2.823885202407837, "learning_rate": 1.1238787420358604e-05, "loss": 0.0583, "step": 158650 }, { "epoch": 5.363940633557592, "grad_norm": 7.3817315101623535, "learning_rate": 1.1233565580536304e-05, "loss": 0.0576, "step": 158660 }, { "epoch": 5.364278711247845, "grad_norm": 1.1772396564483643, "learning_rate": 1.1228344800573753e-05, "loss": 0.051, "step": 158670 }, { "epoch": 5.364616788938098, "grad_norm": 1.4013843536376953, "learning_rate": 1.1223125080613694e-05, "loss": 0.0567, "step": 158680 }, { "epoch": 5.364954866628351, "grad_norm": 0.5551392436027527, "learning_rate": 1.1217906420798812e-05, "loss": 0.051, "step": 158690 }, { "epoch": 5.365292944318605, "grad_norm": 1.5599305629730225, "learning_rate": 1.12126888212718e-05, "loss": 0.0546, "step": 158700 }, { "epoch": 5.365631022008857, "grad_norm": 0.7425277829170227, "learning_rate": 1.1207472282175329e-05, "loss": 0.0578, "step": 158710 }, { "epoch": 5.365969099699111, "grad_norm": 1.6341112852096558, "learning_rate": 1.1202256803651968e-05, "loss": 0.0537, "step": 158720 }, { "epoch": 5.3663071773893645, "grad_norm": 1.182172179222107, "learning_rate": 1.1197042385844342e-05, "loss": 0.0582, "step": 158730 }, { "epoch": 5.366645255079617, "grad_norm": 1.0270757675170898, "learning_rate": 1.1191829028894985e-05, "loss": 0.0577, "step": 158740 }, { "epoch": 5.366983332769871, "grad_norm": 1.4276366233825684, "learning_rate": 1.1186616732946453e-05, "loss": 0.0469, "step": 158750 }, { "epoch": 5.367321410460124, "grad_norm": 1.1160424947738647, "learning_rate": 1.1181405498141245e-05, "loss": 0.055, "step": 158760 }, { "epoch": 5.367659488150377, "grad_norm": 0.8297913670539856, "learning_rate": 1.1176195324621807e-05, "loss": 0.0663, "step": 158770 }, { "epoch": 5.36799756584063, "grad_norm": 2.3390843868255615, "learning_rate": 1.1170986212530621e-05, "loss": 0.0586, "step": 158780 }, { "epoch": 5.368335643530884, "grad_norm": 2.1659626960754395, "learning_rate": 1.1165778162010088e-05, "loss": 0.0605, "step": 158790 }, { "epoch": 5.368673721221136, "grad_norm": 1.2438197135925293, "learning_rate": 1.1160571173202583e-05, "loss": 0.0576, "step": 158800 }, { "epoch": 5.36901179891139, "grad_norm": 0.7954272627830505, "learning_rate": 1.1155365246250466e-05, "loss": 0.0593, "step": 158810 }, { "epoch": 5.369349876601643, "grad_norm": 2.002408742904663, "learning_rate": 1.115016038129606e-05, "loss": 0.0515, "step": 158820 }, { "epoch": 5.369687954291896, "grad_norm": 0.8807024359703064, "learning_rate": 1.1144956578481697e-05, "loss": 0.0606, "step": 158830 }, { "epoch": 5.37002603198215, "grad_norm": 1.6409214735031128, "learning_rate": 1.1139753837949618e-05, "loss": 0.0561, "step": 158840 }, { "epoch": 5.370364109672403, "grad_norm": 0.8153035044670105, "learning_rate": 1.1134552159842072e-05, "loss": 0.0518, "step": 158850 }, { "epoch": 5.370702187362656, "grad_norm": 1.0310020446777344, "learning_rate": 1.1129351544301252e-05, "loss": 0.054, "step": 158860 }, { "epoch": 5.371040265052909, "grad_norm": 2.2278664112091064, "learning_rate": 1.1124151991469378e-05, "loss": 0.0542, "step": 158870 }, { "epoch": 5.371378342743163, "grad_norm": 6.971222877502441, "learning_rate": 1.1118953501488577e-05, "loss": 0.0602, "step": 158880 }, { "epoch": 5.371716420433415, "grad_norm": 1.5675206184387207, "learning_rate": 1.1113756074500969e-05, "loss": 0.0547, "step": 158890 }, { "epoch": 5.372054498123669, "grad_norm": 0.7674347162246704, "learning_rate": 1.1108559710648674e-05, "loss": 0.0453, "step": 158900 }, { "epoch": 5.372392575813922, "grad_norm": 2.3935329914093018, "learning_rate": 1.1103364410073735e-05, "loss": 0.0572, "step": 158910 }, { "epoch": 5.372730653504175, "grad_norm": 1.2767560482025146, "learning_rate": 1.1098170172918227e-05, "loss": 0.0537, "step": 158920 }, { "epoch": 5.3730687311944285, "grad_norm": 5.381278038024902, "learning_rate": 1.10929769993241e-05, "loss": 0.0547, "step": 158930 }, { "epoch": 5.373406808884682, "grad_norm": 1.6176021099090576, "learning_rate": 1.1087784889433372e-05, "loss": 0.0505, "step": 158940 }, { "epoch": 5.373744886574935, "grad_norm": 0.7857402563095093, "learning_rate": 1.1082593843387996e-05, "loss": 0.0517, "step": 158950 }, { "epoch": 5.374082964265188, "grad_norm": 1.774737000465393, "learning_rate": 1.1077403861329882e-05, "loss": 0.0577, "step": 158960 }, { "epoch": 5.374421041955442, "grad_norm": 0.9709895253181458, "learning_rate": 1.1072214943400926e-05, "loss": 0.0537, "step": 158970 }, { "epoch": 5.374759119645694, "grad_norm": 1.469378113746643, "learning_rate": 1.1067027089742976e-05, "loss": 0.0535, "step": 158980 }, { "epoch": 5.375097197335948, "grad_norm": 0.8185970783233643, "learning_rate": 1.1061840300497895e-05, "loss": 0.0533, "step": 158990 }, { "epoch": 5.375435275026201, "grad_norm": 2.1417365074157715, "learning_rate": 1.1056654575807468e-05, "loss": 0.0591, "step": 159000 }, { "epoch": 5.375773352716454, "grad_norm": 2.2752797603607178, "learning_rate": 1.1051469915813462e-05, "loss": 0.0581, "step": 159010 }, { "epoch": 5.3761114304067075, "grad_norm": 9.379144668579102, "learning_rate": 1.1046286320657646e-05, "loss": 0.0554, "step": 159020 }, { "epoch": 5.376449508096961, "grad_norm": 5.1884894371032715, "learning_rate": 1.1041103790481716e-05, "loss": 0.0505, "step": 159030 }, { "epoch": 5.376787585787214, "grad_norm": 1.3738501071929932, "learning_rate": 1.103592232542739e-05, "loss": 0.0583, "step": 159040 }, { "epoch": 5.377125663477467, "grad_norm": 0.8839583992958069, "learning_rate": 1.1030741925636307e-05, "loss": 0.0567, "step": 159050 }, { "epoch": 5.377463741167721, "grad_norm": 1.6153002977371216, "learning_rate": 1.10255625912501e-05, "loss": 0.0514, "step": 159060 }, { "epoch": 5.377801818857973, "grad_norm": 1.1629942655563354, "learning_rate": 1.1020384322410354e-05, "loss": 0.0568, "step": 159070 }, { "epoch": 5.378139896548227, "grad_norm": 1.7942548990249634, "learning_rate": 1.1015207119258675e-05, "loss": 0.0517, "step": 159080 }, { "epoch": 5.37847797423848, "grad_norm": 3.1103131771087646, "learning_rate": 1.1010030981936581e-05, "loss": 0.0543, "step": 159090 }, { "epoch": 5.378816051928733, "grad_norm": 0.878259003162384, "learning_rate": 1.1004855910585582e-05, "loss": 0.0575, "step": 159100 }, { "epoch": 5.379154129618986, "grad_norm": 2.0185117721557617, "learning_rate": 1.0999681905347187e-05, "loss": 0.0566, "step": 159110 }, { "epoch": 5.37949220730924, "grad_norm": 4.286313056945801, "learning_rate": 1.0994508966362827e-05, "loss": 0.0518, "step": 159120 }, { "epoch": 5.379830284999493, "grad_norm": 0.9903103113174438, "learning_rate": 1.0989337093773943e-05, "loss": 0.0566, "step": 159130 }, { "epoch": 5.380168362689746, "grad_norm": 6.631227493286133, "learning_rate": 1.0984166287721931e-05, "loss": 0.0526, "step": 159140 }, { "epoch": 5.38050644038, "grad_norm": 1.2750797271728516, "learning_rate": 1.0978996548348142e-05, "loss": 0.0492, "step": 159150 }, { "epoch": 5.380844518070252, "grad_norm": 4.995594024658203, "learning_rate": 1.0973827875793946e-05, "loss": 0.0543, "step": 159160 }, { "epoch": 5.381182595760506, "grad_norm": 2.3867123126983643, "learning_rate": 1.0968660270200626e-05, "loss": 0.0498, "step": 159170 }, { "epoch": 5.381520673450759, "grad_norm": 1.2932207584381104, "learning_rate": 1.096349373170948e-05, "loss": 0.0553, "step": 159180 }, { "epoch": 5.381858751141012, "grad_norm": 1.3909661769866943, "learning_rate": 1.0958328260461725e-05, "loss": 0.0499, "step": 159190 }, { "epoch": 5.382196828831265, "grad_norm": 3.0157084465026855, "learning_rate": 1.095316385659862e-05, "loss": 0.0655, "step": 159200 }, { "epoch": 5.382534906521519, "grad_norm": 1.0699926614761353, "learning_rate": 1.0948000520261364e-05, "loss": 0.0588, "step": 159210 }, { "epoch": 5.3828729842117715, "grad_norm": 1.0716170072555542, "learning_rate": 1.0942838251591081e-05, "loss": 0.0522, "step": 159220 }, { "epoch": 5.383211061902025, "grad_norm": 2.055518388748169, "learning_rate": 1.093767705072894e-05, "loss": 0.0543, "step": 159230 }, { "epoch": 5.3835491395922785, "grad_norm": 3.0278520584106445, "learning_rate": 1.0932516917816016e-05, "loss": 0.0555, "step": 159240 }, { "epoch": 5.383887217282531, "grad_norm": 1.2671740055084229, "learning_rate": 1.092735785299342e-05, "loss": 0.0501, "step": 159250 }, { "epoch": 5.384225294972785, "grad_norm": 1.1207647323608398, "learning_rate": 1.0922199856402176e-05, "loss": 0.0538, "step": 159260 }, { "epoch": 5.384563372663038, "grad_norm": 3.1288700103759766, "learning_rate": 1.0917042928183297e-05, "loss": 0.0552, "step": 159270 }, { "epoch": 5.384901450353291, "grad_norm": 1.7462610006332397, "learning_rate": 1.09118870684778e-05, "loss": 0.057, "step": 159280 }, { "epoch": 5.385239528043544, "grad_norm": 1.1397857666015625, "learning_rate": 1.0906732277426619e-05, "loss": 0.0561, "step": 159290 }, { "epoch": 5.385577605733798, "grad_norm": 5.074526786804199, "learning_rate": 1.0901578555170693e-05, "loss": 0.0543, "step": 159300 }, { "epoch": 5.3859156834240505, "grad_norm": 1.7466788291931152, "learning_rate": 1.089642590185091e-05, "loss": 0.0504, "step": 159310 }, { "epoch": 5.386253761114304, "grad_norm": 1.1691335439682007, "learning_rate": 1.0891274317608164e-05, "loss": 0.055, "step": 159320 }, { "epoch": 5.3865918388045575, "grad_norm": 1.8352062702178955, "learning_rate": 1.088612380258327e-05, "loss": 0.0545, "step": 159330 }, { "epoch": 5.38692991649481, "grad_norm": 1.971946358680725, "learning_rate": 1.0880974356917067e-05, "loss": 0.0517, "step": 159340 }, { "epoch": 5.387267994185064, "grad_norm": 3.32948637008667, "learning_rate": 1.0875825980750337e-05, "loss": 0.0563, "step": 159350 }, { "epoch": 5.387606071875317, "grad_norm": 0.845162034034729, "learning_rate": 1.087067867422381e-05, "loss": 0.0507, "step": 159360 }, { "epoch": 5.38794414956557, "grad_norm": 1.2756919860839844, "learning_rate": 1.0865532437478237e-05, "loss": 0.0502, "step": 159370 }, { "epoch": 5.388282227255823, "grad_norm": 0.9861913323402405, "learning_rate": 1.086038727065431e-05, "loss": 0.0546, "step": 159380 }, { "epoch": 5.388620304946077, "grad_norm": 2.610004425048828, "learning_rate": 1.0855243173892676e-05, "loss": 0.0554, "step": 159390 }, { "epoch": 5.388958382636329, "grad_norm": 3.7758586406707764, "learning_rate": 1.0850100147334003e-05, "loss": 0.0511, "step": 159400 }, { "epoch": 5.389296460326583, "grad_norm": 1.4324063062667847, "learning_rate": 1.084495819111887e-05, "loss": 0.0574, "step": 159410 }, { "epoch": 5.3896345380168365, "grad_norm": 2.818702220916748, "learning_rate": 1.0839817305387894e-05, "loss": 0.0491, "step": 159420 }, { "epoch": 5.389972615707089, "grad_norm": 0.9299402236938477, "learning_rate": 1.0834677490281576e-05, "loss": 0.0565, "step": 159430 }, { "epoch": 5.390310693397343, "grad_norm": 1.1018776893615723, "learning_rate": 1.0829538745940476e-05, "loss": 0.0594, "step": 159440 }, { "epoch": 5.390648771087596, "grad_norm": 0.8942745327949524, "learning_rate": 1.0824401072505058e-05, "loss": 0.06, "step": 159450 }, { "epoch": 5.39098684877785, "grad_norm": 0.8834397792816162, "learning_rate": 1.0819264470115808e-05, "loss": 0.0577, "step": 159460 }, { "epoch": 5.391324926468102, "grad_norm": 1.5638189315795898, "learning_rate": 1.0814128938913149e-05, "loss": 0.056, "step": 159470 }, { "epoch": 5.391663004158356, "grad_norm": 3.6268720626831055, "learning_rate": 1.080899447903747e-05, "loss": 0.0552, "step": 159480 }, { "epoch": 5.392001081848608, "grad_norm": 2.080427885055542, "learning_rate": 1.0803861090629174e-05, "loss": 0.0522, "step": 159490 }, { "epoch": 5.392339159538862, "grad_norm": 2.7912402153015137, "learning_rate": 1.0798728773828588e-05, "loss": 0.0578, "step": 159500 }, { "epoch": 5.392677237229115, "grad_norm": 1.0778443813323975, "learning_rate": 1.0793597528776034e-05, "loss": 0.0501, "step": 159510 }, { "epoch": 5.393015314919369, "grad_norm": 0.9330847263336182, "learning_rate": 1.0788467355611775e-05, "loss": 0.0554, "step": 159520 }, { "epoch": 5.3933533926096215, "grad_norm": 0.9936987161636353, "learning_rate": 1.0783338254476088e-05, "loss": 0.0566, "step": 159530 }, { "epoch": 5.393691470299875, "grad_norm": 1.1457346677780151, "learning_rate": 1.0778210225509217e-05, "loss": 0.0528, "step": 159540 }, { "epoch": 5.394029547990128, "grad_norm": 0.9652108550071716, "learning_rate": 1.0773083268851341e-05, "loss": 0.0548, "step": 159550 }, { "epoch": 5.394367625680381, "grad_norm": 0.8045342564582825, "learning_rate": 1.0767957384642635e-05, "loss": 0.0514, "step": 159560 }, { "epoch": 5.394705703370635, "grad_norm": 1.5689605474472046, "learning_rate": 1.0762832573023219e-05, "loss": 0.0552, "step": 159570 }, { "epoch": 5.395043781060888, "grad_norm": 1.0969064235687256, "learning_rate": 1.0757708834133224e-05, "loss": 0.0603, "step": 159580 }, { "epoch": 5.395381858751141, "grad_norm": 1.6174366474151611, "learning_rate": 1.075258616811275e-05, "loss": 0.0611, "step": 159590 }, { "epoch": 5.395719936441394, "grad_norm": 0.989016056060791, "learning_rate": 1.0747464575101795e-05, "loss": 0.0565, "step": 159600 }, { "epoch": 5.396058014131647, "grad_norm": 1.7419002056121826, "learning_rate": 1.0742344055240427e-05, "loss": 0.0521, "step": 159610 }, { "epoch": 5.3963960918219005, "grad_norm": 1.4405940771102905, "learning_rate": 1.0737224608668616e-05, "loss": 0.0512, "step": 159620 }, { "epoch": 5.396734169512154, "grad_norm": 1.5694133043289185, "learning_rate": 1.0732106235526357e-05, "loss": 0.0569, "step": 159630 }, { "epoch": 5.3970722472024075, "grad_norm": 4.299824237823486, "learning_rate": 1.0726988935953536e-05, "loss": 0.0535, "step": 159640 }, { "epoch": 5.39741032489266, "grad_norm": 1.3153494596481323, "learning_rate": 1.0721872710090081e-05, "loss": 0.0554, "step": 159650 }, { "epoch": 5.397748402582914, "grad_norm": 1.146078109741211, "learning_rate": 1.0716757558075885e-05, "loss": 0.0537, "step": 159660 }, { "epoch": 5.398086480273167, "grad_norm": 0.9860498905181885, "learning_rate": 1.0711643480050782e-05, "loss": 0.055, "step": 159670 }, { "epoch": 5.39842455796342, "grad_norm": 4.617015838623047, "learning_rate": 1.0706530476154581e-05, "loss": 0.0589, "step": 159680 }, { "epoch": 5.398762635653673, "grad_norm": 0.9486483931541443, "learning_rate": 1.0701418546527065e-05, "loss": 0.0512, "step": 159690 }, { "epoch": 5.399100713343927, "grad_norm": 2.244231939315796, "learning_rate": 1.0696307691308016e-05, "loss": 0.0524, "step": 159700 }, { "epoch": 5.3994387910341795, "grad_norm": 1.5773446559906006, "learning_rate": 1.069119791063713e-05, "loss": 0.0476, "step": 159710 }, { "epoch": 5.399776868724433, "grad_norm": 1.5368788242340088, "learning_rate": 1.0686089204654144e-05, "loss": 0.0534, "step": 159720 }, { "epoch": 5.4001149464146865, "grad_norm": 2.646040439605713, "learning_rate": 1.068098157349871e-05, "loss": 0.0563, "step": 159730 }, { "epoch": 5.400453024104939, "grad_norm": 1.3088852167129517, "learning_rate": 1.0675875017310455e-05, "loss": 0.055, "step": 159740 }, { "epoch": 5.400791101795193, "grad_norm": 2.1832363605499268, "learning_rate": 1.0670769536229013e-05, "loss": 0.0524, "step": 159750 }, { "epoch": 5.401129179485446, "grad_norm": 0.9242690801620483, "learning_rate": 1.0665665130393959e-05, "loss": 0.0546, "step": 159760 }, { "epoch": 5.401467257175699, "grad_norm": 1.5913574695587158, "learning_rate": 1.066056179994483e-05, "loss": 0.0541, "step": 159770 }, { "epoch": 5.401805334865952, "grad_norm": 1.3747261762619019, "learning_rate": 1.0655459545021174e-05, "loss": 0.048, "step": 159780 }, { "epoch": 5.402143412556206, "grad_norm": 3.2363386154174805, "learning_rate": 1.0650358365762465e-05, "loss": 0.0504, "step": 159790 }, { "epoch": 5.402481490246458, "grad_norm": 3.2435364723205566, "learning_rate": 1.0645258262308194e-05, "loss": 0.0582, "step": 159800 }, { "epoch": 5.402819567936712, "grad_norm": 1.6277475357055664, "learning_rate": 1.0640159234797758e-05, "loss": 0.0561, "step": 159810 }, { "epoch": 5.403157645626965, "grad_norm": 34.026546478271484, "learning_rate": 1.0635061283370589e-05, "loss": 0.0552, "step": 159820 }, { "epoch": 5.403495723317218, "grad_norm": 1.1564712524414062, "learning_rate": 1.062996440816605e-05, "loss": 0.0542, "step": 159830 }, { "epoch": 5.403833801007472, "grad_norm": 1.665759801864624, "learning_rate": 1.0624868609323496e-05, "loss": 0.0552, "step": 159840 }, { "epoch": 5.404171878697725, "grad_norm": 1.096245288848877, "learning_rate": 1.0619773886982249e-05, "loss": 0.0503, "step": 159850 }, { "epoch": 5.404509956387978, "grad_norm": 1.9227752685546875, "learning_rate": 1.0614680241281577e-05, "loss": 0.0522, "step": 159860 }, { "epoch": 5.404848034078231, "grad_norm": 1.3020431995391846, "learning_rate": 1.0609587672360761e-05, "loss": 0.0599, "step": 159870 }, { "epoch": 5.405186111768485, "grad_norm": 3.71091365814209, "learning_rate": 1.0604496180359019e-05, "loss": 0.0645, "step": 159880 }, { "epoch": 5.405524189458737, "grad_norm": 2.8615827560424805, "learning_rate": 1.0599405765415554e-05, "loss": 0.0618, "step": 159890 }, { "epoch": 5.405862267148991, "grad_norm": 1.5208536386489868, "learning_rate": 1.059431642766952e-05, "loss": 0.0538, "step": 159900 }, { "epoch": 5.406200344839244, "grad_norm": 2.514737844467163, "learning_rate": 1.0589228167260067e-05, "loss": 0.0615, "step": 159910 }, { "epoch": 5.406538422529497, "grad_norm": 1.3361842632293701, "learning_rate": 1.058414098432633e-05, "loss": 0.0585, "step": 159920 }, { "epoch": 5.4068765002197505, "grad_norm": 1.3758442401885986, "learning_rate": 1.0579054879007366e-05, "loss": 0.0559, "step": 159930 }, { "epoch": 5.407214577910004, "grad_norm": 3.69413423538208, "learning_rate": 1.0573969851442234e-05, "loss": 0.056, "step": 159940 }, { "epoch": 5.407552655600257, "grad_norm": 1.1563597917556763, "learning_rate": 1.0568885901769938e-05, "loss": 0.0522, "step": 159950 }, { "epoch": 5.40789073329051, "grad_norm": 1.2481117248535156, "learning_rate": 1.05638030301295e-05, "loss": 0.0554, "step": 159960 }, { "epoch": 5.408228810980764, "grad_norm": 1.918013334274292, "learning_rate": 1.0558721236659874e-05, "loss": 0.0586, "step": 159970 }, { "epoch": 5.408566888671016, "grad_norm": 0.9727224111557007, "learning_rate": 1.055364052149998e-05, "loss": 0.0504, "step": 159980 }, { "epoch": 5.40890496636127, "grad_norm": 1.5946104526519775, "learning_rate": 1.054856088478875e-05, "loss": 0.0532, "step": 159990 }, { "epoch": 5.409243044051523, "grad_norm": 0.9843226075172424, "learning_rate": 1.0543482326665028e-05, "loss": 0.0555, "step": 160000 }, { "epoch": 5.409581121741776, "grad_norm": 0.879939079284668, "learning_rate": 1.0538404847267702e-05, "loss": 0.065, "step": 160010 }, { "epoch": 5.4099191994320295, "grad_norm": 1.9532194137573242, "learning_rate": 1.0533328446735535e-05, "loss": 0.0574, "step": 160020 }, { "epoch": 5.410257277122283, "grad_norm": 0.840158998966217, "learning_rate": 1.0528253125207343e-05, "loss": 0.0546, "step": 160030 }, { "epoch": 5.410595354812536, "grad_norm": 1.0805222988128662, "learning_rate": 1.05231788828219e-05, "loss": 0.0559, "step": 160040 }, { "epoch": 5.410933432502789, "grad_norm": 0.9456830620765686, "learning_rate": 1.0518105719717907e-05, "loss": 0.052, "step": 160050 }, { "epoch": 5.411271510193043, "grad_norm": 2.945453405380249, "learning_rate": 1.0513033636034075e-05, "loss": 0.0572, "step": 160060 }, { "epoch": 5.411609587883295, "grad_norm": 4.758403301239014, "learning_rate": 1.0507962631909057e-05, "loss": 0.0512, "step": 160070 }, { "epoch": 5.411947665573549, "grad_norm": 2.1235251426696777, "learning_rate": 1.0502892707481515e-05, "loss": 0.0558, "step": 160080 }, { "epoch": 5.412285743263802, "grad_norm": 0.9623025059700012, "learning_rate": 1.0497823862890049e-05, "loss": 0.0559, "step": 160090 }, { "epoch": 5.412623820954055, "grad_norm": 0.9059848189353943, "learning_rate": 1.049275609827322e-05, "loss": 0.0521, "step": 160100 }, { "epoch": 5.412961898644308, "grad_norm": 1.4493836164474487, "learning_rate": 1.048768941376962e-05, "loss": 0.0509, "step": 160110 }, { "epoch": 5.413299976334562, "grad_norm": 2.001570701599121, "learning_rate": 1.0482623809517727e-05, "loss": 0.0506, "step": 160120 }, { "epoch": 5.413638054024815, "grad_norm": 1.6629812717437744, "learning_rate": 1.0477559285656068e-05, "loss": 0.0583, "step": 160130 }, { "epoch": 5.413976131715068, "grad_norm": 1.2086421251296997, "learning_rate": 1.0472495842323093e-05, "loss": 0.0539, "step": 160140 }, { "epoch": 5.414314209405322, "grad_norm": 2.120490550994873, "learning_rate": 1.0467433479657212e-05, "loss": 0.0581, "step": 160150 }, { "epoch": 5.414652287095574, "grad_norm": 1.0412994623184204, "learning_rate": 1.0462372197796866e-05, "loss": 0.0552, "step": 160160 }, { "epoch": 5.414990364785828, "grad_norm": 2.4687509536743164, "learning_rate": 1.0457311996880415e-05, "loss": 0.0548, "step": 160170 }, { "epoch": 5.415328442476081, "grad_norm": 1.5962365865707397, "learning_rate": 1.0452252877046193e-05, "loss": 0.0522, "step": 160180 }, { "epoch": 5.415666520166334, "grad_norm": 0.9643933176994324, "learning_rate": 1.0447194838432506e-05, "loss": 0.0545, "step": 160190 }, { "epoch": 5.416004597856587, "grad_norm": 0.868864893913269, "learning_rate": 1.0442137881177666e-05, "loss": 0.0565, "step": 160200 }, { "epoch": 5.416342675546841, "grad_norm": 1.1141853332519531, "learning_rate": 1.04370820054199e-05, "loss": 0.0556, "step": 160210 }, { "epoch": 5.4166807532370935, "grad_norm": 0.7260059714317322, "learning_rate": 1.0432027211297468e-05, "loss": 0.0486, "step": 160220 }, { "epoch": 5.417018830927347, "grad_norm": 15.067363739013672, "learning_rate": 1.0426973498948545e-05, "loss": 0.054, "step": 160230 }, { "epoch": 5.4173569086176006, "grad_norm": 1.5578768253326416, "learning_rate": 1.0421920868511286e-05, "loss": 0.0508, "step": 160240 }, { "epoch": 5.417694986307853, "grad_norm": 2.275824546813965, "learning_rate": 1.0416869320123856e-05, "loss": 0.0586, "step": 160250 }, { "epoch": 5.418033063998107, "grad_norm": 1.717624306678772, "learning_rate": 1.0411818853924343e-05, "loss": 0.0585, "step": 160260 }, { "epoch": 5.41837114168836, "grad_norm": 2.267165184020996, "learning_rate": 1.0406769470050836e-05, "loss": 0.0514, "step": 160270 }, { "epoch": 5.418709219378613, "grad_norm": 2.873116970062256, "learning_rate": 1.040172116864136e-05, "loss": 0.0579, "step": 160280 }, { "epoch": 5.419047297068866, "grad_norm": 2.294445514678955, "learning_rate": 1.0396673949833952e-05, "loss": 0.0559, "step": 160290 }, { "epoch": 5.41938537475912, "grad_norm": 1.0117263793945312, "learning_rate": 1.039162781376663e-05, "loss": 0.0574, "step": 160300 }, { "epoch": 5.4197234524493725, "grad_norm": 3.1652793884277344, "learning_rate": 1.0386582760577296e-05, "loss": 0.0555, "step": 160310 }, { "epoch": 5.420061530139626, "grad_norm": 1.4613944292068481, "learning_rate": 1.0381538790403917e-05, "loss": 0.0625, "step": 160320 }, { "epoch": 5.4203996078298795, "grad_norm": 3.5913808345794678, "learning_rate": 1.037649590338437e-05, "loss": 0.0509, "step": 160330 }, { "epoch": 5.420737685520132, "grad_norm": 2.4012563228607178, "learning_rate": 1.0371454099656552e-05, "loss": 0.0532, "step": 160340 }, { "epoch": 5.421075763210386, "grad_norm": 1.5938669443130493, "learning_rate": 1.0366413379358297e-05, "loss": 0.0583, "step": 160350 }, { "epoch": 5.421413840900639, "grad_norm": 2.011591672897339, "learning_rate": 1.036137374262739e-05, "loss": 0.0538, "step": 160360 }, { "epoch": 5.421751918590892, "grad_norm": 3.326392889022827, "learning_rate": 1.035633518960165e-05, "loss": 0.0555, "step": 160370 }, { "epoch": 5.422089996281145, "grad_norm": 1.4404702186584473, "learning_rate": 1.0351297720418812e-05, "loss": 0.0489, "step": 160380 }, { "epoch": 5.422428073971399, "grad_norm": 1.4774810075759888, "learning_rate": 1.03462613352166e-05, "loss": 0.0567, "step": 160390 }, { "epoch": 5.422766151661651, "grad_norm": 2.5035791397094727, "learning_rate": 1.0341226034132695e-05, "loss": 0.0524, "step": 160400 }, { "epoch": 5.423104229351905, "grad_norm": 1.8775917291641235, "learning_rate": 1.0336191817304769e-05, "loss": 0.0513, "step": 160410 }, { "epoch": 5.4234423070421585, "grad_norm": 2.6236801147460938, "learning_rate": 1.0331158684870469e-05, "loss": 0.055, "step": 160420 }, { "epoch": 5.423780384732412, "grad_norm": 10.923602104187012, "learning_rate": 1.0326126636967388e-05, "loss": 0.0534, "step": 160430 }, { "epoch": 5.424118462422665, "grad_norm": 0.9996063113212585, "learning_rate": 1.03210956737331e-05, "loss": 0.0601, "step": 160440 }, { "epoch": 5.424456540112918, "grad_norm": 2.2768301963806152, "learning_rate": 1.031606579530514e-05, "loss": 0.0583, "step": 160450 }, { "epoch": 5.424794617803171, "grad_norm": 1.39234459400177, "learning_rate": 1.0311037001821044e-05, "loss": 0.0519, "step": 160460 }, { "epoch": 5.425132695493424, "grad_norm": 1.3151345252990723, "learning_rate": 1.0306009293418283e-05, "loss": 0.0585, "step": 160470 }, { "epoch": 5.425470773183678, "grad_norm": 1.7662394046783447, "learning_rate": 1.0300982670234306e-05, "loss": 0.0553, "step": 160480 }, { "epoch": 5.425808850873931, "grad_norm": 1.135352611541748, "learning_rate": 1.0295957132406559e-05, "loss": 0.0535, "step": 160490 }, { "epoch": 5.426146928564184, "grad_norm": 0.687944233417511, "learning_rate": 1.0290932680072412e-05, "loss": 0.0614, "step": 160500 }, { "epoch": 5.426485006254437, "grad_norm": 13.274025917053223, "learning_rate": 1.028590931336928e-05, "loss": 0.0562, "step": 160510 }, { "epoch": 5.42682308394469, "grad_norm": 2.175504446029663, "learning_rate": 1.028088703243444e-05, "loss": 0.0559, "step": 160520 }, { "epoch": 5.4271611616349436, "grad_norm": 0.7049710154533386, "learning_rate": 1.027586583740524e-05, "loss": 0.0613, "step": 160530 }, { "epoch": 5.427499239325197, "grad_norm": 1.4077272415161133, "learning_rate": 1.0270845728418927e-05, "loss": 0.0487, "step": 160540 }, { "epoch": 5.427837317015451, "grad_norm": 2.0957589149475098, "learning_rate": 1.0265826705612786e-05, "loss": 0.0552, "step": 160550 }, { "epoch": 5.428175394705703, "grad_norm": 0.8459675312042236, "learning_rate": 1.0260808769124009e-05, "loss": 0.0526, "step": 160560 }, { "epoch": 5.428513472395957, "grad_norm": 2.0134365558624268, "learning_rate": 1.025579191908978e-05, "loss": 0.0558, "step": 160570 }, { "epoch": 5.428851550086209, "grad_norm": 2.0231564044952393, "learning_rate": 1.0250776155647284e-05, "loss": 0.0561, "step": 160580 }, { "epoch": 5.429189627776463, "grad_norm": 1.0377415418624878, "learning_rate": 1.0245761478933636e-05, "loss": 0.0581, "step": 160590 }, { "epoch": 5.429527705466716, "grad_norm": 0.7365829348564148, "learning_rate": 1.0240747889085916e-05, "loss": 0.0532, "step": 160600 }, { "epoch": 5.42986578315697, "grad_norm": 0.7107486128807068, "learning_rate": 1.0235735386241223e-05, "loss": 0.0527, "step": 160610 }, { "epoch": 5.4302038608472225, "grad_norm": 1.8873980045318604, "learning_rate": 1.0230723970536581e-05, "loss": 0.0534, "step": 160620 }, { "epoch": 5.430541938537476, "grad_norm": 1.871161937713623, "learning_rate": 1.0225713642109014e-05, "loss": 0.0497, "step": 160630 }, { "epoch": 5.4308800162277295, "grad_norm": 1.4063037633895874, "learning_rate": 1.0220704401095494e-05, "loss": 0.0542, "step": 160640 }, { "epoch": 5.431218093917982, "grad_norm": 1.442403793334961, "learning_rate": 1.0215696247632972e-05, "loss": 0.0509, "step": 160650 }, { "epoch": 5.431556171608236, "grad_norm": 2.508972406387329, "learning_rate": 1.0210689181858352e-05, "loss": 0.0545, "step": 160660 }, { "epoch": 5.431894249298489, "grad_norm": 5.408607006072998, "learning_rate": 1.0205683203908555e-05, "loss": 0.0521, "step": 160670 }, { "epoch": 5.432232326988742, "grad_norm": 0.849375307559967, "learning_rate": 1.0200678313920436e-05, "loss": 0.0576, "step": 160680 }, { "epoch": 5.432570404678995, "grad_norm": 0.9924922585487366, "learning_rate": 1.01956745120308e-05, "loss": 0.0619, "step": 160690 }, { "epoch": 5.432908482369249, "grad_norm": 0.975983202457428, "learning_rate": 1.0190671798376483e-05, "loss": 0.0573, "step": 160700 }, { "epoch": 5.4332465600595015, "grad_norm": 1.3563175201416016, "learning_rate": 1.018567017309423e-05, "loss": 0.0542, "step": 160710 }, { "epoch": 5.433584637749755, "grad_norm": 1.6850165128707886, "learning_rate": 1.0180669636320827e-05, "loss": 0.0521, "step": 160720 }, { "epoch": 5.4339227154400085, "grad_norm": 0.8837896585464478, "learning_rate": 1.0175670188192926e-05, "loss": 0.0597, "step": 160730 }, { "epoch": 5.434260793130261, "grad_norm": 2.453375816345215, "learning_rate": 1.0170671828847244e-05, "loss": 0.0486, "step": 160740 }, { "epoch": 5.434598870820515, "grad_norm": 2.4702906608581543, "learning_rate": 1.016567455842044e-05, "loss": 0.0512, "step": 160750 }, { "epoch": 5.434936948510768, "grad_norm": 5.141760349273682, "learning_rate": 1.0160678377049133e-05, "loss": 0.0489, "step": 160760 }, { "epoch": 5.435275026201021, "grad_norm": 1.9150766134262085, "learning_rate": 1.015568328486991e-05, "loss": 0.0558, "step": 160770 }, { "epoch": 5.435613103891274, "grad_norm": 1.6418217420578003, "learning_rate": 1.0150689282019326e-05, "loss": 0.0522, "step": 160780 }, { "epoch": 5.435951181581528, "grad_norm": 0.9254345893859863, "learning_rate": 1.0145696368633928e-05, "loss": 0.0507, "step": 160790 }, { "epoch": 5.43628925927178, "grad_norm": 1.5447959899902344, "learning_rate": 1.0140704544850239e-05, "loss": 0.055, "step": 160800 }, { "epoch": 5.436627336962034, "grad_norm": 2.6350955963134766, "learning_rate": 1.0135713810804692e-05, "loss": 0.0571, "step": 160810 }, { "epoch": 5.436965414652287, "grad_norm": 0.9464623928070068, "learning_rate": 1.0130724166633764e-05, "loss": 0.049, "step": 160820 }, { "epoch": 5.43730349234254, "grad_norm": 1.5910035371780396, "learning_rate": 1.0125735612473847e-05, "loss": 0.0525, "step": 160830 }, { "epoch": 5.437641570032794, "grad_norm": 1.0357074737548828, "learning_rate": 1.0120748148461351e-05, "loss": 0.055, "step": 160840 }, { "epoch": 5.437979647723047, "grad_norm": 3.1104884147644043, "learning_rate": 1.0115761774732618e-05, "loss": 0.0588, "step": 160850 }, { "epoch": 5.4383177254133, "grad_norm": 2.622241258621216, "learning_rate": 1.0110776491423956e-05, "loss": 0.0565, "step": 160860 }, { "epoch": 5.438655803103553, "grad_norm": 2.5317959785461426, "learning_rate": 1.0105792298671702e-05, "loss": 0.0557, "step": 160870 }, { "epoch": 5.438993880793807, "grad_norm": 0.8812841773033142, "learning_rate": 1.0100809196612093e-05, "loss": 0.0558, "step": 160880 }, { "epoch": 5.439331958484059, "grad_norm": 2.6200859546661377, "learning_rate": 1.0095827185381373e-05, "loss": 0.0546, "step": 160890 }, { "epoch": 5.439670036174313, "grad_norm": 2.5074260234832764, "learning_rate": 1.0090846265115733e-05, "loss": 0.0544, "step": 160900 }, { "epoch": 5.440008113864566, "grad_norm": 0.8927388191223145, "learning_rate": 1.0085866435951374e-05, "loss": 0.0547, "step": 160910 }, { "epoch": 5.440346191554819, "grad_norm": 1.2473335266113281, "learning_rate": 1.0080887698024422e-05, "loss": 0.0609, "step": 160920 }, { "epoch": 5.4406842692450725, "grad_norm": 5.02292537689209, "learning_rate": 1.0075910051471016e-05, "loss": 0.0562, "step": 160930 }, { "epoch": 5.441022346935326, "grad_norm": 1.944431185722351, "learning_rate": 1.0070933496427232e-05, "loss": 0.0576, "step": 160940 }, { "epoch": 5.441360424625579, "grad_norm": 0.8620766401290894, "learning_rate": 1.006595803302911e-05, "loss": 0.0507, "step": 160950 }, { "epoch": 5.441698502315832, "grad_norm": 1.8794702291488647, "learning_rate": 1.0060983661412705e-05, "loss": 0.0545, "step": 160960 }, { "epoch": 5.442036580006086, "grad_norm": 0.9057166576385498, "learning_rate": 1.0056010381714004e-05, "loss": 0.0552, "step": 160970 }, { "epoch": 5.442374657696338, "grad_norm": 0.7574752569198608, "learning_rate": 1.0051038194068956e-05, "loss": 0.0589, "step": 160980 }, { "epoch": 5.442712735386592, "grad_norm": 1.2741031646728516, "learning_rate": 1.0046067098613532e-05, "loss": 0.0581, "step": 160990 }, { "epoch": 5.443050813076845, "grad_norm": 1.2890502214431763, "learning_rate": 1.004109709548361e-05, "loss": 0.0516, "step": 161000 }, { "epoch": 5.443388890767098, "grad_norm": 3.486708164215088, "learning_rate": 1.0036128184815108e-05, "loss": 0.0508, "step": 161010 }, { "epoch": 5.4437269684573515, "grad_norm": 2.0150582790374756, "learning_rate": 1.0031160366743825e-05, "loss": 0.053, "step": 161020 }, { "epoch": 5.444065046147605, "grad_norm": 0.8159366846084595, "learning_rate": 1.0026193641405608e-05, "loss": 0.0484, "step": 161030 }, { "epoch": 5.444403123837858, "grad_norm": 1.1075778007507324, "learning_rate": 1.0021228008936228e-05, "loss": 0.0562, "step": 161040 }, { "epoch": 5.444741201528111, "grad_norm": 0.9597973823547363, "learning_rate": 1.001626346947147e-05, "loss": 0.0604, "step": 161050 }, { "epoch": 5.445079279218365, "grad_norm": 1.3491853475570679, "learning_rate": 1.0011300023147052e-05, "loss": 0.0515, "step": 161060 }, { "epoch": 5.445417356908617, "grad_norm": 0.9915998578071594, "learning_rate": 1.0006337670098648e-05, "loss": 0.0559, "step": 161070 }, { "epoch": 5.445755434598871, "grad_norm": 0.9289300441741943, "learning_rate": 1.0001376410461966e-05, "loss": 0.0537, "step": 161080 }, { "epoch": 5.446093512289124, "grad_norm": 1.118668794631958, "learning_rate": 9.996416244372619e-06, "loss": 0.0546, "step": 161090 }, { "epoch": 5.446431589979377, "grad_norm": 1.0862064361572266, "learning_rate": 9.991457171966228e-06, "loss": 0.0488, "step": 161100 }, { "epoch": 5.44676966766963, "grad_norm": 2.5627193450927734, "learning_rate": 9.986499193378358e-06, "loss": 0.0532, "step": 161110 }, { "epoch": 5.447107745359884, "grad_norm": 1.8584058284759521, "learning_rate": 9.981542308744562e-06, "loss": 0.0545, "step": 161120 }, { "epoch": 5.447445823050137, "grad_norm": 0.9282134771347046, "learning_rate": 9.976586518200377e-06, "loss": 0.0477, "step": 161130 }, { "epoch": 5.44778390074039, "grad_norm": 3.502883195877075, "learning_rate": 9.971631821881282e-06, "loss": 0.0561, "step": 161140 }, { "epoch": 5.448121978430644, "grad_norm": 1.461748480796814, "learning_rate": 9.966678219922737e-06, "loss": 0.055, "step": 161150 }, { "epoch": 5.448460056120896, "grad_norm": 2.779367446899414, "learning_rate": 9.961725712460157e-06, "loss": 0.0566, "step": 161160 }, { "epoch": 5.44879813381115, "grad_norm": 0.8704031109809875, "learning_rate": 9.956774299628962e-06, "loss": 0.0553, "step": 161170 }, { "epoch": 5.449136211501403, "grad_norm": 0.8229036927223206, "learning_rate": 9.951823981564524e-06, "loss": 0.0555, "step": 161180 }, { "epoch": 5.449474289191656, "grad_norm": 10.334455490112305, "learning_rate": 9.946874758402152e-06, "loss": 0.0562, "step": 161190 }, { "epoch": 5.449812366881909, "grad_norm": 1.0344061851501465, "learning_rate": 9.941926630277192e-06, "loss": 0.0495, "step": 161200 }, { "epoch": 5.450150444572163, "grad_norm": 0.9563714265823364, "learning_rate": 9.93697959732489e-06, "loss": 0.0554, "step": 161210 }, { "epoch": 5.4504885222624155, "grad_norm": 1.6059019565582275, "learning_rate": 9.932033659680551e-06, "loss": 0.0578, "step": 161220 }, { "epoch": 5.450826599952669, "grad_norm": 2.7403647899627686, "learning_rate": 9.92708881747933e-06, "loss": 0.0559, "step": 161230 }, { "epoch": 5.451164677642923, "grad_norm": 1.665472388267517, "learning_rate": 9.922145070856443e-06, "loss": 0.0505, "step": 161240 }, { "epoch": 5.451502755333175, "grad_norm": 3.96293568611145, "learning_rate": 9.91720241994707e-06, "loss": 0.0574, "step": 161250 }, { "epoch": 5.451840833023429, "grad_norm": 4.409687519073486, "learning_rate": 9.912260864886319e-06, "loss": 0.0634, "step": 161260 }, { "epoch": 5.452178910713682, "grad_norm": 2.024721622467041, "learning_rate": 9.907320405809301e-06, "loss": 0.0503, "step": 161270 }, { "epoch": 5.452516988403935, "grad_norm": 4.223821640014648, "learning_rate": 9.902381042851061e-06, "loss": 0.0563, "step": 161280 }, { "epoch": 5.452855066094188, "grad_norm": 1.0040180683135986, "learning_rate": 9.897442776146682e-06, "loss": 0.0548, "step": 161290 }, { "epoch": 5.453193143784442, "grad_norm": 1.0762003660202026, "learning_rate": 9.892505605831143e-06, "loss": 0.0486, "step": 161300 }, { "epoch": 5.4535312214746945, "grad_norm": 1.443332314491272, "learning_rate": 9.887569532039426e-06, "loss": 0.048, "step": 161310 }, { "epoch": 5.453869299164948, "grad_norm": 16.611637115478516, "learning_rate": 9.8826345549065e-06, "loss": 0.0618, "step": 161320 }, { "epoch": 5.4542073768552015, "grad_norm": 1.0556834936141968, "learning_rate": 9.877700674567258e-06, "loss": 0.0526, "step": 161330 }, { "epoch": 5.454545454545454, "grad_norm": 1.0896399021148682, "learning_rate": 9.872767891156621e-06, "loss": 0.0518, "step": 161340 }, { "epoch": 5.454883532235708, "grad_norm": 1.928315281867981, "learning_rate": 9.867836204809433e-06, "loss": 0.0568, "step": 161350 }, { "epoch": 5.455221609925961, "grad_norm": 1.0075241327285767, "learning_rate": 9.862905615660518e-06, "loss": 0.0499, "step": 161360 }, { "epoch": 5.455559687616214, "grad_norm": 3.0823814868927, "learning_rate": 9.857976123844698e-06, "loss": 0.0498, "step": 161370 }, { "epoch": 5.455897765306467, "grad_norm": 1.5484412908554077, "learning_rate": 9.853047729496723e-06, "loss": 0.0588, "step": 161380 }, { "epoch": 5.456235842996721, "grad_norm": 1.6489946842193604, "learning_rate": 9.848120432751345e-06, "loss": 0.0583, "step": 161390 }, { "epoch": 5.456573920686974, "grad_norm": 2.6210410594940186, "learning_rate": 9.843194233743258e-06, "loss": 0.0488, "step": 161400 }, { "epoch": 5.456911998377227, "grad_norm": 1.6407893896102905, "learning_rate": 9.838269132607164e-06, "loss": 0.0513, "step": 161410 }, { "epoch": 5.4572500760674805, "grad_norm": 1.3475053310394287, "learning_rate": 9.833345129477694e-06, "loss": 0.0552, "step": 161420 }, { "epoch": 5.457588153757733, "grad_norm": 0.8682962656021118, "learning_rate": 9.828422224489492e-06, "loss": 0.0596, "step": 161430 }, { "epoch": 5.457926231447987, "grad_norm": 1.5126115083694458, "learning_rate": 9.823500417777131e-06, "loss": 0.052, "step": 161440 }, { "epoch": 5.45826430913824, "grad_norm": 2.0959677696228027, "learning_rate": 9.818579709475161e-06, "loss": 0.0566, "step": 161450 }, { "epoch": 5.458602386828494, "grad_norm": 1.9492992162704468, "learning_rate": 9.813660099718148e-06, "loss": 0.0559, "step": 161460 }, { "epoch": 5.458940464518746, "grad_norm": 0.9291053414344788, "learning_rate": 9.808741588640563e-06, "loss": 0.0508, "step": 161470 }, { "epoch": 5.459278542209, "grad_norm": 2.678208827972412, "learning_rate": 9.803824176376886e-06, "loss": 0.056, "step": 161480 }, { "epoch": 5.459616619899252, "grad_norm": 0.7479564547538757, "learning_rate": 9.798907863061541e-06, "loss": 0.0499, "step": 161490 }, { "epoch": 5.459954697589506, "grad_norm": 0.95057612657547, "learning_rate": 9.793992648828953e-06, "loss": 0.0514, "step": 161500 }, { "epoch": 5.460292775279759, "grad_norm": 0.6339238882064819, "learning_rate": 9.789078533813529e-06, "loss": 0.0526, "step": 161510 }, { "epoch": 5.460630852970013, "grad_norm": 1.4257333278656006, "learning_rate": 9.784165518149562e-06, "loss": 0.0606, "step": 161520 }, { "epoch": 5.460968930660266, "grad_norm": 1.5068718194961548, "learning_rate": 9.779253601971416e-06, "loss": 0.0566, "step": 161530 }, { "epoch": 5.461307008350519, "grad_norm": 1.309980034828186, "learning_rate": 9.774342785413349e-06, "loss": 0.0555, "step": 161540 }, { "epoch": 5.461645086040772, "grad_norm": 0.869597852230072, "learning_rate": 9.769433068609657e-06, "loss": 0.0541, "step": 161550 }, { "epoch": 5.461983163731025, "grad_norm": 1.0074609518051147, "learning_rate": 9.764524451694551e-06, "loss": 0.05, "step": 161560 }, { "epoch": 5.462321241421279, "grad_norm": 3.062197685241699, "learning_rate": 9.759616934802213e-06, "loss": 0.0566, "step": 161570 }, { "epoch": 5.462659319111532, "grad_norm": 1.8497040271759033, "learning_rate": 9.754710518066846e-06, "loss": 0.053, "step": 161580 }, { "epoch": 5.462997396801785, "grad_norm": 2.51992130279541, "learning_rate": 9.749805201622574e-06, "loss": 0.0572, "step": 161590 }, { "epoch": 5.463335474492038, "grad_norm": 1.8347622156143188, "learning_rate": 9.744900985603506e-06, "loss": 0.064, "step": 161600 }, { "epoch": 5.463673552182292, "grad_norm": 1.932154655456543, "learning_rate": 9.739997870143713e-06, "loss": 0.0582, "step": 161610 }, { "epoch": 5.4640116298725445, "grad_norm": 0.8886308073997498, "learning_rate": 9.735095855377246e-06, "loss": 0.0566, "step": 161620 }, { "epoch": 5.464349707562798, "grad_norm": 0.5093808174133301, "learning_rate": 9.730194941438153e-06, "loss": 0.0543, "step": 161630 }, { "epoch": 5.4646877852530515, "grad_norm": 0.9296020269393921, "learning_rate": 9.725295128460393e-06, "loss": 0.0568, "step": 161640 }, { "epoch": 5.465025862943304, "grad_norm": 0.8346285820007324, "learning_rate": 9.720396416577937e-06, "loss": 0.0555, "step": 161650 }, { "epoch": 5.465363940633558, "grad_norm": 1.672477126121521, "learning_rate": 9.7154988059247e-06, "loss": 0.0533, "step": 161660 }, { "epoch": 5.465702018323811, "grad_norm": 0.8253046274185181, "learning_rate": 9.710602296634603e-06, "loss": 0.0518, "step": 161670 }, { "epoch": 5.466040096014064, "grad_norm": 11.730449676513672, "learning_rate": 9.705706888841503e-06, "loss": 0.0546, "step": 161680 }, { "epoch": 5.466378173704317, "grad_norm": 3.399731397628784, "learning_rate": 9.700812582679224e-06, "loss": 0.0568, "step": 161690 }, { "epoch": 5.466716251394571, "grad_norm": 0.7124305963516235, "learning_rate": 9.695919378281604e-06, "loss": 0.0567, "step": 161700 }, { "epoch": 5.4670543290848235, "grad_norm": 7.165259838104248, "learning_rate": 9.691027275782394e-06, "loss": 0.0511, "step": 161710 }, { "epoch": 5.467392406775077, "grad_norm": 1.964749813079834, "learning_rate": 9.686136275315383e-06, "loss": 0.0573, "step": 161720 }, { "epoch": 5.4677304844653305, "grad_norm": 0.9683399796485901, "learning_rate": 9.68124637701423e-06, "loss": 0.0589, "step": 161730 }, { "epoch": 5.468068562155583, "grad_norm": 1.3327395915985107, "learning_rate": 9.676357581012657e-06, "loss": 0.0567, "step": 161740 }, { "epoch": 5.468406639845837, "grad_norm": 2.7814526557922363, "learning_rate": 9.671469887444334e-06, "loss": 0.0474, "step": 161750 }, { "epoch": 5.46874471753609, "grad_norm": 4.04279088973999, "learning_rate": 9.666583296442871e-06, "loss": 0.054, "step": 161760 }, { "epoch": 5.469082795226343, "grad_norm": 1.3184326887130737, "learning_rate": 9.661697808141868e-06, "loss": 0.0543, "step": 161770 }, { "epoch": 5.469420872916596, "grad_norm": 2.747445821762085, "learning_rate": 9.656813422674881e-06, "loss": 0.0597, "step": 161780 }, { "epoch": 5.46975895060685, "grad_norm": 1.3208307027816772, "learning_rate": 9.65193014017547e-06, "loss": 0.0539, "step": 161790 }, { "epoch": 5.470097028297102, "grad_norm": 2.245450735092163, "learning_rate": 9.647047960777129e-06, "loss": 0.054, "step": 161800 }, { "epoch": 5.470435105987356, "grad_norm": 1.9535892009735107, "learning_rate": 9.64216688461333e-06, "loss": 0.0519, "step": 161810 }, { "epoch": 5.4707731836776095, "grad_norm": 2.3926706314086914, "learning_rate": 9.637286911817534e-06, "loss": 0.0553, "step": 161820 }, { "epoch": 5.471111261367862, "grad_norm": 2.223684787750244, "learning_rate": 9.632408042523139e-06, "loss": 0.0492, "step": 161830 }, { "epoch": 5.471449339058116, "grad_norm": 6.34535551071167, "learning_rate": 9.627530276863555e-06, "loss": 0.0573, "step": 161840 }, { "epoch": 5.471787416748369, "grad_norm": 3.0877721309661865, "learning_rate": 9.622653614972127e-06, "loss": 0.0515, "step": 161850 }, { "epoch": 5.472125494438622, "grad_norm": 2.297173261642456, "learning_rate": 9.61777805698218e-06, "loss": 0.0559, "step": 161860 }, { "epoch": 5.472463572128875, "grad_norm": 1.9410934448242188, "learning_rate": 9.612903603026996e-06, "loss": 0.0559, "step": 161870 }, { "epoch": 5.472801649819129, "grad_norm": 1.1166496276855469, "learning_rate": 9.608030253239868e-06, "loss": 0.0559, "step": 161880 }, { "epoch": 5.473139727509381, "grad_norm": 0.9925201535224915, "learning_rate": 9.60315800775402e-06, "loss": 0.0511, "step": 161890 }, { "epoch": 5.473477805199635, "grad_norm": 1.4022154808044434, "learning_rate": 9.598286866702644e-06, "loss": 0.0564, "step": 161900 }, { "epoch": 5.473815882889888, "grad_norm": 1.9781428575515747, "learning_rate": 9.593416830218937e-06, "loss": 0.0509, "step": 161910 }, { "epoch": 5.474153960580141, "grad_norm": 1.6164826154708862, "learning_rate": 9.588547898436017e-06, "loss": 0.0595, "step": 161920 }, { "epoch": 5.4744920382703945, "grad_norm": 1.180358648300171, "learning_rate": 9.58368007148705e-06, "loss": 0.0537, "step": 161930 }, { "epoch": 5.474830115960648, "grad_norm": 0.9480934143066406, "learning_rate": 9.578813349505057e-06, "loss": 0.0545, "step": 161940 }, { "epoch": 5.475168193650901, "grad_norm": 1.121569275856018, "learning_rate": 9.573947732623118e-06, "loss": 0.0586, "step": 161950 }, { "epoch": 5.475506271341154, "grad_norm": 3.981109619140625, "learning_rate": 9.569083220974279e-06, "loss": 0.0598, "step": 161960 }, { "epoch": 5.475844349031408, "grad_norm": 1.8647456169128418, "learning_rate": 9.564219814691511e-06, "loss": 0.0496, "step": 161970 }, { "epoch": 5.47618242672166, "grad_norm": 2.5301480293273926, "learning_rate": 9.559357513907785e-06, "loss": 0.0535, "step": 161980 }, { "epoch": 5.476520504411914, "grad_norm": 1.3559726476669312, "learning_rate": 9.554496318756017e-06, "loss": 0.0606, "step": 161990 }, { "epoch": 5.476858582102167, "grad_norm": 1.2277703285217285, "learning_rate": 9.549636229369125e-06, "loss": 0.057, "step": 162000 }, { "epoch": 5.47719665979242, "grad_norm": 3.868725538253784, "learning_rate": 9.54477724588001e-06, "loss": 0.0537, "step": 162010 }, { "epoch": 5.4775347374826735, "grad_norm": 1.5756689310073853, "learning_rate": 9.539919368421456e-06, "loss": 0.0563, "step": 162020 }, { "epoch": 5.477872815172927, "grad_norm": 0.8801721334457397, "learning_rate": 9.535062597126326e-06, "loss": 0.0504, "step": 162030 }, { "epoch": 5.47821089286318, "grad_norm": 1.3169727325439453, "learning_rate": 9.530206932127367e-06, "loss": 0.0539, "step": 162040 }, { "epoch": 5.478548970553433, "grad_norm": 0.7925052046775818, "learning_rate": 9.525352373557356e-06, "loss": 0.0558, "step": 162050 }, { "epoch": 5.478887048243687, "grad_norm": 1.7133798599243164, "learning_rate": 9.52049892154901e-06, "loss": 0.057, "step": 162060 }, { "epoch": 5.479225125933939, "grad_norm": 1.3383948802947998, "learning_rate": 9.515646576235005e-06, "loss": 0.0511, "step": 162070 }, { "epoch": 5.479563203624193, "grad_norm": 0.6095520257949829, "learning_rate": 9.510795337748018e-06, "loss": 0.0545, "step": 162080 }, { "epoch": 5.479901281314446, "grad_norm": 5.038722038269043, "learning_rate": 9.505945206220684e-06, "loss": 0.0459, "step": 162090 }, { "epoch": 5.480239359004699, "grad_norm": 2.916086435317993, "learning_rate": 9.50109618178559e-06, "loss": 0.0506, "step": 162100 }, { "epoch": 5.4805774366949525, "grad_norm": 10.857172012329102, "learning_rate": 9.496248264575302e-06, "loss": 0.0515, "step": 162110 }, { "epoch": 5.480915514385206, "grad_norm": 1.911749005317688, "learning_rate": 9.491401454722382e-06, "loss": 0.0496, "step": 162120 }, { "epoch": 5.481253592075459, "grad_norm": 1.9892500638961792, "learning_rate": 9.486555752359311e-06, "loss": 0.0537, "step": 162130 }, { "epoch": 5.481591669765712, "grad_norm": 0.9224956035614014, "learning_rate": 9.4817111576186e-06, "loss": 0.0559, "step": 162140 }, { "epoch": 5.481929747455966, "grad_norm": 3.1121037006378174, "learning_rate": 9.476867670632688e-06, "loss": 0.0528, "step": 162150 }, { "epoch": 5.482267825146218, "grad_norm": 1.3089241981506348, "learning_rate": 9.472025291533975e-06, "loss": 0.0554, "step": 162160 }, { "epoch": 5.482605902836472, "grad_norm": 1.0287189483642578, "learning_rate": 9.467184020454878e-06, "loss": 0.0569, "step": 162170 }, { "epoch": 5.482943980526725, "grad_norm": 3.04677414894104, "learning_rate": 9.46234385752774e-06, "loss": 0.0506, "step": 162180 }, { "epoch": 5.483282058216978, "grad_norm": 1.5228763818740845, "learning_rate": 9.457504802884882e-06, "loss": 0.055, "step": 162190 }, { "epoch": 5.483620135907231, "grad_norm": 1.178149938583374, "learning_rate": 9.45266685665862e-06, "loss": 0.0552, "step": 162200 }, { "epoch": 5.483958213597485, "grad_norm": 1.6899936199188232, "learning_rate": 9.447830018981202e-06, "loss": 0.0503, "step": 162210 }, { "epoch": 5.4842962912877375, "grad_norm": 2.7819817066192627, "learning_rate": 9.442994289984907e-06, "loss": 0.058, "step": 162220 }, { "epoch": 5.484634368977991, "grad_norm": 11.233778953552246, "learning_rate": 9.438159669801883e-06, "loss": 0.0548, "step": 162230 }, { "epoch": 5.484972446668245, "grad_norm": 1.9638477563858032, "learning_rate": 9.433326158564349e-06, "loss": 0.0555, "step": 162240 }, { "epoch": 5.485310524358497, "grad_norm": 1.1742713451385498, "learning_rate": 9.428493756404422e-06, "loss": 0.0598, "step": 162250 }, { "epoch": 5.485648602048751, "grad_norm": 11.252228736877441, "learning_rate": 9.423662463454247e-06, "loss": 0.058, "step": 162260 }, { "epoch": 5.485986679739004, "grad_norm": 1.862359881401062, "learning_rate": 9.418832279845896e-06, "loss": 0.0539, "step": 162270 }, { "epoch": 5.486324757429257, "grad_norm": 0.698058545589447, "learning_rate": 9.414003205711414e-06, "loss": 0.0504, "step": 162280 }, { "epoch": 5.48666283511951, "grad_norm": 7.951470375061035, "learning_rate": 9.40917524118285e-06, "loss": 0.0599, "step": 162290 }, { "epoch": 5.487000912809764, "grad_norm": 2.1601343154907227, "learning_rate": 9.40434838639218e-06, "loss": 0.0615, "step": 162300 }, { "epoch": 5.4873389905000165, "grad_norm": 0.7448236346244812, "learning_rate": 9.39952264147138e-06, "loss": 0.0534, "step": 162310 }, { "epoch": 5.48767706819027, "grad_norm": 2.1585848331451416, "learning_rate": 9.394698006552365e-06, "loss": 0.0565, "step": 162320 }, { "epoch": 5.4880151458805235, "grad_norm": 1.4183400869369507, "learning_rate": 9.389874481767053e-06, "loss": 0.0567, "step": 162330 }, { "epoch": 5.488353223570776, "grad_norm": 1.4416134357452393, "learning_rate": 9.385052067247324e-06, "loss": 0.0566, "step": 162340 }, { "epoch": 5.48869130126103, "grad_norm": 1.2850826978683472, "learning_rate": 9.380230763125019e-06, "loss": 0.0502, "step": 162350 }, { "epoch": 5.489029378951283, "grad_norm": 0.981234073638916, "learning_rate": 9.375410569531939e-06, "loss": 0.0527, "step": 162360 }, { "epoch": 5.489367456641537, "grad_norm": 1.5111033916473389, "learning_rate": 9.37059148659986e-06, "loss": 0.0537, "step": 162370 }, { "epoch": 5.489705534331789, "grad_norm": 3.8714606761932373, "learning_rate": 9.365773514460564e-06, "loss": 0.0494, "step": 162380 }, { "epoch": 5.490043612022043, "grad_norm": 4.507384300231934, "learning_rate": 9.36095665324575e-06, "loss": 0.0519, "step": 162390 }, { "epoch": 5.4903816897122955, "grad_norm": 0.9415566325187683, "learning_rate": 9.356140903087102e-06, "loss": 0.0504, "step": 162400 }, { "epoch": 5.490719767402549, "grad_norm": 1.61631178855896, "learning_rate": 9.351326264116305e-06, "loss": 0.0533, "step": 162410 }, { "epoch": 5.4910578450928025, "grad_norm": 1.3003532886505127, "learning_rate": 9.346512736464963e-06, "loss": 0.0524, "step": 162420 }, { "epoch": 5.491395922783056, "grad_norm": 1.8895373344421387, "learning_rate": 9.341700320264718e-06, "loss": 0.0563, "step": 162430 }, { "epoch": 5.491734000473309, "grad_norm": 1.3773505687713623, "learning_rate": 9.336889015647087e-06, "loss": 0.056, "step": 162440 }, { "epoch": 5.492072078163562, "grad_norm": 3.598329544067383, "learning_rate": 9.332078822743634e-06, "loss": 0.0542, "step": 162450 }, { "epoch": 5.492410155853815, "grad_norm": 0.9748683571815491, "learning_rate": 9.327269741685879e-06, "loss": 0.0545, "step": 162460 }, { "epoch": 5.492748233544068, "grad_norm": 1.4439111948013306, "learning_rate": 9.322461772605285e-06, "loss": 0.0496, "step": 162470 }, { "epoch": 5.493086311234322, "grad_norm": 4.534870147705078, "learning_rate": 9.317654915633312e-06, "loss": 0.0524, "step": 162480 }, { "epoch": 5.493424388924575, "grad_norm": 1.3249566555023193, "learning_rate": 9.31284917090135e-06, "loss": 0.0545, "step": 162490 }, { "epoch": 5.493762466614828, "grad_norm": 3.4319214820861816, "learning_rate": 9.308044538540822e-06, "loss": 0.0515, "step": 162500 }, { "epoch": 5.494100544305081, "grad_norm": 2.7715659141540527, "learning_rate": 9.303241018683073e-06, "loss": 0.0567, "step": 162510 }, { "epoch": 5.494438621995334, "grad_norm": 0.8633771538734436, "learning_rate": 9.29843861145941e-06, "loss": 0.0511, "step": 162520 }, { "epoch": 5.494776699685588, "grad_norm": 1.3564772605895996, "learning_rate": 9.293637317001152e-06, "loss": 0.0536, "step": 162530 }, { "epoch": 5.495114777375841, "grad_norm": 5.481445789337158, "learning_rate": 9.288837135439548e-06, "loss": 0.0593, "step": 162540 }, { "epoch": 5.495452855066095, "grad_norm": 1.5225919485092163, "learning_rate": 9.284038066905859e-06, "loss": 0.0517, "step": 162550 }, { "epoch": 5.495790932756347, "grad_norm": 5.698759078979492, "learning_rate": 9.279240111531268e-06, "loss": 0.0578, "step": 162560 }, { "epoch": 5.496129010446601, "grad_norm": 1.3358758687973022, "learning_rate": 9.274443269446942e-06, "loss": 0.053, "step": 162570 }, { "epoch": 5.496467088136854, "grad_norm": 1.917863368988037, "learning_rate": 9.269647540784054e-06, "loss": 0.0541, "step": 162580 }, { "epoch": 5.496805165827107, "grad_norm": 1.2832635641098022, "learning_rate": 9.264852925673696e-06, "loss": 0.0529, "step": 162590 }, { "epoch": 5.49714324351736, "grad_norm": 0.9887105822563171, "learning_rate": 9.260059424246959e-06, "loss": 0.059, "step": 162600 }, { "epoch": 5.497481321207614, "grad_norm": 1.8330694437026978, "learning_rate": 9.255267036634879e-06, "loss": 0.0517, "step": 162610 }, { "epoch": 5.4978193988978665, "grad_norm": 1.0049941539764404, "learning_rate": 9.250475762968502e-06, "loss": 0.0583, "step": 162620 }, { "epoch": 5.49815747658812, "grad_norm": 7.484334945678711, "learning_rate": 9.245685603378795e-06, "loss": 0.0548, "step": 162630 }, { "epoch": 5.4984955542783736, "grad_norm": 1.2598696947097778, "learning_rate": 9.240896557996748e-06, "loss": 0.0569, "step": 162640 }, { "epoch": 5.498833631968626, "grad_norm": 0.6825039982795715, "learning_rate": 9.236108626953277e-06, "loss": 0.0479, "step": 162650 }, { "epoch": 5.49917170965888, "grad_norm": 1.0197093486785889, "learning_rate": 9.231321810379267e-06, "loss": 0.052, "step": 162660 }, { "epoch": 5.499509787349133, "grad_norm": 1.6454070806503296, "learning_rate": 9.226536108405615e-06, "loss": 0.0483, "step": 162670 }, { "epoch": 5.499847865039386, "grad_norm": 1.1605091094970703, "learning_rate": 9.221751521163146e-06, "loss": 0.0552, "step": 162680 }, { "epoch": 5.500185942729639, "grad_norm": 1.2255274057388306, "learning_rate": 9.216968048782676e-06, "loss": 0.0595, "step": 162690 }, { "epoch": 5.500524020419893, "grad_norm": 1.644885778427124, "learning_rate": 9.212185691394965e-06, "loss": 0.0572, "step": 162700 }, { "epoch": 5.5008620981101455, "grad_norm": 3.261201858520508, "learning_rate": 9.207404449130769e-06, "loss": 0.0544, "step": 162710 }, { "epoch": 5.501200175800399, "grad_norm": 0.8114252090454102, "learning_rate": 9.202624322120839e-06, "loss": 0.0506, "step": 162720 }, { "epoch": 5.5015382534906525, "grad_norm": 0.8906373381614685, "learning_rate": 9.197845310495812e-06, "loss": 0.056, "step": 162730 }, { "epoch": 5.501876331180905, "grad_norm": 1.9707324504852295, "learning_rate": 9.19306741438637e-06, "loss": 0.0544, "step": 162740 }, { "epoch": 5.502214408871159, "grad_norm": 2.062998056411743, "learning_rate": 9.188290633923124e-06, "loss": 0.0551, "step": 162750 }, { "epoch": 5.502552486561412, "grad_norm": 4.885598659515381, "learning_rate": 9.183514969236696e-06, "loss": 0.0499, "step": 162760 }, { "epoch": 5.502890564251665, "grad_norm": 1.565980315208435, "learning_rate": 9.178740420457633e-06, "loss": 0.0529, "step": 162770 }, { "epoch": 5.503228641941918, "grad_norm": 0.9409155249595642, "learning_rate": 9.173966987716459e-06, "loss": 0.0521, "step": 162780 }, { "epoch": 5.503566719632172, "grad_norm": 10.086396217346191, "learning_rate": 9.1691946711437e-06, "loss": 0.0536, "step": 162790 }, { "epoch": 5.503904797322424, "grad_norm": 9.198476791381836, "learning_rate": 9.164423470869821e-06, "loss": 0.0517, "step": 162800 }, { "epoch": 5.504242875012678, "grad_norm": 3.9152371883392334, "learning_rate": 9.159653387025264e-06, "loss": 0.0547, "step": 162810 }, { "epoch": 5.5045809527029315, "grad_norm": 3.19698429107666, "learning_rate": 9.154884419740428e-06, "loss": 0.0538, "step": 162820 }, { "epoch": 5.504919030393184, "grad_norm": 1.0172157287597656, "learning_rate": 9.150116569145706e-06, "loss": 0.0569, "step": 162830 }, { "epoch": 5.505257108083438, "grad_norm": 1.1786682605743408, "learning_rate": 9.145349835371465e-06, "loss": 0.0491, "step": 162840 }, { "epoch": 5.505595185773691, "grad_norm": 1.1871404647827148, "learning_rate": 9.14058421854801e-06, "loss": 0.0536, "step": 162850 }, { "epoch": 5.505933263463944, "grad_norm": 1.173570990562439, "learning_rate": 9.135819718805627e-06, "loss": 0.0535, "step": 162860 }, { "epoch": 5.506271341154197, "grad_norm": 2.2536165714263916, "learning_rate": 9.131056336274573e-06, "loss": 0.0605, "step": 162870 }, { "epoch": 5.506609418844451, "grad_norm": 2.3504579067230225, "learning_rate": 9.1262940710851e-06, "loss": 0.0524, "step": 162880 }, { "epoch": 5.506947496534703, "grad_norm": 1.9562575817108154, "learning_rate": 9.121532923367387e-06, "loss": 0.0579, "step": 162890 }, { "epoch": 5.507285574224957, "grad_norm": 2.967726469039917, "learning_rate": 9.116772893251595e-06, "loss": 0.0561, "step": 162900 }, { "epoch": 5.50762365191521, "grad_norm": 1.1753926277160645, "learning_rate": 9.112013980867889e-06, "loss": 0.0545, "step": 162910 }, { "epoch": 5.507961729605463, "grad_norm": 0.7172357439994812, "learning_rate": 9.107256186346347e-06, "loss": 0.0503, "step": 162920 }, { "epoch": 5.5082998072957166, "grad_norm": 1.3546795845031738, "learning_rate": 9.102499509817081e-06, "loss": 0.0548, "step": 162930 }, { "epoch": 5.50863788498597, "grad_norm": 1.3086696863174438, "learning_rate": 9.097743951410087e-06, "loss": 0.0509, "step": 162940 }, { "epoch": 5.508975962676223, "grad_norm": 4.810861110687256, "learning_rate": 9.092989511255406e-06, "loss": 0.0545, "step": 162950 }, { "epoch": 5.509314040366476, "grad_norm": 2.053281307220459, "learning_rate": 9.08823618948304e-06, "loss": 0.0552, "step": 162960 }, { "epoch": 5.50965211805673, "grad_norm": 2.6167588233947754, "learning_rate": 9.08348398622293e-06, "loss": 0.0543, "step": 162970 }, { "epoch": 5.509990195746982, "grad_norm": 1.4506996870040894, "learning_rate": 9.078732901604992e-06, "loss": 0.0559, "step": 162980 }, { "epoch": 5.510328273437236, "grad_norm": 1.0158482789993286, "learning_rate": 9.073982935759107e-06, "loss": 0.0546, "step": 162990 }, { "epoch": 5.510666351127489, "grad_norm": 1.6217963695526123, "learning_rate": 9.069234088815171e-06, "loss": 0.0562, "step": 163000 }, { "epoch": 5.511004428817742, "grad_norm": 2.297415018081665, "learning_rate": 9.064486360902996e-06, "loss": 0.0496, "step": 163010 }, { "epoch": 5.5113425065079955, "grad_norm": 0.6580448150634766, "learning_rate": 9.059739752152374e-06, "loss": 0.0513, "step": 163020 }, { "epoch": 5.511680584198249, "grad_norm": 2.0290424823760986, "learning_rate": 9.0549942626931e-06, "loss": 0.059, "step": 163030 }, { "epoch": 5.512018661888502, "grad_norm": 12.835987091064453, "learning_rate": 9.050249892654888e-06, "loss": 0.0503, "step": 163040 }, { "epoch": 5.512356739578755, "grad_norm": 0.9728246331214905, "learning_rate": 9.045506642167468e-06, "loss": 0.0532, "step": 163050 }, { "epoch": 5.512694817269009, "grad_norm": 0.8211826086044312, "learning_rate": 9.040764511360517e-06, "loss": 0.0569, "step": 163060 }, { "epoch": 5.513032894959261, "grad_norm": 1.1548300981521606, "learning_rate": 9.036023500363672e-06, "loss": 0.0544, "step": 163070 }, { "epoch": 5.513370972649515, "grad_norm": 1.5705833435058594, "learning_rate": 9.03128360930654e-06, "loss": 0.0503, "step": 163080 }, { "epoch": 5.513709050339768, "grad_norm": 1.3969517946243286, "learning_rate": 9.02654483831874e-06, "loss": 0.0559, "step": 163090 }, { "epoch": 5.514047128030021, "grad_norm": 1.0875126123428345, "learning_rate": 9.021807187529813e-06, "loss": 0.0558, "step": 163100 }, { "epoch": 5.5143852057202745, "grad_norm": 4.28786039352417, "learning_rate": 9.017070657069265e-06, "loss": 0.0511, "step": 163110 }, { "epoch": 5.514723283410528, "grad_norm": 1.0136090517044067, "learning_rate": 9.012335247066621e-06, "loss": 0.0563, "step": 163120 }, { "epoch": 5.515061361100781, "grad_norm": 1.4433366060256958, "learning_rate": 9.007600957651325e-06, "loss": 0.0579, "step": 163130 }, { "epoch": 5.515399438791034, "grad_norm": 1.6181594133377075, "learning_rate": 9.00286778895284e-06, "loss": 0.0527, "step": 163140 }, { "epoch": 5.515737516481288, "grad_norm": 1.3368210792541504, "learning_rate": 8.998135741100527e-06, "loss": 0.0544, "step": 163150 }, { "epoch": 5.51607559417154, "grad_norm": 1.2443217039108276, "learning_rate": 8.993404814223777e-06, "loss": 0.0484, "step": 163160 }, { "epoch": 5.516413671861794, "grad_norm": 4.387998104095459, "learning_rate": 8.988675008451941e-06, "loss": 0.0626, "step": 163170 }, { "epoch": 5.516751749552047, "grad_norm": 0.8219020366668701, "learning_rate": 8.983946323914333e-06, "loss": 0.0519, "step": 163180 }, { "epoch": 5.5170898272423, "grad_norm": 2.955472230911255, "learning_rate": 8.979218760740215e-06, "loss": 0.0566, "step": 163190 }, { "epoch": 5.517427904932553, "grad_norm": 2.8076446056365967, "learning_rate": 8.974492319058836e-06, "loss": 0.0524, "step": 163200 }, { "epoch": 5.517765982622807, "grad_norm": 1.3858473300933838, "learning_rate": 8.969766998999423e-06, "loss": 0.0551, "step": 163210 }, { "epoch": 5.5181040603130596, "grad_norm": 0.9183757901191711, "learning_rate": 8.965042800691198e-06, "loss": 0.0592, "step": 163220 }, { "epoch": 5.518442138003313, "grad_norm": 1.6725002527236938, "learning_rate": 8.960319724263255e-06, "loss": 0.0546, "step": 163230 }, { "epoch": 5.518780215693567, "grad_norm": 1.83629310131073, "learning_rate": 8.955597769844764e-06, "loss": 0.0573, "step": 163240 }, { "epoch": 5.519118293383819, "grad_norm": 1.4027596712112427, "learning_rate": 8.9508769375648e-06, "loss": 0.0526, "step": 163250 }, { "epoch": 5.519456371074073, "grad_norm": 2.842493772506714, "learning_rate": 8.946157227552453e-06, "loss": 0.0599, "step": 163260 }, { "epoch": 5.519794448764326, "grad_norm": 0.8979275226593018, "learning_rate": 8.941438639936744e-06, "loss": 0.055, "step": 163270 }, { "epoch": 5.52013252645458, "grad_norm": 1.5117346048355103, "learning_rate": 8.936721174846663e-06, "loss": 0.0581, "step": 163280 }, { "epoch": 5.520470604144832, "grad_norm": 1.7412800788879395, "learning_rate": 8.932004832411212e-06, "loss": 0.0507, "step": 163290 }, { "epoch": 5.520808681835086, "grad_norm": 0.943841278553009, "learning_rate": 8.927289612759321e-06, "loss": 0.0549, "step": 163300 }, { "epoch": 5.5211467595253385, "grad_norm": 1.0901069641113281, "learning_rate": 8.922575516019898e-06, "loss": 0.0539, "step": 163310 }, { "epoch": 5.521484837215592, "grad_norm": 1.2910548448562622, "learning_rate": 8.91786254232182e-06, "loss": 0.054, "step": 163320 }, { "epoch": 5.5218229149058455, "grad_norm": 1.070329189300537, "learning_rate": 8.91315069179396e-06, "loss": 0.0526, "step": 163330 }, { "epoch": 5.522160992596099, "grad_norm": 1.123182773590088, "learning_rate": 8.908439964565107e-06, "loss": 0.0528, "step": 163340 }, { "epoch": 5.522499070286352, "grad_norm": 1.5476140975952148, "learning_rate": 8.903730360764085e-06, "loss": 0.0532, "step": 163350 }, { "epoch": 5.522837147976605, "grad_norm": 1.058797836303711, "learning_rate": 8.899021880519637e-06, "loss": 0.0575, "step": 163360 }, { "epoch": 5.523175225666858, "grad_norm": 1.0245147943496704, "learning_rate": 8.894314523960473e-06, "loss": 0.0569, "step": 163370 }, { "epoch": 5.523513303357111, "grad_norm": 1.2015326023101807, "learning_rate": 8.889608291215318e-06, "loss": 0.0566, "step": 163380 }, { "epoch": 5.523851381047365, "grad_norm": 5.517469882965088, "learning_rate": 8.884903182412835e-06, "loss": 0.054, "step": 163390 }, { "epoch": 5.524189458737618, "grad_norm": 1.0467437505722046, "learning_rate": 8.880199197681632e-06, "loss": 0.0476, "step": 163400 }, { "epoch": 5.524527536427871, "grad_norm": 1.132370114326477, "learning_rate": 8.875496337150352e-06, "loss": 0.0468, "step": 163410 }, { "epoch": 5.5248656141181245, "grad_norm": 1.856512427330017, "learning_rate": 8.87079460094754e-06, "loss": 0.06, "step": 163420 }, { "epoch": 5.525203691808377, "grad_norm": 1.1948575973510742, "learning_rate": 8.866093989201778e-06, "loss": 0.0575, "step": 163430 }, { "epoch": 5.525541769498631, "grad_norm": 1.2206993103027344, "learning_rate": 8.861394502041525e-06, "loss": 0.0544, "step": 163440 }, { "epoch": 5.525879847188884, "grad_norm": 1.8037792444229126, "learning_rate": 8.856696139595306e-06, "loss": 0.0565, "step": 163450 }, { "epoch": 5.526217924879138, "grad_norm": 1.2278302907943726, "learning_rate": 8.851998901991542e-06, "loss": 0.0574, "step": 163460 }, { "epoch": 5.52655600256939, "grad_norm": 8.639667510986328, "learning_rate": 8.847302789358685e-06, "loss": 0.0485, "step": 163470 }, { "epoch": 5.526894080259644, "grad_norm": 0.8512406349182129, "learning_rate": 8.84260780182511e-06, "loss": 0.0549, "step": 163480 }, { "epoch": 5.527232157949896, "grad_norm": 0.49846112728118896, "learning_rate": 8.837913939519166e-06, "loss": 0.0484, "step": 163490 }, { "epoch": 5.52757023564015, "grad_norm": 1.5311135053634644, "learning_rate": 8.833221202569198e-06, "loss": 0.0524, "step": 163500 }, { "epoch": 5.5279083133304034, "grad_norm": 1.9502054452896118, "learning_rate": 8.828529591103495e-06, "loss": 0.05, "step": 163510 }, { "epoch": 5.528246391020657, "grad_norm": 1.2644414901733398, "learning_rate": 8.82383910525033e-06, "loss": 0.0522, "step": 163520 }, { "epoch": 5.52858446871091, "grad_norm": 0.9445052742958069, "learning_rate": 8.819149745137922e-06, "loss": 0.0527, "step": 163530 }, { "epoch": 5.528922546401163, "grad_norm": 0.7307634353637695, "learning_rate": 8.81446151089449e-06, "loss": 0.0523, "step": 163540 }, { "epoch": 5.529260624091416, "grad_norm": 1.5087053775787354, "learning_rate": 8.809774402648213e-06, "loss": 0.0539, "step": 163550 }, { "epoch": 5.529598701781669, "grad_norm": 0.9943653345108032, "learning_rate": 8.805088420527235e-06, "loss": 0.051, "step": 163560 }, { "epoch": 5.529936779471923, "grad_norm": 1.4002608060836792, "learning_rate": 8.800403564659665e-06, "loss": 0.0503, "step": 163570 }, { "epoch": 5.530274857162176, "grad_norm": 1.7704037427902222, "learning_rate": 8.795719835173571e-06, "loss": 0.0525, "step": 163580 }, { "epoch": 5.530612934852429, "grad_norm": 0.8662258982658386, "learning_rate": 8.791037232197026e-06, "loss": 0.0588, "step": 163590 }, { "epoch": 5.530951012542682, "grad_norm": 1.2150294780731201, "learning_rate": 8.786355755858044e-06, "loss": 0.0507, "step": 163600 }, { "epoch": 5.531289090232936, "grad_norm": 10.021449089050293, "learning_rate": 8.781675406284596e-06, "loss": 0.0578, "step": 163610 }, { "epoch": 5.5316271679231885, "grad_norm": 1.144860863685608, "learning_rate": 8.776996183604674e-06, "loss": 0.0576, "step": 163620 }, { "epoch": 5.531965245613442, "grad_norm": 1.455489993095398, "learning_rate": 8.772318087946174e-06, "loss": 0.0571, "step": 163630 }, { "epoch": 5.532303323303696, "grad_norm": 1.3964430093765259, "learning_rate": 8.767641119437037e-06, "loss": 0.0516, "step": 163640 }, { "epoch": 5.532641400993948, "grad_norm": 1.249629259109497, "learning_rate": 8.762965278205076e-06, "loss": 0.0474, "step": 163650 }, { "epoch": 5.532979478684202, "grad_norm": 2.7921392917633057, "learning_rate": 8.758290564378152e-06, "loss": 0.0556, "step": 163660 }, { "epoch": 5.533317556374455, "grad_norm": 0.6674310564994812, "learning_rate": 8.753616978084083e-06, "loss": 0.0566, "step": 163670 }, { "epoch": 5.533655634064708, "grad_norm": 0.8175573348999023, "learning_rate": 8.74894451945063e-06, "loss": 0.054, "step": 163680 }, { "epoch": 5.533993711754961, "grad_norm": 0.7773176431655884, "learning_rate": 8.744273188605535e-06, "loss": 0.055, "step": 163690 }, { "epoch": 5.534331789445215, "grad_norm": 2.5579333305358887, "learning_rate": 8.739602985676498e-06, "loss": 0.0514, "step": 163700 }, { "epoch": 5.5346698671354675, "grad_norm": 1.357249140739441, "learning_rate": 8.734933910791232e-06, "loss": 0.0503, "step": 163710 }, { "epoch": 5.535007944825721, "grad_norm": 2.15429949760437, "learning_rate": 8.730265964077367e-06, "loss": 0.0555, "step": 163720 }, { "epoch": 5.5353460225159745, "grad_norm": 1.396806001663208, "learning_rate": 8.725599145662517e-06, "loss": 0.0537, "step": 163730 }, { "epoch": 5.535684100206227, "grad_norm": 0.6659584641456604, "learning_rate": 8.720933455674291e-06, "loss": 0.0535, "step": 163740 }, { "epoch": 5.536022177896481, "grad_norm": 0.7544284462928772, "learning_rate": 8.71626889424022e-06, "loss": 0.0565, "step": 163750 }, { "epoch": 5.536360255586734, "grad_norm": 0.598975419998169, "learning_rate": 8.711605461487865e-06, "loss": 0.0531, "step": 163760 }, { "epoch": 5.536698333276987, "grad_norm": 1.9113898277282715, "learning_rate": 8.706943157544705e-06, "loss": 0.0543, "step": 163770 }, { "epoch": 5.53703641096724, "grad_norm": 3.1392409801483154, "learning_rate": 8.702281982538191e-06, "loss": 0.0555, "step": 163780 }, { "epoch": 5.537374488657494, "grad_norm": 0.6173873543739319, "learning_rate": 8.69762193659579e-06, "loss": 0.0601, "step": 163790 }, { "epoch": 5.5377125663477464, "grad_norm": 0.9197636842727661, "learning_rate": 8.692963019844886e-06, "loss": 0.0549, "step": 163800 }, { "epoch": 5.538050644038, "grad_norm": 1.0222667455673218, "learning_rate": 8.688305232412852e-06, "loss": 0.0524, "step": 163810 }, { "epoch": 5.5383887217282535, "grad_norm": 6.5437211990356445, "learning_rate": 8.683648574427022e-06, "loss": 0.0573, "step": 163820 }, { "epoch": 5.538726799418506, "grad_norm": 1.1692992448806763, "learning_rate": 8.678993046014733e-06, "loss": 0.0522, "step": 163830 }, { "epoch": 5.53906487710876, "grad_norm": 10.087265014648438, "learning_rate": 8.674338647303237e-06, "loss": 0.0633, "step": 163840 }, { "epoch": 5.539402954799013, "grad_norm": 0.920174777507782, "learning_rate": 8.669685378419806e-06, "loss": 0.0591, "step": 163850 }, { "epoch": 5.539741032489266, "grad_norm": 1.2693637609481812, "learning_rate": 8.665033239491655e-06, "loss": 0.0543, "step": 163860 }, { "epoch": 5.540079110179519, "grad_norm": 1.0708165168762207, "learning_rate": 8.660382230645947e-06, "loss": 0.0483, "step": 163870 }, { "epoch": 5.540417187869773, "grad_norm": 1.9211039543151855, "learning_rate": 8.655732352009876e-06, "loss": 0.0571, "step": 163880 }, { "epoch": 5.540755265560025, "grad_norm": 5.110486030578613, "learning_rate": 8.651083603710552e-06, "loss": 0.0517, "step": 163890 }, { "epoch": 5.541093343250279, "grad_norm": 0.723950982093811, "learning_rate": 8.646435985875062e-06, "loss": 0.0555, "step": 163900 }, { "epoch": 5.541431420940532, "grad_norm": 2.500448226928711, "learning_rate": 8.64178949863047e-06, "loss": 0.0492, "step": 163910 }, { "epoch": 5.541769498630785, "grad_norm": 1.010149598121643, "learning_rate": 8.637144142103815e-06, "loss": 0.0534, "step": 163920 }, { "epoch": 5.542107576321039, "grad_norm": 0.9061107039451599, "learning_rate": 8.632499916422121e-06, "loss": 0.0486, "step": 163930 }, { "epoch": 5.542445654011292, "grad_norm": 97.59593200683594, "learning_rate": 8.627856821712316e-06, "loss": 0.0577, "step": 163940 }, { "epoch": 5.542783731701545, "grad_norm": 1.0422604084014893, "learning_rate": 8.623214858101375e-06, "loss": 0.0588, "step": 163950 }, { "epoch": 5.543121809391798, "grad_norm": 0.5068143010139465, "learning_rate": 8.61857402571618e-06, "loss": 0.0495, "step": 163960 }, { "epoch": 5.543459887082052, "grad_norm": 12.216851234436035, "learning_rate": 8.613934324683636e-06, "loss": 0.0515, "step": 163970 }, { "epoch": 5.543797964772304, "grad_norm": 20.504165649414062, "learning_rate": 8.609295755130586e-06, "loss": 0.0615, "step": 163980 }, { "epoch": 5.544136042462558, "grad_norm": 0.6450909972190857, "learning_rate": 8.604658317183823e-06, "loss": 0.0493, "step": 163990 }, { "epoch": 5.544474120152811, "grad_norm": 2.173018217086792, "learning_rate": 8.600022010970165e-06, "loss": 0.0583, "step": 164000 }, { "epoch": 5.544812197843064, "grad_norm": 0.7855659127235413, "learning_rate": 8.595386836616354e-06, "loss": 0.0558, "step": 164010 }, { "epoch": 5.5451502755333175, "grad_norm": 1.4401131868362427, "learning_rate": 8.590752794249108e-06, "loss": 0.0495, "step": 164020 }, { "epoch": 5.545488353223571, "grad_norm": 0.9169313907623291, "learning_rate": 8.586119883995115e-06, "loss": 0.0523, "step": 164030 }, { "epoch": 5.545826430913824, "grad_norm": 3.3329896926879883, "learning_rate": 8.58148810598104e-06, "loss": 0.049, "step": 164040 }, { "epoch": 5.546164508604077, "grad_norm": 1.058032512664795, "learning_rate": 8.576857460333536e-06, "loss": 0.0551, "step": 164050 }, { "epoch": 5.546502586294331, "grad_norm": 1.3331094980239868, "learning_rate": 8.57222794717919e-06, "loss": 0.0562, "step": 164060 }, { "epoch": 5.546840663984583, "grad_norm": 2.27954363822937, "learning_rate": 8.567599566644564e-06, "loss": 0.0568, "step": 164070 }, { "epoch": 5.547178741674837, "grad_norm": 0.9276072978973389, "learning_rate": 8.562972318856189e-06, "loss": 0.0568, "step": 164080 }, { "epoch": 5.54751681936509, "grad_norm": 0.6723279356956482, "learning_rate": 8.558346203940592e-06, "loss": 0.0568, "step": 164090 }, { "epoch": 5.547854897055343, "grad_norm": 2.089801549911499, "learning_rate": 8.553721222024247e-06, "loss": 0.0527, "step": 164100 }, { "epoch": 5.5481929747455965, "grad_norm": 1.8653672933578491, "learning_rate": 8.549097373233578e-06, "loss": 0.056, "step": 164110 }, { "epoch": 5.54853105243585, "grad_norm": 3.591830253601074, "learning_rate": 8.544474657695029e-06, "loss": 0.06, "step": 164120 }, { "epoch": 5.548869130126103, "grad_norm": 1.549335241317749, "learning_rate": 8.539853075534953e-06, "loss": 0.0543, "step": 164130 }, { "epoch": 5.549207207816356, "grad_norm": 1.1598713397979736, "learning_rate": 8.535232626879747e-06, "loss": 0.0559, "step": 164140 }, { "epoch": 5.54954528550661, "grad_norm": 1.194484829902649, "learning_rate": 8.53061331185568e-06, "loss": 0.0536, "step": 164150 }, { "epoch": 5.549883363196862, "grad_norm": 1.064132809638977, "learning_rate": 8.525995130589059e-06, "loss": 0.0561, "step": 164160 }, { "epoch": 5.550221440887116, "grad_norm": 0.7293317914009094, "learning_rate": 8.521378083206167e-06, "loss": 0.0553, "step": 164170 }, { "epoch": 5.550559518577369, "grad_norm": 3.827073574066162, "learning_rate": 8.516762169833215e-06, "loss": 0.0635, "step": 164180 }, { "epoch": 5.550897596267622, "grad_norm": 1.0591002702713013, "learning_rate": 8.512147390596397e-06, "loss": 0.0496, "step": 164190 }, { "epoch": 5.551235673957875, "grad_norm": 2.4835009574890137, "learning_rate": 8.507533745621877e-06, "loss": 0.0491, "step": 164200 }, { "epoch": 5.551573751648129, "grad_norm": 1.1614935398101807, "learning_rate": 8.502921235035805e-06, "loss": 0.0543, "step": 164210 }, { "epoch": 5.551911829338382, "grad_norm": 0.8905147910118103, "learning_rate": 8.498309858964276e-06, "loss": 0.0526, "step": 164220 }, { "epoch": 5.552249907028635, "grad_norm": 4.505281925201416, "learning_rate": 8.493699617533352e-06, "loss": 0.0559, "step": 164230 }, { "epoch": 5.552587984718889, "grad_norm": 1.0616523027420044, "learning_rate": 8.489090510869102e-06, "loss": 0.0532, "step": 164240 }, { "epoch": 5.552926062409142, "grad_norm": 1.8796603679656982, "learning_rate": 8.484482539097505e-06, "loss": 0.05, "step": 164250 }, { "epoch": 5.553264140099395, "grad_norm": 1.9369866847991943, "learning_rate": 8.479875702344575e-06, "loss": 0.0557, "step": 164260 }, { "epoch": 5.553602217789648, "grad_norm": 1.1439738273620605, "learning_rate": 8.475270000736246e-06, "loss": 0.0554, "step": 164270 }, { "epoch": 5.553940295479901, "grad_norm": 1.5291603803634644, "learning_rate": 8.470665434398434e-06, "loss": 0.0551, "step": 164280 }, { "epoch": 5.554278373170154, "grad_norm": 4.248259544372559, "learning_rate": 8.466062003457015e-06, "loss": 0.051, "step": 164290 }, { "epoch": 5.554616450860408, "grad_norm": 0.5946614146232605, "learning_rate": 8.461459708037866e-06, "loss": 0.0557, "step": 164300 }, { "epoch": 5.554954528550661, "grad_norm": 0.9258425831794739, "learning_rate": 8.456858548266805e-06, "loss": 0.0592, "step": 164310 }, { "epoch": 5.555292606240914, "grad_norm": 2.3436524868011475, "learning_rate": 8.452258524269613e-06, "loss": 0.0539, "step": 164320 }, { "epoch": 5.5556306839311675, "grad_norm": 3.081448793411255, "learning_rate": 8.447659636172079e-06, "loss": 0.0522, "step": 164330 }, { "epoch": 5.55596876162142, "grad_norm": 0.9438538551330566, "learning_rate": 8.443061884099906e-06, "loss": 0.0589, "step": 164340 }, { "epoch": 5.556306839311674, "grad_norm": 1.3230804204940796, "learning_rate": 8.43846526817883e-06, "loss": 0.0506, "step": 164350 }, { "epoch": 5.556644917001927, "grad_norm": 1.8685015439987183, "learning_rate": 8.433869788534481e-06, "loss": 0.0579, "step": 164360 }, { "epoch": 5.556982994692181, "grad_norm": 3.395515203475952, "learning_rate": 8.429275445292512e-06, "loss": 0.0548, "step": 164370 }, { "epoch": 5.557321072382433, "grad_norm": 1.1235151290893555, "learning_rate": 8.424682238578546e-06, "loss": 0.0518, "step": 164380 }, { "epoch": 5.557659150072687, "grad_norm": 0.8876902461051941, "learning_rate": 8.420090168518152e-06, "loss": 0.0548, "step": 164390 }, { "epoch": 5.5579972277629395, "grad_norm": 0.6238037347793579, "learning_rate": 8.415499235236874e-06, "loss": 0.0575, "step": 164400 }, { "epoch": 5.558335305453193, "grad_norm": 1.2353483438491821, "learning_rate": 8.41090943886021e-06, "loss": 0.0558, "step": 164410 }, { "epoch": 5.5586733831434465, "grad_norm": 0.6828323602676392, "learning_rate": 8.406320779513655e-06, "loss": 0.0512, "step": 164420 }, { "epoch": 5.5590114608337, "grad_norm": 32.15391159057617, "learning_rate": 8.401733257322692e-06, "loss": 0.0524, "step": 164430 }, { "epoch": 5.559349538523953, "grad_norm": 1.6591328382492065, "learning_rate": 8.397146872412687e-06, "loss": 0.0518, "step": 164440 }, { "epoch": 5.559687616214206, "grad_norm": 1.3679522275924683, "learning_rate": 8.392561624909068e-06, "loss": 0.0543, "step": 164450 }, { "epoch": 5.560025693904459, "grad_norm": 3.6149938106536865, "learning_rate": 8.387977514937174e-06, "loss": 0.0547, "step": 164460 }, { "epoch": 5.560363771594712, "grad_norm": 2.3673927783966064, "learning_rate": 8.383394542622352e-06, "loss": 0.0532, "step": 164470 }, { "epoch": 5.560701849284966, "grad_norm": 3.778350830078125, "learning_rate": 8.378812708089889e-06, "loss": 0.061, "step": 164480 }, { "epoch": 5.561039926975219, "grad_norm": 2.955610513687134, "learning_rate": 8.37423201146504e-06, "loss": 0.0514, "step": 164490 }, { "epoch": 5.561378004665472, "grad_norm": 1.3898546695709229, "learning_rate": 8.369652452873056e-06, "loss": 0.0547, "step": 164500 }, { "epoch": 5.5617160823557255, "grad_norm": 13.389652252197266, "learning_rate": 8.36507403243914e-06, "loss": 0.0506, "step": 164510 }, { "epoch": 5.562054160045978, "grad_norm": 1.3819308280944824, "learning_rate": 8.360496750288454e-06, "loss": 0.0517, "step": 164520 }, { "epoch": 5.562392237736232, "grad_norm": 1.7560070753097534, "learning_rate": 8.355920606546137e-06, "loss": 0.0543, "step": 164530 }, { "epoch": 5.562730315426485, "grad_norm": 0.9098740220069885, "learning_rate": 8.351345601337313e-06, "loss": 0.0524, "step": 164540 }, { "epoch": 5.563068393116739, "grad_norm": 1.019904613494873, "learning_rate": 8.346771734787045e-06, "loss": 0.0575, "step": 164550 }, { "epoch": 5.563406470806991, "grad_norm": 0.49503517150878906, "learning_rate": 8.342199007020396e-06, "loss": 0.0566, "step": 164560 }, { "epoch": 5.563744548497245, "grad_norm": 4.323513031005859, "learning_rate": 8.33762741816238e-06, "loss": 0.057, "step": 164570 }, { "epoch": 5.564082626187498, "grad_norm": 5.615124225616455, "learning_rate": 8.333056968337972e-06, "loss": 0.053, "step": 164580 }, { "epoch": 5.564420703877751, "grad_norm": 1.0868587493896484, "learning_rate": 8.328487657672135e-06, "loss": 0.06, "step": 164590 }, { "epoch": 5.564758781568004, "grad_norm": 1.2265806198120117, "learning_rate": 8.323919486289794e-06, "loss": 0.0552, "step": 164600 }, { "epoch": 5.565096859258258, "grad_norm": 0.8345349431037903, "learning_rate": 8.319352454315826e-06, "loss": 0.058, "step": 164610 }, { "epoch": 5.5654349369485105, "grad_norm": 0.7851117849349976, "learning_rate": 8.314786561875115e-06, "loss": 0.0528, "step": 164620 }, { "epoch": 5.565773014638764, "grad_norm": 0.6399181485176086, "learning_rate": 8.310221809092466e-06, "loss": 0.0538, "step": 164630 }, { "epoch": 5.566111092329018, "grad_norm": 1.9677820205688477, "learning_rate": 8.305658196092719e-06, "loss": 0.0552, "step": 164640 }, { "epoch": 5.56644917001927, "grad_norm": 2.0276176929473877, "learning_rate": 8.301095723000585e-06, "loss": 0.054, "step": 164650 }, { "epoch": 5.566787247709524, "grad_norm": 1.3593534231185913, "learning_rate": 8.29653438994084e-06, "loss": 0.053, "step": 164660 }, { "epoch": 5.567125325399777, "grad_norm": 1.2123677730560303, "learning_rate": 8.291974197038166e-06, "loss": 0.0525, "step": 164670 }, { "epoch": 5.56746340309003, "grad_norm": 0.7265716195106506, "learning_rate": 8.28741514441726e-06, "loss": 0.0521, "step": 164680 }, { "epoch": 5.567801480780283, "grad_norm": 0.7928506731987, "learning_rate": 8.282857232202757e-06, "loss": 0.0534, "step": 164690 }, { "epoch": 5.568139558470537, "grad_norm": 0.8632078766822815, "learning_rate": 8.27830046051925e-06, "loss": 0.057, "step": 164700 }, { "epoch": 5.5684776361607895, "grad_norm": 3.869394302368164, "learning_rate": 8.273744829491348e-06, "loss": 0.0517, "step": 164710 }, { "epoch": 5.568815713851043, "grad_norm": 1.4237266778945923, "learning_rate": 8.269190339243581e-06, "loss": 0.054, "step": 164720 }, { "epoch": 5.5691537915412965, "grad_norm": 0.5498939156532288, "learning_rate": 8.264636989900482e-06, "loss": 0.0567, "step": 164730 }, { "epoch": 5.569491869231549, "grad_norm": 1.5227158069610596, "learning_rate": 8.260084781586508e-06, "loss": 0.054, "step": 164740 }, { "epoch": 5.569829946921803, "grad_norm": 0.997817873954773, "learning_rate": 8.255533714426139e-06, "loss": 0.0592, "step": 164750 }, { "epoch": 5.570168024612056, "grad_norm": 4.564409255981445, "learning_rate": 8.250983788543804e-06, "loss": 0.054, "step": 164760 }, { "epoch": 5.570506102302309, "grad_norm": 0.970122218132019, "learning_rate": 8.24643500406389e-06, "loss": 0.0553, "step": 164770 }, { "epoch": 5.570844179992562, "grad_norm": 0.9826189875602722, "learning_rate": 8.241887361110761e-06, "loss": 0.0517, "step": 164780 }, { "epoch": 5.571182257682816, "grad_norm": 1.7243211269378662, "learning_rate": 8.23734085980873e-06, "loss": 0.055, "step": 164790 }, { "epoch": 5.5715203353730685, "grad_norm": 3.171710729598999, "learning_rate": 8.23279550028212e-06, "loss": 0.0519, "step": 164800 }, { "epoch": 5.571858413063322, "grad_norm": 0.9317070841789246, "learning_rate": 8.228251282655186e-06, "loss": 0.0539, "step": 164810 }, { "epoch": 5.5721964907535755, "grad_norm": 1.4230973720550537, "learning_rate": 8.223708207052167e-06, "loss": 0.0594, "step": 164820 }, { "epoch": 5.572534568443828, "grad_norm": 3.0955986976623535, "learning_rate": 8.219166273597273e-06, "loss": 0.0561, "step": 164830 }, { "epoch": 5.572872646134082, "grad_norm": 1.6155389547348022, "learning_rate": 8.214625482414673e-06, "loss": 0.0543, "step": 164840 }, { "epoch": 5.573210723824335, "grad_norm": 1.1842724084854126, "learning_rate": 8.210085833628534e-06, "loss": 0.0579, "step": 164850 }, { "epoch": 5.573548801514588, "grad_norm": 5.492832660675049, "learning_rate": 8.205547327362923e-06, "loss": 0.0572, "step": 164860 }, { "epoch": 5.573886879204841, "grad_norm": 2.936880111694336, "learning_rate": 8.20100996374195e-06, "loss": 0.0495, "step": 164870 }, { "epoch": 5.574224956895095, "grad_norm": 1.779235601425171, "learning_rate": 8.196473742889672e-06, "loss": 0.0546, "step": 164880 }, { "epoch": 5.574563034585347, "grad_norm": 0.6892966628074646, "learning_rate": 8.191938664930094e-06, "loss": 0.0526, "step": 164890 }, { "epoch": 5.574901112275601, "grad_norm": 0.6813516616821289, "learning_rate": 8.187404729987214e-06, "loss": 0.0583, "step": 164900 }, { "epoch": 5.575239189965854, "grad_norm": 0.9454429745674133, "learning_rate": 8.182871938184961e-06, "loss": 0.058, "step": 164910 }, { "epoch": 5.575577267656107, "grad_norm": 7.808308124542236, "learning_rate": 8.178340289647296e-06, "loss": 0.0562, "step": 164920 }, { "epoch": 5.575915345346361, "grad_norm": 1.6972684860229492, "learning_rate": 8.173809784498093e-06, "loss": 0.0558, "step": 164930 }, { "epoch": 5.576253423036614, "grad_norm": 1.6520357131958008, "learning_rate": 8.169280422861202e-06, "loss": 0.0557, "step": 164940 }, { "epoch": 5.576591500726867, "grad_norm": 1.1304118633270264, "learning_rate": 8.164752204860487e-06, "loss": 0.0685, "step": 164950 }, { "epoch": 5.57692957841712, "grad_norm": 0.8727756142616272, "learning_rate": 8.16022513061972e-06, "loss": 0.0521, "step": 164960 }, { "epoch": 5.577267656107374, "grad_norm": 2.9733529090881348, "learning_rate": 8.155699200262685e-06, "loss": 0.0541, "step": 164970 }, { "epoch": 5.577605733797626, "grad_norm": 2.2384262084960938, "learning_rate": 8.151174413913115e-06, "loss": 0.0519, "step": 164980 }, { "epoch": 5.57794381148788, "grad_norm": 1.92122483253479, "learning_rate": 8.146650771694709e-06, "loss": 0.0565, "step": 164990 }, { "epoch": 5.578281889178133, "grad_norm": 1.4017044305801392, "learning_rate": 8.142128273731153e-06, "loss": 0.0515, "step": 165000 }, { "epoch": 5.578619966868386, "grad_norm": 1.6114110946655273, "learning_rate": 8.137606920146085e-06, "loss": 0.0574, "step": 165010 }, { "epoch": 5.5789580445586395, "grad_norm": 1.3076204061508179, "learning_rate": 8.133086711063114e-06, "loss": 0.0591, "step": 165020 }, { "epoch": 5.579296122248893, "grad_norm": 1.2735533714294434, "learning_rate": 8.128567646605818e-06, "loss": 0.0534, "step": 165030 }, { "epoch": 5.579634199939146, "grad_norm": 0.8107364177703857, "learning_rate": 8.124049726897764e-06, "loss": 0.0579, "step": 165040 }, { "epoch": 5.579972277629399, "grad_norm": 4.292011737823486, "learning_rate": 8.119532952062442e-06, "loss": 0.0559, "step": 165050 }, { "epoch": 5.580310355319653, "grad_norm": 0.9309839010238647, "learning_rate": 8.115017322223372e-06, "loss": 0.052, "step": 165060 }, { "epoch": 5.580648433009905, "grad_norm": 1.4531980752944946, "learning_rate": 8.110502837503986e-06, "loss": 0.0497, "step": 165070 }, { "epoch": 5.580986510700159, "grad_norm": 15.67619514465332, "learning_rate": 8.10598949802771e-06, "loss": 0.0582, "step": 165080 }, { "epoch": 5.581324588390412, "grad_norm": 3.9107296466827393, "learning_rate": 8.101477303917953e-06, "loss": 0.0571, "step": 165090 }, { "epoch": 5.581662666080665, "grad_norm": 1.209499716758728, "learning_rate": 8.096966255298066e-06, "loss": 0.0519, "step": 165100 }, { "epoch": 5.5820007437709185, "grad_norm": 1.6331515312194824, "learning_rate": 8.092456352291378e-06, "loss": 0.0559, "step": 165110 }, { "epoch": 5.582338821461172, "grad_norm": 1.6451150178909302, "learning_rate": 8.087947595021178e-06, "loss": 0.0528, "step": 165120 }, { "epoch": 5.582676899151425, "grad_norm": 2.6884865760803223, "learning_rate": 8.083439983610742e-06, "loss": 0.0555, "step": 165130 }, { "epoch": 5.583014976841678, "grad_norm": 0.9899844527244568, "learning_rate": 8.078933518183334e-06, "loss": 0.0537, "step": 165140 }, { "epoch": 5.583353054531932, "grad_norm": 1.1503548622131348, "learning_rate": 8.074428198862116e-06, "loss": 0.0555, "step": 165150 }, { "epoch": 5.583691132222184, "grad_norm": 2.117021083831787, "learning_rate": 8.069924025770287e-06, "loss": 0.0596, "step": 165160 }, { "epoch": 5.584029209912438, "grad_norm": 0.8034312129020691, "learning_rate": 8.065420999030976e-06, "loss": 0.0531, "step": 165170 }, { "epoch": 5.584367287602691, "grad_norm": 6.422389507293701, "learning_rate": 8.060919118767307e-06, "loss": 0.0551, "step": 165180 }, { "epoch": 5.584705365292944, "grad_norm": 1.7705936431884766, "learning_rate": 8.05641838510236e-06, "loss": 0.0541, "step": 165190 }, { "epoch": 5.585043442983197, "grad_norm": 1.9185328483581543, "learning_rate": 8.051918798159159e-06, "loss": 0.0527, "step": 165200 }, { "epoch": 5.585381520673451, "grad_norm": 1.1304463148117065, "learning_rate": 8.047420358060754e-06, "loss": 0.0482, "step": 165210 }, { "epoch": 5.5857195983637045, "grad_norm": 1.4183924198150635, "learning_rate": 8.042923064930114e-06, "loss": 0.0507, "step": 165220 }, { "epoch": 5.586057676053957, "grad_norm": 1.7784687280654907, "learning_rate": 8.038426918890196e-06, "loss": 0.0629, "step": 165230 }, { "epoch": 5.586395753744211, "grad_norm": 0.7439466714859009, "learning_rate": 8.033931920063909e-06, "loss": 0.0588, "step": 165240 }, { "epoch": 5.586733831434463, "grad_norm": 2.4226744174957275, "learning_rate": 8.029438068574158e-06, "loss": 0.054, "step": 165250 }, { "epoch": 5.587071909124717, "grad_norm": 1.653538465499878, "learning_rate": 8.024945364543812e-06, "loss": 0.0485, "step": 165260 }, { "epoch": 5.58740998681497, "grad_norm": 0.8382819294929504, "learning_rate": 8.020453808095695e-06, "loss": 0.0519, "step": 165270 }, { "epoch": 5.587748064505224, "grad_norm": 3.2284252643585205, "learning_rate": 8.0159633993526e-06, "loss": 0.0529, "step": 165280 }, { "epoch": 5.588086142195476, "grad_norm": 1.1126867532730103, "learning_rate": 8.011474138437274e-06, "loss": 0.0571, "step": 165290 }, { "epoch": 5.58842421988573, "grad_norm": 1.1892380714416504, "learning_rate": 8.006986025472491e-06, "loss": 0.0562, "step": 165300 }, { "epoch": 5.5887622975759825, "grad_norm": 1.1952258348464966, "learning_rate": 8.002499060580932e-06, "loss": 0.0524, "step": 165310 }, { "epoch": 5.589100375266236, "grad_norm": 0.8415890336036682, "learning_rate": 7.998013243885256e-06, "loss": 0.059, "step": 165320 }, { "epoch": 5.5894384529564896, "grad_norm": 2.7247474193573, "learning_rate": 7.99352857550813e-06, "loss": 0.0531, "step": 165330 }, { "epoch": 5.589776530646743, "grad_norm": 4.0166215896606445, "learning_rate": 7.989045055572142e-06, "loss": 0.05, "step": 165340 }, { "epoch": 5.590114608336996, "grad_norm": 0.8092449307441711, "learning_rate": 7.984562684199904e-06, "loss": 0.0577, "step": 165350 }, { "epoch": 5.590452686027249, "grad_norm": 1.8251270055770874, "learning_rate": 7.98008146151391e-06, "loss": 0.0541, "step": 165360 }, { "epoch": 5.590790763717502, "grad_norm": 0.8280951976776123, "learning_rate": 7.975601387636705e-06, "loss": 0.0552, "step": 165370 }, { "epoch": 5.591128841407755, "grad_norm": 1.2750688791275024, "learning_rate": 7.971122462690783e-06, "loss": 0.055, "step": 165380 }, { "epoch": 5.591466919098009, "grad_norm": 11.181873321533203, "learning_rate": 7.966644686798586e-06, "loss": 0.0534, "step": 165390 }, { "epoch": 5.591804996788262, "grad_norm": 1.0541021823883057, "learning_rate": 7.962168060082526e-06, "loss": 0.0471, "step": 165400 }, { "epoch": 5.592143074478515, "grad_norm": 0.8280107378959656, "learning_rate": 7.957692582664989e-06, "loss": 0.0532, "step": 165410 }, { "epoch": 5.5924811521687685, "grad_norm": 0.6698434352874756, "learning_rate": 7.953218254668355e-06, "loss": 0.0519, "step": 165420 }, { "epoch": 5.592819229859021, "grad_norm": 0.6757177710533142, "learning_rate": 7.948745076214936e-06, "loss": 0.0558, "step": 165430 }, { "epoch": 5.593157307549275, "grad_norm": 1.505280613899231, "learning_rate": 7.94427304742702e-06, "loss": 0.0537, "step": 165440 }, { "epoch": 5.593495385239528, "grad_norm": 1.1825929880142212, "learning_rate": 7.939802168426891e-06, "loss": 0.0532, "step": 165450 }, { "epoch": 5.593833462929782, "grad_norm": 341.6217956542969, "learning_rate": 7.935332439336751e-06, "loss": 0.0522, "step": 165460 }, { "epoch": 5.594171540620034, "grad_norm": 2.020317316055298, "learning_rate": 7.930863860278836e-06, "loss": 0.0518, "step": 165470 }, { "epoch": 5.594509618310288, "grad_norm": 0.6934840083122253, "learning_rate": 7.9263964313753e-06, "loss": 0.0506, "step": 165480 }, { "epoch": 5.59484769600054, "grad_norm": 1.4129681587219238, "learning_rate": 7.921930152748275e-06, "loss": 0.0532, "step": 165490 }, { "epoch": 5.595185773690794, "grad_norm": 0.9327914714813232, "learning_rate": 7.917465024519865e-06, "loss": 0.0523, "step": 165500 }, { "epoch": 5.5955238513810475, "grad_norm": 3.152358055114746, "learning_rate": 7.913001046812158e-06, "loss": 0.057, "step": 165510 }, { "epoch": 5.595861929071301, "grad_norm": 0.8613348603248596, "learning_rate": 7.908538219747197e-06, "loss": 0.0486, "step": 165520 }, { "epoch": 5.596200006761554, "grad_norm": 27.283546447753906, "learning_rate": 7.90407654344697e-06, "loss": 0.0568, "step": 165530 }, { "epoch": 5.596538084451807, "grad_norm": 2.322547674179077, "learning_rate": 7.89961601803349e-06, "loss": 0.0547, "step": 165540 }, { "epoch": 5.596876162142061, "grad_norm": 1.139324426651001, "learning_rate": 7.895156643628681e-06, "loss": 0.056, "step": 165550 }, { "epoch": 5.597214239832313, "grad_norm": 2.5026917457580566, "learning_rate": 7.890698420354498e-06, "loss": 0.0573, "step": 165560 }, { "epoch": 5.597552317522567, "grad_norm": 0.9587687849998474, "learning_rate": 7.886241348332768e-06, "loss": 0.051, "step": 165570 }, { "epoch": 5.59789039521282, "grad_norm": 1.4435129165649414, "learning_rate": 7.881785427685384e-06, "loss": 0.0548, "step": 165580 }, { "epoch": 5.598228472903073, "grad_norm": 0.7618182301521301, "learning_rate": 7.877330658534176e-06, "loss": 0.055, "step": 165590 }, { "epoch": 5.598566550593326, "grad_norm": 3.4395318031311035, "learning_rate": 7.87287704100092e-06, "loss": 0.0524, "step": 165600 }, { "epoch": 5.59890462828358, "grad_norm": 1.0867674350738525, "learning_rate": 7.868424575207379e-06, "loss": 0.0571, "step": 165610 }, { "epoch": 5.599242705973833, "grad_norm": 0.5533514022827148, "learning_rate": 7.86397326127527e-06, "loss": 0.0505, "step": 165620 }, { "epoch": 5.599580783664086, "grad_norm": 0.6851968169212341, "learning_rate": 7.859523099326305e-06, "loss": 0.0492, "step": 165630 }, { "epoch": 5.59991886135434, "grad_norm": 1.0120857954025269, "learning_rate": 7.855074089482173e-06, "loss": 0.0502, "step": 165640 }, { "epoch": 5.600256939044592, "grad_norm": 1.4117745161056519, "learning_rate": 7.850626231864455e-06, "loss": 0.0496, "step": 165650 }, { "epoch": 5.600595016734846, "grad_norm": 0.7917114496231079, "learning_rate": 7.846179526594793e-06, "loss": 0.0506, "step": 165660 }, { "epoch": 5.600933094425099, "grad_norm": 3.0660037994384766, "learning_rate": 7.841733973794735e-06, "loss": 0.0501, "step": 165670 }, { "epoch": 5.601271172115352, "grad_norm": 1.1384450197219849, "learning_rate": 7.837289573585843e-06, "loss": 0.0563, "step": 165680 }, { "epoch": 5.601609249805605, "grad_norm": 0.7272506952285767, "learning_rate": 7.832846326089616e-06, "loss": 0.0559, "step": 165690 }, { "epoch": 5.601947327495859, "grad_norm": 1.8600578308105469, "learning_rate": 7.828404231427511e-06, "loss": 0.0587, "step": 165700 }, { "epoch": 5.6022854051861115, "grad_norm": 1.6358742713928223, "learning_rate": 7.82396328972101e-06, "loss": 0.0441, "step": 165710 }, { "epoch": 5.602623482876365, "grad_norm": 4.351259708404541, "learning_rate": 7.819523501091508e-06, "loss": 0.0534, "step": 165720 }, { "epoch": 5.6029615605666185, "grad_norm": 1.1436123847961426, "learning_rate": 7.815084865660383e-06, "loss": 0.055, "step": 165730 }, { "epoch": 5.603299638256871, "grad_norm": 1.0996071100234985, "learning_rate": 7.81064738354898e-06, "loss": 0.0548, "step": 165740 }, { "epoch": 5.603637715947125, "grad_norm": 0.9865462183952332, "learning_rate": 7.806211054878638e-06, "loss": 0.0523, "step": 165750 }, { "epoch": 5.603975793637378, "grad_norm": 0.5068854689598083, "learning_rate": 7.801775879770624e-06, "loss": 0.0542, "step": 165760 }, { "epoch": 5.604313871327631, "grad_norm": 9.45713996887207, "learning_rate": 7.797341858346214e-06, "loss": 0.0491, "step": 165770 }, { "epoch": 5.604651949017884, "grad_norm": 1.3019578456878662, "learning_rate": 7.792908990726627e-06, "loss": 0.0558, "step": 165780 }, { "epoch": 5.604990026708138, "grad_norm": 1.5672824382781982, "learning_rate": 7.788477277033036e-06, "loss": 0.0565, "step": 165790 }, { "epoch": 5.6053281043983905, "grad_norm": 2.2521133422851562, "learning_rate": 7.784046717386634e-06, "loss": 0.0517, "step": 165800 }, { "epoch": 5.605666182088644, "grad_norm": 0.7778737545013428, "learning_rate": 7.779617311908527e-06, "loss": 0.0559, "step": 165810 }, { "epoch": 5.6060042597788975, "grad_norm": 1.2153898477554321, "learning_rate": 7.775189060719817e-06, "loss": 0.0533, "step": 165820 }, { "epoch": 5.60634233746915, "grad_norm": 1.1475610733032227, "learning_rate": 7.770761963941587e-06, "loss": 0.0543, "step": 165830 }, { "epoch": 5.606680415159404, "grad_norm": 3.2098755836486816, "learning_rate": 7.766336021694842e-06, "loss": 0.0536, "step": 165840 }, { "epoch": 5.607018492849657, "grad_norm": 0.9120008945465088, "learning_rate": 7.761911234100633e-06, "loss": 0.0546, "step": 165850 }, { "epoch": 5.60735657053991, "grad_norm": 1.0220545530319214, "learning_rate": 7.757487601279878e-06, "loss": 0.0578, "step": 165860 }, { "epoch": 5.607694648230163, "grad_norm": 1.2786850929260254, "learning_rate": 7.753065123353553e-06, "loss": 0.0552, "step": 165870 }, { "epoch": 5.608032725920417, "grad_norm": 11.669400215148926, "learning_rate": 7.748643800442545e-06, "loss": 0.0579, "step": 165880 }, { "epoch": 5.608370803610669, "grad_norm": 1.1818559169769287, "learning_rate": 7.744223632667752e-06, "loss": 0.0606, "step": 165890 }, { "epoch": 5.608708881300923, "grad_norm": 2.012817144393921, "learning_rate": 7.73980462015001e-06, "loss": 0.056, "step": 165900 }, { "epoch": 5.6090469589911764, "grad_norm": 1.1424744129180908, "learning_rate": 7.735386763010127e-06, "loss": 0.0601, "step": 165910 }, { "epoch": 5.609385036681429, "grad_norm": 1.8491634130477905, "learning_rate": 7.730970061368898e-06, "loss": 0.055, "step": 165920 }, { "epoch": 5.609723114371683, "grad_norm": 1.1399086713790894, "learning_rate": 7.726554515347067e-06, "loss": 0.0567, "step": 165930 }, { "epoch": 5.610061192061936, "grad_norm": 2.4563794136047363, "learning_rate": 7.722140125065358e-06, "loss": 0.057, "step": 165940 }, { "epoch": 5.610399269752189, "grad_norm": 2.453444242477417, "learning_rate": 7.717726890644444e-06, "loss": 0.0517, "step": 165950 }, { "epoch": 5.610737347442442, "grad_norm": 1.1654152870178223, "learning_rate": 7.713314812204992e-06, "loss": 0.0543, "step": 165960 }, { "epoch": 5.611075425132696, "grad_norm": 1.3133642673492432, "learning_rate": 7.708903889867635e-06, "loss": 0.0583, "step": 165970 }, { "epoch": 5.611413502822948, "grad_norm": 4.236378192901611, "learning_rate": 7.70449412375296e-06, "loss": 0.0565, "step": 165980 }, { "epoch": 5.611751580513202, "grad_norm": 1.1486066579818726, "learning_rate": 7.700085513981525e-06, "loss": 0.0502, "step": 165990 }, { "epoch": 5.612089658203455, "grad_norm": 1.113581895828247, "learning_rate": 7.69567806067385e-06, "loss": 0.0509, "step": 166000 }, { "epoch": 5.612427735893708, "grad_norm": 0.9053779244422913, "learning_rate": 7.691271763950453e-06, "loss": 0.0492, "step": 166010 }, { "epoch": 5.6127658135839615, "grad_norm": 1.9090327024459839, "learning_rate": 7.686866623931794e-06, "loss": 0.058, "step": 166020 }, { "epoch": 5.613103891274215, "grad_norm": 1.4751172065734863, "learning_rate": 7.682462640738292e-06, "loss": 0.0527, "step": 166030 }, { "epoch": 5.613441968964468, "grad_norm": 0.9735257029533386, "learning_rate": 7.678059814490373e-06, "loss": 0.0567, "step": 166040 }, { "epoch": 5.613780046654721, "grad_norm": 1.4952508211135864, "learning_rate": 7.673658145308393e-06, "loss": 0.0517, "step": 166050 }, { "epoch": 5.614118124344975, "grad_norm": 1.9059644937515259, "learning_rate": 7.669257633312715e-06, "loss": 0.0553, "step": 166060 }, { "epoch": 5.614456202035227, "grad_norm": 3.4431021213531494, "learning_rate": 7.66485827862361e-06, "loss": 0.0543, "step": 166070 }, { "epoch": 5.614794279725481, "grad_norm": 0.8923485279083252, "learning_rate": 7.660460081361377e-06, "loss": 0.0538, "step": 166080 }, { "epoch": 5.615132357415734, "grad_norm": 0.6692079305648804, "learning_rate": 7.656063041646272e-06, "loss": 0.0545, "step": 166090 }, { "epoch": 5.615470435105987, "grad_norm": 0.8985962867736816, "learning_rate": 7.651667159598491e-06, "loss": 0.0534, "step": 166100 }, { "epoch": 5.6158085127962405, "grad_norm": 1.273293137550354, "learning_rate": 7.647272435338222e-06, "loss": 0.0531, "step": 166110 }, { "epoch": 5.616146590486494, "grad_norm": 0.7820833921432495, "learning_rate": 7.642878868985604e-06, "loss": 0.0561, "step": 166120 }, { "epoch": 5.616484668176747, "grad_norm": 1.1564043760299683, "learning_rate": 7.63848646066077e-06, "loss": 0.0577, "step": 166130 }, { "epoch": 5.616822745867, "grad_norm": 4.737094402313232, "learning_rate": 7.63409521048381e-06, "loss": 0.0536, "step": 166140 }, { "epoch": 5.617160823557254, "grad_norm": 0.5287104249000549, "learning_rate": 7.629705118574754e-06, "loss": 0.0535, "step": 166150 }, { "epoch": 5.617498901247506, "grad_norm": 3.1992788314819336, "learning_rate": 7.625316185053655e-06, "loss": 0.0576, "step": 166160 }, { "epoch": 5.61783697893776, "grad_norm": 0.7132573127746582, "learning_rate": 7.6209284100404785e-06, "loss": 0.0527, "step": 166170 }, { "epoch": 5.618175056628013, "grad_norm": 1.0532101392745972, "learning_rate": 7.616541793655213e-06, "loss": 0.0523, "step": 166180 }, { "epoch": 5.618513134318267, "grad_norm": 0.6955969929695129, "learning_rate": 7.612156336017773e-06, "loss": 0.0538, "step": 166190 }, { "epoch": 5.6188512120085194, "grad_norm": 0.9813593626022339, "learning_rate": 7.6077720372480395e-06, "loss": 0.0509, "step": 166200 }, { "epoch": 5.619189289698773, "grad_norm": 2.931232452392578, "learning_rate": 7.603388897465908e-06, "loss": 0.0554, "step": 166210 }, { "epoch": 5.619527367389026, "grad_norm": 2.215865135192871, "learning_rate": 7.599006916791196e-06, "loss": 0.0513, "step": 166220 }, { "epoch": 5.619865445079279, "grad_norm": 1.2197188138961792, "learning_rate": 7.5946260953436984e-06, "loss": 0.053, "step": 166230 }, { "epoch": 5.620203522769533, "grad_norm": 0.8640730977058411, "learning_rate": 7.5902464332431865e-06, "loss": 0.0534, "step": 166240 }, { "epoch": 5.620541600459786, "grad_norm": 1.1915463209152222, "learning_rate": 7.585867930609419e-06, "loss": 0.0543, "step": 166250 }, { "epoch": 5.620879678150039, "grad_norm": 1.7159016132354736, "learning_rate": 7.581490587562073e-06, "loss": 0.0559, "step": 166260 }, { "epoch": 5.621217755840292, "grad_norm": 1.7629691362380981, "learning_rate": 7.577114404220848e-06, "loss": 0.0531, "step": 166270 }, { "epoch": 5.621555833530545, "grad_norm": 3.0458908081054688, "learning_rate": 7.572739380705379e-06, "loss": 0.0503, "step": 166280 }, { "epoch": 5.621893911220798, "grad_norm": 0.9200678467750549, "learning_rate": 7.568365517135267e-06, "loss": 0.0533, "step": 166290 }, { "epoch": 5.622231988911052, "grad_norm": 1.2402873039245605, "learning_rate": 7.563992813630111e-06, "loss": 0.0568, "step": 166300 }, { "epoch": 5.622570066601305, "grad_norm": 0.7020381689071655, "learning_rate": 7.559621270309447e-06, "loss": 0.0486, "step": 166310 }, { "epoch": 5.622908144291558, "grad_norm": 1.0135418176651, "learning_rate": 7.555250887292797e-06, "loss": 0.0569, "step": 166320 }, { "epoch": 5.623246221981812, "grad_norm": 0.9724254012107849, "learning_rate": 7.550881664699628e-06, "loss": 0.0517, "step": 166330 }, { "epoch": 5.623584299672064, "grad_norm": 1.1953608989715576, "learning_rate": 7.546513602649408e-06, "loss": 0.0523, "step": 166340 }, { "epoch": 5.623922377362318, "grad_norm": 1.09833824634552, "learning_rate": 7.5421467012615766e-06, "loss": 0.0484, "step": 166350 }, { "epoch": 5.624260455052571, "grad_norm": 1.0005395412445068, "learning_rate": 7.537780960655483e-06, "loss": 0.0574, "step": 166360 }, { "epoch": 5.624598532742825, "grad_norm": 0.8177785873413086, "learning_rate": 7.5334163809505195e-06, "loss": 0.0528, "step": 166370 }, { "epoch": 5.624936610433077, "grad_norm": 1.7445863485336304, "learning_rate": 7.529052962265987e-06, "loss": 0.0545, "step": 166380 }, { "epoch": 5.625274688123331, "grad_norm": 2.08638334274292, "learning_rate": 7.524690704721199e-06, "loss": 0.0596, "step": 166390 }, { "epoch": 5.6256127658135835, "grad_norm": 1.4632372856140137, "learning_rate": 7.520329608435411e-06, "loss": 0.0542, "step": 166400 }, { "epoch": 5.625950843503837, "grad_norm": 0.6725773215293884, "learning_rate": 7.51596967352784e-06, "loss": 0.0506, "step": 166410 }, { "epoch": 5.6262889211940905, "grad_norm": 0.7613099217414856, "learning_rate": 7.51161090011771e-06, "loss": 0.0498, "step": 166420 }, { "epoch": 5.626626998884344, "grad_norm": 1.2208503484725952, "learning_rate": 7.507253288324178e-06, "loss": 0.0505, "step": 166430 }, { "epoch": 5.626965076574597, "grad_norm": 0.7145552039146423, "learning_rate": 7.502896838266377e-06, "loss": 0.0513, "step": 166440 }, { "epoch": 5.62730315426485, "grad_norm": 1.053346872329712, "learning_rate": 7.498541550063398e-06, "loss": 0.0549, "step": 166450 }, { "epoch": 5.627641231955103, "grad_norm": 0.692335307598114, "learning_rate": 7.494187423834326e-06, "loss": 0.0526, "step": 166460 }, { "epoch": 5.627979309645356, "grad_norm": 7.571805953979492, "learning_rate": 7.489834459698214e-06, "loss": 0.0513, "step": 166470 }, { "epoch": 5.62831738733561, "grad_norm": 0.8542212247848511, "learning_rate": 7.485482657774057e-06, "loss": 0.054, "step": 166480 }, { "epoch": 5.628655465025863, "grad_norm": 1.3488682508468628, "learning_rate": 7.481132018180831e-06, "loss": 0.0528, "step": 166490 }, { "epoch": 5.628993542716116, "grad_norm": 4.4394211769104, "learning_rate": 7.47678254103747e-06, "loss": 0.0565, "step": 166500 }, { "epoch": 5.6293316204063695, "grad_norm": 2.096007823944092, "learning_rate": 7.4724342264629095e-06, "loss": 0.0542, "step": 166510 }, { "epoch": 5.629669698096623, "grad_norm": 1.9439862966537476, "learning_rate": 7.468087074576025e-06, "loss": 0.0509, "step": 166520 }, { "epoch": 5.630007775786876, "grad_norm": 1.1118061542510986, "learning_rate": 7.463741085495646e-06, "loss": 0.0585, "step": 166530 }, { "epoch": 5.630345853477129, "grad_norm": 6.618424892425537, "learning_rate": 7.4593962593406176e-06, "loss": 0.0547, "step": 166540 }, { "epoch": 5.630683931167383, "grad_norm": 1.5259114503860474, "learning_rate": 7.455052596229706e-06, "loss": 0.0536, "step": 166550 }, { "epoch": 5.631022008857635, "grad_norm": 0.9135314226150513, "learning_rate": 7.450710096281693e-06, "loss": 0.0561, "step": 166560 }, { "epoch": 5.631360086547889, "grad_norm": 0.6988504528999329, "learning_rate": 7.446368759615258e-06, "loss": 0.0544, "step": 166570 }, { "epoch": 5.631698164238142, "grad_norm": 0.7666547298431396, "learning_rate": 7.442028586349115e-06, "loss": 0.0541, "step": 166580 }, { "epoch": 5.632036241928395, "grad_norm": 1.8887535333633423, "learning_rate": 7.437689576601931e-06, "loss": 0.0529, "step": 166590 }, { "epoch": 5.632374319618648, "grad_norm": 0.5436114072799683, "learning_rate": 7.433351730492328e-06, "loss": 0.0484, "step": 166600 }, { "epoch": 5.632712397308902, "grad_norm": 2.784909963607788, "learning_rate": 7.429015048138893e-06, "loss": 0.0476, "step": 166610 }, { "epoch": 5.633050474999155, "grad_norm": 0.8079767227172852, "learning_rate": 7.424679529660183e-06, "loss": 0.0529, "step": 166620 }, { "epoch": 5.633388552689408, "grad_norm": 0.633125364780426, "learning_rate": 7.420345175174753e-06, "loss": 0.0539, "step": 166630 }, { "epoch": 5.633726630379662, "grad_norm": 1.7581428289413452, "learning_rate": 7.416011984801091e-06, "loss": 0.0559, "step": 166640 }, { "epoch": 5.634064708069914, "grad_norm": 0.698306143283844, "learning_rate": 7.4116799586576525e-06, "loss": 0.0534, "step": 166650 }, { "epoch": 5.634402785760168, "grad_norm": 0.7389642596244812, "learning_rate": 7.407349096862892e-06, "loss": 0.0554, "step": 166660 }, { "epoch": 5.634740863450421, "grad_norm": 1.3862465620040894, "learning_rate": 7.403019399535193e-06, "loss": 0.0575, "step": 166670 }, { "epoch": 5.635078941140674, "grad_norm": 4.096547603607178, "learning_rate": 7.398690866792956e-06, "loss": 0.0587, "step": 166680 }, { "epoch": 5.635417018830927, "grad_norm": 5.5926923751831055, "learning_rate": 7.394363498754503e-06, "loss": 0.0554, "step": 166690 }, { "epoch": 5.635755096521181, "grad_norm": 0.9964120388031006, "learning_rate": 7.3900372955381495e-06, "loss": 0.0539, "step": 166700 }, { "epoch": 5.6360931742114335, "grad_norm": 3.484509229660034, "learning_rate": 7.385712257262156e-06, "loss": 0.0572, "step": 166710 }, { "epoch": 5.636431251901687, "grad_norm": 1.0992480516433716, "learning_rate": 7.381388384044791e-06, "loss": 0.0572, "step": 166720 }, { "epoch": 5.6367693295919405, "grad_norm": 1.1653720140457153, "learning_rate": 7.377065676004258e-06, "loss": 0.0548, "step": 166730 }, { "epoch": 5.637107407282193, "grad_norm": 1.3379520177841187, "learning_rate": 7.372744133258719e-06, "loss": 0.0518, "step": 166740 }, { "epoch": 5.637445484972447, "grad_norm": 0.8298848867416382, "learning_rate": 7.368423755926357e-06, "loss": 0.0574, "step": 166750 }, { "epoch": 5.6377835626627, "grad_norm": 2.3018226623535156, "learning_rate": 7.364104544125261e-06, "loss": 0.0516, "step": 166760 }, { "epoch": 5.638121640352953, "grad_norm": 0.7539892792701721, "learning_rate": 7.359786497973558e-06, "loss": 0.0536, "step": 166770 }, { "epoch": 5.638459718043206, "grad_norm": 2.955626964569092, "learning_rate": 7.355469617589245e-06, "loss": 0.0522, "step": 166780 }, { "epoch": 5.63879779573346, "grad_norm": 0.842585027217865, "learning_rate": 7.3511539030903685e-06, "loss": 0.0553, "step": 166790 }, { "epoch": 5.6391358734237125, "grad_norm": 0.9437268376350403, "learning_rate": 7.346839354594936e-06, "loss": 0.0483, "step": 166800 }, { "epoch": 5.639473951113966, "grad_norm": 0.9962479472160339, "learning_rate": 7.342525972220893e-06, "loss": 0.0559, "step": 166810 }, { "epoch": 5.6398120288042195, "grad_norm": 1.4605684280395508, "learning_rate": 7.338213756086165e-06, "loss": 0.0548, "step": 166820 }, { "epoch": 5.640150106494472, "grad_norm": 0.7505987882614136, "learning_rate": 7.333902706308632e-06, "loss": 0.0533, "step": 166830 }, { "epoch": 5.640488184184726, "grad_norm": 1.472512125968933, "learning_rate": 7.329592823006171e-06, "loss": 0.0567, "step": 166840 }, { "epoch": 5.640826261874979, "grad_norm": 0.704188883304596, "learning_rate": 7.3252841062966374e-06, "loss": 0.053, "step": 166850 }, { "epoch": 5.641164339565232, "grad_norm": 0.8405405282974243, "learning_rate": 7.320976556297776e-06, "loss": 0.0497, "step": 166860 }, { "epoch": 5.641502417255485, "grad_norm": 1.2004996538162231, "learning_rate": 7.316670173127394e-06, "loss": 0.0566, "step": 166870 }, { "epoch": 5.641840494945739, "grad_norm": 0.7357946634292603, "learning_rate": 7.312364956903206e-06, "loss": 0.0522, "step": 166880 }, { "epoch": 5.642178572635991, "grad_norm": 0.7649639844894409, "learning_rate": 7.308060907742931e-06, "loss": 0.0581, "step": 166890 }, { "epoch": 5.642516650326245, "grad_norm": 1.0948405265808105, "learning_rate": 7.3037580257642346e-06, "loss": 0.051, "step": 166900 }, { "epoch": 5.6428547280164985, "grad_norm": 0.7199130654335022, "learning_rate": 7.29945631108474e-06, "loss": 0.049, "step": 166910 }, { "epoch": 5.643192805706751, "grad_norm": 3.1599161624908447, "learning_rate": 7.295155763822082e-06, "loss": 0.0501, "step": 166920 }, { "epoch": 5.643530883397005, "grad_norm": 1.0381369590759277, "learning_rate": 7.290856384093825e-06, "loss": 0.0529, "step": 166930 }, { "epoch": 5.643868961087258, "grad_norm": 1.8259506225585938, "learning_rate": 7.286558172017505e-06, "loss": 0.0584, "step": 166940 }, { "epoch": 5.644207038777511, "grad_norm": 0.9325333833694458, "learning_rate": 7.282261127710627e-06, "loss": 0.0538, "step": 166950 }, { "epoch": 5.644545116467764, "grad_norm": 1.3796828985214233, "learning_rate": 7.277965251290691e-06, "loss": 0.0551, "step": 166960 }, { "epoch": 5.644883194158018, "grad_norm": 1.4586825370788574, "learning_rate": 7.273670542875122e-06, "loss": 0.0513, "step": 166970 }, { "epoch": 5.64522127184827, "grad_norm": 2.767040252685547, "learning_rate": 7.269377002581368e-06, "loss": 0.0543, "step": 166980 }, { "epoch": 5.645559349538524, "grad_norm": 1.2392326593399048, "learning_rate": 7.265084630526786e-06, "loss": 0.0521, "step": 166990 }, { "epoch": 5.645897427228777, "grad_norm": 1.2416019439697266, "learning_rate": 7.260793426828727e-06, "loss": 0.0537, "step": 167000 }, { "epoch": 5.64623550491903, "grad_norm": 1.2785835266113281, "learning_rate": 7.25650339160453e-06, "loss": 0.0519, "step": 167010 }, { "epoch": 5.6465735826092835, "grad_norm": 1.464887261390686, "learning_rate": 7.2522145249714735e-06, "loss": 0.0537, "step": 167020 }, { "epoch": 5.646911660299537, "grad_norm": 3.2359983921051025, "learning_rate": 7.2479268270468025e-06, "loss": 0.0533, "step": 167030 }, { "epoch": 5.64724973798979, "grad_norm": 0.9546372294425964, "learning_rate": 7.2436402979477565e-06, "loss": 0.0504, "step": 167040 }, { "epoch": 5.647587815680043, "grad_norm": 1.217228889465332, "learning_rate": 7.239354937791515e-06, "loss": 0.0558, "step": 167050 }, { "epoch": 5.647925893370297, "grad_norm": 2.3834667205810547, "learning_rate": 7.235070746695272e-06, "loss": 0.0539, "step": 167060 }, { "epoch": 5.648263971060549, "grad_norm": 1.066351056098938, "learning_rate": 7.2307877247761e-06, "loss": 0.0545, "step": 167070 }, { "epoch": 5.648602048750803, "grad_norm": 0.8878491520881653, "learning_rate": 7.226505872151135e-06, "loss": 0.0521, "step": 167080 }, { "epoch": 5.648940126441056, "grad_norm": 1.7834147214889526, "learning_rate": 7.22222518893742e-06, "loss": 0.0599, "step": 167090 }, { "epoch": 5.649278204131309, "grad_norm": 0.786751389503479, "learning_rate": 7.217945675252008e-06, "loss": 0.0578, "step": 167100 }, { "epoch": 5.6496162818215625, "grad_norm": 1.7721017599105835, "learning_rate": 7.213667331211893e-06, "loss": 0.0564, "step": 167110 }, { "epoch": 5.649954359511816, "grad_norm": 1.113097071647644, "learning_rate": 7.209390156934021e-06, "loss": 0.0491, "step": 167120 }, { "epoch": 5.650292437202069, "grad_norm": 1.7691826820373535, "learning_rate": 7.205114152535364e-06, "loss": 0.0518, "step": 167130 }, { "epoch": 5.650630514892322, "grad_norm": 1.1780561208724976, "learning_rate": 7.200839318132807e-06, "loss": 0.0554, "step": 167140 }, { "epoch": 5.650968592582576, "grad_norm": 1.9690074920654297, "learning_rate": 7.1965656538432235e-06, "loss": 0.0515, "step": 167150 }, { "epoch": 5.651306670272829, "grad_norm": 2.9118196964263916, "learning_rate": 7.192293159783442e-06, "loss": 0.0488, "step": 167160 }, { "epoch": 5.651644747963082, "grad_norm": 0.88663649559021, "learning_rate": 7.188021836070286e-06, "loss": 0.0475, "step": 167170 }, { "epoch": 5.651982825653335, "grad_norm": 0.8587915897369385, "learning_rate": 7.183751682820539e-06, "loss": 0.0563, "step": 167180 }, { "epoch": 5.652320903343588, "grad_norm": 1.3014209270477295, "learning_rate": 7.17948270015093e-06, "loss": 0.0517, "step": 167190 }, { "epoch": 5.6526589810338415, "grad_norm": 2.7385668754577637, "learning_rate": 7.175214888178184e-06, "loss": 0.051, "step": 167200 }, { "epoch": 5.652997058724095, "grad_norm": 1.6489942073822021, "learning_rate": 7.170948247018966e-06, "loss": 0.057, "step": 167210 }, { "epoch": 5.6533351364143485, "grad_norm": 1.174949049949646, "learning_rate": 7.166682776789935e-06, "loss": 0.0545, "step": 167220 }, { "epoch": 5.653673214104601, "grad_norm": 0.9522877335548401, "learning_rate": 7.162418477607713e-06, "loss": 0.0525, "step": 167230 }, { "epoch": 5.654011291794855, "grad_norm": 0.7823420166969299, "learning_rate": 7.158155349588863e-06, "loss": 0.0484, "step": 167240 }, { "epoch": 5.654349369485107, "grad_norm": 1.0766327381134033, "learning_rate": 7.153893392849959e-06, "loss": 0.0492, "step": 167250 }, { "epoch": 5.654687447175361, "grad_norm": 3.0479440689086914, "learning_rate": 7.149632607507506e-06, "loss": 0.0511, "step": 167260 }, { "epoch": 5.655025524865614, "grad_norm": 0.9174832701683044, "learning_rate": 7.145372993678018e-06, "loss": 0.0494, "step": 167270 }, { "epoch": 5.655363602555868, "grad_norm": 2.8783962726593018, "learning_rate": 7.141114551477912e-06, "loss": 0.0573, "step": 167280 }, { "epoch": 5.65570168024612, "grad_norm": 0.8714588284492493, "learning_rate": 7.1368572810236345e-06, "loss": 0.0556, "step": 167290 }, { "epoch": 5.656039757936374, "grad_norm": 1.1354684829711914, "learning_rate": 7.1326011824315865e-06, "loss": 0.0561, "step": 167300 }, { "epoch": 5.6563778356266265, "grad_norm": 1.5874298810958862, "learning_rate": 7.128346255818119e-06, "loss": 0.048, "step": 167310 }, { "epoch": 5.65671591331688, "grad_norm": 2.796506404876709, "learning_rate": 7.1240925012995565e-06, "loss": 0.0506, "step": 167320 }, { "epoch": 5.657053991007134, "grad_norm": 2.7225770950317383, "learning_rate": 7.119839918992182e-06, "loss": 0.0572, "step": 167330 }, { "epoch": 5.657392068697387, "grad_norm": 2.364534616470337, "learning_rate": 7.1155885090122945e-06, "loss": 0.0559, "step": 167340 }, { "epoch": 5.65773014638764, "grad_norm": 1.3672378063201904, "learning_rate": 7.111338271476098e-06, "loss": 0.0522, "step": 167350 }, { "epoch": 5.658068224077893, "grad_norm": 0.927545428276062, "learning_rate": 7.107089206499795e-06, "loss": 0.0486, "step": 167360 }, { "epoch": 5.658406301768146, "grad_norm": 1.7573243379592896, "learning_rate": 7.102841314199571e-06, "loss": 0.052, "step": 167370 }, { "epoch": 5.658744379458399, "grad_norm": 0.5913464426994324, "learning_rate": 7.098594594691532e-06, "loss": 0.0515, "step": 167380 }, { "epoch": 5.659082457148653, "grad_norm": 1.7757598161697388, "learning_rate": 7.094349048091819e-06, "loss": 0.0507, "step": 167390 }, { "epoch": 5.659420534838906, "grad_norm": 0.5801427960395813, "learning_rate": 7.090104674516479e-06, "loss": 0.0502, "step": 167400 }, { "epoch": 5.659758612529159, "grad_norm": 1.051265001296997, "learning_rate": 7.0858614740815456e-06, "loss": 0.0532, "step": 167410 }, { "epoch": 5.6600966902194125, "grad_norm": 1.7390007972717285, "learning_rate": 7.081619446903049e-06, "loss": 0.0562, "step": 167420 }, { "epoch": 5.660434767909665, "grad_norm": 0.7843343615531921, "learning_rate": 7.077378593096956e-06, "loss": 0.0524, "step": 167430 }, { "epoch": 5.660772845599919, "grad_norm": 1.279402256011963, "learning_rate": 7.073138912779203e-06, "loss": 0.0515, "step": 167440 }, { "epoch": 5.661110923290172, "grad_norm": 2.1225717067718506, "learning_rate": 7.068900406065693e-06, "loss": 0.0541, "step": 167450 }, { "epoch": 5.661449000980426, "grad_norm": 0.9444267153739929, "learning_rate": 7.06466307307233e-06, "loss": 0.0546, "step": 167460 }, { "epoch": 5.661787078670678, "grad_norm": 1.4940948486328125, "learning_rate": 7.060426913914941e-06, "loss": 0.0543, "step": 167470 }, { "epoch": 5.662125156360932, "grad_norm": 1.87342369556427, "learning_rate": 7.05619192870936e-06, "loss": 0.0549, "step": 167480 }, { "epoch": 5.662463234051185, "grad_norm": 0.7920831441879272, "learning_rate": 7.051958117571356e-06, "loss": 0.0459, "step": 167490 }, { "epoch": 5.662801311741438, "grad_norm": 0.9677808284759521, "learning_rate": 7.047725480616668e-06, "loss": 0.0558, "step": 167500 }, { "epoch": 5.6631393894316915, "grad_norm": 1.0519260168075562, "learning_rate": 7.043494017961045e-06, "loss": 0.0554, "step": 167510 }, { "epoch": 5.663477467121945, "grad_norm": 2.293694257736206, "learning_rate": 7.039263729720152e-06, "loss": 0.0528, "step": 167520 }, { "epoch": 5.663815544812198, "grad_norm": 0.9708880186080933, "learning_rate": 7.035034616009656e-06, "loss": 0.0576, "step": 167530 }, { "epoch": 5.664153622502451, "grad_norm": 1.0672098398208618, "learning_rate": 7.030806676945156e-06, "loss": 0.0495, "step": 167540 }, { "epoch": 5.664491700192705, "grad_norm": 2.0020177364349365, "learning_rate": 7.026579912642262e-06, "loss": 0.0517, "step": 167550 }, { "epoch": 5.664829777882957, "grad_norm": 1.0712541341781616, "learning_rate": 7.022354323216546e-06, "loss": 0.0483, "step": 167560 }, { "epoch": 5.665167855573211, "grad_norm": 1.2713130712509155, "learning_rate": 7.0181299087835e-06, "loss": 0.0518, "step": 167570 }, { "epoch": 5.665505933263464, "grad_norm": 1.0343800783157349, "learning_rate": 7.013906669458642e-06, "loss": 0.0526, "step": 167580 }, { "epoch": 5.665844010953717, "grad_norm": 1.8413411378860474, "learning_rate": 7.009684605357419e-06, "loss": 0.0555, "step": 167590 }, { "epoch": 5.66618208864397, "grad_norm": 1.735445499420166, "learning_rate": 7.005463716595279e-06, "loss": 0.0543, "step": 167600 }, { "epoch": 5.666520166334224, "grad_norm": 2.965419054031372, "learning_rate": 7.001244003287605e-06, "loss": 0.0509, "step": 167610 }, { "epoch": 5.666858244024477, "grad_norm": 1.157938003540039, "learning_rate": 6.997025465549756e-06, "loss": 0.0528, "step": 167620 }, { "epoch": 5.66719632171473, "grad_norm": 1.38174569606781, "learning_rate": 6.992808103497089e-06, "loss": 0.0567, "step": 167630 }, { "epoch": 5.667534399404984, "grad_norm": 0.855862021446228, "learning_rate": 6.988591917244891e-06, "loss": 0.0486, "step": 167640 }, { "epoch": 5.667872477095236, "grad_norm": 1.2831943035125732, "learning_rate": 6.9843769069084296e-06, "loss": 0.0531, "step": 167650 }, { "epoch": 5.66821055478549, "grad_norm": 0.8315128087997437, "learning_rate": 6.980163072602935e-06, "loss": 0.0615, "step": 167660 }, { "epoch": 5.668548632475743, "grad_norm": 1.043147325515747, "learning_rate": 6.975950414443616e-06, "loss": 0.0471, "step": 167670 }, { "epoch": 5.668886710165996, "grad_norm": 1.3291782140731812, "learning_rate": 6.971738932545663e-06, "loss": 0.0596, "step": 167680 }, { "epoch": 5.669224787856249, "grad_norm": 0.9035908579826355, "learning_rate": 6.9675286270242e-06, "loss": 0.0534, "step": 167690 }, { "epoch": 5.669562865546503, "grad_norm": 0.8289474248886108, "learning_rate": 6.963319497994342e-06, "loss": 0.0526, "step": 167700 }, { "epoch": 5.6699009432367555, "grad_norm": 0.9083680510520935, "learning_rate": 6.959111545571151e-06, "loss": 0.0564, "step": 167710 }, { "epoch": 5.670239020927009, "grad_norm": 0.7977940440177917, "learning_rate": 6.954904769869686e-06, "loss": 0.0471, "step": 167720 }, { "epoch": 5.670577098617263, "grad_norm": 1.1611498594284058, "learning_rate": 6.950699171004954e-06, "loss": 0.0526, "step": 167730 }, { "epoch": 5.670915176307515, "grad_norm": 1.9784308671951294, "learning_rate": 6.946494749091925e-06, "loss": 0.055, "step": 167740 }, { "epoch": 5.671253253997769, "grad_norm": 0.7321290373802185, "learning_rate": 6.9422915042455685e-06, "loss": 0.0562, "step": 167750 }, { "epoch": 5.671591331688022, "grad_norm": 3.33268404006958, "learning_rate": 6.938089436580769e-06, "loss": 0.0589, "step": 167760 }, { "epoch": 5.671929409378275, "grad_norm": 1.088273525238037, "learning_rate": 6.933888546212452e-06, "loss": 0.0534, "step": 167770 }, { "epoch": 5.672267487068528, "grad_norm": 2.040957450866699, "learning_rate": 6.9296888332554215e-06, "loss": 0.0518, "step": 167780 }, { "epoch": 5.672605564758782, "grad_norm": 1.0037606954574585, "learning_rate": 6.925490297824516e-06, "loss": 0.0529, "step": 167790 }, { "epoch": 5.6729436424490345, "grad_norm": 1.043022871017456, "learning_rate": 6.921292940034541e-06, "loss": 0.0507, "step": 167800 }, { "epoch": 5.673281720139288, "grad_norm": 0.9484987854957581, "learning_rate": 6.917096760000225e-06, "loss": 0.0577, "step": 167810 }, { "epoch": 5.6736197978295415, "grad_norm": 1.8592751026153564, "learning_rate": 6.912901757836298e-06, "loss": 0.0546, "step": 167820 }, { "epoch": 5.673957875519794, "grad_norm": 1.0592079162597656, "learning_rate": 6.9087079336574436e-06, "loss": 0.055, "step": 167830 }, { "epoch": 5.674295953210048, "grad_norm": 0.9456402063369751, "learning_rate": 6.9045152875783336e-06, "loss": 0.0516, "step": 167840 }, { "epoch": 5.674634030900301, "grad_norm": 1.269526481628418, "learning_rate": 6.900323819713589e-06, "loss": 0.0551, "step": 167850 }, { "epoch": 5.674972108590554, "grad_norm": 13.347387313842773, "learning_rate": 6.896133530177784e-06, "loss": 0.0523, "step": 167860 }, { "epoch": 5.675310186280807, "grad_norm": 1.4268851280212402, "learning_rate": 6.891944419085511e-06, "loss": 0.0509, "step": 167870 }, { "epoch": 5.675648263971061, "grad_norm": 4.927706241607666, "learning_rate": 6.887756486551267e-06, "loss": 0.0535, "step": 167880 }, { "epoch": 5.675986341661313, "grad_norm": 1.2542742490768433, "learning_rate": 6.883569732689577e-06, "loss": 0.0515, "step": 167890 }, { "epoch": 5.676324419351567, "grad_norm": 0.9902651309967041, "learning_rate": 6.879384157614887e-06, "loss": 0.0558, "step": 167900 }, { "epoch": 5.6766624970418205, "grad_norm": 1.479204773902893, "learning_rate": 6.875199761441642e-06, "loss": 0.0543, "step": 167910 }, { "epoch": 5.677000574732073, "grad_norm": 0.6529287099838257, "learning_rate": 6.87101654428422e-06, "loss": 0.0503, "step": 167920 }, { "epoch": 5.677338652422327, "grad_norm": 0.8897402286529541, "learning_rate": 6.866834506257014e-06, "loss": 0.0538, "step": 167930 }, { "epoch": 5.67767673011258, "grad_norm": 29.43577766418457, "learning_rate": 6.862653647474348e-06, "loss": 0.0583, "step": 167940 }, { "epoch": 5.678014807802833, "grad_norm": 0.89955735206604, "learning_rate": 6.858473968050516e-06, "loss": 0.0586, "step": 167950 }, { "epoch": 5.678352885493086, "grad_norm": 0.8609687089920044, "learning_rate": 6.854295468099808e-06, "loss": 0.0501, "step": 167960 }, { "epoch": 5.67869096318334, "grad_norm": 1.3804911375045776, "learning_rate": 6.850118147736445e-06, "loss": 0.0551, "step": 167970 }, { "epoch": 5.679029040873592, "grad_norm": 0.7269237041473389, "learning_rate": 6.845942007074646e-06, "loss": 0.048, "step": 167980 }, { "epoch": 5.679367118563846, "grad_norm": 1.6607447862625122, "learning_rate": 6.841767046228587e-06, "loss": 0.0524, "step": 167990 }, { "epoch": 5.679705196254099, "grad_norm": 1.025822401046753, "learning_rate": 6.837593265312387e-06, "loss": 0.0529, "step": 168000 }, { "epoch": 5.680043273944352, "grad_norm": 0.8454108834266663, "learning_rate": 6.833420664440188e-06, "loss": 0.0535, "step": 168010 }, { "epoch": 5.680381351634606, "grad_norm": 3.8385798931121826, "learning_rate": 6.829249243726043e-06, "loss": 0.0575, "step": 168020 }, { "epoch": 5.680719429324859, "grad_norm": 5.170386791229248, "learning_rate": 6.825079003284013e-06, "loss": 0.0556, "step": 168030 }, { "epoch": 5.681057507015112, "grad_norm": 1.9496617317199707, "learning_rate": 6.820909943228088e-06, "loss": 0.0551, "step": 168040 }, { "epoch": 5.681395584705365, "grad_norm": 0.8784122467041016, "learning_rate": 6.816742063672266e-06, "loss": 0.0553, "step": 168050 }, { "epoch": 5.681733662395619, "grad_norm": 1.0312855243682861, "learning_rate": 6.8125753647305115e-06, "loss": 0.0595, "step": 168060 }, { "epoch": 5.682071740085871, "grad_norm": 1.33269202709198, "learning_rate": 6.8084098465167005e-06, "loss": 0.0529, "step": 168070 }, { "epoch": 5.682409817776125, "grad_norm": 0.8665863275527954, "learning_rate": 6.804245509144752e-06, "loss": 0.0521, "step": 168080 }, { "epoch": 5.682747895466378, "grad_norm": 0.9138957262039185, "learning_rate": 6.800082352728487e-06, "loss": 0.0537, "step": 168090 }, { "epoch": 5.683085973156631, "grad_norm": 2.171180248260498, "learning_rate": 6.795920377381754e-06, "loss": 0.048, "step": 168100 }, { "epoch": 5.6834240508468845, "grad_norm": 1.0184953212738037, "learning_rate": 6.7917595832183324e-06, "loss": 0.0588, "step": 168110 }, { "epoch": 5.683762128537138, "grad_norm": 1.4652043581008911, "learning_rate": 6.787599970351955e-06, "loss": 0.0502, "step": 168120 }, { "epoch": 5.6841002062273915, "grad_norm": 1.0743471384048462, "learning_rate": 6.783441538896373e-06, "loss": 0.0552, "step": 168130 }, { "epoch": 5.684438283917644, "grad_norm": 1.397952675819397, "learning_rate": 6.7792842889652595e-06, "loss": 0.0588, "step": 168140 }, { "epoch": 5.684776361607898, "grad_norm": 1.447415828704834, "learning_rate": 6.775128220672283e-06, "loss": 0.0514, "step": 168150 }, { "epoch": 5.68511443929815, "grad_norm": 0.7543726563453674, "learning_rate": 6.770973334131048e-06, "loss": 0.0554, "step": 168160 }, { "epoch": 5.685452516988404, "grad_norm": 0.7595614790916443, "learning_rate": 6.766819629455168e-06, "loss": 0.0529, "step": 168170 }, { "epoch": 5.685790594678657, "grad_norm": 1.8960609436035156, "learning_rate": 6.7626671067581914e-06, "loss": 0.0521, "step": 168180 }, { "epoch": 5.686128672368911, "grad_norm": 0.5160548686981201, "learning_rate": 6.758515766153667e-06, "loss": 0.0555, "step": 168190 }, { "epoch": 5.6864667500591635, "grad_norm": 1.4085590839385986, "learning_rate": 6.75436560775507e-06, "loss": 0.0567, "step": 168200 }, { "epoch": 5.686804827749417, "grad_norm": 1.4822239875793457, "learning_rate": 6.7502166316758595e-06, "loss": 0.0533, "step": 168210 }, { "epoch": 5.68714290543967, "grad_norm": 1.1475664377212524, "learning_rate": 6.746068838029496e-06, "loss": 0.0513, "step": 168220 }, { "epoch": 5.687480983129923, "grad_norm": 0.7275873422622681, "learning_rate": 6.741922226929353e-06, "loss": 0.0496, "step": 168230 }, { "epoch": 5.687819060820177, "grad_norm": 2.235365867614746, "learning_rate": 6.737776798488798e-06, "loss": 0.0538, "step": 168240 }, { "epoch": 5.68815713851043, "grad_norm": 0.5221022367477417, "learning_rate": 6.733632552821184e-06, "loss": 0.0599, "step": 168250 }, { "epoch": 5.688495216200683, "grad_norm": 0.9854851365089417, "learning_rate": 6.729489490039786e-06, "loss": 0.0552, "step": 168260 }, { "epoch": 5.688833293890936, "grad_norm": 13.76037883758545, "learning_rate": 6.72534761025792e-06, "loss": 0.056, "step": 168270 }, { "epoch": 5.689171371581189, "grad_norm": 0.5593861937522888, "learning_rate": 6.7212069135887625e-06, "loss": 0.0561, "step": 168280 }, { "epoch": 5.689509449271442, "grad_norm": 0.9379264712333679, "learning_rate": 6.717067400145566e-06, "loss": 0.049, "step": 168290 }, { "epoch": 5.689847526961696, "grad_norm": 3.361693859100342, "learning_rate": 6.7129290700414685e-06, "loss": 0.0494, "step": 168300 }, { "epoch": 5.6901856046519494, "grad_norm": 1.3615669012069702, "learning_rate": 6.708791923389646e-06, "loss": 0.0494, "step": 168310 }, { "epoch": 5.690523682342202, "grad_norm": 0.7042991518974304, "learning_rate": 6.704655960303185e-06, "loss": 0.0523, "step": 168320 }, { "epoch": 5.690861760032456, "grad_norm": 1.165539026260376, "learning_rate": 6.700521180895153e-06, "loss": 0.0526, "step": 168330 }, { "epoch": 5.691199837722708, "grad_norm": 1.5646708011627197, "learning_rate": 6.696387585278618e-06, "loss": 0.0556, "step": 168340 }, { "epoch": 5.691537915412962, "grad_norm": 1.2269747257232666, "learning_rate": 6.6922551735665755e-06, "loss": 0.0527, "step": 168350 }, { "epoch": 5.691875993103215, "grad_norm": 2.387122631072998, "learning_rate": 6.688123945872011e-06, "loss": 0.0531, "step": 168360 }, { "epoch": 5.692214070793469, "grad_norm": 0.5030258893966675, "learning_rate": 6.68399390230785e-06, "loss": 0.0537, "step": 168370 }, { "epoch": 5.692552148483721, "grad_norm": 1.0948292016983032, "learning_rate": 6.6798650429870265e-06, "loss": 0.0508, "step": 168380 }, { "epoch": 5.692890226173975, "grad_norm": 3.179629325866699, "learning_rate": 6.675737368022428e-06, "loss": 0.0523, "step": 168390 }, { "epoch": 5.6932283038642275, "grad_norm": 2.316720485687256, "learning_rate": 6.67161087752689e-06, "loss": 0.0477, "step": 168400 }, { "epoch": 5.693566381554481, "grad_norm": 2.201608180999756, "learning_rate": 6.667485571613235e-06, "loss": 0.0549, "step": 168410 }, { "epoch": 5.6939044592447345, "grad_norm": 1.3025052547454834, "learning_rate": 6.66336145039424e-06, "loss": 0.0547, "step": 168420 }, { "epoch": 5.694242536934988, "grad_norm": 0.8128229379653931, "learning_rate": 6.659238513982668e-06, "loss": 0.0582, "step": 168430 }, { "epoch": 5.694580614625241, "grad_norm": 2.8322014808654785, "learning_rate": 6.655116762491231e-06, "loss": 0.0497, "step": 168440 }, { "epoch": 5.694918692315494, "grad_norm": 2.0911688804626465, "learning_rate": 6.6509961960326105e-06, "loss": 0.0531, "step": 168450 }, { "epoch": 5.695256770005748, "grad_norm": 0.9262492060661316, "learning_rate": 6.646876814719477e-06, "loss": 0.0561, "step": 168460 }, { "epoch": 5.695594847696, "grad_norm": 1.5266410112380981, "learning_rate": 6.642758618664435e-06, "loss": 0.0495, "step": 168470 }, { "epoch": 5.695932925386254, "grad_norm": 1.3060314655303955, "learning_rate": 6.638641607980106e-06, "loss": 0.0538, "step": 168480 }, { "epoch": 5.696271003076507, "grad_norm": 1.333431601524353, "learning_rate": 6.6345257827789996e-06, "loss": 0.0489, "step": 168490 }, { "epoch": 5.69660908076676, "grad_norm": 1.3252601623535156, "learning_rate": 6.63041114317367e-06, "loss": 0.0571, "step": 168500 }, { "epoch": 5.6969471584570135, "grad_norm": 1.6691460609436035, "learning_rate": 6.626297689276617e-06, "loss": 0.0572, "step": 168510 }, { "epoch": 5.697285236147267, "grad_norm": 0.6977753043174744, "learning_rate": 6.622185421200294e-06, "loss": 0.0536, "step": 168520 }, { "epoch": 5.69762331383752, "grad_norm": 2.6117794513702393, "learning_rate": 6.6180743390571235e-06, "loss": 0.0622, "step": 168530 }, { "epoch": 5.697961391527773, "grad_norm": 0.9501315355300903, "learning_rate": 6.613964442959486e-06, "loss": 0.0517, "step": 168540 }, { "epoch": 5.698299469218027, "grad_norm": 0.9572442770004272, "learning_rate": 6.609855733019776e-06, "loss": 0.0573, "step": 168550 }, { "epoch": 5.698637546908279, "grad_norm": 2.138232707977295, "learning_rate": 6.60574820935031e-06, "loss": 0.0583, "step": 168560 }, { "epoch": 5.698975624598533, "grad_norm": 1.0010966062545776, "learning_rate": 6.601641872063374e-06, "loss": 0.0562, "step": 168570 }, { "epoch": 5.699313702288786, "grad_norm": 0.9331434369087219, "learning_rate": 6.597536721271258e-06, "loss": 0.0529, "step": 168580 }, { "epoch": 5.699651779979039, "grad_norm": 2.5990607738494873, "learning_rate": 6.593432757086171e-06, "loss": 0.0542, "step": 168590 }, { "epoch": 5.6999898576692924, "grad_norm": 1.494061827659607, "learning_rate": 6.589329979620334e-06, "loss": 0.054, "step": 168600 }, { "epoch": 5.700327935359546, "grad_norm": 0.7155889868736267, "learning_rate": 6.585228388985909e-06, "loss": 0.0508, "step": 168610 }, { "epoch": 5.700666013049799, "grad_norm": 1.933110237121582, "learning_rate": 6.581127985295016e-06, "loss": 0.0524, "step": 168620 }, { "epoch": 5.701004090740052, "grad_norm": 1.266298532485962, "learning_rate": 6.577028768659788e-06, "loss": 0.0528, "step": 168630 }, { "epoch": 5.701342168430306, "grad_norm": 1.291789174079895, "learning_rate": 6.572930739192273e-06, "loss": 0.0563, "step": 168640 }, { "epoch": 5.701680246120558, "grad_norm": 1.8295713663101196, "learning_rate": 6.5688338970045225e-06, "loss": 0.0531, "step": 168650 }, { "epoch": 5.702018323810812, "grad_norm": 1.9474549293518066, "learning_rate": 6.564738242208524e-06, "loss": 0.0519, "step": 168660 }, { "epoch": 5.702356401501065, "grad_norm": 1.235599398612976, "learning_rate": 6.5606437749162755e-06, "loss": 0.0568, "step": 168670 }, { "epoch": 5.702694479191318, "grad_norm": 1.3365041017532349, "learning_rate": 6.556550495239694e-06, "loss": 0.056, "step": 168680 }, { "epoch": 5.703032556881571, "grad_norm": 1.1077483892440796, "learning_rate": 6.552458403290717e-06, "loss": 0.0521, "step": 168690 }, { "epoch": 5.703370634571825, "grad_norm": 2.5895278453826904, "learning_rate": 6.548367499181201e-06, "loss": 0.0556, "step": 168700 }, { "epoch": 5.7037087122620775, "grad_norm": 2.241652727127075, "learning_rate": 6.544277783022984e-06, "loss": 0.0538, "step": 168710 }, { "epoch": 5.704046789952331, "grad_norm": 1.151398777961731, "learning_rate": 6.540189254927897e-06, "loss": 0.0519, "step": 168720 }, { "epoch": 5.704384867642585, "grad_norm": 2.1960983276367188, "learning_rate": 6.536101915007708e-06, "loss": 0.054, "step": 168730 }, { "epoch": 5.704722945332837, "grad_norm": 0.6254653334617615, "learning_rate": 6.532015763374161e-06, "loss": 0.0579, "step": 168740 }, { "epoch": 5.705061023023091, "grad_norm": 0.8219990134239197, "learning_rate": 6.527930800138971e-06, "loss": 0.0528, "step": 168750 }, { "epoch": 5.705399100713344, "grad_norm": 1.7613669633865356, "learning_rate": 6.523847025413815e-06, "loss": 0.0525, "step": 168760 }, { "epoch": 5.705737178403597, "grad_norm": 0.9263336062431335, "learning_rate": 6.5197644393103665e-06, "loss": 0.0518, "step": 168770 }, { "epoch": 5.70607525609385, "grad_norm": 0.7580674290657043, "learning_rate": 6.515683041940207e-06, "loss": 0.0515, "step": 168780 }, { "epoch": 5.706413333784104, "grad_norm": 1.5527089834213257, "learning_rate": 6.511602833414948e-06, "loss": 0.0565, "step": 168790 }, { "epoch": 5.7067514114743565, "grad_norm": 1.6419092416763306, "learning_rate": 6.507523813846117e-06, "loss": 0.0518, "step": 168800 }, { "epoch": 5.70708948916461, "grad_norm": 1.728731632232666, "learning_rate": 6.503445983345257e-06, "loss": 0.0497, "step": 168810 }, { "epoch": 5.7074275668548635, "grad_norm": 1.3425514698028564, "learning_rate": 6.499369342023842e-06, "loss": 0.0578, "step": 168820 }, { "epoch": 5.707765644545116, "grad_norm": 3.946986436843872, "learning_rate": 6.495293889993314e-06, "loss": 0.054, "step": 168830 }, { "epoch": 5.70810372223537, "grad_norm": 3.243276834487915, "learning_rate": 6.4912196273651245e-06, "loss": 0.0526, "step": 168840 }, { "epoch": 5.708441799925623, "grad_norm": 1.63675856590271, "learning_rate": 6.487146554250639e-06, "loss": 0.054, "step": 168850 }, { "epoch": 5.708779877615876, "grad_norm": 6.004101276397705, "learning_rate": 6.483074670761225e-06, "loss": 0.0513, "step": 168860 }, { "epoch": 5.709117955306129, "grad_norm": 0.6988430023193359, "learning_rate": 6.479003977008185e-06, "loss": 0.0619, "step": 168870 }, { "epoch": 5.709456032996383, "grad_norm": 1.5820844173431396, "learning_rate": 6.474934473102829e-06, "loss": 0.0533, "step": 168880 }, { "epoch": 5.7097941106866354, "grad_norm": 0.6663766503334045, "learning_rate": 6.470866159156419e-06, "loss": 0.0515, "step": 168890 }, { "epoch": 5.710132188376889, "grad_norm": 0.7386287450790405, "learning_rate": 6.46679903528018e-06, "loss": 0.052, "step": 168900 }, { "epoch": 5.7104702660671425, "grad_norm": 1.0913538932800293, "learning_rate": 6.4627331015853e-06, "loss": 0.0512, "step": 168910 }, { "epoch": 5.710808343757395, "grad_norm": 0.6669141054153442, "learning_rate": 6.458668358182929e-06, "loss": 0.0484, "step": 168920 }, { "epoch": 5.711146421447649, "grad_norm": 3.709047317504883, "learning_rate": 6.4546048051842165e-06, "loss": 0.0503, "step": 168930 }, { "epoch": 5.711484499137902, "grad_norm": 0.7940764427185059, "learning_rate": 6.450542442700252e-06, "loss": 0.0547, "step": 168940 }, { "epoch": 5.711822576828155, "grad_norm": 1.5948765277862549, "learning_rate": 6.446481270842086e-06, "loss": 0.0539, "step": 168950 }, { "epoch": 5.712160654518408, "grad_norm": 2.4750287532806396, "learning_rate": 6.442421289720768e-06, "loss": 0.0543, "step": 168960 }, { "epoch": 5.712498732208662, "grad_norm": 0.688712477684021, "learning_rate": 6.438362499447281e-06, "loss": 0.0575, "step": 168970 }, { "epoch": 5.712836809898914, "grad_norm": 1.1385033130645752, "learning_rate": 6.434304900132615e-06, "loss": 0.0479, "step": 168980 }, { "epoch": 5.713174887589168, "grad_norm": 0.9873184561729431, "learning_rate": 6.430248491887669e-06, "loss": 0.0507, "step": 168990 }, { "epoch": 5.713512965279421, "grad_norm": 3.0013978481292725, "learning_rate": 6.426193274823361e-06, "loss": 0.0573, "step": 169000 }, { "epoch": 5.713851042969674, "grad_norm": 0.4975447654724121, "learning_rate": 6.422139249050568e-06, "loss": 