{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 20000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 2.4760658740997314, "learning_rate": 9e-07, "loss": 1.1686, "step": 10 }, { "grad_norm": 1.8600856065750122, "learning_rate": 1.9e-06, "loss": 1.1581, "step": 20 }, { "grad_norm": 1.0571194887161255, "learning_rate": 2.9e-06, "loss": 1.1525, "step": 30 }, { "grad_norm": 0.4121485948562622, "learning_rate": 3.9e-06, "loss": 1.11, "step": 40 }, { "grad_norm": 0.39927810430526733, "learning_rate": 4.9000000000000005e-06, "loss": 1.1001, "step": 50 }, { "grad_norm": 0.31044280529022217, "learning_rate": 5.9e-06, "loss": 1.0949, "step": 60 }, { "grad_norm": 0.3371363878250122, "learning_rate": 6.900000000000001e-06, "loss": 1.086, "step": 70 }, { "grad_norm": 0.4314686357975006, "learning_rate": 7.9e-06, "loss": 1.0716, "step": 80 }, { "grad_norm": 0.41141462326049805, "learning_rate": 8.9e-06, "loss": 1.0468, "step": 90 }, { "grad_norm": 0.3785245418548584, "learning_rate": 9.900000000000002e-06, "loss": 1.0554, "step": 100 }, { "grad_norm": 0.35778874158859253, "learning_rate": 1.09e-05, "loss": 1.0566, "step": 110 }, { "grad_norm": 0.5309902429580688, "learning_rate": 1.19e-05, "loss": 1.0363, "step": 120 }, { "grad_norm": 0.3853893280029297, "learning_rate": 1.29e-05, "loss": 1.0311, "step": 130 }, { "grad_norm": 0.4217294752597809, "learning_rate": 1.3900000000000002e-05, "loss": 1.0335, "step": 140 }, { "grad_norm": 0.48050689697265625, "learning_rate": 1.49e-05, "loss": 1.028, "step": 150 }, { "grad_norm": 0.938592255115509, "learning_rate": 1.59e-05, "loss": 0.9942, "step": 160 }, { "grad_norm": 3.108050584793091, "learning_rate": 1.69e-05, "loss": 0.9407, "step": 170 }, { "grad_norm": 1.752050518989563, "learning_rate": 1.79e-05, "loss": 0.8978, "step": 180 }, { "grad_norm": 2.8531410694122314, "learning_rate": 1.8900000000000002e-05, "loss": 0.7982, "step": 190 }, { "grad_norm": 2.2095277309417725, "learning_rate": 1.9900000000000003e-05, "loss": 0.6942, "step": 200 }, { "grad_norm": 2.044131278991699, "learning_rate": 2.09e-05, "loss": 0.6207, "step": 210 }, { "grad_norm": 3.021023750305176, "learning_rate": 2.19e-05, "loss": 0.5206, "step": 220 }, { "grad_norm": 2.687671184539795, "learning_rate": 2.29e-05, "loss": 0.4123, "step": 230 }, { "grad_norm": 2.8484673500061035, "learning_rate": 2.39e-05, "loss": 0.3109, "step": 240 }, { "grad_norm": 2.6023690700531006, "learning_rate": 2.4900000000000002e-05, "loss": 0.2442, "step": 250 }, { "grad_norm": 2.3004181385040283, "learning_rate": 2.5900000000000003e-05, "loss": 0.206, "step": 260 }, { "grad_norm": 1.675736665725708, "learning_rate": 2.6900000000000003e-05, "loss": 0.1806, "step": 270 }, { "grad_norm": 2.281965494155884, "learning_rate": 2.7900000000000004e-05, "loss": 0.163, "step": 280 }, { "grad_norm": 1.4266277551651, "learning_rate": 2.8899999999999998e-05, "loss": 0.1532, "step": 290 }, { "grad_norm": 1.3119257688522339, "learning_rate": 2.9900000000000002e-05, "loss": 0.154, "step": 300 }, { "grad_norm": 1.5665065050125122, "learning_rate": 3.09e-05, "loss": 0.1485, "step": 310 }, { "grad_norm": 1.5950955152511597, "learning_rate": 3.19e-05, "loss": 0.1293, "step": 320 }, { "grad_norm": 1.8489373922348022, "learning_rate": 3.29e-05, "loss": 0.1264, "step": 330 }, { "grad_norm": 1.7027757167816162, "learning_rate": 3.3900000000000004e-05, "loss": 0.1385, "step": 340 }, { "grad_norm": 1.9789204597473145, "learning_rate": 3.49e-05, "loss": 0.1252, "step": 350 }, { "grad_norm": 1.3323863744735718, "learning_rate": 3.59e-05, "loss": 0.1229, "step": 360 }, { "grad_norm": 1.8391609191894531, "learning_rate": 3.69e-05, "loss": 0.1114, "step": 370 }, { "grad_norm": 1.3688678741455078, "learning_rate": 3.79e-05, "loss": 0.1095, "step": 380 }, { "grad_norm": 0.8539863228797913, "learning_rate": 3.8900000000000004e-05, "loss": 0.1063, "step": 390 }, { "grad_norm": 1.3404130935668945, "learning_rate": 3.99e-05, "loss": 0.1037, "step": 400 }, { "grad_norm": 1.3641117811203003, "learning_rate": 4.09e-05, "loss": 0.1055, "step": 410 }, { "grad_norm": 1.0738294124603271, "learning_rate": 4.19e-05, "loss": 0.1048, "step": 420 }, { "grad_norm": 1.149735689163208, "learning_rate": 4.29e-05, "loss": 0.112, "step": 430 }, { "grad_norm": 1.025214433670044, "learning_rate": 4.39e-05, "loss": 0.0955, "step": 440 }, { "grad_norm": 1.1336759328842163, "learning_rate": 4.49e-05, "loss": 0.0949, "step": 450 }, { "grad_norm": 1.0962635278701782, "learning_rate": 4.5900000000000004e-05, "loss": 0.0967, "step": 460 }, { "grad_norm": 1.014551043510437, "learning_rate": 4.69e-05, "loss": 0.0931, "step": 470 }, { "grad_norm": 0.814590334892273, "learning_rate": 4.79e-05, "loss": 0.0948, "step": 480 }, { "grad_norm": 1.1694084405899048, "learning_rate": 4.89e-05, "loss": 0.0917, "step": 490 }, { "grad_norm": 1.330562949180603, "learning_rate": 4.99e-05, "loss": 0.0907, "step": 500 }, { "grad_norm": 1.4349883794784546, "learning_rate": 5.0900000000000004e-05, "loss": 0.0888, "step": 510 }, { "grad_norm": 1.6133650541305542, "learning_rate": 5.19e-05, "loss": 0.0915, "step": 520 }, { "grad_norm": 1.006913185119629, "learning_rate": 5.2900000000000005e-05, "loss": 0.0918, "step": 530 }, { "grad_norm": 0.7664061784744263, "learning_rate": 5.390000000000001e-05, "loss": 0.0888, "step": 540 }, { "grad_norm": 0.8765918016433716, "learning_rate": 5.4900000000000006e-05, "loss": 0.0787, "step": 550 }, { "grad_norm": 0.807823121547699, "learning_rate": 5.590000000000001e-05, "loss": 0.0846, "step": 560 }, { "grad_norm": 1.0837857723236084, "learning_rate": 5.69e-05, "loss": 0.0899, "step": 570 }, { "grad_norm": 0.979864239692688, "learning_rate": 5.79e-05, "loss": 0.0862, "step": 580 }, { "grad_norm": 0.9319878220558167, "learning_rate": 5.89e-05, "loss": 0.0891, "step": 590 }, { "grad_norm": 0.8744037747383118, "learning_rate": 5.99e-05, "loss": 0.0865, "step": 600 }, { "grad_norm": 0.9416177868843079, "learning_rate": 6.09e-05, "loss": 0.0823, "step": 610 }, { "grad_norm": 0.7182490229606628, "learning_rate": 6.19e-05, "loss": 0.0793, "step": 620 }, { "grad_norm": 0.65711909532547, "learning_rate": 6.29e-05, "loss": 0.0806, "step": 630 }, { "grad_norm": 0.8147222399711609, "learning_rate": 6.390000000000001e-05, "loss": 0.0812, "step": 640 }, { "grad_norm": 0.6686334013938904, "learning_rate": 6.49e-05, "loss": 0.0841, "step": 650 }, { "grad_norm": 0.9076411128044128, "learning_rate": 6.59e-05, "loss": 0.0778, "step": 660 }, { "grad_norm": 0.7595466375350952, "learning_rate": 6.690000000000001e-05, "loss": 0.0752, "step": 670 }, { "grad_norm": 0.7147392630577087, "learning_rate": 6.790000000000001e-05, "loss": 0.0779, "step": 680 }, { "grad_norm": 0.6534247994422913, "learning_rate": 6.89e-05, "loss": 0.0671, "step": 690 }, { "grad_norm": 0.6289360523223877, "learning_rate": 6.99e-05, "loss": 0.0698, "step": 700 }, { "grad_norm": 0.618780791759491, "learning_rate": 7.09e-05, "loss": 0.0657, "step": 710 }, { "grad_norm": 0.6702368259429932, "learning_rate": 7.19e-05, "loss": 0.0692, "step": 720 }, { "grad_norm": 0.6759467720985413, "learning_rate": 7.29e-05, "loss": 0.0749, "step": 730 }, { "grad_norm": 0.37818241119384766, "learning_rate": 7.390000000000001e-05, "loss": 0.0733, "step": 740 }, { "grad_norm": 0.827802300453186, "learning_rate": 7.49e-05, "loss": 0.0739, "step": 750 }, { "grad_norm": 0.8170549869537354, "learning_rate": 7.59e-05, "loss": 0.0721, "step": 760 }, { "grad_norm": 0.9417720437049866, "learning_rate": 7.69e-05, "loss": 0.0763, "step": 770 }, { "grad_norm": 0.6940046548843384, "learning_rate": 7.790000000000001e-05, "loss": 0.0782, "step": 780 }, { "grad_norm": 0.6347594261169434, "learning_rate": 7.890000000000001e-05, "loss": 0.0712, "step": 790 }, { "grad_norm": 0.7504335641860962, "learning_rate": 7.99e-05, "loss": 0.0733, "step": 800 }, { "grad_norm": 0.8423503041267395, "learning_rate": 8.090000000000001e-05, "loss": 0.0655, "step": 810 }, { "grad_norm": 