0.0488, "step": 169010 }, { "epoch": 5.714189120659928, "grad_norm": 1.144407868385315, "learning_rate": 6.418086414680114e-06, "loss": 0.0465, "step": 169020 }, { "epoch": 5.714527198350181, "grad_norm": 3.4260289669036865, "learning_rate": 6.414034771822808e-06, "loss": 0.0525, "step": 169030 }, { "epoch": 5.714865276040434, "grad_norm": 0.9183563590049744, "learning_rate": 6.4099843205894014e-06, "loss": 0.0524, "step": 169040 }, { "epoch": 5.715203353730687, "grad_norm": 1.8444697856903076, "learning_rate": 6.405935061090662e-06, "loss": 0.0578, "step": 169050 }, { "epoch": 5.715541431420941, "grad_norm": 0.7942444086074829, "learning_rate": 6.4018869934372835e-06, "loss": 0.0505, "step": 169060 }, { "epoch": 5.715879509111193, "grad_norm": 1.6317089796066284, "learning_rate": 6.397840117739917e-06, "loss": 0.0498, "step": 169070 }, { "epoch": 5.716217586801447, "grad_norm": 0.8038350343704224, "learning_rate": 6.393794434109241e-06, "loss": 0.0522, "step": 169080 }, { "epoch": 5.7165556644917, "grad_norm": 2.9224228858947754, "learning_rate": 6.389749942655826e-06, "loss": 0.0493, "step": 169090 }, { "epoch": 5.716893742181954, "grad_norm": 1.0522617101669312, "learning_rate": 6.385706643490275e-06, "loss": 0.0547, "step": 169100 }, { "epoch": 5.7172318198722065, "grad_norm": 1.2873024940490723, "learning_rate": 6.38166453672312e-06, "loss": 0.0534, "step": 169110 }, { "epoch": 5.71756989756246, "grad_norm": 1.5797712802886963, "learning_rate": 6.377623622464868e-06, "loss": 0.0541, "step": 169120 }, { "epoch": 5.717907975252713, "grad_norm": 1.982438325881958, "learning_rate": 6.373583900825991e-06, "loss": 0.056, "step": 169130 }, { "epoch": 5.718246052942966, "grad_norm": 0.7442300319671631, "learning_rate": 6.369545371916951e-06, "loss": 0.0571, "step": 169140 }, { "epoch": 5.71858413063322, "grad_norm": 0.9678658843040466, "learning_rate": 6.36550803584815e-06, "loss": 0.0498, "step": 169150 }, { "epoch": 5.718922208323473, "grad_norm": 1.7314774990081787, "learning_rate": 6.361471892729954e-06, "loss": 0.0591, "step": 169160 }, { "epoch": 5.719260286013726, "grad_norm": 0.5665121078491211, "learning_rate": 6.357436942672729e-06, "loss": 0.0607, "step": 169170 }, { "epoch": 5.719598363703979, "grad_norm": 0.5879282355308533, "learning_rate": 6.353403185786777e-06, "loss": 0.0541, "step": 169180 }, { "epoch": 5.719936441394232, "grad_norm": 2.0745160579681396, "learning_rate": 6.349370622182393e-06, "loss": 0.0503, "step": 169190 }, { "epoch": 5.7202745190844855, "grad_norm": 0.8610880374908447, "learning_rate": 6.345339251969817e-06, "loss": 0.0522, "step": 169200 }, { "epoch": 5.720612596774739, "grad_norm": 1.9194644689559937, "learning_rate": 6.341309075259255e-06, "loss": 0.0543, "step": 169210 }, { "epoch": 5.7209506744649925, "grad_norm": 0.9380202889442444, "learning_rate": 6.337280092160913e-06, "loss": 0.0564, "step": 169220 }, { "epoch": 5.721288752155245, "grad_norm": 0.6950068473815918, "learning_rate": 6.333252302784931e-06, "loss": 0.0539, "step": 169230 }, { "epoch": 5.721626829845499, "grad_norm": 0.9807898998260498, "learning_rate": 6.329225707241426e-06, "loss": 0.0539, "step": 169240 }, { "epoch": 5.721964907535751, "grad_norm": 1.0682612657546997, "learning_rate": 6.325200305640472e-06, "loss": 0.0543, "step": 169250 }, { "epoch": 5.722302985226005, "grad_norm": 3.9157168865203857, "learning_rate": 6.321176098092135e-06, "loss": 0.0525, "step": 169260 }, { "epoch": 5.722641062916258, "grad_norm": 1.316924810409546, "learning_rate": 6.317153084706457e-06, "loss": 0.0516, "step": 169270 }, { "epoch": 5.722979140606512, "grad_norm": 1.009401798248291, "learning_rate": 6.313131265593375e-06, "loss": 0.0483, "step": 169280 }, { "epoch": 5.723317218296764, "grad_norm": 0.5499650835990906, "learning_rate": 6.309110640862887e-06, "loss": 0.0525, "step": 169290 }, { "epoch": 5.723655295987018, "grad_norm": 4.875337600708008, "learning_rate": 6.305091210624886e-06, "loss": 0.0521, "step": 169300 }, { "epoch": 5.723993373677271, "grad_norm": 1.1070572137832642, "learning_rate": 6.301072974989291e-06, "loss": 0.0503, "step": 169310 }, { "epoch": 5.724331451367524, "grad_norm": 0.8019848465919495, "learning_rate": 6.297055934065943e-06, "loss": 0.059, "step": 169320 }, { "epoch": 5.724669529057778, "grad_norm": 0.7783927917480469, "learning_rate": 6.293040087964658e-06, "loss": 0.0549, "step": 169330 }, { "epoch": 5.725007606748031, "grad_norm": 2.127046823501587, "learning_rate": 6.289025436795243e-06, "loss": 0.0553, "step": 169340 }, { "epoch": 5.725345684438284, "grad_norm": 1.07439124584198, "learning_rate": 6.285011980667454e-06, "loss": 0.0476, "step": 169350 }, { "epoch": 5.725683762128537, "grad_norm": 1.115317463874817, "learning_rate": 6.280999719691016e-06, "loss": 0.0544, "step": 169360 }, { "epoch": 5.72602183981879, "grad_norm": 0.9017830491065979, "learning_rate": 6.276988653975607e-06, "loss": 0.0516, "step": 169370 }, { "epoch": 5.726359917509043, "grad_norm": 0.9223926663398743, "learning_rate": 6.272978783630917e-06, "loss": 0.0487, "step": 169380 }, { "epoch": 5.726697995199297, "grad_norm": 1.2085864543914795, "learning_rate": 6.268970108766542e-06, "loss": 0.0533, "step": 169390 }, { "epoch": 5.72703607288955, "grad_norm": 0.8535234332084656, "learning_rate": 6.26496262949211e-06, "loss": 0.0481, "step": 169400 }, { "epoch": 5.727374150579803, "grad_norm": 0.6019396185874939, "learning_rate": 6.260956345917173e-06, "loss": 0.0605, "step": 169410 }, { "epoch": 5.7277122282700565, "grad_norm": 0.9012186527252197, "learning_rate": 6.256951258151245e-06, "loss": 0.0541, "step": 169420 }, { "epoch": 5.72805030596031, "grad_norm": 0.951196551322937, "learning_rate": 6.252947366303841e-06, "loss": 0.0571, "step": 169430 }, { "epoch": 5.728388383650563, "grad_norm": 0.9634355306625366, "learning_rate": 6.24894467048443e-06, "loss": 0.0485, "step": 169440 }, { "epoch": 5.728726461340816, "grad_norm": 0.8825631737709045, "learning_rate": 6.2449431708024195e-06, "loss": 0.0553, "step": 169450 }, { "epoch": 5.72906453903107, "grad_norm": 4.140778064727783, "learning_rate": 6.240942867367233e-06, "loss": 0.0582, "step": 169460 }, { "epoch": 5.729402616721322, "grad_norm": 0.644932210445404, "learning_rate": 6.236943760288222e-06, "loss": 0.0512, "step": 169470 }, { "epoch": 5.729740694411576, "grad_norm": 5.085697650909424, "learning_rate": 6.232945849674748e-06, "loss": 0.0519, "step": 169480 }, { "epoch": 5.730078772101829, "grad_norm": 1.839107871055603, "learning_rate": 6.22894913563607e-06, "loss": 0.053, "step": 169490 }, { "epoch": 5.730416849792082, "grad_norm": 1.0557963848114014, "learning_rate": 6.2249536182814926e-06, "loss": 0.0572, "step": 169500 }, { "epoch": 5.7307549274823355, "grad_norm": 0.9138804078102112, "learning_rate": 6.220959297720219e-06, "loss": 0.0562, "step": 169510 }, { "epoch": 5.731093005172589, "grad_norm": 0.9347957968711853, "learning_rate": 6.216966174061489e-06, "loss": 0.0495, "step": 169520 }, { "epoch": 5.731431082862842, "grad_norm": 0.9671232104301453, "learning_rate": 6.212974247414455e-06, "loss": 0.0513, "step": 169530 }, { "epoch": 5.731769160553095, "grad_norm": 1.255172610282898, "learning_rate": 6.208983517888239e-06, "loss": 0.0563, "step": 169540 }, { "epoch": 5.732107238243349, "grad_norm": 1.3315951824188232, "learning_rate": 6.204993985591978e-06, "loss": 0.0543, "step": 169550 }, { "epoch": 5.732445315933601, "grad_norm": 1.222800374031067, "learning_rate": 6.201005650634728e-06, "loss": 0.0613, "step": 169560 }, { "epoch": 5.732783393623855, "grad_norm": 0.516313374042511, "learning_rate": 6.19701851312553e-06, "loss": 0.0589, "step": 169570 }, { "epoch": 5.733121471314108, "grad_norm": 0.37833911180496216, "learning_rate": 6.193032573173374e-06, "loss": 0.0489, "step": 169580 }, { "epoch": 5.733459549004361, "grad_norm": 1.0132784843444824, "learning_rate": 6.1890478308872564e-06, "loss": 0.0475, "step": 169590 }, { "epoch": 5.7337976266946145, "grad_norm": 1.0791124105453491, "learning_rate": 6.185064286376119e-06, "loss": 0.0529, "step": 169600 }, { "epoch": 5.734135704384868, "grad_norm": 0.8078890442848206, "learning_rate": 6.181081939748862e-06, "loss": 0.0497, "step": 169610 }, { "epoch": 5.734473782075121, "grad_norm": 0.8416943550109863, "learning_rate": 6.177100791114365e-06, "loss": 0.0486, "step": 169620 }, { "epoch": 5.734811859765374, "grad_norm": 1.0496653318405151, "learning_rate": 6.173120840581459e-06, "loss": 0.0536, "step": 169630 }, { "epoch": 5.735149937455628, "grad_norm": 1.218367338180542, "learning_rate": 6.169142088258972e-06, "loss": 0.0485, "step": 169640 }, { "epoch": 5.73548801514588, "grad_norm": 1.073349118232727, "learning_rate": 6.16516453425568e-06, "loss": 0.0578, "step": 169650 }, { "epoch": 5.735826092836134, "grad_norm": 0.7504022121429443, "learning_rate": 6.161188178680305e-06, "loss": 0.0521, "step": 169660 }, { "epoch": 5.736164170526387, "grad_norm": 2.151552438735962, "learning_rate": 6.157213021641589e-06, "loss": 0.0606, "step": 169670 }, { "epoch": 5.73650224821664, "grad_norm": 1.2431048154830933, "learning_rate": 6.153239063248184e-06, "loss": 0.0465, "step": 169680 }, { "epoch": 5.736840325906893, "grad_norm": 0.527899444103241, "learning_rate": 6.149266303608775e-06, "loss": 0.055, "step": 169690 }, { "epoch": 5.737178403597147, "grad_norm": 1.2364088296890259, "learning_rate": 6.14529474283193e-06, "loss": 0.0563, "step": 169700 }, { "epoch": 5.7375164812873996, "grad_norm": 1.2901034355163574, "learning_rate": 6.141324381026248e-06, "loss": 0.0516, "step": 169710 }, { "epoch": 5.737854558977653, "grad_norm": 1.8237512111663818, "learning_rate": 6.137355218300295e-06, "loss": 0.0518, "step": 169720 }, { "epoch": 5.738192636667907, "grad_norm": 0.7379084825515747, "learning_rate": 6.133387254762568e-06, "loss": 0.0517, "step": 169730 }, { "epoch": 5.738530714358159, "grad_norm": 1.3547555208206177, "learning_rate": 6.129420490521559e-06, "loss": 0.0549, "step": 169740 }, { "epoch": 5.738868792048413, "grad_norm": 1.397633671760559, "learning_rate": 6.125454925685698e-06, "loss": 0.0568, "step": 169750 }, { "epoch": 5.739206869738666, "grad_norm": 1.1192981004714966, "learning_rate": 6.12149056036343e-06, "loss": 0.0548, "step": 169760 }, { "epoch": 5.739544947428919, "grad_norm": 1.031078815460205, "learning_rate": 6.11752739466312e-06, "loss": 0.0603, "step": 169770 }, { "epoch": 5.739883025119172, "grad_norm": 1.2601631879806519, "learning_rate": 6.113565428693119e-06, "loss": 0.0497, "step": 169780 }, { "epoch": 5.740221102809426, "grad_norm": 4.007390975952148, "learning_rate": 6.109604662561763e-06, "loss": 0.0571, "step": 169790 }, { "epoch": 5.7405591804996785, "grad_norm": 1.26991605758667, "learning_rate": 6.105645096377316e-06, "loss": 0.0533, "step": 169800 }, { "epoch": 5.740897258189932, "grad_norm": 1.9308693408966064, "learning_rate": 6.101686730248052e-06, "loss": 0.0533, "step": 169810 }, { "epoch": 5.7412353358801855, "grad_norm": 1.0655975341796875, "learning_rate": 6.097729564282178e-06, "loss": 0.0528, "step": 169820 }, { "epoch": 5.741573413570438, "grad_norm": 1.3007704019546509, "learning_rate": 6.093773598587877e-06, "loss": 0.0579, "step": 169830 }, { "epoch": 5.741911491260692, "grad_norm": 0.7326316833496094, "learning_rate": 6.0898188332733195e-06, "loss": 0.0558, "step": 169840 }, { "epoch": 5.742249568950945, "grad_norm": 2.2689785957336426, "learning_rate": 6.085865268446617e-06, "loss": 0.0531, "step": 169850 }, { "epoch": 5.742587646641198, "grad_norm": 0.5092983245849609, "learning_rate": 6.081912904215864e-06, "loss": 0.054, "step": 169860 }, { "epoch": 5.742925724331451, "grad_norm": 1.8328654766082764, "learning_rate": 6.0779617406891034e-06, "loss": 0.0532, "step": 169870 }, { "epoch": 5.743263802021705, "grad_norm": 1.3256046772003174, "learning_rate": 6.0740117779743755e-06, "loss": 0.0574, "step": 169880 }, { "epoch": 5.7436018797119575, "grad_norm": 1.1666628122329712, "learning_rate": 6.070063016179656e-06, "loss": 0.0562, "step": 169890 }, { "epoch": 5.743939957402211, "grad_norm": 1.226798176765442, "learning_rate": 6.0661154554129126e-06, "loss": 0.0494, "step": 169900 }, { "epoch": 5.7442780350924645, "grad_norm": 1.8240277767181396, "learning_rate": 6.062169095782078e-06, "loss": 0.0583, "step": 169910 }, { "epoch": 5.744616112782717, "grad_norm": 1.0629539489746094, "learning_rate": 6.058223937395019e-06, "loss": 0.0528, "step": 169920 }, { "epoch": 5.744954190472971, "grad_norm": 0.7111765146255493, "learning_rate": 6.054279980359618e-06, "loss": 0.0502, "step": 169930 }, { "epoch": 5.745292268163224, "grad_norm": 0.6171987652778625, "learning_rate": 6.050337224783692e-06, "loss": 0.0504, "step": 169940 }, { "epoch": 5.745630345853477, "grad_norm": 1.3789403438568115, "learning_rate": 6.046395670775034e-06, "loss": 0.0616, "step": 169950 }, { "epoch": 5.74596842354373, "grad_norm": 1.564448595046997, "learning_rate": 6.042455318441403e-06, "loss": 0.0571, "step": 169960 }, { "epoch": 5.746306501233984, "grad_norm": 1.0549383163452148, "learning_rate": 6.038516167890524e-06, "loss": 0.0533, "step": 169970 }, { "epoch": 5.746644578924236, "grad_norm": 0.7346466183662415, "learning_rate": 6.034578219230119e-06, "loss": 0.0489, "step": 169980 }, { "epoch": 5.74698265661449, "grad_norm": 1.0926798582077026, "learning_rate": 6.030641472567805e-06, "loss": 0.0536, "step": 169990 }, { "epoch": 5.747320734304743, "grad_norm": 1.5841974020004272, "learning_rate": 6.02670592801125e-06, "loss": 0.0625, "step": 170000 }, { "epoch": 5.747658811994996, "grad_norm": 0.5867480039596558, "learning_rate": 6.022771585668024e-06, "loss": 0.0524, "step": 170010 }, { "epoch": 5.74799688968525, "grad_norm": 0.7233244180679321, "learning_rate": 6.018838445645708e-06, "loss": 0.0523, "step": 170020 }, { "epoch": 5.748334967375503, "grad_norm": 1.4184736013412476, "learning_rate": 6.014906508051827e-06, "loss": 0.0543, "step": 170030 }, { "epoch": 5.748673045065756, "grad_norm": 0.8845331072807312, "learning_rate": 6.010975772993871e-06, "loss": 0.0543, "step": 170040 }, { "epoch": 5.749011122756009, "grad_norm": 4.051862716674805, "learning_rate": 6.0070462405793175e-06, "loss": 0.0556, "step": 170050 }, { "epoch": 5.749349200446263, "grad_norm": 0.6956710815429688, "learning_rate": 6.003117910915595e-06, "loss": 0.05, "step": 170060 }, { "epoch": 5.749687278136516, "grad_norm": 1.1074557304382324, "learning_rate": 5.999190784110098e-06, "loss": 0.0484, "step": 170070 }, { "epoch": 5.750025355826769, "grad_norm": 0.7608702778816223, "learning_rate": 5.995264860270183e-06, "loss": 0.0494, "step": 170080 }, { "epoch": 5.750363433517022, "grad_norm": 4.7027788162231445, "learning_rate": 5.991340139503199e-06, "loss": 0.0494, "step": 170090 }, { "epoch": 5.750701511207275, "grad_norm": 2.162485122680664, "learning_rate": 5.987416621916453e-06, "loss": 0.0462, "step": 170100 }, { "epoch": 5.7510395888975285, "grad_norm": 6.8195953369140625, "learning_rate": 5.983494307617199e-06, "loss": 0.0555, "step": 170110 }, { "epoch": 5.751377666587782, "grad_norm": 1.9078068733215332, "learning_rate": 5.979573196712673e-06, "loss": 0.0545, "step": 170120 }, { "epoch": 5.751715744278036, "grad_norm": 2.0547897815704346, "learning_rate": 5.975653289310068e-06, "loss": 0.0524, "step": 170130 }, { "epoch": 5.752053821968288, "grad_norm": 0.6125230193138123, "learning_rate": 5.971734585516575e-06, "loss": 0.0453, "step": 170140 }, { "epoch": 5.752391899658542, "grad_norm": 0.8059685826301575, "learning_rate": 5.96781708543932e-06, "loss": 0.0511, "step": 170150 }, { "epoch": 5.752729977348794, "grad_norm": 1.2841237783432007, "learning_rate": 5.963900789185389e-06, "loss": 0.0579, "step": 170160 }, { "epoch": 5.753068055039048, "grad_norm": 1.759848713874817, "learning_rate": 5.959985696861881e-06, "loss": 0.0538, "step": 170170 }, { "epoch": 5.753406132729301, "grad_norm": 18.228647232055664, "learning_rate": 5.956071808575809e-06, "loss": 0.0513, "step": 170180 }, { "epoch": 5.753744210419555, "grad_norm": 2.4785070419311523, "learning_rate": 5.952159124434203e-06, "loss": 0.0536, "step": 170190 }, { "epoch": 5.7540822881098075, "grad_norm": 1.148848533630371, "learning_rate": 5.948247644544002e-06, "loss": 0.0572, "step": 170200 }, { "epoch": 5.754420365800061, "grad_norm": 1.3902533054351807, "learning_rate": 5.9443373690121595e-06, "loss": 0.056, "step": 170210 }, { "epoch": 5.754758443490314, "grad_norm": 3.418376922607422, "learning_rate": 5.940428297945594e-06, "loss": 0.052, "step": 170220 }, { "epoch": 5.755096521180567, "grad_norm": 0.776760458946228, "learning_rate": 5.936520431451164e-06, "loss": 0.0582, "step": 170230 }, { "epoch": 5.755434598870821, "grad_norm": 1.0060791969299316, "learning_rate": 5.932613769635714e-06, "loss": 0.0522, "step": 170240 }, { "epoch": 5.755772676561074, "grad_norm": 2.6210744380950928, "learning_rate": 5.928708312606035e-06, "loss": 0.054, "step": 170250 }, { "epoch": 5.756110754251327, "grad_norm": 1.425484299659729, "learning_rate": 5.9248040604689246e-06, "loss": 0.0556, "step": 170260 }, { "epoch": 5.75644883194158, "grad_norm": 1.17145836353302, "learning_rate": 5.920901013331115e-06, "loss": 0.0544, "step": 170270 }, { "epoch": 5.756786909631833, "grad_norm": 2.935429811477661, "learning_rate": 5.916999171299303e-06, "loss": 0.0578, "step": 170280 }, { "epoch": 5.757124987322086, "grad_norm": 1.1662967205047607, "learning_rate": 5.913098534480177e-06, "loss": 0.0538, "step": 170290 }, { "epoch": 5.75746306501234, "grad_norm": 3.4681122303009033, "learning_rate": 5.909199102980367e-06, "loss": 0.0528, "step": 170300 }, { "epoch": 5.7578011427025935, "grad_norm": 1.0642327070236206, "learning_rate": 5.905300876906505e-06, "loss": 0.0521, "step": 170310 }, { "epoch": 5.758139220392846, "grad_norm": 0.6168573498725891, "learning_rate": 5.901403856365145e-06, "loss": 0.0515, "step": 170320 }, { "epoch": 5.7584772980831, "grad_norm": 1.65897536277771, "learning_rate": 5.8975080414628405e-06, "loss": 0.0562, "step": 170330 }, { "epoch": 5.758815375773352, "grad_norm": 1.6375750303268433, "learning_rate": 5.893613432306083e-06, "loss": 0.0555, "step": 170340 }, { "epoch": 5.759153453463606, "grad_norm": 0.9830319881439209, "learning_rate": 5.8897200290013775e-06, "loss": 0.0527, "step": 170350 }, { "epoch": 5.759491531153859, "grad_norm": 0.7844643592834473, "learning_rate": 5.885827831655155e-06, "loss": 0.0528, "step": 170360 }, { "epoch": 5.759829608844113, "grad_norm": 1.2364972829818726, "learning_rate": 5.881936840373814e-06, "loss": 0.053, "step": 170370 }, { "epoch": 5.760167686534365, "grad_norm": 0.633744478225708, "learning_rate": 5.878047055263758e-06, "loss": 0.0561, "step": 170380 }, { "epoch": 5.760505764224619, "grad_norm": 1.0407235622406006, "learning_rate": 5.874158476431308e-06, "loss": 0.0587, "step": 170390 }, { "epoch": 5.760843841914872, "grad_norm": 0.7434836626052856, "learning_rate": 5.8702711039828e-06, "loss": 0.0492, "step": 170400 }, { "epoch": 5.761181919605125, "grad_norm": 1.2685226202011108, "learning_rate": 5.866384938024499e-06, "loss": 0.0542, "step": 170410 }, { "epoch": 5.761519997295379, "grad_norm": 1.2980703115463257, "learning_rate": 5.862499978662645e-06, "loss": 0.0534, "step": 170420 }, { "epoch": 5.761858074985632, "grad_norm": 1.3725295066833496, "learning_rate": 5.858616226003466e-06, "loss": 0.054, "step": 170430 }, { "epoch": 5.762196152675885, "grad_norm": 1.3441840410232544, "learning_rate": 5.85473368015314e-06, "loss": 0.0552, "step": 170440 }, { "epoch": 5.762534230366138, "grad_norm": 0.6667277812957764, "learning_rate": 5.85085234121781e-06, "loss": 0.0574, "step": 170450 }, { "epoch": 5.762872308056392, "grad_norm": 0.6578407883644104, "learning_rate": 5.84697220930358e-06, "loss": 0.0493, "step": 170460 }, { "epoch": 5.763210385746644, "grad_norm": 1.3526954650878906, "learning_rate": 5.843093284516543e-06, "loss": 0.0542, "step": 170470 }, { "epoch": 5.763548463436898, "grad_norm": 1.3799400329589844, "learning_rate": 5.839215566962769e-06, "loss": 0.05, "step": 170480 }, { "epoch": 5.763886541127151, "grad_norm": 0.7168686985969543, "learning_rate": 5.835339056748229e-06, "loss": 0.0518, "step": 170490 }, { "epoch": 5.764224618817404, "grad_norm": 0.7710129618644714, "learning_rate": 5.831463753978938e-06, "loss": 0.0553, "step": 170500 }, { "epoch": 5.7645626965076575, "grad_norm": 1.2664986848831177, "learning_rate": 5.827589658760824e-06, "loss": 0.0544, "step": 170510 }, { "epoch": 5.764900774197911, "grad_norm": 3.5757791996002197, "learning_rate": 5.823716771199822e-06, "loss": 0.0454, "step": 170520 }, { "epoch": 5.765238851888164, "grad_norm": 1.2979010343551636, "learning_rate": 5.8198450914018095e-06, "loss": 0.0508, "step": 170530 }, { "epoch": 5.765576929578417, "grad_norm": 1.277574062347412, "learning_rate": 5.815974619472625e-06, "loss": 0.0509, "step": 170540 }, { "epoch": 5.765915007268671, "grad_norm": 1.0458863973617554, "learning_rate": 5.8121053555181045e-06, "loss": 0.0541, "step": 170550 }, { "epoch": 5.766253084958923, "grad_norm": 0.8984849452972412, "learning_rate": 5.808237299644021e-06, "loss": 0.0496, "step": 170560 }, { "epoch": 5.766591162649177, "grad_norm": 1.4173669815063477, "learning_rate": 5.804370451956131e-06, "loss": 0.0525, "step": 170570 }, { "epoch": 5.76692924033943, "grad_norm": 2.706470251083374, "learning_rate": 5.800504812560137e-06, "loss": 0.0586, "step": 170580 }, { "epoch": 5.767267318029683, "grad_norm": 1.6007429361343384, "learning_rate": 5.796640381561747e-06, "loss": 0.0524, "step": 170590 }, { "epoch": 5.7676053957199365, "grad_norm": 0.8535202145576477, "learning_rate": 5.7927771590665926e-06, "loss": 0.0557, "step": 170600 }, { "epoch": 5.76794347341019, "grad_norm": 0.8667658567428589, "learning_rate": 5.788915145180313e-06, "loss": 0.0499, "step": 170610 }, { "epoch": 5.768281551100443, "grad_norm": 1.0098683834075928, "learning_rate": 5.785054340008484e-06, "loss": 0.055, "step": 170620 }, { "epoch": 5.768619628790696, "grad_norm": 1.8659378290176392, "learning_rate": 5.781194743656649e-06, "loss": 0.0501, "step": 170630 }, { "epoch": 5.76895770648095, "grad_norm": 1.0947030782699585, "learning_rate": 5.777336356230345e-06, "loss": 0.059, "step": 170640 }, { "epoch": 5.769295784171202, "grad_norm": 1.2076221704483032, "learning_rate": 5.773479177835056e-06, "loss": 0.0506, "step": 170650 }, { "epoch": 5.769633861861456, "grad_norm": 0.9244847893714905, "learning_rate": 5.769623208576219e-06, "loss": 0.0586, "step": 170660 }, { "epoch": 5.769971939551709, "grad_norm": 2.3168132305145264, "learning_rate": 5.765768448559272e-06, "loss": 0.0505, "step": 170670 }, { "epoch": 5.770310017241962, "grad_norm": 5.531295299530029, "learning_rate": 5.761914897889592e-06, "loss": 0.0585, "step": 170680 }, { "epoch": 5.770648094932215, "grad_norm": 1.07534921169281, "learning_rate": 5.758062556672561e-06, "loss": 0.0514, "step": 170690 }, { "epoch": 5.770986172622469, "grad_norm": 1.4092046022415161, "learning_rate": 5.754211425013456e-06, "loss": 0.0545, "step": 170700 }, { "epoch": 5.771324250312722, "grad_norm": 1.4493135213851929, "learning_rate": 5.750361503017599e-06, "loss": 0.0521, "step": 170710 }, { "epoch": 5.771662328002975, "grad_norm": 1.0213384628295898, "learning_rate": 5.746512790790226e-06, "loss": 0.0502, "step": 170720 }, { "epoch": 5.772000405693229, "grad_norm": 1.110887885093689, "learning_rate": 5.742665288436577e-06, "loss": 0.0498, "step": 170730 }, { "epoch": 5.772338483383481, "grad_norm": 1.0340499877929688, "learning_rate": 5.738818996061834e-06, "loss": 0.0573, "step": 170740 }, { "epoch": 5.772676561073735, "grad_norm": 0.5653747320175171, "learning_rate": 5.734973913771135e-06, "loss": 0.0459, "step": 170750 }, { "epoch": 5.773014638763988, "grad_norm": 2.086620569229126, "learning_rate": 5.7311300416696355e-06, "loss": 0.0484, "step": 170760 }, { "epoch": 5.773352716454241, "grad_norm": 1.8668090105056763, "learning_rate": 5.727287379862412e-06, "loss": 0.0574, "step": 170770 }, { "epoch": 5.773690794144494, "grad_norm": 0.8581480979919434, "learning_rate": 5.723445928454513e-06, "loss": 0.0552, "step": 170780 }, { "epoch": 5.774028871834748, "grad_norm": 0.9103877544403076, "learning_rate": 5.7196056875509604e-06, "loss": 0.0488, "step": 170790 }, { "epoch": 5.7743669495250005, "grad_norm": 0.9292710423469543, "learning_rate": 5.715766657256754e-06, "loss": 0.05, "step": 170800 }, { "epoch": 5.774705027215254, "grad_norm": 1.3865647315979004, "learning_rate": 5.711928837676861e-06, "loss": 0.0474, "step": 170810 }, { "epoch": 5.7750431049055075, "grad_norm": 0.932735025882721, "learning_rate": 5.7080922289161955e-06, "loss": 0.0568, "step": 170820 }, { "epoch": 5.77538118259576, "grad_norm": 0.8395851850509644, "learning_rate": 5.704256831079646e-06, "loss": 0.0515, "step": 170830 }, { "epoch": 5.775719260286014, "grad_norm": 1.4647197723388672, "learning_rate": 5.700422644272063e-06, "loss": 0.0529, "step": 170840 }, { "epoch": 5.776057337976267, "grad_norm": 0.4879355728626251, "learning_rate": 5.696589668598296e-06, "loss": 0.0506, "step": 170850 }, { "epoch": 5.77639541566652, "grad_norm": 0.7621909379959106, "learning_rate": 5.6927579041631205e-06, "loss": 0.0478, "step": 170860 }, { "epoch": 5.776733493356773, "grad_norm": 1.5755561590194702, "learning_rate": 5.6889273510712885e-06, "loss": 0.0541, "step": 170870 }, { "epoch": 5.777071571047027, "grad_norm": 1.7896989583969116, "learning_rate": 5.685098009427547e-06, "loss": 0.0571, "step": 170880 }, { "epoch": 5.7774096487372795, "grad_norm": 1.0177935361862183, "learning_rate": 5.68126987933657e-06, "loss": 0.0483, "step": 170890 }, { "epoch": 5.777747726427533, "grad_norm": 1.0682274103164673, "learning_rate": 5.67744296090304e-06, "loss": 0.0547, "step": 170900 }, { "epoch": 5.7780858041177865, "grad_norm": 1.0331618785858154, "learning_rate": 5.67361725423155e-06, "loss": 0.0537, "step": 170910 }, { "epoch": 5.778423881808039, "grad_norm": 2.676410436630249, "learning_rate": 5.669792759426712e-06, "loss": 0.0538, "step": 170920 }, { "epoch": 5.778761959498293, "grad_norm": 0.7572657465934753, "learning_rate": 5.665969476593097e-06, "loss": 0.054, "step": 170930 }, { "epoch": 5.779100037188546, "grad_norm": 0.6915484666824341, "learning_rate": 5.662147405835222e-06, "loss": 0.0525, "step": 170940 }, { "epoch": 5.779438114878799, "grad_norm": 0.7820326089859009, "learning_rate": 5.6583265472575755e-06, "loss": 0.0558, "step": 170950 }, { "epoch": 5.779776192569052, "grad_norm": 2.664303779602051, "learning_rate": 5.654506900964618e-06, "loss": 0.0484, "step": 170960 }, { "epoch": 5.780114270259306, "grad_norm": 1.0315395593643188, "learning_rate": 5.650688467060789e-06, "loss": 0.051, "step": 170970 }, { "epoch": 5.780452347949558, "grad_norm": 2.219815969467163, "learning_rate": 5.6468712456504754e-06, "loss": 0.0564, "step": 170980 }, { "epoch": 5.780790425639812, "grad_norm": 0.7542764544487, "learning_rate": 5.64305523683803e-06, "loss": 0.0542, "step": 170990 }, { "epoch": 5.7811285033300654, "grad_norm": 0.809927225112915, "learning_rate": 5.6392404407278e-06, "loss": 0.055, "step": 171000 }, { "epoch": 5.781466581020318, "grad_norm": 1.7396310567855835, "learning_rate": 5.635426857424058e-06, "loss": 0.0551, "step": 171010 }, { "epoch": 5.781804658710572, "grad_norm": 1.7954217195510864, "learning_rate": 5.631614487031095e-06, "loss": 0.0501, "step": 171020 }, { "epoch": 5.782142736400825, "grad_norm": 0.8331809639930725, "learning_rate": 5.627803329653119e-06, "loss": 0.0499, "step": 171030 }, { "epoch": 5.782480814091079, "grad_norm": 1.1833114624023438, "learning_rate": 5.623993385394322e-06, "loss": 0.0493, "step": 171040 }, { "epoch": 5.782818891781331, "grad_norm": 0.6662466526031494, "learning_rate": 5.620184654358884e-06, "loss": 0.0527, "step": 171050 }, { "epoch": 5.783156969471585, "grad_norm": 0.9770718216896057, "learning_rate": 5.616377136650925e-06, "loss": 0.0591, "step": 171060 }, { "epoch": 5.783495047161837, "grad_norm": 2.0796518325805664, "learning_rate": 5.612570832374542e-06, "loss": 0.0562, "step": 171070 }, { "epoch": 5.783833124852091, "grad_norm": 1.2694958448410034, "learning_rate": 5.608765741633787e-06, "loss": 0.0528, "step": 171080 }, { "epoch": 5.784171202542344, "grad_norm": 1.2377620935440063, "learning_rate": 5.6049618645327095e-06, "loss": 0.0555, "step": 171090 }, { "epoch": 5.784509280232598, "grad_norm": 0.5995837450027466, "learning_rate": 5.601159201175282e-06, "loss": 0.0476, "step": 171100 }, { "epoch": 5.7848473579228505, "grad_norm": 0.8174798488616943, "learning_rate": 5.597357751665494e-06, "loss": 0.0485, "step": 171110 }, { "epoch": 5.785185435613104, "grad_norm": 1.2754719257354736, "learning_rate": 5.593557516107267e-06, "loss": 0.0491, "step": 171120 }, { "epoch": 5.785523513303357, "grad_norm": 2.6647632122039795, "learning_rate": 5.58975849460448e-06, "loss": 0.0511, "step": 171130 }, { "epoch": 5.78586159099361, "grad_norm": 1.3278775215148926, "learning_rate": 5.585960687261027e-06, "loss": 0.0524, "step": 171140 }, { "epoch": 5.786199668683864, "grad_norm": 0.624832034111023, "learning_rate": 5.58216409418072e-06, "loss": 0.0569, "step": 171150 }, { "epoch": 5.786537746374117, "grad_norm": 2.0595643520355225, "learning_rate": 5.578368715467358e-06, "loss": 0.0565, "step": 171160 }, { "epoch": 5.78687582406437, "grad_norm": 1.1935988664627075, "learning_rate": 5.574574551224698e-06, "loss": 0.0545, "step": 171170 }, { "epoch": 5.787213901754623, "grad_norm": 0.9831138253211975, "learning_rate": 5.570781601556485e-06, "loss": 0.0617, "step": 171180 }, { "epoch": 5.787551979444876, "grad_norm": 1.1384162902832031, "learning_rate": 5.566989866566424e-06, "loss": 0.0583, "step": 171190 }, { "epoch": 5.7878900571351295, "grad_norm": 0.9169137477874756, "learning_rate": 5.5631993463581555e-06, "loss": 0.0519, "step": 171200 }, { "epoch": 5.788228134825383, "grad_norm": 0.6743333339691162, "learning_rate": 5.559410041035329e-06, "loss": 0.0517, "step": 171210 }, { "epoch": 5.7885662125156365, "grad_norm": 1.5572623014450073, "learning_rate": 5.555621950701523e-06, "loss": 0.0499, "step": 171220 }, { "epoch": 5.788904290205889, "grad_norm": 0.9889369606971741, "learning_rate": 5.551835075460326e-06, "loss": 0.0527, "step": 171230 }, { "epoch": 5.789242367896143, "grad_norm": 3.5654220581054688, "learning_rate": 5.548049415415263e-06, "loss": 0.0556, "step": 171240 }, { "epoch": 5.789580445586395, "grad_norm": 0.5770269632339478, "learning_rate": 5.544264970669821e-06, "loss": 0.0481, "step": 171250 }, { "epoch": 5.789918523276649, "grad_norm": 1.1458135843276978, "learning_rate": 5.54048174132748e-06, "loss": 0.0512, "step": 171260 }, { "epoch": 5.790256600966902, "grad_norm": 1.7055203914642334, "learning_rate": 5.536699727491667e-06, "loss": 0.0521, "step": 171270 }, { "epoch": 5.790594678657156, "grad_norm": 0.5588108897209167, "learning_rate": 5.532918929265779e-06, "loss": 0.058, "step": 171280 }, { "epoch": 5.7909327563474084, "grad_norm": 1.6896008253097534, "learning_rate": 5.52913934675317e-06, "loss": 0.0516, "step": 171290 }, { "epoch": 5.791270834037662, "grad_norm": 0.661369264125824, "learning_rate": 5.525360980057187e-06, "loss": 0.0494, "step": 171300 }, { "epoch": 5.791608911727915, "grad_norm": 1.835874080657959, "learning_rate": 5.521583829281135e-06, "loss": 0.0523, "step": 171310 }, { "epoch": 5.791946989418168, "grad_norm": 0.8618054986000061, "learning_rate": 5.517807894528271e-06, "loss": 0.0554, "step": 171320 }, { "epoch": 5.792285067108422, "grad_norm": 1.3072503805160522, "learning_rate": 5.514033175901828e-06, "loss": 0.0569, "step": 171330 }, { "epoch": 5.792623144798675, "grad_norm": 0.7556434273719788, "learning_rate": 5.510259673504997e-06, "loss": 0.0554, "step": 171340 }, { "epoch": 5.792961222488928, "grad_norm": 0.9235063195228577, "learning_rate": 5.506487387440962e-06, "loss": 0.0556, "step": 171350 }, { "epoch": 5.793299300179181, "grad_norm": 0.9276642203330994, "learning_rate": 5.502716317812845e-06, "loss": 0.056, "step": 171360 }, { "epoch": 5.793637377869435, "grad_norm": 1.1763628721237183, "learning_rate": 5.498946464723742e-06, "loss": 0.0564, "step": 171370 }, { "epoch": 5.793975455559687, "grad_norm": 1.5626368522644043, "learning_rate": 5.4951778282767265e-06, "loss": 0.0567, "step": 171380 }, { "epoch": 5.794313533249941, "grad_norm": 2.998506784439087, "learning_rate": 5.491410408574826e-06, "loss": 0.057, "step": 171390 }, { "epoch": 5.794651610940194, "grad_norm": 1.1889785528182983, "learning_rate": 5.48764420572106e-06, "loss": 0.0589, "step": 171400 }, { "epoch": 5.794989688630447, "grad_norm": 0.7938758134841919, "learning_rate": 5.483879219818361e-06, "loss": 0.0474, "step": 171410 }, { "epoch": 5.795327766320701, "grad_norm": 4.286575794219971, "learning_rate": 5.480115450969681e-06, "loss": 0.0537, "step": 171420 }, { "epoch": 5.795665844010954, "grad_norm": 3.310896873474121, "learning_rate": 5.476352899277925e-06, "loss": 0.0585, "step": 171430 }, { "epoch": 5.796003921701207, "grad_norm": 1.6652535200119019, "learning_rate": 5.472591564845958e-06, "loss": 0.0562, "step": 171440 }, { "epoch": 5.79634199939146, "grad_norm": 1.5508018732070923, "learning_rate": 5.468831447776607e-06, "loss": 0.0525, "step": 171450 }, { "epoch": 5.796680077081714, "grad_norm": 2.3696746826171875, "learning_rate": 5.465072548172667e-06, "loss": 0.0557, "step": 171460 }, { "epoch": 5.797018154771966, "grad_norm": 1.8239985704421997, "learning_rate": 5.461314866136919e-06, "loss": 0.0566, "step": 171470 }, { "epoch": 5.79735623246222, "grad_norm": 1.2360674142837524, "learning_rate": 5.4575584017720945e-06, "loss": 0.0539, "step": 171480 }, { "epoch": 5.797694310152473, "grad_norm": 2.6730172634124756, "learning_rate": 5.453803155180875e-06, "loss": 0.0502, "step": 171490 }, { "epoch": 5.798032387842726, "grad_norm": 2.182311773300171, "learning_rate": 5.450049126465956e-06, "loss": 0.0566, "step": 171500 }, { "epoch": 5.7983704655329795, "grad_norm": 0.9855180382728577, "learning_rate": 5.4462963157299486e-06, "loss": 0.0487, "step": 171510 }, { "epoch": 5.798708543223233, "grad_norm": 0.5436849594116211, "learning_rate": 5.442544723075465e-06, "loss": 0.0556, "step": 171520 }, { "epoch": 5.799046620913486, "grad_norm": 2.9878454208374023, "learning_rate": 5.438794348605075e-06, "loss": 0.0547, "step": 171530 }, { "epoch": 5.799384698603739, "grad_norm": 0.7557013034820557, "learning_rate": 5.435045192421307e-06, "loss": 0.0551, "step": 171540 }, { "epoch": 5.799722776293993, "grad_norm": 0.6081267595291138, "learning_rate": 5.431297254626644e-06, "loss": 0.0503, "step": 171550 }, { "epoch": 5.800060853984245, "grad_norm": 2.3420820236206055, "learning_rate": 5.427550535323578e-06, "loss": 0.0595, "step": 171560 }, { "epoch": 5.800398931674499, "grad_norm": 0.5940603613853455, "learning_rate": 5.4238050346145545e-06, "loss": 0.0581, "step": 171570 }, { "epoch": 5.800737009364752, "grad_norm": 1.1496820449829102, "learning_rate": 5.4200607526019295e-06, "loss": 0.0598, "step": 171580 }, { "epoch": 5.801075087055005, "grad_norm": 1.2730385065078735, "learning_rate": 5.416317689388112e-06, "loss": 0.0518, "step": 171590 }, { "epoch": 5.8014131647452585, "grad_norm": 0.9365721344947815, "learning_rate": 5.412575845075401e-06, "loss": 0.059, "step": 171600 }, { "epoch": 5.801751242435512, "grad_norm": 1.0431640148162842, "learning_rate": 5.408835219766134e-06, "loss": 0.0522, "step": 171610 }, { "epoch": 5.802089320125765, "grad_norm": 1.9606080055236816, "learning_rate": 5.405095813562555e-06, "loss": 0.0486, "step": 171620 }, { "epoch": 5.802427397816018, "grad_norm": 0.5146839618682861, "learning_rate": 5.4013576265668996e-06, "loss": 0.0543, "step": 171630 }, { "epoch": 5.802765475506272, "grad_norm": 1.3413991928100586, "learning_rate": 5.397620658881375e-06, "loss": 0.0528, "step": 171640 }, { "epoch": 5.803103553196524, "grad_norm": 1.860143780708313, "learning_rate": 5.393884910608149e-06, "loss": 0.05, "step": 171650 }, { "epoch": 5.803441630886778, "grad_norm": 0.8014330267906189, "learning_rate": 5.390150381849351e-06, "loss": 0.0493, "step": 171660 }, { "epoch": 5.803779708577031, "grad_norm": 1.2136949300765991, "learning_rate": 5.3864170727070725e-06, "loss": 0.0564, "step": 171670 }, { "epoch": 5.804117786267284, "grad_norm": 0.5195823311805725, "learning_rate": 5.382684983283387e-06, "loss": 0.0542, "step": 171680 }, { "epoch": 5.804455863957537, "grad_norm": 0.9335038661956787, "learning_rate": 5.378954113680346e-06, "loss": 0.054, "step": 171690 }, { "epoch": 5.804793941647791, "grad_norm": 8.021363258361816, "learning_rate": 5.3752244639999335e-06, "loss": 0.0548, "step": 171700 }, { "epoch": 5.805132019338044, "grad_norm": 0.5235339403152466, "learning_rate": 5.371496034344126e-06, "loss": 0.0558, "step": 171710 }, { "epoch": 5.805470097028297, "grad_norm": 1.0077519416809082, "learning_rate": 5.367768824814834e-06, "loss": 0.053, "step": 171720 }, { "epoch": 5.805808174718551, "grad_norm": 4.0054931640625, "learning_rate": 5.364042835513988e-06, "loss": 0.0553, "step": 171730 }, { "epoch": 5.806146252408803, "grad_norm": 1.6227456331253052, "learning_rate": 5.3603180665434405e-06, "loss": 0.0505, "step": 171740 }, { "epoch": 5.806484330099057, "grad_norm": 1.451981544494629, "learning_rate": 5.356594518005015e-06, "loss": 0.0539, "step": 171750 }, { "epoch": 5.80682240778931, "grad_norm": 2.1942973136901855, "learning_rate": 5.3528721900005355e-06, "loss": 0.0525, "step": 171760 }, { "epoch": 5.807160485479563, "grad_norm": 0.7544739842414856, "learning_rate": 5.349151082631748e-06, "loss": 0.0547, "step": 171770 }, { "epoch": 5.807498563169816, "grad_norm": 3.1915366649627686, "learning_rate": 5.345431196000411e-06, "loss": 0.0531, "step": 171780 }, { "epoch": 5.80783664086007, "grad_norm": 0.8869893550872803, "learning_rate": 5.341712530208193e-06, "loss": 0.0536, "step": 171790 }, { "epoch": 5.8081747185503225, "grad_norm": 2.560354471206665, "learning_rate": 5.337995085356789e-06, "loss": 0.0503, "step": 171800 }, { "epoch": 5.808512796240576, "grad_norm": 0.976087212562561, "learning_rate": 5.334278861547809e-06, "loss": 0.053, "step": 171810 }, { "epoch": 5.8088508739308296, "grad_norm": 0.866163432598114, "learning_rate": 5.330563858882875e-06, "loss": 0.0576, "step": 171820 }, { "epoch": 5.809188951621082, "grad_norm": 1.6988409757614136, "learning_rate": 5.326850077463541e-06, "loss": 0.0535, "step": 171830 }, { "epoch": 5.809527029311336, "grad_norm": 2.0267202854156494, "learning_rate": 5.3231375173913355e-06, "loss": 0.0515, "step": 171840 }, { "epoch": 5.809865107001589, "grad_norm": 1.2695562839508057, "learning_rate": 5.319426178767772e-06, "loss": 0.0502, "step": 171850 }, { "epoch": 5.810203184691842, "grad_norm": 0.5900347828865051, "learning_rate": 5.315716061694315e-06, "loss": 0.0544, "step": 171860 }, { "epoch": 5.810541262382095, "grad_norm": 0.5699080228805542, "learning_rate": 5.312007166272387e-06, "loss": 0.0573, "step": 171870 }, { "epoch": 5.810879340072349, "grad_norm": 1.3305565118789673, "learning_rate": 5.3082994926033965e-06, "loss": 0.0517, "step": 171880 }, { "epoch": 5.8112174177626015, "grad_norm": 0.5595580339431763, "learning_rate": 5.304593040788702e-06, "loss": 0.0507, "step": 171890 }, { "epoch": 5.811555495452855, "grad_norm": 3.2677342891693115, "learning_rate": 5.300887810929656e-06, "loss": 0.0542, "step": 171900 }, { "epoch": 5.8118935731431085, "grad_norm": 0.9050723910331726, "learning_rate": 5.2971838031275425e-06, "loss": 0.0533, "step": 171910 }, { "epoch": 5.812231650833361, "grad_norm": 0.558906078338623, "learning_rate": 5.293481017483626e-06, "loss": 0.0505, "step": 171920 }, { "epoch": 5.812569728523615, "grad_norm": 1.1106171607971191, "learning_rate": 5.289779454099137e-06, "loss": 0.0564, "step": 171930 }, { "epoch": 5.812907806213868, "grad_norm": 1.319470763206482, "learning_rate": 5.286079113075288e-06, "loss": 0.0537, "step": 171940 }, { "epoch": 5.813245883904121, "grad_norm": 0.6753986477851868, "learning_rate": 5.282379994513237e-06, "loss": 0.0522, "step": 171950 }, { "epoch": 5.813583961594374, "grad_norm": 0.9926959276199341, "learning_rate": 5.278682098514109e-06, "loss": 0.0561, "step": 171960 }, { "epoch": 5.813922039284628, "grad_norm": 1.0205273628234863, "learning_rate": 5.274985425179024e-06, "loss": 0.0527, "step": 171970 }, { "epoch": 5.81426011697488, "grad_norm": 0.8167818784713745, "learning_rate": 5.271289974609023e-06, "loss": 0.0527, "step": 171980 }, { "epoch": 5.814598194665134, "grad_norm": 0.991873562335968, "learning_rate": 5.267595746905169e-06, "loss": 0.0493, "step": 171990 }, { "epoch": 5.8149362723553875, "grad_norm": 0.5546700954437256, "learning_rate": 5.263902742168425e-06, "loss": 0.049, "step": 172000 }, { "epoch": 5.815274350045641, "grad_norm": 4.822112083435059, "learning_rate": 5.260210960499773e-06, "loss": 0.0454, "step": 172010 }, { "epoch": 5.815612427735894, "grad_norm": 1.145258903503418, "learning_rate": 5.256520402000154e-06, "loss": 0.054, "step": 172020 }, { "epoch": 5.815950505426147, "grad_norm": 1.4689269065856934, "learning_rate": 5.252831066770458e-06, "loss": 0.0517, "step": 172030 }, { "epoch": 5.8162885831164, "grad_norm": 1.4062689542770386, "learning_rate": 5.2491429549115455e-06, "loss": 0.0549, "step": 172040 }, { "epoch": 5.816626660806653, "grad_norm": 2.187431812286377, "learning_rate": 5.245456066524251e-06, "loss": 0.049, "step": 172050 }, { "epoch": 5.816964738496907, "grad_norm": 0.7668362855911255, "learning_rate": 5.241770401709367e-06, "loss": 0.0528, "step": 172060 }, { "epoch": 5.81730281618716, "grad_norm": 0.9994479417800903, "learning_rate": 5.2380859605676906e-06, "loss": 0.0515, "step": 172070 }, { "epoch": 5.817640893877413, "grad_norm": 1.6416031122207642, "learning_rate": 5.234402743199906e-06, "loss": 0.052, "step": 172080 }, { "epoch": 5.817978971567666, "grad_norm": 0.7546430826187134, "learning_rate": 5.230720749706741e-06, "loss": 0.0529, "step": 172090 }, { "epoch": 5.818317049257919, "grad_norm": 0.9863873720169067, "learning_rate": 5.227039980188842e-06, "loss": 0.0493, "step": 172100 }, { "epoch": 5.8186551269481726, "grad_norm": 1.177747130393982, "learning_rate": 5.22336043474686e-06, "loss": 0.0522, "step": 172110 }, { "epoch": 5.818993204638426, "grad_norm": 1.6391944885253906, "learning_rate": 5.219682113481383e-06, "loss": 0.0546, "step": 172120 }, { "epoch": 5.81933128232868, "grad_norm": 0.7249484658241272, "learning_rate": 5.216005016492964e-06, "loss": 0.0575, "step": 172130 }, { "epoch": 5.819669360018932, "grad_norm": 1.2223104238510132, "learning_rate": 5.212329143882156e-06, "loss": 0.0534, "step": 172140 }, { "epoch": 5.820007437709186, "grad_norm": 0.6676709651947021, "learning_rate": 5.208654495749437e-06, "loss": 0.0539, "step": 172150 }, { "epoch": 5.820345515399438, "grad_norm": 0.9836439490318298, "learning_rate": 5.204981072195281e-06, "loss": 0.056, "step": 172160 }, { "epoch": 5.820683593089692, "grad_norm": 1.3606594800949097, "learning_rate": 5.201308873320099e-06, "loss": 0.062, "step": 172170 }, { "epoch": 5.821021670779945, "grad_norm": 1.1826467514038086, "learning_rate": 5.197637899224317e-06, "loss": 0.0499, "step": 172180 }, { "epoch": 5.821359748470199, "grad_norm": 1.0530825853347778, "learning_rate": 5.193968150008272e-06, "loss": 0.0509, "step": 172190 }, { "epoch": 5.8216978261604515, "grad_norm": 0.7834356427192688, "learning_rate": 5.190299625772316e-06, "loss": 0.0541, "step": 172200 }, { "epoch": 5.822035903850705, "grad_norm": 1.2263935804367065, "learning_rate": 5.186632326616731e-06, "loss": 0.0521, "step": 172210 }, { "epoch": 5.822373981540958, "grad_norm": 1.280246376991272, "learning_rate": 5.182966252641774e-06, "loss": 0.0579, "step": 172220 }, { "epoch": 5.822712059231211, "grad_norm": 1.432349681854248, "learning_rate": 5.179301403947695e-06, "loss": 0.0517, "step": 172230 }, { "epoch": 5.823050136921465, "grad_norm": 1.369063138961792, "learning_rate": 5.175637780634674e-06, "loss": 0.0508, "step": 172240 }, { "epoch": 5.823388214611718, "grad_norm": 0.8906575441360474, "learning_rate": 5.171975382802868e-06, "loss": 0.0567, "step": 172250 }, { "epoch": 5.823726292301971, "grad_norm": 0.6913078427314758, "learning_rate": 5.168314210552427e-06, "loss": 0.0512, "step": 172260 }, { "epoch": 5.824064369992224, "grad_norm": 1.9087915420532227, "learning_rate": 5.16465426398342e-06, "loss": 0.0503, "step": 172270 }, { "epoch": 5.824402447682477, "grad_norm": 5.552338600158691, "learning_rate": 5.16099554319594e-06, "loss": 0.0559, "step": 172280 }, { "epoch": 5.8247405253727305, "grad_norm": 2.1527810096740723, "learning_rate": 5.157338048289978e-06, "loss": 0.0463, "step": 172290 }, { "epoch": 5.825078603062984, "grad_norm": 1.3827555179595947, "learning_rate": 5.153681779365554e-06, "loss": 0.0506, "step": 172300 }, { "epoch": 5.8254166807532375, "grad_norm": 5.238071918487549, "learning_rate": 5.150026736522617e-06, "loss": 0.0568, "step": 172310 }, { "epoch": 5.82575475844349, "grad_norm": 1.010898232460022, "learning_rate": 5.146372919861109e-06, "loss": 0.0595, "step": 172320 }, { "epoch": 5.826092836133744, "grad_norm": 34.14173889160156, "learning_rate": 5.14272032948091e-06, "loss": 0.051, "step": 172330 }, { "epoch": 5.826430913823997, "grad_norm": 2.8112449645996094, "learning_rate": 5.13906896548188e-06, "loss": 0.0497, "step": 172340 }, { "epoch": 5.82676899151425, "grad_norm": 3.424182176589966, "learning_rate": 5.1354188279638614e-06, "loss": 0.0483, "step": 172350 }, { "epoch": 5.827107069204503, "grad_norm": 1.7552183866500854, "learning_rate": 5.1317699170266356e-06, "loss": 0.0487, "step": 172360 }, { "epoch": 5.827445146894757, "grad_norm": 1.031107783317566, "learning_rate": 5.128122232769966e-06, "loss": 0.0483, "step": 172370 }, { "epoch": 5.827783224585009, "grad_norm": 1.011046051979065, "learning_rate": 5.1244757752935635e-06, "loss": 0.055, "step": 172380 }, { "epoch": 5.828121302275263, "grad_norm": 2.3918023109436035, "learning_rate": 5.120830544697136e-06, "loss": 0.0501, "step": 172390 }, { "epoch": 5.828459379965516, "grad_norm": 1.0847383737564087, "learning_rate": 5.117186541080349e-06, "loss": 0.0505, "step": 172400 }, { "epoch": 5.828797457655769, "grad_norm": 1.268578290939331, "learning_rate": 5.113543764542822e-06, "loss": 0.0519, "step": 172410 }, { "epoch": 5.829135535346023, "grad_norm": 1.036278486251831, "learning_rate": 5.1099022151841484e-06, "loss": 0.0511, "step": 172420 }, { "epoch": 5.829473613036276, "grad_norm": 1.150181770324707, "learning_rate": 5.106261893103869e-06, "loss": 0.0524, "step": 172430 }, { "epoch": 5.829811690726529, "grad_norm": 1.3674414157867432, "learning_rate": 5.102622798401535e-06, "loss": 0.0536, "step": 172440 }, { "epoch": 5.830149768416782, "grad_norm": 0.8163677453994751, "learning_rate": 5.098984931176626e-06, "loss": 0.0511, "step": 172450 }, { "epoch": 5.830487846107036, "grad_norm": 1.1685162782669067, "learning_rate": 5.095348291528595e-06, "loss": 0.0539, "step": 172460 }, { "epoch": 5.830825923797288, "grad_norm": 1.6833292245864868, "learning_rate": 5.091712879556876e-06, "loss": 0.0512, "step": 172470 }, { "epoch": 5.831164001487542, "grad_norm": 0.9514369964599609, "learning_rate": 5.088078695360849e-06, "loss": 0.0565, "step": 172480 }, { "epoch": 5.831502079177795, "grad_norm": 0.8807774186134338, "learning_rate": 5.084445739039894e-06, "loss": 0.0543, "step": 172490 }, { "epoch": 5.831840156868048, "grad_norm": 2.128786087036133, "learning_rate": 5.080814010693302e-06, "loss": 0.0552, "step": 172500 }, { "epoch": 5.8321782345583015, "grad_norm": 1.9301115274429321, "learning_rate": 5.07718351042038e-06, "loss": 0.0551, "step": 172510 }, { "epoch": 5.832516312248555, "grad_norm": 1.3134709596633911, "learning_rate": 5.073554238320394e-06, "loss": 0.0574, "step": 172520 }, { "epoch": 5.832854389938808, "grad_norm": 1.903609275817871, "learning_rate": 5.0699261944925584e-06, "loss": 0.0501, "step": 172530 }, { "epoch": 5.833192467629061, "grad_norm": 0.83930903673172, "learning_rate": 5.0662993790360544e-06, "loss": 0.0515, "step": 172540 }, { "epoch": 5.833530545319315, "grad_norm": 1.0089761018753052, "learning_rate": 5.062673792050043e-06, "loss": 0.0507, "step": 172550 }, { "epoch": 5.833868623009567, "grad_norm": 1.1995365619659424, "learning_rate": 5.05904943363365e-06, "loss": 0.0502, "step": 172560 }, { "epoch": 5.834206700699821, "grad_norm": 1.1057645082473755, "learning_rate": 5.055426303885969e-06, "loss": 0.0506, "step": 172570 }, { "epoch": 5.834544778390074, "grad_norm": 0.7580112218856812, "learning_rate": 5.051804402906035e-06, "loss": 0.0478, "step": 172580 }, { "epoch": 5.834882856080327, "grad_norm": 0.716708242893219, "learning_rate": 5.048183730792893e-06, "loss": 0.0505, "step": 172590 }, { "epoch": 5.8352209337705805, "grad_norm": 4.152706623077393, "learning_rate": 5.044564287645509e-06, "loss": 0.0531, "step": 172600 }, { "epoch": 5.835559011460834, "grad_norm": 0.5904433131217957, "learning_rate": 5.040946073562858e-06, "loss": 0.0504, "step": 172610 }, { "epoch": 5.835897089151087, "grad_norm": 1.4208793640136719, "learning_rate": 5.037329088643855e-06, "loss": 0.0572, "step": 172620 }, { "epoch": 5.83623516684134, "grad_norm": 1.1010292768478394, "learning_rate": 5.033713332987366e-06, "loss": 0.0576, "step": 172630 }, { "epoch": 5.836573244531594, "grad_norm": 0.6635333299636841, "learning_rate": 5.03009880669228e-06, "loss": 0.0513, "step": 172640 }, { "epoch": 5.836911322221846, "grad_norm": 1.3084361553192139, "learning_rate": 5.026485509857393e-06, "loss": 0.0537, "step": 172650 }, { "epoch": 5.8372493999121, "grad_norm": 1.661271572113037, "learning_rate": 5.022873442581494e-06, "loss": 0.056, "step": 172660 }, { "epoch": 5.837587477602353, "grad_norm": 1.7780208587646484, "learning_rate": 5.019262604963332e-06, "loss": 0.0589, "step": 172670 }, { "epoch": 5.837925555292606, "grad_norm": 1.0643056631088257, "learning_rate": 5.015652997101639e-06, "loss": 0.0484, "step": 172680 }, { "epoch": 5.838263632982859, "grad_norm": 0.7257704138755798, "learning_rate": 5.012044619095085e-06, "loss": 0.0585, "step": 172690 }, { "epoch": 5.838601710673113, "grad_norm": 0.9672261476516724, "learning_rate": 5.0084374710423435e-06, "loss": 0.0524, "step": 172700 }, { "epoch": 5.838939788363366, "grad_norm": 0.6773024797439575, "learning_rate": 5.0048315530420166e-06, "loss": 0.0522, "step": 172710 }, { "epoch": 5.839277866053619, "grad_norm": 0.5626498460769653, "learning_rate": 5.001226865192676e-06, "loss": 0.0548, "step": 172720 }, { "epoch": 5.839615943743873, "grad_norm": 1.7309192419052124, "learning_rate": 4.997623407592905e-06, "loss": 0.0546, "step": 172730 }, { "epoch": 5.839954021434125, "grad_norm": 0.7441580891609192, "learning_rate": 4.994021180341196e-06, "loss": 0.0523, "step": 172740 }, { "epoch": 5.840292099124379, "grad_norm": 0.9412990808486938, "learning_rate": 4.990420183536049e-06, "loss": 0.0567, "step": 172750 }, { "epoch": 5.840630176814632, "grad_norm": 0.917733371257782, "learning_rate": 4.986820417275889e-06, "loss": 0.052, "step": 172760 }, { "epoch": 5.840968254504885, "grad_norm": 0.8286101818084717, "learning_rate": 4.98322188165915e-06, "loss": 0.0547, "step": 172770 }, { "epoch": 5.841306332195138, "grad_norm": 1.1184937953948975, "learning_rate": 4.9796245767842305e-06, "loss": 0.0508, "step": 172780 }, { "epoch": 5.841644409885392, "grad_norm": 1.0608136653900146, "learning_rate": 4.976028502749441e-06, "loss": 0.0513, "step": 172790 }, { "epoch": 5.8419824875756445, "grad_norm": 0.8248029947280884, "learning_rate": 4.9724336596531304e-06, "loss": 0.0546, "step": 172800 }, { "epoch": 5.842320565265898, "grad_norm": 1.7384649515151978, "learning_rate": 4.968840047593559e-06, "loss": 0.0566, "step": 172810 }, { "epoch": 5.842658642956152, "grad_norm": 0.6763592958450317, "learning_rate": 4.965247666668988e-06, "loss": 0.056, "step": 172820 }, { "epoch": 5.842996720646404, "grad_norm": 0.9137069582939148, "learning_rate": 4.961656516977631e-06, "loss": 0.0515, "step": 172830 }, { "epoch": 5.843334798336658, "grad_norm": 1.1464695930480957, "learning_rate": 4.958066598617656e-06, "loss": 0.0485, "step": 172840 }, { "epoch": 5.843672876026911, "grad_norm": 0.743202805519104, "learning_rate": 4.954477911687222e-06, "loss": 0.0477, "step": 172850 }, { "epoch": 5.844010953717164, "grad_norm": 1.1331042051315308, "learning_rate": 4.950890456284446e-06, "loss": 0.0503, "step": 172860 }, { "epoch": 5.844349031407417, "grad_norm": 0.4159351587295532, "learning_rate": 4.947304232507399e-06, "loss": 0.0565, "step": 172870 }, { "epoch": 5.844687109097671, "grad_norm": 1.0418541431427002, "learning_rate": 4.943719240454115e-06, "loss": 0.0521, "step": 172880 }, { "epoch": 5.8450251867879235, "grad_norm": 2.1762943267822266, "learning_rate": 4.940135480222624e-06, "loss": 0.0522, "step": 172890 }, { "epoch": 5.845363264478177, "grad_norm": 2.344165325164795, "learning_rate": 4.936552951910911e-06, "loss": 0.0522, "step": 172900 }, { "epoch": 5.8457013421684305, "grad_norm": 0.9306358695030212, "learning_rate": 4.932971655616908e-06, "loss": 0.0513, "step": 172910 }, { "epoch": 5.846039419858683, "grad_norm": 0.9430601000785828, "learning_rate": 4.929391591438531e-06, "loss": 0.0552, "step": 172920 }, { "epoch": 5.846377497548937, "grad_norm": 1.6436896324157715, "learning_rate": 4.925812759473647e-06, "loss": 0.0526, "step": 172930 }, { "epoch": 5.84671557523919, "grad_norm": 0.7509148120880127, "learning_rate": 4.922235159820115e-06, "loss": 0.0558, "step": 172940 }, { "epoch": 5.847053652929443, "grad_norm": 11.545515060424805, "learning_rate": 4.918658792575742e-06, "loss": 0.0472, "step": 172950 }, { "epoch": 5.847391730619696, "grad_norm": 1.6196224689483643, "learning_rate": 4.915083657838287e-06, "loss": 0.051, "step": 172960 }, { "epoch": 5.84772980830995, "grad_norm": 0.888683557510376, "learning_rate": 4.911509755705524e-06, "loss": 0.059, "step": 172970 }, { "epoch": 5.848067886000203, "grad_norm": 2.2715516090393066, "learning_rate": 4.9079370862751325e-06, "loss": 0.0534, "step": 172980 }, { "epoch": 5.848405963690456, "grad_norm": 0.7973757386207581, "learning_rate": 4.904365649644815e-06, "loss": 0.0557, "step": 172990 }, { "epoch": 5.8487440413807095, "grad_norm": 0.7891026735305786, "learning_rate": 4.900795445912188e-06, "loss": 0.0579, "step": 173000 }, { "epoch": 5.849082119070962, "grad_norm": 4.492432594299316, "learning_rate": 4.897226475174877e-06, "loss": 0.0544, "step": 173010 }, { "epoch": 5.849420196761216, "grad_norm": 0.5704372525215149, "learning_rate": 4.893658737530437e-06, "loss": 0.0556, "step": 173020 }, { "epoch": 5.849758274451469, "grad_norm": 1.577797293663025, "learning_rate": 4.890092233076432e-06, "loss": 0.0529, "step": 173030 }, { "epoch": 5.850096352141723, "grad_norm": 1.6274813413619995, "learning_rate": 4.886526961910359e-06, "loss": 0.0547, "step": 173040 }, { "epoch": 5.850434429831975, "grad_norm": 0.9246727228164673, "learning_rate": 4.882962924129686e-06, "loss": 0.0488, "step": 173050 }, { "epoch": 5.850772507522229, "grad_norm": 18.206321716308594, "learning_rate": 4.879400119831862e-06, "loss": 0.0561, "step": 173060 }, { "epoch": 5.851110585212481, "grad_norm": 1.5012723207473755, "learning_rate": 4.875838549114287e-06, "loss": 0.0494, "step": 173070 }, { "epoch": 5.851448662902735, "grad_norm": 0.8905030488967896, "learning_rate": 4.872278212074321e-06, "loss": 0.0491, "step": 173080 }, { "epoch": 5.851786740592988, "grad_norm": 0.5761411190032959, "learning_rate": 4.868719108809333e-06, "loss": 0.0558, "step": 173090 }, { "epoch": 5.852124818283242, "grad_norm": 0.7287878394126892, "learning_rate": 4.865161239416594e-06, "loss": 0.0518, "step": 173100 }, { "epoch": 5.852462895973495, "grad_norm": 0.9706095457077026, "learning_rate": 4.861604603993403e-06, "loss": 0.0522, "step": 173110 }, { "epoch": 5.852800973663748, "grad_norm": 1.1290336847305298, "learning_rate": 4.858049202636983e-06, "loss": 0.0565, "step": 173120 }, { "epoch": 5.853139051354001, "grad_norm": 0.5292786359786987, "learning_rate": 4.8544950354445396e-06, "loss": 0.0514, "step": 173130 }, { "epoch": 5.853477129044254, "grad_norm": 1.537233591079712, "learning_rate": 4.850942102513228e-06, "loss": 0.0569, "step": 173140 }, { "epoch": 5.853815206734508, "grad_norm": 0.9116700887680054, "learning_rate": 4.84739040394021e-06, "loss": 0.0561, "step": 173150 }, { "epoch": 5.854153284424761, "grad_norm": 0.7589449882507324, "learning_rate": 4.843839939822575e-06, "loss": 0.0497, "step": 173160 }, { "epoch": 5.854491362115014, "grad_norm": 1.1555339097976685, "learning_rate": 4.840290710257383e-06, "loss": 0.0561, "step": 173170 }, { "epoch": 5.854829439805267, "grad_norm": 0.7501056790351868, "learning_rate": 4.836742715341685e-06, "loss": 0.0533, "step": 173180 }, { "epoch": 5.85516751749552, "grad_norm": 1.3035773038864136, "learning_rate": 4.833195955172471e-06, "loss": 0.0535, "step": 173190 }, { "epoch": 5.8555055951857735, "grad_norm": 0.6774801015853882, "learning_rate": 4.829650429846721e-06, "loss": 0.0549, "step": 173200 }, { "epoch": 5.855843672876027, "grad_norm": 0.6085599064826965, "learning_rate": 4.826106139461345e-06, "loss": 0.0542, "step": 173210 }, { "epoch": 5.8561817505662805, "grad_norm": 5.512900352478027, "learning_rate": 4.822563084113257e-06, "loss": 0.0532, "step": 173220 }, { "epoch": 5.856519828256533, "grad_norm": 0.9373773336410522, "learning_rate": 4.819021263899332e-06, "loss": 0.0556, "step": 173230 }, { "epoch": 5.856857905946787, "grad_norm": 1.8708934783935547, "learning_rate": 4.815480678916384e-06, "loss": 0.0553, "step": 173240 }, { "epoch": 5.857195983637039, "grad_norm": 1.079128623008728, "learning_rate": 4.811941329261227e-06, "loss": 0.0527, "step": 173250 }, { "epoch": 5.857534061327293, "grad_norm": 0.9096693396568298, "learning_rate": 4.808403215030605e-06, "loss": 0.0502, "step": 173260 }, { "epoch": 5.857872139017546, "grad_norm": 0.923366904258728, "learning_rate": 4.804866336321262e-06, "loss": 0.0531, "step": 173270 }, { "epoch": 5.8582102167078, "grad_norm": 1.1234910488128662, "learning_rate": 4.8013306932299164e-06, "loss": 0.0564, "step": 173280 }, { "epoch": 5.8585482943980525, "grad_norm": 1.255722165107727, "learning_rate": 4.797796285853184e-06, "loss": 0.0591, "step": 173290 }, { "epoch": 5.858886372088306, "grad_norm": 1.2598636150360107, "learning_rate": 4.7942631142877335e-06, "loss": 0.053, "step": 173300 }, { "epoch": 5.8592244497785595, "grad_norm": 39.933834075927734, "learning_rate": 4.7907311786301366e-06, "loss": 0.0557, "step": 173310 }, { "epoch": 5.859562527468812, "grad_norm": 1.0291069746017456, "learning_rate": 4.787200478976972e-06, "loss": 0.0477, "step": 173320 }, { "epoch": 5.859900605159066, "grad_norm": 1.275896430015564, "learning_rate": 4.7836710154247625e-06, "loss": 0.0531, "step": 173330 }, { "epoch": 5.860238682849319, "grad_norm": 0.5958139300346375, "learning_rate": 4.780142788069991e-06, "loss": 0.0471, "step": 173340 }, { "epoch": 5.860576760539572, "grad_norm": 4.318897247314453, "learning_rate": 4.7766157970091375e-06, "loss": 0.0465, "step": 173350 }, { "epoch": 5.860914838229825, "grad_norm": 1.770158290863037, "learning_rate": 4.7730900423386185e-06, "loss": 0.0582, "step": 173360 }, { "epoch": 5.861252915920079, "grad_norm": 1.3091249465942383, "learning_rate": 4.769565524154823e-06, "loss": 0.0525, "step": 173370 }, { "epoch": 5.861590993610331, "grad_norm": 4.015211582183838, "learning_rate": 4.766042242554108e-06, "loss": 0.0523, "step": 173380 }, { "epoch": 5.861929071300585, "grad_norm": 1.2050423622131348, "learning_rate": 4.762520197632814e-06, "loss": 0.0544, "step": 173390 }, { "epoch": 5.8622671489908385, "grad_norm": 0.9379584193229675, "learning_rate": 4.758999389487212e-06, "loss": 0.0555, "step": 173400 }, { "epoch": 5.862605226681091, "grad_norm": 1.1847220659255981, "learning_rate": 4.755479818213582e-06, "loss": 0.0449, "step": 173410 }, { "epoch": 5.862943304371345, "grad_norm": 0.5750625133514404, "learning_rate": 4.751961483908135e-06, "loss": 0.0548, "step": 173420 }, { "epoch": 5.863281382061598, "grad_norm": 0.9491074085235596, "learning_rate": 4.74844438666705e-06, "loss": 0.0547, "step": 173430 }, { "epoch": 5.863619459751851, "grad_norm": 3.958883047103882, "learning_rate": 4.744928526586506e-06, "loss": 0.0524, "step": 173440 }, { "epoch": 5.863957537442104, "grad_norm": 0.5022224187850952, "learning_rate": 4.7414139037626144e-06, "loss": 0.056, "step": 173450 }, { "epoch": 5.864295615132358, "grad_norm": 1.8470580577850342, "learning_rate": 4.737900518291455e-06, "loss": 0.0484, "step": 173460 }, { "epoch": 5.86463369282261, "grad_norm": 1.4226926565170288, "learning_rate": 4.734388370269105e-06, "loss": 0.0462, "step": 173470 }, { "epoch": 5.864971770512864, "grad_norm": 1.5989030599594116, "learning_rate": 4.730877459791555e-06, "loss": 0.0544, "step": 173480 }, { "epoch": 5.865309848203117, "grad_norm": 1.183243989944458, "learning_rate": 4.727367786954834e-06, "loss": 0.0573, "step": 173490 }, { "epoch": 5.86564792589337, "grad_norm": 0.9672117233276367, "learning_rate": 4.723859351854848e-06, "loss": 0.0565, "step": 173500 }, { "epoch": 5.8659860035836235, "grad_norm": 1.0519733428955078, "learning_rate": 4.720352154587543e-06, "loss": 0.0534, "step": 173510 }, { "epoch": 5.866324081273877, "grad_norm": 0.7249627113342285, "learning_rate": 4.716846195248792e-06, "loss": 0.0521, "step": 173520 }, { "epoch": 5.86666215896413, "grad_norm": 1.146183729171753, "learning_rate": 4.713341473934468e-06, "loss": 0.0499, "step": 173530 }, { "epoch": 5.867000236654383, "grad_norm": 0.7114830613136292, "learning_rate": 4.709837990740368e-06, "loss": 0.0515, "step": 173540 }, { "epoch": 5.867338314344637, "grad_norm": 4.343298435211182, "learning_rate": 4.70633574576228e-06, "loss": 0.0503, "step": 173550 }, { "epoch": 5.867676392034889, "grad_norm": 0.7905638217926025, "learning_rate": 4.702834739095963e-06, "loss": 0.0583, "step": 173560 }, { "epoch": 5.868014469725143, "grad_norm": 1.074663519859314, "learning_rate": 4.699334970837127e-06, "loss": 0.0532, "step": 173570 }, { "epoch": 5.868352547415396, "grad_norm": 0.9022141695022583, "learning_rate": 4.695836441081453e-06, "loss": 0.0576, "step": 173580 }, { "epoch": 5.868690625105649, "grad_norm": 0.8150424957275391, "learning_rate": 4.692339149924579e-06, "loss": 0.054, "step": 173590 }, { "epoch": 5.8690287027959025, "grad_norm": 0.6735371947288513, "learning_rate": 4.688843097462137e-06, "loss": 0.0538, "step": 173600 }, { "epoch": 5.869366780486156, "grad_norm": 1.7607389688491821, "learning_rate": 4.685348283789709e-06, "loss": 0.0533, "step": 173610 }, { "epoch": 5.869704858176409, "grad_norm": 0.7715208530426025, "learning_rate": 4.681854709002837e-06, "loss": 0.0577, "step": 173620 }, { "epoch": 5.870042935866662, "grad_norm": 1.125281810760498, "learning_rate": 4.678362373197026e-06, "loss": 0.053, "step": 173630 }, { "epoch": 5.870381013556916, "grad_norm": 1.5849989652633667, "learning_rate": 4.674871276467757e-06, "loss": 0.0486, "step": 173640 }, { "epoch": 5.870719091247168, "grad_norm": 0.733907163143158, "learning_rate": 4.671381418910487e-06, "loss": 0.0499, "step": 173650 }, { "epoch": 5.871057168937422, "grad_norm": 1.1649181842803955, "learning_rate": 4.667892800620621e-06, "loss": 0.0552, "step": 173660 }, { "epoch": 5.871395246627675, "grad_norm": 0.5608481764793396, "learning_rate": 4.664405421693524e-06, "loss": 0.0477, "step": 173670 }, { "epoch": 5.871733324317928, "grad_norm": 1.9432834386825562, "learning_rate": 4.660919282224563e-06, "loss": 0.0548, "step": 173680 }, { "epoch": 5.8720714020081815, "grad_norm": 0.8762124180793762, "learning_rate": 4.657434382309023e-06, "loss": 0.0504, "step": 173690 }, { "epoch": 5.872409479698435, "grad_norm": 1.0230321884155273, "learning_rate": 4.6539507220422166e-06, "loss": 0.0526, "step": 173700 }, { "epoch": 5.872747557388688, "grad_norm": 1.0441293716430664, "learning_rate": 4.6504683015193395e-06, "loss": 0.05, "step": 173710 }, { "epoch": 5.873085635078941, "grad_norm": 0.9081701636314392, "learning_rate": 4.646987120835622e-06, "loss": 0.0511, "step": 173720 }, { "epoch": 5.873423712769195, "grad_norm": 1.791756510734558, "learning_rate": 4.643507180086248e-06, "loss": 0.0515, "step": 173730 }, { "epoch": 5.873761790459447, "grad_norm": 0.8809263110160828, "learning_rate": 4.640028479366349e-06, "loss": 0.0525, "step": 173740 }, { "epoch": 5.874099868149701, "grad_norm": 0.901773989200592, "learning_rate": 4.636551018771029e-06, "loss": 0.058, "step": 173750 }, { "epoch": 5.874437945839954, "grad_norm": 1.4879416227340698, "learning_rate": 4.6330747983953535e-06, "loss": 0.0499, "step": 173760 }, { "epoch": 5.874776023530207, "grad_norm": 0.670550525188446, "learning_rate": 4.6295998183343795e-06, "loss": 0.0498, "step": 173770 }, { "epoch": 5.87511410122046, "grad_norm": 0.848380982875824, "learning_rate": 4.626126078683096e-06, "loss": 0.0511, "step": 173780 }, { "epoch": 5.875452178910714, "grad_norm": 7.91378927230835, "learning_rate": 4.6226535795364786e-06, "loss": 0.0576, "step": 173790 }, { "epoch": 5.8757902566009665, "grad_norm": 0.743212878704071, "learning_rate": 4.619182320989468e-06, "loss": 0.0537, "step": 173800 }, { "epoch": 5.87612833429122, "grad_norm": 1.0640631914138794, "learning_rate": 4.615712303136954e-06, "loss": 0.0519, "step": 173810 }, { "epoch": 5.876466411981474, "grad_norm": 1.2530752420425415, "learning_rate": 4.6122435260738284e-06, "loss": 0.056, "step": 173820 }, { "epoch": 5.876804489671726, "grad_norm": 1.031394362449646, "learning_rate": 4.608775989894915e-06, "loss": 0.0522, "step": 173830 }, { "epoch": 5.87714256736198, "grad_norm": 0.5974265336990356, "learning_rate": 4.605309694694998e-06, "loss": 0.0566, "step": 173840 }, { "epoch": 5.877480645052233, "grad_norm": 1.594619631767273, "learning_rate": 4.601844640568876e-06, "loss": 0.0514, "step": 173850 }, { "epoch": 5.877818722742486, "grad_norm": 1.0511035919189453, "learning_rate": 4.598380827611265e-06, "loss": 0.0504, "step": 173860 }, { "epoch": 5.878156800432739, "grad_norm": 1.0897350311279297, "learning_rate": 4.594918255916864e-06, "loss": 0.0502, "step": 173870 }, { "epoch": 5.878494878122993, "grad_norm": 0.8419625163078308, "learning_rate": 4.591456925580329e-06, "loss": 0.0578, "step": 173880 }, { "epoch": 5.8788329558132455, "grad_norm": 0.8826836943626404, "learning_rate": 4.587996836696312e-06, "loss": 0.0538, "step": 173890 }, { "epoch": 5.879171033503499, "grad_norm": 2.298039674758911, "learning_rate": 4.584537989359394e-06, "loss": 0.0527, "step": 173900 }, { "epoch": 5.8795091111937525, "grad_norm": 1.3134440183639526, "learning_rate": 4.581080383664149e-06, "loss": 0.0551, "step": 173910 }, { "epoch": 5.879847188884005, "grad_norm": 0.6548305153846741, "learning_rate": 4.577624019705113e-06, "loss": 0.0515, "step": 173920 }, { "epoch": 5.880185266574259, "grad_norm": 0.5026528835296631, "learning_rate": 4.574168897576753e-06, "loss": 0.0523, "step": 173930 }, { "epoch": 5.880523344264512, "grad_norm": 1.682919979095459, "learning_rate": 4.570715017373567e-06, "loss": 0.0549, "step": 173940 }, { "epoch": 5.880861421954766, "grad_norm": 0.6373582482337952, "learning_rate": 4.567262379189963e-06, "loss": 0.0487, "step": 173950 }, { "epoch": 5.881199499645018, "grad_norm": 0.7124146223068237, "learning_rate": 4.563810983120331e-06, "loss": 0.0601, "step": 173960 }, { "epoch": 5.881537577335272, "grad_norm": 1.1096186637878418, "learning_rate": 4.560360829259036e-06, "loss": 0.0537, "step": 173970 }, { "epoch": 5.8818756550255245, "grad_norm": 0.8162810802459717, "learning_rate": 4.556911917700402e-06, "loss": 0.0534, "step": 173980 }, { "epoch": 5.882213732715778, "grad_norm": 0.5059170722961426, "learning_rate": 4.553464248538741e-06, "loss": 0.053, "step": 173990 }, { "epoch": 5.8825518104060315, "grad_norm": 1.5092676877975464, "learning_rate": 4.55001782186828e-06, "loss": 0.0519, "step": 174000 }, { "epoch": 5.882889888096285, "grad_norm": 2.333562135696411, "learning_rate": 4.5465726377832585e-06, "loss": 0.0524, "step": 174010 }, { "epoch": 5.883227965786538, "grad_norm": 1.6504673957824707, "learning_rate": 4.543128696377863e-06, "loss": 0.0547, "step": 174020 }, { "epoch": 5.883566043476791, "grad_norm": 2.920976400375366, "learning_rate": 4.539685997746257e-06, "loss": 0.0547, "step": 174030 }, { "epoch": 5.883904121167044, "grad_norm": 1.3658075332641602, "learning_rate": 4.5362445419825595e-06, "loss": 0.0542, "step": 174040 }, { "epoch": 5.884242198857297, "grad_norm": 3.677006244659424, "learning_rate": 4.532804329180845e-06, "loss": 0.0517, "step": 174050 }, { "epoch": 5.884580276547551, "grad_norm": 0.5723469257354736, "learning_rate": 4.529365359435189e-06, "loss": 0.0526, "step": 174060 }, { "epoch": 5.884918354237804, "grad_norm": 1.5644876956939697, "learning_rate": 4.525927632839605e-06, "loss": 0.0511, "step": 174070 }, { "epoch": 5.885256431928057, "grad_norm": 2.52508807182312, "learning_rate": 4.5224911494880725e-06, "loss": 0.0513, "step": 174080 }, { "epoch": 5.88559450961831, "grad_norm": 1.698973298072815, "learning_rate": 4.519055909474535e-06, "loss": 0.0535, "step": 174090 }, { "epoch": 5.885932587308563, "grad_norm": 1.0343081951141357, "learning_rate": 4.5156219128929255e-06, "loss": 0.0554, "step": 174100 }, { "epoch": 5.886270664998817, "grad_norm": 0.8758142590522766, "learning_rate": 4.512189159837133e-06, "loss": 0.0575, "step": 174110 }, { "epoch": 5.88660874268907, "grad_norm": 0.9238566160202026, "learning_rate": 4.508757650401002e-06, "loss": 0.0507, "step": 174120 }, { "epoch": 5.886946820379324, "grad_norm": 2.7235090732574463, "learning_rate": 4.505327384678348e-06, "loss": 0.0547, "step": 174130 }, { "epoch": 5.887284898069576, "grad_norm": 0.6108474135398865, "learning_rate": 4.501898362762935e-06, "loss": 0.0571, "step": 174140 }, { "epoch": 5.88762297575983, "grad_norm": 2.9529800415039062, "learning_rate": 4.498470584748543e-06, "loss": 0.0581, "step": 174150 }, { "epoch": 5.887961053450082, "grad_norm": 4.4112958908081055, "learning_rate": 4.495044050728875e-06, "loss": 0.052, "step": 174160 }, { "epoch": 5.888299131140336, "grad_norm": 7.773270606994629, "learning_rate": 4.491618760797589e-06, "loss": 0.0537, "step": 174170 }, { "epoch": 5.888637208830589, "grad_norm": 1.9195432662963867, "learning_rate": 4.4881947150483665e-06, "loss": 0.0577, "step": 174180 }, { "epoch": 5.888975286520843, "grad_norm": 1.923094630241394, "learning_rate": 4.484771913574787e-06, "loss": 0.0522, "step": 174190 }, { "epoch": 5.8893133642110955, "grad_norm": 2.3555190563201904, "learning_rate": 4.481350356470465e-06, "loss": 0.0509, "step": 174200 }, { "epoch": 5.889651441901349, "grad_norm": 0.9041480422019958, "learning_rate": 4.477930043828904e-06, "loss": 0.0507, "step": 174210 }, { "epoch": 5.889989519591602, "grad_norm": 3.5157501697540283, "learning_rate": 4.47451097574364e-06, "loss": 0.0506, "step": 174220 }, { "epoch": 5.890327597281855, "grad_norm": 2.6209497451782227, "learning_rate": 4.471093152308142e-06, "loss": 0.06, "step": 174230 }, { "epoch": 5.890665674972109, "grad_norm": 1.8385850191116333, "learning_rate": 4.467676573615859e-06, "loss": 0.0509, "step": 174240 }, { "epoch": 5.891003752662362, "grad_norm": 0.7942824959754944, "learning_rate": 4.464261239760192e-06, "loss": 0.0549, "step": 174250 }, { "epoch": 5.891341830352615, "grad_norm": 0.6406533718109131, "learning_rate": 4.460847150834508e-06, "loss": 0.0594, "step": 174260 }, { "epoch": 5.891679908042868, "grad_norm": 0.7484568953514099, "learning_rate": 4.457434306932162e-06, "loss": 0.0581, "step": 174270 }, { "epoch": 5.892017985733122, "grad_norm": 1.0713539123535156, "learning_rate": 4.454022708146454e-06, "loss": 0.0547, "step": 174280 }, { "epoch": 5.8923560634233745, "grad_norm": 0.8648759722709656, "learning_rate": 4.4506123545706435e-06, "loss": 0.0536, "step": 174290 }, { "epoch": 5.892694141113628, "grad_norm": 0.5092030167579651, "learning_rate": 4.4472032462979875e-06, "loss": 0.0529, "step": 174300 }, { "epoch": 5.8930322188038815, "grad_norm": 0.9209978580474854, "learning_rate": 4.443795383421673e-06, "loss": 0.0536, "step": 174310 }, { "epoch": 5.893370296494134, "grad_norm": 0.696305513381958, "learning_rate": 4.440388766034887e-06, "loss": 0.0489, "step": 174320 }, { "epoch": 5.893708374184388, "grad_norm": 1.1336737871170044, "learning_rate": 4.4369833942307595e-06, "loss": 0.0588, "step": 174330 }, { "epoch": 5.894046451874641, "grad_norm": 1.5272992849349976, "learning_rate": 4.433579268102383e-06, "loss": 0.0534, "step": 174340 }, { "epoch": 5.894384529564894, "grad_norm": 2.1996407508850098, "learning_rate": 4.430176387742824e-06, "loss": 0.049, "step": 174350 }, { "epoch": 5.894722607255147, "grad_norm": 1.1322497129440308, "learning_rate": 4.426774753245133e-06, "loss": 0.0503, "step": 174360 }, { "epoch": 5.895060684945401, "grad_norm": 1.1202443838119507, "learning_rate": 4.423374364702298e-06, "loss": 0.0544, "step": 174370 }, { "epoch": 5.895398762635653, "grad_norm": 1.298111081123352, "learning_rate": 4.4199752222072784e-06, "loss": 0.0544, "step": 174380 }, { "epoch": 5.895736840325907, "grad_norm": 1.1004714965820312, "learning_rate": 4.4165773258530215e-06, "loss": 0.0516, "step": 174390 }, { "epoch": 5.8960749180161605, "grad_norm": 1.7788281440734863, "learning_rate": 4.413180675732404e-06, "loss": 0.0586, "step": 174400 }, { "epoch": 5.896412995706413, "grad_norm": 5.430677890777588, "learning_rate": 4.409785271938321e-06, "loss": 0.0498, "step": 174410 }, { "epoch": 5.896751073396667, "grad_norm": 1.718908429145813, "learning_rate": 4.406391114563563e-06, "loss": 0.049, "step": 174420 }, { "epoch": 5.89708915108692, "grad_norm": 0.9301692247390747, "learning_rate": 4.402998203700942e-06, "loss": 0.0519, "step": 174430 }, { "epoch": 5.897427228777173, "grad_norm": 1.7243393659591675, "learning_rate": 4.39960653944323e-06, "loss": 0.0496, "step": 174440 }, { "epoch": 5.897765306467426, "grad_norm": 1.3003352880477905, "learning_rate": 4.396216121883145e-06, "loss": 0.0512, "step": 174450 }, { "epoch": 5.89810338415768, "grad_norm": 1.3320423364639282, "learning_rate": 4.392826951113377e-06, "loss": 0.0583, "step": 174460 }, { "epoch": 5.898441461847932, "grad_norm": 1.1493773460388184, "learning_rate": 4.389439027226577e-06, "loss": 0.0528, "step": 174470 }, { "epoch": 5.898779539538186, "grad_norm": 0.5402300953865051, "learning_rate": 4.386052350315379e-06, "loss": 0.0519, "step": 174480 }, { "epoch": 5.899117617228439, "grad_norm": 0.9001967310905457, "learning_rate": 4.382666920472389e-06, "loss": 0.0555, "step": 174490 }, { "epoch": 5.899455694918692, "grad_norm": 0.9473523497581482, "learning_rate": 4.379282737790136e-06, "loss": 0.0524, "step": 174500 }, { "epoch": 5.8997937726089456, "grad_norm": 1.846271276473999, "learning_rate": 4.375899802361161e-06, "loss": 0.0448, "step": 174510 }, { "epoch": 5.900131850299199, "grad_norm": 6.563230514526367, "learning_rate": 4.372518114277935e-06, "loss": 0.0516, "step": 174520 }, { "epoch": 5.900469927989452, "grad_norm": 1.3732668161392212, "learning_rate": 4.369137673632928e-06, "loss": 0.0465, "step": 174530 }, { "epoch": 5.900808005679705, "grad_norm": 1.169911503791809, "learning_rate": 4.365758480518561e-06, "loss": 0.0526, "step": 174540 }, { "epoch": 5.901146083369959, "grad_norm": 1.1027004718780518, "learning_rate": 4.362380535027199e-06, "loss": 0.0498, "step": 174550 }, { "epoch": 5.901484161060211, "grad_norm": 2.8099396228790283, "learning_rate": 4.3590038372512185e-06, "loss": 0.0574, "step": 174560 }, { "epoch": 5.901822238750465, "grad_norm": 0.733039915561676, "learning_rate": 4.3556283872829285e-06, "loss": 0.0505, "step": 174570 }, { "epoch": 5.902160316440718, "grad_norm": 0.7317398190498352, "learning_rate": 4.352254185214605e-06, "loss": 0.0479, "step": 174580 }, { "epoch": 5.902498394130971, "grad_norm": 0.7128865122795105, "learning_rate": 4.348881231138502e-06, "loss": 0.0514, "step": 174590 }, { "epoch": 5.9028364718212245, "grad_norm": 0.5173023343086243, "learning_rate": 4.345509525146841e-06, "loss": 0.0545, "step": 174600 }, { "epoch": 5.903174549511478, "grad_norm": 0.8302408456802368, "learning_rate": 4.342139067331785e-06, "loss": 0.0538, "step": 174610 }, { "epoch": 5.903512627201731, "grad_norm": 1.0575497150421143, "learning_rate": 4.338769857785507e-06, "loss": 0.052, "step": 174620 }, { "epoch": 5.903850704891984, "grad_norm": 1.54192316532135, "learning_rate": 4.335401896600105e-06, "loss": 0.0524, "step": 174630 }, { "epoch": 5.904188782582238, "grad_norm": 0.6919353008270264, "learning_rate": 4.332035183867655e-06, "loss": 0.0511, "step": 174640 }, { "epoch": 5.90452686027249, "grad_norm": 0.7774945497512817, "learning_rate": 4.328669719680212e-06, "loss": 0.0589, "step": 174650 }, { "epoch": 5.904864937962744, "grad_norm": 0.8517903089523315, "learning_rate": 4.3253055041297845e-06, "loss": 0.0559, "step": 174660 }, { "epoch": 5.905203015652997, "grad_norm": 0.7146642208099365, "learning_rate": 4.321942537308332e-06, "loss": 0.0521, "step": 174670 }, { "epoch": 5.90554109334325, "grad_norm": 0.9276451468467712, "learning_rate": 4.3185808193078215e-06, "loss": 0.0457, "step": 174680 }, { "epoch": 5.9058791710335035, "grad_norm": 0.4660193920135498, "learning_rate": 4.31522035022014e-06, "loss": 0.0498, "step": 174690 }, { "epoch": 5.906217248723757, "grad_norm": 2.459573745727539, "learning_rate": 4.311861130137185e-06, "loss": 0.0578, "step": 174700 }, { "epoch": 5.90655532641401, "grad_norm": 2.2511887550354004, "learning_rate": 4.308503159150773e-06, "loss": 0.0501, "step": 174710 }, { "epoch": 5.906893404104263, "grad_norm": 0.8179429769515991, "learning_rate": 4.305146437352719e-06, "loss": 0.0498, "step": 174720 }, { "epoch": 5.907231481794517, "grad_norm": 1.1878376007080078, "learning_rate": 4.30179096483479e-06, "loss": 0.0496, "step": 174730 }, { "epoch": 5.907569559484769, "grad_norm": 2.4683148860931396, "learning_rate": 4.298436741688738e-06, "loss": 0.0497, "step": 174740 }, { "epoch": 5.907907637175023, "grad_norm": 0.8813265562057495, "learning_rate": 4.295083768006258e-06, "loss": 0.0519, "step": 174750 }, { "epoch": 5.908245714865276, "grad_norm": 1.3591547012329102, "learning_rate": 4.2917320438790044e-06, "loss": 0.0598, "step": 174760 }, { "epoch": 5.908583792555529, "grad_norm": 1.6151434183120728, "learning_rate": 4.288381569398636e-06, "loss": 0.0569, "step": 174770 }, { "epoch": 5.908921870245782, "grad_norm": 1.1995640993118286, "learning_rate": 4.285032344656742e-06, "loss": 0.0529, "step": 174780 }, { "epoch": 5.909259947936036, "grad_norm": 1.1027806997299194, "learning_rate": 4.281684369744887e-06, "loss": 0.0588, "step": 174790 }, { "epoch": 5.9095980256262886, "grad_norm": 1.3819372653961182, "learning_rate": 4.278337644754599e-06, "loss": 0.0501, "step": 174800 }, { "epoch": 5.909936103316542, "grad_norm": 1.2731051445007324, "learning_rate": 4.274992169777381e-06, "loss": 0.0536, "step": 174810 }, { "epoch": 5.910274181006796, "grad_norm": 1.064671516418457, "learning_rate": 4.271647944904711e-06, "loss": 0.0585, "step": 174820 }, { "epoch": 5.910612258697048, "grad_norm": 0.7372093796730042, "learning_rate": 4.268304970228005e-06, "loss": 0.0527, "step": 174830 }, { "epoch": 5.910950336387302, "grad_norm": 0.8172779679298401, "learning_rate": 4.26496324583866e-06, "loss": 0.0537, "step": 174840 }, { "epoch": 5.911288414077555, "grad_norm": 3.474588632583618, "learning_rate": 4.261622771828034e-06, "loss": 0.0608, "step": 174850 }, { "epoch": 5.911626491767808, "grad_norm": 2.3551738262176514, "learning_rate": 4.258283548287461e-06, "loss": 0.0562, "step": 174860 }, { "epoch": 5.911964569458061, "grad_norm": 6.119458198547363, "learning_rate": 4.254945575308233e-06, "loss": 0.0536, "step": 174870 }, { "epoch": 5.912302647148315, "grad_norm": 1.3490790128707886, "learning_rate": 4.251608852981603e-06, "loss": 0.0516, "step": 174880 }, { "epoch": 5.9126407248385675, "grad_norm": 1.2702897787094116, "learning_rate": 4.248273381398804e-06, "loss": 0.0536, "step": 174890 }, { "epoch": 5.912978802528821, "grad_norm": 2.1312813758850098, "learning_rate": 4.244939160651012e-06, "loss": 0.0468, "step": 174900 }, { "epoch": 5.9133168802190745, "grad_norm": 0.9768792986869812, "learning_rate": 4.241606190829417e-06, "loss": 0.058, "step": 174910 }, { "epoch": 5.913654957909328, "grad_norm": 2.0684335231781006, "learning_rate": 4.2382744720250996e-06, "loss": 0.0543, "step": 174920 }, { "epoch": 5.913993035599581, "grad_norm": 0.4374644458293915, "learning_rate": 4.234944004329161e-06, "loss": 0.0569, "step": 174930 }, { "epoch": 5.914331113289834, "grad_norm": 0.6985024809837341, "learning_rate": 4.231614787832677e-06, "loss": 0.0521, "step": 174940 }, { "epoch": 5.914669190980087, "grad_norm": 0.895413875579834, "learning_rate": 4.228286822626642e-06, "loss": 0.0544, "step": 174950 }, { "epoch": 5.91500726867034, "grad_norm": 0.5778948664665222, "learning_rate": 4.224960108802056e-06, "loss": 0.0497, "step": 174960 }, { "epoch": 5.915345346360594, "grad_norm": 1.0441021919250488, "learning_rate": 4.221634646449852e-06, "loss": 0.0501, "step": 174970 }, { "epoch": 5.915683424050847, "grad_norm": 1.5178664922714233, "learning_rate": 4.218310435660966e-06, "loss": 0.0514, "step": 174980 }, { "epoch": 5.9160215017411, "grad_norm": 0.7533966302871704, "learning_rate": 4.214987476526278e-06, "loss": 0.0491, "step": 174990 }, { "epoch": 5.9163595794313535, "grad_norm": 1.885515809059143, "learning_rate": 4.211665769136619e-06, "loss": 0.0545, "step": 175000 }, { "epoch": 5.916697657121606, "grad_norm": 2.002535820007324, "learning_rate": 4.2083453135828285e-06, "loss": 0.0562, "step": 175010 }, { "epoch": 5.91703573481186, "grad_norm": 1.1307125091552734, "learning_rate": 4.205026109955662e-06, "loss": 0.0528, "step": 175020 }, { "epoch": 5.917373812502113, "grad_norm": 2.6090474128723145, "learning_rate": 4.201708158345885e-06, "loss": 0.0516, "step": 175030 }, { "epoch": 5.917711890192367, "grad_norm": 0.625956654548645, "learning_rate": 4.198391458844203e-06, "loss": 0.0504, "step": 175040 }, { "epoch": 5.918049967882619, "grad_norm": 2.6914989948272705, "learning_rate": 4.195076011541282e-06, "loss": 0.0522, "step": 175050 }, { "epoch": 5.918388045572873, "grad_norm": 0.9585439562797546, "learning_rate": 4.1917618165277894e-06, "loss": 0.0503, "step": 175060 }, { "epoch": 5.918726123263125, "grad_norm": 1.8233078718185425, "learning_rate": 4.188448873894313e-06, "loss": 0.0588, "step": 175070 }, { "epoch": 5.919064200953379, "grad_norm": 1.2823973894119263, "learning_rate": 4.185137183731436e-06, "loss": 0.0497, "step": 175080 }, { "epoch": 5.919402278643632, "grad_norm": 0.9563882350921631, "learning_rate": 4.1818267461296856e-06, "loss": 0.0533, "step": 175090 }, { "epoch": 5.919740356333886, "grad_norm": 1.0956668853759766, "learning_rate": 4.17851756117959e-06, "loss": 0.0514, "step": 175100 }, { "epoch": 5.920078434024139, "grad_norm": 0.8269746899604797, "learning_rate": 4.175209628971599e-06, "loss": 0.0505, "step": 175110 }, { "epoch": 5.920416511714392, "grad_norm": 4.333338737487793, "learning_rate": 4.171902949596173e-06, "loss": 0.0571, "step": 175120 }, { "epoch": 5.920754589404645, "grad_norm": 1.48030686378479, "learning_rate": 4.168597523143702e-06, "loss": 0.0519, "step": 175130 }, { "epoch": 5.921092667094898, "grad_norm": 9.785242080688477, "learning_rate": 4.165293349704547e-06, "loss": 0.0533, "step": 175140 }, { "epoch": 5.921430744785152, "grad_norm": 2.360243320465088, "learning_rate": 4.161990429369062e-06, "loss": 0.0562, "step": 175150 }, { "epoch": 5.921768822475405, "grad_norm": 3.1179113388061523, "learning_rate": 4.158688762227536e-06, "loss": 0.0621, "step": 175160 }, { "epoch": 5.922106900165658, "grad_norm": 0.783032238483429, "learning_rate": 4.155388348370243e-06, "loss": 0.0548, "step": 175170 }, { "epoch": 5.922444977855911, "grad_norm": 0.8733497858047485, "learning_rate": 4.152089187887398e-06, "loss": 0.0526, "step": 175180 }, { "epoch": 5.922783055546164, "grad_norm": 1.1146334409713745, "learning_rate": 4.148791280869208e-06, "loss": 0.0516, "step": 175190 }, { "epoch": 5.9231211332364175, "grad_norm": 1.3701419830322266, "learning_rate": 4.1454946274058605e-06, "loss": 0.0472, "step": 175200 }, { "epoch": 5.923459210926671, "grad_norm": 1.5342648029327393, "learning_rate": 4.14219922758744e-06, "loss": 0.0547, "step": 175210 }, { "epoch": 5.923797288616925, "grad_norm": 2.349832534790039, "learning_rate": 4.13890508150408e-06, "loss": 0.0527, "step": 175220 }, { "epoch": 5.924135366307177, "grad_norm": 1.024430274963379, "learning_rate": 4.135612189245813e-06, "loss": 0.0477, "step": 175230 }, { "epoch": 5.924473443997431, "grad_norm": 2.0558133125305176, "learning_rate": 4.132320550902685e-06, "loss": 0.0576, "step": 175240 }, { "epoch": 5.924811521687684, "grad_norm": 1.3029589653015137, "learning_rate": 4.129030166564685e-06, "loss": 0.0511, "step": 175250 }, { "epoch": 5.925149599377937, "grad_norm": 1.2971380949020386, "learning_rate": 4.125741036321756e-06, "loss": 0.0497, "step": 175260 }, { "epoch": 5.92548767706819, "grad_norm": 1.208728551864624, "learning_rate": 4.122453160263845e-06, "loss": 0.0484, "step": 175270 }, { "epoch": 5.925825754758444, "grad_norm": 1.2558380365371704, "learning_rate": 4.119166538480828e-06, "loss": 0.0511, "step": 175280 }, { "epoch": 5.9261638324486965, "grad_norm": 0.9519052505493164, "learning_rate": 4.115881171062563e-06, "loss": 0.0542, "step": 175290 }, { "epoch": 5.92650191013895, "grad_norm": 2.6433417797088623, "learning_rate": 4.112597058098855e-06, "loss": 0.0503, "step": 175300 }, { "epoch": 5.9268399878292035, "grad_norm": 12.112420082092285, "learning_rate": 4.1093141996795095e-06, "loss": 0.0571, "step": 175310 }, { "epoch": 5.927178065519456, "grad_norm": 1.1724010705947876, "learning_rate": 4.106032595894277e-06, "loss": 0.052, "step": 175320 }, { "epoch": 5.92751614320971, "grad_norm": 0.7883080840110779, "learning_rate": 4.102752246832875e-06, "loss": 0.0501, "step": 175330 }, { "epoch": 5.927854220899963, "grad_norm": 2.623192310333252, "learning_rate": 4.0994731525849875e-06, "loss": 0.059, "step": 175340 }, { "epoch": 5.928192298590216, "grad_norm": 1.330926775932312, "learning_rate": 4.0961953132402474e-06, "loss": 0.0524, "step": 175350 }, { "epoch": 5.928530376280469, "grad_norm": 0.6081628203392029, "learning_rate": 4.09291872888829e-06, "loss": 0.0499, "step": 175360 }, { "epoch": 5.928868453970723, "grad_norm": 1.1303573846817017, "learning_rate": 4.089643399618687e-06, "loss": 0.0523, "step": 175370 }, { "epoch": 5.929206531660975, "grad_norm": 2.811758041381836, "learning_rate": 4.086369325520978e-06, "loss": 0.047, "step": 175380 }, { "epoch": 5.929544609351229, "grad_norm": 1.5261483192443848, "learning_rate": 4.083096506684697e-06, "loss": 0.0536, "step": 175390 }, { "epoch": 5.9298826870414825, "grad_norm": 11.00935173034668, "learning_rate": 4.07982494319929e-06, "loss": 0.0576, "step": 175400 }, { "epoch": 5.930220764731735, "grad_norm": 2.615626573562622, "learning_rate": 4.0765546351542396e-06, "loss": 0.0516, "step": 175410 }, { "epoch": 5.930558842421989, "grad_norm": 1.2582613229751587, "learning_rate": 4.073285582638914e-06, "loss": 0.0544, "step": 175420 }, { "epoch": 5.930896920112242, "grad_norm": 1.0027892589569092, "learning_rate": 4.070017785742708e-06, "loss": 0.0547, "step": 175430 }, { "epoch": 5.931234997802495, "grad_norm": 7.603614330291748, "learning_rate": 4.066751244554973e-06, "loss": 0.0582, "step": 175440 }, { "epoch": 5.931573075492748, "grad_norm": 1.1826221942901611, "learning_rate": 4.0634859591649975e-06, "loss": 0.0607, "step": 175450 }, { "epoch": 5.931911153183002, "grad_norm": 0.916907548904419, "learning_rate": 4.060221929662061e-06, "loss": 0.0504, "step": 175460 }, { "epoch": 5.932249230873254, "grad_norm": 0.9388952255249023, "learning_rate": 4.056959156135387e-06, "loss": 0.053, "step": 175470 }, { "epoch": 5.932587308563508, "grad_norm": 0.8675338625907898, "learning_rate": 4.0536976386741975e-06, "loss": 0.0513, "step": 175480 }, { "epoch": 5.932925386253761, "grad_norm": 0.6994855999946594, "learning_rate": 4.050437377367655e-06, "loss": 0.0516, "step": 175490 }, { "epoch": 5.933263463944014, "grad_norm": 0.9672110676765442, "learning_rate": 4.047178372304883e-06, "loss": 0.0481, "step": 175500 }, { "epoch": 5.933601541634268, "grad_norm": 0.85157710313797, "learning_rate": 4.0439206235749985e-06, "loss": 0.053, "step": 175510 }, { "epoch": 5.933939619324521, "grad_norm": 2.227482795715332, "learning_rate": 4.040664131267052e-06, "loss": 0.0484, "step": 175520 }, { "epoch": 5.934277697014774, "grad_norm": 2.1908504962921143, "learning_rate": 4.037408895470091e-06, "loss": 0.0542, "step": 175530 }, { "epoch": 5.934615774705027, "grad_norm": 2.3791451454162598, "learning_rate": 4.034154916273103e-06, "loss": 0.062, "step": 175540 }, { "epoch": 5.934953852395281, "grad_norm": 1.5495496988296509, "learning_rate": 4.030902193765052e-06, "loss": 0.0577, "step": 175550 }, { "epoch": 5.935291930085533, "grad_norm": 1.104900598526001, "learning_rate": 4.02765072803486e-06, "loss": 0.0525, "step": 175560 }, { "epoch": 5.935630007775787, "grad_norm": 0.6605034470558167, "learning_rate": 4.024400519171428e-06, "loss": 0.054, "step": 175570 }, { "epoch": 5.93596808546604, "grad_norm": 0.9658768773078918, "learning_rate": 4.021151567263614e-06, "loss": 0.045, "step": 175580 }, { "epoch": 5.936306163156293, "grad_norm": 1.5248565673828125, "learning_rate": 4.017903872400236e-06, "loss": 0.0541, "step": 175590 }, { "epoch": 5.9366442408465465, "grad_norm": 2.201854705810547, "learning_rate": 4.014657434670105e-06, "loss": 0.0565, "step": 175600 }, { "epoch": 5.9369823185368, "grad_norm": 0.8831501603126526, "learning_rate": 4.01141225416195e-06, "loss": 0.052, "step": 175610 }, { "epoch": 5.937320396227053, "grad_norm": 0.4978671073913574, "learning_rate": 4.008168330964523e-06, "loss": 0.0479, "step": 175620 }, { "epoch": 5.937658473917306, "grad_norm": 3.174859046936035, "learning_rate": 4.004925665166482e-06, "loss": 0.0539, "step": 175630 }, { "epoch": 5.93799655160756, "grad_norm": 9.255265235900879, "learning_rate": 4.0016842568564865e-06, "loss": 0.053, "step": 175640 }, { "epoch": 5.938334629297812, "grad_norm": 0.9777920842170715, "learning_rate": 3.998444106123178e-06, "loss": 0.0538, "step": 175650 }, { "epoch": 5.938672706988066, "grad_norm": 0.8015970587730408, "learning_rate": 3.995205213055125e-06, "loss": 0.051, "step": 175660 }, { "epoch": 5.939010784678319, "grad_norm": 1.1208018064498901, "learning_rate": 3.991967577740879e-06, "loss": 0.0531, "step": 175670 }, { "epoch": 5.939348862368572, "grad_norm": 0.5995882153511047, "learning_rate": 3.9887312002689416e-06, "loss": 0.0482, "step": 175680 }, { "epoch": 5.9396869400588255, "grad_norm": 1.8145248889923096, "learning_rate": 3.985496080727813e-06, "loss": 0.0527, "step": 175690 }, { "epoch": 5.940025017749079, "grad_norm": 1.008434534072876, "learning_rate": 3.982262219205945e-06, "loss": 0.0447, "step": 175700 }, { "epoch": 5.940363095439332, "grad_norm": 1.3308671712875366, "learning_rate": 3.979029615791729e-06, "loss": 0.054, "step": 175710 }, { "epoch": 5.940701173129585, "grad_norm": 6.574393272399902, "learning_rate": 3.975798270573561e-06, "loss": 0.0499, "step": 175720 }, { "epoch": 5.941039250819839, "grad_norm": 0.7945867776870728, "learning_rate": 3.972568183639768e-06, "loss": 0.0535, "step": 175730 }, { "epoch": 5.941377328510091, "grad_norm": 0.6180662512779236, "learning_rate": 3.9693393550786826e-06, "loss": 0.0494, "step": 175740 }, { "epoch": 5.941715406200345, "grad_norm": 1.2714482545852661, "learning_rate": 3.96611178497856e-06, "loss": 0.0502, "step": 175750 }, { "epoch": 5.942053483890598, "grad_norm": 0.5852651596069336, "learning_rate": 3.9628854734276414e-06, "loss": 0.0541, "step": 175760 }, { "epoch": 5.942391561580851, "grad_norm": 1.0105451345443726, "learning_rate": 3.959660420514149e-06, "loss": 0.0486, "step": 175770 }, { "epoch": 5.942729639271104, "grad_norm": 1.2127013206481934, "learning_rate": 3.956436626326238e-06, "loss": 0.0548, "step": 175780 }, { "epoch": 5.943067716961358, "grad_norm": 2.0024046897888184, "learning_rate": 3.953214090952056e-06, "loss": 0.0545, "step": 175790 }, { "epoch": 5.943405794651611, "grad_norm": 1.495911717414856, "learning_rate": 3.9499928144796975e-06, "loss": 0.0502, "step": 175800 }, { "epoch": 5.943743872341864, "grad_norm": 1.857650876045227, "learning_rate": 3.946772796997239e-06, "loss": 0.05, "step": 175810 }, { "epoch": 5.944081950032118, "grad_norm": 1.1440123319625854, "learning_rate": 3.9435540385927005e-06, "loss": 0.0538, "step": 175820 }, { "epoch": 5.94442002772237, "grad_norm": 0.7162172794342041, "learning_rate": 3.9403365393541e-06, "loss": 0.0489, "step": 175830 }, { "epoch": 5.944758105412624, "grad_norm": 0.5813451409339905, "learning_rate": 3.937120299369401e-06, "loss": 0.0546, "step": 175840 }, { "epoch": 5.945096183102877, "grad_norm": 1.388360857963562, "learning_rate": 3.933905318726516e-06, "loss": 0.0518, "step": 175850 }, { "epoch": 5.94543426079313, "grad_norm": 0.6097222566604614, "learning_rate": 3.930691597513358e-06, "loss": 0.0503, "step": 175860 }, { "epoch": 5.945772338483383, "grad_norm": 0.8973327875137329, "learning_rate": 3.927479135817791e-06, "loss": 0.051, "step": 175870 }, { "epoch": 5.946110416173637, "grad_norm": 0.6335031986236572, "learning_rate": 3.924267933727621e-06, "loss": 0.0554, "step": 175880 }, { "epoch": 5.94644849386389, "grad_norm": 0.7866342663764954, "learning_rate": 3.9210579913306675e-06, "loss": 0.0492, "step": 175890 }, { "epoch": 5.946786571554143, "grad_norm": 0.932715117931366, "learning_rate": 3.917849308714672e-06, "loss": 0.0568, "step": 175900 }, { "epoch": 5.9471246492443965, "grad_norm": 0.9268988966941833, "learning_rate": 3.914641885967379e-06, "loss": 0.0547, "step": 175910 }, { "epoch": 5.947462726934649, "grad_norm": 0.7966722249984741, "learning_rate": 3.911435723176443e-06, "loss": 0.0569, "step": 175920 }, { "epoch": 5.947800804624903, "grad_norm": 0.6245222687721252, "learning_rate": 3.908230820429554e-06, "loss": 0.0545, "step": 175930 }, { "epoch": 5.948138882315156, "grad_norm": 1.3709664344787598, "learning_rate": 3.9050271778143075e-06, "loss": 0.0563, "step": 175940 }, { "epoch": 5.94847696000541, "grad_norm": 1.6535601615905762, "learning_rate": 3.901824795418307e-06, "loss": 0.0539, "step": 175950 }, { "epoch": 5.948815037695662, "grad_norm": 1.5648014545440674, "learning_rate": 3.8986236733290994e-06, "loss": 0.051, "step": 175960 }, { "epoch": 5.949153115385916, "grad_norm": 0.6010932326316833, "learning_rate": 3.8954238116341914e-06, "loss": 0.0459, "step": 175970 }, { "epoch": 5.9494911930761685, "grad_norm": 0.8152380585670471, "learning_rate": 3.892225210421086e-06, "loss": 0.0506, "step": 175980 }, { "epoch": 5.949829270766422, "grad_norm": 1.793116569519043, "learning_rate": 3.889027869777223e-06, "loss": 0.0535, "step": 175990 }, { "epoch": 5.9501673484566755, "grad_norm": 0.8463582992553711, "learning_rate": 3.885831789790012e-06, "loss": 0.0523, "step": 176000 }, { "epoch": 5.950505426146929, "grad_norm": 1.2165303230285645, "learning_rate": 3.8826369705468326e-06, "loss": 0.0548, "step": 176010 }, { "epoch": 5.950843503837182, "grad_norm": 0.6177223324775696, "learning_rate": 3.8794434121350255e-06, "loss": 0.0582, "step": 176020 }, { "epoch": 5.951181581527435, "grad_norm": 0.711452305316925, "learning_rate": 3.876251114641921e-06, "loss": 0.0583, "step": 176030 }, { "epoch": 5.951519659217688, "grad_norm": 1.7291070222854614, "learning_rate": 3.873060078154783e-06, "loss": 0.0556, "step": 176040 }, { "epoch": 5.951857736907941, "grad_norm": 0.8794509768486023, "learning_rate": 3.869870302760853e-06, "loss": 0.05, "step": 176050 }, { "epoch": 5.952195814598195, "grad_norm": 1.5279831886291504, "learning_rate": 3.8666817885473325e-06, "loss": 0.0515, "step": 176060 }, { "epoch": 5.952533892288448, "grad_norm": 0.5355485081672668, "learning_rate": 3.863494535601403e-06, "loss": 0.055, "step": 176070 }, { "epoch": 5.952871969978701, "grad_norm": 1.1897412538528442, "learning_rate": 3.860308544010205e-06, "loss": 0.0528, "step": 176080 }, { "epoch": 5.9532100476689545, "grad_norm": 0.7049497961997986, "learning_rate": 3.85712381386083e-06, "loss": 0.0572, "step": 176090 }, { "epoch": 5.953548125359207, "grad_norm": 2.085759162902832, "learning_rate": 3.853940345240359e-06, "loss": 0.0559, "step": 176100 }, { "epoch": 5.953886203049461, "grad_norm": 1.2143230438232422, "learning_rate": 3.850758138235816e-06, "loss": 0.0519, "step": 176110 }, { "epoch": 5.954224280739714, "grad_norm": 0.6692935824394226, "learning_rate": 3.847577192934226e-06, "loss": 0.0501, "step": 176120 }, { "epoch": 5.954562358429968, "grad_norm": 3.8358399868011475, "learning_rate": 3.8443975094225205e-06, "loss": 0.0533, "step": 176130 }, { "epoch": 5.95490043612022, "grad_norm": 2.974072217941284, "learning_rate": 3.841219087787645e-06, "loss": 0.0541, "step": 176140 }, { "epoch": 5.955238513810474, "grad_norm": 1.3749788999557495, "learning_rate": 3.838041928116509e-06, "loss": 0.0538, "step": 176150 }, { "epoch": 5.955576591500726, "grad_norm": 1.3157120943069458, "learning_rate": 3.834866030495965e-06, "loss": 0.0523, "step": 176160 }, { "epoch": 5.95591466919098, "grad_norm": 0.8437052965164185, "learning_rate": 3.831691395012837e-06, "loss": 0.0531, "step": 176170 }, { "epoch": 5.956252746881233, "grad_norm": 1.4062399864196777, "learning_rate": 3.8285180217539185e-06, "loss": 0.0574, "step": 176180 }, { "epoch": 5.956590824571487, "grad_norm": 1.3441663980484009, "learning_rate": 3.825345910805972e-06, "loss": 0.0505, "step": 176190 }, { "epoch": 5.9569289022617395, "grad_norm": 0.9620845913887024, "learning_rate": 3.8221750622557285e-06, "loss": 0.0541, "step": 176200 }, { "epoch": 5.957266979951993, "grad_norm": 0.7706760764122009, "learning_rate": 3.8190054761898575e-06, "loss": 0.0464, "step": 176210 }, { "epoch": 5.957605057642247, "grad_norm": 1.5069082975387573, "learning_rate": 3.81583715269504e-06, "loss": 0.056, "step": 176220 }, { "epoch": 5.957943135332499, "grad_norm": 1.1473886966705322, "learning_rate": 3.812670091857873e-06, "loss": 0.0564, "step": 176230 }, { "epoch": 5.958281213022753, "grad_norm": 0.6870758533477783, "learning_rate": 3.809504293764965e-06, "loss": 0.0547, "step": 176240 }, { "epoch": 5.958619290713006, "grad_norm": 0.6102941632270813, "learning_rate": 3.8063397585028524e-06, "loss": 0.0484, "step": 176250 }, { "epoch": 5.958957368403259, "grad_norm": 2.424913167953491, "learning_rate": 3.80317648615805e-06, "loss": 0.0527, "step": 176260 }, { "epoch": 5.959295446093512, "grad_norm": 1.0422394275665283, "learning_rate": 3.8000144768170544e-06, "loss": 0.0504, "step": 176270 }, { "epoch": 5.959633523783766, "grad_norm": 2.5228097438812256, "learning_rate": 3.7968537305663077e-06, "loss": 0.0563, "step": 176280 }, { "epoch": 5.9599716014740185, "grad_norm": 0.7635006308555603, "learning_rate": 3.793694247492224e-06, "loss": 0.0563, "step": 176290 }, { "epoch": 5.960309679164272, "grad_norm": 5.336712837219238, "learning_rate": 3.7905360276811685e-06, "loss": 0.0555, "step": 176300 }, { "epoch": 5.9606477568545255, "grad_norm": 4.452243328094482, "learning_rate": 3.7873790712195102e-06, "loss": 0.0508, "step": 176310 }, { "epoch": 5.960985834544778, "grad_norm": 1.2091169357299805, "learning_rate": 3.7842233781935356e-06, "loss": 0.0528, "step": 176320 }, { "epoch": 5.961323912235032, "grad_norm": 1.0399771928787231, "learning_rate": 3.781068948689542e-06, "loss": 0.0543, "step": 176330 }, { "epoch": 5.961661989925285, "grad_norm": 1.2853022813796997, "learning_rate": 3.7779157827937615e-06, "loss": 0.0546, "step": 176340 }, { "epoch": 5.962000067615538, "grad_norm": 0.8480104804039001, "learning_rate": 3.7747638805923857e-06, "loss": 0.0578, "step": 176350 }, { "epoch": 5.962338145305791, "grad_norm": 2.4661614894866943, "learning_rate": 3.7716132421716123e-06, "loss": 0.0517, "step": 176360 }, { "epoch": 5.962676222996045, "grad_norm": 0.9247278571128845, "learning_rate": 3.768463867617561e-06, "loss": 0.0502, "step": 176370 }, { "epoch": 5.9630143006862975, "grad_norm": 0.8474062085151672, "learning_rate": 3.765315757016341e-06, "loss": 0.0495, "step": 176380 }, { "epoch": 5.963352378376551, "grad_norm": 0.8410447239875793, "learning_rate": 3.7621689104540116e-06, "loss": 0.0596, "step": 176390 }, { "epoch": 5.9636904560668045, "grad_norm": 2.1819369792938232, "learning_rate": 3.759023328016609e-06, "loss": 0.0516, "step": 176400 }, { "epoch": 5.964028533757057, "grad_norm": 0.60042405128479, "learning_rate": 3.755879009790159e-06, "loss": 0.0499, "step": 176410 }, { "epoch": 5.964366611447311, "grad_norm": 0.8819296956062317, "learning_rate": 3.752735955860587e-06, "loss": 0.0541, "step": 176420 }, { "epoch": 5.964704689137564, "grad_norm": 0.9742637276649475, "learning_rate": 3.7495941663138412e-06, "loss": 0.0556, "step": 176430 }, { "epoch": 5.965042766827817, "grad_norm": 10.010607719421387, "learning_rate": 3.7464536412358143e-06, "loss": 0.0503, "step": 176440 }, { "epoch": 5.96538084451807, "grad_norm": 1.7709412574768066, "learning_rate": 3.7433143807123707e-06, "loss": 0.0501, "step": 176450 }, { "epoch": 5.965718922208324, "grad_norm": 2.195117950439453, "learning_rate": 3.7401763848293415e-06, "loss": 0.0521, "step": 176460 }, { "epoch": 5.966056999898576, "grad_norm": 2.15633225440979, "learning_rate": 3.7370396536724973e-06, "loss": 0.0514, "step": 176470 }, { "epoch": 5.96639507758883, "grad_norm": 3.6843371391296387, "learning_rate": 3.733904187327619e-06, "loss": 0.0538, "step": 176480 }, { "epoch": 5.966733155279083, "grad_norm": 0.4708469808101654, "learning_rate": 3.730769985880417e-06, "loss": 0.0552, "step": 176490 }, { "epoch": 5.967071232969336, "grad_norm": 1.408060908317566, "learning_rate": 3.7276370494165824e-06, "loss": 0.0524, "step": 176500 }, { "epoch": 5.96740931065959, "grad_norm": 0.5675322413444519, "learning_rate": 3.724505378021759e-06, "loss": 0.0485, "step": 176510 }, { "epoch": 5.967747388349843, "grad_norm": 0.8457247018814087, "learning_rate": 3.7213749717815717e-06, "loss": 0.0558, "step": 176520 }, { "epoch": 5.968085466040096, "grad_norm": 1.7386716604232788, "learning_rate": 3.718245830781614e-06, "loss": 0.0514, "step": 176530 }, { "epoch": 5.968423543730349, "grad_norm": 1.5349600315093994, "learning_rate": 3.7151179551074332e-06, "loss": 0.0589, "step": 176540 }, { "epoch": 5.968761621420603, "grad_norm": 1.0147522687911987, "learning_rate": 3.7119913448445343e-06, "loss": 0.0523, "step": 176550 }, { "epoch": 5.969099699110855, "grad_norm": 11.902310371398926, "learning_rate": 3.708866000078398e-06, "loss": 0.0533, "step": 176560 }, { "epoch": 5.969437776801109, "grad_norm": 0.529301643371582, "learning_rate": 3.7057419208944786e-06, "loss": 0.0495, "step": 176570 }, { "epoch": 5.969775854491362, "grad_norm": 1.9103697538375854, "learning_rate": 3.7026191073781857e-06, "loss": 0.0438, "step": 176580 }, { "epoch": 5.970113932181615, "grad_norm": 2.1401851177215576, "learning_rate": 3.699497559614884e-06, "loss": 0.057, "step": 176590 }, { "epoch": 5.9704520098718685, "grad_norm": 1.0663458108901978, "learning_rate": 3.6963772776899388e-06, "loss": 0.0521, "step": 176600 }, { "epoch": 5.970790087562122, "grad_norm": 0.8199230432510376, "learning_rate": 3.693258261688626e-06, "loss": 0.0493, "step": 176610 }, { "epoch": 5.971128165252375, "grad_norm": 0.9470874071121216, "learning_rate": 3.690140511696261e-06, "loss": 0.0575, "step": 176620 }, { "epoch": 5.971466242942628, "grad_norm": 0.6932397484779358, "learning_rate": 3.6870240277980304e-06, "loss": 0.0486, "step": 176630 }, { "epoch": 5.971804320632882, "grad_norm": 0.5661075115203857, "learning_rate": 3.683908810079173e-06, "loss": 0.0525, "step": 176640 }, { "epoch": 5.972142398323134, "grad_norm": 1.0836631059646606, "learning_rate": 3.680794858624853e-06, "loss": 0.0484, "step": 176650 }, { "epoch": 5.972480476013388, "grad_norm": 0.8969037532806396, "learning_rate": 3.6776821735201972e-06, "loss": 0.0478, "step": 176660 }, { "epoch": 5.972818553703641, "grad_norm": 2.344226598739624, "learning_rate": 3.6745707548503096e-06, "loss": 0.054, "step": 176670 }, { "epoch": 5.973156631393894, "grad_norm": 1.13894784450531, "learning_rate": 3.6714606027002497e-06, "loss": 0.0501, "step": 176680 }, { "epoch": 5.9734947090841475, "grad_norm": 4.4794135093688965, "learning_rate": 3.6683517171550553e-06, "loss": 0.0475, "step": 176690 }, { "epoch": 5.973832786774401, "grad_norm": 0.5956399440765381, "learning_rate": 3.665244098299725e-06, "loss": 0.0551, "step": 176700 }, { "epoch": 5.974170864464654, "grad_norm": 0.9360091686248779, "learning_rate": 3.662137746219202e-06, "loss": 0.0548, "step": 176710 }, { "epoch": 5.974508942154907, "grad_norm": 0.7717128992080688, "learning_rate": 3.6590326609984294e-06, "loss": 0.0504, "step": 176720 }, { "epoch": 5.974847019845161, "grad_norm": 3.2000739574432373, "learning_rate": 3.6559288427222894e-06, "loss": 0.0509, "step": 176730 }, { "epoch": 5.975185097535413, "grad_norm": 0.9215921759605408, "learning_rate": 3.652826291475653e-06, "loss": 0.056, "step": 176740 }, { "epoch": 5.975523175225667, "grad_norm": 1.6064485311508179, "learning_rate": 3.6497250073433352e-06, "loss": 0.0615, "step": 176750 }, { "epoch": 5.97586125291592, "grad_norm": 1.4101089239120483, "learning_rate": 3.646624990410119e-06, "loss": 0.0513, "step": 176760 }, { "epoch": 5.976199330606173, "grad_norm": 18.68932342529297, "learning_rate": 3.6435262407607582e-06, "loss": 0.0564, "step": 176770 }, { "epoch": 5.976537408296426, "grad_norm": 1.2043499946594238, "learning_rate": 3.64042875847998e-06, "loss": 0.0481, "step": 176780 }, { "epoch": 5.97687548598668, "grad_norm": 1.1714786291122437, "learning_rate": 3.637332543652461e-06, "loss": 0.0525, "step": 176790 }, { "epoch": 5.977213563676933, "grad_norm": 0.8322468996047974, "learning_rate": 3.6342375963628493e-06, "loss": 0.0515, "step": 176800 }, { "epoch": 5.977551641367186, "grad_norm": 0.7853884100914001, "learning_rate": 3.631143916695767e-06, "loss": 0.0477, "step": 176810 }, { "epoch": 5.97788971905744, "grad_norm": 1.094855546951294, "learning_rate": 3.6280515047357856e-06, "loss": 0.0494, "step": 176820 }, { "epoch": 5.978227796747692, "grad_norm": 0.7849136590957642, "learning_rate": 3.62496036056747e-06, "loss": 0.0468, "step": 176830 }, { "epoch": 5.978565874437946, "grad_norm": 0.9673792719841003, "learning_rate": 3.6218704842752972e-06, "loss": 0.0568, "step": 176840 }, { "epoch": 5.978903952128199, "grad_norm": 0.9799716472625732, "learning_rate": 3.6187818759437607e-06, "loss": 0.0583, "step": 176850 }, { "epoch": 5.979242029818453, "grad_norm": 1.7123844623565674, "learning_rate": 3.615694535657316e-06, "loss": 0.052, "step": 176860 }, { "epoch": 5.979580107508705, "grad_norm": 0.9609081149101257, "learning_rate": 3.6126084635003553e-06, "loss": 0.054, "step": 176870 }, { "epoch": 5.979918185198959, "grad_norm": 1.1352026462554932, "learning_rate": 3.6095236595572514e-06, "loss": 0.0543, "step": 176880 }, { "epoch": 5.9802562628892115, "grad_norm": 0.42542481422424316, "learning_rate": 3.60644012391233e-06, "loss": 0.0565, "step": 176890 }, { "epoch": 5.980594340579465, "grad_norm": 0.7738592624664307, "learning_rate": 3.6033578566499137e-06, "loss": 0.0512, "step": 176900 }, { "epoch": 5.9809324182697186, "grad_norm": 0.648725152015686, "learning_rate": 3.6002768578542788e-06, "loss": 0.0498, "step": 176910 }, { "epoch": 5.981270495959972, "grad_norm": 0.8655660152435303, "learning_rate": 3.5971971276096197e-06, "loss": 0.0484, "step": 176920 }, { "epoch": 5.981608573650225, "grad_norm": 1.3460090160369873, "learning_rate": 3.5941186660001734e-06, "loss": 0.0519, "step": 176930 }, { "epoch": 5.981946651340478, "grad_norm": 9.486783027648926, "learning_rate": 3.5910414731100796e-06, "loss": 0.0525, "step": 176940 }, { "epoch": 5.982284729030731, "grad_norm": 2.8897602558135986, "learning_rate": 3.5879655490234866e-06, "loss": 0.0526, "step": 176950 }, { "epoch": 5.982622806720984, "grad_norm": 2.2650184631347656, "learning_rate": 3.5848908938244774e-06, "loss": 0.0503, "step": 176960 }, { "epoch": 5.982960884411238, "grad_norm": 2.943432331085205, "learning_rate": 3.5818175075971017e-06, "loss": 0.0526, "step": 176970 }, { "epoch": 5.983298962101491, "grad_norm": 1.2715319395065308, "learning_rate": 3.5787453904254086e-06, "loss": 0.0512, "step": 176980 }, { "epoch": 5.983637039791744, "grad_norm": 6.024862766265869, "learning_rate": 3.5756745423933758e-06, "loss": 0.05, "step": 176990 }, { "epoch": 5.9839751174819975, "grad_norm": 1.1213430166244507, "learning_rate": 3.5726049635849633e-06, "loss": 0.0529, "step": 177000 }, { "epoch": 5.98431319517225, "grad_norm": 0.5253880023956299, "learning_rate": 3.569536654084077e-06, "loss": 0.0522, "step": 177010 }, { "epoch": 5.984651272862504, "grad_norm": 1.4074110984802246, "learning_rate": 3.566469613974627e-06, "loss": 0.0589, "step": 177020 }, { "epoch": 5.984989350552757, "grad_norm": 1.2978700399398804, "learning_rate": 3.563403843340446e-06, "loss": 0.0572, "step": 177030 }, { "epoch": 5.985327428243011, "grad_norm": 0.8025118708610535, "learning_rate": 3.560339342265362e-06, "loss": 0.0506, "step": 177040 }, { "epoch": 5.985665505933263, "grad_norm": 1.1821000576019287, "learning_rate": 3.5572761108331575e-06, "loss": 0.0577, "step": 177050 }, { "epoch": 5.986003583623517, "grad_norm": 0.9177402257919312, "learning_rate": 3.554214149127566e-06, "loss": 0.0547, "step": 177060 }, { "epoch": 5.986341661313769, "grad_norm": 1.763792634010315, "learning_rate": 3.55115345723232e-06, "loss": 0.0523, "step": 177070 }, { "epoch": 5.986679739004023, "grad_norm": 1.0922785997390747, "learning_rate": 3.5480940352310867e-06, "loss": 0.0548, "step": 177080 }, { "epoch": 5.9870178166942765, "grad_norm": 0.457952618598938, "learning_rate": 3.5450358832075093e-06, "loss": 0.0535, "step": 177090 }, { "epoch": 5.98735589438453, "grad_norm": 4.031879901885986, "learning_rate": 3.541979001245199e-06, "loss": 0.0525, "step": 177100 }, { "epoch": 5.987693972074783, "grad_norm": 1.690330982208252, "learning_rate": 3.5389233894277285e-06, "loss": 0.0504, "step": 177110 }, { "epoch": 5.988032049765036, "grad_norm": 0.5457683205604553, "learning_rate": 3.535869047838647e-06, "loss": 0.05, "step": 177120 }, { "epoch": 5.988370127455289, "grad_norm": 1.83842933177948, "learning_rate": 3.532815976561438e-06, "loss": 0.054, "step": 177130 }, { "epoch": 5.988708205145542, "grad_norm": 0.6812132596969604, "learning_rate": 3.5297641756795896e-06, "loss": 0.051, "step": 177140 }, { "epoch": 5.989046282835796, "grad_norm": 0.7436808347702026, "learning_rate": 3.526713645276525e-06, "loss": 0.0515, "step": 177150 }, { "epoch": 5.989384360526049, "grad_norm": 0.9853156208992004, "learning_rate": 3.523664385435654e-06, "loss": 0.0491, "step": 177160 }, { "epoch": 5.989722438216302, "grad_norm": 2.631256341934204, "learning_rate": 3.5206163962403392e-06, "loss": 0.05, "step": 177170 }, { "epoch": 5.990060515906555, "grad_norm": 0.8670761585235596, "learning_rate": 3.5175696777739078e-06, "loss": 0.056, "step": 177180 }, { "epoch": 5.990398593596809, "grad_norm": 0.9917459487915039, "learning_rate": 3.5145242301196646e-06, "loss": 0.0469, "step": 177190 }, { "epoch": 5.9907366712870616, "grad_norm": 0.8075500726699829, "learning_rate": 3.5114800533608603e-06, "loss": 0.0561, "step": 177200 }, { "epoch": 5.991074748977315, "grad_norm": 1.2204899787902832, "learning_rate": 3.5084371475807287e-06, "loss": 0.0541, "step": 177210 }, { "epoch": 5.991412826667569, "grad_norm": 1.1431175470352173, "learning_rate": 3.5053955128624525e-06, "loss": 0.0499, "step": 177220 }, { "epoch": 5.991750904357821, "grad_norm": 0.7991443276405334, "learning_rate": 3.5023551492891995e-06, "loss": 0.0558, "step": 177230 }, { "epoch": 5.992088982048075, "grad_norm": 1.126251459121704, "learning_rate": 3.499316056944091e-06, "loss": 0.061, "step": 177240 }, { "epoch": 5.992427059738328, "grad_norm": 0.6760119199752808, "learning_rate": 3.4962782359102163e-06, "loss": 0.0586, "step": 177250 }, { "epoch": 5.992765137428581, "grad_norm": 1.0468720197677612, "learning_rate": 3.4932416862706208e-06, "loss": 0.0477, "step": 177260 }, { "epoch": 5.993103215118834, "grad_norm": 1.440169334411621, "learning_rate": 3.490206408108315e-06, "loss": 0.0509, "step": 177270 }, { "epoch": 5.993441292809088, "grad_norm": 1.5333809852600098, "learning_rate": 3.4871724015063055e-06, "loss": 0.0551, "step": 177280 }, { "epoch": 5.9937793704993405, "grad_norm": 1.571926236152649, "learning_rate": 3.484139666547526e-06, "loss": 0.0535, "step": 177290 }, { "epoch": 5.994117448189594, "grad_norm": 0.9815561175346375, "learning_rate": 3.481108203314887e-06, "loss": 0.0528, "step": 177300 }, { "epoch": 5.9944555258798475, "grad_norm": 1.829502820968628, "learning_rate": 3.478078011891284e-06, "loss": 0.0544, "step": 177310 }, { "epoch": 5.9947936035701, "grad_norm": 0.9678701758384705, "learning_rate": 3.475049092359539e-06, "loss": 0.0561, "step": 177320 }, { "epoch": 5.995131681260354, "grad_norm": 0.9106041193008423, "learning_rate": 3.472021444802487e-06, "loss": 0.0509, "step": 177330 }, { "epoch": 5.995469758950607, "grad_norm": 0.6729151010513306, "learning_rate": 3.4689950693028773e-06, "loss": 0.0563, "step": 177340 }, { "epoch": 5.99580783664086, "grad_norm": 0.6301792860031128, "learning_rate": 3.4659699659434606e-06, "loss": 0.0449, "step": 177350 }, { "epoch": 5.996145914331113, "grad_norm": 0.768196702003479, "learning_rate": 3.462946134806949e-06, "loss": 0.051, "step": 177360 }, { "epoch": 5.996483992021367, "grad_norm": 1.3492717742919922, "learning_rate": 3.4599235759760096e-06, "loss": 0.0467, "step": 177370 }, { "epoch": 5.9968220697116195, "grad_norm": 3.339792490005493, "learning_rate": 3.4569022895332704e-06, "loss": 0.0529, "step": 177380 }, { "epoch": 5.997160147401873, "grad_norm": 2.7739152908325195, "learning_rate": 3.453882275561332e-06, "loss": 0.0579, "step": 177390 }, { "epoch": 5.9974982250921265, "grad_norm": 1.3790876865386963, "learning_rate": 3.4508635341427788e-06, "loss": 0.0545, "step": 177400 }, { "epoch": 5.997836302782379, "grad_norm": 3.907989263534546, "learning_rate": 3.447846065360122e-06, "loss": 0.0513, "step": 177410 }, { "epoch": 5.998174380472633, "grad_norm": 1.4583795070648193, "learning_rate": 3.444829869295857e-06, "loss": 0.0501, "step": 177420 }, { "epoch": 5.998512458162886, "grad_norm": 0.6259157657623291, "learning_rate": 3.4418149460324676e-06, "loss": 0.0503, "step": 177430 }, { "epoch": 5.998850535853139, "grad_norm": 1.0125871896743774, "learning_rate": 3.4388012956523553e-06, "loss": 0.0517, "step": 177440 }, { "epoch": 5.999188613543392, "grad_norm": 1.2417525053024292, "learning_rate": 3.435788918237931e-06, "loss": 0.0505, "step": 177450 }, { "epoch": 5.999526691233646, "grad_norm": 0.5610377788543701, "learning_rate": 3.4327778138715404e-06, "loss": 0.0547, "step": 177460 }, { "epoch": 5.999864768923898, "grad_norm": 2.056043863296509, "learning_rate": 3.4297679826355067e-06, "loss": 0.0614, "step": 177470 }, { "epoch": 6.000202846614152, "grad_norm": 0.5809886455535889, "learning_rate": 3.426759424612125e-06, "loss": 0.0559, "step": 177480 }, { "epoch": 6.000540924304405, "grad_norm": 0.6476240754127502, "learning_rate": 3.423752139883646e-06, "loss": 0.0524, "step": 177490 }, { "epoch": 6.000879001994658, "grad_norm": 2.6009180545806885, "learning_rate": 3.420746128532287e-06, "loss": 0.0525, "step": 177500 }, { "epoch": 6.001217079684912, "grad_norm": 0.6811365485191345, "learning_rate": 3.417741390640222e-06, "loss": 0.0544, "step": 177510 }, { "epoch": 6.001555157375165, "grad_norm": 1.3972327709197998, "learning_rate": 3.4147379262896127e-06, "loss": 0.0556, "step": 177520 }, { "epoch": 6.001893235065418, "grad_norm": 0.8767895698547363, "learning_rate": 3.41173573556256e-06, "loss": 0.0499, "step": 177530 }, { "epoch": 6.002231312755671, "grad_norm": 1.8344488143920898, "learning_rate": 3.4087348185411596e-06, "loss": 0.0519, "step": 177540 }, { "epoch": 6.002569390445925, "grad_norm": 0.5025607347488403, "learning_rate": 3.405735175307445e-06, "loss": 0.0531, "step": 177550 }, { "epoch": 6.002907468136177, "grad_norm": 1.7989842891693115, "learning_rate": 3.4027368059434186e-06, "loss": 0.0525, "step": 177560 }, { "epoch": 6.003245545826431, "grad_norm": 0.656892716884613, "learning_rate": 3.399739710531069e-06, "loss": 0.051, "step": 177570 }, { "epoch": 6.003583623516684, "grad_norm": 0.6949807405471802, "learning_rate": 3.396743889152326e-06, "loss": 0.0514, "step": 177580 }, { "epoch": 6.003921701206937, "grad_norm": 1.4980647563934326, "learning_rate": 3.393749341889102e-06, "loss": 0.0566, "step": 177590 }, { "epoch": 6.0042597788971905, "grad_norm": 1.3102824687957764, "learning_rate": 3.390756068823253e-06, "loss": 0.055, "step": 177600 }, { "epoch": 6.004597856587444, "grad_norm": 1.1378448009490967, "learning_rate": 3.3877640700366197e-06, "loss": 0.048, "step": 177610 }, { "epoch": 6.004935934277697, "grad_norm": 0.728407084941864, "learning_rate": 3.3847733456110254e-06, "loss": 0.0507, "step": 177620 }, { "epoch": 6.00527401196795, "grad_norm": 0.5744236707687378, "learning_rate": 3.381783895628193e-06, "loss": 0.0536, "step": 177630 }, { "epoch": 6.005612089658204, "grad_norm": 0.5726353526115417, "learning_rate": 3.378795720169892e-06, "loss": 0.059, "step": 177640 }, { "epoch": 6.005950167348456, "grad_norm": 0.8111822605133057, "learning_rate": 3.3758088193177884e-06, "loss": 0.0546, "step": 177650 }, { "epoch": 6.00628824503871, "grad_norm": 1.7824972867965698, "learning_rate": 3.3728231931535627e-06, "loss": 0.0544, "step": 177660 }, { "epoch": 6.006626322728963, "grad_norm": 2.7810704708099365, "learning_rate": 3.3698388417588323e-06, "loss": 0.053, "step": 177670 }, { "epoch": 6.006964400419216, "grad_norm": 0.8481602072715759, "learning_rate": 3.3668557652151877e-06, "loss": 0.0515, "step": 177680 }, { "epoch": 6.0073024781094695, "grad_norm": 0.5348073840141296, "learning_rate": 3.3638739636041918e-06, "loss": 0.0534, "step": 177690 }, { "epoch": 6.007640555799723, "grad_norm": 1.4670922756195068, "learning_rate": 3.360893437007362e-06, "loss": 0.0549, "step": 177700 }, { "epoch": 6.007978633489976, "grad_norm": 0.8462061882019043, "learning_rate": 3.3579141855061837e-06, "loss": 0.0549, "step": 177710 }, { "epoch": 6.008316711180229, "grad_norm": 1.4669020175933838, "learning_rate": 3.3549362091821026e-06, "loss": 0.05, "step": 177720 }, { "epoch": 6.008654788870483, "grad_norm": 1.0631874799728394, "learning_rate": 3.351959508116537e-06, "loss": 0.0464, "step": 177730 }, { "epoch": 6.008992866560735, "grad_norm": 8.15190601348877, "learning_rate": 3.348984082390888e-06, "loss": 0.0538, "step": 177740 }, { "epoch": 6.009330944250989, "grad_norm": 0.6179962158203125, "learning_rate": 3.3460099320864857e-06, "loss": 0.0542, "step": 177750 }, { "epoch": 6.009669021941242, "grad_norm": 3.405562400817871, "learning_rate": 3.3430370572846425e-06, "loss": 0.0528, "step": 177760 }, { "epoch": 6.010007099631495, "grad_norm": 1.1916911602020264, "learning_rate": 3.3400654580666325e-06, "loss": 0.0523, "step": 177770 }, { "epoch": 6.0103451773217484, "grad_norm": 0.8406286239624023, "learning_rate": 3.3370951345137125e-06, "loss": 0.0507, "step": 177780 }, { "epoch": 6.010683255012002, "grad_norm": 0.8242563009262085, "learning_rate": 3.334126086707079e-06, "loss": 0.0574, "step": 177790 }, { "epoch": 6.011021332702255, "grad_norm": 1.1409507989883423, "learning_rate": 3.3311583147278945e-06, "loss": 0.0515, "step": 177800 }, { "epoch": 6.011359410392508, "grad_norm": 1.030656099319458, "learning_rate": 3.3281918186573225e-06, "loss": 0.0485, "step": 177810 }, { "epoch": 6.011697488082762, "grad_norm": 1.8863776922225952, "learning_rate": 3.3252265985764475e-06, "loss": 0.0524, "step": 177820 }, { "epoch": 6.012035565773014, "grad_norm": 1.2408145666122437, "learning_rate": 3.3222626545663493e-06, "loss": 0.0525, "step": 177830 }, { "epoch": 6.012373643463268, "grad_norm": 0.9297131299972534, "learning_rate": 3.3192999867080466e-06, "loss": 0.0587, "step": 177840 }, { "epoch": 6.012711721153521, "grad_norm": 0.6416146159172058, "learning_rate": 3.316338595082541e-06, "loss": 0.0531, "step": 177850 }, { "epoch": 6.013049798843774, "grad_norm": 0.8100929260253906, "learning_rate": 3.3133784797708066e-06, "loss": 0.0501, "step": 177860 }, { "epoch": 6.013387876534027, "grad_norm": 0.812581479549408, "learning_rate": 3.310419640853768e-06, "loss": 0.0477, "step": 177870 }, { "epoch": 6.013725954224281, "grad_norm": 1.4870941638946533, "learning_rate": 3.3074620784123157e-06, "loss": 0.0497, "step": 177880 }, { "epoch": 6.0140640319145335, "grad_norm": 0.7027384638786316, "learning_rate": 3.3045057925273014e-06, "loss": 0.054, "step": 177890 }, { "epoch": 6.014402109604787, "grad_norm": 1.0885708332061768, "learning_rate": 3.3015507832795613e-06, "loss": 0.0551, "step": 177900 }, { "epoch": 6.014740187295041, "grad_norm": 0.6212389469146729, "learning_rate": 3.2985970507498797e-06, "loss": 0.0517, "step": 177910 }, { "epoch": 6.015078264985293, "grad_norm": 1.0112438201904297, "learning_rate": 3.295644595019004e-06, "loss": 0.0505, "step": 177920 }, { "epoch": 6.015416342675547, "grad_norm": 1.6844661235809326, "learning_rate": 3.2926934161676694e-06, "loss": 0.0485, "step": 177930 }, { "epoch": 6.0157544203658, "grad_norm": 0.609917402267456, "learning_rate": 3.289743514276533e-06, "loss": 0.0527, "step": 177940 }, { "epoch": 6.016092498056053, "grad_norm": 0.5230181217193604, "learning_rate": 3.2867948894262757e-06, "loss": 0.0563, "step": 177950 }, { "epoch": 6.016430575746306, "grad_norm": 2.203000068664551, "learning_rate": 3.2838475416974933e-06, "loss": 0.0523, "step": 177960 }, { "epoch": 6.01676865343656, "grad_norm": 0.7991521954536438, "learning_rate": 3.280901471170772e-06, "loss": 0.0497, "step": 177970 }, { "epoch": 6.017106731126813, "grad_norm": 0.571869969367981, "learning_rate": 3.2779566779266414e-06, "loss": 0.0475, "step": 177980 }, { "epoch": 6.017444808817066, "grad_norm": 0.8259674906730652, "learning_rate": 3.2750131620456314e-06, "loss": 0.0514, "step": 177990 }, { "epoch": 6.0177828865073195, "grad_norm": 0.818871796131134, "learning_rate": 3.2720709236082116e-06, "loss": 0.0545, "step": 178000 }, { "epoch": 6.018120964197573, "grad_norm": 4.747012138366699, "learning_rate": 3.2691299626948057e-06, "loss": 0.0593, "step": 178010 }, { "epoch": 6.018459041887826, "grad_norm": 0.9738027453422546, "learning_rate": 3.266190279385839e-06, "loss": 0.0533, "step": 178020 }, { "epoch": 6.018797119578079, "grad_norm": 0.71379554271698, "learning_rate": 3.263251873761669e-06, "loss": 0.0575, "step": 178030 }, { "epoch": 6.019135197268333, "grad_norm": 1.2574520111083984, "learning_rate": 3.2603147459026482e-06, "loss": 0.0476, "step": 178040 }, { "epoch": 6.019473274958585, "grad_norm": 0.6790558099746704, "learning_rate": 3.257378895889046e-06, "loss": 0.0506, "step": 178050 }, { "epoch": 6.019811352648839, "grad_norm": 1.3716367483139038, "learning_rate": 3.2544443238011425e-06, "loss": 0.0532, "step": 178060 }, { "epoch": 6.020149430339092, "grad_norm": 1.0306942462921143, "learning_rate": 3.2515110297191786e-06, "loss": 0.0552, "step": 178070 }, { "epoch": 6.020487508029345, "grad_norm": 0.835770308971405, "learning_rate": 3.248579013723341e-06, "loss": 0.0535, "step": 178080 }, { "epoch": 6.0208255857195985, "grad_norm": 0.7697169184684753, "learning_rate": 3.2456482758937923e-06, "loss": 0.0505, "step": 178090 }, { "epoch": 6.021163663409852, "grad_norm": 0.4611896872520447, "learning_rate": 3.242718816310647e-06, "loss": 0.0549, "step": 178100 }, { "epoch": 6.021501741100105, "grad_norm": 1.4944427013397217, "learning_rate": 3.239790635054002e-06, "loss": 0.0596, "step": 178110 }, { "epoch": 6.021839818790358, "grad_norm": 1.658787727355957, "learning_rate": 3.236863732203926e-06, "loss": 0.049, "step": 178120 }, { "epoch": 6.022177896480612, "grad_norm": 0.9955359101295471, "learning_rate": 3.2339381078404175e-06, "loss": 0.0559, "step": 178130 }, { "epoch": 6.022515974170864, "grad_norm": 21.523012161254883, "learning_rate": 3.231013762043472e-06, "loss": 0.0497, "step": 178140 }, { "epoch": 6.022854051861118, "grad_norm": 0.9992521405220032, "learning_rate": 3.228090694893038e-06, "loss": 0.0523, "step": 178150 }, { "epoch": 6.023192129551371, "grad_norm": 2.2413549423217773, "learning_rate": 3.225168906469045e-06, "loss": 0.0554, "step": 178160 }, { "epoch": 6.023530207241624, "grad_norm": 3.1604979038238525, "learning_rate": 3.2222483968513516e-06, "loss": 0.0561, "step": 178170 }, { "epoch": 6.023868284931877, "grad_norm": 2.1894772052764893, "learning_rate": 3.219329166119811e-06, "loss": 0.0483, "step": 178180 }, { "epoch": 6.024206362622131, "grad_norm": 1.080214023590088, "learning_rate": 3.216411214354248e-06, "loss": 0.0547, "step": 178190 }, { "epoch": 6.024544440312384, "grad_norm": 0.8102483153343201, "learning_rate": 3.2134945416344207e-06, "loss": 0.0513, "step": 178200 }, { "epoch": 6.024882518002637, "grad_norm": 0.9219042658805847, "learning_rate": 3.210579148040077e-06, "loss": 0.0529, "step": 178210 }, { "epoch": 6.025220595692891, "grad_norm": 1.830567479133606, "learning_rate": 3.207665033650914e-06, "loss": 0.0518, "step": 178220 }, { "epoch": 6.025558673383143, "grad_norm": 0.7587426900863647, "learning_rate": 3.2047521985466177e-06, "loss": 0.0539, "step": 178230 }, { "epoch": 6.025896751073397, "grad_norm": 1.0046552419662476, "learning_rate": 3.201840642806808e-06, "loss": 0.0499, "step": 178240 }, { "epoch": 6.02623482876365, "grad_norm": 0.9825737476348877, "learning_rate": 3.1989303665111046e-06, "loss": 0.05, "step": 178250 }, { "epoch": 6.026572906453903, "grad_norm": 1.221390962600708, "learning_rate": 3.1960213697390607e-06, "loss": 0.0572, "step": 178260 }, { "epoch": 6.026910984144156, "grad_norm": 0.6956121921539307, "learning_rate": 3.193113652570201e-06, "loss": 0.0489, "step": 178270 }, { "epoch": 6.02724906183441, "grad_norm": 2.1043052673339844, "learning_rate": 3.1902072150840345e-06, "loss": 0.0519, "step": 178280 }, { "epoch": 6.0275871395246625, "grad_norm": 0.9818426966667175, "learning_rate": 3.18730205736002e-06, "loss": 0.056, "step": 178290 }, { "epoch": 6.027925217214916, "grad_norm": 0.8236175775527954, "learning_rate": 3.1843981794775714e-06, "loss": 0.0528, "step": 178300 }, { "epoch": 6.0282632949051695, "grad_norm": 0.49681565165519714, "learning_rate": 3.1814955815160975e-06, "loss": 0.0491, "step": 178310 }, { "epoch": 6.028601372595422, "grad_norm": 1.3164501190185547, "learning_rate": 3.1785942635549404e-06, "loss": 0.056, "step": 178320 }, { "epoch": 6.028939450285676, "grad_norm": 0.6460851430892944, "learning_rate": 3.1756942256734366e-06, "loss": 0.0515, "step": 178330 }, { "epoch": 6.029277527975929, "grad_norm": 0.7022082805633545, "learning_rate": 3.172795467950851e-06, "loss": 0.0519, "step": 178340 }, { "epoch": 6.029615605666182, "grad_norm": 0.6900461316108704, "learning_rate": 3.1698979904664526e-06, "loss": 0.0601, "step": 178350 }, { "epoch": 6.029953683356435, "grad_norm": 0.8628765344619751, "learning_rate": 3.16700179329944e-06, "loss": 0.0479, "step": 178360 }, { "epoch": 6.030291761046689, "grad_norm": 0.9421212077140808, "learning_rate": 3.1641068765290105e-06, "loss": 0.0585, "step": 178370 }, { "epoch": 6.0306298387369415, "grad_norm": 0.5536237955093384, "learning_rate": 3.1612132402343063e-06, "loss": 0.0491, "step": 178380 }, { "epoch": 6.030967916427195, "grad_norm": 0.8132035136222839, "learning_rate": 3.1583208844944257e-06, "loss": 0.0524, "step": 178390 }, { "epoch": 6.0313059941174485, "grad_norm": 1.2329976558685303, "learning_rate": 3.1554298093884602e-06, "loss": 0.0601, "step": 178400 }, { "epoch": 6.031644071807701, "grad_norm": 0.8507082462310791, "learning_rate": 3.1525400149954475e-06, "loss": 0.0539, "step": 178410 }, { "epoch": 6.031982149497955, "grad_norm": 0.948523759841919, "learning_rate": 3.1496515013943906e-06, "loss": 0.0603, "step": 178420 }, { "epoch": 6.032320227188208, "grad_norm": 0.9673991799354553, "learning_rate": 3.146764268664254e-06, "loss": 0.0603, "step": 178430 }, { "epoch": 6.032658304878461, "grad_norm": 0.44959402084350586, "learning_rate": 3.1438783168839803e-06, "loss": 0.054, "step": 178440 }, { "epoch": 6.032996382568714, "grad_norm": 0.6491873264312744, "learning_rate": 3.1409936461324786e-06, "loss": 0.0484, "step": 178450 }, { "epoch": 6.033334460258968, "grad_norm": 1.5861867666244507, "learning_rate": 3.138110256488608e-06, "loss": 0.0547, "step": 178460 }, { "epoch": 6.03367253794922, "grad_norm": 1.244141936302185, "learning_rate": 3.1352281480311942e-06, "loss": 0.0517, "step": 178470 }, { "epoch": 6.034010615639474, "grad_norm": 0.728415310382843, "learning_rate": 3.1323473208390295e-06, "loss": 0.0577, "step": 178480 }, { "epoch": 6.0343486933297275, "grad_norm": 0.7793962955474854, "learning_rate": 3.1294677749908905e-06, "loss": 0.0568, "step": 178490 }, { "epoch": 6.03468677101998, "grad_norm": 1.3360543251037598, "learning_rate": 3.1265895105654975e-06, "loss": 0.0552, "step": 178500 }, { "epoch": 6.035024848710234, "grad_norm": 1.014633297920227, "learning_rate": 3.123712527641526e-06, "loss": 0.0574, "step": 178510 }, { "epoch": 6.035362926400487, "grad_norm": 1.4506789445877075, "learning_rate": 3.120836826297652e-06, "loss": 0.0519, "step": 178520 }, { "epoch": 6.03570100409074, "grad_norm": 2.010566473007202, "learning_rate": 3.117962406612479e-06, "loss": 0.0494, "step": 178530 }, { "epoch": 6.036039081780993, "grad_norm": 0.9144390821456909, "learning_rate": 3.1150892686646173e-06, "loss": 0.0526, "step": 178540 }, { "epoch": 6.036377159471247, "grad_norm": 1.4981215000152588, "learning_rate": 3.112217412532581e-06, "loss": 0.061, "step": 178550 }, { "epoch": 6.036715237161499, "grad_norm": 0.8681074380874634, "learning_rate": 3.109346838294913e-06, "loss": 0.0485, "step": 178560 }, { "epoch": 6.037053314851753, "grad_norm": 1.2522622346878052, "learning_rate": 3.10647754603009e-06, "loss": 0.0518, "step": 178570 }, { "epoch": 6.037391392542006, "grad_norm": 0.9860970377922058, "learning_rate": 3.103609535816554e-06, "loss": 0.0518, "step": 178580 }, { "epoch": 6.037729470232259, "grad_norm": 0.679002046585083, "learning_rate": 3.100742807732715e-06, "loss": 0.0515, "step": 178590 }, { "epoch": 6.0380675479225125, "grad_norm": 0.9941109418869019, "learning_rate": 3.0978773618569436e-06, "loss": 0.0525, "step": 178600 }, { "epoch": 6.038405625612766, "grad_norm": 0.9601516723632812, "learning_rate": 3.095013198267593e-06, "loss": 0.0618, "step": 178610 }, { "epoch": 6.038743703303019, "grad_norm": 1.1056287288665771, "learning_rate": 3.0921503170429576e-06, "loss": 0.0492, "step": 178620 }, { "epoch": 6.039081780993272, "grad_norm": 0.7540740966796875, "learning_rate": 3.0892887182613018e-06, "loss": 0.051, "step": 178630 }, { "epoch": 6.039419858683526, "grad_norm": 0.552236020565033, "learning_rate": 3.086428402000874e-06, "loss": 0.0518, "step": 178640 }, { "epoch": 6.039757936373778, "grad_norm": 1.3476396799087524, "learning_rate": 3.083569368339867e-06, "loss": 0.0554, "step": 178650 }, { "epoch": 6.040096014064032, "grad_norm": 0.5010314583778381, "learning_rate": 3.080711617356452e-06, "loss": 0.0524, "step": 178660 }, { "epoch": 6.040434091754285, "grad_norm": 5.986785411834717, "learning_rate": 3.0778551491287554e-06, "loss": 0.0543, "step": 178670 }, { "epoch": 6.040772169444538, "grad_norm": 2.1258199214935303, "learning_rate": 3.0749999637348648e-06, "loss": 0.0596, "step": 178680 }, { "epoch": 6.0411102471347915, "grad_norm": 1.0118902921676636, "learning_rate": 3.072146061252856e-06, "loss": 0.0556, "step": 178690 }, { "epoch": 6.041448324825045, "grad_norm": 2.0317533016204834, "learning_rate": 3.06929344176074e-06, "loss": 0.0502, "step": 178700 }, { "epoch": 6.041786402515298, "grad_norm": 6.753607749938965, "learning_rate": 3.066442105336509e-06, "loss": 0.0585, "step": 178710 }, { "epoch": 6.042124480205551, "grad_norm": 1.2225497961044312, "learning_rate": 3.0635920520581117e-06, "loss": 0.0504, "step": 178720 }, { "epoch": 6.042462557895805, "grad_norm": 0.9212021231651306, "learning_rate": 3.0607432820034867e-06, "loss": 0.0562, "step": 178730 }, { "epoch": 6.042800635586057, "grad_norm": 0.859891951084137, "learning_rate": 3.0578957952504982e-06, "loss": 0.0552, "step": 178740 }, { "epoch": 6.043138713276311, "grad_norm": 1.4094488620758057, "learning_rate": 3.0550495918770073e-06, "loss": 0.0547, "step": 178750 }, { "epoch": 6.043476790966564, "grad_norm": 1.449729561805725, "learning_rate": 3.0522046719608232e-06, "loss": 0.0519, "step": 178760 }, { "epoch": 6.043814868656817, "grad_norm": 0.7700001001358032, "learning_rate": 3.049361035579723e-06, "loss": 0.0483, "step": 178770 }, { "epoch": 6.0441529463470705, "grad_norm": 0.7067578434944153, "learning_rate": 3.0465186828114556e-06, "loss": 0.0529, "step": 178780 }, { "epoch": 6.044491024037324, "grad_norm": 2.7661917209625244, "learning_rate": 3.043677613733731e-06, "loss": 0.06, "step": 178790 }, { "epoch": 6.044829101727577, "grad_norm": 1.3928117752075195, "learning_rate": 3.04083782842422e-06, "loss": 0.0575, "step": 178800 }, { "epoch": 6.04516717941783, "grad_norm": 0.6333429217338562, "learning_rate": 3.03799932696055e-06, "loss": 0.0505, "step": 178810 }, { "epoch": 6.045505257108084, "grad_norm": 0.44583678245544434, "learning_rate": 3.035162109420342e-06, "loss": 0.044, "step": 178820 }, { "epoch": 6.045843334798336, "grad_norm": 1.0460882186889648, "learning_rate": 3.032326175881167e-06, "loss": 0.052, "step": 178830 }, { "epoch": 6.04618141248859, "grad_norm": 1.2394890785217285, "learning_rate": 3.0294915264205416e-06, "loss": 0.0514, "step": 178840 }, { "epoch": 6.046519490178843, "grad_norm": 2.069431781768799, "learning_rate": 3.0266581611159693e-06, "loss": 0.0576, "step": 178850 }, { "epoch": 6.046857567869096, "grad_norm": 0.8604061007499695, "learning_rate": 3.023826080044917e-06, "loss": 0.0472, "step": 178860 }, { "epoch": 6.047195645559349, "grad_norm": 0.8324203491210938, "learning_rate": 3.0209952832848164e-06, "loss": 0.0554, "step": 178870 }, { "epoch": 6.047533723249603, "grad_norm": 0.9602800011634827, "learning_rate": 3.018165770913056e-06, "loss": 0.0508, "step": 178880 }, { "epoch": 6.0478718009398555, "grad_norm": 0.8881412744522095, "learning_rate": 3.015337543006985e-06, "loss": 0.0547, "step": 178890 }, { "epoch": 6.048209878630109, "grad_norm": 0.6796255707740784, "learning_rate": 3.012510599643936e-06, "loss": 0.0498, "step": 178900 }, { "epoch": 6.048547956320363, "grad_norm": 1.5376228094100952, "learning_rate": 3.009684940901203e-06, "loss": 0.0535, "step": 178910 }, { "epoch": 6.048886034010616, "grad_norm": 0.8985132575035095, "learning_rate": 3.006860566856029e-06, "loss": 0.053, "step": 178920 }, { "epoch": 6.049224111700869, "grad_norm": 3.699305772781372, "learning_rate": 3.004037477585625e-06, "loss": 0.0467, "step": 178930 }, { "epoch": 6.049562189391122, "grad_norm": 1.2918304204940796, "learning_rate": 3.001215673167179e-06, "loss": 0.05, "step": 178940 }, { "epoch": 6.049900267081376, "grad_norm": 1.4417152404785156, "learning_rate": 2.9983951536778466e-06, "loss": 0.0601, "step": 178950 }, { "epoch": 6.050238344771628, "grad_norm": 2.266017436981201, "learning_rate": 2.9955759191947318e-06, "loss": 0.0571, "step": 178960 }, { "epoch": 6.050576422461882, "grad_norm": 1.518209457397461, "learning_rate": 2.9927579697949182e-06, "loss": 0.0539, "step": 178970 }, { "epoch": 6.050914500152135, "grad_norm": 0.7190278172492981, "learning_rate": 2.989941305555427e-06, "loss": 0.0509, "step": 178980 }, { "epoch": 6.051252577842388, "grad_norm": 0.7923188805580139, "learning_rate": 2.9871259265532912e-06, "loss": 0.057, "step": 178990 }, { "epoch": 6.0515906555326415, "grad_norm": 0.9207851886749268, "learning_rate": 2.9843118328654717e-06, "loss": 0.048, "step": 179000 }, { "epoch": 6.051928733222895, "grad_norm": 1.0774075984954834, "learning_rate": 2.9814990245688957e-06, "loss": 0.0537, "step": 179010 }, { "epoch": 6.052266810913148, "grad_norm": 0.8028969168663025, "learning_rate": 2.978687501740479e-06, "loss": 0.0561, "step": 179020 }, { "epoch": 6.052604888603401, "grad_norm": 0.9144867658615112, "learning_rate": 2.9758772644570775e-06, "loss": 0.0522, "step": 179030 }, { "epoch": 6.052942966293655, "grad_norm": 2.305222272872925, "learning_rate": 2.973068312795535e-06, "loss": 0.0496, "step": 179040 }, { "epoch": 6.053281043983907, "grad_norm": 0.6984835267066956, "learning_rate": 2.970260646832623e-06, "loss": 0.0548, "step": 179050 }, { "epoch": 6.053619121674161, "grad_norm": 1.0023685693740845, "learning_rate": 2.96745426664512e-06, "loss": 0.0508, "step": 179060 }, { "epoch": 6.053957199364414, "grad_norm": 1.6054987907409668, "learning_rate": 2.9646491723097524e-06, "loss": 0.056, "step": 179070 }, { "epoch": 6.054295277054667, "grad_norm": 0.8072948455810547, "learning_rate": 2.9618453639032095e-06, "loss": 0.0592, "step": 179080 }, { "epoch": 6.0546333547449205, "grad_norm": 1.770811676979065, "learning_rate": 2.959042841502141e-06, "loss": 0.0473, "step": 179090 }, { "epoch": 6.054971432435174, "grad_norm": 7.620509624481201, "learning_rate": 2.9562416051831575e-06, "loss": 0.0575, "step": 179100 }, { "epoch": 6.055309510125427, "grad_norm": 1.7304853200912476, "learning_rate": 2.9534416550228707e-06, "loss": 0.0566, "step": 179110 }, { "epoch": 6.05564758781568, "grad_norm": 1.429223656654358, "learning_rate": 2.9506429910978073e-06, "loss": 0.0569, "step": 179120 }, { "epoch": 6.055985665505934, "grad_norm": 0.8860852122306824, "learning_rate": 2.9478456134844844e-06, "loss": 0.05, "step": 179130 }, { "epoch": 6.056323743196186, "grad_norm": 0.6366641521453857, "learning_rate": 2.945049522259391e-06, "loss": 0.0516, "step": 179140 }, { "epoch": 6.05666182088644, "grad_norm": 0.7981538772583008, "learning_rate": 2.94225471749896e-06, "loss": 0.054, "step": 179150 }, { "epoch": 6.056999898576693, "grad_norm": 1.42210054397583, "learning_rate": 2.939461199279614e-06, "loss": 0.0558, "step": 179160 }, { "epoch": 6.057337976266946, "grad_norm": 0.6258522868156433, "learning_rate": 2.936668967677719e-06, "loss": 0.0544, "step": 179170 }, { "epoch": 6.057676053957199, "grad_norm": 0.5897513031959534, "learning_rate": 2.933878022769615e-06, "loss": 0.0503, "step": 179180 }, { "epoch": 6.058014131647453, "grad_norm": 1.0700592994689941, "learning_rate": 2.9310883646315955e-06, "loss": 0.0518, "step": 179190 }, { "epoch": 6.058352209337706, "grad_norm": 1.0319291353225708, "learning_rate": 2.9282999933399447e-06, "loss": 0.0533, "step": 179200 }, { "epoch": 6.058690287027959, "grad_norm": 1.28691828250885, "learning_rate": 2.9255129089708844e-06, "loss": 0.0477, "step": 179210 }, { "epoch": 6.059028364718213, "grad_norm": 1.2840173244476318, "learning_rate": 2.922727111600615e-06, "loss": 0.058, "step": 179220 }, { "epoch": 6.059366442408465, "grad_norm": 1.1608105897903442, "learning_rate": 2.9199426013053033e-06, "loss": 0.0499, "step": 179230 }, { "epoch": 6.059704520098719, "grad_norm": 0.5166975259780884, "learning_rate": 2.9171593781610663e-06, "loss": 0.0525, "step": 179240 }, { "epoch": 6.060042597788972, "grad_norm": 2.265268564224243, "learning_rate": 2.914377442244015e-06, "loss": 0.0554, "step": 179250 }, { "epoch": 6.060380675479225, "grad_norm": 1.4019956588745117, "learning_rate": 2.911596793630189e-06, "loss": 0.0584, "step": 179260 }, { "epoch": 6.060718753169478, "grad_norm": 0.43241086602211, "learning_rate": 2.908817432395616e-06, "loss": 0.051, "step": 179270 }, { "epoch": 6.061056830859732, "grad_norm": 0.9690984487533569, "learning_rate": 2.906039358616286e-06, "loss": 0.0526, "step": 179280 }, { "epoch": 6.0613949085499845, "grad_norm": 2.566405773162842, "learning_rate": 2.9032625723681484e-06, "loss": 0.052, "step": 179290 }, { "epoch": 6.061732986240238, "grad_norm": 1.6760808229446411, "learning_rate": 2.9004870737271206e-06, "loss": 0.0535, "step": 179300 }, { "epoch": 6.0620710639304916, "grad_norm": 0.6252039074897766, "learning_rate": 2.8977128627690696e-06, "loss": 0.0522, "step": 179310 }, { "epoch": 6.062409141620744, "grad_norm": 0.9166063070297241, "learning_rate": 2.8949399395698572e-06, "loss": 0.0483, "step": 179320 }, { "epoch": 6.062747219310998, "grad_norm": 1.0757009983062744, "learning_rate": 2.8921683042053007e-06, "loss": 0.0525, "step": 179330 }, { "epoch": 6.063085297001251, "grad_norm": 0.9171271920204163, "learning_rate": 2.889397956751155e-06, "loss": 0.0567, "step": 179340 }, { "epoch": 6.063423374691504, "grad_norm": 0.5983816385269165, "learning_rate": 2.886628897283178e-06, "loss": 0.0537, "step": 179350 }, { "epoch": 6.063761452381757, "grad_norm": 1.6147890090942383, "learning_rate": 2.8838611258770577e-06, "loss": 0.0514, "step": 179360 }, { "epoch": 6.064099530072011, "grad_norm": 