0.6170629262924194, "learning_rate": 8.19e-05, "loss": 0.0695, "step": 820 }, { "grad_norm": 0.48295333981513977, "learning_rate": 8.29e-05, "loss": 0.0647, "step": 830 }, { "grad_norm": 0.6359453797340393, "learning_rate": 8.39e-05, "loss": 0.0653, "step": 840 }, { "grad_norm": 0.6045881509780884, "learning_rate": 8.49e-05, "loss": 0.0638, "step": 850 }, { "grad_norm": 0.6148896217346191, "learning_rate": 8.59e-05, "loss": 0.0636, "step": 860 }, { "grad_norm": 0.5587151646614075, "learning_rate": 8.69e-05, "loss": 0.0665, "step": 870 }, { "grad_norm": 0.41081568598747253, "learning_rate": 8.790000000000001e-05, "loss": 0.066, "step": 880 }, { "grad_norm": 0.42127594351768494, "learning_rate": 8.89e-05, "loss": 0.0642, "step": 890 }, { "grad_norm": 0.6480566263198853, "learning_rate": 8.99e-05, "loss": 0.069, "step": 900 }, { "grad_norm": 0.5283571481704712, "learning_rate": 9.090000000000001e-05, "loss": 0.0615, "step": 910 }, { "grad_norm": 0.6150722503662109, "learning_rate": 9.190000000000001e-05, "loss": 0.0601, "step": 920 }, { "grad_norm": 0.532468855381012, "learning_rate": 9.290000000000001e-05, "loss": 0.0639, "step": 930 }, { "grad_norm": 0.5080350041389465, "learning_rate": 9.39e-05, "loss": 0.0617, "step": 940 }, { "grad_norm": 0.40383249521255493, "learning_rate": 9.49e-05, "loss": 0.0608, "step": 950 }, { "grad_norm": 0.6272702217102051, "learning_rate": 9.59e-05, "loss": 0.0627, "step": 960 }, { "grad_norm": 0.6091583967208862, "learning_rate": 9.69e-05, "loss": 0.064, "step": 970 }, { "grad_norm": 0.3984649181365967, "learning_rate": 9.790000000000001e-05, "loss": 0.0593, "step": 980 }, { "grad_norm": 0.6853853464126587, "learning_rate": 9.89e-05, "loss": 0.0721, "step": 990 }, { "grad_norm": 0.4686420261859894, "learning_rate": 9.99e-05, "loss": 0.0573, "step": 1000 }, { "grad_norm": 0.7333941459655762, "learning_rate": 9.999994463727085e-05, "loss": 0.0574, "step": 1010 }, { "grad_norm": 0.6852871179580688, "learning_rate": 9.999975326009292e-05, "loss": 0.0666, "step": 1020 }, { "grad_norm": 0.3442159593105316, "learning_rate": 9.999942518549879e-05, "loss": 0.0571, "step": 1030 }, { "grad_norm": 0.5390961170196533, "learning_rate": 9.999896041438544e-05, "loss": 0.0543, "step": 1040 }, { "grad_norm": 0.37518876791000366, "learning_rate": 9.999835894802353e-05, "loss": 0.0524, "step": 1050 }, { "grad_norm": 0.4920927584171295, "learning_rate": 9.999762078805743e-05, "loss": 0.051, "step": 1060 }, { "grad_norm": 0.5402520895004272, "learning_rate": 9.999674593650526e-05, "loss": 0.0562, "step": 1070 }, { "grad_norm": 0.49340325593948364, "learning_rate": 9.99957343957588e-05, "loss": 0.0566, "step": 1080 }, { "grad_norm": 0.4530620574951172, "learning_rate": 9.99945861685836e-05, "loss": 0.0561, "step": 1090 }, { "grad_norm": 0.34514665603637695, "learning_rate": 9.999330125811884e-05, "loss": 0.0477, "step": 1100 }, { "grad_norm": 0.5346470475196838, "learning_rate": 9.999187966787744e-05, "loss": 0.0487, "step": 1110 }, { "grad_norm": 0.5709505081176758, "learning_rate": 9.999032140174595e-05, "loss": 0.0535, "step": 1120 }, { "grad_norm": 0.36133873462677, "learning_rate": 9.998862646398464e-05, "loss": 0.0564, "step": 1130 }, { "grad_norm": 0.4472231864929199, "learning_rate": 9.998679485922739e-05, "loss": 0.0522, "step": 1140 }, { "grad_norm": 0.5858401656150818, "learning_rate": 9.998482659248174e-05, "loss": 0.0515, "step": 1150 }, { "grad_norm": 0.4985833764076233, "learning_rate": 9.998272166912883e-05, "loss": 0.0555, "step": 1160 }, { "grad_norm": 0.3400823771953583, "learning_rate": 9.998048009492347e-05, "loss": 0.056, "step": 1170 }, { "grad_norm": 0.426344633102417, "learning_rate": 9.997810187599403e-05, "loss": 0.0525, "step": 1180 }, { "grad_norm": 0.5356988310813904, "learning_rate": 9.997558701884249e-05, "loss": 0.0476, "step": 1190 }, { "grad_norm": 0.3942427337169647, "learning_rate": 9.997293553034433e-05, "loss": 0.0502, "step": 1200 }, { "grad_norm": 0.2943044602870941, "learning_rate": 9.997014741774866e-05, "loss": 0.0459, "step": 1210 }, { "grad_norm": 0.3055969774723053, "learning_rate": 9.996722268867803e-05, "loss": 0.0483, "step": 1220 }, { "grad_norm": 0.34858521819114685, "learning_rate": 9.996416135112858e-05, "loss": 0.0507, "step": 1230 }, { "grad_norm": 0.48210063576698303, "learning_rate": 9.996096341346988e-05, "loss": 0.0474, "step": 1240 }, { "grad_norm": 0.39170143008232117, "learning_rate": 9.995762888444495e-05, "loss": 0.0481, "step": 1250 }, { "grad_norm": 0.3783523738384247, "learning_rate": 9.995415777317027e-05, "loss": 0.0479, "step": 1260 }, { "grad_norm": 0.5220600962638855, "learning_rate": 9.995055008913574e-05, "loss": 0.0478, "step": 1270 }, { "grad_norm": 0.38259047269821167, "learning_rate": 9.994680584220463e-05, "loss": 0.0516, "step": 1280 }, { "grad_norm": 0.3639887273311615, "learning_rate": 9.994292504261355e-05, "loss": 0.0471, "step": 1290 }, { "grad_norm": 0.3751421570777893, "learning_rate": 9.993890770097247e-05, "loss": 0.0462, "step": 1300 }, { "grad_norm": 0.4323166310787201, "learning_rate": 9.993475382826467e-05, "loss": 0.0445, "step": 1310 }, { "grad_norm": 0.3960559070110321, "learning_rate": 9.993046343584664e-05, "loss": 0.04, "step": 1320 }, { "grad_norm": 0.3886789381504059, "learning_rate": 9.992603653544816e-05, "loss": 0.0534, "step": 1330 }, { "grad_norm": 0.4407539665699005, "learning_rate": 9.992147313917222e-05, "loss": 0.0475, "step": 1340 }, { "grad_norm": 0.4733286201953888, "learning_rate": 9.991677325949497e-05, "loss": 0.0473, "step": 1350 }, { "grad_norm": 0.2731863856315613, "learning_rate": 9.991193690926568e-05, "loss": 0.0425, "step": 1360 }, { "grad_norm": 0.4292997419834137, "learning_rate": 9.990696410170678e-05, "loss": 0.0446, "step": 1370 }, { "grad_norm": 0.41229552030563354, "learning_rate": 9.990185485041371e-05, "loss": 0.0404, "step": 1380 }, { "grad_norm": 0.3411741554737091, "learning_rate": 9.989660916935498e-05, "loss": 0.0424, "step": 1390 }, { "grad_norm": 0.4606143832206726, "learning_rate": 9.989122707287208e-05, "loss": 0.0436, "step": 1400 }, { "grad_norm": 0.363071084022522, "learning_rate": 9.988570857567945e-05, "loss": 0.0425, "step": 1410 }, { "grad_norm": 0.4887494444847107, "learning_rate": 9.988005369286446e-05, "loss": 0.042, "step": 1420 }, { "grad_norm": 0.4026138484477997, "learning_rate": 9.987426243988734e-05, "loss": 0.0461, "step": 1430 }, { "grad_norm": 0.4510866105556488, "learning_rate": 9.986833483258114e-05, "loss": 0.0415, "step": 1440 }, { "grad_norm": 0.3407531678676605, "learning_rate": 9.986227088715173e-05, "loss": 0.0422, "step": 1450 }, { "grad_norm": 0.4207153022289276, "learning_rate": 9.98560706201777e-05, "loss": 0.05, "step": 1460 }, { "grad_norm": 0.6353428959846497, "learning_rate": 9.984973404861036e-05, "loss": 0.0398, "step": 1470 }, { "grad_norm": 0.49047642946243286, "learning_rate": 9.984326118977361e-05, "loss": 0.0449, "step": 1480 }, { "grad_norm": 0.49542757868766785, "learning_rate": 9.983665206136406e-05, "loss": 0.0417, "step": 1490 }, { "grad_norm": 0.2879747152328491, "learning_rate": 9.982990668145075e-05, "loss": 0.0409, "step": 1500 }, { "grad_norm": 0.6042370796203613, "learning_rate": 9.982302506847534e-05, "loss": 0.0459, "step": 1510 }, { "grad_norm": 0.39689815044403076, "learning_rate": 9.981600724125189e-05, "loss": 0.0388, "step": 1520 }, { "grad_norm": 0.35404685139656067, "learning_rate": 9.980885321896685e-05, "loss": 0.0426, "step": 1530 }, { "grad_norm": 0.31749141216278076, "learning_rate": 9.980156302117905e-05, "loss": 0.0426, "step": 1540 }, { "grad_norm": 0.29580625891685486, "learning_rate": 9.979413666781963e-05, "loss": 0.0433, "step": 1550 }, { "grad_norm": 0.3974514603614807, "learning_rate": 9.978657417919193e-05, "loss": 0.0435, "step": 1560 }, { "grad_norm": 0.42137080430984497, "learning_rate": 9.977887557597153e-05, "loss": 0.0453, "step": 1570 }, { "grad_norm": 0.38182738423347473, "learning_rate": 