1.3467656373977661, "learning_rate": 2.881094642608484e-06, "loss": 0.0595, "step": 179370 }, { "epoch": 6.0644376077622635, "grad_norm": 1.851123571395874, "learning_rate": 2.8783294475530797e-06, "loss": 0.052, "step": 179380 }, { "epoch": 6.064775685452517, "grad_norm": 0.9818781614303589, "learning_rate": 2.8755655407864345e-06, "loss": 0.0528, "step": 179390 }, { "epoch": 6.0651137631427705, "grad_norm": 1.2004215717315674, "learning_rate": 2.8728029223841325e-06, "loss": 0.0538, "step": 179400 }, { "epoch": 6.065451840833023, "grad_norm": 0.9094123840332031, "learning_rate": 2.870041592421696e-06, "loss": 0.053, "step": 179410 }, { "epoch": 6.065789918523277, "grad_norm": 0.9730353951454163, "learning_rate": 2.8672815509746154e-06, "loss": 0.0544, "step": 179420 }, { "epoch": 6.06612799621353, "grad_norm": 1.4987707138061523, "learning_rate": 2.864522798118341e-06, "loss": 0.0493, "step": 179430 }, { "epoch": 6.066466073903783, "grad_norm": 0.8583269119262695, "learning_rate": 2.861765333928318e-06, "loss": 0.0542, "step": 179440 }, { "epoch": 6.066804151594036, "grad_norm": 4.575313568115234, "learning_rate": 2.859009158479908e-06, "loss": 0.058, "step": 179450 }, { "epoch": 6.06714222928429, "grad_norm": 5.261446475982666, "learning_rate": 2.856254271848491e-06, "loss": 0.0541, "step": 179460 }, { "epoch": 6.067480306974542, "grad_norm": 1.0220386981964111, "learning_rate": 2.8535006741093663e-06, "loss": 0.0608, "step": 179470 }, { "epoch": 6.067818384664796, "grad_norm": 1.0099186897277832, "learning_rate": 2.850748365337819e-06, "loss": 0.0497, "step": 179480 }, { "epoch": 6.0681564623550495, "grad_norm": 1.4945263862609863, "learning_rate": 2.847997345609099e-06, "loss": 0.0563, "step": 179490 }, { "epoch": 6.068494540045302, "grad_norm": 0.7844577431678772, "learning_rate": 2.8452476149984198e-06, "loss": 0.0526, "step": 179500 }, { "epoch": 6.068832617735556, "grad_norm": 1.825154185295105, "learning_rate": 2.842499173580948e-06, "loss": 0.057, "step": 179510 }, { "epoch": 6.069170695425809, "grad_norm": 1.0337709188461304, "learning_rate": 2.8397520214318405e-06, "loss": 0.0497, "step": 179520 }, { "epoch": 6.069508773116062, "grad_norm": 1.4769037961959839, "learning_rate": 2.837006158626193e-06, "loss": 0.0514, "step": 179530 }, { "epoch": 6.069846850806315, "grad_norm": 1.2683650255203247, "learning_rate": 2.834261585239084e-06, "loss": 0.0518, "step": 179540 }, { "epoch": 6.070184928496569, "grad_norm": 0.9741569757461548, "learning_rate": 2.831518301345537e-06, "loss": 0.0512, "step": 179550 }, { "epoch": 6.070523006186821, "grad_norm": 1.283935785293579, "learning_rate": 2.8287763070205643e-06, "loss": 0.054, "step": 179560 }, { "epoch": 6.070861083877075, "grad_norm": 6.36748743057251, "learning_rate": 2.8260356023391167e-06, "loss": 0.0569, "step": 179570 }, { "epoch": 6.071199161567328, "grad_norm": 0.6821145415306091, "learning_rate": 2.8232961873761398e-06, "loss": 0.0522, "step": 179580 }, { "epoch": 6.071537239257581, "grad_norm": 1.2804776430130005, "learning_rate": 2.8205580622065243e-06, "loss": 0.0511, "step": 179590 }, { "epoch": 6.0718753169478346, "grad_norm": 0.8923646807670593, "learning_rate": 2.817821226905121e-06, "loss": 0.0491, "step": 179600 }, { "epoch": 6.072213394638088, "grad_norm": 1.1288526058197021, "learning_rate": 2.815085681546764e-06, "loss": 0.0521, "step": 179610 }, { "epoch": 6.072551472328341, "grad_norm": 0.9961531162261963, "learning_rate": 2.812351426206239e-06, "loss": 0.0546, "step": 179620 }, { "epoch": 6.072889550018594, "grad_norm": 1.4326972961425781, "learning_rate": 2.809618460958302e-06, "loss": 0.0539, "step": 179630 }, { "epoch": 6.073227627708848, "grad_norm": 1.5274769067764282, "learning_rate": 2.806886785877655e-06, "loss": 0.0546, "step": 179640 }, { "epoch": 6.0735657053991, "grad_norm": 0.8805164694786072, "learning_rate": 2.804156401038993e-06, "loss": 0.0559, "step": 179650 }, { "epoch": 6.073903783089354, "grad_norm": 1.8564183712005615, "learning_rate": 2.8014273065169795e-06, "loss": 0.0512, "step": 179660 }, { "epoch": 6.074241860779607, "grad_norm": 0.47049665451049805, "learning_rate": 2.798699502386204e-06, "loss": 0.0482, "step": 179670 }, { "epoch": 6.07457993846986, "grad_norm": 0.9891312718391418, "learning_rate": 2.7959729887212518e-06, "loss": 0.0512, "step": 179680 }, { "epoch": 6.0749180161601135, "grad_norm": 2.063007116317749, "learning_rate": 2.793247765596663e-06, "loss": 0.0465, "step": 179690 }, { "epoch": 6.075256093850367, "grad_norm": 0.4836382567882538, "learning_rate": 2.7905238330869443e-06, "loss": 0.0499, "step": 179700 }, { "epoch": 6.07559417154062, "grad_norm": 3.2539725303649902, "learning_rate": 2.7878011912665757e-06, "loss": 0.0528, "step": 179710 }, { "epoch": 6.075932249230873, "grad_norm": 0.8014851212501526, "learning_rate": 2.785079840209975e-06, "loss": 0.057, "step": 179720 }, { "epoch": 6.076270326921127, "grad_norm": 1.2669163942337036, "learning_rate": 2.7823597799915615e-06, "loss": 0.0514, "step": 179730 }, { "epoch": 6.076608404611379, "grad_norm": 1.215580940246582, "learning_rate": 2.779641010685685e-06, "loss": 0.051, "step": 179740 }, { "epoch": 6.076946482301633, "grad_norm": 1.8084938526153564, "learning_rate": 2.7769235323666986e-06, "loss": 0.0573, "step": 179750 }, { "epoch": 6.077284559991886, "grad_norm": 0.652382493019104, "learning_rate": 2.7742073451088703e-06, "loss": 0.0493, "step": 179760 }, { "epoch": 6.077622637682139, "grad_norm": 0.6212230920791626, "learning_rate": 2.771492448986468e-06, "loss": 0.0499, "step": 179770 }, { "epoch": 6.0779607153723925, "grad_norm": 1.8570263385772705, "learning_rate": 2.7687788440737274e-06, "loss": 0.0514, "step": 179780 }, { "epoch": 6.078298793062646, "grad_norm": 4.845217704772949, "learning_rate": 2.7660665304448274e-06, "loss": 0.0542, "step": 179790 }, { "epoch": 6.078636870752899, "grad_norm": 1.26335608959198, "learning_rate": 2.763355508173926e-06, "loss": 0.0546, "step": 179800 }, { "epoch": 6.078974948443152, "grad_norm": 0.854367196559906, "learning_rate": 2.76064577733513e-06, "loss": 0.0525, "step": 179810 }, { "epoch": 6.079313026133406, "grad_norm": 0.5549156665802002, "learning_rate": 2.7579373380025365e-06, "loss": 0.0522, "step": 179820 }, { "epoch": 6.079651103823658, "grad_norm": 0.9509512186050415, "learning_rate": 2.755230190250191e-06, "loss": 0.0534, "step": 179830 }, { "epoch": 6.079989181513912, "grad_norm": 2.4324841499328613, "learning_rate": 2.7525243341520956e-06, "loss": 0.0564, "step": 179840 }, { "epoch": 6.080327259204165, "grad_norm": 0.9079005122184753, "learning_rate": 2.7498197697822412e-06, "loss": 0.0595, "step": 179850 }, { "epoch": 6.080665336894418, "grad_norm": 1.3757994174957275, "learning_rate": 2.747116497214558e-06, "loss": 0.0549, "step": 179860 }, { "epoch": 6.081003414584671, "grad_norm": 0.5688043832778931, "learning_rate": 2.744414516522964e-06, "loss": 0.0498, "step": 179870 }, { "epoch": 6.081341492274925, "grad_norm": 1.0259978771209717, "learning_rate": 2.7417138277813225e-06, "loss": 0.056, "step": 179880 }, { "epoch": 6.0816795699651784, "grad_norm": 2.064326524734497, "learning_rate": 2.739014431063469e-06, "loss": 0.0555, "step": 179890 }, { "epoch": 6.082017647655431, "grad_norm": 1.0036871433258057, "learning_rate": 2.736316326443206e-06, "loss": 0.0487, "step": 179900 }, { "epoch": 6.082355725345685, "grad_norm": 0.5526680946350098, "learning_rate": 2.733619513994301e-06, "loss": 0.0572, "step": 179910 }, { "epoch": 6.082693803035937, "grad_norm": 0.7095121145248413, "learning_rate": 2.730923993790485e-06, "loss": 0.0536, "step": 179920 }, { "epoch": 6.083031880726191, "grad_norm": 0.8095587491989136, "learning_rate": 2.728229765905438e-06, "loss": 0.0527, "step": 179930 }, { "epoch": 6.083369958416444, "grad_norm": 0.7082990407943726, "learning_rate": 2.725536830412839e-06, "loss": 0.0537, "step": 179940 }, { "epoch": 6.083708036106698, "grad_norm": 1.901552677154541, "learning_rate": 2.7228451873862916e-06, "loss": 0.0497, "step": 179950 }, { "epoch": 6.08404611379695, "grad_norm": 0.8639147877693176, "learning_rate": 2.720154836899408e-06, "loss": 0.0518, "step": 179960 }, { "epoch": 6.084384191487204, "grad_norm": 0.9416313171386719, "learning_rate": 2.7174657790257243e-06, "loss": 0.0535, "step": 179970 }, { "epoch": 6.084722269177457, "grad_norm": 1.3403651714324951, "learning_rate": 2.714778013838759e-06, "loss": 0.055, "step": 179980 }, { "epoch": 6.08506034686771, "grad_norm": 0.7038953304290771, "learning_rate": 2.712091541412004e-06, "loss": 0.0455, "step": 179990 }, { "epoch": 6.0853984245579635, "grad_norm": 1.5922603607177734, "learning_rate": 2.7094063618189004e-06, "loss": 0.0504, "step": 180000 }, { "epoch": 6.085736502248217, "grad_norm": 0.7841039299964905, "learning_rate": 2.7067224751328666e-06, "loss": 0.0527, "step": 180010 }, { "epoch": 6.08607457993847, "grad_norm": 1.184516429901123, "learning_rate": 2.7040398814272616e-06, "loss": 0.0526, "step": 180020 }, { "epoch": 6.086412657628723, "grad_norm": 0.6835324168205261, "learning_rate": 2.7013585807754373e-06, "loss": 0.0487, "step": 180030 }, { "epoch": 6.086750735318977, "grad_norm": 0.7408879399299622, "learning_rate": 2.6986785732507126e-06, "loss": 0.0493, "step": 180040 }, { "epoch": 6.087088813009229, "grad_norm": 1.17965567111969, "learning_rate": 2.6959998589263345e-06, "loss": 0.0531, "step": 180050 }, { "epoch": 6.087426890699483, "grad_norm": 1.3111064434051514, "learning_rate": 2.693322437875556e-06, "loss": 0.0505, "step": 180060 }, { "epoch": 6.087764968389736, "grad_norm": 2.863312005996704, "learning_rate": 2.6906463101715574e-06, "loss": 0.0561, "step": 180070 }, { "epoch": 6.088103046079989, "grad_norm": 1.2024294137954712, "learning_rate": 2.6879714758875295e-06, "loss": 0.0535, "step": 180080 }, { "epoch": 6.0884411237702425, "grad_norm": 0.5102964639663696, "learning_rate": 2.6852979350965814e-06, "loss": 0.0513, "step": 180090 }, { "epoch": 6.088779201460496, "grad_norm": 0.7138341069221497, "learning_rate": 2.682625687871809e-06, "loss": 0.0503, "step": 180100 }, { "epoch": 6.089117279150749, "grad_norm": 0.8662034869194031, "learning_rate": 2.679954734286283e-06, "loss": 0.0526, "step": 180110 }, { "epoch": 6.089455356841002, "grad_norm": 0.8265975713729858, "learning_rate": 2.6772850744130163e-06, "loss": 0.0522, "step": 180120 }, { "epoch": 6.089793434531256, "grad_norm": 1.35858952999115, "learning_rate": 2.674616708324995e-06, "loss": 0.0563, "step": 180130 }, { "epoch": 6.090131512221508, "grad_norm": 1.1152005195617676, "learning_rate": 2.671949636095167e-06, "loss": 0.0508, "step": 180140 }, { "epoch": 6.090469589911762, "grad_norm": 2.1810944080352783, "learning_rate": 2.6692838577964564e-06, "loss": 0.0638, "step": 180150 }, { "epoch": 6.090807667602015, "grad_norm": 0.911478579044342, "learning_rate": 2.6666193735017496e-06, "loss": 0.0594, "step": 180160 }, { "epoch": 6.091145745292268, "grad_norm": 0.9392303228378296, "learning_rate": 2.6639561832838833e-06, "loss": 0.0506, "step": 180170 }, { "epoch": 6.0914838229825214, "grad_norm": 1.5059863328933716, "learning_rate": 2.6612942872156766e-06, "loss": 0.0536, "step": 180180 }, { "epoch": 6.091821900672775, "grad_norm": 2.0189614295959473, "learning_rate": 2.6586336853698933e-06, "loss": 0.0584, "step": 180190 }, { "epoch": 6.092159978363028, "grad_norm": 0.6495495438575745, "learning_rate": 2.655974377819287e-06, "loss": 0.0595, "step": 180200 }, { "epoch": 6.092498056053281, "grad_norm": 2.230443239212036, "learning_rate": 2.6533163646365543e-06, "loss": 0.0505, "step": 180210 }, { "epoch": 6.092836133743535, "grad_norm": 0.6396477222442627, "learning_rate": 2.6506596458943545e-06, "loss": 0.0563, "step": 180220 }, { "epoch": 6.093174211433787, "grad_norm": 0.6060940623283386, "learning_rate": 2.64800422166534e-06, "loss": 0.0547, "step": 180230 }, { "epoch": 6.093512289124041, "grad_norm": 1.643263816833496, "learning_rate": 2.6453500920220973e-06, "loss": 0.0584, "step": 180240 }, { "epoch": 6.093850366814294, "grad_norm": 1.4402483701705933, "learning_rate": 2.6426972570372023e-06, "loss": 0.0508, "step": 180250 }, { "epoch": 6.094188444504547, "grad_norm": 1.0590085983276367, "learning_rate": 2.640045716783163e-06, "loss": 0.0545, "step": 180260 }, { "epoch": 6.0945265221948, "grad_norm": 1.736641526222229, "learning_rate": 2.6373954713324767e-06, "loss": 0.0528, "step": 180270 }, { "epoch": 6.094864599885054, "grad_norm": 0.7700615525245667, "learning_rate": 2.634746520757614e-06, "loss": 0.0486, "step": 180280 }, { "epoch": 6.0952026775753065, "grad_norm": 0.7126889228820801, "learning_rate": 2.6320988651309888e-06, "loss": 0.0503, "step": 180290 }, { "epoch": 6.09554075526556, "grad_norm": 0.8542036414146423, "learning_rate": 2.6294525045249873e-06, "loss": 0.0569, "step": 180300 }, { "epoch": 6.095878832955814, "grad_norm": 1.1833171844482422, "learning_rate": 2.626807439011947e-06, "loss": 0.0592, "step": 180310 }, { "epoch": 6.096216910646066, "grad_norm": 0.9843077659606934, "learning_rate": 2.6241636686642036e-06, "loss": 0.0534, "step": 180320 }, { "epoch": 6.09655498833632, "grad_norm": 2.20038104057312, "learning_rate": 2.6215211935540217e-06, "loss": 0.0583, "step": 180330 }, { "epoch": 6.096893066026573, "grad_norm": 0.7672187089920044, "learning_rate": 2.6188800137536495e-06, "loss": 0.0512, "step": 180340 }, { "epoch": 6.097231143716826, "grad_norm": 2.2825331687927246, "learning_rate": 2.616240129335301e-06, "loss": 0.0496, "step": 180350 }, { "epoch": 6.097569221407079, "grad_norm": 1.5974658727645874, "learning_rate": 2.613601540371141e-06, "loss": 0.0515, "step": 180360 }, { "epoch": 6.097907299097333, "grad_norm": 0.5876410603523254, "learning_rate": 2.610964246933323e-06, "loss": 0.0475, "step": 180370 }, { "epoch": 6.0982453767875855, "grad_norm": 1.2868051528930664, "learning_rate": 2.608328249093933e-06, "loss": 0.0555, "step": 180380 }, { "epoch": 6.098583454477839, "grad_norm": 1.2876514196395874, "learning_rate": 2.6056935469250476e-06, "loss": 0.0505, "step": 180390 }, { "epoch": 6.0989215321680925, "grad_norm": 1.6281230449676514, "learning_rate": 2.603060140498692e-06, "loss": 0.0591, "step": 180400 }, { "epoch": 6.099259609858345, "grad_norm": 1.115909457206726, "learning_rate": 2.6004280298868643e-06, "loss": 0.0526, "step": 180410 }, { "epoch": 6.099597687548599, "grad_norm": 0.9427900910377502, "learning_rate": 2.5977972151615348e-06, "loss": 0.0483, "step": 180420 }, { "epoch": 6.099935765238852, "grad_norm": 1.8944718837738037, "learning_rate": 2.5951676963946126e-06, "loss": 0.0547, "step": 180430 }, { "epoch": 6.100273842929105, "grad_norm": 1.103496789932251, "learning_rate": 2.592539473658001e-06, "loss": 0.0588, "step": 180440 }, { "epoch": 6.100611920619358, "grad_norm": 1.1903305053710938, "learning_rate": 2.589912547023543e-06, "loss": 0.0493, "step": 180450 }, { "epoch": 6.100949998309612, "grad_norm": 1.3066754341125488, "learning_rate": 2.5872869165630755e-06, "loss": 0.0543, "step": 180460 }, { "epoch": 6.1012880759998644, "grad_norm": 0.9040312170982361, "learning_rate": 2.584662582348368e-06, "loss": 0.0506, "step": 180470 }, { "epoch": 6.101626153690118, "grad_norm": 0.6441068053245544, "learning_rate": 2.5820395444511648e-06, "loss": 0.0566, "step": 180480 }, { "epoch": 6.1019642313803715, "grad_norm": 0.9231178164482117, "learning_rate": 2.579417802943196e-06, "loss": 0.0565, "step": 180490 }, { "epoch": 6.102302309070624, "grad_norm": 0.7715988755226135, "learning_rate": 2.5767973578961326e-06, "loss": 0.0571, "step": 180500 }, { "epoch": 6.102640386760878, "grad_norm": 0.7265868186950684, "learning_rate": 2.5741782093816056e-06, "loss": 0.0497, "step": 180510 }, { "epoch": 6.102978464451131, "grad_norm": 7.121551036834717, "learning_rate": 2.5715603574712253e-06, "loss": 0.0517, "step": 180520 }, { "epoch": 6.103316542141384, "grad_norm": 1.5355050563812256, "learning_rate": 2.5689438022365674e-06, "loss": 0.0501, "step": 180530 }, { "epoch": 6.103654619831637, "grad_norm": 1.0735704898834229, "learning_rate": 2.56632854374918e-06, "loss": 0.0524, "step": 180540 }, { "epoch": 6.103992697521891, "grad_norm": 0.5644404888153076, "learning_rate": 2.5637145820805396e-06, "loss": 0.0531, "step": 180550 }, { "epoch": 6.104330775212143, "grad_norm": 0.5700413584709167, "learning_rate": 2.5611019173021278e-06, "loss": 0.0499, "step": 180560 }, { "epoch": 6.104668852902397, "grad_norm": 0.7977573871612549, "learning_rate": 2.5584905494853597e-06, "loss": 0.0548, "step": 180570 }, { "epoch": 6.10500693059265, "grad_norm": 1.0905914306640625, "learning_rate": 2.555880478701639e-06, "loss": 0.0501, "step": 180580 }, { "epoch": 6.105345008282903, "grad_norm": 2.165210723876953, "learning_rate": 2.553271705022331e-06, "loss": 0.0542, "step": 180590 }, { "epoch": 6.105683085973157, "grad_norm": 1.0309590101242065, "learning_rate": 2.5506642285187342e-06, "loss": 0.0539, "step": 180600 }, { "epoch": 6.10602116366341, "grad_norm": 1.288831114768982, "learning_rate": 2.5480580492621643e-06, "loss": 0.0501, "step": 180610 }, { "epoch": 6.106359241353663, "grad_norm": 1.6745699644088745, "learning_rate": 2.545453167323858e-06, "loss": 0.0527, "step": 180620 }, { "epoch": 6.106697319043916, "grad_norm": 1.8890273571014404, "learning_rate": 2.5428495827750365e-06, "loss": 0.0505, "step": 180630 }, { "epoch": 6.10703539673417, "grad_norm": 0.6903654336929321, "learning_rate": 2.5402472956868704e-06, "loss": 0.0501, "step": 180640 }, { "epoch": 6.107373474424422, "grad_norm": 0.8456857800483704, "learning_rate": 2.53764630613052e-06, "loss": 0.0552, "step": 180650 }, { "epoch": 6.107711552114676, "grad_norm": 0.6535930037498474, "learning_rate": 2.535046614177078e-06, "loss": 0.0486, "step": 180660 }, { "epoch": 6.108049629804929, "grad_norm": 0.9014965891838074, "learning_rate": 2.532448219897643e-06, "loss": 0.0576, "step": 180670 }, { "epoch": 6.108387707495182, "grad_norm": 1.2401480674743652, "learning_rate": 2.529851123363236e-06, "loss": 0.0505, "step": 180680 }, { "epoch": 6.1087257851854355, "grad_norm": 0.911722719669342, "learning_rate": 2.5272553246448615e-06, "loss": 0.055, "step": 180690 }, { "epoch": 6.109063862875689, "grad_norm": 1.0618791580200195, "learning_rate": 2.5246608238135015e-06, "loss": 0.0513, "step": 180700 }, { "epoch": 6.109401940565942, "grad_norm": 0.7146616578102112, "learning_rate": 2.5220676209400717e-06, "loss": 0.0496, "step": 180710 }, { "epoch": 6.109740018256195, "grad_norm": 0.7802998423576355, "learning_rate": 2.5194757160954765e-06, "loss": 0.0475, "step": 180720 }, { "epoch": 6.110078095946449, "grad_norm": 1.1353837251663208, "learning_rate": 2.5168851093505807e-06, "loss": 0.0584, "step": 180730 }, { "epoch": 6.110416173636701, "grad_norm": 0.8982781171798706, "learning_rate": 2.514295800776201e-06, "loss": 0.0475, "step": 180740 }, { "epoch": 6.110754251326955, "grad_norm": 2.0353076457977295, "learning_rate": 2.5117077904431517e-06, "loss": 0.0527, "step": 180750 }, { "epoch": 6.111092329017208, "grad_norm": 0.7385593056678772, "learning_rate": 2.509121078422155e-06, "loss": 0.0485, "step": 180760 }, { "epoch": 6.111430406707461, "grad_norm": 0.9842239022254944, "learning_rate": 2.5065356647839487e-06, "loss": 0.0528, "step": 180770 }, { "epoch": 6.1117684843977145, "grad_norm": 1.321737289428711, "learning_rate": 2.503951549599215e-06, "loss": 0.0493, "step": 180780 }, { "epoch": 6.112106562087968, "grad_norm": 0.8430019617080688, "learning_rate": 2.5013687329386027e-06, "loss": 0.0481, "step": 180790 }, { "epoch": 6.112444639778221, "grad_norm": 0.5205417275428772, "learning_rate": 2.4987872148727333e-06, "loss": 0.0539, "step": 180800 }, { "epoch": 6.112782717468474, "grad_norm": 0.639661967754364, "learning_rate": 2.496206995472161e-06, "loss": 0.0453, "step": 180810 }, { "epoch": 6.113120795158728, "grad_norm": 0.9761170148849487, "learning_rate": 2.4936280748074525e-06, "loss": 0.0515, "step": 180820 }, { "epoch": 6.11345887284898, "grad_norm": 0.8899286389350891, "learning_rate": 2.4910504529491064e-06, "loss": 0.0502, "step": 180830 }, { "epoch": 6.113796950539234, "grad_norm": 2.48462176322937, "learning_rate": 2.488474129967594e-06, "loss": 0.0506, "step": 180840 }, { "epoch": 6.114135028229487, "grad_norm": 0.45105960965156555, "learning_rate": 2.4858991059333425e-06, "loss": 0.0518, "step": 180850 }, { "epoch": 6.114473105919741, "grad_norm": 1.3556599617004395, "learning_rate": 2.4833253809167566e-06, "loss": 0.0474, "step": 180860 }, { "epoch": 6.114811183609993, "grad_norm": 1.0537958145141602, "learning_rate": 2.480752954988208e-06, "loss": 0.0539, "step": 180870 }, { "epoch": 6.115149261300247, "grad_norm": 1.5195387601852417, "learning_rate": 2.4781818282180292e-06, "loss": 0.0539, "step": 180880 }, { "epoch": 6.1154873389905, "grad_norm": 0.6077849268913269, "learning_rate": 2.4756120006764972e-06, "loss": 0.0482, "step": 180890 }, { "epoch": 6.115825416680753, "grad_norm": 1.5589507818222046, "learning_rate": 2.473043472433878e-06, "loss": 0.0625, "step": 180900 }, { "epoch": 6.116163494371007, "grad_norm": 1.8241304159164429, "learning_rate": 2.4704762435604044e-06, "loss": 0.0539, "step": 180910 }, { "epoch": 6.11650157206126, "grad_norm": 3.5213615894317627, "learning_rate": 2.467910314126248e-06, "loss": 0.0482, "step": 180920 }, { "epoch": 6.116839649751513, "grad_norm": 1.0417150259017944, "learning_rate": 2.4653456842015642e-06, "loss": 0.0543, "step": 180930 }, { "epoch": 6.117177727441766, "grad_norm": 0.8332805633544922, "learning_rate": 2.462782353856474e-06, "loss": 0.0522, "step": 180940 }, { "epoch": 6.11751580513202, "grad_norm": 1.011803150177002, "learning_rate": 2.4602203231610556e-06, "loss": 0.053, "step": 180950 }, { "epoch": 6.117853882822272, "grad_norm": 1.7829231023788452, "learning_rate": 2.4576595921853627e-06, "loss": 0.0517, "step": 180960 }, { "epoch": 6.118191960512526, "grad_norm": 0.8369168639183044, "learning_rate": 2.4551001609993796e-06, "loss": 0.0515, "step": 180970 }, { "epoch": 6.118530038202779, "grad_norm": 2.136622905731201, "learning_rate": 2.4525420296730995e-06, "loss": 0.0585, "step": 180980 }, { "epoch": 6.118868115893032, "grad_norm": 0.4525745213031769, "learning_rate": 2.449985198276461e-06, "loss": 0.0489, "step": 180990 }, { "epoch": 6.1192061935832855, "grad_norm": 0.5802786350250244, "learning_rate": 2.4474296668793696e-06, "loss": 0.0553, "step": 181000 }, { "epoch": 6.119544271273539, "grad_norm": 1.0269899368286133, "learning_rate": 2.44487543555168e-06, "loss": 0.0525, "step": 181010 }, { "epoch": 6.119882348963792, "grad_norm": 0.6006488800048828, "learning_rate": 2.4423225043632312e-06, "loss": 0.0509, "step": 181020 }, { "epoch": 6.120220426654045, "grad_norm": 0.9054251313209534, "learning_rate": 2.4397708733838178e-06, "loss": 0.0552, "step": 181030 }, { "epoch": 6.120558504344299, "grad_norm": 0.941110372543335, "learning_rate": 2.4372205426832053e-06, "loss": 0.054, "step": 181040 }, { "epoch": 6.120896582034551, "grad_norm": 1.7095316648483276, "learning_rate": 2.434671512331105e-06, "loss": 0.0586, "step": 181050 }, { "epoch": 6.121234659724805, "grad_norm": 0.6882056593894958, "learning_rate": 2.4321237823972276e-06, "loss": 0.0549, "step": 181060 }, { "epoch": 6.121572737415058, "grad_norm": 1.1405138969421387, "learning_rate": 2.4295773529512066e-06, "loss": 0.0593, "step": 181070 }, { "epoch": 6.121910815105311, "grad_norm": 1.0896049737930298, "learning_rate": 2.427032224062675e-06, "loss": 0.0518, "step": 181080 }, { "epoch": 6.1222488927955645, "grad_norm": 2.5784127712249756, "learning_rate": 2.424488395801211e-06, "loss": 0.0544, "step": 181090 }, { "epoch": 6.122586970485818, "grad_norm": 1.4734216928482056, "learning_rate": 2.4219458682363527e-06, "loss": 0.0492, "step": 181100 }, { "epoch": 6.122925048176071, "grad_norm": 0.5174538493156433, "learning_rate": 2.4194046414376283e-06, "loss": 0.0529, "step": 181110 }, { "epoch": 6.123263125866324, "grad_norm": 0.6796219944953918, "learning_rate": 2.4168647154745037e-06, "loss": 0.0441, "step": 181120 }, { "epoch": 6.123601203556578, "grad_norm": 1.598462700843811, "learning_rate": 2.4143260904164246e-06, "loss": 0.0538, "step": 181130 }, { "epoch": 6.12393928124683, "grad_norm": 1.247078776359558, "learning_rate": 2.4117887663327845e-06, "loss": 0.0538, "step": 181140 }, { "epoch": 6.124277358937084, "grad_norm": 0.7685192823410034, "learning_rate": 2.409252743292967e-06, "loss": 0.0521, "step": 181150 }, { "epoch": 6.124615436627337, "grad_norm": 0.6895242929458618, "learning_rate": 2.4067180213663e-06, "loss": 0.0493, "step": 181160 }, { "epoch": 6.12495351431759, "grad_norm": 0.6331155896186829, "learning_rate": 2.4041846006220837e-06, "loss": 0.0541, "step": 181170 }, { "epoch": 6.1252915920078435, "grad_norm": 0.7200593948364258, "learning_rate": 2.4016524811295792e-06, "loss": 0.0572, "step": 181180 }, { "epoch": 6.125629669698097, "grad_norm": 1.1906232833862305, "learning_rate": 2.399121662958009e-06, "loss": 0.0525, "step": 181190 }, { "epoch": 6.12596774738835, "grad_norm": 0.9647044539451599, "learning_rate": 2.396592146176574e-06, "loss": 0.0543, "step": 181200 }, { "epoch": 6.126305825078603, "grad_norm": 1.713754415512085, "learning_rate": 2.3940639308544287e-06, "loss": 0.058, "step": 181210 }, { "epoch": 6.126643902768857, "grad_norm": 0.6618756651878357, "learning_rate": 2.391537017060691e-06, "loss": 0.0481, "step": 181220 }, { "epoch": 6.126981980459109, "grad_norm": 0.9751932621002197, "learning_rate": 2.3890114048644384e-06, "loss": 0.053, "step": 181230 }, { "epoch": 6.127320058149363, "grad_norm": 0.7089768648147583, "learning_rate": 2.3864870943347274e-06, "loss": 0.0532, "step": 181240 }, { "epoch": 6.127658135839616, "grad_norm": 6.3886494636535645, "learning_rate": 2.3839640855405853e-06, "loss": 0.0544, "step": 181250 }, { "epoch": 6.127996213529869, "grad_norm": 1.1790697574615479, "learning_rate": 2.3814423785509576e-06, "loss": 0.055, "step": 181260 }, { "epoch": 6.128334291220122, "grad_norm": 0.6320131421089172, "learning_rate": 2.3789219734348224e-06, "loss": 0.0536, "step": 181270 }, { "epoch": 6.128672368910376, "grad_norm": 1.4632099866867065, "learning_rate": 2.3764028702610575e-06, "loss": 0.0626, "step": 181280 }, { "epoch": 6.1290104466006285, "grad_norm": 0.9715215563774109, "learning_rate": 2.3738850690985525e-06, "loss": 0.0516, "step": 181290 }, { "epoch": 6.129348524290882, "grad_norm": 1.5378113985061646, "learning_rate": 2.3713685700161415e-06, "loss": 0.0479, "step": 181300 }, { "epoch": 6.129686601981136, "grad_norm": 1.9146630764007568, "learning_rate": 2.3688533730826078e-06, "loss": 0.0524, "step": 181310 }, { "epoch": 6.130024679671388, "grad_norm": 1.0806372165679932, "learning_rate": 2.3663394783667415e-06, "loss": 0.0523, "step": 181320 }, { "epoch": 6.130362757361642, "grad_norm": 1.0994969606399536, "learning_rate": 2.3638268859372535e-06, "loss": 0.0541, "step": 181330 }, { "epoch": 6.130700835051895, "grad_norm": 0.5586904287338257, "learning_rate": 2.361315595862845e-06, "loss": 0.053, "step": 181340 }, { "epoch": 6.131038912742148, "grad_norm": 2.168884038925171, "learning_rate": 2.3588056082121667e-06, "loss": 0.0524, "step": 181350 }, { "epoch": 6.131376990432401, "grad_norm": 0.922812819480896, "learning_rate": 2.3562969230538414e-06, "loss": 0.0567, "step": 181360 }, { "epoch": 6.131715068122655, "grad_norm": 0.7524814605712891, "learning_rate": 2.353789540456469e-06, "loss": 0.0581, "step": 181370 }, { "epoch": 6.1320531458129075, "grad_norm": 4.380501747131348, "learning_rate": 2.35128346048859e-06, "loss": 0.0501, "step": 181380 }, { "epoch": 6.132391223503161, "grad_norm": 2.030726432800293, "learning_rate": 2.3487786832187163e-06, "loss": 0.0513, "step": 181390 }, { "epoch": 6.1327293011934145, "grad_norm": 0.7988460063934326, "learning_rate": 2.3462752087153318e-06, "loss": 0.0512, "step": 181400 }, { "epoch": 6.133067378883667, "grad_norm": 0.9325039982795715, "learning_rate": 2.3437730370468814e-06, "loss": 0.0549, "step": 181410 }, { "epoch": 6.133405456573921, "grad_norm": 0.6396743655204773, "learning_rate": 2.341272168281777e-06, "loss": 0.052, "step": 181420 }, { "epoch": 6.133743534264174, "grad_norm": 1.0065507888793945, "learning_rate": 2.338772602488376e-06, "loss": 0.0526, "step": 181430 }, { "epoch": 6.134081611954427, "grad_norm": 0.7744230031967163, "learning_rate": 2.3362743397350338e-06, "loss": 0.0486, "step": 181440 }, { "epoch": 6.13441968964468, "grad_norm": 0.6495693922042847, "learning_rate": 2.333777380090041e-06, "loss": 0.0557, "step": 181450 }, { "epoch": 6.134757767334934, "grad_norm": 1.7550456523895264, "learning_rate": 2.331281723621681e-06, "loss": 0.0514, "step": 181460 }, { "epoch": 6.1350958450251865, "grad_norm": 1.379722237586975, "learning_rate": 2.3287873703981556e-06, "loss": 0.0469, "step": 181470 }, { "epoch": 6.13543392271544, "grad_norm": 5.582474231719971, "learning_rate": 2.326294320487671e-06, "loss": 0.0523, "step": 181480 }, { "epoch": 6.1357720004056935, "grad_norm": 0.681952178478241, "learning_rate": 2.3238025739584e-06, "loss": 0.0574, "step": 181490 }, { "epoch": 6.136110078095946, "grad_norm": 1.8386176824569702, "learning_rate": 2.3213121308784556e-06, "loss": 0.053, "step": 181500 }, { "epoch": 6.1364481557862, "grad_norm": 1.469472050666809, "learning_rate": 2.318822991315922e-06, "loss": 0.0519, "step": 181510 }, { "epoch": 6.136786233476453, "grad_norm": 3.213057041168213, "learning_rate": 2.316335155338856e-06, "loss": 0.0552, "step": 181520 }, { "epoch": 6.137124311166706, "grad_norm": 0.720257043838501, "learning_rate": 2.313848623015269e-06, "loss": 0.0471, "step": 181530 }, { "epoch": 6.137462388856959, "grad_norm": 0.7512444853782654, "learning_rate": 2.3113633944131518e-06, "loss": 0.0515, "step": 181540 }, { "epoch": 6.137800466547213, "grad_norm": 1.6452031135559082, "learning_rate": 2.308879469600439e-06, "loss": 0.0519, "step": 181550 }, { "epoch": 6.138138544237465, "grad_norm": 1.4006117582321167, "learning_rate": 2.3063968486450483e-06, "loss": 0.0518, "step": 181560 }, { "epoch": 6.138476621927719, "grad_norm": 0.7276862859725952, "learning_rate": 2.303915531614842e-06, "loss": 0.0496, "step": 181570 }, { "epoch": 6.138814699617972, "grad_norm": 0.887779951095581, "learning_rate": 2.301435518577677e-06, "loss": 0.0493, "step": 181580 }, { "epoch": 6.139152777308225, "grad_norm": 0.898162841796875, "learning_rate": 2.2989568096013437e-06, "loss": 0.0493, "step": 181590 }, { "epoch": 6.139490854998479, "grad_norm": 1.8834383487701416, "learning_rate": 2.2964794047536096e-06, "loss": 0.0529, "step": 181600 }, { "epoch": 6.139828932688732, "grad_norm": 0.578299880027771, "learning_rate": 2.2940033041021993e-06, "loss": 0.0517, "step": 181610 }, { "epoch": 6.140167010378985, "grad_norm": 0.6040945649147034, "learning_rate": 2.291528507714824e-06, "loss": 0.0537, "step": 181620 }, { "epoch": 6.140505088069238, "grad_norm": 0.6670138835906982, "learning_rate": 2.2890550156591418e-06, "loss": 0.0489, "step": 181630 }, { "epoch": 6.140843165759492, "grad_norm": 1.1567809581756592, "learning_rate": 2.286582828002759e-06, "loss": 0.0593, "step": 181640 }, { "epoch": 6.141181243449744, "grad_norm": 1.0665024518966675, "learning_rate": 2.284111944813283e-06, "loss": 0.0535, "step": 181650 }, { "epoch": 6.141519321139998, "grad_norm": 0.552495002746582, "learning_rate": 2.281642366158254e-06, "loss": 0.057, "step": 181660 }, { "epoch": 6.141857398830251, "grad_norm": 1.4464013576507568, "learning_rate": 2.2791740921052017e-06, "loss": 0.0551, "step": 181670 }, { "epoch": 6.142195476520504, "grad_norm": 0.300858736038208, "learning_rate": 2.276707122721605e-06, "loss": 0.0513, "step": 181680 }, { "epoch": 6.1425335542107575, "grad_norm": 1.2466403245925903, "learning_rate": 2.2742414580748993e-06, "loss": 0.0511, "step": 181690 }, { "epoch": 6.142871631901011, "grad_norm": 0.7873505353927612, "learning_rate": 2.271777098232508e-06, "loss": 0.0573, "step": 181700 }, { "epoch": 6.143209709591264, "grad_norm": 1.1745411157608032, "learning_rate": 2.2693140432618e-06, "loss": 0.0528, "step": 181710 }, { "epoch": 6.143547787281517, "grad_norm": 1.146576166152954, "learning_rate": 2.266852293230115e-06, "loss": 0.0467, "step": 181720 }, { "epoch": 6.143885864971771, "grad_norm": 0.810383677482605, "learning_rate": 2.26439184820475e-06, "loss": 0.0549, "step": 181730 }, { "epoch": 6.144223942662023, "grad_norm": 4.146225452423096, "learning_rate": 2.2619327082529786e-06, "loss": 0.0489, "step": 181740 }, { "epoch": 6.144562020352277, "grad_norm": 1.1344163417816162, "learning_rate": 2.259474873442041e-06, "loss": 0.0493, "step": 181750 }, { "epoch": 6.14490009804253, "grad_norm": 5.140076160430908, "learning_rate": 2.2570183438391177e-06, "loss": 0.054, "step": 181760 }, { "epoch": 6.145238175732783, "grad_norm": 1.1959006786346436, "learning_rate": 2.2545631195113827e-06, "loss": 0.0472, "step": 181770 }, { "epoch": 6.1455762534230365, "grad_norm": 0.8297168612480164, "learning_rate": 2.252109200525948e-06, "loss": 0.0499, "step": 181780 }, { "epoch": 6.14591433111329, "grad_norm": 1.365597128868103, "learning_rate": 2.2496565869499107e-06, "loss": 0.053, "step": 181790 }, { "epoch": 6.146252408803543, "grad_norm": 6.6638264656066895, "learning_rate": 2.2472052788503285e-06, "loss": 0.0569, "step": 181800 }, { "epoch": 6.146590486493796, "grad_norm": 0.9630396962165833, "learning_rate": 2.244755276294208e-06, "loss": 0.0501, "step": 181810 }, { "epoch": 6.14692856418405, "grad_norm": 0.5429942011833191, "learning_rate": 2.242306579348541e-06, "loss": 0.0523, "step": 181820 }, { "epoch": 6.147266641874303, "grad_norm": 1.1959178447723389, "learning_rate": 2.2398591880802734e-06, "loss": 0.0521, "step": 181830 }, { "epoch": 6.147604719564556, "grad_norm": 0.8155199289321899, "learning_rate": 2.237413102556313e-06, "loss": 0.055, "step": 181840 }, { "epoch": 6.147942797254809, "grad_norm": 0.6354270577430725, "learning_rate": 2.234968322843528e-06, "loss": 0.0539, "step": 181850 }, { "epoch": 6.148280874945062, "grad_norm": 1.048546314239502, "learning_rate": 2.2325248490087715e-06, "loss": 0.0547, "step": 181860 }, { "epoch": 6.148618952635315, "grad_norm": 2.9057533740997314, "learning_rate": 2.230082681118828e-06, "loss": 0.0488, "step": 181870 }, { "epoch": 6.148957030325569, "grad_norm": 0.8193291425704956, "learning_rate": 2.227641819240489e-06, "loss": 0.0555, "step": 181880 }, { "epoch": 6.1492951080158225, "grad_norm": 2.1968979835510254, "learning_rate": 2.2252022634404734e-06, "loss": 0.0523, "step": 181890 }, { "epoch": 6.149633185706075, "grad_norm": 0.6132403016090393, "learning_rate": 2.2227640137854777e-06, "loss": 0.0516, "step": 181900 }, { "epoch": 6.149971263396329, "grad_norm": 0.6794495582580566, "learning_rate": 2.220327070342165e-06, "loss": 0.0535, "step": 181910 }, { "epoch": 6.150309341086582, "grad_norm": 0.812583327293396, "learning_rate": 2.217891433177166e-06, "loss": 0.0532, "step": 181920 }, { "epoch": 6.150647418776835, "grad_norm": 1.0952256917953491, "learning_rate": 2.2154571023570547e-06, "loss": 0.0526, "step": 181930 }, { "epoch": 6.150985496467088, "grad_norm": 0.7684354782104492, "learning_rate": 2.2130240779484e-06, "loss": 0.0483, "step": 181940 }, { "epoch": 6.151323574157342, "grad_norm": 1.3499568700790405, "learning_rate": 2.2105923600177105e-06, "loss": 0.0513, "step": 181950 }, { "epoch": 6.151661651847594, "grad_norm": 2.3204472064971924, "learning_rate": 2.2081619486314875e-06, "loss": 0.0475, "step": 181960 }, { "epoch": 6.151999729537848, "grad_norm": 0.9847950339317322, "learning_rate": 2.2057328438561455e-06, "loss": 0.0527, "step": 181970 }, { "epoch": 6.152337807228101, "grad_norm": 0.5486800670623779, "learning_rate": 2.2033050457581196e-06, "loss": 0.051, "step": 181980 }, { "epoch": 6.152675884918354, "grad_norm": 1.0558412075042725, "learning_rate": 2.2008785544037682e-06, "loss": 0.057, "step": 181990 }, { "epoch": 6.153013962608608, "grad_norm": 0.7300518751144409, "learning_rate": 2.198453369859449e-06, "loss": 0.0538, "step": 182000 }, { "epoch": 6.153352040298861, "grad_norm": 1.332355260848999, "learning_rate": 2.1960294921914537e-06, "loss": 0.0509, "step": 182010 }, { "epoch": 6.153690117989114, "grad_norm": 0.6644890904426575, "learning_rate": 2.1936069214660516e-06, "loss": 0.0539, "step": 182020 }, { "epoch": 6.154028195679367, "grad_norm": 0.8629720211029053, "learning_rate": 2.1911856577494783e-06, "loss": 0.0514, "step": 182030 }, { "epoch": 6.154366273369621, "grad_norm": 1.6386619806289673, "learning_rate": 2.188765701107931e-06, "loss": 0.0647, "step": 182040 }, { "epoch": 6.154704351059873, "grad_norm": 1.238825798034668, "learning_rate": 2.186347051607562e-06, "loss": 0.056, "step": 182050 }, { "epoch": 6.155042428750127, "grad_norm": 0.563820481300354, "learning_rate": 2.183929709314503e-06, "loss": 0.0499, "step": 182060 }, { "epoch": 6.15538050644038, "grad_norm": 1.0648441314697266, "learning_rate": 2.1815136742948382e-06, "loss": 0.0519, "step": 182070 }, { "epoch": 6.155718584130633, "grad_norm": 2.466769218444824, "learning_rate": 2.1790989466146327e-06, "loss": 0.0564, "step": 182080 }, { "epoch": 6.1560566618208865, "grad_norm": 1.5602736473083496, "learning_rate": 2.1766855263398945e-06, "loss": 0.0545, "step": 182090 }, { "epoch": 6.15639473951114, "grad_norm": 0.7268685698509216, "learning_rate": 2.1742734135366094e-06, "loss": 0.0541, "step": 182100 }, { "epoch": 6.156732817201393, "grad_norm": 0.9521709680557251, "learning_rate": 2.171862608270714e-06, "loss": 0.0579, "step": 182110 }, { "epoch": 6.157070894891646, "grad_norm": 0.5029749274253845, "learning_rate": 2.169453110608133e-06, "loss": 0.0514, "step": 182120 }, { "epoch": 6.1574089725819, "grad_norm": 1.4907375574111938, "learning_rate": 2.167044920614736e-06, "loss": 0.0515, "step": 182130 }, { "epoch": 6.157747050272152, "grad_norm": 0.7802935242652893, "learning_rate": 2.1646380383563537e-06, "loss": 0.054, "step": 182140 }, { "epoch": 6.158085127962406, "grad_norm": 1.235382318496704, "learning_rate": 2.1622324638988057e-06, "loss": 0.0579, "step": 182150 }, { "epoch": 6.158423205652659, "grad_norm": 0.5318626165390015, "learning_rate": 2.1598281973078395e-06, "loss": 0.0521, "step": 182160 }, { "epoch": 6.158761283342912, "grad_norm": 1.5341368913650513, "learning_rate": 2.1574252386492132e-06, "loss": 0.0526, "step": 182170 }, { "epoch": 6.1590993610331655, "grad_norm": 1.8814938068389893, "learning_rate": 2.1550235879885914e-06, "loss": 0.0605, "step": 182180 }, { "epoch": 6.159437438723419, "grad_norm": 0.8195576667785645, "learning_rate": 2.152623245391655e-06, "loss": 0.059, "step": 182190 }, { "epoch": 6.159775516413672, "grad_norm": 0.5064899325370789, "learning_rate": 2.150224210924029e-06, "loss": 0.0545, "step": 182200 }, { "epoch": 6.160113594103925, "grad_norm": 0.44297441840171814, "learning_rate": 2.147826484651294e-06, "loss": 0.0532, "step": 182210 }, { "epoch": 6.160451671794179, "grad_norm": 0.7746579051017761, "learning_rate": 2.145430066639009e-06, "loss": 0.0535, "step": 182220 }, { "epoch": 6.160789749484431, "grad_norm": 0.4750208258628845, "learning_rate": 2.1430349569526777e-06, "loss": 0.0495, "step": 182230 }, { "epoch": 6.161127827174685, "grad_norm": 0.9481210708618164, "learning_rate": 2.140641155657802e-06, "loss": 0.0522, "step": 182240 }, { "epoch": 6.161465904864938, "grad_norm": 1.28354811668396, "learning_rate": 2.138248662819814e-06, "loss": 0.0544, "step": 182250 }, { "epoch": 6.161803982555191, "grad_norm": 1.0011858940124512, "learning_rate": 2.1358574785041217e-06, "loss": 0.0538, "step": 182260 }, { "epoch": 6.162142060245444, "grad_norm": 0.4792812466621399, "learning_rate": 2.1334676027761126e-06, "loss": 0.0564, "step": 182270 }, { "epoch": 6.162480137935698, "grad_norm": 1.7535998821258545, "learning_rate": 2.1310790357011058e-06, "loss": 0.0541, "step": 182280 }, { "epoch": 6.162818215625951, "grad_norm": 0.7245689630508423, "learning_rate": 2.1286917773444216e-06, "loss": 0.0494, "step": 182290 }, { "epoch": 6.163156293316204, "grad_norm": 1.0293598175048828, "learning_rate": 2.1263058277713242e-06, "loss": 0.0508, "step": 182300 }, { "epoch": 6.163494371006458, "grad_norm": 1.4906872510910034, "learning_rate": 2.123921187047029e-06, "loss": 0.0566, "step": 182310 }, { "epoch": 6.16383244869671, "grad_norm": 1.9156792163848877, "learning_rate": 2.1215378552367492e-06, "loss": 0.0537, "step": 182320 }, { "epoch": 6.164170526386964, "grad_norm": 0.7494385242462158, "learning_rate": 2.1191558324056393e-06, "loss": 0.052, "step": 182330 }, { "epoch": 6.164508604077217, "grad_norm": 2.1337735652923584, "learning_rate": 2.1167751186188244e-06, "loss": 0.0488, "step": 182340 }, { "epoch": 6.16484668176747, "grad_norm": 1.1118444204330444, "learning_rate": 2.114395713941375e-06, "loss": 0.0533, "step": 182350 }, { "epoch": 6.165184759457723, "grad_norm": 1.074526071548462, "learning_rate": 2.1120176184383667e-06, "loss": 0.0577, "step": 182360 }, { "epoch": 6.165522837147977, "grad_norm": 1.8938227891921997, "learning_rate": 2.109640832174803e-06, "loss": 0.0578, "step": 182370 }, { "epoch": 6.1658609148382295, "grad_norm": 0.9231767058372498, "learning_rate": 2.1072653552156706e-06, "loss": 0.0604, "step": 182380 }, { "epoch": 6.166198992528483, "grad_norm": 1.6928701400756836, "learning_rate": 2.104891187625907e-06, "loss": 0.0544, "step": 182390 }, { "epoch": 6.1665370702187365, "grad_norm": 1.2733350992202759, "learning_rate": 2.1025183294704265e-06, "loss": 0.0523, "step": 182400 }, { "epoch": 6.166875147908989, "grad_norm": 0.7843475937843323, "learning_rate": 2.1001467808140993e-06, "loss": 0.0508, "step": 182410 }, { "epoch": 6.167213225599243, "grad_norm": 28.006969451904297, "learning_rate": 2.0977765417217686e-06, "loss": 0.0494, "step": 182420 }, { "epoch": 6.167551303289496, "grad_norm": 0.7293487191200256, "learning_rate": 2.0954076122582324e-06, "loss": 0.0537, "step": 182430 }, { "epoch": 6.167889380979749, "grad_norm": 0.5609843134880066, "learning_rate": 2.093039992488249e-06, "loss": 0.0488, "step": 182440 }, { "epoch": 6.168227458670002, "grad_norm": 0.7119935750961304, "learning_rate": 2.0906736824765515e-06, "loss": 0.0552, "step": 182450 }, { "epoch": 6.168565536360256, "grad_norm": 2.856782913208008, "learning_rate": 2.0883086822878484e-06, "loss": 0.054, "step": 182460 }, { "epoch": 6.1689036140505085, "grad_norm": 1.018445372581482, "learning_rate": 2.085944991986777e-06, "loss": 0.0511, "step": 182470 }, { "epoch": 6.169241691740762, "grad_norm": 1.571494460105896, "learning_rate": 2.083582611637974e-06, "loss": 0.0509, "step": 182480 }, { "epoch": 6.1695797694310155, "grad_norm": 0.8171431422233582, "learning_rate": 2.081221541306011e-06, "loss": 0.0563, "step": 182490 }, { "epoch": 6.169917847121268, "grad_norm": 0.9793990254402161, "learning_rate": 2.078861781055463e-06, "loss": 0.0499, "step": 182500 }, { "epoch": 6.170255924811522, "grad_norm": 0.7783454060554504, "learning_rate": 2.076503330950824e-06, "loss": 0.0553, "step": 182510 }, { "epoch": 6.170594002501775, "grad_norm": 1.0584087371826172, "learning_rate": 2.074146191056575e-06, "loss": 0.0536, "step": 182520 }, { "epoch": 6.170932080192028, "grad_norm": 0.5061303377151489, "learning_rate": 2.071790361437176e-06, "loss": 0.0527, "step": 182530 }, { "epoch": 6.171270157882281, "grad_norm": 0.4737567901611328, "learning_rate": 2.069435842157014e-06, "loss": 0.0494, "step": 182540 }, { "epoch": 6.171608235572535, "grad_norm": 0.9226948022842407, "learning_rate": 2.067082633280476e-06, "loss": 0.0518, "step": 182550 }, { "epoch": 6.171946313262787, "grad_norm": 2.1797239780426025, "learning_rate": 2.0647307348718834e-06, "loss": 0.0539, "step": 182560 }, { "epoch": 6.172284390953041, "grad_norm": 0.5443108677864075, "learning_rate": 2.0623801469955404e-06, "loss": 0.0473, "step": 182570 }, { "epoch": 6.1726224686432944, "grad_norm": 1.1408735513687134, "learning_rate": 2.060030869715723e-06, "loss": 0.0506, "step": 182580 }, { "epoch": 6.172960546333547, "grad_norm": 0.9627345204353333, "learning_rate": 2.0576829030966525e-06, "loss": 0.0471, "step": 182590 }, { "epoch": 6.173298624023801, "grad_norm": 0.6275210976600647, "learning_rate": 2.0553362472025217e-06, "loss": 0.053, "step": 182600 }, { "epoch": 6.173636701714054, "grad_norm": 1.3760426044464111, "learning_rate": 2.052990902097479e-06, "loss": 0.0573, "step": 182610 }, { "epoch": 6.173974779404307, "grad_norm": 1.0719077587127686, "learning_rate": 2.050646867845657e-06, "loss": 0.0497, "step": 182620 }, { "epoch": 6.17431285709456, "grad_norm": 1.1263556480407715, "learning_rate": 2.048304144511137e-06, "loss": 0.0486, "step": 182630 }, { "epoch": 6.174650934784814, "grad_norm": 0.765651285648346, "learning_rate": 2.0459627321579635e-06, "loss": 0.0513, "step": 182640 }, { "epoch": 6.174989012475066, "grad_norm": 0.9480125904083252, "learning_rate": 2.043622630850156e-06, "loss": 0.059, "step": 182650 }, { "epoch": 6.17532709016532, "grad_norm": 0.7145214080810547, "learning_rate": 2.041283840651681e-06, "loss": 0.0511, "step": 182660 }, { "epoch": 6.175665167855573, "grad_norm": 0.4123634099960327, "learning_rate": 2.0389463616265037e-06, "loss": 0.0495, "step": 182670 }, { "epoch": 6.176003245545826, "grad_norm": 0.7354621291160583, "learning_rate": 2.0366101938385064e-06, "loss": 0.0533, "step": 182680 }, { "epoch": 6.1763413232360795, "grad_norm": 0.5316206216812134, "learning_rate": 2.03427533735156e-06, "loss": 0.054, "step": 182690 }, { "epoch": 6.176679400926333, "grad_norm": 0.8331571817398071, "learning_rate": 2.031941792229519e-06, "loss": 0.053, "step": 182700 }, { "epoch": 6.177017478616586, "grad_norm": 0.8642734289169312, "learning_rate": 2.0296095585361606e-06, "loss": 0.0542, "step": 182710 }, { "epoch": 6.177355556306839, "grad_norm": 2.7934725284576416, "learning_rate": 2.027278636335261e-06, "loss": 0.0573, "step": 182720 }, { "epoch": 6.177693633997093, "grad_norm": 0.6733826994895935, "learning_rate": 2.024949025690537e-06, "loss": 0.0523, "step": 182730 }, { "epoch": 6.178031711687345, "grad_norm": 4.6539740562438965, "learning_rate": 2.022620726665686e-06, "loss": 0.0518, "step": 182740 }, { "epoch": 6.178369789377599, "grad_norm": 0.8195956945419312, "learning_rate": 2.0202937393243584e-06, "loss": 0.0586, "step": 182750 }, { "epoch": 6.178707867067852, "grad_norm": 0.9506604075431824, "learning_rate": 2.017968063730169e-06, "loss": 0.0561, "step": 182760 }, { "epoch": 6.179045944758105, "grad_norm": 0.7223851680755615, "learning_rate": 2.0156436999467178e-06, "loss": 0.0496, "step": 182770 }, { "epoch": 6.1793840224483585, "grad_norm": 0.9129946231842041, "learning_rate": 2.013320648037531e-06, "loss": 0.0547, "step": 182780 }, { "epoch": 6.179722100138612, "grad_norm": 0.640178382396698, "learning_rate": 2.010998908066136e-06, "loss": 0.048, "step": 182790 }, { "epoch": 6.1800601778288655, "grad_norm": 1.348408579826355, "learning_rate": 2.008678480095999e-06, "loss": 0.0534, "step": 182800 }, { "epoch": 6.180398255519118, "grad_norm": 2.115943193435669, "learning_rate": 2.006359364190569e-06, "loss": 0.0489, "step": 182810 }, { "epoch": 6.180736333209372, "grad_norm": 0.8470109105110168, "learning_rate": 2.0040415604132333e-06, "loss": 0.0627, "step": 182820 }, { "epoch": 6.181074410899624, "grad_norm": 0.49929872155189514, "learning_rate": 2.0017250688273815e-06, "loss": 0.0473, "step": 182830 }, { "epoch": 6.181412488589878, "grad_norm": 1.0630757808685303, "learning_rate": 1.9994098894963288e-06, "loss": 0.0474, "step": 182840 }, { "epoch": 6.181750566280131, "grad_norm": 0.5879382491111755, "learning_rate": 1.997096022483369e-06, "loss": 0.057, "step": 182850 }, { "epoch": 6.182088643970385, "grad_norm": 0.7857234477996826, "learning_rate": 1.9947834678517795e-06, "loss": 0.0545, "step": 182860 }, { "epoch": 6.1824267216606374, "grad_norm": 1.1934936046600342, "learning_rate": 1.992472225664771e-06, "loss": 0.0535, "step": 182870 }, { "epoch": 6.182764799350891, "grad_norm": 1.0067682266235352, "learning_rate": 1.9901622959855425e-06, "loss": 0.053, "step": 182880 }, { "epoch": 6.1831028770411445, "grad_norm": 2.7772278785705566, "learning_rate": 1.9878536788772385e-06, "loss": 0.0564, "step": 182890 }, { "epoch": 6.183440954731397, "grad_norm": 0.8973185420036316, "learning_rate": 1.985546374402969e-06, "loss": 0.0483, "step": 182900 }, { "epoch": 6.183779032421651, "grad_norm": 1.0480753183364868, "learning_rate": 1.9832403826258338e-06, "loss": 0.0535, "step": 182910 }, { "epoch": 6.184117110111904, "grad_norm": 1.8529013395309448, "learning_rate": 1.980935703608866e-06, "loss": 0.0557, "step": 182920 }, { "epoch": 6.184455187802157, "grad_norm": 0.8807229399681091, "learning_rate": 1.978632337415076e-06, "loss": 0.0509, "step": 182930 }, { "epoch": 6.18479326549241, "grad_norm": 0.8169487714767456, "learning_rate": 1.9763302841074307e-06, "loss": 0.0536, "step": 182940 }, { "epoch": 6.185131343182664, "grad_norm": 1.0195767879486084, "learning_rate": 1.974029543748873e-06, "loss": 0.0472, "step": 182950 }, { "epoch": 6.185469420872916, "grad_norm": 0.888970673084259, "learning_rate": 1.9717301164023204e-06, "loss": 0.0551, "step": 182960 }, { "epoch": 6.18580749856317, "grad_norm": 0.5984423756599426, "learning_rate": 1.9694320021306057e-06, "loss": 0.0563, "step": 182970 }, { "epoch": 6.186145576253423, "grad_norm": 1.280652403831482, "learning_rate": 1.967135200996584e-06, "loss": 0.048, "step": 182980 }, { "epoch": 6.186483653943676, "grad_norm": 0.5714284777641296, "learning_rate": 1.964839713063038e-06, "loss": 0.0571, "step": 182990 }, { "epoch": 6.18682173163393, "grad_norm": 2.028578281402588, "learning_rate": 1.9625455383927294e-06, "loss": 0.0558, "step": 183000 }, { "epoch": 6.187159809324183, "grad_norm": 2.016611337661743, "learning_rate": 1.9602526770483853e-06, "loss": 0.0552, "step": 183010 }, { "epoch": 6.187497887014436, "grad_norm": 0.6070467233657837, "learning_rate": 1.9579611290926727e-06, "loss": 0.0523, "step": 183020 }, { "epoch": 6.187835964704689, "grad_norm": 1.0920698642730713, "learning_rate": 1.955670894588263e-06, "loss": 0.0504, "step": 183030 }, { "epoch": 6.188174042394943, "grad_norm": 0.828780472278595, "learning_rate": 1.953381973597762e-06, "loss": 0.0524, "step": 183040 }, { "epoch": 6.188512120085195, "grad_norm": 0.7920927405357361, "learning_rate": 1.9510943661837476e-06, "loss": 0.0621, "step": 183050 }, { "epoch": 6.188850197775449, "grad_norm": 1.088557481765747, "learning_rate": 1.948808072408753e-06, "loss": 0.0497, "step": 183060 }, { "epoch": 6.189188275465702, "grad_norm": 3.5809133052825928, "learning_rate": 1.9465230923353063e-06, "loss": 0.0562, "step": 183070 }, { "epoch": 6.189526353155955, "grad_norm": 2.1586785316467285, "learning_rate": 1.944239426025851e-06, "loss": 0.0522, "step": 183080 }, { "epoch": 6.1898644308462085, "grad_norm": 1.0046658515930176, "learning_rate": 1.9419570735428493e-06, "loss": 0.0538, "step": 183090 }, { "epoch": 6.190202508536462, "grad_norm": 1.8780463933944702, "learning_rate": 1.939676034948684e-06, "loss": 0.0536, "step": 183100 }, { "epoch": 6.190540586226715, "grad_norm": 1.0284254550933838, "learning_rate": 1.937396310305717e-06, "loss": 0.0456, "step": 183110 }, { "epoch": 6.190878663916968, "grad_norm": 1.0100682973861694, "learning_rate": 1.935117899676281e-06, "loss": 0.0572, "step": 183120 }, { "epoch": 6.191216741607222, "grad_norm": 0.9750216603279114, "learning_rate": 1.9328408031226654e-06, "loss": 0.0467, "step": 183130 }, { "epoch": 6.191554819297474, "grad_norm": 1.04281747341156, "learning_rate": 1.9305650207071145e-06, "loss": 0.0517, "step": 183140 }, { "epoch": 6.191892896987728, "grad_norm": 1.0730890035629272, "learning_rate": 1.928290552491868e-06, "loss": 0.0539, "step": 183150 }, { "epoch": 6.192230974677981, "grad_norm": 0.6280877590179443, "learning_rate": 1.926017398539093e-06, "loss": 0.0486, "step": 183160 }, { "epoch": 6.192569052368234, "grad_norm": 1.0875507593154907, "learning_rate": 1.92374555891095e-06, "loss": 0.051, "step": 183170 }, { "epoch": 6.1929071300584875, "grad_norm": 0.8574487566947937, "learning_rate": 1.9214750336695343e-06, "loss": 0.0545, "step": 183180 }, { "epoch": 6.193245207748741, "grad_norm": 0.795377790927887, "learning_rate": 1.9192058228769293e-06, "loss": 0.0547, "step": 183190 }, { "epoch": 6.193583285438994, "grad_norm": 1.1154849529266357, "learning_rate": 1.916937926595175e-06, "loss": 0.0546, "step": 183200 }, { "epoch": 6.193921363129247, "grad_norm": 0.9409171342849731, "learning_rate": 1.9146713448862763e-06, "loss": 0.0537, "step": 183210 }, { "epoch": 6.194259440819501, "grad_norm": 0.6301463842391968, "learning_rate": 1.9124060778121953e-06, "loss": 0.049, "step": 183220 }, { "epoch": 6.194597518509753, "grad_norm": 0.7357832193374634, "learning_rate": 1.9101421254348607e-06, "loss": 0.0501, "step": 183230 }, { "epoch": 6.194935596200007, "grad_norm": 1.5832624435424805, "learning_rate": 1.9078794878161834e-06, "loss": 0.0535, "step": 183240 }, { "epoch": 6.19527367389026, "grad_norm": 0.5474095344543457, "learning_rate": 1.9056181650180138e-06, "loss": 0.0512, "step": 183250 }, { "epoch": 6.195611751580513, "grad_norm": 0.6206044554710388, "learning_rate": 1.9033581571021753e-06, "loss": 0.0483, "step": 183260 }, { "epoch": 6.195949829270766, "grad_norm": 1.0936532020568848, "learning_rate": 1.9010994641304459e-06, "loss": 0.0547, "step": 183270 }, { "epoch": 6.19628790696102, "grad_norm": 0.9687653183937073, "learning_rate": 1.8988420861645873e-06, "loss": 0.0537, "step": 183280 }, { "epoch": 6.196625984651273, "grad_norm": 3.3689141273498535, "learning_rate": 1.896586023266328e-06, "loss": 0.0503, "step": 183290 }, { "epoch": 6.196964062341526, "grad_norm": 1.5085397958755493, "learning_rate": 1.8943312754973297e-06, "loss": 0.0534, "step": 183300 }, { "epoch": 6.19730214003178, "grad_norm": 0.6787968277931213, "learning_rate": 1.892077842919243e-06, "loss": 0.0498, "step": 183310 }, { "epoch": 6.197640217722032, "grad_norm": 1.423010230064392, "learning_rate": 1.889825725593669e-06, "loss": 0.0536, "step": 183320 }, { "epoch": 6.197978295412286, "grad_norm": 1.2480008602142334, "learning_rate": 1.887574923582186e-06, "loss": 0.0519, "step": 183330 }, { "epoch": 6.198316373102539, "grad_norm": 0.6440044045448303, "learning_rate": 1.8853254369463392e-06, "loss": 0.0524, "step": 183340 }, { "epoch": 6.198654450792792, "grad_norm": 0.58999103307724, "learning_rate": 1.883077265747607e-06, "loss": 0.0557, "step": 183350 }, { "epoch": 6.198992528483045, "grad_norm": 0.9120384454727173, "learning_rate": 1.880830410047474e-06, "loss": 0.0575, "step": 183360 }, { "epoch": 6.199330606173299, "grad_norm": 1.6204782724380493, "learning_rate": 1.8785848699073516e-06, "loss": 0.059, "step": 183370 }, { "epoch": 6.1996686838635515, "grad_norm": 0.3467116057872772, "learning_rate": 1.8763406453886578e-06, "loss": 0.0491, "step": 183380 }, { "epoch": 6.200006761553805, "grad_norm": 0.6948140263557434, "learning_rate": 1.8740977365527157e-06, "loss": 0.0543, "step": 183390 }, { "epoch": 6.2003448392440585, "grad_norm": 1.962418556213379, "learning_rate": 1.8718561434608595e-06, "loss": 0.0495, "step": 183400 }, { "epoch": 6.200682916934311, "grad_norm": 0.7715311050415039, "learning_rate": 1.8696158661743846e-06, "loss": 0.0528, "step": 183410 }, { "epoch": 6.201020994624565, "grad_norm": 0.7165629863739014, "learning_rate": 1.8673769047545308e-06, "loss": 0.0496, "step": 183420 }, { "epoch": 6.201359072314818, "grad_norm": 0.5634532570838928, "learning_rate": 1.8651392592625105e-06, "loss": 0.0529, "step": 183430 }, { "epoch": 6.201697150005071, "grad_norm": 0.6536450386047363, "learning_rate": 1.8629029297594914e-06, "loss": 0.0592, "step": 183440 }, { "epoch": 6.202035227695324, "grad_norm": 1.4946860074996948, "learning_rate": 1.8606679163066243e-06, "loss": 0.0612, "step": 183450 }, { "epoch": 6.202373305385578, "grad_norm": 0.6432825326919556, "learning_rate": 1.8584342189650106e-06, "loss": 0.0566, "step": 183460 }, { "epoch": 6.2027113830758305, "grad_norm": 1.3818882703781128, "learning_rate": 1.8562018377957236e-06, "loss": 0.0521, "step": 183470 }, { "epoch": 6.203049460766084, "grad_norm": 1.2963718175888062, "learning_rate": 1.8539707728597922e-06, "loss": 0.0519, "step": 183480 }, { "epoch": 6.2033875384563375, "grad_norm": 0.6185532808303833, "learning_rate": 1.8517410242182064e-06, "loss": 0.0504, "step": 183490 }, { "epoch": 6.20372561614659, "grad_norm": 1.3769304752349854, "learning_rate": 1.8495125919319345e-06, "loss": 0.0571, "step": 183500 }, { "epoch": 6.204063693836844, "grad_norm": 0.770796000957489, "learning_rate": 1.847285476061905e-06, "loss": 0.0544, "step": 183510 }, { "epoch": 6.204401771527097, "grad_norm": 0.6114866733551025, "learning_rate": 1.845059676668992e-06, "loss": 0.0551, "step": 183520 }, { "epoch": 6.20473984921735, "grad_norm": 1.4260870218276978, "learning_rate": 1.8428351938140576e-06, "loss": 0.0527, "step": 183530 }, { "epoch": 6.205077926907603, "grad_norm": 2.0422909259796143, "learning_rate": 1.8406120275579198e-06, "loss": 0.0543, "step": 183540 }, { "epoch": 6.205416004597857, "grad_norm": 0.4691859781742096, "learning_rate": 1.8383901779613632e-06, "loss": 0.0464, "step": 183550 }, { "epoch": 6.205754082288109, "grad_norm": 0.5979186296463013, "learning_rate": 1.8361696450851173e-06, "loss": 0.0543, "step": 183560 }, { "epoch": 6.206092159978363, "grad_norm": 1.2237095832824707, "learning_rate": 1.8339504289899002e-06, "loss": 0.0442, "step": 183570 }, { "epoch": 6.2064302376686165, "grad_norm": 0.9572486877441406, "learning_rate": 1.8317325297363797e-06, "loss": 0.0513, "step": 183580 }, { "epoch": 6.206768315358869, "grad_norm": 0.8903906941413879, "learning_rate": 1.829515947385202e-06, "loss": 0.0511, "step": 183590 }, { "epoch": 6.207106393049123, "grad_norm": 0.8007000684738159, "learning_rate": 1.827300681996963e-06, "loss": 0.0566, "step": 183600 }, { "epoch": 6.207444470739376, "grad_norm": 0.6771592497825623, "learning_rate": 1.8250867336322141e-06, "loss": 0.0525, "step": 183610 }, { "epoch": 6.207782548429629, "grad_norm": 1.0670697689056396, "learning_rate": 1.8228741023515073e-06, "loss": 0.0535, "step": 183620 }, { "epoch": 6.208120626119882, "grad_norm": 2.5513510704040527, "learning_rate": 1.8206627882153216e-06, "loss": 0.0491, "step": 183630 }, { "epoch": 6.208458703810136, "grad_norm": 0.7343792915344238, "learning_rate": 1.8184527912841143e-06, "loss": 0.053, "step": 183640 }, { "epoch": 6.208796781500388, "grad_norm": 2.7462384700775146, "learning_rate": 1.8162441116183037e-06, "loss": 0.0505, "step": 183650 }, { "epoch": 6.209134859190642, "grad_norm": 1.093817114830017, "learning_rate": 1.814036749278275e-06, "loss": 0.0468, "step": 183660 }, { "epoch": 6.209472936880895, "grad_norm": 1.1451926231384277, "learning_rate": 1.81183070432438e-06, "loss": 0.0527, "step": 183670 }, { "epoch": 6.209811014571148, "grad_norm": 0.5110088586807251, "learning_rate": 1.8096259768169366e-06, "loss": 0.0501, "step": 183680 }, { "epoch": 6.2101490922614015, "grad_norm": 0.8775824308395386, "learning_rate": 1.8074225668162136e-06, "loss": 0.0563, "step": 183690 }, { "epoch": 6.210487169951655, "grad_norm": 1.332356572151184, "learning_rate": 1.8052204743824464e-06, "loss": 0.0492, "step": 183700 }, { "epoch": 6.210825247641908, "grad_norm": 0.6934872269630432, "learning_rate": 1.803019699575853e-06, "loss": 0.0505, "step": 183710 }, { "epoch": 6.211163325332161, "grad_norm": 0.698573648929596, "learning_rate": 1.8008202424565911e-06, "loss": 0.0584, "step": 183720 }, { "epoch": 6.211501403022415, "grad_norm": 0.9974992275238037, "learning_rate": 1.7986221030847906e-06, "loss": 0.054, "step": 183730 }, { "epoch": 6.211839480712667, "grad_norm": 0.3547292947769165, "learning_rate": 1.7964252815205585e-06, "loss": 0.0536, "step": 183740 }, { "epoch": 6.212177558402921, "grad_norm": 0.8215861916542053, "learning_rate": 1.7942297778239469e-06, "loss": 0.0547, "step": 183750 }, { "epoch": 6.212515636093174, "grad_norm": 1.0301520824432373, "learning_rate": 1.7920355920549914e-06, "loss": 0.052, "step": 183760 }, { "epoch": 6.212853713783428, "grad_norm": 0.5007059574127197, "learning_rate": 1.789842724273666e-06, "loss": 0.0549, "step": 183770 }, { "epoch": 6.2131917914736805, "grad_norm": 2.1101272106170654, "learning_rate": 1.7876511745399227e-06, "loss": 0.0503, "step": 183780 }, { "epoch": 6.213529869163934, "grad_norm": 0.576339066028595, "learning_rate": 1.7854609429136916e-06, "loss": 0.0463, "step": 183790 }, { "epoch": 6.213867946854187, "grad_norm": 1.1161020994186401, "learning_rate": 1.7832720294548465e-06, "loss": 0.0567, "step": 183800 }, { "epoch": 6.21420602454444, "grad_norm": 0.7828927636146545, "learning_rate": 1.781084434223229e-06, "loss": 0.0512, "step": 183810 }, { "epoch": 6.214544102234694, "grad_norm": 0.8709369897842407, "learning_rate": 1.7788981572786412e-06, "loss": 0.0498, "step": 183820 }, { "epoch": 6.214882179924947, "grad_norm": 1.0502068996429443, "learning_rate": 1.7767131986808683e-06, "loss": 0.0472, "step": 183830 }, { "epoch": 6.2152202576152, "grad_norm": 0.7728190422058105, "learning_rate": 1.7745295584896405e-06, "loss": 0.0451, "step": 183840 }, { "epoch": 6.215558335305453, "grad_norm": 1.0872619152069092, "learning_rate": 1.7723472367646543e-06, "loss": 0.0497, "step": 183850 }, { "epoch": 6.215896412995707, "grad_norm": 1.1700289249420166, "learning_rate": 1.770166233565579e-06, "loss": 0.0513, "step": 183860 }, { "epoch": 6.2162344906859595, "grad_norm": 0.5963453054428101, "learning_rate": 1.767986548952033e-06, "loss": 0.0517, "step": 183870 }, { "epoch": 6.216572568376213, "grad_norm": 0.6392315626144409, "learning_rate": 1.7658081829836193e-06, "loss": 0.053, "step": 183880 }, { "epoch": 6.2169106460664665, "grad_norm": 0.8353294730186462, "learning_rate": 1.7636311357198953e-06, "loss": 0.0479, "step": 183890 }, { "epoch": 6.217248723756719, "grad_norm": 0.7765493988990784, "learning_rate": 1.7614554072203638e-06, "loss": 0.0511, "step": 183900 }, { "epoch": 6.217586801446973, "grad_norm": 0.9017364382743835, "learning_rate": 1.7592809975445268e-06, "loss": 0.0523, "step": 183910 }, { "epoch": 6.217924879137226, "grad_norm": 0.7084877490997314, "learning_rate": 1.7571079067518203e-06, "loss": 0.0527, "step": 183920 }, { "epoch": 6.218262956827479, "grad_norm": 0.6820888519287109, "learning_rate": 1.7549361349016635e-06, "loss": 0.0491, "step": 183930 }, { "epoch": 6.218601034517732, "grad_norm": 1.153182864189148, "learning_rate": 1.75276568205342e-06, "loss": 0.0513, "step": 183940 }, { "epoch": 6.218939112207986, "grad_norm": 0.9240016937255859, "learning_rate": 1.7505965482664478e-06, "loss": 0.0494, "step": 183950 }, { "epoch": 6.219277189898238, "grad_norm": 0.5421381592750549, "learning_rate": 1.7484287336000271e-06, "loss": 0.0541, "step": 183960 }, { "epoch": 6.219615267588492, "grad_norm": 0.7682020664215088, "learning_rate": 1.7462622381134496e-06, "loss": 0.0493, "step": 183970 }, { "epoch": 6.219953345278745, "grad_norm": 0.9652904868125916, "learning_rate": 1.7440970618659292e-06, "loss": 0.0523, "step": 183980 }, { "epoch": 6.220291422968998, "grad_norm": 0.5475218892097473, "learning_rate": 1.7419332049166626e-06, "loss": 0.0475, "step": 183990 }, { "epoch": 6.220629500659252, "grad_norm": 0.9395572543144226, "learning_rate": 1.7397706673248194e-06, "loss": 0.0516, "step": 184000 }, { "epoch": 6.220967578349505, "grad_norm": 0.9639958739280701, "learning_rate": 1.7376094491495131e-06, "loss": 0.0522, "step": 184010 }, { "epoch": 6.221305656039758, "grad_norm": 0.9191983938217163, "learning_rate": 1.7354495504498303e-06, "loss": 0.0517, "step": 184020 }, { "epoch": 6.221643733730011, "grad_norm": 1.0835341215133667, "learning_rate": 1.7332909712848233e-06, "loss": 0.0528, "step": 184030 }, { "epoch": 6.221981811420265, "grad_norm": 1.9523298740386963, "learning_rate": 1.7311337117135118e-06, "loss": 0.0586, "step": 184040 }, { "epoch": 6.222319889110517, "grad_norm": 0.949202835559845, "learning_rate": 1.7289777717948763e-06, "loss": 0.0497, "step": 184050 }, { "epoch": 6.222657966800771, "grad_norm": 0.8667237758636475, "learning_rate": 1.7268231515878419e-06, "loss": 0.0537, "step": 184060 }, { "epoch": 6.222996044491024, "grad_norm": 0.9992071390151978, "learning_rate": 1.7246698511513394e-06, "loss": 0.0518, "step": 184070 }, { "epoch": 6.223334122181277, "grad_norm": 1.5384055376052856, "learning_rate": 1.7225178705442158e-06, "loss": 0.0511, "step": 184080 }, { "epoch": 6.2236721998715305, "grad_norm": 0.3697916865348816, "learning_rate": 1.7203672098253244e-06, "loss": 0.0475, "step": 184090 }, { "epoch": 6.224010277561784, "grad_norm": 1.2837125062942505, "learning_rate": 1.7182178690534567e-06, "loss": 0.0585, "step": 184100 }, { "epoch": 6.224348355252037, "grad_norm": 0.4887196123600006, "learning_rate": 1.716069848287366e-06, "loss": 0.0465, "step": 184110 }, { "epoch": 6.22468643294229, "grad_norm": 0.3950463831424713, "learning_rate": 1.7139231475857887e-06, "loss": 0.0525, "step": 184120 }, { "epoch": 6.225024510632544, "grad_norm": 0.9467363953590393, "learning_rate": 1.7117777670074164e-06, "loss": 0.0521, "step": 184130 }, { "epoch": 6.225362588322796, "grad_norm": 0.7736162543296814, "learning_rate": 1.7096337066109026e-06, "loss": 0.0553, "step": 184140 }, { "epoch": 6.22570066601305, "grad_norm": 1.0325348377227783, "learning_rate": 1.70749096645485e-06, "loss": 0.0579, "step": 184150 }, { "epoch": 6.226038743703303, "grad_norm": 0.5789021849632263, "learning_rate": 1.7053495465978508e-06, "loss": 0.0562, "step": 184160 }, { "epoch": 6.226376821393556, "grad_norm": 0.5605279803276062, "learning_rate": 1.7032094470984584e-06, "loss": 0.0558, "step": 184170 }, { "epoch": 6.2267148990838095, "grad_norm": 1.2318466901779175, "learning_rate": 1.7010706680151812e-06, "loss": 0.0558, "step": 184180 }, { "epoch": 6.227052976774063, "grad_norm": 0.9775013327598572, "learning_rate": 1.698933209406478e-06, "loss": 0.0581, "step": 184190 }, { "epoch": 6.227391054464316, "grad_norm": 1.684445858001709, "learning_rate": 1.6967970713307967e-06, "loss": 0.0534, "step": 184200 }, { "epoch": 6.227729132154569, "grad_norm": 1.726082682609558, "learning_rate": 1.6946622538465406e-06, "loss": 0.0471, "step": 184210 }, { "epoch": 6.228067209844823, "grad_norm": 3.4985153675079346, "learning_rate": 1.6925287570120684e-06, "loss": 0.0507, "step": 184220 }, { "epoch": 6.228405287535075, "grad_norm": 7.258920669555664, "learning_rate": 1.6903965808857057e-06, "loss": 0.0515, "step": 184230 }, { "epoch": 6.228743365225329, "grad_norm": 2.0750350952148438, "learning_rate": 1.6882657255257617e-06, "loss": 0.056, "step": 184240 }, { "epoch": 6.229081442915582, "grad_norm": 0.5105446577072144, "learning_rate": 1.6861361909904727e-06, "loss": 0.0507, "step": 184250 }, { "epoch": 6.229419520605835, "grad_norm": 0.6368921995162964, "learning_rate": 1.6840079773380868e-06, "loss": 0.0509, "step": 184260 }, { "epoch": 6.229757598296088, "grad_norm": 1.1296586990356445, "learning_rate": 1.6818810846267518e-06, "loss": 0.0522, "step": 184270 }, { "epoch": 6.230095675986342, "grad_norm": 8.192983627319336, "learning_rate": 1.6797555129146492e-06, "loss": 0.0563, "step": 184280 }, { "epoch": 6.230433753676595, "grad_norm": 0.6245958209037781, "learning_rate": 1.6776312622598656e-06, "loss": 0.0482, "step": 184290 }, { "epoch": 6.230771831366848, "grad_norm": 1.217317819595337, "learning_rate": 1.675508332720499e-06, "loss": 0.0533, "step": 184300 }, { "epoch": 6.231109909057102, "grad_norm": 0.9160075187683105, "learning_rate": 1.673386724354581e-06, "loss": 0.0536, "step": 184310 }, { "epoch": 6.231447986747354, "grad_norm": 2.535438299179077, "learning_rate": 1.6712664372201037e-06, "loss": 0.0472, "step": 184320 }, { "epoch": 6.231786064437608, "grad_norm": 1.0507805347442627, "learning_rate": 1.6691474713750543e-06, "loss": 0.0517, "step": 184330 }, { "epoch": 6.232124142127861, "grad_norm": 0.965094804763794, "learning_rate": 1.6670298268773587e-06, "loss": 0.0511, "step": 184340 }, { "epoch": 6.232462219818114, "grad_norm": 0.6841169595718384, "learning_rate": 1.6649135037848983e-06, "loss": 0.0514, "step": 184350 }, { "epoch": 6.232800297508367, "grad_norm": 1.0904399156570435, "learning_rate": 1.6627985021555547e-06, "loss": 0.0472, "step": 184360 }, { "epoch": 6.233138375198621, "grad_norm": 1.2262505292892456, "learning_rate": 1.6606848220471317e-06, "loss": 0.0504, "step": 184370 }, { "epoch": 6.2334764528888735, "grad_norm": 1.8486136198043823, "learning_rate": 1.6585724635174326e-06, "loss": 0.0539, "step": 184380 }, { "epoch": 6.233814530579127, "grad_norm": 0.6978463530540466, "learning_rate": 1.6564614266242006e-06, "loss": 0.0497, "step": 184390 }, { "epoch": 6.234152608269381, "grad_norm": 0.5745195150375366, "learning_rate": 1.6543517114251505e-06, "loss": 0.0547, "step": 184400 }, { "epoch": 6.234490685959633, "grad_norm": 0.7277806997299194, "learning_rate": 1.6522433179779583e-06, "loss": 0.0475, "step": 184410 }, { "epoch": 6.234828763649887, "grad_norm": 0.8573551177978516, "learning_rate": 1.650136246340278e-06, "loss": 0.0547, "step": 184420 }, { "epoch": 6.23516684134014, "grad_norm": 0.6103766560554504, "learning_rate": 1.6480304965697025e-06, "loss": 0.051, "step": 184430 }, { "epoch": 6.235504919030393, "grad_norm": 0.971655011177063, "learning_rate": 1.6459260687238075e-06, "loss": 0.0528, "step": 184440 }, { "epoch": 6.235842996720646, "grad_norm": 0.9212433099746704, "learning_rate": 1.643822962860131e-06, "loss": 0.0569, "step": 184450 }, { "epoch": 6.2361810744109, "grad_norm": 0.6934645771980286, "learning_rate": 1.6417211790361653e-06, "loss": 0.0496, "step": 184460 }, { "epoch": 6.2365191521011525, "grad_norm": 1.0719718933105469, "learning_rate": 1.6396207173093869e-06, "loss": 0.0557, "step": 184470 }, { "epoch": 6.236857229791406, "grad_norm": 0.5588176846504211, "learning_rate": 1.6375215777371944e-06, "loss": 0.0537, "step": 184480 }, { "epoch": 6.2371953074816595, "grad_norm": 0.7949521541595459, "learning_rate": 1.6354237603769973e-06, "loss": 0.0556, "step": 184490 }, { "epoch": 6.237533385171912, "grad_norm": 1.2613681554794312, "learning_rate": 1.6333272652861498e-06, "loss": 0.0505, "step": 184500 }, { "epoch": 6.237871462862166, "grad_norm": 0.8895339965820312, "learning_rate": 1.6312320925219615e-06, "loss": 0.0529, "step": 184510 }, { "epoch": 6.238209540552419, "grad_norm": 1.0923075675964355, "learning_rate": 1.62913824214172e-06, "loss": 0.0556, "step": 184520 }, { "epoch": 6.238547618242672, "grad_norm": 0.9375727772712708, "learning_rate": 1.6270457142026574e-06, "loss": 0.0512, "step": 184530 }, { "epoch": 6.238885695932925, "grad_norm": 1.4676297903060913, "learning_rate": 1.6249545087619945e-06, "loss": 0.0522, "step": 184540 }, { "epoch": 6.239223773623179, "grad_norm": 0.9899007081985474, "learning_rate": 1.6228646258769132e-06, "loss": 0.0569, "step": 184550 }, { "epoch": 6.239561851313431, "grad_norm": 0.35381758213043213, "learning_rate": 1.6207760656045235e-06, "loss": 0.0483, "step": 184560 }, { "epoch": 6.239899929003685, "grad_norm": 0.6278788447380066, "learning_rate": 1.618688828001952e-06, "loss": 0.0561, "step": 184570 }, { "epoch": 6.2402380066939385, "grad_norm": 1.1157232522964478, "learning_rate": 1.6166029131262418e-06, "loss": 0.053, "step": 184580 }, { "epoch": 6.240576084384191, "grad_norm": 1.7763973474502563, "learning_rate": 1.6145183210344417e-06, "loss": 0.0577, "step": 184590 }, { "epoch": 6.240914162074445, "grad_norm": 3.5030813217163086, "learning_rate": 1.6124350517835286e-06, "loss": 0.054, "step": 184600 }, { "epoch": 6.241252239764698, "grad_norm": 1.0221710205078125, "learning_rate": 1.6103531054304565e-06, "loss": 0.0541, "step": 184610 }, { "epoch": 6.241590317454951, "grad_norm": 0.7126814723014832, "learning_rate": 1.6082724820321638e-06, "loss": 0.0529, "step": 184620 }, { "epoch": 6.241928395145204, "grad_norm": 1.2909029722213745, "learning_rate": 1.6061931816455157e-06, "loss": 0.0585, "step": 184630 }, { "epoch": 6.242266472835458, "grad_norm": 1.2538645267486572, "learning_rate": 1.6041152043273667e-06, "loss": 0.0495, "step": 184640 }, { "epoch": 6.24260455052571, "grad_norm": 1.0841425657272339, "learning_rate": 1.6020385501345215e-06, "loss": 0.0511, "step": 184650 }, { "epoch": 6.242942628215964, "grad_norm": 1.0720927715301514, "learning_rate": 1.5999632191237623e-06, "loss": 0.0508, "step": 184660 }, { "epoch": 6.243280705906217, "grad_norm": 0.7003243565559387, "learning_rate": 1.5978892113518273e-06, "loss": 0.0445, "step": 184670 }, { "epoch": 6.24361878359647, "grad_norm": 1.7711371183395386, "learning_rate": 1.5958165268754156e-06, "loss": 0.05, "step": 184680 }, { "epoch": 6.243956861286724, "grad_norm": 0.7446427941322327, "learning_rate": 1.593745165751198e-06, "loss": 0.0542, "step": 184690 }, { "epoch": 6.244294938976977, "grad_norm": 0.5721840858459473, "learning_rate": 1.5916751280357966e-06, "loss": 0.0505, "step": 184700 }, { "epoch": 6.24463301666723, "grad_norm": 2.1857521533966064, "learning_rate": 1.5896064137858158e-06, "loss": 0.0593, "step": 184710 }, { "epoch": 6.244971094357483, "grad_norm": 6.180686950683594, "learning_rate": 1.5875390230578103e-06, "loss": 0.0501, "step": 184720 }, { "epoch": 6.245309172047737, "grad_norm": 1.0727370977401733, "learning_rate": 1.5854729559082958e-06, "loss": 0.0532, "step": 184730 }, { "epoch": 6.24564724973799, "grad_norm": 0.9299927949905396, "learning_rate": 1.5834082123937667e-06, "loss": 0.0522, "step": 184740 }, { "epoch": 6.245985327428243, "grad_norm": 0.6949784755706787, "learning_rate": 1.5813447925706604e-06, "loss": 0.0476, "step": 184750 }, { "epoch": 6.246323405118496, "grad_norm": 1.3275922536849976, "learning_rate": 1.5792826964954099e-06, "loss": 0.0546, "step": 184760 }, { "epoch": 6.246661482808749, "grad_norm": 0.572039783000946, "learning_rate": 1.5772219242243702e-06, "loss": 0.0532, "step": 184770 }, { "epoch": 6.2469995604990025, "grad_norm": 0.7590565085411072, "learning_rate": 1.575162475813896e-06, "loss": 0.0568, "step": 184780 }, { "epoch": 6.247337638189256, "grad_norm": 1.1619808673858643, "learning_rate": 1.5731043513202814e-06, "loss": 0.052, "step": 184790 }, { "epoch": 6.2476757158795095, "grad_norm": 0.4872724115848541, "learning_rate": 1.5710475507998034e-06, "loss": 0.0472, "step": 184800 }, { "epoch": 6.248013793569762, "grad_norm": 0.8758885264396667, "learning_rate": 1.568992074308695e-06, "loss": 0.0511, "step": 184810 }, { "epoch": 6.248351871260016, "grad_norm": 0.41252681612968445, "learning_rate": 1.5669379219031444e-06, "loss": 0.0481, "step": 184820 }, { "epoch": 6.248689948950269, "grad_norm": 0.7241824269294739, "learning_rate": 1.5648850936393179e-06, "loss": 0.053, "step": 184830 }, { "epoch": 6.249028026640522, "grad_norm": 1.4629522562026978, "learning_rate": 1.562833589573337e-06, "loss": 0.0497, "step": 184840 }, { "epoch": 6.249366104330775, "grad_norm": 2.883695602416992, "learning_rate": 1.5607834097612906e-06, "loss": 0.051, "step": 184850 }, { "epoch": 6.249704182021029, "grad_norm": 0.5993130803108215, "learning_rate": 1.5587345542592223e-06, "loss": 0.0544, "step": 184860 }, { "epoch": 6.2500422597112815, "grad_norm": 0.6630851626396179, "learning_rate": 1.556687023123149e-06, "loss": 0.0509, "step": 184870 }, { "epoch": 6.250380337401535, "grad_norm": 1.1858911514282227, "learning_rate": 1.5546408164090641e-06, "loss": 0.0493, "step": 184880 }, { "epoch": 6.2507184150917885, "grad_norm": 2.449842929840088, "learning_rate": 1.5525959341728958e-06, "loss": 0.0549, "step": 184890 }, { "epoch": 6.251056492782041, "grad_norm": 0.7463679313659668, "learning_rate": 1.5505523764705542e-06, "loss": 0.0606, "step": 184900 }, { "epoch": 6.251394570472295, "grad_norm": 0.9189892411231995, "learning_rate": 1.5485101433579063e-06, "loss": 0.0481, "step": 184910 }, { "epoch": 6.251732648162548, "grad_norm": 0.5532091856002808, "learning_rate": 1.5464692348907906e-06, "loss": 0.0498, "step": 184920 }, { "epoch": 6.252070725852801, "grad_norm": 0.781971275806427, "learning_rate": 1.5444296511250012e-06, "loss": 0.0593, "step": 184930 }, { "epoch": 6.252408803543054, "grad_norm": 0.834518313407898, "learning_rate": 1.542391392116299e-06, "loss": 0.0527, "step": 184940 }, { "epoch": 6.252746881233308, "grad_norm": 0.5816661715507507, "learning_rate": 1.5403544579204177e-06, "loss": 0.0573, "step": 184950 }, { "epoch": 6.25308495892356, "grad_norm": 1.0831321477890015, "learning_rate": 1.5383188485930344e-06, "loss": 0.0477, "step": 184960 }, { "epoch": 6.253423036613814, "grad_norm": 0.6463578939437866, "learning_rate": 1.536284564189816e-06, "loss": 0.057, "step": 184970 }, { "epoch": 6.2537611143040674, "grad_norm": 1.1543976068496704, "learning_rate": 1.5342516047663624e-06, "loss": 0.0542, "step": 184980 }, { "epoch": 6.25409919199432, "grad_norm": 0.498735249042511, "learning_rate": 1.5322199703782624e-06, "loss": 0.0503, "step": 184990 }, { "epoch": 6.254437269684574, "grad_norm": 1.1584267616271973, "learning_rate": 1.5301896610810607e-06, "loss": 0.0558, "step": 185000 }, { "epoch": 6.254775347374827, "grad_norm": 1.496731162071228, "learning_rate": 1.528160676930268e-06, "loss": 0.0488, "step": 185010 }, { "epoch": 6.25511342506508, "grad_norm": 0.6868177056312561, "learning_rate": 1.5261330179813515e-06, "loss": 0.0569, "step": 185020 }, { "epoch": 6.255451502755333, "grad_norm": 0.8010616898536682, "learning_rate": 1.5241066842897444e-06, "loss": 0.0473, "step": 185030 }, { "epoch": 6.255789580445587, "grad_norm": 0.8347423076629639, "learning_rate": 1.5220816759108526e-06, "loss": 0.0492, "step": 185040 }, { "epoch": 6.256127658135839, "grad_norm": 0.7163659930229187, "learning_rate": 1.5200579929000368e-06, "loss": 0.05, "step": 185050 }, { "epoch": 6.256465735826093, "grad_norm": 1.4834634065628052, "learning_rate": 1.5180356353126146e-06, "loss": 0.0578, "step": 185060 }, { "epoch": 6.256803813516346, "grad_norm": 0.5904650092124939, "learning_rate": 1.5160146032038913e-06, "loss": 0.0519, "step": 185070 }, { "epoch": 6.257141891206599, "grad_norm": 1.3694332838058472, "learning_rate": 1.5139948966291061e-06, "loss": 0.0522, "step": 185080 }, { "epoch": 6.2574799688968525, "grad_norm": 0.9627716541290283, "learning_rate": 1.5119765156434928e-06, "loss": 0.0526, "step": 185090 }, { "epoch": 6.257818046587106, "grad_norm": 0.8617656230926514, "learning_rate": 1.5099594603022292e-06, "loss": 0.059, "step": 185100 }, { "epoch": 6.258156124277359, "grad_norm": 0.7214646935462952, "learning_rate": 1.5079437306604494e-06, "loss": 0.0488, "step": 185110 }, { "epoch": 6.258494201967612, "grad_norm": 1.1667290925979614, "learning_rate": 1.5059293267732699e-06, "loss": 0.0527, "step": 185120 }, { "epoch": 6.258832279657866, "grad_norm": 1.790575623512268, "learning_rate": 1.5039162486957691e-06, "loss": 0.0518, "step": 185130 }, { "epoch": 6.259170357348118, "grad_norm": 0.5917167663574219, "learning_rate": 1.501904496482981e-06, "loss": 0.0554, "step": 185140 }, { "epoch": 6.259508435038372, "grad_norm": 1.0722899436950684, "learning_rate": 1.4998940701898945e-06, "loss": 0.056, "step": 185150 }, { "epoch": 6.259846512728625, "grad_norm": 0.5145789384841919, "learning_rate": 1.4978849698714882e-06, "loss": 0.0518, "step": 185160 }, { "epoch": 6.260184590418878, "grad_norm": 0.5808742642402649, "learning_rate": 1.4958771955826845e-06, "loss": 0.0489, "step": 185170 }, { "epoch": 6.2605226681091315, "grad_norm": 1.255885362625122, "learning_rate": 1.4938707473783786e-06, "loss": 0.0595, "step": 185180 }, { "epoch": 6.260860745799385, "grad_norm": 0.6713712811470032, "learning_rate": 1.4918656253134267e-06, "loss": 0.0526, "step": 185190 }, { "epoch": 6.261198823489638, "grad_norm": 1.1377512216567993, "learning_rate": 1.4898618294426347e-06, "loss": 0.0475, "step": 185200 }, { "epoch": 6.261536901179891, "grad_norm": 0.9726547002792358, "learning_rate": 1.4878593598208036e-06, "loss": 0.0512, "step": 185210 }, { "epoch": 6.261874978870145, "grad_norm": 0.5121480822563171, "learning_rate": 1.4858582165026724e-06, "loss": 0.06, "step": 185220 }, { "epoch": 6.262213056560397, "grad_norm": 2.1988089084625244, "learning_rate": 1.4838583995429533e-06, "loss": 0.0472, "step": 185230 }, { "epoch": 6.262551134250651, "grad_norm": 0.536404550075531, "learning_rate": 1.4818599089963136e-06, "loss": 0.0501, "step": 185240 }, { "epoch": 6.262889211940904, "grad_norm": 0.7689090967178345, "learning_rate": 1.479862744917393e-06, "loss": 0.0542, "step": 185250 }, { "epoch": 6.263227289631157, "grad_norm": 1.7325074672698975, "learning_rate": 1.4778669073608087e-06, "loss": 0.0494, "step": 185260 }, { "epoch": 6.2635653673214104, "grad_norm": 0.5686958432197571, "learning_rate": 1.4758723963811061e-06, "loss": 0.0483, "step": 185270 }, { "epoch": 6.263903445011664, "grad_norm": 0.638749897480011, "learning_rate": 1.4738792120328248e-06, "loss": 0.0571, "step": 185280 }, { "epoch": 6.264241522701917, "grad_norm": 3.1622138023376465, "learning_rate": 1.4718873543704549e-06, "loss": 0.0511, "step": 185290 }, { "epoch": 6.26457960039217, "grad_norm": 0.6309669613838196, "learning_rate": 1.4698968234484522e-06, "loss": 0.0453, "step": 185300 }, { "epoch": 6.264917678082424, "grad_norm": 2.800731897354126, "learning_rate": 1.4679076193212404e-06, "loss": 0.0512, "step": 185310 }, { "epoch": 6.265255755772676, "grad_norm": 0.8084607124328613, "learning_rate": 1.465919742043198e-06, "loss": 0.0497, "step": 185320 }, { "epoch": 6.26559383346293, "grad_norm": 1.6437416076660156, "learning_rate": 1.4639331916686816e-06, "loss": 0.0556, "step": 185330 }, { "epoch": 6.265931911153183, "grad_norm": 0.9185938835144043, "learning_rate": 1.4619479682519977e-06, "loss": 0.051, "step": 185340 }, { "epoch": 6.266269988843436, "grad_norm": 0.6730356812477112, "learning_rate": 1.4599640718474194e-06, "loss": 0.0482, "step": 185350 }, { "epoch": 6.266608066533689, "grad_norm": 0.8287341594696045, "learning_rate": 1.457981502509187e-06, "loss": 0.0543, "step": 185360 }, { "epoch": 6.266946144223943, "grad_norm": 1.450772762298584, "learning_rate": 1.4560002602915013e-06, "loss": 0.0513, "step": 185370 }, { "epoch": 6.2672842219141955, "grad_norm": 0.6955794095993042, "learning_rate": 1.4540203452485357e-06, "loss": 0.0493, "step": 185380 }, { "epoch": 6.267622299604449, "grad_norm": 0.8818496465682983, "learning_rate": 1.4520417574344135e-06, "loss": 0.0521, "step": 185390 }, { "epoch": 6.267960377294703, "grad_norm": 1.0132980346679688, "learning_rate": 1.4500644969032362e-06, "loss": 0.0536, "step": 185400 }, { "epoch": 6.268298454984955, "grad_norm": 1.2320562601089478, "learning_rate": 1.448088563709049e-06, "loss": 0.0549, "step": 185410 }, { "epoch": 6.268636532675209, "grad_norm": 1.1374152898788452, "learning_rate": 1.4461139579058868e-06, "loss": 0.0503, "step": 185420 }, { "epoch": 6.268974610365462, "grad_norm": 0.7962634563446045, "learning_rate": 1.444140679547723e-06, "loss": 0.0532, "step": 185430 }, { "epoch": 6.269312688055715, "grad_norm": 0.49959900975227356, "learning_rate": 1.4421687286885143e-06, "loss": 0.0539, "step": 185440 }, { "epoch": 6.269650765745968, "grad_norm": 2.541229724884033, "learning_rate": 1.4401981053821677e-06, "loss": 0.0512, "step": 185450 }, { "epoch": 6.269988843436222, "grad_norm": 2.249781608581543, "learning_rate": 1.4382288096825625e-06, "loss": 0.0574, "step": 185460 }, { "epoch": 6.2703269211264745, "grad_norm": 1.5950905084609985, "learning_rate": 1.4362608416435441e-06, "loss": 0.0539, "step": 185470 }, { "epoch": 6.270664998816728, "grad_norm": 1.193993330001831, "learning_rate": 1.434294201318903e-06, "loss": 0.0519, "step": 185480 }, { "epoch": 6.2710030765069815, "grad_norm": 3.9572722911834717, "learning_rate": 1.432328888762413e-06, "loss": 0.0597, "step": 185490 }, { "epoch": 6.271341154197234, "grad_norm": 0.6159037947654724, "learning_rate": 1.4303649040278088e-06, "loss": 0.0544, "step": 185500 }, { "epoch": 6.271679231887488, "grad_norm": 0.6935245990753174, "learning_rate": 1.4284022471687808e-06, "loss": 0.0523, "step": 185510 }, { "epoch": 6.272017309577741, "grad_norm": 0.906400740146637, "learning_rate": 1.426440918238986e-06, "loss": 0.0582, "step": 185520 }, { "epoch": 6.272355387267994, "grad_norm": 1.1575419902801514, "learning_rate": 1.4244809172920426e-06, "loss": 0.0532, "step": 185530 }, { "epoch": 6.272693464958247, "grad_norm": 1.407778263092041, "learning_rate": 1.4225222443815523e-06, "loss": 0.054, "step": 185540 }, { "epoch": 6.273031542648501, "grad_norm": 0.8489707708358765, "learning_rate": 1.42056489956105e-06, "loss": 0.0553, "step": 185550 }, { "epoch": 6.2733696203387534, "grad_norm": 0.6072512865066528, "learning_rate": 1.4186088828840482e-06, "loss": 0.0469, "step": 185560 }, { "epoch": 6.273707698029007, "grad_norm": 1.5175807476043701, "learning_rate": 1.4166541944040379e-06, "loss": 0.0502, "step": 185570 }, { "epoch": 6.2740457757192605, "grad_norm": 1.4659987688064575, "learning_rate": 1.4147008341744427e-06, "loss": 0.0536, "step": 185580 }, { "epoch": 6.274383853409513, "grad_norm": 0.8883557319641113, "learning_rate": 1.4127488022486756e-06, "loss": 0.0566, "step": 185590 }, { "epoch": 6.274721931099767, "grad_norm": 0.754587709903717, "learning_rate": 1.410798098680105e-06, "loss": 0.0574, "step": 185600 }, { "epoch": 6.27506000879002, "grad_norm": 0.8502720594406128, "learning_rate": 1.4088487235220605e-06, "loss": 0.057, "step": 185610 }, { "epoch": 6.275398086480273, "grad_norm": 0.3981549143791199, "learning_rate": 1.4069006768278326e-06, "loss": 0.0513, "step": 185620 }, { "epoch": 6.275736164170526, "grad_norm": 0.6419553756713867, "learning_rate": 1.4049539586506898e-06, "loss": 0.0485, "step": 185630 }, { "epoch": 6.27607424186078, "grad_norm": 0.5964678525924683, "learning_rate": 1.4030085690438511e-06, "loss": 0.0532, "step": 185640 }, { "epoch": 6.276412319551033, "grad_norm": 1.2256397008895874, "learning_rate": 1.40106450806049e-06, "loss": 0.0579, "step": 185650 }, { "epoch": 6.276750397241286, "grad_norm": 0.8507282137870789, "learning_rate": 1.3991217757537812e-06, "loss": 0.0541, "step": 185660 }, { "epoch": 6.277088474931539, "grad_norm": 0.59525066614151, "learning_rate": 1.3971803721768095e-06, "loss": 0.047, "step": 185670 }, { "epoch": 6.277426552621792, "grad_norm": 1.3036320209503174, "learning_rate": 1.3952402973826829e-06, "loss": 0.0525, "step": 185680 }, { "epoch": 6.277764630312046, "grad_norm": 0.8064653873443604, "learning_rate": 1.3933015514244197e-06, "loss": 0.051, "step": 185690 }, { "epoch": 6.278102708002299, "grad_norm": 0.5583375096321106, "learning_rate": 1.3913641343550276e-06, "loss": 0.0559, "step": 185700 }, { "epoch": 6.278440785692553, "grad_norm": 0.7397626042366028, "learning_rate": 1.3894280462274812e-06, "loss": 0.0509, "step": 185710 }, { "epoch": 6.278778863382805, "grad_norm": 1.0340198278427124, "learning_rate": 1.3874932870947154e-06, "loss": 0.0488, "step": 185720 }, { "epoch": 6.279116941073059, "grad_norm": 0.9228925704956055, "learning_rate": 1.3855598570096162e-06, "loss": 0.0548, "step": 185730 }, { "epoch": 6.279455018763311, "grad_norm": 0.6256652474403381, "learning_rate": 1.3836277560250411e-06, "loss": 0.0518, "step": 185740 }, { "epoch": 6.279793096453565, "grad_norm": 1.6614407300949097, "learning_rate": 1.3816969841938199e-06, "loss": 0.051, "step": 185750 }, { "epoch": 6.280131174143818, "grad_norm": 0.5073131918907166, "learning_rate": 1.3797675415687494e-06, "loss": 0.0567, "step": 185760 }, { "epoch": 6.280469251834072, "grad_norm": 0.7665131688117981, "learning_rate": 1.3778394282025597e-06, "loss": 0.0514, "step": 185770 }, { "epoch": 6.2808073295243245, "grad_norm": 0.7136300802230835, "learning_rate": 1.3759126441479753e-06, "loss": 0.0481, "step": 185780 }, { "epoch": 6.281145407214578, "grad_norm": 2.6970317363739014, "learning_rate": 1.3739871894576706e-06, "loss": 0.0555, "step": 185790 }, { "epoch": 6.281483484904831, "grad_norm": 17.20843505859375, "learning_rate": 1.3720630641842924e-06, "loss": 0.0586, "step": 185800 }, { "epoch": 6.281821562595084, "grad_norm": 1.3165031671524048, "learning_rate": 1.3701402683804376e-06, "loss": 0.0404, "step": 185810 }, { "epoch": 6.282159640285338, "grad_norm": 0.6288724541664124, "learning_rate": 1.3682188020986752e-06, "loss": 0.0497, "step": 185820 }, { "epoch": 6.282497717975591, "grad_norm": 0.5951597690582275, "learning_rate": 1.3662986653915466e-06, "loss": 0.0529, "step": 185830 }, { "epoch": 6.282835795665844, "grad_norm": 1.0786988735198975, "learning_rate": 1.364379858311543e-06, "loss": 0.0522, "step": 185840 }, { "epoch": 6.283173873356097, "grad_norm": 2.9620249271392822, "learning_rate": 1.3624623809111226e-06, "loss": 0.0515, "step": 185850 }, { "epoch": 6.283511951046351, "grad_norm": 2.1180739402770996, "learning_rate": 1.360546233242699e-06, "loss": 0.0571, "step": 185860 }, { "epoch": 6.2838500287366035, "grad_norm": 1.5694748163223267, "learning_rate": 1.3586314153586798e-06, "loss": 0.0515, "step": 185870 }, { "epoch": 6.284188106426857, "grad_norm": 3.685896396636963, "learning_rate": 1.3567179273113962e-06, "loss": 0.0502, "step": 185880 }, { "epoch": 6.2845261841171105, "grad_norm": 2.7284090518951416, "learning_rate": 1.3548057691531724e-06, "loss": 0.057, "step": 185890 }, { "epoch": 6.284864261807363, "grad_norm": 1.1782376766204834, "learning_rate": 1.3528949409362834e-06, "loss": 0.055, "step": 185900 }, { "epoch": 6.285202339497617, "grad_norm": 1.2841660976409912, "learning_rate": 1.350985442712971e-06, "loss": 0.0524, "step": 185910 }, { "epoch": 6.28554041718787, "grad_norm": 0.6396021842956543, "learning_rate": 1.349077274535443e-06, "loss": 0.0569, "step": 185920 }, { "epoch": 6.285878494878123, "grad_norm": 2.3684446811676025, "learning_rate": 1.3471704364558635e-06, "loss": 0.0618, "step": 185930 }, { "epoch": 6.286216572568376, "grad_norm": 1.023138403892517, "learning_rate": 1.345264928526363e-06, "loss": 0.048, "step": 185940 }, { "epoch": 6.28655465025863, "grad_norm": 0.5019248127937317, "learning_rate": 1.3433607507990441e-06, "loss": 0.05, "step": 185950 }, { "epoch": 6.286892727948882, "grad_norm": 0.6813050508499146, "learning_rate": 1.3414579033259544e-06, "loss": 0.0503, "step": 185960 }, { "epoch": 6.287230805639136, "grad_norm": 2.877772569656372, "learning_rate": 1.339556386159141e-06, "loss": 0.0572, "step": 185970 }, { "epoch": 6.2875688833293895, "grad_norm": 0.9200043082237244, "learning_rate": 1.3376561993505621e-06, "loss": 0.051, "step": 185980 }, { "epoch": 6.287906961019642, "grad_norm": 0.6870385408401489, "learning_rate": 1.3357573429521875e-06, "loss": 0.0546, "step": 185990 }, { "epoch": 6.288245038709896, "grad_norm": 0.9037074446678162, "learning_rate": 1.33385981701592e-06, "loss": 0.0559, "step": 186000 }, { "epoch": 6.288583116400149, "grad_norm": 0.7217369079589844, "learning_rate": 1.331963621593646e-06, "loss": 0.0542, "step": 186010 }, { "epoch": 6.288921194090402, "grad_norm": 0.7720704078674316, "learning_rate": 1.3300687567372072e-06, "loss": 0.0563, "step": 186020 }, { "epoch": 6.289259271780655, "grad_norm": 0.7703233957290649, "learning_rate": 1.3281752224983956e-06, "loss": 0.0496, "step": 186030 }, { "epoch": 6.289597349470909, "grad_norm": 0.6388902068138123, "learning_rate": 1.326283018928992e-06, "loss": 0.0545, "step": 186040 }, { "epoch": 6.289935427161161, "grad_norm": 1.2962720394134521, "learning_rate": 1.3243921460807274e-06, "loss": 0.0479, "step": 186050 }, { "epoch": 6.290273504851415, "grad_norm": 0.3475991189479828, "learning_rate": 1.3225026040052935e-06, "loss": 0.0536, "step": 186060 }, { "epoch": 6.290611582541668, "grad_norm": 1.1752121448516846, "learning_rate": 1.3206143927543434e-06, "loss": 0.0472, "step": 186070 }, { "epoch": 6.290949660231921, "grad_norm": 0.7970879077911377, "learning_rate": 1.318727512379514e-06, "loss": 0.0556, "step": 186080 }, { "epoch": 6.2912877379221745, "grad_norm": 0.6041489243507385, "learning_rate": 1.316841962932386e-06, "loss": 0.0515, "step": 186090 }, { "epoch": 6.291625815612428, "grad_norm": 1.0451810359954834, "learning_rate": 1.3149577444645122e-06, "loss": 0.0578, "step": 186100 }, { "epoch": 6.291963893302681, "grad_norm": 0.7534788250923157, "learning_rate": 1.313074857027402e-06, "loss": 0.0478, "step": 186110 }, { "epoch": 6.292301970992934, "grad_norm": 0.569382905960083, "learning_rate": 1.3111933006725252e-06, "loss": 0.0495, "step": 186120 }, { "epoch": 6.292640048683188, "grad_norm": 0.7170382738113403, "learning_rate": 1.3093130754513462e-06, "loss": 0.0516, "step": 186130 }, { "epoch": 6.29297812637344, "grad_norm": 0.8830835819244385, "learning_rate": 1.3074341814152458e-06, "loss": 0.057, "step": 186140 }, { "epoch": 6.293316204063694, "grad_norm": 0.784534752368927, "learning_rate": 1.3055566186156053e-06, "loss": 0.0529, "step": 186150 }, { "epoch": 6.293654281753947, "grad_norm": 1.5542166233062744, "learning_rate": 1.3036803871037505e-06, "loss": 0.0515, "step": 186160 }, { "epoch": 6.2939923594442, "grad_norm": 1.3759313821792603, "learning_rate": 1.3018054869309792e-06, "loss": 0.0538, "step": 186170 }, { "epoch": 6.2943304371344535, "grad_norm": 0.6060729622840881, "learning_rate": 1.2999319181485613e-06, "loss": 0.0541, "step": 186180 }, { "epoch": 6.294668514824707, "grad_norm": 0.5167579650878906, "learning_rate": 1.2980596808077005e-06, "loss": 0.0588, "step": 186190 }, { "epoch": 6.29500659251496, "grad_norm": 0.8593753576278687, "learning_rate": 1.2961887749595891e-06, "loss": 0.0511, "step": 186200 }, { "epoch": 6.295344670205213, "grad_norm": 0.7243024110794067, "learning_rate": 1.2943192006553805e-06, "loss": 0.0512, "step": 186210 }, { "epoch": 6.295682747895467, "grad_norm": 0.9234288334846497, "learning_rate": 1.2924509579461953e-06, "loss": 0.0581, "step": 186220 }, { "epoch": 6.296020825585719, "grad_norm": 0.9999457597732544, "learning_rate": 1.2905840468830976e-06, "loss": 0.053, "step": 186230 }, { "epoch": 6.296358903275973, "grad_norm": 1.3408414125442505, "learning_rate": 1.2887184675171249e-06, "loss": 0.0542, "step": 186240 }, { "epoch": 6.296696980966226, "grad_norm": 0.6423962116241455, "learning_rate": 1.2868542198992972e-06, "loss": 0.054, "step": 186250 }, { "epoch": 6.297035058656479, "grad_norm": 0.9355167150497437, "learning_rate": 1.2849913040805738e-06, "loss": 0.0508, "step": 186260 }, { "epoch": 6.2973731363467325, "grad_norm": 0.9713393449783325, "learning_rate": 1.2831297201118863e-06, "loss": 0.0555, "step": 186270 }, { "epoch": 6.297711214036986, "grad_norm": 3.4311084747314453, "learning_rate": 1.2812694680441329e-06, "loss": 0.0505, "step": 186280 }, { "epoch": 6.298049291727239, "grad_norm": 34.1024055480957, "learning_rate": 1.2794105479281616e-06, "loss": 0.0529, "step": 186290 }, { "epoch": 6.298387369417492, "grad_norm": 0.6835463047027588, "learning_rate": 1.2775529598148095e-06, "loss": 0.0437, "step": 186300 }, { "epoch": 6.298725447107746, "grad_norm": 0.559886634349823, "learning_rate": 1.2756967037548528e-06, "loss": 0.0531, "step": 186310 }, { "epoch": 6.299063524797998, "grad_norm": 0.6122680902481079, "learning_rate": 1.2738417797990399e-06, "loss": 0.0536, "step": 186320 }, { "epoch": 6.299401602488252, "grad_norm": 0.6376776695251465, "learning_rate": 1.2719881879980967e-06, "loss": 0.0494, "step": 186330 }, { "epoch": 6.299739680178505, "grad_norm": 0.9116863012313843, "learning_rate": 1.2701359284026827e-06, "loss": 0.0519, "step": 186340 }, { "epoch": 6.300077757868758, "grad_norm": 0.5702790021896362, "learning_rate": 1.2682850010634518e-06, "loss": 0.054, "step": 186350 }, { "epoch": 6.300415835559011, "grad_norm": 3.7070436477661133, "learning_rate": 1.2664354060309914e-06, "loss": 0.0512, "step": 186360 }, { "epoch": 6.300753913249265, "grad_norm": 3.2367517948150635, "learning_rate": 1.2645871433558832e-06, "loss": 0.0457, "step": 186370 }, { "epoch": 6.3010919909395176, "grad_norm": 0.592043936252594, "learning_rate": 1.2627402130886535e-06, "loss": 0.0539, "step": 186380 }, { "epoch": 6.301430068629771, "grad_norm": 0.7918109893798828, "learning_rate": 1.2608946152798007e-06, "loss": 0.0548, "step": 186390 }, { "epoch": 6.301768146320025, "grad_norm": 0.7079563736915588, "learning_rate": 1.2590503499797735e-06, "loss": 0.0561, "step": 186400 }, { "epoch": 6.302106224010277, "grad_norm": 0.9930834770202637, "learning_rate": 1.257207417238998e-06, "loss": 0.0507, "step": 186410 }, { "epoch": 6.302444301700531, "grad_norm": 1.182485580444336, "learning_rate": 1.2553658171078674e-06, "loss": 0.0503, "step": 186420 }, { "epoch": 6.302782379390784, "grad_norm": 1.2893826961517334, "learning_rate": 1.2535255496367193e-06, "loss": 0.0542, "step": 186430 }, { "epoch": 6.303120457081037, "grad_norm": 1.1989878416061401, "learning_rate": 1.2516866148758689e-06, "loss": 0.0533, "step": 186440 }, { "epoch": 6.30345853477129, "grad_norm": 3.1021296977996826, "learning_rate": 1.2498490128755869e-06, "loss": 0.0603, "step": 186450 }, { "epoch": 6.303796612461544, "grad_norm": 1.2545562982559204, "learning_rate": 1.2480127436861222e-06, "loss": 0.0572, "step": 186460 }, { "epoch": 6.3041346901517965, "grad_norm": 2.1075661182403564, "learning_rate": 1.246177807357679e-06, "loss": 0.0473, "step": 186470 }, { "epoch": 6.30447276784205, "grad_norm": 0.9056797027587891, "learning_rate": 1.2443442039404118e-06, "loss": 0.0591, "step": 186480 }, { "epoch": 6.3048108455323035, "grad_norm": 1.4512649774551392, "learning_rate": 1.2425119334844581e-06, "loss": 0.0502, "step": 186490 }, { "epoch": 6.305148923222556, "grad_norm": 0.7983725666999817, "learning_rate": 1.2406809960399113e-06, "loss": 0.0556, "step": 186500 }, { "epoch": 6.30548700091281, "grad_norm": 0.8903429508209229, "learning_rate": 1.2388513916568311e-06, "loss": 0.0504, "step": 186510 }, { "epoch": 6.305825078603063, "grad_norm": 1.9600547552108765, "learning_rate": 1.2370231203852333e-06, "loss": 0.0495, "step": 186520 }, { "epoch": 6.306163156293316, "grad_norm": 0.5023627877235413, "learning_rate": 1.2351961822751057e-06, "loss": 0.0514, "step": 186530 }, { "epoch": 6.306501233983569, "grad_norm": 1.9376780986785889, "learning_rate": 1.2333705773763916e-06, "loss": 0.0485, "step": 186540 }, { "epoch": 6.306839311673823, "grad_norm": 0.5882744193077087, "learning_rate": 1.2315463057390065e-06, "loss": 0.0531, "step": 186550 }, { "epoch": 6.3071773893640755, "grad_norm": 1.5235437154769897, "learning_rate": 1.2297233674128271e-06, "loss": 0.051, "step": 186560 }, { "epoch": 6.307515467054329, "grad_norm": 0.8115285634994507, "learning_rate": 1.2279017624476808e-06, "loss": 0.0533, "step": 186570 }, { "epoch": 6.3078535447445825, "grad_norm": 0.9819762110710144, "learning_rate": 1.2260814908933826e-06, "loss": 0.0479, "step": 186580 }, { "epoch": 6.308191622434835, "grad_norm": 1.1403907537460327, "learning_rate": 1.224262552799693e-06, "loss": 0.0515, "step": 186590 }, { "epoch": 6.308529700125089, "grad_norm": 0.5775164365768433, "learning_rate": 1.2224449482163391e-06, "loss": 0.0477, "step": 186600 }, { "epoch": 6.308867777815342, "grad_norm": 0.5464187264442444, "learning_rate": 1.2206286771930197e-06, "loss": 0.0549, "step": 186610 }, { "epoch": 6.309205855505596, "grad_norm": 0.8120267391204834, "learning_rate": 1.2188137397793786e-06, "loss": 0.0503, "step": 186620 }, { "epoch": 6.309543933195848, "grad_norm": 1.5004515647888184, "learning_rate": 1.2170001360250538e-06, "loss": 0.0488, "step": 186630 }, { "epoch": 6.309882010886102, "grad_norm": 0.5310854315757751, "learning_rate": 1.2151878659796111e-06, "loss": 0.0523, "step": 186640 }, { "epoch": 6.310220088576354, "grad_norm": 0.9634395241737366, "learning_rate": 1.2133769296926057e-06, "loss": 0.0563, "step": 186650 }, { "epoch": 6.310558166266608, "grad_norm": 0.7189825773239136, "learning_rate": 1.2115673272135475e-06, "loss": 0.053, "step": 186660 }, { "epoch": 6.310896243956861, "grad_norm": 0.8646373152732849, "learning_rate": 1.2097590585919083e-06, "loss": 0.0533, "step": 186670 }, { "epoch": 6.311234321647115, "grad_norm": 0.7527883052825928, "learning_rate": 1.2079521238771373e-06, "loss": 0.0574, "step": 186680 }, { "epoch": 6.311572399337368, "grad_norm": 1.3544063568115234, "learning_rate": 1.206146523118612e-06, "loss": 0.0476, "step": 186690 }, { "epoch": 6.311910477027621, "grad_norm": 0.8068625330924988, "learning_rate": 1.2043422563657093e-06, "loss": 0.047, "step": 186700 }, { "epoch": 6.312248554717874, "grad_norm": 1.7864677906036377, "learning_rate": 1.2025393236677618e-06, "loss": 0.0529, "step": 186710 }, { "epoch": 6.312586632408127, "grad_norm": 1.5204012393951416, "learning_rate": 1.2007377250740582e-06, "loss": 0.0542, "step": 186720 }, { "epoch": 6.312924710098381, "grad_norm": 0.7244189381599426, "learning_rate": 1.1989374606338532e-06, "loss": 0.0533, "step": 186730 }, { "epoch": 6.313262787788634, "grad_norm": 0.6284661293029785, "learning_rate": 1.1971385303963633e-06, "loss": 0.0475, "step": 186740 }, { "epoch": 6.313600865478887, "grad_norm": 0.9649743437767029, "learning_rate": 1.1953409344107712e-06, "loss": 0.0543, "step": 186750 }, { "epoch": 6.31393894316914, "grad_norm": 0.8017266392707825, "learning_rate": 1.1935446727262212e-06, "loss": 0.0541, "step": 186760 }, { "epoch": 6.314277020859393, "grad_norm": 0.9115057587623596, "learning_rate": 1.1917497453918236e-06, "loss": 0.0527, "step": 186770 }, { "epoch": 6.3146150985496465, "grad_norm": 1.6026902198791504, "learning_rate": 1.189956152456656e-06, "loss": 0.052, "step": 186780 }, { "epoch": 6.3149531762399, "grad_norm": 0.6904276609420776, "learning_rate": 1.1881638939697404e-06, "loss": 0.0509, "step": 186790 }, { "epoch": 6.315291253930154, "grad_norm": 0.915697455406189, "learning_rate": 1.1863729699800985e-06, "loss": 0.0486, "step": 186800 }, { "epoch": 6.315629331620406, "grad_norm": 0.7697750329971313, "learning_rate": 1.1845833805366747e-06, "loss": 0.0552, "step": 186810 }, { "epoch": 6.31596740931066, "grad_norm": 0.6055203080177307, "learning_rate": 1.1827951256884073e-06, "loss": 0.0456, "step": 186820 }, { "epoch": 6.316305487000913, "grad_norm": 0.5820989608764648, "learning_rate": 1.1810082054841742e-06, "loss": 0.0468, "step": 186830 }, { "epoch": 6.316643564691166, "grad_norm": 0.5126862525939941, "learning_rate": 1.1792226199728418e-06, "loss": 0.0557, "step": 186840 }, { "epoch": 6.316981642381419, "grad_norm": 3.749297618865967, "learning_rate": 1.1774383692032209e-06, "loss": 0.0527, "step": 186850 }, { "epoch": 6.317319720071673, "grad_norm": 0.8586016297340393, "learning_rate": 1.1756554532240837e-06, "loss": 0.057, "step": 186860 }, { "epoch": 6.3176577977619255, "grad_norm": 1.0257619619369507, "learning_rate": 1.173873872084197e-06, "loss": 0.0488, "step": 186870 }, { "epoch": 6.317995875452179, "grad_norm": 0.8021496534347534, "learning_rate": 1.1720936258322436e-06, "loss": 0.0474, "step": 186880 }, { "epoch": 6.3183339531424325, "grad_norm": 2.6233444213867188, "learning_rate": 1.1703147145169235e-06, "loss": 0.05, "step": 186890 }, { "epoch": 6.318672030832685, "grad_norm": 1.8453973531723022, "learning_rate": 1.1685371381868427e-06, "loss": 0.0497, "step": 186900 }, { "epoch": 6.319010108522939, "grad_norm": 0.4018305242061615, "learning_rate": 1.1667608968906062e-06, "loss": 0.0463, "step": 186910 }, { "epoch": 6.319348186213192, "grad_norm": 0.6939331293106079, "learning_rate": 1.1649859906767923e-06, "loss": 0.0567, "step": 186920 }, { "epoch": 6.319686263903445, "grad_norm": 0.7826530337333679, "learning_rate": 1.163212419593912e-06, "loss": 0.0473, "step": 186930 }, { "epoch": 6.320024341593698, "grad_norm": 0.9033558964729309, "learning_rate": 1.1614401836904597e-06, "loss": 0.055, "step": 186940 }, { "epoch": 6.320362419283952, "grad_norm": 0.6722220182418823, "learning_rate": 1.1596692830148747e-06, "loss": 0.0602, "step": 186950 }, { "epoch": 6.320700496974204, "grad_norm": 0.5841608047485352, "learning_rate": 1.1578997176155849e-06, "loss": 0.0541, "step": 186960 }, { "epoch": 6.321038574664458, "grad_norm": 0.5769639611244202, "learning_rate": 1.156131487540979e-06, "loss": 0.0537, "step": 186970 }, { "epoch": 6.3213766523547115, "grad_norm": 0.5119314193725586, "learning_rate": 1.1543645928393798e-06, "loss": 0.0493, "step": 186980 }, { "epoch": 6.321714730044964, "grad_norm": 0.6800058484077454, "learning_rate": 1.1525990335591042e-06, "loss": 0.049, "step": 186990 }, { "epoch": 6.322052807735218, "grad_norm": 0.9676329493522644, "learning_rate": 1.1508348097484188e-06, "loss": 0.0504, "step": 187000 }, { "epoch": 6.322390885425471, "grad_norm": 0.9111858606338501, "learning_rate": 1.1490719214555579e-06, "loss": 0.0535, "step": 187010 }, { "epoch": 6.322728963115724, "grad_norm": 0.9077770709991455, "learning_rate": 1.1473103687287212e-06, "loss": 0.0572, "step": 187020 }, { "epoch": 6.323067040805977, "grad_norm": 1.1404379606246948, "learning_rate": 1.1455501516160649e-06, "loss": 0.0487, "step": 187030 }, { "epoch": 6.323405118496231, "grad_norm": 0.44899019598960876, "learning_rate": 1.1437912701657117e-06, "loss": 0.0508, "step": 187040 }, { "epoch": 6.323743196186483, "grad_norm": 1.3327447175979614, "learning_rate": 1.1420337244257562e-06, "loss": 0.0506, "step": 187050 }, { "epoch": 6.324081273876737, "grad_norm": 0.6724000573158264, "learning_rate": 1.140277514444238e-06, "loss": 0.0561, "step": 187060 }, { "epoch": 6.32441935156699, "grad_norm": 0.8078383207321167, "learning_rate": 1.1385226402691739e-06, "loss": 0.0561, "step": 187070 }, { "epoch": 6.324757429257243, "grad_norm": 0.7066652774810791, "learning_rate": 1.1367691019485538e-06, "loss": 0.0496, "step": 187080 }, { "epoch": 6.325095506947497, "grad_norm": 1.8493475914001465, "learning_rate": 1.1350168995303002e-06, "loss": 0.0499, "step": 187090 }, { "epoch": 6.32543358463775, "grad_norm": 1.1500492095947266, "learning_rate": 1.1332660330623301e-06, "loss": 0.0527, "step": 187100 }, { "epoch": 6.325771662328003, "grad_norm": 1.1509160995483398, "learning_rate": 1.1315165025925112e-06, "loss": 0.0552, "step": 187110 }, { "epoch": 6.326109740018256, "grad_norm": 0.5736871361732483, "learning_rate": 1.129768308168666e-06, "loss": 0.0522, "step": 187120 }, { "epoch": 6.32644781770851, "grad_norm": 1.2692186832427979, "learning_rate": 1.128021449838601e-06, "loss": 0.0561, "step": 187130 }, { "epoch": 6.326785895398762, "grad_norm": 0.9655890464782715, "learning_rate": 1.1262759276500723e-06, "loss": 0.0511, "step": 187140 }, { "epoch": 6.327123973089016, "grad_norm": 0.9244374632835388, "learning_rate": 1.1245317416507918e-06, "loss": 0.0587, "step": 187150 }, { "epoch": 6.327462050779269, "grad_norm": 1.2335386276245117, "learning_rate": 1.1227888918884545e-06, "loss": 0.0581, "step": 187160 }, { "epoch": 6.327800128469522, "grad_norm": 0.7085201740264893, "learning_rate": 1.1210473784107e-06, "loss": 0.0537, "step": 187170 }, { "epoch": 6.3281382061597755, "grad_norm": 1.2534700632095337, "learning_rate": 1.1193072012651628e-06, "loss": 0.0522, "step": 187180 }, { "epoch": 6.328476283850029, "grad_norm": 1.0302754640579224, "learning_rate": 1.117568360499388e-06, "loss": 0.0602, "step": 187190 }, { "epoch": 6.328814361540282, "grad_norm": 3.808004379272461, "learning_rate": 1.1158308561609376e-06, "loss": 0.0516, "step": 187200 }, { "epoch": 6.329152439230535, "grad_norm": 1.5234040021896362, "learning_rate": 1.1140946882973014e-06, "loss": 0.0518, "step": 187210 }, { "epoch": 6.329490516920789, "grad_norm": 1.4507997035980225, "learning_rate": 1.112359856955958e-06, "loss": 0.0633, "step": 187220 }, { "epoch": 6.329828594611041, "grad_norm": 0.6252875924110413, "learning_rate": 1.1106263621843249e-06, "loss": 0.0524, "step": 187230 }, { "epoch": 6.330166672301295, "grad_norm": 0.8458039164543152, "learning_rate": 1.1088942040297978e-06, "loss": 0.0518, "step": 187240 }, { "epoch": 6.330504749991548, "grad_norm": 1.4346872568130493, "learning_rate": 1.1071633825397443e-06, "loss": 0.0566, "step": 187250 }, { "epoch": 6.330842827681801, "grad_norm": 1.2977184057235718, "learning_rate": 1.1054338977614708e-06, "loss": 0.0563, "step": 187260 }, { "epoch": 6.3311809053720545, "grad_norm": 0.43155720829963684, "learning_rate": 1.1037057497422676e-06, "loss": 0.0517, "step": 187270 }, { "epoch": 6.331518983062308, "grad_norm": 0.6466832160949707, "learning_rate": 1.1019789385293745e-06, "loss": 0.049, "step": 187280 }, { "epoch": 6.331857060752561, "grad_norm": 0.4640486538410187, "learning_rate": 1.1002534641700036e-06, "loss": 0.0466, "step": 187290 }, { "epoch": 6.332195138442814, "grad_norm": 1.0139460563659668, "learning_rate": 1.0985293267113396e-06, "loss": 0.0535, "step": 187300 }, { "epoch": 6.332533216133068, "grad_norm": 2.681487560272217, "learning_rate": 1.0968065262005166e-06, "loss": 0.0479, "step": 187310 }, { "epoch": 6.33287129382332, "grad_norm": 1.115515112876892, "learning_rate": 1.0950850626846253e-06, "loss": 0.0569, "step": 187320 }, { "epoch": 6.333209371513574, "grad_norm": 0.6319331526756287, "learning_rate": 1.0933649362107334e-06, "loss": 0.0508, "step": 187330 }, { "epoch": 6.333547449203827, "grad_norm": 0.4652867317199707, "learning_rate": 1.0916461468258755e-06, "loss": 0.0553, "step": 187340 }, { "epoch": 6.33388552689408, "grad_norm": 1.2585371732711792, "learning_rate": 1.089928694577036e-06, "loss": 0.0553, "step": 187350 }, { "epoch": 6.334223604584333, "grad_norm": 0.9043365120887756, "learning_rate": 1.088212579511172e-06, "loss": 0.0546, "step": 187360 }, { "epoch": 6.334561682274587, "grad_norm": 1.8023473024368286, "learning_rate": 1.0864978016752014e-06, "loss": 0.0553, "step": 187370 }, { "epoch": 6.33489975996484, "grad_norm": 0.6404841542243958, "learning_rate": 1.084784361115998e-06, "loss": 0.0531, "step": 187380 }, { "epoch": 6.335237837655093, "grad_norm": 0.7667039036750793, "learning_rate": 1.0830722578804297e-06, "loss": 0.0525, "step": 187390 }, { "epoch": 6.335575915345347, "grad_norm": 0.5537599921226501, "learning_rate": 1.0813614920152758e-06, "loss": 0.0524, "step": 187400 }, { "epoch": 6.335913993035599, "grad_norm": 1.0314058065414429, "learning_rate": 1.0796520635673156e-06, "loss": 0.0508, "step": 187410 }, { "epoch": 6.336252070725853, "grad_norm": 0.8089837431907654, "learning_rate": 1.0779439725833007e-06, "loss": 0.0545, "step": 187420 }, { "epoch": 6.336590148416106, "grad_norm": 0.7868688106536865, "learning_rate": 1.0762372191099156e-06, "loss": 0.0486, "step": 187430 }, { "epoch": 6.336928226106359, "grad_norm": 0.8279412388801575, "learning_rate": 1.074531803193829e-06, "loss": 0.0507, "step": 187440 }, { "epoch": 6.337266303796612, "grad_norm": 1.003164529800415, "learning_rate": 1.0728277248816532e-06, "loss": 0.0529, "step": 187450 }, { "epoch": 6.337604381486866, "grad_norm": 1.084873080253601, "learning_rate": 1.07112498421999e-06, "loss": 0.0453, "step": 187460 }, { "epoch": 6.3379424591771185, "grad_norm": 1.7471656799316406, "learning_rate": 1.069423581255391e-06, "loss": 0.0514, "step": 187470 }, { "epoch": 6.338280536867372, "grad_norm": 0.4377981722354889, "learning_rate": 1.0677235160343634e-06, "loss": 0.0513, "step": 187480 }, { "epoch": 6.3386186145576255, "grad_norm": 0.7115517854690552, "learning_rate": 1.066024788603398e-06, "loss": 0.0534, "step": 187490 }, { "epoch": 6.338956692247878, "grad_norm": 0.5954616069793701, "learning_rate": 1.0643273990089297e-06, "loss": 0.0516, "step": 187500 }, { "epoch": 6.339294769938132, "grad_norm": 1.0216127634048462, "learning_rate": 1.0626313472973658e-06, "loss": 0.0569, "step": 187510 }, { "epoch": 6.339632847628385, "grad_norm": 1.85592520236969, "learning_rate": 1.060936633515075e-06, "loss": 0.0583, "step": 187520 }, { "epoch": 6.339970925318638, "grad_norm": 0.8029412031173706, "learning_rate": 1.0592432577083866e-06, "loss": 0.0473, "step": 187530 }, { "epoch": 6.340309003008891, "grad_norm": 0.8403282761573792, "learning_rate": 1.0575512199236081e-06, "loss": 0.0484, "step": 187540 }, { "epoch": 6.340647080699145, "grad_norm": 0.8959871530532837, "learning_rate": 1.0558605202069916e-06, "loss": 0.0561, "step": 187550 }, { "epoch": 6.3409851583893975, "grad_norm": 1.5390655994415283, "learning_rate": 1.054171158604761e-06, "loss": 0.0544, "step": 187560 }, { "epoch": 6.341323236079651, "grad_norm": 2.4702563285827637, "learning_rate": 1.0524831351631015e-06, "loss": 0.0529, "step": 187570 }, { "epoch": 6.3416613137699045, "grad_norm": 2.4544007778167725, "learning_rate": 1.0507964499281653e-06, "loss": 0.0488, "step": 187580 }, { "epoch": 6.341999391460158, "grad_norm": 1.2425384521484375, "learning_rate": 1.0491111029460599e-06, "loss": 0.0503, "step": 187590 }, { "epoch": 6.342337469150411, "grad_norm": 0.9334409236907959, "learning_rate": 1.0474270942628705e-06, "loss": 0.0507, "step": 187600 }, { "epoch": 6.342675546840664, "grad_norm": 0.4361629784107208, "learning_rate": 1.0457444239246384e-06, "loss": 0.0571, "step": 187610 }, { "epoch": 6.343013624530917, "grad_norm": 1.9534451961517334, "learning_rate": 1.0440630919773542e-06, "loss": 0.048, "step": 187620 }, { "epoch": 6.34335170222117, "grad_norm": 3.0066077709198, "learning_rate": 1.042383098466998e-06, "loss": 0.0535, "step": 187630 }, { "epoch": 6.343689779911424, "grad_norm": 0.7234485149383545, "learning_rate": 1.0407044434394997e-06, "loss": 0.0501, "step": 187640 }, { "epoch": 6.344027857601677, "grad_norm": 0.8161541223526001, "learning_rate": 1.0390271269407447e-06, "loss": 0.0499, "step": 187650 }, { "epoch": 6.34436593529193, "grad_norm": 0.7534410953521729, "learning_rate": 1.037351149016591e-06, "loss": 0.0535, "step": 187660 }, { "epoch": 6.3447040129821834, "grad_norm": 0.32166197896003723, "learning_rate": 1.0356765097128573e-06, "loss": 0.0508, "step": 187670 }, { "epoch": 6.345042090672436, "grad_norm": 0.9501951336860657, "learning_rate": 1.034003209075346e-06, "loss": 0.0572, "step": 187680 }, { "epoch": 6.34538016836269, "grad_norm": 0.6263871788978577, "learning_rate": 1.032331247149787e-06, "loss": 0.0516, "step": 187690 }, { "epoch": 6.345718246052943, "grad_norm": 0.8802043795585632, "learning_rate": 1.0306606239818939e-06, "loss": 0.0525, "step": 187700 }, { "epoch": 6.346056323743197, "grad_norm": 0.4127708375453949, "learning_rate": 1.0289913396173412e-06, "loss": 0.0462, "step": 187710 }, { "epoch": 6.346394401433449, "grad_norm": 0.776373565196991, "learning_rate": 1.0273233941017702e-06, "loss": 0.0555, "step": 187720 }, { "epoch": 6.346732479123703, "grad_norm": 0.6510003209114075, "learning_rate": 1.0256567874807776e-06, "loss": 0.0547, "step": 187730 }, { "epoch": 6.347070556813955, "grad_norm": 0.9030652046203613, "learning_rate": 1.0239915197999272e-06, "loss": 0.0446, "step": 187740 }, { "epoch": 6.347408634504209, "grad_norm": 0.3990611732006073, "learning_rate": 1.0223275911047546e-06, "loss": 0.0481, "step": 187750 }, { "epoch": 6.347746712194462, "grad_norm": 0.705586314201355, "learning_rate": 1.020665001440746e-06, "loss": 0.0556, "step": 187760 }, { "epoch": 6.348084789884716, "grad_norm": 1.0961003303527832, "learning_rate": 1.0190037508533535e-06, "loss": 0.0602, "step": 187770 }, { "epoch": 6.3484228675749685, "grad_norm": 0.8404494524002075, "learning_rate": 1.0173438393879907e-06, "loss": 0.051, "step": 187780 }, { "epoch": 6.348760945265222, "grad_norm": 0.8198268413543701, "learning_rate": 1.0156852670900497e-06, "loss": 0.0583, "step": 187790 }, { "epoch": 6.349099022955476, "grad_norm": 0.588851273059845, "learning_rate": 1.0140280340048768e-06, "loss": 0.0504, "step": 187800 }, { "epoch": 6.349437100645728, "grad_norm": 0.7060059309005737, "learning_rate": 1.0123721401777698e-06, "loss": 0.0521, "step": 187810 }, { "epoch": 6.349775178335982, "grad_norm": 0.8685484528541565, "learning_rate": 1.0107175856540029e-06, "loss": 0.0493, "step": 187820 }, { "epoch": 6.350113256026235, "grad_norm": 0.5267302989959717, "learning_rate": 1.0090643704788128e-06, "loss": 0.0474, "step": 187830 }, { "epoch": 6.350451333716488, "grad_norm": 0.9925631880760193, "learning_rate": 1.0074124946974017e-06, "loss": 0.054, "step": 187840 }, { "epoch": 6.350789411406741, "grad_norm": 0.707171618938446, "learning_rate": 1.0057619583549282e-06, "loss": 0.0575, "step": 187850 }, { "epoch": 6.351127489096995, "grad_norm": 0.5877361297607422, "learning_rate": 1.0041127614965117e-06, "loss": 0.0576, "step": 187860 }, { "epoch": 6.3514655667872475, "grad_norm": 1.0138397216796875, "learning_rate": 1.0024649041672495e-06, "loss": 0.0553, "step": 187870 }, { "epoch": 6.351803644477501, "grad_norm": 1.015879511833191, "learning_rate": 1.0008183864121834e-06, "loss": 0.0551, "step": 187880 }, { "epoch": 6.3521417221677545, "grad_norm": 0.5739387273788452, "learning_rate": 9.991732082763438e-07, "loss": 0.0572, "step": 187890 }, { "epoch": 6.352479799858007, "grad_norm": 0.6256886720657349, "learning_rate": 9.975293698046895e-07, "loss": 0.0532, "step": 187900 }, { "epoch": 6.352817877548261, "grad_norm": 0.5012191534042358, "learning_rate": 9.958868710421731e-07, "loss": 0.0465, "step": 187910 }, { "epoch": 6.353155955238514, "grad_norm": 0.6655175685882568, "learning_rate": 9.94245712033709e-07, "loss": 0.0493, "step": 187920 }, { "epoch": 6.353494032928767, "grad_norm": 0.646649181842804, "learning_rate": 9.9260589282415e-07, "loss": 0.0551, "step": 187930 }, { "epoch": 6.35383211061902, "grad_norm": 1.1678001880645752, "learning_rate": 9.90967413458338e-07, "loss": 0.0556, "step": 187940 }, { "epoch": 6.354170188309274, "grad_norm": 1.4648525714874268, "learning_rate": 9.893302739810594e-07, "loss": 0.0518, "step": 187950 }, { "epoch": 6.3545082659995265, "grad_norm": 1.6491225957870483, "learning_rate": 9.876944744370897e-07, "loss": 0.0534, "step": 187960 }, { "epoch": 6.35484634368978, "grad_norm": 1.4147379398345947, "learning_rate": 9.860600148711319e-07, "loss": 0.0448, "step": 187970 }, { "epoch": 6.3551844213800335, "grad_norm": 0.4837760031223297, "learning_rate": 9.844268953278779e-07, "loss": 0.0511, "step": 187980 }, { "epoch": 6.355522499070286, "grad_norm": 1.0687477588653564, "learning_rate": 9.82795115851981e-07, "loss": 0.0526, "step": 187990 }, { "epoch": 6.35586057676054, "grad_norm": 0.5677381157875061, "learning_rate": 9.8116467648805e-07, "loss": 0.0529, "step": 188000 }, { "epoch": 6.356198654450793, "grad_norm": 1.1531013250350952, "learning_rate": 9.795355772806602e-07, "loss": 0.0562, "step": 188010 }, { "epoch": 6.356536732141046, "grad_norm": 1.5930451154708862, "learning_rate": 9.779078182743594e-07, "loss": 0.0553, "step": 188020 }, { "epoch": 6.356874809831299, "grad_norm": 0.888006865978241, "learning_rate": 9.762813995136345e-07, "loss": 0.0526, "step": 188030 }, { "epoch": 6.357212887521553, "grad_norm": 1.6276464462280273, "learning_rate": 9.746563210429605e-07, "loss": 0.0595, "step": 188040 }, { "epoch": 6.357550965211805, "grad_norm": 0.4280475676059723, "learning_rate": 9.730325829067632e-07, "loss": 0.049, "step": 188050 }, { "epoch": 6.357889042902059, "grad_norm": 0.4811166822910309, "learning_rate": 9.714101851494407e-07, "loss": 0.0479, "step": 188060 }, { "epoch": 6.358227120592312, "grad_norm": 0.9359447360038757, "learning_rate": 9.697891278153404e-07, "loss": 0.0503, "step": 188070 }, { "epoch": 6.358565198282565, "grad_norm": 0.7054978013038635, "learning_rate": 9.681694109487882e-07, "loss": 0.0548, "step": 188080 }, { "epoch": 6.358903275972819, "grad_norm": 0.638861358165741, "learning_rate": 9.665510345940598e-07, "loss": 0.0501, "step": 188090 }, { "epoch": 6.359241353663072, "grad_norm": 1.4958213567733765, "learning_rate": 9.649339987954143e-07, "loss": 0.0531, "step": 188100 }, { "epoch": 6.359579431353325, "grad_norm": 0.897727906703949, "learning_rate": 9.633183035970383e-07, "loss": 0.0527, "step": 188110 }, { "epoch": 6.359917509043578, "grad_norm": 0.5869240760803223, "learning_rate": 9.617039490431246e-07, "loss": 0.0615, "step": 188120 }, { "epoch": 6.360255586733832, "grad_norm": 0.597170889377594, "learning_rate": 9.600909351778043e-07, "loss": 0.0539, "step": 188130 }, { "epoch": 6.360593664424084, "grad_norm": 0.9562898874282837, "learning_rate": 9.5847926204517e-07, "loss": 0.0532, "step": 188140 }, { "epoch": 6.360931742114338, "grad_norm": 0.8470253944396973, "learning_rate": 9.568689296892974e-07, "loss": 0.0461, "step": 188150 }, { "epoch": 6.361269819804591, "grad_norm": 0.8051943182945251, "learning_rate": 9.552599381541904e-07, "loss": 0.0533, "step": 188160 }, { "epoch": 6.361607897494844, "grad_norm": 0.7188144326210022, "learning_rate": 9.536522874838583e-07, "loss": 0.0505, "step": 188170 }, { "epoch": 6.3619459751850975, "grad_norm": 1.206139087677002, "learning_rate": 9.52045977722249e-07, "loss": 0.0518, "step": 188180 }, { "epoch": 6.362284052875351, "grad_norm": 0.9389896392822266, "learning_rate": 9.50441008913272e-07, "loss": 0.047, "step": 188190 }, { "epoch": 6.362622130565604, "grad_norm": 0.9228979349136353, "learning_rate": 9.488373811008144e-07, "loss": 0.0493, "step": 188200 }, { "epoch": 6.362960208255857, "grad_norm": 1.3188990354537964, "learning_rate": 9.472350943287134e-07, "loss": 0.0538, "step": 188210 }, { "epoch": 6.363298285946111, "grad_norm": 2.1236071586608887, "learning_rate": 9.456341486407783e-07, "loss": 0.0501, "step": 188220 }, { "epoch": 6.363636363636363, "grad_norm": 1.187909722328186, "learning_rate": 9.44034544080774e-07, "loss": 0.0554, "step": 188230 }, { "epoch": 6.363974441326617, "grad_norm": 0.7641802430152893, "learning_rate": 9.424362806924325e-07, "loss": 0.0483, "step": 188240 }, { "epoch": 6.36431251901687, "grad_norm": 0.5406213998794556, "learning_rate": 9.408393585194575e-07, "loss": 0.0518, "step": 188250 }, { "epoch": 6.364650596707123, "grad_norm": 1.0114719867706299, "learning_rate": 9.392437776055085e-07, "loss": 0.0527, "step": 188260 }, { "epoch": 6.3649886743973765, "grad_norm": 1.410369634628296, "learning_rate": 9.376495379942007e-07, "loss": 0.0541, "step": 188270 }, { "epoch": 6.36532675208763, "grad_norm": 0.7314886450767517, "learning_rate": 9.36056639729116e-07, "loss": 0.052, "step": 188280 }, { "epoch": 6.365664829777883, "grad_norm": 0.5789607763290405, "learning_rate": 9.344650828538137e-07, "loss": 0.0558, "step": 188290 }, { "epoch": 6.366002907468136, "grad_norm": 0.7011852860450745, "learning_rate": 9.328748674118038e-07, "loss": 0.0547, "step": 188300 }, { "epoch": 6.36634098515839, "grad_norm": 1.314398169517517, "learning_rate": 9.312859934465623e-07, "loss": 0.0573, "step": 188310 }, { "epoch": 6.366679062848642, "grad_norm": 0.812744140625, "learning_rate": 9.296984610015269e-07, "loss": 0.0528, "step": 188320 }, { "epoch": 6.367017140538896, "grad_norm": 0.9568145871162415, "learning_rate": 9.281122701201017e-07, "loss": 0.0523, "step": 188330 }, { "epoch": 6.367355218229149, "grad_norm": 0.7756704092025757, "learning_rate": 9.265274208456521e-07, "loss": 0.0576, "step": 188340 }, { "epoch": 6.367693295919402, "grad_norm": 0.8717159032821655, "learning_rate": 9.249439132215043e-07, "loss": 0.0552, "step": 188350 }, { "epoch": 6.368031373609655, "grad_norm": 0.6917339563369751, "learning_rate": 9.233617472909518e-07, "loss": 0.0565, "step": 188360 }, { "epoch": 6.368369451299909, "grad_norm": 0.6714615225791931, "learning_rate": 9.21780923097254e-07, "loss": 0.0478, "step": 188370 }, { "epoch": 6.368707528990162, "grad_norm": 2.5048677921295166, "learning_rate": 9.202014406836268e-07, "loss": 0.0486, "step": 188380 }, { "epoch": 6.369045606680415, "grad_norm": 0.5386804938316345, "learning_rate": 9.186233000932576e-07, "loss": 0.0584, "step": 188390 }, { "epoch": 6.369383684370669, "grad_norm": 0.8405048847198486, "learning_rate": 9.170465013692841e-07, "loss": 0.0491, "step": 188400 }, { "epoch": 6.369721762060921, "grad_norm": 0.9517069458961487, "learning_rate": 9.15471044554822e-07, "loss": 0.0504, "step": 188410 }, { "epoch": 6.370059839751175, "grad_norm": 1.315366506576538, "learning_rate": 9.138969296929422e-07, "loss": 0.0489, "step": 188420 }, { "epoch": 6.370397917441428, "grad_norm": 2.88344669342041, "learning_rate": 9.123241568266772e-07, "loss": 0.055, "step": 188430 }, { "epoch": 6.370735995131681, "grad_norm": 0.9369712471961975, "learning_rate": 9.107527259990312e-07, "loss": 0.0502, "step": 188440 }, { "epoch": 6.371074072821934, "grad_norm": 1.6415185928344727, "learning_rate": 9.091826372529588e-07, "loss": 0.0489, "step": 188450 }, { "epoch": 6.371412150512188, "grad_norm": 1.0049684047698975, "learning_rate": 9.076138906313924e-07, "loss": 0.0561, "step": 188460 }, { "epoch": 6.3717502282024405, "grad_norm": 0.7247353196144104, "learning_rate": 9.060464861772255e-07, "loss": 0.0557, "step": 188470 }, { "epoch": 6.372088305892694, "grad_norm": 1.289335012435913, "learning_rate": 9.044804239332961e-07, "loss": 0.0556, "step": 188480 }, { "epoch": 6.3724263835829476, "grad_norm": 0.43906667828559875, "learning_rate": 9.029157039424307e-07, "loss": 0.0474, "step": 188490 }, { "epoch": 6.3727644612732, "grad_norm": 0.6684073805809021, "learning_rate": 9.013523262474011e-07, "loss": 0.0609, "step": 188500 }, { "epoch": 6.373102538963454, "grad_norm": 1.0315630435943604, "learning_rate": 8.997902908909561e-07, "loss": 0.0558, "step": 188510 }, { "epoch": 6.373440616653707, "grad_norm": 0.693229615688324, "learning_rate": 8.982295979158006e-07, "loss": 0.0496, "step": 188520 }, { "epoch": 6.37377869434396, "grad_norm": 0.48452043533325195, "learning_rate": 8.966702473646005e-07, "loss": 0.0513, "step": 188530 }, { "epoch": 6.374116772034213, "grad_norm": 2.0637094974517822, "learning_rate": 8.951122392799883e-07, "loss": 0.0531, "step": 188540 }, { "epoch": 6.374454849724467, "grad_norm": 1.79352605342865, "learning_rate": 8.935555737045576e-07, "loss": 0.0509, "step": 188550 }, { "epoch": 6.37479292741472, "grad_norm": 1.0696791410446167, "learning_rate": 8.920002506808745e-07, "loss": 0.0534, "step": 188560 }, { "epoch": 6.375131005104973, "grad_norm": 0.9293838143348694, "learning_rate": 8.904462702514494e-07, "loss": 0.046, "step": 188570 }, { "epoch": 6.3754690827952265, "grad_norm": 2.055058240890503, "learning_rate": 8.88893632458776e-07, "loss": 0.0482, "step": 188580 }, { "epoch": 6.375807160485479, "grad_norm": 1.0123904943466187, "learning_rate": 8.873423373452983e-07, "loss": 0.0466, "step": 188590 }, { "epoch": 6.376145238175733, "grad_norm": 1.2084987163543701, "learning_rate": 8.857923849534377e-07, "loss": 0.0516, "step": 188600 }, { "epoch": 6.376483315865986, "grad_norm": 0.4303727447986603, "learning_rate": 8.84243775325555e-07, "loss": 0.0514, "step": 188610 }, { "epoch": 6.37682139355624, "grad_norm": 1.1571067571640015, "learning_rate": 8.82696508503994e-07, "loss": 0.0522, "step": 188620 }, { "epoch": 6.377159471246492, "grad_norm": 1.5730448961257935, "learning_rate": 8.811505845310597e-07, "loss": 0.0428, "step": 188630 }, { "epoch": 6.377497548936746, "grad_norm": 3.3598716259002686, "learning_rate": 8.796060034490183e-07, "loss": 0.0518, "step": 188640 }, { "epoch": 6.377835626626998, "grad_norm": 0.4490405023097992, "learning_rate": 8.780627653000972e-07, "loss": 0.0523, "step": 188650 }, { "epoch": 6.378173704317252, "grad_norm": 0.7482168674468994, "learning_rate": 8.765208701264793e-07, "loss": 0.0524, "step": 188660 }, { "epoch": 6.3785117820075055, "grad_norm": 0.7946632504463196, "learning_rate": 8.749803179703308e-07, "loss": 0.0512, "step": 188670 }, { "epoch": 6.378849859697759, "grad_norm": 2.5395519733428955, "learning_rate": 8.734411088737627e-07, "loss": 0.0513, "step": 188680 }, { "epoch": 6.379187937388012, "grad_norm": 1.8725565671920776, "learning_rate": 8.719032428788576e-07, "loss": 0.0531, "step": 188690 }, { "epoch": 6.379526015078265, "grad_norm": 0.9024413228034973, "learning_rate": 8.703667200276599e-07, "loss": 0.055, "step": 188700 }, { "epoch": 6.379864092768518, "grad_norm": 1.037782073020935, "learning_rate": 8.688315403621805e-07, "loss": 0.0543, "step": 188710 }, { "epoch": 6.380202170458771, "grad_norm": 0.6315953135490417, "learning_rate": 8.672977039243913e-07, "loss": 0.0457, "step": 188720 }, { "epoch": 6.380540248149025, "grad_norm": 3.659771680831909, "learning_rate": 8.657652107562253e-07, "loss": 0.0585, "step": 188730 }, { "epoch": 6.380878325839278, "grad_norm": 0.9801493883132935, "learning_rate": 8.64234060899577e-07, "loss": 0.0524, "step": 188740 }, { "epoch": 6.381216403529531, "grad_norm": 5.45262336730957, "learning_rate": 8.627042543963126e-07, "loss": 0.0532, "step": 188750 }, { "epoch": 6.381554481219784, "grad_norm": 0.5459614396095276, "learning_rate": 8.611757912882545e-07, "loss": 0.051, "step": 188760 }, { "epoch": 6.381892558910038, "grad_norm": 2.539144515991211, "learning_rate": 8.596486716171858e-07, "loss": 0.051, "step": 188770 }, { "epoch": 6.3822306366002906, "grad_norm": 0.4005071818828583, "learning_rate": 8.581228954248621e-07, "loss": 0.0553, "step": 188780 }, { "epoch": 6.382568714290544, "grad_norm": 0.6727320551872253, "learning_rate": 8.565984627529999e-07, "loss": 0.0529, "step": 188790 }, { "epoch": 6.382906791980798, "grad_norm": 0.7927709817886353, "learning_rate": 8.550753736432659e-07, "loss": 0.0538, "step": 188800 }, { "epoch": 6.38324486967105, "grad_norm": 3.742114305496216, "learning_rate": 8.535536281373158e-07, "loss": 0.0513, "step": 188810 }, { "epoch": 6.383582947361304, "grad_norm": 2.0802650451660156, "learning_rate": 8.520332262767439e-07, "loss": 0.052, "step": 188820 }, { "epoch": 6.383921025051557, "grad_norm": 0.921557605266571, "learning_rate": 8.50514168103117e-07, "loss": 0.0559, "step": 188830 }, { "epoch": 6.38425910274181, "grad_norm": 1.0200308561325073, "learning_rate": 8.489964536579686e-07, "loss": 0.0526, "step": 188840 }, { "epoch": 6.384597180432063, "grad_norm": 0.9997226595878601, "learning_rate": 8.474800829827989e-07, "loss": 0.0534, "step": 188850 }, { "epoch": 6.384935258122317, "grad_norm": 0.6508844494819641, "learning_rate": 8.459650561190524e-07, "loss": 0.0475, "step": 188860 }, { "epoch": 6.3852733358125695, "grad_norm": 0.8843724131584167, "learning_rate": 8.444513731081516e-07, "loss": 0.0557, "step": 188870 }, { "epoch": 6.385611413502823, "grad_norm": 0.9246588349342346, "learning_rate": 8.429390339914856e-07, "loss": 0.0464, "step": 188880 }, { "epoch": 6.3859494911930765, "grad_norm": 0.672838032245636, "learning_rate": 8.414280388104046e-07, "loss": 0.0541, "step": 188890 }, { "epoch": 6.386287568883329, "grad_norm": 0.6316890120506287, "learning_rate": 8.399183876062089e-07, "loss": 0.0538, "step": 188900 }, { "epoch": 6.386625646573583, "grad_norm": 1.8066874742507935, "learning_rate": 8.384100804201766e-07, "loss": 0.0486, "step": 188910 }, { "epoch": 6.386963724263836, "grad_norm": 0.6985694766044617, "learning_rate": 8.369031172935416e-07, "loss": 0.0548, "step": 188920 }, { "epoch": 6.387301801954089, "grad_norm": 0.7616019248962402, "learning_rate": 8.353974982675039e-07, "loss": 0.0483, "step": 188930 }, { "epoch": 6.387639879644342, "grad_norm": 0.7604362368583679, "learning_rate": 8.338932233832364e-07, "loss": 0.0579, "step": 188940 }, { "epoch": 6.387977957334596, "grad_norm": 1.162209153175354, "learning_rate": 8.323902926818505e-07, "loss": 0.0561, "step": 188950 }, { "epoch": 6.3883160350248485, "grad_norm": 1.4892407655715942, "learning_rate": 8.308887062044412e-07, "loss": 0.0542, "step": 188960 }, { "epoch": 6.388654112715102, "grad_norm": 1.2375460863113403, "learning_rate": 8.293884639920701e-07, "loss": 0.0503, "step": 188970 }, { "epoch": 6.3889921904053555, "grad_norm": 0.8556011319160461, "learning_rate": 8.278895660857433e-07, "loss": 0.0493, "step": 188980 }, { "epoch": 6.389330268095608, "grad_norm": 0.6889888644218445, "learning_rate": 8.263920125264391e-07, "loss": 0.0587, "step": 188990 }, { "epoch": 6.389668345785862, "grad_norm": 0.5023332834243774, "learning_rate": 8.248958033551024e-07, "loss": 0.0491, "step": 189000 }, { "epoch": 6.390006423476115, "grad_norm": 0.7611451745033264, "learning_rate": 8.234009386126451e-07, "loss": 0.0507, "step": 189010 }, { "epoch": 6.390344501166368, "grad_norm": 0.5232903361320496, "learning_rate": 8.219074183399289e-07, "loss": 0.0517, "step": 189020 }, { "epoch": 6.390682578856621, "grad_norm": 0.4937470853328705, "learning_rate": 8.204152425777878e-07, "loss": 0.0558, "step": 189030 }, { "epoch": 6.391020656546875, "grad_norm": 0.6858423948287964, "learning_rate": 8.189244113670169e-07, "loss": 0.0485, "step": 189040 }, { "epoch": 6.391358734237127, "grad_norm": 0.6145347356796265, "learning_rate": 8.174349247483836e-07, "loss": 0.0541, "step": 189050 }, { "epoch": 6.391696811927381, "grad_norm": 0.9739745259284973, "learning_rate": 8.159467827625944e-07, "loss": 0.0495, "step": 189060 }, { "epoch": 6.392034889617634, "grad_norm": 0.5953440070152283, "learning_rate": 8.144599854503443e-07, "loss": 0.0497, "step": 189070 }, { "epoch": 6.392372967307887, "grad_norm": 1.4900493621826172, "learning_rate": 8.129745328522842e-07, "loss": 0.0564, "step": 189080 }, { "epoch": 6.392711044998141, "grad_norm": 2.709564208984375, "learning_rate": 8.11490425009015e-07, "loss": 0.0564, "step": 189090 }, { "epoch": 6.393049122688394, "grad_norm": 0.5393505692481995, "learning_rate": 8.100076619611263e-07, "loss": 0.0516, "step": 189100 }, { "epoch": 6.393387200378647, "grad_norm": 1.093497633934021, "learning_rate": 8.085262437491414e-07, "loss": 0.0537, "step": 189110 }, { "epoch": 6.3937252780689, "grad_norm": 4.263219356536865, "learning_rate": 8.070461704135724e-07, "loss": 0.0532, "step": 189120 }, { "epoch": 6.394063355759154, "grad_norm": 0.6238735914230347, "learning_rate": 8.055674419948811e-07, "loss": 0.0487, "step": 189130 }, { "epoch": 6.394401433449406, "grad_norm": 0.7859389185905457, "learning_rate": 8.04090058533491e-07, "loss": 0.0458, "step": 189140 }, { "epoch": 6.39473951113966, "grad_norm": 0.3618536591529846, "learning_rate": 8.026140200697974e-07, "loss": 0.0504, "step": 189150 }, { "epoch": 6.395077588829913, "grad_norm": 0.5417601466178894, "learning_rate": 8.011393266441569e-07, "loss": 0.0563, "step": 189160 }, { "epoch": 6.395415666520166, "grad_norm": 0.9076443314552307, "learning_rate": 7.996659782968819e-07, "loss": 0.0496, "step": 189170 }, { "epoch": 6.3957537442104195, "grad_norm": 0.47194409370422363, "learning_rate": 7.981939750682566e-07, "loss": 0.0531, "step": 189180 }, { "epoch": 6.396091821900673, "grad_norm": 0.8294535875320435, "learning_rate": 7.967233169985156e-07, "loss": 0.0572, "step": 189190 }, { "epoch": 6.396429899590926, "grad_norm": 1.0812726020812988, "learning_rate": 7.952540041278877e-07, "loss": 0.0512, "step": 189200 }, { "epoch": 6.396767977281179, "grad_norm": 1.3768373727798462, "learning_rate": 7.937860364965188e-07, "loss": 0.0498, "step": 189210 }, { "epoch": 6.397106054971433, "grad_norm": 0.5641200542449951, "learning_rate": 7.9231941414456e-07, "loss": 0.0578, "step": 189220 }, { "epoch": 6.397444132661685, "grad_norm": 1.1708680391311646, "learning_rate": 7.908541371121014e-07, "loss": 0.0532, "step": 189230 }, { "epoch": 6.397782210351939, "grad_norm": 2.1625609397888184, "learning_rate": 7.893902054392055e-07, "loss": 0.0539, "step": 189240 }, { "epoch": 6.398120288042192, "grad_norm": 1.132177710533142, "learning_rate": 7.879276191658847e-07, "loss": 0.0568, "step": 189250 }, { "epoch": 6.398458365732445, "grad_norm": 0.7410096526145935, "learning_rate": 7.864663783321402e-07, "loss": 0.0501, "step": 189260 }, { "epoch": 6.3987964434226985, "grad_norm": 9.21816635131836, "learning_rate": 7.850064829779236e-07, "loss": 0.0541, "step": 189270 }, { "epoch": 6.399134521112952, "grad_norm": 1.6070274114608765, "learning_rate": 7.835479331431306e-07, "loss": 0.065, "step": 189280 }, { "epoch": 6.399472598803205, "grad_norm": 1.5846878290176392, "learning_rate": 7.820907288676516e-07, "loss": 0.0455, "step": 189290 }, { "epoch": 6.399810676493458, "grad_norm": 0.9466418027877808, "learning_rate": 7.806348701913158e-07, "loss": 0.0491, "step": 189300 }, { "epoch": 6.400148754183712, "grad_norm": 1.028701663017273, "learning_rate": 7.791803571539413e-07, "loss": 0.0486, "step": 189310 }, { "epoch": 6.400486831873964, "grad_norm": 0.582534909248352, "learning_rate": 7.777271897952798e-07, "loss": 0.0511, "step": 189320 }, { "epoch": 6.400824909564218, "grad_norm": 1.699400782585144, "learning_rate": 7.762753681550605e-07, "loss": 0.0538, "step": 189330 }, { "epoch": 6.401162987254471, "grad_norm": 0.9184331893920898, "learning_rate": 7.74824892272985e-07, "loss": 0.0439, "step": 189340 }, { "epoch": 6.401501064944724, "grad_norm": 0.6779996156692505, "learning_rate": 7.73375762188705e-07, "loss": 0.0534, "step": 189350 }, { "epoch": 6.401839142634977, "grad_norm": 3.812755584716797, "learning_rate": 7.719279779418387e-07, "loss": 0.0543, "step": 189360 }, { "epoch": 6.402177220325231, "grad_norm": 0.5921809077262878, "learning_rate": 7.704815395719656e-07, "loss": 0.052, "step": 189370 }, { "epoch": 6.402515298015484, "grad_norm": 0.8252399563789368, "learning_rate": 7.690364471186263e-07, "loss": 0.0494, "step": 189380 }, { "epoch": 6.402853375705737, "grad_norm": 0.5462896227836609, "learning_rate": 7.675927006213446e-07, "loss": 0.0534, "step": 189390 }, { "epoch": 6.403191453395991, "grad_norm": 1.1384354829788208, "learning_rate": 7.66150300119578e-07, "loss": 0.0542, "step": 189400 }, { "epoch": 6.403529531086243, "grad_norm": 0.47556057572364807, "learning_rate": 7.647092456527671e-07, "loss": 0.0489, "step": 189410 }, { "epoch": 6.403867608776497, "grad_norm": 1.0632288455963135, "learning_rate": 7.632695372603083e-07, "loss": 0.0541, "step": 189420 }, { "epoch": 6.40420568646675, "grad_norm": 0.37927496433258057, "learning_rate": 7.618311749815643e-07, "loss": 0.0498, "step": 189430 }, { "epoch": 6.404543764157003, "grad_norm": 0.6289421319961548, "learning_rate": 7.603941588558594e-07, "loss": 0.049, "step": 189440 }, { "epoch": 6.404881841847256, "grad_norm": 0.7345463037490845, "learning_rate": 7.589584889224732e-07, "loss": 0.0516, "step": 189450 }, { "epoch": 6.40521991953751, "grad_norm": 1.7901785373687744, "learning_rate": 7.575241652206689e-07, "loss": 0.05, "step": 189460 }, { "epoch": 6.4055579972277625, "grad_norm": 5.246009349822998, "learning_rate": 7.560911877896537e-07, "loss": 0.0534, "step": 189470 }, { "epoch": 6.405896074918016, "grad_norm": 0.7292713522911072, "learning_rate": 7.546595566686077e-07, "loss": 0.0512, "step": 189480 }, { "epoch": 6.40623415260827, "grad_norm": 0.7649489641189575, "learning_rate": 7.532292718966605e-07, "loss": 0.0462, "step": 189490 }, { "epoch": 6.406572230298522, "grad_norm": 2.306814670562744, "learning_rate": 7.518003335129308e-07, "loss": 0.0555, "step": 189500 }, { "epoch": 6.406910307988776, "grad_norm": 0.5532640814781189, "learning_rate": 7.503727415564766e-07, "loss": 0.0493, "step": 189510 }, { "epoch": 6.407248385679029, "grad_norm": 5.488307476043701, "learning_rate": 7.489464960663273e-07, "loss": 0.0537, "step": 189520 }, { "epoch": 6.407586463369283, "grad_norm": 1.0937227010726929, "learning_rate": 7.475215970814852e-07, "loss": 0.0527, "step": 189530 }, { "epoch": 6.407924541059535, "grad_norm": 1.536260724067688, "learning_rate": 7.460980446408916e-07, "loss": 0.052, "step": 189540 }, { "epoch": 6.408262618749789, "grad_norm": 0.8388839364051819, "learning_rate": 7.446758387834762e-07, "loss": 0.0531, "step": 189550 }, { "epoch": 6.4086006964400415, "grad_norm": 2.3313090801239014, "learning_rate": 7.432549795481191e-07, "loss": 0.0514, "step": 189560 }, { "epoch": 6.408938774130295, "grad_norm": 1.1081198453903198, "learning_rate": 7.418354669736616e-07, "loss": 0.0493, "step": 189570 }, { "epoch": 6.4092768518205485, "grad_norm": 0.8165484666824341, "learning_rate": 7.404173010989223e-07, "loss": 0.0498, "step": 189580 }, { "epoch": 6.409614929510802, "grad_norm": 2.0093672275543213, "learning_rate": 7.39000481962665e-07, "loss": 0.0518, "step": 189590 }, { "epoch": 6.409953007201055, "grad_norm": 0.627255380153656, "learning_rate": 7.375850096036363e-07, "loss": 0.054, "step": 189600 }, { "epoch": 6.410291084891308, "grad_norm": 1.0688008069992065, "learning_rate": 7.361708840605164e-07, "loss": 0.0536, "step": 189610 }, { "epoch": 6.410629162581561, "grad_norm": 0.7042531967163086, "learning_rate": 7.347581053719798e-07, "loss": 0.0542, "step": 189620 }, { "epoch": 6.410967240271814, "grad_norm": 0.6596340537071228, "learning_rate": 7.333466735766459e-07, "loss": 0.0524, "step": 189630 }, { "epoch": 6.411305317962068, "grad_norm": 0.49929845333099365, "learning_rate": 7.319365887131113e-07, "loss": 0.0506, "step": 189640 }, { "epoch": 6.411643395652321, "grad_norm": 0.8154290318489075, "learning_rate": 7.305278508199232e-07, "loss": 0.0512, "step": 189650 }, { "epoch": 6.411981473342574, "grad_norm": 1.7831529378890991, "learning_rate": 7.291204599355838e-07, "loss": 0.0516, "step": 189660 }, { "epoch": 6.4123195510328275, "grad_norm": 0.5744960308074951, "learning_rate": 7.277144160985904e-07, "loss": 0.0482, "step": 189670 }, { "epoch": 6.41265762872308, "grad_norm": 0.8164200782775879, "learning_rate": 7.263097193473678e-07, "loss": 0.0572, "step": 189680 }, { "epoch": 6.412995706413334, "grad_norm": 0.6796416640281677, "learning_rate": 7.249063697203296e-07, "loss": 