9.97710408792061e-05, "loss": 0.0413, "step": 1580 }, { "grad_norm": 0.27168598771095276, "learning_rate": 9.976307011031542e-05, "loss": 0.0399, "step": 1590 }, { "grad_norm": 0.37684035301208496, "learning_rate": 9.975496329109126e-05, "loss": 0.0423, "step": 1600 }, { "grad_norm": 0.42773041129112244, "learning_rate": 9.974672044369732e-05, "loss": 0.0389, "step": 1610 }, { "grad_norm": 0.38486358523368835, "learning_rate": 9.97383415906693e-05, "loss": 0.0381, "step": 1620 }, { "grad_norm": 0.31259557604789734, "learning_rate": 9.97298267549146e-05, "loss": 0.04, "step": 1630 }, { "grad_norm": 0.2967011332511902, "learning_rate": 9.972117595971249e-05, "loss": 0.039, "step": 1640 }, { "grad_norm": 0.39362311363220215, "learning_rate": 9.971238922871391e-05, "loss": 0.0367, "step": 1650 }, { "grad_norm": 0.3642995059490204, "learning_rate": 9.970346658594142e-05, "loss": 0.0358, "step": 1660 }, { "grad_norm": 0.3209611475467682, "learning_rate": 9.969440805578923e-05, "loss": 0.0367, "step": 1670 }, { "grad_norm": 0.22932590544223785, "learning_rate": 9.968521366302298e-05, "loss": 0.0381, "step": 1680 }, { "grad_norm": 0.4837380349636078, "learning_rate": 9.967588343277981e-05, "loss": 0.0476, "step": 1690 }, { "grad_norm": 0.41174453496932983, "learning_rate": 9.966641739056818e-05, "loss": 0.0451, "step": 1700 }, { "grad_norm": 0.25535985827445984, "learning_rate": 9.965681556226793e-05, "loss": 0.0399, "step": 1710 }, { "grad_norm": 0.27589040994644165, "learning_rate": 9.964707797413006e-05, "loss": 0.0375, "step": 1720 }, { "grad_norm": 0.31984713673591614, "learning_rate": 9.963720465277679e-05, "loss": 0.0381, "step": 1730 }, { "grad_norm": 0.37560322880744934, "learning_rate": 9.96271956252014e-05, "loss": 0.037, "step": 1740 }, { "grad_norm": 0.32271909713745117, "learning_rate": 9.961705091876816e-05, "loss": 0.0421, "step": 1750 }, { "grad_norm": 0.3897894322872162, "learning_rate": 9.960677056121235e-05, "loss": 0.0374, "step": 1760 }, { "grad_norm": 0.3455497622489929, "learning_rate": 9.959635458064005e-05, "loss": 0.0399, "step": 1770 }, { "grad_norm": 0.31158310174942017, "learning_rate": 9.958580300552815e-05, "loss": 0.0379, "step": 1780 }, { "grad_norm": 0.41133132576942444, "learning_rate": 9.957511586472426e-05, "loss": 0.0395, "step": 1790 }, { "grad_norm": 0.24093501269817352, "learning_rate": 9.956429318744662e-05, "loss": 0.0378, "step": 1800 }, { "grad_norm": 0.34598711133003235, "learning_rate": 9.955333500328404e-05, "loss": 0.0342, "step": 1810 }, { "grad_norm": 0.36232852935791016, "learning_rate": 9.95422413421957e-05, "loss": 0.0377, "step": 1820 }, { "grad_norm": 0.3515927791595459, "learning_rate": 9.953101223451133e-05, "loss": 0.0352, "step": 1830 }, { "grad_norm": 0.35568922758102417, "learning_rate": 9.951964771093085e-05, "loss": 0.0409, "step": 1840 }, { "grad_norm": 0.4155760705471039, "learning_rate": 9.950814780252442e-05, "loss": 0.041, "step": 1850 }, { "grad_norm": 0.3880210220813751, "learning_rate": 9.949651254073236e-05, "loss": 0.0418, "step": 1860 }, { "grad_norm": 0.3192378282546997, "learning_rate": 9.948474195736504e-05, "loss": 0.0381, "step": 1870 }, { "grad_norm": 0.32938939332962036, "learning_rate": 9.947283608460277e-05, "loss": 0.0386, "step": 1880 }, { "grad_norm": 0.35251763463020325, "learning_rate": 9.946079495499577e-05, "loss": 0.043, "step": 1890 }, { "grad_norm": 0.2882311940193176, "learning_rate": 9.944861860146401e-05, "loss": 0.0384, "step": 1900 }, { "grad_norm": 0.3445051610469818, "learning_rate": 9.943630705729719e-05, "loss": 0.036, "step": 1910 }, { "grad_norm": 0.29421016573905945, "learning_rate": 9.942386035615459e-05, "loss": 0.037, "step": 1920 }, { "grad_norm": 0.288557231426239, "learning_rate": 9.941127853206503e-05, "loss": 0.0392, "step": 1930 }, { "grad_norm": 0.3950519859790802, "learning_rate": 9.939856161942673e-05, "loss": 0.0409, "step": 1940 }, { "grad_norm": 0.3286008834838867, "learning_rate": 9.938570965300724e-05, "loss": 0.039, "step": 1950 }, { "grad_norm": 0.2765180468559265, "learning_rate": 9.937272266794335e-05, "loss": 0.0369, "step": 1960 }, { "grad_norm": 0.3880895674228668, "learning_rate": 9.935960069974096e-05, "loss": 0.0419, "step": 1970 }, { "grad_norm": 0.27147775888442993, "learning_rate": 9.934634378427506e-05, "loss": 0.0405, "step": 1980 }, { "grad_norm": 0.39060819149017334, "learning_rate": 9.933295195778954e-05, "loss": 0.0387, "step": 1990 }, { "grad_norm": 0.27577319741249084, "learning_rate": 9.931942525689715e-05, "loss": 0.0365, "step": 2000 }, { "grad_norm": 0.2607966363430023, "learning_rate": 9.930576371857936e-05, "loss": 0.036, "step": 2010 }, { "grad_norm": 0.3863831162452698, "learning_rate": 9.929196738018629e-05, "loss": 0.0392, "step": 2020 }, { "grad_norm": 0.2632920444011688, "learning_rate": 9.927803627943662e-05, "loss": 0.037, "step": 2030 }, { "grad_norm": 0.2446487545967102, "learning_rate": 9.926397045441744e-05, "loss": 0.0357, "step": 2040 }, { "grad_norm": 0.37675121426582336, "learning_rate": 9.924976994358417e-05, "loss": 0.0395, "step": 2050 }, { "grad_norm": 0.2831648290157318, "learning_rate": 9.923543478576048e-05, "loss": 0.0359, "step": 2060 }, { "grad_norm": 0.2651529014110565, "learning_rate": 9.922096502013813e-05, "loss": 0.0379, "step": 2070 }, { "grad_norm": 0.2821183204650879, "learning_rate": 9.92063606862769e-05, "loss": 0.032, "step": 2080 }, { "grad_norm": 0.29363521933555603, "learning_rate": 9.919162182410453e-05, "loss": 0.0396, "step": 2090 }, { "grad_norm": 0.345195472240448, "learning_rate": 9.917674847391645e-05, "loss": 0.0336, "step": 2100 }, { "grad_norm": 0.45705685019493103, "learning_rate": 9.916174067637584e-05, "loss": 0.0307, "step": 2110 }, { "grad_norm": 0.2671544551849365, "learning_rate": 9.914659847251348e-05, "loss": 0.033, "step": 2120 }, { "grad_norm": 0.3651313781738281, "learning_rate": 9.913132190372753e-05, "loss": 0.0356, "step": 2130 }, { "grad_norm": 0.3620646893978119, "learning_rate": 9.911591101178359e-05, "loss": 0.0334, "step": 2140 }, { "grad_norm": 0.22402408719062805, "learning_rate": 9.910036583881443e-05, "loss": 0.0354, "step": 2150 }, { "grad_norm": 0.23167388141155243, "learning_rate": 9.908468642731995e-05, "loss": 0.031, "step": 2160 }, { "grad_norm": 0.34602999687194824, "learning_rate": 9.906887282016707e-05, "loss": 0.0347, "step": 2170 }, { "grad_norm": 0.39348194003105164, "learning_rate": 9.90529250605896e-05, "loss": 0.0319, "step": 2180 }, { "grad_norm": 0.30921998620033264, "learning_rate": 9.903684319218809e-05, "loss": 0.0316, "step": 2190 }, { "grad_norm": 0.28107547760009766, "learning_rate": 9.902062725892976e-05, "loss": 0.0394, "step": 2200 }, { "grad_norm": 0.23763656616210938, "learning_rate": 9.900427730514834e-05, "loss": 0.0342, "step": 2210 }, { "grad_norm": 0.3290095925331116, "learning_rate": 9.8987793375544e-05, "loss": 0.0371, "step": 2220 }, { "grad_norm": 0.301687091588974, "learning_rate": 9.897117551518318e-05, "loss": 0.0322, "step": 2230 }, { "grad_norm": 0.33845826983451843, "learning_rate": 9.895442376949844e-05, "loss": 0.0328, "step": 2240 }, { "grad_norm": 0.26240310072898865, "learning_rate": 9.893753818428845e-05, "loss": 0.0343, "step": 2250 }, { "grad_norm": 0.29250526428222656, "learning_rate": 9.892051880571773e-05, "loss": 0.0365, "step": 2260 }, { "grad_norm": 0.20801101624965668, "learning_rate": 9.890336568031663e-05, "loss": 0.0322, "step": 2270 }, { "grad_norm": 0.21533405780792236, "learning_rate": 9.888607885498113e-05, "loss": 0.0302, "step": 2280 }, { "grad_norm": 0.2218470573425293, "learning_rate": 9.886865837697275e-05, "loss": 0.0323, "step": 2290 }, { "grad_norm": 0.2593401074409485, "learning_rate": 9.88511042939184e-05, "loss": 0.0302, "step": 2300 }, { "grad_norm": 0.2596835792064667, "learning_rate": 9.883341665381028e-05, "loss": 0.0348, "step": 2310 }, { "grad_norm": 0.326631635427475, "learning_rate": 9.881559550500575e-05, "loss": 0.0344, "step": 2320 }, { "grad_norm": 0.2603054642677307, "learning_rate": 9.879764089622712e-05, "loss": 0.0381, "step": 2330 }, { "grad_norm": 0.2734562158584595, "learning_rate": 9.87795528765616e-05, "loss": 0.03, "step": 2340 }, { "grad_norm": 0.2510208189487457, "learning_rate": 9.876133149546118e-05, "loss": 0.0362, "step": 2350 }, { "grad_norm": 0.24825884401798248, "learning_rate": 9.874297680274238e-05, "loss": 0.0318, "step": 2360 }, { "grad_norm": 0.28092172741889954, "learning_rate": 9.872448884858624e-05, "loss": 0.0303, "step": 2370 }, { "grad_norm": 0.3570152521133423, "learning_rate": 9.870586768353815e-05, "loss": 0.0318, "step": 2380 }, { "grad_norm": 0.2720980644226074, "learning_rate": 9.868711335850764e-05, "loss": 0.0328, "step": 2390 }, { "grad_norm": 0.1665506809949875, "learning_rate": 9.866822592476833e-05, "loss": 0.0311, "step": 2400 }, { "grad_norm": 0.35338079929351807, "learning_rate": 9.86492054339577e-05, "loss": 0.0276, "step": 2410 }, { "grad_norm": 0.24944140017032623, "learning_rate": 9.863005193807711e-05, "loss": 0.0408, "step": 2420 }, { "grad_norm": 0.19900642335414886, "learning_rate": 9.861076548949143e-05, "loss": 0.0312, "step": 2430 }, { "grad_norm": 0.22543023526668549, "learning_rate": 9.859134614092912e-05, "loss": 0.0332, "step": 2440 }, { "grad_norm": 0.33223649859428406, "learning_rate": 9.857179394548191e-05, "loss": 0.0328, "step": 2450 }, { "grad_norm": 0.34036460518836975, "learning_rate": 9.855210895660477e-05, "loss": 0.0304, "step": 2460 }, { "grad_norm": 0.244074746966362, "learning_rate": 9.853229122811568e-05, "loss": 0.0334, "step": 2470 }, { "grad_norm": 0.2671298086643219, "learning_rate": 9.851234081419559e-05, "loss": 0.0295, "step": 2480 }, { "grad_norm": 0.2725508511066437, "learning_rate": 9.849225776938814e-05, "loss": 0.0375, "step": 2490 }, { "grad_norm": 0.3493804931640625, "learning_rate": 9.847204214859964e-05, "loss": 0.0333, "step": 2500 }, { "grad_norm": 0.3126058280467987, "learning_rate": 9.845169400709879e-05, "loss": 0.0348, "step": 2510 }, { "grad_norm": 0.3028257191181183, "learning_rate": 9.843121340051664e-05, "loss": 0.0344, "step": 2520 }, { "grad_norm": 0.34032270312309265, "learning_rate": 9.841060038484641e-05, "loss": 0.0343, "step": 2530 }, { "grad_norm": 0.2512003183364868, "learning_rate": 9.838985501644328e-05, "loss": 0.031, "step": 2540 }, { "grad_norm": 0.2976381778717041, "learning_rate": 9.83689773520243e-05, "loss": 0.0355, "step": 2550 }, { "grad_norm": 0.293525755405426, "learning_rate": 9.834796744866819e-05, "loss": 0.037, "step": 2560 }, { "grad_norm": 0.2686387598514557, "learning_rate": 9.832682536381525e-05, "loss": 0.031, "step": 2570 }, { "grad_norm": 0.2771541178226471, "learning_rate": 9.830555115526711e-05, "loss": 0.0323, "step": 2580 }, { "grad_norm": 0.1876172125339508, "learning_rate": 9.828414488118667e-05, "loss": 0.0284, "step": 2590 }, { "grad_norm": 0.23483826220035553, "learning_rate": 9.826260660009785e-05, "loss": 0.0345, "step": 2600 }, { "grad_norm": 0.29043036699295044, "learning_rate": 9.824093637088547e-05, "loss": 0.0336, "step": 2610 }, { "grad_norm": 0.24006465077400208, "learning_rate": 9.821913425279514e-05, "loss": 0.0305, "step": 2620 }, { "grad_norm": 0.21774016320705414, "learning_rate": 9.8197200305433e-05, "loss": 0.0303, "step": 2630 }, { "grad_norm": 0.19311004877090454, "learning_rate": 9.817513458876564e-05, "loss": 0.0351, "step": 2640 }, { "grad_norm": 0.27010485529899597, "learning_rate": 9.815293716311987e-05, "loss": 0.0314, "step": 2650 }, { "grad_norm": 0.2535725235939026, "learning_rate": 9.813060808918262e-05, "loss": 0.0309, "step": 2660 }, { "grad_norm": 0.30278968811035156, "learning_rate": 9.810814742800069e-05, "loss": 0.0312, "step": 2670 }, { "grad_norm": 0.298956036567688, "learning_rate": 9.808555524098074e-05, "loss": 0.0308, "step": 2680 }, { "grad_norm": 0.24597777426242828, "learning_rate": 9.806283158988887e-05, "loss": 0.033, "step": 2690 }, { "grad_norm": 