0.0538, "step": 189690 }, { "epoch": 6.413333784103587, "grad_norm": 0.6549639701843262, "learning_rate": 7.235043672558395e-07, "loss": 0.0552, "step": 189700 }, { "epoch": 6.413671861793841, "grad_norm": 0.4905422031879425, "learning_rate": 7.221037119922225e-07, "loss": 0.054, "step": 189710 }, { "epoch": 6.414009939484093, "grad_norm": 1.4075970649719238, "learning_rate": 7.207044039677812e-07, "loss": 0.0528, "step": 189720 }, { "epoch": 6.414348017174347, "grad_norm": 0.8275675177574158, "learning_rate": 7.193064432207686e-07, "loss": 0.0495, "step": 189730 }, { "epoch": 6.4146860948646, "grad_norm": 1.149388074874878, "learning_rate": 7.179098297894038e-07, "loss": 0.0477, "step": 189740 }, { "epoch": 6.415024172554853, "grad_norm": 1.2303287982940674, "learning_rate": 7.165145637118675e-07, "loss": 0.0495, "step": 189750 }, { "epoch": 6.415362250245106, "grad_norm": 0.7118922472000122, "learning_rate": 7.151206450263126e-07, "loss": 0.0539, "step": 189760 }, { "epoch": 6.41570032793536, "grad_norm": 0.6503867506980896, "learning_rate": 7.137280737708418e-07, "loss": 0.0519, "step": 189770 }, { "epoch": 6.416038405625613, "grad_norm": 0.8462639451026917, "learning_rate": 7.123368499835304e-07, "loss": 0.0509, "step": 189780 }, { "epoch": 6.416376483315866, "grad_norm": 1.260799527168274, "learning_rate": 7.109469737024088e-07, "loss": 0.0512, "step": 189790 }, { "epoch": 6.41671456100612, "grad_norm": 0.7761067748069763, "learning_rate": 7.095584449654802e-07, "loss": 0.0496, "step": 189800 }, { "epoch": 6.417052638696372, "grad_norm": 0.6974138021469116, "learning_rate": 7.081712638107141e-07, "loss": 0.0495, "step": 189810 }, { "epoch": 6.417390716386626, "grad_norm": 3.586663007736206, "learning_rate": 7.067854302760191e-07, "loss": 0.0514, "step": 189820 }, { "epoch": 6.417728794076879, "grad_norm": 1.1867907047271729, "learning_rate": 7.054009443992926e-07, "loss": 0.055, "step": 189830 }, { "epoch": 6.418066871767132, "grad_norm": 0.8373696804046631, "learning_rate": 7.040178062183877e-07, "loss": 0.0528, "step": 189840 }, { "epoch": 6.418404949457385, "grad_norm": 0.6314054727554321, "learning_rate": 7.026360157711187e-07, "loss": 0.0535, "step": 189850 }, { "epoch": 6.418743027147639, "grad_norm": 1.829714298248291, "learning_rate": 7.012555730952608e-07, "loss": 0.053, "step": 189860 }, { "epoch": 6.4190811048378915, "grad_norm": 1.630477786064148, "learning_rate": 6.998764782285505e-07, "loss": 0.0493, "step": 189870 }, { "epoch": 6.419419182528145, "grad_norm": 1.1211578845977783, "learning_rate": 6.984987312086965e-07, "loss": 0.0512, "step": 189880 }, { "epoch": 6.4197572602183985, "grad_norm": 0.4842367172241211, "learning_rate": 6.971223320733688e-07, "loss": 0.0509, "step": 189890 }, { "epoch": 6.420095337908651, "grad_norm": 1.1104902029037476, "learning_rate": 6.957472808601928e-07, "loss": 0.0494, "step": 189900 }, { "epoch": 6.420433415598905, "grad_norm": 0.6655954718589783, "learning_rate": 6.943735776067606e-07, "loss": 0.0495, "step": 189910 }, { "epoch": 6.420771493289158, "grad_norm": 0.7941566705703735, "learning_rate": 6.93001222350631e-07, "loss": 0.0566, "step": 189920 }, { "epoch": 6.421109570979411, "grad_norm": 0.8062461018562317, "learning_rate": 6.916302151293296e-07, "loss": 0.0536, "step": 189930 }, { "epoch": 6.421447648669664, "grad_norm": 1.176967740058899, "learning_rate": 6.902605559803322e-07, "loss": 0.0571, "step": 189940 }, { "epoch": 6.421785726359918, "grad_norm": 0.7271298766136169, "learning_rate": 6.888922449410862e-07, "loss": 0.0522, "step": 189950 }, { "epoch": 6.4221238040501705, "grad_norm": 0.7318764328956604, "learning_rate": 6.875252820490008e-07, "loss": 0.0531, "step": 189960 }, { "epoch": 6.422461881740424, "grad_norm": 0.5050644278526306, "learning_rate": 6.861596673414461e-07, "loss": 0.056, "step": 189970 }, { "epoch": 6.4227999594306775, "grad_norm": 0.9170066118240356, "learning_rate": 6.847954008557644e-07, "loss": 0.054, "step": 189980 }, { "epoch": 6.42313803712093, "grad_norm": 0.5205515027046204, "learning_rate": 6.834324826292426e-07, "loss": 0.0562, "step": 189990 }, { "epoch": 6.423476114811184, "grad_norm": 1.3666208982467651, "learning_rate": 6.820709126991565e-07, "loss": 0.051, "step": 190000 }, { "epoch": 6.423814192501437, "grad_norm": 0.8362797498703003, "learning_rate": 6.807106911027151e-07, "loss": 0.052, "step": 190010 }, { "epoch": 6.42415227019169, "grad_norm": 0.8380873203277588, "learning_rate": 6.793518178771219e-07, "loss": 0.0529, "step": 190020 }, { "epoch": 6.424490347881943, "grad_norm": 3.9048566818237305, "learning_rate": 6.779942930595196e-07, "loss": 0.0529, "step": 190030 }, { "epoch": 6.424828425572197, "grad_norm": 0.947770357131958, "learning_rate": 6.766381166870228e-07, "loss": 0.0545, "step": 190040 }, { "epoch": 6.425166503262449, "grad_norm": 1.1347098350524902, "learning_rate": 6.752832887967131e-07, "loss": 0.0536, "step": 190050 }, { "epoch": 6.425504580952703, "grad_norm": 1.647541880607605, "learning_rate": 6.739298094256274e-07, "loss": 0.0578, "step": 190060 }, { "epoch": 6.4258426586429565, "grad_norm": 0.8433789610862732, "learning_rate": 6.725776786107696e-07, "loss": 0.0546, "step": 190070 }, { "epoch": 6.426180736333209, "grad_norm": 0.5820905566215515, "learning_rate": 6.712268963891045e-07, "loss": 0.0474, "step": 190080 }, { "epoch": 6.426518814023463, "grad_norm": 0.9077852368354797, "learning_rate": 6.698774627975635e-07, "loss": 0.049, "step": 190090 }, { "epoch": 6.426856891713716, "grad_norm": 0.6293759346008301, "learning_rate": 6.68529377873045e-07, "loss": 0.0533, "step": 190100 }, { "epoch": 6.427194969403969, "grad_norm": 1.1496638059616089, "learning_rate": 6.671826416523918e-07, "loss": 0.052, "step": 190110 }, { "epoch": 6.427533047094222, "grad_norm": 0.9685332179069519, "learning_rate": 6.658372541724412e-07, "loss": 0.0612, "step": 190120 }, { "epoch": 6.427871124784476, "grad_norm": 0.6944194436073303, "learning_rate": 6.644932154699579e-07, "loss": 0.0488, "step": 190130 }, { "epoch": 6.428209202474728, "grad_norm": 1.5966007709503174, "learning_rate": 6.631505255817016e-07, "loss": 0.0426, "step": 190140 }, { "epoch": 6.428547280164982, "grad_norm": 0.9886109232902527, "learning_rate": 6.618091845443708e-07, "loss": 0.0557, "step": 190150 }, { "epoch": 6.428885357855235, "grad_norm": 0.7130035758018494, "learning_rate": 6.604691923946416e-07, "loss": 0.0547, "step": 190160 }, { "epoch": 6.429223435545488, "grad_norm": 0.7114317417144775, "learning_rate": 6.591305491691512e-07, "loss": 0.052, "step": 190170 }, { "epoch": 6.4295615132357415, "grad_norm": 1.9672433137893677, "learning_rate": 6.577932549044985e-07, "loss": 0.0492, "step": 190180 }, { "epoch": 6.429899590925995, "grad_norm": 0.9182571768760681, "learning_rate": 6.564573096372373e-07, "loss": 0.0514, "step": 190190 }, { "epoch": 6.430237668616248, "grad_norm": 1.6536670923233032, "learning_rate": 6.55122713403894e-07, "loss": 0.0482, "step": 190200 }, { "epoch": 6.430575746306501, "grad_norm": 0.9179980754852295, "learning_rate": 6.537894662409561e-07, "loss": 0.0542, "step": 190210 }, { "epoch": 6.430913823996755, "grad_norm": 1.575294852256775, "learning_rate": 6.524575681848777e-07, "loss": 0.054, "step": 190220 }, { "epoch": 6.431251901687007, "grad_norm": 2.8524329662323, "learning_rate": 6.511270192720741e-07, "loss": 0.0484, "step": 190230 }, { "epoch": 6.431589979377261, "grad_norm": 1.7442821264266968, "learning_rate": 6.497978195389165e-07, "loss": 0.049, "step": 190240 }, { "epoch": 6.431928057067514, "grad_norm": 1.1405270099639893, "learning_rate": 6.484699690217422e-07, "loss": 0.0503, "step": 190250 }, { "epoch": 6.432266134757767, "grad_norm": 0.7021884322166443, "learning_rate": 6.471434677568666e-07, "loss": 0.049, "step": 190260 }, { "epoch": 6.4326042124480205, "grad_norm": 0.7695399522781372, "learning_rate": 6.458183157805386e-07, "loss": 0.0492, "step": 190270 }, { "epoch": 6.432942290138274, "grad_norm": 0.645574688911438, "learning_rate": 6.444945131290015e-07, "loss": 0.0521, "step": 190280 }, { "epoch": 6.433280367828527, "grad_norm": 1.3170771598815918, "learning_rate": 6.431720598384428e-07, "loss": 0.0498, "step": 190290 }, { "epoch": 6.43361844551878, "grad_norm": 1.0416321754455566, "learning_rate": 6.418509559450114e-07, "loss": 0.0473, "step": 190300 }, { "epoch": 6.433956523209034, "grad_norm": 1.6951375007629395, "learning_rate": 6.405312014848452e-07, "loss": 0.0525, "step": 190310 }, { "epoch": 6.434294600899286, "grad_norm": 1.3458400964736938, "learning_rate": 6.392127964939987e-07, "loss": 0.0492, "step": 190320 }, { "epoch": 6.43463267858954, "grad_norm": 1.454734206199646, "learning_rate": 6.378957410085318e-07, "loss": 0.0561, "step": 190330 }, { "epoch": 6.434970756279793, "grad_norm": 0.5041961669921875, "learning_rate": 6.365800350644601e-07, "loss": 0.0506, "step": 190340 }, { "epoch": 6.435308833970046, "grad_norm": 0.597133219242096, "learning_rate": 6.352656786977384e-07, "loss": 0.0467, "step": 190350 }, { "epoch": 6.4356469116602995, "grad_norm": 1.564103126525879, "learning_rate": 6.339526719443101e-07, "loss": 0.0497, "step": 190360 }, { "epoch": 6.435984989350553, "grad_norm": 0.9452283382415771, "learning_rate": 6.326410148400685e-07, "loss": 0.0534, "step": 190370 }, { "epoch": 6.436323067040806, "grad_norm": 2.1789143085479736, "learning_rate": 6.31330707420874e-07, "loss": 0.0606, "step": 190380 }, { "epoch": 6.436661144731059, "grad_norm": 0.7422147393226624, "learning_rate": 6.300217497225535e-07, "loss": 0.0581, "step": 190390 }, { "epoch": 6.436999222421313, "grad_norm": 1.244295597076416, "learning_rate": 6.287141417808895e-07, "loss": 0.0544, "step": 190400 }, { "epoch": 6.437337300111565, "grad_norm": 1.391189455986023, "learning_rate": 6.27407883631631e-07, "loss": 0.0536, "step": 190410 }, { "epoch": 6.437675377801819, "grad_norm": 0.5462530255317688, "learning_rate": 6.261029753104941e-07, "loss": 0.0524, "step": 190420 }, { "epoch": 6.438013455492072, "grad_norm": 1.2584848403930664, "learning_rate": 6.247994168531557e-07, "loss": 0.056, "step": 190430 }, { "epoch": 6.438351533182325, "grad_norm": 0.7296465635299683, "learning_rate": 6.234972082952483e-07, "loss": 0.0561, "step": 190440 }, { "epoch": 6.438689610872578, "grad_norm": 1.5243844985961914, "learning_rate": 6.221963496723826e-07, "loss": 0.0517, "step": 190450 }, { "epoch": 6.439027688562832, "grad_norm": 0.8399065136909485, "learning_rate": 6.208968410201078e-07, "loss": 0.0507, "step": 190460 }, { "epoch": 6.4393657662530845, "grad_norm": 0.5227516889572144, "learning_rate": 6.19598682373973e-07, "loss": 0.0521, "step": 190470 }, { "epoch": 6.439703843943338, "grad_norm": 2.826162338256836, "learning_rate": 6.183018737694557e-07, "loss": 0.0564, "step": 190480 }, { "epoch": 6.440041921633592, "grad_norm": 5.341006278991699, "learning_rate": 6.170064152420107e-07, "loss": 0.0508, "step": 190490 }, { "epoch": 6.440379999323845, "grad_norm": 0.8949779272079468, "learning_rate": 6.157123068270599e-07, "loss": 0.0555, "step": 190500 }, { "epoch": 6.440718077014098, "grad_norm": 0.8272786140441895, "learning_rate": 6.144195485599802e-07, "loss": 0.0549, "step": 190510 }, { "epoch": 6.441056154704351, "grad_norm": 0.7670996189117432, "learning_rate": 6.131281404761214e-07, "loss": 0.0479, "step": 190520 }, { "epoch": 6.441394232394604, "grad_norm": 1.9323179721832275, "learning_rate": 6.118380826107828e-07, "loss": 0.0528, "step": 190530 }, { "epoch": 6.441732310084857, "grad_norm": 1.1733492612838745, "learning_rate": 6.105493749992308e-07, "loss": 0.0577, "step": 190540 }, { "epoch": 6.442070387775111, "grad_norm": 1.40725839138031, "learning_rate": 6.09262017676715e-07, "loss": 0.0506, "step": 190550 }, { "epoch": 6.442408465465364, "grad_norm": 0.7416399717330933, "learning_rate": 6.079760106784127e-07, "loss": 0.0527, "step": 190560 }, { "epoch": 6.442746543155617, "grad_norm": 1.6704739332199097, "learning_rate": 6.066913540394958e-07, "loss": 0.0562, "step": 190570 }, { "epoch": 6.4430846208458705, "grad_norm": 2.1198840141296387, "learning_rate": 6.054080477950752e-07, "loss": 0.0549, "step": 190580 }, { "epoch": 6.443422698536123, "grad_norm": 2.5456702709198, "learning_rate": 6.041260919802449e-07, "loss": 0.0467, "step": 190590 }, { "epoch": 6.443760776226377, "grad_norm": 0.731929361820221, "learning_rate": 6.028454866300548e-07, "loss": 0.053, "step": 190600 }, { "epoch": 6.44409885391663, "grad_norm": 1.2194726467132568, "learning_rate": 6.015662317795045e-07, "loss": 0.0498, "step": 190610 }, { "epoch": 6.444436931606884, "grad_norm": 6.134307861328125, "learning_rate": 6.002883274635773e-07, "loss": 0.0539, "step": 190620 }, { "epoch": 6.444775009297136, "grad_norm": 1.1132545471191406, "learning_rate": 5.990117737172063e-07, "loss": 0.0534, "step": 190630 }, { "epoch": 6.44511308698739, "grad_norm": 0.7858620285987854, "learning_rate": 5.977365705752968e-07, "loss": 0.0517, "step": 190640 }, { "epoch": 6.4454511646776425, "grad_norm": 1.0172933340072632, "learning_rate": 5.9646271807271e-07, "loss": 0.0529, "step": 190650 }, { "epoch": 6.445789242367896, "grad_norm": 3.1401138305664062, "learning_rate": 5.951902162442735e-07, "loss": 0.054, "step": 190660 }, { "epoch": 6.4461273200581495, "grad_norm": 0.8557301759719849, "learning_rate": 5.939190651247762e-07, "loss": 0.0514, "step": 190670 }, { "epoch": 6.446465397748403, "grad_norm": 1.0471750497817993, "learning_rate": 5.926492647489678e-07, "loss": 0.053, "step": 190680 }, { "epoch": 6.446803475438656, "grad_norm": 1.0223515033721924, "learning_rate": 5.91380815151571e-07, "loss": 0.0504, "step": 190690 }, { "epoch": 6.447141553128909, "grad_norm": 0.8112385272979736, "learning_rate": 5.901137163672577e-07, "loss": 0.0553, "step": 190700 }, { "epoch": 6.447479630819163, "grad_norm": 0.9010557532310486, "learning_rate": 5.888479684306724e-07, "loss": 0.0552, "step": 190710 }, { "epoch": 6.447817708509415, "grad_norm": 1.7538307905197144, "learning_rate": 5.87583571376421e-07, "loss": 0.0585, "step": 190720 }, { "epoch": 6.448155786199669, "grad_norm": 0.7655194997787476, "learning_rate": 5.863205252390702e-07, "loss": 0.055, "step": 190730 }, { "epoch": 6.448493863889922, "grad_norm": 0.8245247602462769, "learning_rate": 5.85058830053159e-07, "loss": 0.0525, "step": 190740 }, { "epoch": 6.448831941580175, "grad_norm": 0.5280334949493408, "learning_rate": 5.837984858531653e-07, "loss": 0.0525, "step": 190750 }, { "epoch": 6.449170019270428, "grad_norm": 0.8307613134384155, "learning_rate": 5.825394926735617e-07, "loss": 0.0585, "step": 190760 }, { "epoch": 6.449508096960682, "grad_norm": 0.9753327965736389, "learning_rate": 5.81281850548765e-07, "loss": 0.0529, "step": 190770 }, { "epoch": 6.449846174650935, "grad_norm": 1.2063289880752563, "learning_rate": 5.800255595131532e-07, "loss": 0.0551, "step": 190780 }, { "epoch": 6.450184252341188, "grad_norm": 0.6605727076530457, "learning_rate": 5.787706196010767e-07, "loss": 0.0545, "step": 190790 }, { "epoch": 6.450522330031442, "grad_norm": 2.999466896057129, "learning_rate": 5.775170308468469e-07, "loss": 0.0573, "step": 190800 }, { "epoch": 6.450860407721694, "grad_norm": 0.7919487357139587, "learning_rate": 5.762647932847364e-07, "loss": 0.0541, "step": 190810 }, { "epoch": 6.451198485411948, "grad_norm": 1.4438260793685913, "learning_rate": 5.750139069489735e-07, "loss": 0.0544, "step": 190820 }, { "epoch": 6.451536563102201, "grad_norm": 1.2486704587936401, "learning_rate": 5.73764371873764e-07, "loss": 0.0542, "step": 190830 }, { "epoch": 6.451874640792454, "grad_norm": 0.8351584672927856, "learning_rate": 5.725161880932694e-07, "loss": 0.0543, "step": 190840 }, { "epoch": 6.452212718482707, "grad_norm": 0.8511788845062256, "learning_rate": 5.712693556416127e-07, "loss": 0.0596, "step": 190850 }, { "epoch": 6.452550796172961, "grad_norm": 2.0140163898468018, "learning_rate": 5.70023874552883e-07, "loss": 0.0498, "step": 190860 }, { "epoch": 6.4528888738632135, "grad_norm": 3.6578938961029053, "learning_rate": 5.687797448611254e-07, "loss": 0.0618, "step": 190870 }, { "epoch": 6.453226951553467, "grad_norm": 1.0394819974899292, "learning_rate": 5.675369666003627e-07, "loss": 0.0536, "step": 190880 }, { "epoch": 6.4535650292437206, "grad_norm": 0.5738291144371033, "learning_rate": 5.662955398045733e-07, "loss": 0.0463, "step": 190890 }, { "epoch": 6.453903106933973, "grad_norm": 1.0714666843414307, "learning_rate": 5.650554645076855e-07, "loss": 0.0476, "step": 190900 }, { "epoch": 6.454241184624227, "grad_norm": 0.7091912627220154, "learning_rate": 5.638167407436112e-07, "loss": 0.0524, "step": 190910 }, { "epoch": 6.45457926231448, "grad_norm": 2.2051703929901123, "learning_rate": 5.625793685462122e-07, "loss": 0.0613, "step": 190920 }, { "epoch": 6.454917340004733, "grad_norm": 1.767057180404663, "learning_rate": 5.61343347949328e-07, "loss": 0.0496, "step": 190930 }, { "epoch": 6.455255417694986, "grad_norm": 3.8323845863342285, "learning_rate": 5.601086789867371e-07, "loss": 0.0506, "step": 190940 }, { "epoch": 6.45559349538524, "grad_norm": 0.667465329170227, "learning_rate": 5.588753616922071e-07, "loss": 0.055, "step": 190950 }, { "epoch": 6.4559315730754925, "grad_norm": 0.7290300130844116, "learning_rate": 5.576433960994443e-07, "loss": 0.0516, "step": 190960 }, { "epoch": 6.456269650765746, "grad_norm": 0.8496538996696472, "learning_rate": 5.564127822421383e-07, "loss": 0.0507, "step": 190970 }, { "epoch": 6.4566077284559995, "grad_norm": 0.7255159616470337, "learning_rate": 5.551835201539346e-07, "loss": 0.057, "step": 190980 }, { "epoch": 6.456945806146252, "grad_norm": 0.5669187903404236, "learning_rate": 5.53955609868434e-07, "loss": 0.0482, "step": 190990 }, { "epoch": 6.457283883836506, "grad_norm": 0.5274244546890259, "learning_rate": 5.527290514192152e-07, "loss": 0.0472, "step": 191000 }, { "epoch": 6.457621961526759, "grad_norm": 0.6889172792434692, "learning_rate": 5.515038448398013e-07, "loss": 0.052, "step": 191010 }, { "epoch": 6.457960039217012, "grad_norm": 1.0300663709640503, "learning_rate": 5.502799901637046e-07, "loss": 0.0583, "step": 191020 }, { "epoch": 6.458298116907265, "grad_norm": 1.0470441579818726, "learning_rate": 5.490574874243648e-07, "loss": 0.0519, "step": 191030 }, { "epoch": 6.458636194597519, "grad_norm": 0.4338615834712982, "learning_rate": 5.47836336655222e-07, "loss": 0.0505, "step": 191040 }, { "epoch": 6.458974272287771, "grad_norm": 0.5309895277023315, "learning_rate": 5.466165378896548e-07, "loss": 0.0498, "step": 191050 }, { "epoch": 6.459312349978025, "grad_norm": 0.8123757243156433, "learning_rate": 5.453980911610146e-07, "loss": 0.0481, "step": 191060 }, { "epoch": 6.4596504276682785, "grad_norm": 4.946199893951416, "learning_rate": 5.441809965026079e-07, "loss": 0.0532, "step": 191070 }, { "epoch": 6.459988505358531, "grad_norm": 0.7804654836654663, "learning_rate": 5.429652539477137e-07, "loss": 0.0493, "step": 191080 }, { "epoch": 6.460326583048785, "grad_norm": 0.6236430406570435, "learning_rate": 5.417508635295665e-07, "loss": 0.0546, "step": 191090 }, { "epoch": 6.460664660739038, "grad_norm": 0.5536018013954163, "learning_rate": 5.405378252813787e-07, "loss": 0.0521, "step": 191100 }, { "epoch": 6.461002738429291, "grad_norm": 0.9452600479125977, "learning_rate": 5.393261392362959e-07, "loss": 0.0533, "step": 191110 }, { "epoch": 6.461340816119544, "grad_norm": 2.5876808166503906, "learning_rate": 5.381158054274638e-07, "loss": 0.0526, "step": 191120 }, { "epoch": 6.461678893809798, "grad_norm": 1.2085508108139038, "learning_rate": 5.369068238879559e-07, "loss": 0.0593, "step": 191130 }, { "epoch": 6.46201697150005, "grad_norm": 0.6018030643463135, "learning_rate": 5.35699194650835e-07, "loss": 0.0528, "step": 191140 }, { "epoch": 6.462355049190304, "grad_norm": 0.9163025617599487, "learning_rate": 5.344929177491186e-07, "loss": 0.0517, "step": 191150 }, { "epoch": 6.462693126880557, "grad_norm": 1.0133823156356812, "learning_rate": 5.332879932157808e-07, "loss": 0.0573, "step": 191160 }, { "epoch": 6.46303120457081, "grad_norm": 0.7821488380432129, "learning_rate": 5.320844210837617e-07, "loss": 0.0493, "step": 191170 }, { "epoch": 6.4633692822610636, "grad_norm": 3.3460874557495117, "learning_rate": 5.308822013859794e-07, "loss": 0.0532, "step": 191180 }, { "epoch": 6.463707359951317, "grad_norm": 0.9137221574783325, "learning_rate": 5.296813341552853e-07, "loss": 0.0505, "step": 191190 }, { "epoch": 6.46404543764157, "grad_norm": 0.6967272758483887, "learning_rate": 5.2848181942452e-07, "loss": 0.057, "step": 191200 }, { "epoch": 6.464383515331823, "grad_norm": 0.6009944081306458, "learning_rate": 5.272836572264739e-07, "loss": 0.0514, "step": 191210 }, { "epoch": 6.464721593022077, "grad_norm": 0.7591487169265747, "learning_rate": 5.260868475939096e-07, "loss": 0.059, "step": 191220 }, { "epoch": 6.465059670712329, "grad_norm": 1.299678921699524, "learning_rate": 5.248913905595454e-07, "loss": 0.0518, "step": 191230 }, { "epoch": 6.465397748402583, "grad_norm": 0.9174524545669556, "learning_rate": 5.236972861560607e-07, "loss": 0.054, "step": 191240 }, { "epoch": 6.465735826092836, "grad_norm": 0.6162327527999878, "learning_rate": 5.225045344161073e-07, "loss": 0.0499, "step": 191250 }, { "epoch": 6.466073903783089, "grad_norm": 1.8962042331695557, "learning_rate": 5.213131353722923e-07, "loss": 0.0513, "step": 191260 }, { "epoch": 6.4664119814733425, "grad_norm": 0.5044132471084595, "learning_rate": 5.201230890571896e-07, "loss": 0.0508, "step": 191270 }, { "epoch": 6.466750059163596, "grad_norm": 1.308775782585144, "learning_rate": 5.189343955033287e-07, "loss": 0.0497, "step": 191280 }, { "epoch": 6.467088136853849, "grad_norm": 0.6342587471008301, "learning_rate": 5.177470547432117e-07, "loss": 0.0551, "step": 191290 }, { "epoch": 6.467426214544102, "grad_norm": 0.7783909440040588, "learning_rate": 5.165610668093013e-07, "loss": 0.054, "step": 191300 }, { "epoch": 6.467764292234356, "grad_norm": 0.5681828260421753, "learning_rate": 5.15376431734027e-07, "loss": 0.0535, "step": 191310 }, { "epoch": 6.468102369924608, "grad_norm": 0.7346116900444031, "learning_rate": 5.141931495497631e-07, "loss": 0.0531, "step": 191320 }, { "epoch": 6.468440447614862, "grad_norm": 0.8986855745315552, "learning_rate": 5.130112202888726e-07, "loss": 0.056, "step": 191330 }, { "epoch": 6.468778525305115, "grad_norm": 1.7830055952072144, "learning_rate": 5.118306439836574e-07, "loss": 0.0511, "step": 191340 }, { "epoch": 6.469116602995368, "grad_norm": 0.6027148365974426, "learning_rate": 5.106514206664081e-07, "loss": 0.0588, "step": 191350 }, { "epoch": 6.4694546806856215, "grad_norm": 1.1606459617614746, "learning_rate": 5.094735503693549e-07, "loss": 0.055, "step": 191360 }, { "epoch": 6.469792758375875, "grad_norm": 1.104992151260376, "learning_rate": 5.082970331246995e-07, "loss": 0.0531, "step": 191370 }, { "epoch": 6.470130836066128, "grad_norm": 2.256622076034546, "learning_rate": 5.071218689646107e-07, "loss": 0.0577, "step": 191380 }, { "epoch": 6.470468913756381, "grad_norm": 0.8231524229049683, "learning_rate": 5.059480579212184e-07, "loss": 0.0539, "step": 191390 }, { "epoch": 6.470806991446635, "grad_norm": 0.754773736000061, "learning_rate": 5.047756000266135e-07, "loss": 0.0531, "step": 191400 }, { "epoch": 6.471145069136887, "grad_norm": 0.4915730059146881, "learning_rate": 5.036044953128427e-07, "loss": 0.0503, "step": 191410 }, { "epoch": 6.471483146827141, "grad_norm": 0.9155041575431824, "learning_rate": 5.024347438119303e-07, "loss": 0.0454, "step": 191420 }, { "epoch": 6.471821224517394, "grad_norm": 1.6740608215332031, "learning_rate": 5.01266345555862e-07, "loss": 0.0585, "step": 191430 }, { "epoch": 6.472159302207647, "grad_norm": 0.7472429275512695, "learning_rate": 5.000993005765786e-07, "loss": 0.0532, "step": 191440 }, { "epoch": 6.4724973798979, "grad_norm": 0.806283175945282, "learning_rate": 4.989336089059771e-07, "loss": 0.0551, "step": 191450 }, { "epoch": 6.472835457588154, "grad_norm": 2.4746181964874268, "learning_rate": 4.977692705759373e-07, "loss": 0.0501, "step": 191460 }, { "epoch": 6.473173535278407, "grad_norm": 2.340486764907837, "learning_rate": 4.966062856182896e-07, "loss": 0.0546, "step": 191470 }, { "epoch": 6.47351161296866, "grad_norm": 0.7688802480697632, "learning_rate": 4.95444654064825e-07, "loss": 0.0506, "step": 191480 }, { "epoch": 6.473849690658914, "grad_norm": 2.6784796714782715, "learning_rate": 4.942843759473071e-07, "loss": 0.0585, "step": 191490 }, { "epoch": 6.474187768349166, "grad_norm": 1.5861924886703491, "learning_rate": 4.931254512974548e-07, "loss": 0.0567, "step": 191500 }, { "epoch": 6.47452584603942, "grad_norm": 0.4858865737915039, "learning_rate": 4.91967880146954e-07, "loss": 0.0529, "step": 191510 }, { "epoch": 6.474863923729673, "grad_norm": 0.5660439729690552, "learning_rate": 4.908116625274572e-07, "loss": 0.0565, "step": 191520 }, { "epoch": 6.475202001419927, "grad_norm": 0.8237901926040649, "learning_rate": 4.896567984705669e-07, "loss": 0.0471, "step": 191530 }, { "epoch": 6.475540079110179, "grad_norm": 0.9989697933197021, "learning_rate": 4.885032880078521e-07, "loss": 0.047, "step": 191540 }, { "epoch": 6.475878156800433, "grad_norm": 6.527319431304932, "learning_rate": 4.873511311708657e-07, "loss": 0.0542, "step": 191550 }, { "epoch": 6.4762162344906855, "grad_norm": 1.4190456867218018, "learning_rate": 4.862003279910988e-07, "loss": 0.0564, "step": 191560 }, { "epoch": 6.476554312180939, "grad_norm": 0.960256040096283, "learning_rate": 4.850508785000096e-07, "loss": 0.0481, "step": 191570 }, { "epoch": 6.4768923898711925, "grad_norm": 0.8698275089263916, "learning_rate": 4.839027827290287e-07, "loss": 0.0542, "step": 191580 }, { "epoch": 6.477230467561446, "grad_norm": 6.309744834899902, "learning_rate": 4.827560407095422e-07, "loss": 0.0529, "step": 191590 }, { "epoch": 6.477568545251699, "grad_norm": 0.8578100204467773, "learning_rate": 4.816106524729025e-07, "loss": 0.0545, "step": 191600 }, { "epoch": 6.477906622941952, "grad_norm": 0.4354330003261566, "learning_rate": 4.804666180504236e-07, "loss": 0.0518, "step": 191610 }, { "epoch": 6.478244700632205, "grad_norm": 0.6456639766693115, "learning_rate": 4.79323937473386e-07, "loss": 0.0526, "step": 191620 }, { "epoch": 6.478582778322458, "grad_norm": 1.2137374877929688, "learning_rate": 4.781826107730258e-07, "loss": 0.0495, "step": 191630 }, { "epoch": 6.478920856012712, "grad_norm": 0.716309666633606, "learning_rate": 4.770426379805515e-07, "loss": 0.0608, "step": 191640 }, { "epoch": 6.479258933702965, "grad_norm": 1.4561628103256226, "learning_rate": 4.759040191271269e-07, "loss": 0.0525, "step": 191650 }, { "epoch": 6.479597011393218, "grad_norm": 1.0756983757019043, "learning_rate": 4.7476675424388274e-07, "loss": 0.0515, "step": 191660 }, { "epoch": 6.4799350890834715, "grad_norm": 0.8937394618988037, "learning_rate": 4.7363084336190523e-07, "loss": 0.0515, "step": 191670 }, { "epoch": 6.480273166773725, "grad_norm": 0.5195163488388062, "learning_rate": 4.7249628651225287e-07, "loss": 0.0506, "step": 191680 }, { "epoch": 6.480611244463978, "grad_norm": 3.069458246231079, "learning_rate": 4.713630837259508e-07, "loss": 0.0572, "step": 191690 }, { "epoch": 6.480949322154231, "grad_norm": 1.5250381231307983, "learning_rate": 4.702312350339688e-07, "loss": 0.0531, "step": 191700 }, { "epoch": 6.481287399844485, "grad_norm": 0.5542913675308228, "learning_rate": 4.691007404672654e-07, "loss": 0.0518, "step": 191710 }, { "epoch": 6.481625477534737, "grad_norm": 0.9119589924812317, "learning_rate": 4.679716000567325e-07, "loss": 0.0521, "step": 191720 }, { "epoch": 6.481963555224991, "grad_norm": 0.88971346616745, "learning_rate": 4.668438138332509e-07, "loss": 0.0528, "step": 191730 }, { "epoch": 6.482301632915244, "grad_norm": 2.5795819759368896, "learning_rate": 4.6571738182765166e-07, "loss": 0.0557, "step": 191740 }, { "epoch": 6.482639710605497, "grad_norm": 1.0725516080856323, "learning_rate": 4.645923040707267e-07, "loss": 0.0467, "step": 191750 }, { "epoch": 6.48297778829575, "grad_norm": 1.767280101776123, "learning_rate": 4.634685805932404e-07, "loss": 0.0481, "step": 191760 }, { "epoch": 6.483315865986004, "grad_norm": 0.5589783191680908, "learning_rate": 4.623462114259125e-07, "loss": 0.0518, "step": 191770 }, { "epoch": 6.483653943676257, "grad_norm": 1.6742420196533203, "learning_rate": 4.612251965994352e-07, "loss": 0.0524, "step": 191780 }, { "epoch": 6.48399202136651, "grad_norm": 1.5155829191207886, "learning_rate": 4.601055361444395e-07, "loss": 0.0536, "step": 191790 }, { "epoch": 6.484330099056764, "grad_norm": 1.4621765613555908, "learning_rate": 4.58987230091551e-07, "loss": 0.0514, "step": 191800 }, { "epoch": 6.484668176747016, "grad_norm": 0.7352020740509033, "learning_rate": 4.578702784713451e-07, "loss": 0.0488, "step": 191810 }, { "epoch": 6.48500625443727, "grad_norm": 0.9862872362136841, "learning_rate": 4.5675468131434753e-07, "loss": 0.0528, "step": 191820 }, { "epoch": 6.485344332127523, "grad_norm": 1.9355462789535522, "learning_rate": 4.556404386510671e-07, "loss": 0.056, "step": 191830 }, { "epoch": 6.485682409817776, "grad_norm": 4.024762153625488, "learning_rate": 4.5452755051196285e-07, "loss": 0.0495, "step": 191840 }, { "epoch": 6.486020487508029, "grad_norm": 1.678024411201477, "learning_rate": 4.5341601692746036e-07, "loss": 0.0561, "step": 191850 }, { "epoch": 6.486358565198283, "grad_norm": 0.5491859316825867, "learning_rate": 4.5230583792795767e-07, "loss": 0.0531, "step": 191860 }, { "epoch": 6.4866966428885355, "grad_norm": 0.8417643308639526, "learning_rate": 4.5119701354379153e-07, "loss": 0.0525, "step": 191870 }, { "epoch": 6.487034720578789, "grad_norm": 1.3698583841323853, "learning_rate": 4.500895438052877e-07, "loss": 0.0549, "step": 191880 }, { "epoch": 6.487372798269043, "grad_norm": 0.8789479732513428, "learning_rate": 4.4898342874272193e-07, "loss": 0.0522, "step": 191890 }, { "epoch": 6.487710875959295, "grad_norm": 0.9256457686424255, "learning_rate": 4.478786683863312e-07, "loss": 0.0487, "step": 191900 }, { "epoch": 6.488048953649549, "grad_norm": 1.7003751993179321, "learning_rate": 4.4677526276631907e-07, "loss": 0.0575, "step": 191910 }, { "epoch": 6.488387031339802, "grad_norm": 0.6501390933990479, "learning_rate": 4.456732119128615e-07, "loss": 0.0483, "step": 191920 }, { "epoch": 6.488725109030055, "grad_norm": 0.5662667751312256, "learning_rate": 4.445725158560732e-07, "loss": 0.0507, "step": 191930 }, { "epoch": 6.489063186720308, "grad_norm": 0.5867604613304138, "learning_rate": 4.434731746260634e-07, "loss": 0.0554, "step": 191940 }, { "epoch": 6.489401264410562, "grad_norm": 0.958499014377594, "learning_rate": 4.4237518825287484e-07, "loss": 0.0508, "step": 191950 }, { "epoch": 6.4897393421008145, "grad_norm": 1.065325379371643, "learning_rate": 4.412785567665334e-07, "loss": 0.0536, "step": 191960 }, { "epoch": 6.490077419791068, "grad_norm": 1.0327401161193848, "learning_rate": 4.4018328019701515e-07, "loss": 0.0547, "step": 191970 }, { "epoch": 6.4904154974813215, "grad_norm": 1.4427498579025269, "learning_rate": 4.3908935857427394e-07, "loss": 0.0502, "step": 191980 }, { "epoch": 6.490753575171574, "grad_norm": 0.846097469329834, "learning_rate": 4.379967919282024e-07, "loss": 0.0545, "step": 191990 }, { "epoch": 6.491091652861828, "grad_norm": 0.7944255471229553, "learning_rate": 4.3690558028868235e-07, "loss": 0.0515, "step": 192000 }, { "epoch": 6.491429730552081, "grad_norm": 1.8765335083007812, "learning_rate": 4.358157236855398e-07, "loss": 0.0493, "step": 192010 }, { "epoch": 6.491767808242334, "grad_norm": 0.7423710227012634, "learning_rate": 4.347272221485843e-07, "loss": 0.0623, "step": 192020 }, { "epoch": 6.492105885932587, "grad_norm": 0.6698752045631409, "learning_rate": 4.336400757075587e-07, "loss": 0.0507, "step": 192030 }, { "epoch": 6.492443963622841, "grad_norm": 1.78199303150177, "learning_rate": 4.325542843921948e-07, "loss": 0.0498, "step": 192040 }, { "epoch": 6.492782041313093, "grad_norm": 0.7105020880699158, "learning_rate": 4.314698482321744e-07, "loss": 0.0522, "step": 192050 }, { "epoch": 6.493120119003347, "grad_norm": 1.512948989868164, "learning_rate": 4.3038676725714597e-07, "loss": 0.0492, "step": 192060 }, { "epoch": 6.4934581966936005, "grad_norm": 4.6250433921813965, "learning_rate": 4.2930504149671927e-07, "loss": 0.0574, "step": 192070 }, { "epoch": 6.493796274383853, "grad_norm": 0.6770449876785278, "learning_rate": 4.282246709804705e-07, "loss": 0.0507, "step": 192080 }, { "epoch": 6.494134352074107, "grad_norm": 0.9437431693077087, "learning_rate": 4.271456557379372e-07, "loss": 0.0529, "step": 192090 }, { "epoch": 6.49447242976436, "grad_norm": 1.187446117401123, "learning_rate": 4.26067995798618e-07, "loss": 0.0507, "step": 192100 }, { "epoch": 6.494810507454613, "grad_norm": 0.9517366290092468, "learning_rate": 4.2499169119197826e-07, "loss": 0.056, "step": 192110 }, { "epoch": 6.495148585144866, "grad_norm": 1.1574311256408691, "learning_rate": 4.2391674194743323e-07, "loss": 0.0445, "step": 192120 }, { "epoch": 6.49548666283512, "grad_norm": 0.6555088758468628, "learning_rate": 4.228431480943873e-07, "loss": 0.0541, "step": 192130 }, { "epoch": 6.495824740525372, "grad_norm": 0.9663295745849609, "learning_rate": 4.217709096621836e-07, "loss": 0.0553, "step": 192140 }, { "epoch": 6.496162818215626, "grad_norm": 0.7250504493713379, "learning_rate": 4.20700026680132e-07, "loss": 0.0523, "step": 192150 }, { "epoch": 6.496500895905879, "grad_norm": 0.42108485102653503, "learning_rate": 4.196304991775202e-07, "loss": 0.0519, "step": 192160 }, { "epoch": 6.496838973596132, "grad_norm": 0.49947866797447205, "learning_rate": 4.185623271835859e-07, "loss": 0.0509, "step": 192170 }, { "epoch": 6.497177051286386, "grad_norm": 2.715416193008423, "learning_rate": 4.17495510727528e-07, "loss": 0.049, "step": 192180 }, { "epoch": 6.497515128976639, "grad_norm": 1.2054327726364136, "learning_rate": 4.164300498385176e-07, "loss": 0.0521, "step": 192190 }, { "epoch": 6.497853206666892, "grad_norm": 0.6700245141983032, "learning_rate": 4.153659445456759e-07, "loss": 0.0531, "step": 192200 }, { "epoch": 6.498191284357145, "grad_norm": 1.6553937196731567, "learning_rate": 4.143031948781073e-07, "loss": 0.0577, "step": 192210 }, { "epoch": 6.498529362047399, "grad_norm": 0.5002315640449524, "learning_rate": 4.132418008648553e-07, "loss": 0.0484, "step": 192220 }, { "epoch": 6.498867439737651, "grad_norm": 0.5246706008911133, "learning_rate": 4.121817625349522e-07, "loss": 0.0486, "step": 192230 }, { "epoch": 6.499205517427905, "grad_norm": 1.1151485443115234, "learning_rate": 4.111230799173638e-07, "loss": 0.0499, "step": 192240 }, { "epoch": 6.499543595118158, "grad_norm": 1.6557170152664185, "learning_rate": 4.1006575304103903e-07, "loss": 0.0542, "step": 192250 }, { "epoch": 6.499881672808411, "grad_norm": 0.6289327144622803, "learning_rate": 4.090097819348937e-07, "loss": 0.0534, "step": 192260 }, { "epoch": 6.5002197504986645, "grad_norm": 0.8862856030464172, "learning_rate": 4.079551666277881e-07, "loss": 0.0524, "step": 192270 }, { "epoch": 6.500557828188918, "grad_norm": 1.4119950532913208, "learning_rate": 4.069019071485547e-07, "loss": 0.052, "step": 192280 }, { "epoch": 6.500895905879171, "grad_norm": 0.43895480036735535, "learning_rate": 4.058500035259927e-07, "loss": 0.0523, "step": 192290 }, { "epoch": 6.501233983569424, "grad_norm": 1.858514428138733, "learning_rate": 4.047994557888568e-07, "loss": 0.0573, "step": 192300 }, { "epoch": 6.501572061259678, "grad_norm": 2.2755281925201416, "learning_rate": 4.0375026396587966e-07, "loss": 0.0499, "step": 192310 }, { "epoch": 6.50191013894993, "grad_norm": 0.6867233514785767, "learning_rate": 4.0270242808573275e-07, "loss": 0.0515, "step": 192320 }, { "epoch": 6.502248216640184, "grad_norm": 0.9545533657073975, "learning_rate": 4.0165594817706544e-07, "loss": 0.0572, "step": 192330 }, { "epoch": 6.502586294330437, "grad_norm": 1.0278449058532715, "learning_rate": 4.006108242684936e-07, "loss": 0.0539, "step": 192340 }, { "epoch": 6.50292437202069, "grad_norm": 0.4238431453704834, "learning_rate": 3.995670563885945e-07, "loss": 0.0523, "step": 192350 }, { "epoch": 6.5032624497109435, "grad_norm": 0.6947270035743713, "learning_rate": 3.985246445658952e-07, "loss": 0.0528, "step": 192360 }, { "epoch": 6.503600527401197, "grad_norm": 1.7599693536758423, "learning_rate": 3.974835888288897e-07, "loss": 0.0566, "step": 192370 }, { "epoch": 6.5039386050914505, "grad_norm": 2.2309439182281494, "learning_rate": 3.964438892060551e-07, "loss": 0.0594, "step": 192380 }, { "epoch": 6.504276682781703, "grad_norm": 0.8265098929405212, "learning_rate": 3.954055457258132e-07, "loss": 0.0484, "step": 192390 }, { "epoch": 6.504614760471957, "grad_norm": 0.7190181016921997, "learning_rate": 3.943685584165413e-07, "loss": 0.0551, "step": 192400 }, { "epoch": 6.504952838162209, "grad_norm": 0.635051965713501, "learning_rate": 3.933329273066e-07, "loss": 0.0555, "step": 192410 }, { "epoch": 6.505290915852463, "grad_norm": 0.47218671441078186, "learning_rate": 3.922986524243e-07, "loss": 0.052, "step": 192420 }, { "epoch": 6.505628993542716, "grad_norm": 0.5332768559455872, "learning_rate": 3.912657337979131e-07, "loss": 0.0478, "step": 192430 }, { "epoch": 6.50596707123297, "grad_norm": 0.6556293964385986, "learning_rate": 3.9023417145568897e-07, "loss": 0.055, "step": 192440 }, { "epoch": 6.506305148923222, "grad_norm": 1.4680073261260986, "learning_rate": 3.892039654258273e-07, "loss": 0.0544, "step": 192450 }, { "epoch": 6.506643226613476, "grad_norm": 0.870566725730896, "learning_rate": 3.881751157364832e-07, "loss": 0.0472, "step": 192460 }, { "epoch": 6.506981304303729, "grad_norm": 0.5014211535453796, "learning_rate": 3.8714762241580103e-07, "loss": 0.0472, "step": 192470 }, { "epoch": 6.507319381993982, "grad_norm": 0.670028805732727, "learning_rate": 3.861214854918638e-07, "loss": 0.0518, "step": 192480 }, { "epoch": 6.507657459684236, "grad_norm": 0.5862621068954468, "learning_rate": 3.850967049927212e-07, "loss": 0.047, "step": 192490 }, { "epoch": 6.507995537374489, "grad_norm": 0.7618396282196045, "learning_rate": 3.840732809463954e-07, "loss": 0.0529, "step": 192500 }, { "epoch": 6.508333615064742, "grad_norm": 0.9588840007781982, "learning_rate": 3.8305121338086946e-07, "loss": 0.0522, "step": 192510 }, { "epoch": 6.508671692754995, "grad_norm": 0.6393513679504395, "learning_rate": 3.820305023240878e-07, "loss": 0.0549, "step": 192520 }, { "epoch": 6.509009770445248, "grad_norm": 1.2733712196350098, "learning_rate": 3.810111478039391e-07, "loss": 0.0589, "step": 192530 }, { "epoch": 6.509347848135501, "grad_norm": 0.5547070503234863, "learning_rate": 3.799931498483122e-07, "loss": 0.0527, "step": 192540 }, { "epoch": 6.509685925825755, "grad_norm": 1.3134022951126099, "learning_rate": 3.789765084850294e-07, "loss": 0.0529, "step": 192550 }, { "epoch": 6.510024003516008, "grad_norm": 0.6237563490867615, "learning_rate": 3.7796122374189056e-07, "loss": 0.0571, "step": 192560 }, { "epoch": 6.510362081206261, "grad_norm": 0.5869527459144592, "learning_rate": 3.7694729564665135e-07, "loss": 0.0545, "step": 192570 }, { "epoch": 6.5107001588965145, "grad_norm": 1.031884789466858, "learning_rate": 3.759347242270228e-07, "loss": 0.0518, "step": 192580 }, { "epoch": 6.511038236586767, "grad_norm": 0.7891429662704468, "learning_rate": 3.7492350951069956e-07, "loss": 0.0523, "step": 192590 }, { "epoch": 6.511376314277021, "grad_norm": 0.9291334748268127, "learning_rate": 3.739136515253261e-07, "loss": 0.056, "step": 192600 }, { "epoch": 6.511714391967274, "grad_norm": 0.7977811694145203, "learning_rate": 3.729051502985081e-07, "loss": 0.0483, "step": 192610 }, { "epoch": 6.512052469657528, "grad_norm": 0.7049038410186768, "learning_rate": 3.7189800585781806e-07, "loss": 0.0526, "step": 192620 }, { "epoch": 6.51239054734778, "grad_norm": 0.934531569480896, "learning_rate": 3.7089221823078943e-07, "loss": 0.0507, "step": 192630 }, { "epoch": 6.512728625038034, "grad_norm": 1.5644018650054932, "learning_rate": 3.6988778744492804e-07, "loss": 0.0521, "step": 192640 }, { "epoch": 6.5130667027282865, "grad_norm": 1.0808285474777222, "learning_rate": 3.688847135276896e-07, "loss": 0.0544, "step": 192650 }, { "epoch": 6.51340478041854, "grad_norm": 1.041740894317627, "learning_rate": 3.678829965064967e-07, "loss": 0.0512, "step": 192660 }, { "epoch": 6.5137428581087935, "grad_norm": 1.9820486307144165, "learning_rate": 3.6688263640873855e-07, "loss": 0.0521, "step": 192670 }, { "epoch": 6.514080935799047, "grad_norm": 1.0442554950714111, "learning_rate": 3.658836332617599e-07, "loss": 0.052, "step": 192680 }, { "epoch": 6.5144190134893, "grad_norm": 0.9970294833183289, "learning_rate": 3.648859870928778e-07, "loss": 0.0508, "step": 192690 }, { "epoch": 6.514757091179553, "grad_norm": 0.7028115391731262, "learning_rate": 3.6388969792935933e-07, "loss": 0.0558, "step": 192700 }, { "epoch": 6.515095168869807, "grad_norm": 1.5650179386138916, "learning_rate": 3.6289476579845494e-07, "loss": 0.0513, "step": 192710 }, { "epoch": 6.515433246560059, "grad_norm": 2.9955992698669434, "learning_rate": 3.619011907273595e-07, "loss": 0.0501, "step": 192720 }, { "epoch": 6.515771324250313, "grad_norm": 0.7925661206245422, "learning_rate": 3.6090897274324023e-07, "loss": 0.0441, "step": 192730 }, { "epoch": 6.516109401940566, "grad_norm": 1.5674954652786255, "learning_rate": 3.5991811187321425e-07, "loss": 0.055, "step": 192740 }, { "epoch": 6.516447479630819, "grad_norm": 0.5986725091934204, "learning_rate": 3.589286081443821e-07, "loss": 0.0547, "step": 192750 }, { "epoch": 6.5167855573210725, "grad_norm": 0.8587197065353394, "learning_rate": 3.5794046158379445e-07, "loss": 0.0509, "step": 192760 }, { "epoch": 6.517123635011326, "grad_norm": 1.6412339210510254, "learning_rate": 3.56953672218463e-07, "loss": 0.0556, "step": 192770 }, { "epoch": 6.517461712701579, "grad_norm": 0.6131212711334229, "learning_rate": 3.5596824007536613e-07, "loss": 0.0494, "step": 192780 }, { "epoch": 6.517799790391832, "grad_norm": 2.014441967010498, "learning_rate": 3.5498416518144895e-07, "loss": 0.0554, "step": 192790 }, { "epoch": 6.518137868082086, "grad_norm": 1.3918969631195068, "learning_rate": 3.540014475636122e-07, "loss": 0.0531, "step": 192800 }, { "epoch": 6.518475945772338, "grad_norm": 1.2428710460662842, "learning_rate": 3.530200872487288e-07, "loss": 0.0538, "step": 192810 }, { "epoch": 6.518814023462592, "grad_norm": 0.535417377948761, "learning_rate": 3.520400842636218e-07, "loss": 0.0473, "step": 192820 }, { "epoch": 6.519152101152845, "grad_norm": 0.8597002029418945, "learning_rate": 3.5106143863508635e-07, "loss": 0.0475, "step": 192830 }, { "epoch": 6.519490178843098, "grad_norm": 1.1757858991622925, "learning_rate": 3.500841503898788e-07, "loss": 0.0493, "step": 192840 }, { "epoch": 6.519828256533351, "grad_norm": 0.8189327716827393, "learning_rate": 3.491082195547224e-07, "loss": 0.0525, "step": 192850 }, { "epoch": 6.520166334223605, "grad_norm": 0.6493304371833801, "learning_rate": 3.481336461562901e-07, "loss": 0.0513, "step": 192860 }, { "epoch": 6.5205044119138575, "grad_norm": 0.763713538646698, "learning_rate": 3.471604302212328e-07, "loss": 0.0527, "step": 192870 }, { "epoch": 6.520842489604111, "grad_norm": 0.6098436117172241, "learning_rate": 3.461885717761515e-07, "loss": 0.0601, "step": 192880 }, { "epoch": 6.521180567294365, "grad_norm": 0.6679440140724182, "learning_rate": 3.452180708476249e-07, "loss": 0.0552, "step": 192890 }, { "epoch": 6.521518644984617, "grad_norm": 0.2721174359321594, "learning_rate": 3.4424892746218183e-07, "loss": 0.0467, "step": 192900 }, { "epoch": 6.521856722674871, "grad_norm": 1.1930147409439087, "learning_rate": 3.4328114164631774e-07, "loss": 0.0513, "step": 192910 }, { "epoch": 6.522194800365124, "grad_norm": 0.7738348841667175, "learning_rate": 3.4231471342648924e-07, "loss": 0.0509, "step": 192920 }, { "epoch": 6.522532878055377, "grad_norm": 0.5572313666343689, "learning_rate": 3.413496428291196e-07, "loss": 0.0498, "step": 192930 }, { "epoch": 6.52287095574563, "grad_norm": 0.8296469449996948, "learning_rate": 3.403859298805989e-07, "loss": 0.0509, "step": 192940 }, { "epoch": 6.523209033435884, "grad_norm": 1.2013601064682007, "learning_rate": 3.394235746072727e-07, "loss": 0.0529, "step": 192950 }, { "epoch": 6.5235471111261365, "grad_norm": 0.7823550701141357, "learning_rate": 3.384625770354477e-07, "loss": 0.0555, "step": 192960 }, { "epoch": 6.52388518881639, "grad_norm": 1.3096840381622314, "learning_rate": 3.375029371913974e-07, "loss": 0.0563, "step": 192970 }, { "epoch": 6.5242232665066435, "grad_norm": 1.2746765613555908, "learning_rate": 3.365446551013618e-07, "loss": 0.052, "step": 192980 }, { "epoch": 6.524561344196896, "grad_norm": 1.3078665733337402, "learning_rate": 3.355877307915367e-07, "loss": 0.0513, "step": 192990 }, { "epoch": 6.52489942188715, "grad_norm": 0.7897768616676331, "learning_rate": 3.3463216428807896e-07, "loss": 0.0528, "step": 193000 }, { "epoch": 6.525237499577403, "grad_norm": 1.7549635171890259, "learning_rate": 3.336779556171232e-07, "loss": 0.0507, "step": 193010 }, { "epoch": 6.525575577267656, "grad_norm": 0.9896904230117798, "learning_rate": 3.327251048047597e-07, "loss": 0.0483, "step": 193020 }, { "epoch": 6.525913654957909, "grad_norm": 0.9956214427947998, "learning_rate": 3.3177361187702314e-07, "loss": 0.0515, "step": 193030 }, { "epoch": 6.526251732648163, "grad_norm": 0.48621493577957153, "learning_rate": 3.3082347685994274e-07, "loss": 0.05, "step": 193040 }, { "epoch": 6.5265898103384155, "grad_norm": 4.553131103515625, "learning_rate": 3.298746997794866e-07, "loss": 0.0517, "step": 193050 }, { "epoch": 6.526927888028669, "grad_norm": 4.629168510437012, "learning_rate": 3.2892728066159504e-07, "loss": 0.051, "step": 193060 }, { "epoch": 6.5272659657189225, "grad_norm": 1.4510589838027954, "learning_rate": 3.279812195321752e-07, "loss": 0.049, "step": 193070 }, { "epoch": 6.527604043409175, "grad_norm": 0.753279447555542, "learning_rate": 3.270365164170841e-07, "loss": 0.0475, "step": 193080 }, { "epoch": 6.527942121099429, "grad_norm": 0.950043797492981, "learning_rate": 3.260931713421567e-07, "loss": 0.0486, "step": 193090 }, { "epoch": 6.528280198789682, "grad_norm": 0.8799218535423279, "learning_rate": 3.251511843331778e-07, "loss": 0.0491, "step": 193100 }, { "epoch": 6.528618276479935, "grad_norm": 0.4527425765991211, "learning_rate": 3.2421055541590474e-07, "loss": 0.0544, "step": 193110 }, { "epoch": 6.528956354170188, "grad_norm": 0.6425865888595581, "learning_rate": 3.2327128461605014e-07, "loss": 0.0555, "step": 193120 }, { "epoch": 6.529294431860442, "grad_norm": 0.7345015406608582, "learning_rate": 3.223333719592991e-07, "loss": 0.0477, "step": 193130 }, { "epoch": 6.529632509550694, "grad_norm": 1.0295501947402954, "learning_rate": 3.213968174712867e-07, "loss": 0.0531, "step": 193140 }, { "epoch": 6.529970587240948, "grad_norm": 0.8001723289489746, "learning_rate": 3.2046162117762567e-07, "loss": 0.0545, "step": 193150 }, { "epoch": 6.530308664931201, "grad_norm": 0.8564051985740662, "learning_rate": 3.1952778310387896e-07, "loss": 0.0508, "step": 193160 }, { "epoch": 6.530646742621454, "grad_norm": 0.9558455944061279, "learning_rate": 3.185953032755762e-07, "loss": 0.0591, "step": 193170 }, { "epoch": 6.530984820311708, "grad_norm": 1.9584115743637085, "learning_rate": 3.1766418171821356e-07, "loss": 0.0487, "step": 193180 }, { "epoch": 6.531322898001961, "grad_norm": 0.7873365879058838, "learning_rate": 3.1673441845725405e-07, "loss": 0.0525, "step": 193190 }, { "epoch": 6.531660975692214, "grad_norm": 0.8931537866592407, "learning_rate": 3.1580601351809957e-07, "loss": 0.0537, "step": 193200 }, { "epoch": 6.531999053382467, "grad_norm": 0.7038564085960388, "learning_rate": 3.1487896692615205e-07, "loss": 0.0517, "step": 193210 }, { "epoch": 6.532337131072721, "grad_norm": 2.426429271697998, "learning_rate": 3.139532787067467e-07, "loss": 0.0525, "step": 193220 }, { "epoch": 6.532675208762973, "grad_norm": 0.6205397248268127, "learning_rate": 3.130289488851912e-07, "loss": 0.0463, "step": 193230 }, { "epoch": 6.533013286453227, "grad_norm": 0.8308476805686951, "learning_rate": 3.1210597748675407e-07, "loss": 0.0519, "step": 193240 }, { "epoch": 6.53335136414348, "grad_norm": 0.863861083984375, "learning_rate": 3.111843645366819e-07, "loss": 0.0526, "step": 193250 }, { "epoch": 6.533689441833733, "grad_norm": 0.40420395135879517, "learning_rate": 3.1026411006015443e-07, "loss": 0.0546, "step": 193260 }, { "epoch": 6.5340275195239865, "grad_norm": 0.7653226256370544, "learning_rate": 3.093452140823405e-07, "loss": 0.0566, "step": 193270 }, { "epoch": 6.53436559721424, "grad_norm": 1.0552858114242554, "learning_rate": 3.084276766283589e-07, "loss": 0.0521, "step": 193280 }, { "epoch": 6.534703674904493, "grad_norm": 0.80390864610672, "learning_rate": 3.075114977232951e-07, "loss": 0.0518, "step": 193290 }, { "epoch": 6.535041752594746, "grad_norm": 1.2102493047714233, "learning_rate": 3.065966773922013e-07, "loss": 0.0527, "step": 193300 }, { "epoch": 6.535379830285, "grad_norm": 0.5696956515312195, "learning_rate": 3.0568321566009085e-07, "loss": 0.0497, "step": 193310 }, { "epoch": 6.535717907975252, "grad_norm": 0.7280354499816895, "learning_rate": 3.0477111255192704e-07, "loss": 0.053, "step": 193320 }, { "epoch": 6.536055985665506, "grad_norm": 1.3072761297225952, "learning_rate": 3.0386036809264553e-07, "loss": 0.0557, "step": 193330 }, { "epoch": 6.536394063355759, "grad_norm": 1.0888971090316772, "learning_rate": 3.029509823071541e-07, "loss": 0.0485, "step": 193340 }, { "epoch": 6.536732141046013, "grad_norm": 0.34678173065185547, "learning_rate": 3.0204295522031634e-07, "loss": 0.05, "step": 193350 }, { "epoch": 6.5370702187362655, "grad_norm": 0.6940165162086487, "learning_rate": 3.011362868569567e-07, "loss": 0.0483, "step": 193360 }, { "epoch": 6.537408296426519, "grad_norm": 0.5239622592926025, "learning_rate": 3.0023097724185547e-07, "loss": 0.0537, "step": 193370 }, { "epoch": 6.537746374116772, "grad_norm": 6.732631206512451, "learning_rate": 2.9932702639976496e-07, "loss": 0.0471, "step": 193380 }, { "epoch": 6.538084451807025, "grad_norm": 0.8129543662071228, "learning_rate": 2.9842443435540437e-07, "loss": 0.0496, "step": 193390 }, { "epoch": 6.538422529497279, "grad_norm": 0.8194596767425537, "learning_rate": 2.9752320113344834e-07, "loss": 0.0555, "step": 193400 }, { "epoch": 6.538760607187532, "grad_norm": 0.6635087728500366, "learning_rate": 2.9662332675853835e-07, "loss": 0.05, "step": 193410 }, { "epoch": 6.539098684877785, "grad_norm": 2.2670042514801025, "learning_rate": 2.957248112552713e-07, "loss": 0.0527, "step": 193420 }, { "epoch": 6.539436762568038, "grad_norm": 1.0114024877548218, "learning_rate": 2.9482765464821093e-07, "loss": 0.0547, "step": 193430 }, { "epoch": 6.539774840258291, "grad_norm": 0.7536861896514893, "learning_rate": 2.939318569618987e-07, "loss": 0.0529, "step": 193440 }, { "epoch": 6.540112917948544, "grad_norm": 0.4899666905403137, "learning_rate": 2.9303741822080956e-07, "loss": 0.0559, "step": 193450 }, { "epoch": 6.540450995638798, "grad_norm": 0.8205596208572388, "learning_rate": 2.9214433844940157e-07, "loss": 0.0502, "step": 193460 }, { "epoch": 6.5407890733290515, "grad_norm": 0.5175672769546509, "learning_rate": 2.912526176720942e-07, "loss": 0.0538, "step": 193470 }, { "epoch": 6.541127151019304, "grad_norm": 1.1660255193710327, "learning_rate": 2.9036225591326795e-07, "loss": 0.0545, "step": 193480 }, { "epoch": 6.541465228709558, "grad_norm": 0.6832950115203857, "learning_rate": 2.894732531972644e-07, "loss": 0.0572, "step": 193490 }, { "epoch": 6.54180330639981, "grad_norm": 0.8075547218322754, "learning_rate": 2.8858560954838075e-07, "loss": 0.0521, "step": 193500 }, { "epoch": 6.542141384090064, "grad_norm": 1.2963284254074097, "learning_rate": 2.876993249908921e-07, "loss": 0.0557, "step": 193510 }, { "epoch": 6.542479461780317, "grad_norm": 0.5682944059371948, "learning_rate": 2.868143995490291e-07, "loss": 0.0538, "step": 193520 }, { "epoch": 6.542817539470571, "grad_norm": 1.3567591905593872, "learning_rate": 2.8593083324698345e-07, "loss": 0.0479, "step": 193530 }, { "epoch": 6.543155617160823, "grad_norm": 1.0111624002456665, "learning_rate": 2.850486261089136e-07, "loss": 0.0467, "step": 193540 }, { "epoch": 6.543493694851077, "grad_norm": 1.0229275226593018, "learning_rate": 2.8416777815893914e-07, "loss": 0.047, "step": 193550 }, { "epoch": 6.5438317725413295, "grad_norm": 0.49530088901519775, "learning_rate": 2.832882894211353e-07, "loss": 0.0492, "step": 193560 }, { "epoch": 6.544169850231583, "grad_norm": 0.4870278239250183, "learning_rate": 2.82410159919555e-07, "loss": 0.0511, "step": 193570 }, { "epoch": 6.5445079279218366, "grad_norm": 1.0059664249420166, "learning_rate": 2.815333896782013e-07, "loss": 0.0512, "step": 193580 }, { "epoch": 6.54484600561209, "grad_norm": 1.2167439460754395, "learning_rate": 2.806579787210495e-07, "loss": 0.0526, "step": 193590 }, { "epoch": 6.545184083302343, "grad_norm": 0.9691763520240784, "learning_rate": 2.797839270720304e-07, "loss": 0.0558, "step": 193600 }, { "epoch": 6.545522160992596, "grad_norm": 0.4514923691749573, "learning_rate": 2.78911234755036e-07, "loss": 0.0466, "step": 193610 }, { "epoch": 6.545860238682849, "grad_norm": 1.1224026679992676, "learning_rate": 2.780399017939306e-07, "loss": 0.0534, "step": 193620 }, { "epoch": 6.546198316373102, "grad_norm": 1.8155550956726074, "learning_rate": 2.7716992821253395e-07, "loss": 0.0579, "step": 193630 }, { "epoch": 6.546536394063356, "grad_norm": 1.3720351457595825, "learning_rate": 2.7630131403463264e-07, "loss": 0.0526, "step": 193640 }, { "epoch": 6.546874471753609, "grad_norm": 4.793300151824951, "learning_rate": 2.7543405928396883e-07, "loss": 0.0503, "step": 193650 }, { "epoch": 6.547212549443862, "grad_norm": 1.4857800006866455, "learning_rate": 2.7456816398426235e-07, "loss": 0.0497, "step": 193660 }, { "epoch": 6.5475506271341155, "grad_norm": 0.8316403031349182, "learning_rate": 2.737036281591776e-07, "loss": 0.053, "step": 193670 }, { "epoch": 6.547888704824369, "grad_norm": 0.7452545762062073, "learning_rate": 2.728404518323568e-07, "loss": 0.0484, "step": 193680 }, { "epoch": 6.548226782514622, "grad_norm": 0.5233886241912842, "learning_rate": 2.7197863502739787e-07, "loss": 0.0527, "step": 193690 }, { "epoch": 6.548564860204875, "grad_norm": 2.5806355476379395, "learning_rate": 2.711181777678595e-07, "loss": 0.0499, "step": 193700 }, { "epoch": 6.548902937895129, "grad_norm": 0.96687912940979, "learning_rate": 2.7025908007726755e-07, "loss": 0.0429, "step": 193710 }, { "epoch": 6.549241015585381, "grad_norm": 0.9554963111877441, "learning_rate": 2.694013419791086e-07, "loss": 0.0516, "step": 193720 }, { "epoch": 6.549579093275635, "grad_norm": 1.2431343793869019, "learning_rate": 2.685449634968362e-07, "loss": 0.0558, "step": 193730 }, { "epoch": 6.549917170965888, "grad_norm": 0.8968905806541443, "learning_rate": 2.6768994465385945e-07, "loss": 0.0523, "step": 193740 }, { "epoch": 6.550255248656141, "grad_norm": 0.9338793754577637, "learning_rate": 2.6683628547355955e-07, "loss": 0.0525, "step": 193750 }, { "epoch": 6.5505933263463945, "grad_norm": 2.7356691360473633, "learning_rate": 2.659839859792623e-07, "loss": 0.0507, "step": 193760 }, { "epoch": 6.550931404036648, "grad_norm": 0.7622181177139282, "learning_rate": 2.65133046194288e-07, "loss": 0.0516, "step": 193770 }, { "epoch": 6.551269481726901, "grad_norm": 0.565604031085968, "learning_rate": 2.6428346614189026e-07, "loss": 0.0561, "step": 193780 }, { "epoch": 6.551607559417154, "grad_norm": 0.4511149525642395, "learning_rate": 2.634352458452949e-07, "loss": 0.0475, "step": 193790 }, { "epoch": 6.551945637107408, "grad_norm": 0.5251466631889343, "learning_rate": 2.6258838532770004e-07, "loss": 0.0482, "step": 193800 }, { "epoch": 6.55228371479766, "grad_norm": 0.9832610487937927, "learning_rate": 2.6174288461224827e-07, "loss": 0.0541, "step": 193810 }, { "epoch": 6.552621792487914, "grad_norm": 1.6445817947387695, "learning_rate": 2.6089874372206e-07, "loss": 0.0612, "step": 193820 }, { "epoch": 6.552959870178167, "grad_norm": 0.48308902978897095, "learning_rate": 2.6005596268021683e-07, "loss": 0.0558, "step": 193830 }, { "epoch": 6.55329794786842, "grad_norm": 1.794666051864624, "learning_rate": 2.5921454150975577e-07, "loss": 0.0524, "step": 193840 }, { "epoch": 6.553636025558673, "grad_norm": 0.5877527594566345, "learning_rate": 2.583744802336807e-07, "loss": 0.0567, "step": 193850 }, { "epoch": 6.553974103248927, "grad_norm": 0.7193230390548706, "learning_rate": 2.5753577887496216e-07, "loss": 0.0521, "step": 193860 }, { "epoch": 6.5543121809391796, "grad_norm": 2.6898691654205322, "learning_rate": 2.5669843745653176e-07, "loss": 0.0538, "step": 193870 }, { "epoch": 6.554650258629433, "grad_norm": 1.2033814191818237, "learning_rate": 2.558624560012712e-07, "loss": 0.0534, "step": 193880 }, { "epoch": 6.554988336319687, "grad_norm": 0.5719309449195862, "learning_rate": 2.550278345320456e-07, "loss": 0.0528, "step": 193890 }, { "epoch": 6.555326414009939, "grad_norm": 1.0716944932937622, "learning_rate": 2.541945730716755e-07, "loss": 0.0577, "step": 193900 }, { "epoch": 6.555664491700193, "grad_norm": 0.9006562829017639, "learning_rate": 2.533626716429316e-07, "loss": 0.0501, "step": 193910 }, { "epoch": 6.556002569390446, "grad_norm": 3.1371450424194336, "learning_rate": 2.525321302685624e-07, "loss": 0.056, "step": 193920 }, { "epoch": 6.556340647080699, "grad_norm": 2.9526047706604004, "learning_rate": 2.5170294897127744e-07, "loss": 0.0567, "step": 193930 }, { "epoch": 6.556678724770952, "grad_norm": 0.9506266117095947, "learning_rate": 2.5087512777374754e-07, "loss": 0.0525, "step": 193940 }, { "epoch": 6.557016802461206, "grad_norm": 1.024609923362732, "learning_rate": 2.500486666985991e-07, "loss": 0.0483, "step": 193950 }, { "epoch": 6.5573548801514585, "grad_norm": 2.244114398956299, "learning_rate": 2.4922356576843053e-07, "loss": 0.052, "step": 193960 }, { "epoch": 6.557692957841712, "grad_norm": 1.0037033557891846, "learning_rate": 2.4839982500579617e-07, "loss": 0.0511, "step": 193970 }, { "epoch": 6.5580310355319655, "grad_norm": 1.030739188194275, "learning_rate": 2.475774444332224e-07, "loss": 0.0566, "step": 193980 }, { "epoch": 6.558369113222218, "grad_norm": 0.5224007368087769, "learning_rate": 2.4675642407319124e-07, "loss": 0.0571, "step": 193990 }, { "epoch": 6.558707190912472, "grad_norm": 0.9292539954185486, "learning_rate": 2.459367639481458e-07, "loss": 0.0482, "step": 194000 }, { "epoch": 6.559045268602725, "grad_norm": 0.48402687907218933, "learning_rate": 2.4511846408050154e-07, "loss": 0.0522, "step": 194010 }, { "epoch": 6.559383346292978, "grad_norm": 0.602703332901001, "learning_rate": 2.443015244926239e-07, "loss": 0.0424, "step": 194020 }, { "epoch": 6.559721423983231, "grad_norm": 0.9492541551589966, "learning_rate": 2.43485945206845e-07, "loss": 0.0534, "step": 194030 }, { "epoch": 6.560059501673485, "grad_norm": 0.6888300776481628, "learning_rate": 2.4267172624547476e-07, "loss": 0.0502, "step": 194040 }, { "epoch": 6.5603975793637375, "grad_norm": 1.734143614768982, "learning_rate": 2.41858867630762e-07, "loss": 0.0519, "step": 194050 }, { "epoch": 6.560735657053991, "grad_norm": 2.403397798538208, "learning_rate": 2.410473693849391e-07, "loss": 0.0464, "step": 194060 }, { "epoch": 6.5610737347442445, "grad_norm": 0.5906816720962524, "learning_rate": 2.402372315301882e-07, "loss": 0.0518, "step": 194070 }, { "epoch": 6.561411812434497, "grad_norm": 1.528954029083252, "learning_rate": 2.3942845408865266e-07, "loss": 0.0501, "step": 194080 }, { "epoch": 6.561749890124751, "grad_norm": 0.7427257895469666, "learning_rate": 2.386210370824482e-07, "loss": 0.0537, "step": 194090 }, { "epoch": 6.562087967815004, "grad_norm": 0.42150068283081055, "learning_rate": 2.3781498053365714e-07, "loss": 0.0513, "step": 194100 }, { "epoch": 6.562426045505257, "grad_norm": 1.9127949476242065, "learning_rate": 2.370102844643063e-07, "loss": 0.0527, "step": 194110 }, { "epoch": 6.56276412319551, "grad_norm": 0.8175609111785889, "learning_rate": 2.362069488963947e-07, "loss": 0.05, "step": 194120 }, { "epoch": 6.563102200885764, "grad_norm": 0.6638332009315491, "learning_rate": 2.3540497385189374e-07, "loss": 0.05, "step": 194130 }, { "epoch": 6.563440278576016, "grad_norm": 2.840848684310913, "learning_rate": 2.3460435935272472e-07, "loss": 0.0566, "step": 194140 }, { "epoch": 6.56377835626627, "grad_norm": 0.9957028031349182, "learning_rate": 2.3380510542077572e-07, "loss": 0.053, "step": 194150 }, { "epoch": 6.5641164339565234, "grad_norm": 0.49560031294822693, "learning_rate": 2.3300721207790143e-07, "loss": 0.0501, "step": 194160 }, { "epoch": 6.564454511646776, "grad_norm": 1.509741187095642, "learning_rate": 2.3221067934590667e-07, "loss": 0.0485, "step": 194170 }, { "epoch": 6.56479258933703, "grad_norm": 1.0920755863189697, "learning_rate": 2.314155072465851e-07, "loss": 0.0485, "step": 194180 }, { "epoch": 6.565130667027283, "grad_norm": 3.2706737518310547, "learning_rate": 2.3062169580165826e-07, "loss": 0.055, "step": 194190 }, { "epoch": 6.565468744717536, "grad_norm": 10.516642570495605, "learning_rate": 2.298292450328421e-07, "loss": 0.0508, "step": 194200 }, { "epoch": 6.565806822407789, "grad_norm": 1.3063243627548218, "learning_rate": 2.2903815496179148e-07, "loss": 0.0549, "step": 194210 }, { "epoch": 6.566144900098043, "grad_norm": 0.727551281452179, "learning_rate": 2.2824842561013914e-07, "loss": 0.049, "step": 194220 }, { "epoch": 6.566482977788295, "grad_norm": 0.8847121596336365, "learning_rate": 2.2746005699948447e-07, "loss": 0.0576, "step": 194230 }, { "epoch": 6.566821055478549, "grad_norm": 2.2546706199645996, "learning_rate": 2.2667304915136577e-07, "loss": 0.0531, "step": 194240 }, { "epoch": 6.567159133168802, "grad_norm": 1.056874394416809, "learning_rate": 2.258874020873103e-07, "loss": 0.0538, "step": 194250 }, { "epoch": 6.567497210859055, "grad_norm": 0.632982075214386, "learning_rate": 2.251031158287953e-07, "loss": 0.0506, "step": 194260 }, { "epoch": 6.5678352885493085, "grad_norm": 1.7704088687896729, "learning_rate": 2.2432019039725915e-07, "loss": 0.0494, "step": 194270 }, { "epoch": 6.568173366239562, "grad_norm": 0.6117491722106934, "learning_rate": 2.23538625814107e-07, "loss": 0.0524, "step": 194280 }, { "epoch": 6.568511443929815, "grad_norm": 0.7038144469261169, "learning_rate": 2.2275842210071064e-07, "loss": 0.0554, "step": 194290 }, { "epoch": 6.568849521620068, "grad_norm": 1.424088478088379, "learning_rate": 2.2197957927840296e-07, "loss": 0.053, "step": 194300 }, { "epoch": 6.569187599310322, "grad_norm": 5.385214805603027, "learning_rate": 2.21202097368467e-07, "loss": 0.0509, "step": 194310 }, { "epoch": 6.569525677000575, "grad_norm": 0.6670940518379211, "learning_rate": 2.2042597639216899e-07, "loss": 0.056, "step": 194320 }, { "epoch": 6.569863754690828, "grad_norm": 0.935141921043396, "learning_rate": 2.1965121637071983e-07, "loss": 0.0546, "step": 194330 }, { "epoch": 6.570201832381081, "grad_norm": 2.897820472717285, "learning_rate": 2.188778173253081e-07, "loss": 0.0579, "step": 194340 }, { "epoch": 6.570539910071334, "grad_norm": 1.0154889822006226, "learning_rate": 2.1810577927706689e-07, "loss": 0.0478, "step": 194350 }, { "epoch": 6.5708779877615875, "grad_norm": 1.0224032402038574, "learning_rate": 2.1733510224711816e-07, "loss": 0.051, "step": 194360 }, { "epoch": 6.571216065451841, "grad_norm": 1.1970677375793457, "learning_rate": 2.1656578625652845e-07, "loss": 0.055, "step": 194370 }, { "epoch": 6.5715541431420945, "grad_norm": 0.6226996779441833, "learning_rate": 2.1579783132631982e-07, "loss": 0.0516, "step": 194380 }, { "epoch": 6.571892220832347, "grad_norm": 1.9371099472045898, "learning_rate": 2.150312374775032e-07, "loss": 0.0519, "step": 194390 }, { "epoch": 6.572230298522601, "grad_norm": 0.7516041398048401, "learning_rate": 2.1426600473102298e-07, "loss": 0.0503, "step": 194400 }, { "epoch": 6.572568376212853, "grad_norm": 1.7887301445007324, "learning_rate": 2.1350213310780687e-07, "loss": 0.0544, "step": 194410 }, { "epoch": 6.572906453903107, "grad_norm": 1.1806838512420654, "learning_rate": 2.127396226287437e-07, "loss": 0.054, "step": 194420 }, { "epoch": 6.57324453159336, "grad_norm": 0.5699807405471802, "learning_rate": 2.1197847331467236e-07, "loss": 0.0566, "step": 194430 }, { "epoch": 6.573582609283614, "grad_norm": 0.7933188080787659, "learning_rate": 2.1121868518640398e-07, "loss": 0.0492, "step": 194440 }, { "epoch": 6.5739206869738664, "grad_norm": 0.5114880800247192, "learning_rate": 2.1046025826471082e-07, "loss": 0.052, "step": 194450 }, { "epoch": 6.57425876466412, "grad_norm": 1.23056161403656, "learning_rate": 2.097031925703319e-07, "loss": 0.0557, "step": 194460 }, { "epoch": 6.574596842354373, "grad_norm": 1.079542636871338, "learning_rate": 2.089474881239617e-07, "loss": 0.0461, "step": 194470 }, { "epoch": 6.574934920044626, "grad_norm": 0.8209384679794312, "learning_rate": 2.0819314494626152e-07, "loss": 0.05, "step": 194480 }, { "epoch": 6.57527299773488, "grad_norm": 0.8305664658546448, "learning_rate": 2.0744016305785375e-07, "loss": 0.0514, "step": 194490 }, { "epoch": 6.575611075425133, "grad_norm": 1.494944453239441, "learning_rate": 2.0668854247932746e-07, "loss": 0.0589, "step": 194500 }, { "epoch": 6.575949153115386, "grad_norm": 1.1425856351852417, "learning_rate": 2.0593828323122734e-07, "loss": 0.0574, "step": 194510 }, { "epoch": 6.576287230805639, "grad_norm": 0.526010274887085, "learning_rate": 2.051893853340703e-07, "loss": 0.052, "step": 194520 }, { "epoch": 6.576625308495892, "grad_norm": 0.8004413843154907, "learning_rate": 2.0444184880832884e-07, "loss": 0.0508, "step": 194530 }, { "epoch": 6.576963386186145, "grad_norm": 0.633499801158905, "learning_rate": 2.0369567367443665e-07, "loss": 0.0572, "step": 194540 }, { "epoch": 6.577301463876399, "grad_norm": 1.386307716369629, "learning_rate": 2.0295085995279406e-07, "loss": 0.0493, "step": 194550 }, { "epoch": 6.577639541566652, "grad_norm": 0.5117813944816589, "learning_rate": 2.0220740766377366e-07, "loss": 0.0496, "step": 194560 }, { "epoch": 6.577977619256905, "grad_norm": 0.5620211362838745, "learning_rate": 2.0146531682769253e-07, "loss": 0.0454, "step": 194570 }, { "epoch": 6.578315696947159, "grad_norm": 1.355325698852539, "learning_rate": 2.0072458746484557e-07, "loss": 0.0509, "step": 194580 }, { "epoch": 6.578653774637411, "grad_norm": 1.0150221586227417, "learning_rate": 1.9998521959547213e-07, "loss": 0.0516, "step": 194590 }, { "epoch": 6.578991852327665, "grad_norm": 1.4497535228729248, "learning_rate": 1.9924721323980044e-07, "loss": 0.0534, "step": 194600 }, { "epoch": 6.579329930017918, "grad_norm": 0.8462178707122803, "learning_rate": 1.9851056841799775e-07, "loss": 0.0558, "step": 194610 }, { "epoch": 6.579668007708172, "grad_norm": 0.5204684138298035, "learning_rate": 1.977752851502035e-07, "loss": 0.0484, "step": 194620 }, { "epoch": 6.580006085398424, "grad_norm": 7.833293437957764, "learning_rate": 1.9704136345652936e-07, "loss": 0.0544, "step": 194630 }, { "epoch": 6.580344163088678, "grad_norm": 0.9709531664848328, "learning_rate": 1.9630880335702594e-07, "loss": 0.0554, "step": 194640 }, { "epoch": 6.580682240778931, "grad_norm": 0.6639246940612793, "learning_rate": 1.9557760487173838e-07, "loss": 0.0492, "step": 194650 }, { "epoch": 6.581020318469184, "grad_norm": 2.12272572517395, "learning_rate": 1.9484776802064508e-07, "loss": 0.0533, "step": 194660 }, { "epoch": 6.5813583961594375, "grad_norm": 0.6806599497795105, "learning_rate": 1.9411929282370234e-07, "loss": 0.0544, "step": 194670 }, { "epoch": 6.581696473849691, "grad_norm": 0.9886118173599243, "learning_rate": 1.9339217930082754e-07, "loss": 0.049, "step": 194680 }, { "epoch": 6.582034551539944, "grad_norm": 1.1934247016906738, "learning_rate": 1.9266642747189923e-07, "loss": 0.0563, "step": 194690 }, { "epoch": 6.582372629230197, "grad_norm": 0.9726387858390808, "learning_rate": 1.9194203735675708e-07, "loss": 0.0501, "step": 194700 }, { "epoch": 6.582710706920451, "grad_norm": 0.5248615741729736, "learning_rate": 1.912190089752075e-07, "loss": 0.0489, "step": 194710 }, { "epoch": 6.583048784610703, "grad_norm": 1.3223683834075928, "learning_rate": 1.90497342347018e-07, "loss": 0.0495, "step": 194720 }, { "epoch": 6.583386862300957, "grad_norm": 1.3738824129104614, "learning_rate": 1.8977703749192277e-07, "loss": 0.0493, "step": 194730 }, { "epoch": 6.58372493999121, "grad_norm": 1.1453152894973755, "learning_rate": 1.8905809442960608e-07, "loss": 0.0473, "step": 194740 }, { "epoch": 6.584063017681463, "grad_norm": 0.5071082711219788, "learning_rate": 1.8834051317973e-07, "loss": 0.0468, "step": 194750 }, { "epoch": 6.5844010953717165, "grad_norm": 2.073516845703125, "learning_rate": 1.876242937619066e-07, "loss": 0.0494, "step": 194760 }, { "epoch": 6.58473917306197, "grad_norm": 0.9397956132888794, "learning_rate": 1.8690943619572577e-07, "loss": 0.0563, "step": 194770 }, { "epoch": 6.585077250752223, "grad_norm": 0.43922218680381775, "learning_rate": 1.8619594050072742e-07, "loss": 0.0493, "step": 194780 }, { "epoch": 6.585415328442476, "grad_norm": 0.9993041157722473, "learning_rate": 1.854838066964182e-07, "loss": 0.0555, "step": 194790 }, { "epoch": 6.58575340613273, "grad_norm": 2.3976235389709473, "learning_rate": 1.8477303480226583e-07, "loss": 0.0559, "step": 194800 }, { "epoch": 6.586091483822982, "grad_norm": 4.873337268829346, "learning_rate": 1.840636248377048e-07, "loss": 0.0507, "step": 194810 }, { "epoch": 6.586429561513236, "grad_norm": 0.8393068909645081, "learning_rate": 1.8335557682213067e-07, "loss": 0.0519, "step": 194820 }, { "epoch": 6.586767639203489, "grad_norm": 0.4927089214324951, "learning_rate": 1.8264889077489465e-07, "loss": 0.0532, "step": 194830 }, { "epoch": 6.587105716893742, "grad_norm": 0.887253999710083, "learning_rate": 1.8194356671532575e-07, "loss": 0.0486, "step": 194840 }, { "epoch": 6.587443794583995, "grad_norm": 0.6028037071228027, "learning_rate": 1.8123960466270295e-07, "loss": 0.0561, "step": 194850 }, { "epoch": 6.587781872274249, "grad_norm": 0.5846290588378906, "learning_rate": 1.80537004636272e-07, "loss": 0.0518, "step": 194860 }, { "epoch": 6.588119949964502, "grad_norm": 0.5833695530891418, "learning_rate": 1.7983576665524525e-07, "loss": 0.0466, "step": 194870 }, { "epoch": 6.588458027654755, "grad_norm": 0.8826213479042053, "learning_rate": 1.791358907387908e-07, "loss": 0.0507, "step": 194880 }, { "epoch": 6.588796105345009, "grad_norm": 0.7537087202072144, "learning_rate": 1.784373769060488e-07, "loss": 0.0582, "step": 194890 }, { "epoch": 6.589134183035261, "grad_norm": 0.923712968826294, "learning_rate": 1.7774022517610956e-07, "loss": 0.053, "step": 194900 }, { "epoch": 6.589472260725515, "grad_norm": 1.1952019929885864, "learning_rate": 1.7704443556803562e-07, "loss": 0.047, "step": 194910 }, { "epoch": 6.589810338415768, "grad_norm": 2.778444528579712, "learning_rate": 1.763500081008451e-07, "loss": 0.0534, "step": 194920 }, { "epoch": 6.590148416106021, "grad_norm": 0.764269232749939, "learning_rate": 1.756569427935284e-07, "loss": 0.0548, "step": 194930 }, { "epoch": 6.590486493796274, "grad_norm": 1.1467509269714355, "learning_rate": 1.7496523966503699e-07, "loss": 0.0578, "step": 194940 }, { "epoch": 6.590824571486528, "grad_norm": 0.8731182217597961, "learning_rate": 1.742748987342724e-07, "loss": 0.054, "step": 194950 }, { "epoch": 6.5911626491767805, "grad_norm": 9.034318923950195, "learning_rate": 1.7358592002011952e-07, "loss": 0.0599, "step": 194960 }, { "epoch": 6.591500726867034, "grad_norm": 0.7258070111274719, "learning_rate": 1.7289830354140224e-07, "loss": 0.0546, "step": 194970 }, { "epoch": 6.5918388045572875, "grad_norm": 1.5387812852859497, "learning_rate": 1.722120493169277e-07, "loss": 0.0507, "step": 194980 }, { "epoch": 6.59217688224754, "grad_norm": 1.2833573818206787, "learning_rate": 1.7152715736545865e-07, "loss": 0.0584, "step": 194990 }, { "epoch": 6.592514959937794, "grad_norm": 0.8171665072441101, "learning_rate": 1.7084362770571904e-07, "loss": 0.0468, "step": 195000 }, { "epoch": 6.592853037628047, "grad_norm": 0.41201263666152954, "learning_rate": 1.7016146035638836e-07, "loss": 0.0499, "step": 195010 }, { "epoch": 6.5931911153183, "grad_norm": 0.6913360953330994, "learning_rate": 1.6948065533612943e-07, "loss": 0.0528, "step": 195020 }, { "epoch": 6.593529193008553, "grad_norm": 0.625942587852478, "learning_rate": 1.6880121266354965e-07, "loss": 0.051, "step": 195030 }, { "epoch": 6.593867270698807, "grad_norm": 1.0466418266296387, "learning_rate": 1.6812313235721744e-07, "loss": 0.0497, "step": 195040 }, { "epoch": 6.5942053483890595, "grad_norm": 1.4841338396072388, "learning_rate": 1.674464144356791e-07, "loss": 0.0497, "step": 195050 }, { "epoch": 6.594543426079313, "grad_norm": 0.5182213187217712, "learning_rate": 1.667710589174365e-07, "loss": 0.0525, "step": 195060 }, { "epoch": 6.5948815037695665, "grad_norm": 0.8832610845565796, "learning_rate": 1.6609706582095262e-07, "loss": 0.0506, "step": 195070 }, { "epoch": 6.595219581459819, "grad_norm": 4.519568920135498, "learning_rate": 1.6542443516465167e-07, "loss": 0.0521, "step": 195080 }, { "epoch": 6.595557659150073, "grad_norm": 0.8940855264663696, "learning_rate": 1.6475316696692444e-07, "loss": 0.0566, "step": 195090 }, { "epoch": 6.595895736840326, "grad_norm": 0.7949044704437256, "learning_rate": 1.640832612461285e-07, "loss": 0.0467, "step": 195100 }, { "epoch": 6.596233814530579, "grad_norm": 1.524204969406128, "learning_rate": 1.634147180205714e-07, "loss": 0.0461, "step": 195110 }, { "epoch": 6.596571892220832, "grad_norm": 0.5521076321601868, "learning_rate": 1.6274753730852742e-07, "loss": 0.0504, "step": 195120 }, { "epoch": 6.596909969911086, "grad_norm": 0.6430985927581787, "learning_rate": 1.6208171912824864e-07, "loss": 0.057, "step": 195130 }, { "epoch": 6.597248047601338, "grad_norm": 0.7231525182723999, "learning_rate": 1.6141726349792606e-07, "loss": 0.043, "step": 195140 }, { "epoch": 6.597586125291592, "grad_norm": 0.8242990970611572, "learning_rate": 1.607541704357396e-07, "loss": 0.0504, "step": 195150 }, { "epoch": 6.5979242029818455, "grad_norm": 1.3787646293640137, "learning_rate": 1.6009243995980806e-07, "loss": 0.0564, "step": 195160 }, { "epoch": 6.598262280672098, "grad_norm": 0.49150246381759644, "learning_rate": 1.5943207208822252e-07, "loss": 0.0527, "step": 195170 }, { "epoch": 6.598600358362352, "grad_norm": 1.0820603370666504, "learning_rate": 1.587730668390408e-07, "loss": 0.05, "step": 195180 }, { "epoch": 6.598938436052605, "grad_norm": 2.4244816303253174, "learning_rate": 1.581154242302818e-07, "loss": 0.052, "step": 195190 }, { "epoch": 6.599276513742858, "grad_norm": 0.8478790521621704, "learning_rate": 1.5745914427992004e-07, "loss": 0.0511, "step": 195200 }, { "epoch": 6.599614591433111, "grad_norm": 1.1392287015914917, "learning_rate": 1.5680422700590224e-07, "loss": 0.0524, "step": 195210 }, { "epoch": 6.599952669123365, "grad_norm": 3.9498744010925293, "learning_rate": 1.561506724261308e-07, "loss": 0.056, "step": 195220 }, { "epoch": 6.600290746813617, "grad_norm": 0.5976879596710205, "learning_rate": 1.554984805584747e-07, "loss": 0.0555, "step": 195230 }, { "epoch": 6.600628824503871, "grad_norm": 0.6121658682823181, "learning_rate": 1.5484765142075863e-07, "loss": 0.0465, "step": 195240 }, { "epoch": 6.600966902194124, "grad_norm": 1.0277711153030396, "learning_rate": 1.5419818503079052e-07, "loss": 0.0545, "step": 195250 }, { "epoch": 6.601304979884377, "grad_norm": 0.648291289806366, "learning_rate": 1.5355008140631178e-07, "loss": 0.0519, "step": 195260 }, { "epoch": 6.6016430575746305, "grad_norm": 0.926682710647583, "learning_rate": 1.5290334056505263e-07, "loss": 0.0568, "step": 195270 }, { "epoch": 6.601981135264884, "grad_norm": 1.2900930643081665, "learning_rate": 1.5225796252468228e-07, "loss": 0.0502, "step": 195280 }, { "epoch": 6.602319212955138, "grad_norm": 0.4723204970359802, "learning_rate": 1.5161394730285882e-07, "loss": 0.0515, "step": 195290 }, { "epoch": 6.60265729064539, "grad_norm": 2.391369581222534, "learning_rate": 1.509712949171793e-07, "loss": 0.0565, "step": 195300 }, { "epoch": 6.602995368335644, "grad_norm": 1.6289616823196411, "learning_rate": 1.5033000538521858e-07, "loss": 0.0558, "step": 195310 }, { "epoch": 6.603333446025896, "grad_norm": 0.9673998355865479, "learning_rate": 1.4969007872451257e-07, "loss": 0.0537, "step": 195320 }, { "epoch": 6.60367152371615, "grad_norm": 0.9945845007896423, "learning_rate": 1.4905151495254177e-07, "loss": 0.0581, "step": 195330 }, { "epoch": 6.604009601406403, "grad_norm": 1.4097309112548828, "learning_rate": 1.484143140867811e-07, "loss": 0.0537, "step": 195340 }, { "epoch": 6.604347679096657, "grad_norm": 0.725770890712738, "learning_rate": 1.4777847614464434e-07, "loss": 0.0545, "step": 195350 }, { "epoch": 6.6046857567869095, "grad_norm": 8.121767044067383, "learning_rate": 1.4714400114351212e-07, "loss": 0.0488, "step": 195360 }, { "epoch": 6.605023834477163, "grad_norm": 0.4448821544647217, "learning_rate": 1.4651088910073718e-07, "loss": 0.0504, "step": 195370 }, { "epoch": 6.605361912167416, "grad_norm": 0.8952879905700684, "learning_rate": 1.458791400336279e-07, "loss": 0.0496, "step": 195380 }, { "epoch": 6.605699989857669, "grad_norm": 0.7174628973007202, "learning_rate": 1.4524875395944826e-07, "loss": 0.0439, "step": 195390 }, { "epoch": 6.606038067547923, "grad_norm": 0.8036051988601685, "learning_rate": 1.4461973089544001e-07, "loss": 0.0533, "step": 195400 }, { "epoch": 6.606376145238176, "grad_norm": 1.0143665075302124, "learning_rate": 1.439920708588005e-07, "loss": 0.0495, "step": 195410 }, { "epoch": 6.606714222928429, "grad_norm": 1.140236735343933, "learning_rate": 1.433657738666827e-07, "loss": 0.0556, "step": 195420 }, { "epoch": 6.607052300618682, "grad_norm": 1.808193564414978, "learning_rate": 1.4274083993621733e-07, "loss": 0.054, "step": 195430 }, { "epoch": 6.607390378308935, "grad_norm": 0.8444710373878479, "learning_rate": 1.4211726908449074e-07, "loss": 0.0593, "step": 195440 }, { "epoch": 6.6077284559991885, "grad_norm": 0.6053309440612793, "learning_rate": 1.4149506132854483e-07, "loss": 0.0548, "step": 195450 }, { "epoch": 6.608066533689442, "grad_norm": 1.1502091884613037, "learning_rate": 1.4087421668538826e-07, "loss": 0.0536, "step": 195460 }, { "epoch": 6.6084046113796955, "grad_norm": 2.8293306827545166, "learning_rate": 1.4025473517200182e-07, "loss": 0.0516, "step": 195470 }, { "epoch": 6.608742689069948, "grad_norm": 0.9593486785888672, "learning_rate": 1.3963661680532203e-07, "loss": 0.0476, "step": 195480 }, { "epoch": 6.609080766760202, "grad_norm": 0.4877096712589264, "learning_rate": 1.3901986160224644e-07, "loss": 0.0479, "step": 195490 }, { "epoch": 6.609418844450454, "grad_norm": 0.4325549006462097, "learning_rate": 1.384044695796338e-07, "loss": 0.0523, "step": 195500 }, { "epoch": 6.609756922140708, "grad_norm": 1.1516954898834229, "learning_rate": 1.3779044075430957e-07, "loss": 0.0614, "step": 195510 }, { "epoch": 6.610094999830961, "grad_norm": 1.0357118844985962, "learning_rate": 1.3717777514306028e-07, "loss": 0.0529, "step": 195520 }, { "epoch": 6.610433077521215, "grad_norm": 0.855965256690979, "learning_rate": 1.3656647276264477e-07, "loss": 0.048, "step": 195530 }, { "epoch": 6.610771155211467, "grad_norm": 0.9753471612930298, "learning_rate": 1.3595653362976635e-07, "loss": 0.0556, "step": 195540 }, { "epoch": 6.611109232901721, "grad_norm": 1.1205023527145386, "learning_rate": 1.3534795776110053e-07, "loss": 0.0574, "step": 195550 }, { "epoch": 6.6114473105919735, "grad_norm": 0.7470988631248474, "learning_rate": 1.347407451732896e-07, "loss": 0.0565, "step": 195560 }, { "epoch": 6.611785388282227, "grad_norm": 0.8578531742095947, "learning_rate": 1.3413489588293692e-07, "loss": 0.0517, "step": 195570 }, { "epoch": 6.612123465972481, "grad_norm": 0.7441496253013611, "learning_rate": 1.3353040990659594e-07, "loss": 0.0561, "step": 195580 }, { "epoch": 6.612461543662734, "grad_norm": 0.8782192468643188, "learning_rate": 1.329272872608034e-07, "loss": 0.0557, "step": 195590 }, { "epoch": 6.612799621352987, "grad_norm": 0.4891115725040436, "learning_rate": 1.3232552796204612e-07, "loss": 0.0498, "step": 195600 }, { "epoch": 6.61313769904324, "grad_norm": 1.0255736112594604, "learning_rate": 1.3172513202677206e-07, "loss": 0.0524, "step": 195610 }, { "epoch": 6.613475776733494, "grad_norm": 1.355777382850647, "learning_rate": 1.3112609947139588e-07, "loss": 0.0518, "step": 195620 }, { "epoch": 6.613813854423746, "grad_norm": 0.7251209020614624, "learning_rate": 1.3052843031230445e-07, "loss": 0.0479, "step": 195630 }, { "epoch": 6.614151932114, "grad_norm": 0.4621625542640686, "learning_rate": 1.299321245658236e-07, "loss": 0.051, "step": 195640 }, { "epoch": 6.614490009804253, "grad_norm": 0.7558961510658264, "learning_rate": 1.2933718224826807e-07, "loss": 0.0559, "step": 195650 }, { "epoch": 6.614828087494506, "grad_norm": 1.401940107345581, "learning_rate": 1.2874360337589707e-07, "loss": 0.0485, "step": 195660 }, { "epoch": 6.6151661651847595, "grad_norm": 0.5877429842948914, "learning_rate": 1.2815138796493654e-07, "loss": 0.0555, "step": 195670 }, { "epoch": 6.615504242875013, "grad_norm": 0.5760998725891113, "learning_rate": 1.2756053603158459e-07, "loss": 0.0548, "step": 195680 }, { "epoch": 6.615842320565266, "grad_norm": 0.8340374231338501, "learning_rate": 1.269710475919894e-07, "loss": 0.0516, "step": 195690 }, { "epoch": 6.616180398255519, "grad_norm": 4.676015853881836, "learning_rate": 1.2638292266227149e-07, "loss": 0.0529, "step": 195700 }, { "epoch": 6.616518475945773, "grad_norm": 1.0291424989700317, "learning_rate": 1.257961612585068e-07, "loss": 0.0502, "step": 195710 }, { "epoch": 6.616856553636025, "grad_norm": 0.5679627656936646, "learning_rate": 1.252107633967381e-07, "loss": 0.0506, "step": 195720 }, { "epoch": 6.617194631326279, "grad_norm": 0.9017906188964844, "learning_rate": 1.2462672909296924e-07, "loss": 0.0546, "step": 195730 }, { "epoch": 6.617532709016532, "grad_norm": 0.9053221344947815, "learning_rate": 1.240440583631708e-07, "loss": 0.0603, "step": 195740 }, { "epoch": 6.617870786706785, "grad_norm": 0.8521177172660828, "learning_rate": 1.2346275122326888e-07, "loss": 0.0486, "step": 195750 }, { "epoch": 6.6182088643970385, "grad_norm": 0.7698767781257629, "learning_rate": 1.228828076891564e-07, "loss": 0.0572, "step": 195760 }, { "epoch": 6.618546942087292, "grad_norm": 1.2777351140975952, "learning_rate": 1.2230422777669283e-07, "loss": 0.0504, "step": 195770 }, { "epoch": 6.618885019777545, "grad_norm": 1.4258413314819336, "learning_rate": 1.217270115016933e-07, "loss": 0.0498, "step": 195780 }, { "epoch": 6.619223097467798, "grad_norm": 0.7961062788963318, "learning_rate": 1.211511588799341e-07, "loss": 0.0495, "step": 195790 }, { "epoch": 6.619561175158052, "grad_norm": 0.8373651504516602, "learning_rate": 1.205766699271693e-07, "loss": 0.0516, "step": 195800 }, { "epoch": 6.619899252848304, "grad_norm": 1.0377525091171265, "learning_rate": 1.2000354465909747e-07, "loss": 0.0568, "step": 195810 }, { "epoch": 6.620237330538558, "grad_norm": 0.7747291326522827, "learning_rate": 1.1943178309138936e-07, "loss": 0.0556, "step": 195820 }, { "epoch": 6.620575408228811, "grad_norm": 0.7985766530036926, "learning_rate": 1.1886138523967693e-07, "loss": 0.0504, "step": 195830 }, { "epoch": 6.620913485919064, "grad_norm": 1.9352049827575684, "learning_rate": 1.1829235111955327e-07, "loss": 0.0511, "step": 195840 }, { "epoch": 6.621251563609317, "grad_norm": 0.9589974880218506, "learning_rate": 1.1772468074657817e-07, "loss": 0.0582, "step": 195850 }, { "epoch": 6.621589641299571, "grad_norm": 0.7325490713119507, "learning_rate": 1.171583741362725e-07, "loss": 0.0484, "step": 195860 }, { "epoch": 6.621927718989824, "grad_norm": 0.9657131433486938, "learning_rate": 1.1659343130411283e-07, "loss": 0.0573, "step": 195870 }, { "epoch": 6.622265796680077, "grad_norm": 0.5492414832115173, "learning_rate": 1.1602985226555341e-07, "loss": 0.0505, "step": 195880 }, { "epoch": 6.622603874370331, "grad_norm": 0.6308180093765259, "learning_rate": 1.1546763703599306e-07, "loss": 0.0513, "step": 195890 }, { "epoch": 6.622941952060583, "grad_norm": 0.4218248426914215, "learning_rate": 1.1490678563080837e-07, "loss": 0.0545, "step": 195900 }, { "epoch": 6.623280029750837, "grad_norm": 0.5204079747200012, "learning_rate": 1.1434729806533151e-07, "loss": 0.0536, "step": 195910 }, { "epoch": 6.62361810744109, "grad_norm": 0.7072861790657043, "learning_rate": 1.1378917435485581e-07, "loss": 0.0484, "step": 195920 }, { "epoch": 6.623956185131343, "grad_norm": 0.6939792633056641, "learning_rate": 1.1323241451464684e-07, "loss": 0.0519, "step": 195930 }, { "epoch": 6.624294262821596, "grad_norm": 0.8092311024665833, "learning_rate": 1.126770185599202e-07, "loss": 0.0535, "step": 195940 }, { "epoch": 6.62463234051185, "grad_norm": 1.6185600757598877, "learning_rate": 1.1212298650586373e-07, "loss": 0.0598, "step": 195950 }, { "epoch": 6.6249704182021025, "grad_norm": 0.9031388163566589, "learning_rate": 1.1157031836762089e-07, "loss": 0.0535, "step": 195960 }, { "epoch": 6.625308495892356, "grad_norm": 0.6154934763908386, "learning_rate": 1.110190141603018e-07, "loss": 0.0493, "step": 195970 }, { "epoch": 6.6256465735826096, "grad_norm": 1.4121545553207397, "learning_rate": 1.1046907389898331e-07, "loss": 0.0553, "step": 195980 }, { "epoch": 6.625984651272862, "grad_norm": 0.8154696822166443, "learning_rate": 1.0992049759869783e-07, "loss": 0.0523, "step": 195990 }, { "epoch": 6.626322728963116, "grad_norm": 2.6894264221191406, "learning_rate": 1.0937328527443891e-07, "loss": 0.0492, "step": 196000 }, { "epoch": 6.626660806653369, "grad_norm": 0.9843581914901733, "learning_rate": 1.0882743694117237e-07, "loss": 0.0561, "step": 196010 }, { "epoch": 6.626998884343622, "grad_norm": 2.2294046878814697, "learning_rate": 1.0828295261382515e-07, "loss": 0.0524, "step": 196020 }, { "epoch": 6.627336962033875, "grad_norm": 2.0452234745025635, "learning_rate": 1.0773983230727425e-07, "loss": 0.0507, "step": 196030 }, { "epoch": 6.627675039724129, "grad_norm": 2.29695725440979, "learning_rate": 1.0719807603637444e-07, "loss": 0.0568, "step": 196040 }, { "epoch": 6.6280131174143815, "grad_norm": 1.038407564163208, "learning_rate": 1.0665768381593611e-07, "loss": 0.0459, "step": 196050 }, { "epoch": 6.628351195104635, "grad_norm": 1.0526098012924194, "learning_rate": 1.0611865566073076e-07, "loss": 0.0526, "step": 196060 }, { "epoch": 6.6286892727948885, "grad_norm": 1.0870612859725952, "learning_rate": 1.0558099158549661e-07, "loss": 0.0539, "step": 196070 }, { "epoch": 6.629027350485141, "grad_norm": 1.329846739768982, "learning_rate": 1.0504469160493857e-07, "loss": 0.0553, "step": 196080 }, { "epoch": 6.629365428175395, "grad_norm": 1.4445217847824097, "learning_rate": 1.0450975573371158e-07, "loss": 0.0465, "step": 196090 }, { "epoch": 6.629703505865648, "grad_norm": 1.1989250183105469, "learning_rate": 1.0397618398643727e-07, "loss": 0.0512, "step": 196100 }, { "epoch": 6.630041583555901, "grad_norm": 1.18532133102417, "learning_rate": 1.0344397637771507e-07, "loss": 0.0468, "step": 196110 }, { "epoch": 6.630379661246154, "grad_norm": 0.5836151838302612, "learning_rate": 1.0291313292208893e-07, "loss": 0.0544, "step": 196120 }, { "epoch": 6.630717738936408, "grad_norm": 0.9192200303077698, "learning_rate": 1.0238365363406943e-07, "loss": 0.0595, "step": 196130 }, { "epoch": 6.63105581662666, "grad_norm": 0.48953568935394287, "learning_rate": 1.0185553852813944e-07, "loss": 0.0498, "step": 196140 }, { "epoch": 6.631393894316914, "grad_norm": 3.8004183769226074, "learning_rate": 1.0132878761873188e-07, "loss": 0.0593, "step": 196150 }, { "epoch": 6.6317319720071675, "grad_norm": 0.9197167754173279, "learning_rate": 1.0080340092024632e-07, "loss": 0.05, "step": 196160 }, { "epoch": 6.63207004969742, "grad_norm": 0.8298338055610657, "learning_rate": 1.0027937844704905e-07, "loss": 0.0567, "step": 196170 }, { "epoch": 6.632408127387674, "grad_norm": 0.6545505523681641, "learning_rate": 9.975672021347304e-08, "loss": 0.047, "step": 196180 }, { "epoch": 6.632746205077927, "grad_norm": 0.661324679851532, "learning_rate": 9.923542623379578e-08, "loss": 0.0502, "step": 196190 }, { "epoch": 6.63308428276818, "grad_norm": 12.701275825500488, "learning_rate": 9.871549652227807e-08, "loss": 0.0528, "step": 196200 }, { "epoch": 6.633422360458433, "grad_norm": 0.9190269112586975, "learning_rate": 9.819693109313077e-08, "loss": 0.0479, "step": 196210 }, { "epoch": 6.633760438148687, "grad_norm": 1.8353936672210693, "learning_rate": 9.767972996053143e-08, "loss": 0.0509, "step": 196220 }, { "epoch": 6.634098515838939, "grad_norm": 0.7314825654029846, "learning_rate": 9.716389313861873e-08, "loss": 0.054, "step": 196230 }, { "epoch": 6.634436593529193, "grad_norm": 1.4683761596679688, "learning_rate": 9.664942064150361e-08, "loss": 0.0535, "step": 196240 }, { "epoch": 6.634774671219446, "grad_norm": 1.4163851737976074, "learning_rate": 9.613631248324151e-08, "loss": 0.0518, "step": 196250 }, { "epoch": 6.6351127489097, "grad_norm": 2.1525864601135254, "learning_rate": 9.562456867786562e-08, "loss": 0.0569, "step": 196260 }, { "epoch": 6.635450826599953, "grad_norm": 0.817582905292511, "learning_rate": 9.511418923936477e-08, "loss": 0.0508, "step": 196270 }, { "epoch": 6.635788904290206, "grad_norm": 1.0617233514785767, "learning_rate": 9.460517418169446e-08, "loss": 0.05, "step": 196280 }, { "epoch": 6.636126981980459, "grad_norm": 0.540246844291687, "learning_rate": 9.409752351877133e-08, "loss": 0.0531, "step": 196290 }, { "epoch": 6.636465059670712, "grad_norm": 0.7045294046401978, "learning_rate": 9.359123726447317e-08, "loss": 0.0486, "step": 196300 }, { "epoch": 6.636803137360966, "grad_norm": 1.210660457611084, "learning_rate": 9.308631543263891e-08, "loss": 0.0518, "step": 196310 }, { "epoch": 6.637141215051219, "grad_norm": 1.3689719438552856, "learning_rate": 9.258275803707973e-08, "loss": 0.0541, "step": 196320 }, { "epoch": 6.637479292741472, "grad_norm": 0.6800330877304077, "learning_rate": 9.208056509155682e-08, "loss": 0.0516, "step": 196330 }, { "epoch": 6.637817370431725, "grad_norm": 0.880108118057251, "learning_rate": 9.157973660979813e-08, "loss": 0.0487, "step": 196340 }, { "epoch": 6.638155448121978, "grad_norm": 2.158991813659668, "learning_rate": 9.108027260550378e-08, "loss": 0.0522, "step": 196350 }, { "epoch": 6.6384935258122315, "grad_norm": 2.1978237628936768, "learning_rate": 9.058217309232397e-08, "loss": 0.0531, "step": 196360 }, { "epoch": 6.638831603502485, "grad_norm": 0.594275951385498, "learning_rate": 9.00854380838756e-08, "loss": 0.0509, "step": 196370 }, { "epoch": 6.6391696811927385, "grad_norm": 3.2032222747802734, "learning_rate": 8.959006759374223e-08, "loss": 0.0549, "step": 196380 }, { "epoch": 6.639507758882991, "grad_norm": 2.8661320209503174, "learning_rate": 8.909606163546858e-08, "loss": 0.0573, "step": 196390 }, { "epoch": 6.639845836573245, "grad_norm": 0.4519179165363312, "learning_rate": 8.860342022255497e-08, "loss": 0.0591, "step": 196400 }, { "epoch": 6.640183914263497, "grad_norm": 1.0539225339889526, "learning_rate": 8.811214336847395e-08, "loss": 0.0484, "step": 196410 }, { "epoch": 6.640521991953751, "grad_norm": 0.4953633248806, "learning_rate": 8.762223108665369e-08, "loss": 0.0558, "step": 196420 }, { "epoch": 6.640860069644004, "grad_norm": 1.0470669269561768, "learning_rate": 8.713368339049454e-08, "loss": 0.0568, "step": 196430 }, { "epoch": 6.641198147334258, "grad_norm": 13.94489574432373, "learning_rate": 8.664650029334143e-08, "loss": 0.0533, "step": 196440 }, { "epoch": 6.6415362250245105, "grad_norm": 2.526510000228882, "learning_rate": 8.61606818085281e-08, "loss": 0.0544, "step": 196450 }, { "epoch": 6.641874302714764, "grad_norm": 0.7105660438537598, "learning_rate": 8.567622794933283e-08, "loss": 0.052, "step": 196460 }, { "epoch": 6.642212380405017, "grad_norm": 0.5023939609527588, "learning_rate": 8.519313872898948e-08, "loss": 0.0493, "step": 196470 }, { "epoch": 6.64255045809527, "grad_norm": 0.5123836994171143, "learning_rate": 8.471141416072081e-08, "loss": 0.0525, "step": 196480 }, { "epoch": 6.642888535785524, "grad_norm": 1.1259756088256836, "learning_rate": 8.423105425768297e-08, "loss": 0.0484, "step": 196490 }, { "epoch": 6.643226613475777, "grad_norm": 1.3602733612060547, "learning_rate": 8.375205903302097e-08, "loss": 0.0601, "step": 196500 }, { "epoch": 6.64356469116603, "grad_norm": 1.6596883535385132, "learning_rate": 8.327442849982436e-08, "loss": 0.0499, "step": 196510 }, { "epoch": 6.643902768856283, "grad_norm": 1.0240212678909302, "learning_rate": 8.279816267114936e-08, "loss": 0.0579, "step": 196520 }, { "epoch": 6.644240846546536, "grad_norm": 1.2738761901855469, "learning_rate": 8.232326156002446e-08, "loss": 0.0525, "step": 196530 }, { "epoch": 6.644578924236789, "grad_norm": 2.0937952995300293, "learning_rate": 8.184972517942257e-08, "loss": 0.051, "step": 196540 }, { "epoch": 6.644917001927043, "grad_norm": 1.2611727714538574, "learning_rate": 8.137755354230003e-08, "loss": 0.0522, "step": 196550 }, { "epoch": 6.6452550796172964, "grad_norm": 0.5495986342430115, "learning_rate": 8.090674666155762e-08, "loss": 0.0513, "step": 196560 }, { "epoch": 6.645593157307549, "grad_norm": 0.8033874034881592, "learning_rate": 8.043730455007393e-08, "loss": 0.0563, "step": 196570 }, { "epoch": 6.645931234997803, "grad_norm": 0.4890439212322235, "learning_rate": 7.996922722067757e-08, "loss": 0.0507, "step": 196580 }, { "epoch": 6.646269312688056, "grad_norm": 1.2831486463546753, "learning_rate": 7.950251468616943e-08, "loss": 0.0555, "step": 196590 }, { "epoch": 6.646607390378309, "grad_norm": 0.5789970755577087, "learning_rate": 7.903716695931152e-08, "loss": 0.0545, "step": 196600 }, { "epoch": 6.646945468068562, "grad_norm": 1.6212971210479736, "learning_rate": 7.857318405282144e-08, "loss": 0.0525, "step": 196610 }, { "epoch": 6.647283545758816, "grad_norm": 0.4799235761165619, "learning_rate": 7.811056597938903e-08, "loss": 0.0494, "step": 196620 }, { "epoch": 6.647621623449068, "grad_norm": 0.7119836807250977, "learning_rate": 7.764931275165421e-08, "loss": 0.049, "step": 196630 }, { "epoch": 6.647959701139322, "grad_norm": 1.645775318145752, "learning_rate": 7.718942438222909e-08, "loss": 0.0517, "step": 196640 }, { "epoch": 6.648297778829575, "grad_norm": 3.284456729888916, "learning_rate": 7.673090088369805e-08, "loss": 0.0544, "step": 196650 }, { "epoch": 6.648635856519828, "grad_norm": 1.5090851783752441, "learning_rate": 7.627374226858441e-08, "loss": 0.0519, "step": 196660 }, { "epoch": 6.6489739342100815, "grad_norm": 0.8548426628112793, "learning_rate": 7.581794854939483e-08, "loss": 0.0519, "step": 196670 }, { "epoch": 6.649312011900335, "grad_norm": 0.885834813117981, "learning_rate": 7.536351973858046e-08, "loss": 0.0535, "step": 196680 }, { "epoch": 6.649650089590588, "grad_norm": 0.39055758714675903, "learning_rate": 7.491045584857581e-08, "loss": 0.0446, "step": 196690 }, { "epoch": 6.649988167280841, "grad_norm": 0.8248982429504395, "learning_rate": 7.44587568917654e-08, "loss": 0.0516, "step": 196700 }, { "epoch": 6.650326244971095, "grad_norm": 1.3296619653701782, "learning_rate": 7.400842288048936e-08, "loss": 0.0525, "step": 196710 }, { "epoch": 6.650664322661347, "grad_norm": 0.8605393767356873, "learning_rate": 7.355945382707119e-08, "loss": 0.0484, "step": 196720 }, { "epoch": 6.651002400351601, "grad_norm": 1.1117656230926514, "learning_rate": 7.311184974377883e-08, "loss": 0.0519, "step": 196730 }, { "epoch": 6.651340478041854, "grad_norm": 1.5103799104690552, "learning_rate": 7.266561064285804e-08, "loss": 0.0485, "step": 196740 }, { "epoch": 6.651678555732107, "grad_norm": 1.3630582094192505, "learning_rate": 7.222073653649908e-08, "loss": 0.0535, "step": 196750 }, { "epoch": 6.6520166334223605, "grad_norm": 0.896356999874115, "learning_rate": 7.177722743686444e-08, "loss": 0.048, "step": 196760 }, { "epoch": 6.652354711112614, "grad_norm": 1.0427796840667725, "learning_rate": 7.133508335608885e-08, "loss": 0.0537, "step": 196770 }, { "epoch": 6.652692788802867, "grad_norm": 0.6489933133125305, "learning_rate": 7.089430430625154e-08, "loss": 0.0483, "step": 196780 }, { "epoch": 6.65303086649312, "grad_norm": 1.0642569065093994, "learning_rate": 7.045489029940955e-08, "loss": 0.0509, "step": 196790 }, { "epoch": 6.653368944183374, "grad_norm": 0.7619485855102539, "learning_rate": 7.001684134756991e-08, "loss": 0.0513, "step": 196800 }, { "epoch": 6.653707021873626, "grad_norm": 7.878578186035156, "learning_rate": 6.958015746271196e-08, "loss": 0.0465, "step": 196810 }, { "epoch": 6.65404509956388, "grad_norm": 0.9228857755661011, "learning_rate": 6.914483865677612e-08, "loss": 0.0553, "step": 196820 }, { "epoch": 6.654383177254133, "grad_norm": 0.9772029519081116, "learning_rate": 6.871088494165845e-08, "loss": 0.0466, "step": 196830 }, { "epoch": 6.654721254944386, "grad_norm": 0.6622556447982788, "learning_rate": 6.827829632923277e-08, "loss": 0.0478, "step": 196840 }, { "epoch": 6.6550593326346394, "grad_norm": 0.43495914340019226, "learning_rate": 6.784707283131741e-08, "loss": 0.0478, "step": 196850 }, { "epoch": 6.655397410324893, "grad_norm": 0.484622985124588, "learning_rate": 6.741721445970295e-08, "loss": 0.0486, "step": 196860 }, { "epoch": 6.655735488015146, "grad_norm": 0.9318879842758179, "learning_rate": 6.698872122614664e-08, "loss": 0.055, "step": 196870 }, { "epoch": 6.656073565705399, "grad_norm": 2.2089180946350098, "learning_rate": 6.656159314235578e-08, "loss": 0.0576, "step": 196880 }, { "epoch": 6.656411643395653, "grad_norm": 0.9166100025177002, "learning_rate": 6.613583022001545e-08, "loss": 0.057, "step": 196890 }, { "epoch": 6.656749721085905, "grad_norm": 1.529449224472046, "learning_rate": 6.571143247076084e-08, "loss": 0.0578, "step": 196900 }, { "epoch": 6.657087798776159, "grad_norm": 2.3882956504821777, "learning_rate": 6.528839990619373e-08, "loss": 0.0541, "step": 196910 }, { "epoch": 6.657425876466412, "grad_norm": 1.188313364982605, "learning_rate": 6.486673253788267e-08, "loss": 0.0491, "step": 196920 }, { "epoch": 6.657763954156665, "grad_norm": 6.829041004180908, "learning_rate": 6.444643037735731e-08, "loss": 0.0535, "step": 196930 }, { "epoch": 6.658102031846918, "grad_norm": 1.1636555194854736, "learning_rate": 6.402749343610848e-08, "loss": 0.0578, "step": 196940 }, { "epoch": 6.658440109537172, "grad_norm": 0.8750718235969543, "learning_rate": 6.36099217255881e-08, "loss": 0.0493, "step": 196950 }, { "epoch": 6.6587781872274245, "grad_norm": 1.0477999448776245, "learning_rate": 6.31937152572093e-08, "loss": 0.0518, "step": 196960 }, { "epoch": 6.659116264917678, "grad_norm": 0.9897826910018921, "learning_rate": 6.277887404235183e-08, "loss": 0.0596, "step": 196970 }, { "epoch": 6.659454342607932, "grad_norm": 0.6464735865592957, "learning_rate": 6.23653980923622e-08, "loss": 0.0508, "step": 196980 }, { "epoch": 6.659792420298184, "grad_norm": 0.7422618865966797, "learning_rate": 6.195328741854245e-08, "loss": 0.056, "step": 196990 }, { "epoch": 6.660130497988438, "grad_norm": 0.564324140548706, "learning_rate": 6.154254203215581e-08, "loss": 0.0481, "step": 197000 }, { "epoch": 6.660468575678691, "grad_norm": 0.9998356699943542, "learning_rate": 6.113316194443775e-08, "loss": 0.0507, "step": 197010 }, { "epoch": 6.660806653368944, "grad_norm": 0.6168570518493652, "learning_rate": 6.072514716657929e-08, "loss": 0.053, "step": 197020 }, { "epoch": 6.661144731059197, "grad_norm": 1.1279284954071045, "learning_rate": 6.031849770973264e-08, "loss": 0.055, "step": 197030 }, { "epoch": 6.661482808749451, "grad_norm": 0.8135918974876404, "learning_rate": 5.991321358501667e-08, "loss": 0.0545, "step": 197040 }, { "epoch": 6.6618208864397035, "grad_norm": 2.6717422008514404, "learning_rate": 5.9509294803511416e-08, "loss": 0.0548, "step": 197050 }, { "epoch": 6.662158964129957, "grad_norm": 1.1645145416259766, "learning_rate": 5.9106741376258045e-08, "loss": 0.0545, "step": 197060 }, { "epoch": 6.6624970418202105, "grad_norm": 2.2105166912078857, "learning_rate": 5.8705553314269965e-08, "loss": 0.0518, "step": 197070 }, { "epoch": 6.662835119510463, "grad_norm": 1.0503933429718018, "learning_rate": 5.8305730628505084e-08, "loss": 0.0532, "step": 197080 }, { "epoch": 6.663173197200717, "grad_norm": 0.8397661447525024, "learning_rate": 5.790727332989909e-08, "loss": 0.0564, "step": 197090 }, { "epoch": 6.66351127489097, "grad_norm": 0.40113502740859985, "learning_rate": 5.751018142934883e-08, "loss": 0.0535, "step": 197100 }, { "epoch": 6.663849352581223, "grad_norm": 1.050046443939209, "learning_rate": 5.711445493770673e-08, "loss": 0.0552, "step": 197110 }, { "epoch": 6.664187430271476, "grad_norm": 0.9184209108352661, "learning_rate": 5.672009386579746e-08, "loss": 0.0513, "step": 197120 }, { "epoch": 6.66452550796173, "grad_norm": 0.8607033491134644, "learning_rate": 5.6327098224390195e-08, "loss": 0.0513, "step": 197130 }, { "epoch": 6.6648635856519824, "grad_norm": 0.8910428881645203, "learning_rate": 5.5935468024242984e-08, "loss": 0.0597, "step": 197140 }, { "epoch": 6.665201663342236, "grad_norm": 1.4415959119796753, "learning_rate": 5.5545203276052836e-08, "loss": 0.0564, "step": 197150 }, { "epoch": 6.6655397410324895, "grad_norm": 2.8898727893829346, "learning_rate": 5.515630399049454e-08, "loss": 0.0497, "step": 197160 }, { "epoch": 6.665877818722742, "grad_norm": 0.8006846904754639, "learning_rate": 5.476877017819848e-08, "loss": 0.0459, "step": 197170 }, { "epoch": 6.666215896412996, "grad_norm": 0.4069165587425232, "learning_rate": 5.4382601849761737e-08, "loss": 0.0544, "step": 197180 }, { "epoch": 6.666553974103249, "grad_norm": 1.0647997856140137, "learning_rate": 5.3997799015742534e-08, "loss": 0.054, "step": 197190 }, { "epoch": 6.666892051793502, "grad_norm": 3.296674966812134, "learning_rate": 5.361436168666023e-08, "loss": 0.0547, "step": 197200 }, { "epoch": 6.667230129483755, "grad_norm": 0.6921969056129456, "learning_rate": 5.3232289872995334e-08, "loss": 0.0531, "step": 197210 }, { "epoch": 6.667568207174009, "grad_norm": 1.122334599494934, "learning_rate": 5.2851583585195044e-08, "loss": 0.0539, "step": 197220 }, { "epoch": 6.667906284864262, "grad_norm": 0.455041766166687, "learning_rate": 5.247224283366769e-08, "loss": 0.0491, "step": 197230 }, { "epoch": 6.668244362554515, "grad_norm": 0.4458043575286865, "learning_rate": 5.209426762878833e-08, "loss": 0.0527, "step": 197240 }, { "epoch": 6.668582440244768, "grad_norm": 1.379408836364746, "learning_rate": 5.171765798088202e-08, "loss": 0.048, "step": 197250 }, { "epoch": 6.668920517935021, "grad_norm": 1.3074088096618652, "learning_rate": 5.134241390025163e-08, "loss": 0.0495, "step": 197260 }, { "epoch": 6.669258595625275, "grad_norm": 1.0587685108184814, "learning_rate": 5.096853539715562e-08, "loss": 0.0584, "step": 197270 }, { "epoch": 6.669596673315528, "grad_norm": 2.2257723808288574, "learning_rate": 5.059602248181916e-08, "loss": 0.0579, "step": 197280 }, { "epoch": 6.669934751005782, "grad_norm": 0.49571025371551514, "learning_rate": 5.022487516441743e-08, "loss": 0.0487, "step": 197290 }, { "epoch": 6.670272828696034, "grad_norm": 0.6988752484321594, "learning_rate": 4.9855093455103416e-08, "loss": 0.05, "step": 197300 }, { "epoch": 6.670610906386288, "grad_norm": 1.583167552947998, "learning_rate": 4.948667736398571e-08, "loss": 0.0557, "step": 197310 }, { "epoch": 6.67094898407654, "grad_norm": 2.2185847759246826, "learning_rate": 4.9119626901139585e-08, "loss": 0.0528, "step": 197320 }, { "epoch": 6.671287061766794, "grad_norm": 0.4552812874317169, "learning_rate": 4.875394207659034e-08, "loss": 0.05, "step": 197330 }, { "epoch": 6.671625139457047, "grad_norm": 0.6185017824172974, "learning_rate": 4.8389622900346654e-08, "loss": 0.0518, "step": 197340 }, { "epoch": 6.671963217147301, "grad_norm": 1.860960602760315, "learning_rate": 4.802666938236722e-08, "loss": 0.052, "step": 197350 }, { "epoch": 6.6723012948375535, "grad_norm": 0.7808802127838135, "learning_rate": 4.766508153256633e-08, "loss": 0.0517, "step": 197360 }, { "epoch": 6.672639372527807, "grad_norm": 1.05451238155365, "learning_rate": 4.730485936084161e-08, "loss": 0.0497, "step": 197370 }, { "epoch": 6.67297745021806, "grad_norm": 1.032463788986206, "learning_rate": 4.694600287703521e-08, "loss": 0.0468, "step": 197380 }, { "epoch": 6.673315527908313, "grad_norm": 0.5335149765014648, "learning_rate": 4.658851209095594e-08, "loss": 0.0524, "step": 197390 }, { "epoch": 6.673653605598567, "grad_norm": 1.1630061864852905, "learning_rate": 4.623238701237931e-08, "loss": 0.0629, "step": 197400 }, { "epoch": 6.67399168328882, "grad_norm": 1.0064914226531982, "learning_rate": 4.587762765104753e-08, "loss": 0.0493, "step": 197410 }, { "epoch": 6.674329760979073, "grad_norm": 0.9799712896347046, "learning_rate": 4.5524234016647294e-08, "loss": 0.0436, "step": 197420 }, { "epoch": 6.674667838669326, "grad_norm": 0.5440787076950073, "learning_rate": 4.5172206118854196e-08, "loss": 0.0523, "step": 197430 }, { "epoch": 6.675005916359579, "grad_norm": 0.8162798285484314, "learning_rate": 4.482154396728277e-08, "loss": 0.0518, "step": 197440 }, { "epoch": 6.6753439940498325, "grad_norm": 0.9186971783638, "learning_rate": 4.447224757152535e-08, "loss": 0.0549, "step": 197450 }, { "epoch": 6.675682071740086, "grad_norm": 1.0318865776062012, "learning_rate": 4.4124316941124285e-08, "loss": 0.0459, "step": 197460 }, { "epoch": 6.6760201494303395, "grad_norm": 0.6842336654663086, "learning_rate": 4.377775208559976e-08, "loss": 0.0525, "step": 197470 }, { "epoch": 6.676358227120592, "grad_norm": 0.8892393112182617, "learning_rate": 4.34325530144275e-08, "loss": 0.0526, "step": 197480 }, { "epoch": 6.676696304810846, "grad_norm": 0.8855428099632263, "learning_rate": 4.308871973703887e-08, "loss": 0.0572, "step": 197490 }, { "epoch": 6.677034382501098, "grad_norm": 0.6876776218414307, "learning_rate": 4.274625226283191e-08, "loss": 0.0564, "step": 197500 }, { "epoch": 6.677372460191352, "grad_norm": 5.826941967010498, "learning_rate": 4.240515060117689e-08, "loss": 0.0508, "step": 197510 }, { "epoch": 6.677710537881605, "grad_norm": 0.8921298384666443, "learning_rate": 4.206541476139969e-08, "loss": 0.05, "step": 197520 }, { "epoch": 6.678048615571859, "grad_norm": 2.194129705429077, "learning_rate": 4.1727044752787325e-08, "loss": 0.0489, "step": 197530 }, { "epoch": 6.678386693262111, "grad_norm": 0.4972230792045593, "learning_rate": 4.1390040584582404e-08, "loss": 0.0618, "step": 197540 }, { "epoch": 6.678724770952365, "grad_norm": 1.8596047163009644, "learning_rate": 4.105440226601087e-08, "loss": 0.0529, "step": 197550 }, { "epoch": 6.6790628486426185, "grad_norm": 0.5827292799949646, "learning_rate": 4.0720129806237625e-08, "loss": 0.0502, "step": 197560 }, { "epoch": 6.679400926332871, "grad_norm": 0.818236768245697, "learning_rate": 4.038722321441091e-08, "loss": 0.0522, "step": 197570 }, { "epoch": 6.679739004023125, "grad_norm": 1.9760204553604126, "learning_rate": 4.005568249962899e-08, "loss": 0.0512, "step": 197580 }, { "epoch": 6.680077081713378, "grad_norm": 0.7246283888816833, "learning_rate": 3.972550767095684e-08, "loss": 0.0509, "step": 197590 }, { "epoch": 6.680415159403631, "grad_norm": 0.5457397699356079, "learning_rate": 3.939669873742058e-08, "loss": 0.0526, "step": 197600 }, { "epoch": 6.680753237093884, "grad_norm": 2.1927602291107178, "learning_rate": 3.906925570800746e-08, "loss": 0.0564, "step": 197610 }, { "epoch": 6.681091314784138, "grad_norm": 0.7143915295600891, "learning_rate": 3.874317859167143e-08, "loss": 0.0526, "step": 197620 }, { "epoch": 6.68142939247439, "grad_norm": 0.4934554398059845, "learning_rate": 3.841846739732202e-08, "loss": 0.0555, "step": 197630 }, { "epoch": 6.681767470164644, "grad_norm": 0.7938443422317505, "learning_rate": 3.809512213384658e-08, "loss": 0.0569, "step": 197640 }, { "epoch": 6.682105547854897, "grad_norm": 1.0545226335525513, "learning_rate": 3.777314281008249e-08, "loss": 0.0473, "step": 197650 }, { "epoch": 6.68244362554515, "grad_norm": 0.9222877025604248, "learning_rate": 3.7452529434828245e-08, "loss": 0.0506, "step": 197660 }, { "epoch": 6.6827817032354035, "grad_norm": 1.1588304042816162, "learning_rate": 3.713328201685462e-08, "loss": 0.0527, "step": 197670 }, { "epoch": 6.683119780925657, "grad_norm": 1.3651058673858643, "learning_rate": 3.681540056488242e-08, "loss": 0.051, "step": 197680 }, { "epoch": 6.68345785861591, "grad_norm": 0.898932158946991, "learning_rate": 3.6498885087610234e-08, "loss": 0.0546, "step": 197690 }, { "epoch": 6.683795936306163, "grad_norm": 1.0099114179611206, "learning_rate": 3.618373559368671e-08, "loss": 0.0514, "step": 197700 }, { "epoch": 6.684134013996417, "grad_norm": 1.0695581436157227, "learning_rate": 3.5869952091732716e-08, "loss": 0.0553, "step": 197710 }, { "epoch": 6.684472091686669, "grad_norm": 0.6659783124923706, "learning_rate": 3.555753459031919e-08, "loss": 0.0579, "step": 197720 }, { "epoch": 6.684810169376923, "grad_norm": 0.587421178817749, "learning_rate": 3.524648309799483e-08, "loss": 0.0514, "step": 197730 }, { "epoch": 6.685148247067176, "grad_norm": 0.8582931756973267, "learning_rate": 3.493679762325841e-08, "loss": 0.0526, "step": 197740 }, { "epoch": 6.685486324757429, "grad_norm": 4.525390625, "learning_rate": 3.4628478174580926e-08, "loss": 0.0543, "step": 197750 }, { "epoch": 6.6858244024476825, "grad_norm": 1.3355202674865723, "learning_rate": 3.432152476038897e-08, "loss": 0.0483, "step": 197760 }, { "epoch": 6.686162480137936, "grad_norm": 0.9418061375617981, "learning_rate": 3.4015937389075824e-08, "loss": 0.0579, "step": 197770 }, { "epoch": 6.686500557828189, "grad_norm": 0.7574859857559204, "learning_rate": 3.371171606899592e-08, "loss": 0.0491, "step": 197780 }, { "epoch": 6.686838635518442, "grad_norm": 0.9312171936035156, "learning_rate": 3.340886080846484e-08, "loss": 0.0489, "step": 197790 }, { "epoch": 6.687176713208696, "grad_norm": 0.8604090213775635, "learning_rate": 3.310737161575927e-08, "loss": 0.0521, "step": 197800 }, { "epoch": 6.687514790898948, "grad_norm": 0.9190970063209534, "learning_rate": 3.280724849913375e-08, "loss": 0.0517, "step": 197810 }, { "epoch": 6.687852868589202, "grad_norm": 0.5388791561126709, "learning_rate": 3.250849146678725e-08, "loss": 0.0516, "step": 197820 }, { "epoch": 6.688190946279455, "grad_norm": 1.8172205686569214, "learning_rate": 3.2211100526879926e-08, "loss": 0.0494, "step": 197830 }, { "epoch": 6.688529023969708, "grad_norm": 0.9385444521903992, "learning_rate": 3.191507568755525e-08, "loss": 0.0516, "step": 197840 }, { "epoch": 6.6888671016599615, "grad_norm": 1.7869553565979004, "learning_rate": 3.162041695689566e-08, "loss": 0.0511, "step": 197850 }, { "epoch": 6.689205179350215, "grad_norm": 1.5383371114730835, "learning_rate": 3.132712434296137e-08, "loss": 0.0479, "step": 197860 }, { "epoch": 6.689543257040468, "grad_norm": 0.6176388263702393, "learning_rate": 3.1035197853773734e-08, "loss": 0.0567, "step": 197870 }, { "epoch": 6.689881334730721, "grad_norm": 0.684188187122345, "learning_rate": 3.0744637497315266e-08, "loss": 0.0502, "step": 197880 }, { "epoch": 6.690219412420975, "grad_norm": 2.8386385440826416, "learning_rate": 3.04554432815185e-08, "loss": 0.0508, "step": 197890 }, { "epoch": 6.690557490111227, "grad_norm": 1.2530725002288818, "learning_rate": 3.016761521429934e-08, "loss": 0.0526, "step": 197900 }, { "epoch": 6.690895567801481, "grad_norm": 0.7238720059394836, "learning_rate": 2.9881153303529254e-08, "loss": 0.0496, "step": 197910 }, { "epoch": 6.691233645491734, "grad_norm": 2.4340386390686035, "learning_rate": 2.9596057557029765e-08, "loss": 0.0491, "step": 197920 }, { "epoch": 6.691571723181987, "grad_norm": 1.6906448602676392, "learning_rate": 2.9312327982605748e-08, "loss": 0.0553, "step": 197930 }, { "epoch": 6.69190980087224, "grad_norm": 1.9433075189590454, "learning_rate": 2.9029964588006554e-08, "loss": 0.0506, "step": 197940 }, { "epoch": 6.692247878562494, "grad_norm": 0.9644364714622498, "learning_rate": 2.8748967380953784e-08, "loss": 0.0525, "step": 197950 }, { "epoch": 6.6925859562527465, "grad_norm": 0.7110255360603333, "learning_rate": 2.8469336369130185e-08, "loss": 0.0538, "step": 197960 }, { "epoch": 6.692924033943, "grad_norm": 1.432680606842041, "learning_rate": 2.819107156018519e-08, "loss": 0.0525, "step": 197970 }, { "epoch": 6.693262111633254, "grad_norm": 1.3417044878005981, "learning_rate": 2.791417296172383e-08, "loss": 0.0511, "step": 197980 }, { "epoch": 6.693600189323506, "grad_norm": 0.4856000244617462, "learning_rate": 2.763864058131227e-08, "loss": 0.0535, "step": 197990 }, { "epoch": 6.69393826701376, "grad_norm": 1.471973180770874, "learning_rate": 2.7364474426483377e-08, "loss": 0.0524, "step": 198000 }, { "epoch": 6.694276344704013, "grad_norm": 0.8933945894241333, "learning_rate": 2.709167450474226e-08, "loss": 0.0556, "step": 198010 }, { "epoch": 6.694614422394266, "grad_norm": 1.1933649778366089, "learning_rate": 2.682024082353851e-08, "loss": 0.0494, "step": 198020 }, { "epoch": 6.694952500084519, "grad_norm": 0.7308223247528076, "learning_rate": 2.6550173390293974e-08, "loss": 0.05, "step": 198030 }, { "epoch": 6.695290577774773, "grad_norm": 1.4732184410095215, "learning_rate": 2.6281472212397185e-08, "loss": 0.0536, "step": 198040 }, { "epoch": 6.6956286554650255, "grad_norm": 4.214659690856934, "learning_rate": 2.6014137297186712e-08, "loss": 0.0485, "step": 198050 }, { "epoch": 6.695966733155279, "grad_norm": 2.8590340614318848, "learning_rate": 2.574816865197338e-08, "loss": 0.05, "step": 198060 }, { "epoch": 6.6963048108455325, "grad_norm": 0.9345232844352722, "learning_rate": 2.54835662840347e-08, "loss": 0.0501, "step": 198070 }, { "epoch": 6.696642888535785, "grad_norm": 1.6535524129867554, "learning_rate": 2.522033020060377e-08, "loss": 0.0501, "step": 198080 }, { "epoch": 6.696980966226039, "grad_norm": 0.676356852054596, "learning_rate": 2.4958460408869287e-08, "loss": 0.0573, "step": 198090 }, { "epoch": 6.697319043916292, "grad_norm": 0.5184271335601807, "learning_rate": 2.4697956915997744e-08, "loss": 0.0524, "step": 198100 }, { "epoch": 6.697657121606545, "grad_norm": 1.3594945669174194, "learning_rate": 2.4438819729111217e-08, "loss": 0.0516, "step": 198110 }, { "epoch": 6.697995199296798, "grad_norm": 1.103162407875061, "learning_rate": 2.418104885528738e-08, "loss": 0.0531, "step": 198120 }, { "epoch": 6.698333276987052, "grad_norm": 1.1166338920593262, "learning_rate": 2.39246443015817e-08, "loss": 0.0556, "step": 198130 }, { "epoch": 6.6986713546773045, "grad_norm": 0.9334307909011841, "learning_rate": 2.3669606074999685e-08, "loss": 0.0531, "step": 198140 }, { "epoch": 6.699009432367558, "grad_norm": 1.1158462762832642, "learning_rate": 2.3415934182513532e-08, "loss": 0.0566, "step": 198150 }, { "epoch": 6.6993475100578115, "grad_norm": 0.9621734619140625, "learning_rate": 2.3163628631067692e-08, "loss": 0.0505, "step": 198160 }, { "epoch": 6.699685587748064, "grad_norm": 0.7116493582725525, "learning_rate": 2.2912689427545543e-08, "loss": 0.0524, "step": 198170 }, { "epoch": 6.700023665438318, "grad_norm": 1.900698184967041, "learning_rate": 2.2663116578819365e-08, "loss": 0.0498, "step": 198180 }, { "epoch": 6.700361743128571, "grad_norm": 0.6798558235168457, "learning_rate": 2.241491009170593e-08, "loss": 0.0478, "step": 198190 }, { "epoch": 6.700699820818825, "grad_norm": 0.5926616191864014, "learning_rate": 2.2168069972994253e-08, "loss": 0.0486, "step": 198200 }, { "epoch": 6.701037898509077, "grad_norm": 0.4796461760997772, "learning_rate": 2.1922596229428936e-08, "loss": 0.0488, "step": 198210 }, { "epoch": 6.701375976199331, "grad_norm": 0.9298915863037109, "learning_rate": 2.1678488867726833e-08, "loss": 0.0519, "step": 198220 }, { "epoch": 6.701714053889583, "grad_norm": 0.8086984753608704, "learning_rate": 2.1435747894560376e-08, "loss": 0.0511, "step": 198230 }, { "epoch": 6.702052131579837, "grad_norm": 1.4883204698562622, "learning_rate": 2.1194373316557604e-08, "loss": 0.0546, "step": 198240 }, { "epoch": 6.70239020927009, "grad_norm": 0.5593591928482056, "learning_rate": 2.095436514032989e-08, "loss": 0.0531, "step": 198250 }, { "epoch": 6.702728286960344, "grad_norm": 1.3603371381759644, "learning_rate": 2.0715723372433104e-08, "loss": 0.0475, "step": 198260 }, { "epoch": 6.703066364650597, "grad_norm": 0.6333616971969604, "learning_rate": 2.0478448019389805e-08, "loss": 0.053, "step": 198270 }, { "epoch": 6.70340444234085, "grad_norm": 0.7259138822555542, "learning_rate": 2.02425390876948e-08, "loss": 0.0468, "step": 198280 }, { "epoch": 6.703742520031103, "grad_norm": 0.8087891340255737, "learning_rate": 2.000799658378738e-08, "loss": 0.0593, "step": 198290 }, { "epoch": 6.704080597721356, "grad_norm": 0.6972510814666748, "learning_rate": 1.9774820514084637e-08, "loss": 0.0533, "step": 198300 }, { "epoch": 6.70441867541161, "grad_norm": 1.4237452745437622, "learning_rate": 1.9543010884964795e-08, "loss": 0.0509, "step": 198310 }, { "epoch": 6.704756753101863, "grad_norm": 1.1139686107635498, "learning_rate": 1.9312567702761687e-08, "loss": 0.0534, "step": 198320 }, { "epoch": 6.705094830792116, "grad_norm": 1.3966816663742065, "learning_rate": 1.908349097377582e-08, "loss": 0.0541, "step": 198330 }, { "epoch": 6.705432908482369, "grad_norm": 0.9905025362968445, "learning_rate": 1.8855780704268856e-08, "loss": 0.0559, "step": 198340 }, { "epoch": 6.705770986172622, "grad_norm": 1.1349105834960938, "learning_rate": 1.8629436900469143e-08, "loss": 0.0514, "step": 198350 }, { "epoch": 6.7061090638628755, "grad_norm": 2.150001049041748, "learning_rate": 1.8404459568560627e-08, "loss": 0.0572, "step": 198360 }, { "epoch": 6.706447141553129, "grad_norm": 1.577765941619873, "learning_rate": 1.8180848714699495e-08, "loss": 0.0523, "step": 198370 }, { "epoch": 6.706785219243383, "grad_norm": 0.9331658482551575, "learning_rate": 1.795860434499752e-08, "loss": 0.0543, "step": 198380 }, { "epoch": 6.707123296933635, "grad_norm": 2.3019940853118896, "learning_rate": 1.7737726465527625e-08, "loss": 0.0445, "step": 198390 }, { "epoch": 6.707461374623889, "grad_norm": 0.6073833107948303, "learning_rate": 1.7518215082329426e-08, "loss": 0.0481, "step": 198400 }, { "epoch": 6.707799452314141, "grad_norm": 0.773814857006073, "learning_rate": 1.7300070201409223e-08, "loss": 0.0539, "step": 198410 }, { "epoch": 6.708137530004395, "grad_norm": 0.9807007312774658, "learning_rate": 1.708329182872892e-08, "loss": 0.0522, "step": 198420 }, { "epoch": 6.708475607694648, "grad_norm": 2.0844991207122803, "learning_rate": 1.6867879970206e-08, "loss": 0.0496, "step": 198430 }, { "epoch": 6.708813685384902, "grad_norm": 0.8981025218963623, "learning_rate": 1.6653834631746856e-08, "loss": 0.0533, "step": 198440 }, { "epoch": 6.7091517630751545, "grad_norm": 0.9776649475097656, "learning_rate": 1.6441155819185705e-08, "loss": 0.0589, "step": 198450 }, { "epoch": 6.709489840765408, "grad_norm": 0.9845037460327148, "learning_rate": 1.622984353835122e-08, "loss": 0.0613, "step": 198460 }, { "epoch": 6.709827918455661, "grad_norm": 0.5152836441993713, "learning_rate": 1.6019897795011007e-08, "loss": 0.0524, "step": 198470 }, { "epoch": 6.710165996145914, "grad_norm": 0.9470195770263672, "learning_rate": 1.5811318594910474e-08, "loss": 0.0462, "step": 198480 }, { "epoch": 6.710504073836168, "grad_norm": 0.7608082890510559, "learning_rate": 1.5604105943750613e-08, "loss": 0.059, "step": 198490 }, { "epoch": 6.710842151526421, "grad_norm": 1.056369423866272, "learning_rate": 1.5398259847199114e-08, "loss": 0.0489, "step": 198500 }, { "epoch": 6.711180229216674, "grad_norm": 0.6445088386535645, "learning_rate": 1.5193780310879256e-08, "loss": 0.0548, "step": 198510 }, { "epoch": 6.711518306906927, "grad_norm": 1.192801833152771, "learning_rate": 1.4990667340381015e-08, "loss": 0.0564, "step": 198520 }, { "epoch": 6.711856384597181, "grad_norm": 0.4219675064086914, "learning_rate": 1.4788920941261053e-08, "loss": 0.0535, "step": 198530 }, { "epoch": 6.712194462287433, "grad_norm": 0.5254861116409302, "learning_rate": 1.458854111903163e-08, "loss": 0.0623, "step": 198540 }, { "epoch": 6.712532539977687, "grad_norm": 0.9982551336288452, "learning_rate": 1.4389527879171693e-08, "loss": 0.0533, "step": 198550 }, { "epoch": 6.7128706176679405, "grad_norm": 1.1969408988952637, "learning_rate": 1.4191881227126891e-08, "loss": 0.0586, "step": 198560 }, { "epoch": 6.713208695358193, "grad_norm": 0.7341405749320984, "learning_rate": 1.3995601168292905e-08, "loss": 0.0511, "step": 198570 }, { "epoch": 6.713546773048447, "grad_norm": 0.9522653222084045, "learning_rate": 1.3800687708043213e-08, "loss": 0.0531, "step": 198580 }, { "epoch": 6.7138848507387, "grad_norm": 1.5054198503494263, "learning_rate": 1.360714085170689e-08, "loss": 0.0485, "step": 198590 }, { "epoch": 6.714222928428953, "grad_norm": 0.5617581605911255, "learning_rate": 1.3414960604568593e-08, "loss": 0.0525, "step": 198600 }, { "epoch": 6.714561006119206, "grad_norm": 0.5487586855888367, "learning_rate": 1.3224146971885231e-08, "loss": 0.0538, "step": 198610 }, { "epoch": 6.71489908380946, "grad_norm": 0.610407292842865, "learning_rate": 1.3034699958874852e-08, "loss": 0.0516, "step": 198620 }, { "epoch": 6.715237161499712, "grad_norm": 0.7205397486686707, "learning_rate": 1.2846619570716645e-08, "loss": 0.0488, "step": 198630 }, { "epoch": 6.715575239189966, "grad_norm": 1.0355844497680664, "learning_rate": 1.2659905812556494e-08, "loss": 0.052, "step": 198640 }, { "epoch": 6.715913316880219, "grad_norm": 0.5324693918228149, "learning_rate": 1.2474558689490324e-08, "loss": 0.052, "step": 198650 }, { "epoch": 6.716251394570472, "grad_norm": 2.481492757797241, "learning_rate": 1.2290578206591852e-08, "loss": 0.0488, "step": 198660 }, { "epoch": 6.716589472260726, "grad_norm": 1.1936508417129517, "learning_rate": 1.210796436889039e-08, "loss": 0.0533, "step": 198670 }, { "epoch": 6.716927549950979, "grad_norm": 1.230623483657837, "learning_rate": 1.192671718137639e-08, "loss": 0.0484, "step": 198680 }, { "epoch": 6.717265627641232, "grad_norm": 4.194586277008057, "learning_rate": 1.1746836649006999e-08, "loss": 0.0534, "step": 198690 }, { "epoch": 6.717603705331485, "grad_norm": 0.3478817939758301, "learning_rate": 1.1568322776700503e-08, "loss": 0.0543, "step": 198700 }, { "epoch": 6.717941783021739, "grad_norm": 0.6669262647628784, "learning_rate": 1.1391175569336333e-08, "loss": 0.053, "step": 198710 }, { "epoch": 6.718279860711991, "grad_norm": 0.9087468981742859, "learning_rate": 1.121539503175506e-08, "loss": 0.051, "step": 198720 }, { "epoch": 6.718617938402245, "grad_norm": 0.6008139848709106, "learning_rate": 1.1040981168763953e-08, "loss": 0.0503, "step": 198730 }, { "epoch": 6.718956016092498, "grad_norm": 0.6223452091217041, "learning_rate": 1.086793398513697e-08, "loss": 0.0463, "step": 198740 }, { "epoch": 6.719294093782751, "grad_norm": 0.7199804186820984, "learning_rate": 1.0696253485598107e-08, "loss": 0.0565, "step": 198750 }, { "epoch": 6.7196321714730045, "grad_norm": 0.8033273816108704, "learning_rate": 1.052593967484361e-08, "loss": 0.0561, "step": 198760 }, { "epoch": 6.719970249163258, "grad_norm": 1.0586010217666626, "learning_rate": 1.0356992557530864e-08, "loss": 0.0495, "step": 198770 }, { "epoch": 6.720308326853511, "grad_norm": 1.3123232126235962, "learning_rate": 1.0189412138272846e-08, "loss": 0.0542, "step": 198780 }, { "epoch": 6.720646404543764, "grad_norm": 0.7575350403785706, "learning_rate": 1.0023198421660328e-08, "loss": 0.0526, "step": 198790 }, { "epoch": 6.720984482234018, "grad_norm": 0.9250807762145996, "learning_rate": 9.858351412234123e-09, "loss": 0.0551, "step": 198800 }, { "epoch": 6.72132255992427, "grad_norm": 1.5937095880508423, "learning_rate": 9.694871114501736e-09, "loss": 0.0496, "step": 198810 }, { "epoch": 6.721660637614524, "grad_norm": 0.7836332321166992, "learning_rate": 9.532757532926262e-09, "loss": 0.0566, "step": 198820 }, { "epoch": 6.721998715304777, "grad_norm": 0.7600414156913757, "learning_rate": 9.372010671948594e-09, "loss": 0.0529, "step": 198830 }, { "epoch": 6.72233679299503, "grad_norm": 0.7275059223175049, "learning_rate": 9.212630535954114e-09, "loss": 0.0553, "step": 198840 }, { "epoch": 6.7226748706852835, "grad_norm": 0.5815807580947876, "learning_rate": 9.05461712931155e-09, "loss": 0.0522, "step": 198850 }, { "epoch": 6.723012948375537, "grad_norm": 0.937476634979248, "learning_rate": 8.897970456328563e-09, "loss": 0.057, "step": 198860 }, { "epoch": 6.72335102606579, "grad_norm": 0.9891679286956787, "learning_rate": 8.742690521301722e-09, "loss": 0.0567, "step": 198870 }, { "epoch": 6.723689103756043, "grad_norm": 0.6859243512153625, "learning_rate": 8.588777328466524e-09, "loss": 0.0515, "step": 198880 }, { "epoch": 6.724027181446297, "grad_norm": 0.5184829235076904, "learning_rate": 8.436230882030715e-09, "loss": 0.0525, "step": 198890 }, { "epoch": 6.724365259136549, "grad_norm": 0.7449501156806946, "learning_rate": 8.285051186168735e-09, "loss": 0.0504, "step": 198900 }, { "epoch": 6.724703336826803, "grad_norm": 0.7121244072914124, "learning_rate": 8.135238245010612e-09, "loss": 0.0514, "step": 198910 }, { "epoch": 6.725041414517056, "grad_norm": 0.9380607604980469, "learning_rate": 7.986792062658621e-09, "loss": 0.0562, "step": 198920 }, { "epoch": 6.725379492207309, "grad_norm": 1.2390458583831787, "learning_rate": 7.839712643159525e-09, "loss": 0.0495, "step": 198930 }, { "epoch": 6.725717569897562, "grad_norm": 1.3023239374160767, "learning_rate": 7.693999990548983e-09, "loss": 0.0496, "step": 198940 }, { "epoch": 6.726055647587816, "grad_norm": 0.49979764223098755, "learning_rate": 7.549654108801596e-09, "loss": 0.0536, "step": 198950 }, { "epoch": 6.726393725278069, "grad_norm": 1.0950372219085693, "learning_rate": 7.406675001864205e-09, "loss": 0.0544, "step": 198960 }, { "epoch": 6.726731802968322, "grad_norm": 1.87173330783844, "learning_rate": 7.265062673644796e-09, "loss": 0.0528, "step": 198970 }, { "epoch": 6.727069880658576, "grad_norm": 1.0617661476135254, "learning_rate": 7.124817128018046e-09, "loss": 0.0494, "step": 198980 }, { "epoch": 6.727407958348828, "grad_norm": 0.871381938457489, "learning_rate": 6.985938368819777e-09, "loss": 0.055, "step": 198990 }, { "epoch": 6.727746036039082, "grad_norm": 0.5120272040367126, "learning_rate": 6.848426399841401e-09, "loss": 0.0502, "step": 199000 }, { "epoch": 6.728084113729335, "grad_norm": 0.8005236983299255, "learning_rate": 6.712281224846573e-09, "loss": 0.0566, "step": 199010 }, { "epoch": 6.728422191419588, "grad_norm": 1.448419451713562, "learning_rate": 6.577502847560091e-09, "loss": 0.0531, "step": 199020 }, { "epoch": 6.728760269109841, "grad_norm": 2.8867621421813965, "learning_rate": 6.4440912716567934e-09, "loss": 0.0537, "step": 199030 }, { "epoch": 6.729098346800095, "grad_norm": 1.0435147285461426, "learning_rate": 6.312046500794866e-09, "loss": 0.0563, "step": 199040 }, { "epoch": 6.7294364244903475, "grad_norm": 4.035772323608398, "learning_rate": 6.1813685385769816e-09, "loss": 0.0567, "step": 199050 }, { "epoch": 6.729774502180601, "grad_norm": 0.5347825884819031, "learning_rate": 6.05205738858361e-09, "loss": 0.0496, "step": 199060 }, { "epoch": 6.7301125798708545, "grad_norm": 2.43942928314209, "learning_rate": 5.924113054339708e-09, "loss": 0.0546, "step": 199070 }, { "epoch": 6.730450657561107, "grad_norm": 0.9521628618240356, "learning_rate": 5.797535539348031e-09, "loss": 0.0538, "step": 199080 }, { "epoch": 6.730788735251361, "grad_norm": 0.6956602931022644, "learning_rate": 5.672324847072474e-09, "loss": 0.0563, "step": 199090 }, { "epoch": 6.731126812941614, "grad_norm": 1.968308448791504, "learning_rate": 5.548480980938076e-09, "loss": 0.051, "step": 199100 }, { "epoch": 6.731464890631867, "grad_norm": 0.7636434435844421, "learning_rate": 5.4260039443199125e-09, "loss": 0.051, "step": 199110 }, { "epoch": 6.73180296832212, "grad_norm": 0.71909499168396, "learning_rate": 5.30489374057086e-09, "loss": 0.0512, "step": 199120 }, { "epoch": 6.732141046012374, "grad_norm": 0.7549579739570618, "learning_rate": 5.1851503730104835e-09, "loss": 0.0565, "step": 199130 }, { "epoch": 6.7324791237026265, "grad_norm": 1.4200077056884766, "learning_rate": 5.066773844902839e-09, "loss": 0.0548, "step": 199140 }, { "epoch": 6.73281720139288, "grad_norm": 0.518115758895874, "learning_rate": 4.949764159484227e-09, "loss": 0.0558, "step": 199150 }, { "epoch": 6.7331552790831335, "grad_norm": 0.9385985136032104, "learning_rate": 4.834121319963192e-09, "loss": 0.0555, "step": 199160 }, { "epoch": 6.733493356773387, "grad_norm": 3.9051358699798584, "learning_rate": 4.719845329492767e-09, "loss": 0.0517, "step": 199170 }, { "epoch": 6.73383143446364, "grad_norm": 0.6534960865974426, "learning_rate": 4.60693619119823e-09, "loss": 0.0512, "step": 199180 }, { "epoch": 6.734169512153893, "grad_norm": 0.7201407551765442, "learning_rate": 4.495393908166001e-09, "loss": 0.0504, "step": 199190 }, { "epoch": 6.734507589844146, "grad_norm": 0.6396350860595703, "learning_rate": 4.385218483449194e-09, "loss": 0.0512, "step": 199200 }, { "epoch": 6.734845667534399, "grad_norm": 0.803153932094574, "learning_rate": 4.276409920056512e-09, "loss": 0.0568, "step": 199210 }, { "epoch": 6.735183745224653, "grad_norm": 0.6833146810531616, "learning_rate": 4.168968220963354e-09, "loss": 0.0517, "step": 199220 }, { "epoch": 6.735521822914906, "grad_norm": 1.553361177444458, "learning_rate": 4.0628933891118105e-09, "loss": 0.0561, "step": 199230 }, { "epoch": 6.735859900605159, "grad_norm": 0.8569018244743347, "learning_rate": 3.958185427399563e-09, "loss": 0.0523, "step": 199240 }, { "epoch": 6.7361979782954124, "grad_norm": 1.1684520244598389, "learning_rate": 3.854844338685437e-09, "loss": 0.0547, "step": 199250 }, { "epoch": 6.736536055985665, "grad_norm": 0.782025933265686, "learning_rate": 3.75287012579495e-09, "loss": 0.0521, "step": 199260 }, { "epoch": 6.736874133675919, "grad_norm": 2.7066707611083984, "learning_rate": 3.652262791520311e-09, "loss": 0.0515, "step": 199270 }, { "epoch": 6.737212211366172, "grad_norm": 0.44072166085243225, "learning_rate": 3.5530223386093244e-09, "loss": 0.0481, "step": 199280 }, { "epoch": 6.737550289056426, "grad_norm": 1.3240886926651, "learning_rate": 3.455148769776484e-09, "loss": 0.0494, "step": 199290 }, { "epoch": 6.737888366746678, "grad_norm": 0.6299409866333008, "learning_rate": 3.3586420876974277e-09, "loss": 0.0544, "step": 199300 }, { "epoch": 6.738226444436932, "grad_norm": 1.115028977394104, "learning_rate": 3.2635022950089358e-09, "loss": 0.0508, "step": 199310 }, { "epoch": 6.738564522127184, "grad_norm": 2.3014280796051025, "learning_rate": 3.16972939431448e-09, "loss": 0.0523, "step": 199320 }, { "epoch": 6.738902599817438, "grad_norm": 0.74518221616745, "learning_rate": 3.0773233881731256e-09, "loss": 0.0522, "step": 199330 }, { "epoch": 6.739240677507691, "grad_norm": 0.6121488213539124, "learning_rate": 2.9862842791217317e-09, "loss": 0.0539, "step": 199340 }, { "epoch": 6.739578755197945, "grad_norm": 0.9112241864204407, "learning_rate": 2.8966120696360954e-09, "loss": 0.0488, "step": 199350 }, { "epoch": 6.7399168328881975, "grad_norm": 0.42784038186073303, "learning_rate": 2.808306762175361e-09, "loss": 0.0498, "step": 199360 }, { "epoch": 6.740254910578451, "grad_norm": 2.127473831176758, "learning_rate": 2.721368359154264e-09, "loss": 0.0553, "step": 199370 }, { "epoch": 6.740592988268704, "grad_norm": 0.668156087398529, "learning_rate": 2.6357968629486805e-09, "loss": 0.0516, "step": 199380 }, { "epoch": 6.740931065958957, "grad_norm": 0.7972316145896912, "learning_rate": 2.5515922758956312e-09, "loss": 0.0493, "step": 199390 }, { "epoch": 6.741269143649211, "grad_norm": 0.6090880036354065, "learning_rate": 2.468754600298828e-09, "loss": 0.0574, "step": 199400 }, { "epoch": 6.741607221339464, "grad_norm": 0.9971704483032227, "learning_rate": 2.3872838384231268e-09, "loss": 0.0552, "step": 199410 }, { "epoch": 6.741945299029717, "grad_norm": 1.0591068267822266, "learning_rate": 2.3071799924945237e-09, "loss": 0.0529, "step": 199420 }, { "epoch": 6.74228337671997, "grad_norm": 1.0220553874969482, "learning_rate": 2.22844306470571e-09, "loss": 0.0523, "step": 199430 }, { "epoch": 6.742621454410223, "grad_norm": 0.815981388092041, "learning_rate": 2.1510730572105175e-09, "loss": 0.0558, "step": 199440 }, { "epoch": 6.7429595321004765, "grad_norm": 0.881450355052948, "learning_rate": 2.0750699721183707e-09, "loss": 0.0535, "step": 199450 }, { "epoch": 6.74329760979073, "grad_norm": 1.2829071283340454, "learning_rate": 2.0004338115109377e-09, "loss": 0.0509, "step": 199460 }, { "epoch": 6.7436356874809835, "grad_norm": 1.0210210084915161, "learning_rate": 1.927164577431029e-09, "loss": 0.0551, "step": 199470 }, { "epoch": 6.743973765171236, "grad_norm": 0.5048279762268066, "learning_rate": 1.8552622718770451e-09, "loss": 0.0496, "step": 199480 }, { "epoch": 6.74431184286149, "grad_norm": 0.9806437492370605, "learning_rate": 1.7847268968140818e-09, "loss": 0.0592, "step": 199490 }, { "epoch": 6.744649920551743, "grad_norm": 0.468797504901886, "learning_rate": 1.7155584541739266e-09, "loss": 0.0543, "step": 199500 }, { "epoch": 6.744987998241996, "grad_norm": 1.681342601776123, "learning_rate": 1.6477569458495102e-09, "loss": 0.0506, "step": 199510 }, { "epoch": 6.745326075932249, "grad_norm": 0.500458300113678, "learning_rate": 1.5813223736949046e-09, "loss": 0.0497, "step": 199520 }, { "epoch": 6.745664153622503, "grad_norm": 0.8891076445579529, "learning_rate": 1.5162547395142223e-09, "loss": 0.05, "step": 199530 }, { "epoch": 6.7460022313127554, "grad_norm": 1.079127311706543, "learning_rate": 1.452554045106025e-09, "loss": 0.0558, "step": 199540 }, { "epoch": 6.746340309003009, "grad_norm": 1.1412527561187744, "learning_rate": 1.3902202921911578e-09, "loss": 0.0488, "step": 199550 }, { "epoch": 6.7466783866932625, "grad_norm": 0.5512163043022156, "learning_rate": 1.3292534824904668e-09, "loss": 0.0476, "step": 199560 }, { "epoch": 6.747016464383515, "grad_norm": 0.6289430856704712, "learning_rate": 1.2696536176637353e-09, "loss": 0.0499, "step": 199570 }, { "epoch": 6.747354542073769, "grad_norm": 2.0752594470977783, "learning_rate": 1.211420699342991e-09, "loss": 0.0483, "step": 199580 }, { "epoch": 6.747692619764022, "grad_norm": 0.7926977872848511, "learning_rate": 1.1545547291158532e-09, "loss": 0.0613, "step": 199590 }, { "epoch": 6.748030697454275, "grad_norm": 0.7741811871528625, "learning_rate": 1.0990557085366338e-09, "loss": 0.0504, "step": 199600 }, { "epoch": 6.748368775144528, "grad_norm": 1.308089017868042, "learning_rate": 1.0449236391263385e-09, "loss": 0.0483, "step": 199610 }, { "epoch": 6.748706852834782, "grad_norm": 2.6051599979400635, "learning_rate": 9.921585223671149e-10, "loss": 0.0537, "step": 199620 }, { "epoch": 6.749044930525034, "grad_norm": 0.8222975134849548, "learning_rate": 9.40760359696702e-10, "loss": 0.0549, "step": 199630 }, { "epoch": 6.749383008215288, "grad_norm": 0.6673756837844849, "learning_rate": 8.907291525250827e-10, "loss": 0.0474, "step": 199640 }, { "epoch": 6.749721085905541, "grad_norm": 0.9188992977142334, "learning_rate": 8.420649022122806e-10, "loss": 0.0527, "step": 199650 }, { "epoch": 6.750059163595794, "grad_norm": 0.45536962151527405, "learning_rate": 7.947676100961143e-10, "loss": 0.0558, "step": 199660 }, { "epoch": 6.750397241286048, "grad_norm": 1.638851523399353, "learning_rate": 7.488372774699937e-10, "loss": 0.0505, "step": 199670 }, { "epoch": 6.750735318976301, "grad_norm": 1.2994359731674194, "learning_rate": 7.042739055829195e-10, "loss": 0.0569, "step": 199680 }, { "epoch": 6.751073396666554, "grad_norm": 0.6949967741966248, "learning_rate": 6.610774956561372e-10, "loss": 0.0565, "step": 199690 }, { "epoch": 6.751411474356807, "grad_norm": 1.1405576467514038, "learning_rate": 6.192480488775854e-10, "loss": 0.05, "step": 199700 }, { "epoch": 6.751749552047061, "grad_norm": 1.0327447652816772, "learning_rate": 5.787855663796915e-10, "loss": 0.0541, "step": 199710 }, { "epoch": 6.752087629737313, "grad_norm": 5.720043182373047, "learning_rate": 5.396900492726786e-10, "loss": 0.0508, "step": 199720 }, { "epoch": 6.752425707427567, "grad_norm": 0.47028905153274536, "learning_rate": 5.01961498633463e-10, "loss": 0.0524, "step": 199730 }, { "epoch": 6.75276378511782, "grad_norm": 1.0418813228607178, "learning_rate": 4.6559991547789896e-10, "loss": 0.051, "step": 199740 }, { "epoch": 6.753101862808073, "grad_norm": 1.032275915145874, "learning_rate": 4.3060530081628915e-10, "loss": 0.0504, "step": 199750 }, { "epoch": 6.7534399404983265, "grad_norm": 1.0525505542755127, "learning_rate": 3.9697765559232326e-10, "loss": 0.0518, "step": 199760 }, { "epoch": 6.75377801818858, "grad_norm": 1.532634973526001, "learning_rate": 3.6471698073858864e-10, "loss": 0.0516, "step": 199770 }, { "epoch": 6.754116095878833, "grad_norm": 0.8051497340202332, "learning_rate": 3.338232771210592e-10, "loss": 0.0555, "step": 199780 }, { "epoch": 6.754454173569086, "grad_norm": 0.9378225207328796, "learning_rate": 3.0429654560015783e-10, "loss": 0.0522, "step": 199790 }, { "epoch": 6.75479225125934, "grad_norm": 0.7875657081604004, "learning_rate": 2.7613678696969404e-10, "loss": 0.0473, "step": 199800 }, { "epoch": 6.755130328949592, "grad_norm": 0.7846224904060364, "learning_rate": 2.4934400200682384e-10, "loss": 0.0488, "step": 199810 }, { "epoch": 6.755468406639846, "grad_norm": 0.4933334290981293, "learning_rate": 2.2391819144429448e-10, "loss": 0.0499, "step": 199820 }, { "epoch": 6.755806484330099, "grad_norm": 2.8653581142425537, "learning_rate": 1.9985935597599536e-10, "loss": 0.0537, "step": 199830 }, { "epoch": 6.756144562020352, "grad_norm": 0.904286801815033, "learning_rate": 1.7716749625695804e-10, "loss": 0.0517, "step": 199840 }, { "epoch": 6.7564826397106055, "grad_norm": 2.0518083572387695, "learning_rate": 1.5584261291445857e-10, "loss": 0.0562, "step": 199850 }, { "epoch": 6.756820717400859, "grad_norm": 1.2057631015777588, "learning_rate": 1.3588470652581286e-10, "loss": 0.0508, "step": 199860 }, { "epoch": 6.757158795091112, "grad_norm": 3.2838566303253174, "learning_rate": 1.1729377763503025e-10, "loss": 0.052, "step": 199870 }, { "epoch": 6.757496872781365, "grad_norm": 0.9159718751907349, "learning_rate": 1.0006982675836441e-10, "loss": 0.0488, "step": 199880 }, { "epoch": 6.757834950471619, "grad_norm": 0.9881700873374939, "learning_rate": 8.42128543565579e-11, "loss": 0.0464, "step": 199890 }, { "epoch": 6.758173028161871, "grad_norm": 1.0027574300765991, "learning_rate": 6.972286086814883e-11, "loss": 0.0548, "step": 199900 }, { "epoch": 6.758511105852125, "grad_norm": 0.8048615455627441, "learning_rate": 5.659984669281748e-11, "loss": 0.0539, "step": 199910 }, { "epoch": 6.758849183542378, "grad_norm": 1.834045171737671, "learning_rate": 4.4843812185835224e-11, "loss": 0.0576, "step": 199920 }, { "epoch": 6.759187261232631, "grad_norm": 0.577520489692688, "learning_rate": 3.4454757669166724e-11, "loss": 0.0488, "step": 199930 }, { "epoch": 6.759525338922884, "grad_norm": 0.49702170491218567, "learning_rate": 2.543268342591887e-11, "loss": 0.0532, "step": 199940 }, { "epoch": 6.759863416613138, "grad_norm": 0.926273763179779, "learning_rate": 1.7777589700340715e-11, "loss": 0.0563, "step": 199950 }, { "epoch": 6.760201494303391, "grad_norm": 1.5751529932022095, "learning_rate": 1.1489476703374636e-11, "loss": 0.0531, "step": 199960 }, { "epoch": 6.760539571993644, "grad_norm": 1.6181734800338745, "learning_rate": 6.568344612656319e-12, "loss": 0.0556, "step": 199970 }, { "epoch": 6.760877649683898, "grad_norm": 1.0868849754333496, "learning_rate": 3.014193555861411e-12, "loss": 0.0501, "step": 199980 }, { "epoch": 6.76121572737415, "grad_norm": 0.39606398344039917, "learning_rate": 8.270236329099845e-13, "loss": 0.0551, "step": 199990 }, { "epoch": 6.761553805064404, "grad_norm": 2.7938971519470215, "learning_rate": 6.834904864305713e-15, "loss": 0.0521, "step": 200000 }, { "epoch": 6.761553805064404, "step": 200000, "total_flos": 0.0, "train_loss": 0.06004954213961959, "train_runtime": 82460.757, "train_samples_per_second": 155.225, "train_steps_per_second": 2.425 } ], "logging_steps": 10, "max_steps": 200000, "num_input_tokens_seen": 0, "num_train_epochs": 7, "save_steps": 50000, "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": 64, "trial_name": null, "trial_params": null }