0.2095940262079239, "learning_rate": 9.803997653685072e-05, "loss": 0.0295, "step": 2700 }, { "grad_norm": 0.48583412170410156, "learning_rate": 9.801699014435112e-05, "loss": 0.0325, "step": 2710 }, { "grad_norm": 0.29490694403648376, "learning_rate": 9.799387247523398e-05, "loss": 0.0317, "step": 2720 }, { "grad_norm": 0.22433573007583618, "learning_rate": 9.797062359270215e-05, "loss": 0.0339, "step": 2730 }, { "grad_norm": 0.2470979243516922, "learning_rate": 9.794724356031715e-05, "loss": 0.0287, "step": 2740 }, { "grad_norm": 0.28273653984069824, "learning_rate": 9.792373244199913e-05, "loss": 0.0263, "step": 2750 }, { "grad_norm": 0.2537808418273926, "learning_rate": 9.790009030202658e-05, "loss": 0.0285, "step": 2760 }, { "grad_norm": 0.25292709469795227, "learning_rate": 9.78763172050362e-05, "loss": 0.0325, "step": 2770 }, { "grad_norm": 0.2579091489315033, "learning_rate": 9.785241321602274e-05, "loss": 0.0279, "step": 2780 }, { "grad_norm": 0.21910031139850616, "learning_rate": 9.782837840033879e-05, "loss": 0.0271, "step": 2790 }, { "grad_norm": 0.3830093443393707, "learning_rate": 9.780421282369461e-05, "loss": 0.0313, "step": 2800 }, { "grad_norm": 0.2738351821899414, "learning_rate": 9.777991655215797e-05, "loss": 0.0319, "step": 2810 }, { "grad_norm": 0.27481529116630554, "learning_rate": 9.775548965215394e-05, "loss": 0.0305, "step": 2820 }, { "grad_norm": 0.28763121366500854, "learning_rate": 9.773093219046474e-05, "loss": 0.0267, "step": 2830 }, { "grad_norm": 0.3087739050388336, "learning_rate": 9.770624423422954e-05, "loss": 0.0293, "step": 2840 }, { "grad_norm": 0.2544531524181366, "learning_rate": 9.768142585094426e-05, "loss": 0.0295, "step": 2850 }, { "grad_norm": 0.3114573359489441, "learning_rate": 9.765647710846142e-05, "loss": 0.0343, "step": 2860 }, { "grad_norm": 0.2625702917575836, "learning_rate": 9.763139807498991e-05, "loss": 0.0319, "step": 2870 }, { "grad_norm": 0.20481164753437042, "learning_rate": 9.760618881909487e-05, "loss": 0.0291, "step": 2880 }, { "grad_norm": 0.3278879225254059, "learning_rate": 9.758084940969744e-05, "loss": 0.0305, "step": 2890 }, { "grad_norm": 0.23843884468078613, "learning_rate": 9.755537991607459e-05, "loss": 0.0306, "step": 2900 }, { "grad_norm": 0.19844265282154083, "learning_rate": 9.752978040785895e-05, "loss": 0.027, "step": 2910 }, { "grad_norm": 0.30732959508895874, "learning_rate": 9.750405095503859e-05, "loss": 0.0333, "step": 2920 }, { "grad_norm": 0.217977374792099, "learning_rate": 9.747819162795686e-05, "loss": 0.0326, "step": 2930 }, { "grad_norm": 0.2740035653114319, "learning_rate": 9.745220249731217e-05, "loss": 0.0354, "step": 2940 }, { "grad_norm": 0.3002717196941376, "learning_rate": 9.742608363415781e-05, "loss": 0.0295, "step": 2950 }, { "grad_norm": 0.22700101137161255, "learning_rate": 9.739983510990176e-05, "loss": 0.0306, "step": 2960 }, { "grad_norm": 0.320560097694397, "learning_rate": 9.737345699630647e-05, "loss": 0.029, "step": 2970 }, { "grad_norm": 0.23764711618423462, "learning_rate": 9.734694936548869e-05, "loss": 0.027, "step": 2980 }, { "grad_norm": 0.24355937540531158, "learning_rate": 9.732031228991932e-05, "loss": 0.0276, "step": 2990 }, { "grad_norm": 0.28828638792037964, "learning_rate": 9.729354584242302e-05, "loss": 0.0298, "step": 3000 }, { "grad_norm": 0.25023505091667175, "learning_rate": 9.726665009617832e-05, "loss": 0.0265, "step": 3010 }, { "grad_norm": 0.24580618739128113, "learning_rate": 9.723962512471714e-05, "loss": 0.0282, "step": 3020 }, { "grad_norm": 0.32479482889175415, "learning_rate": 9.72124710019247e-05, "loss": 0.0252, "step": 3030 }, { "grad_norm": 0.26837533712387085, "learning_rate": 9.718518780203934e-05, "loss": 0.0289, "step": 3040 }, { "grad_norm": 0.288056880235672, "learning_rate": 9.715777559965228e-05, "loss": 0.0262, "step": 3050 }, { "grad_norm": 0.3466870188713074, "learning_rate": 9.713023446970746e-05, "loss": 0.0259, "step": 3060 }, { "grad_norm": 0.22235284745693207, "learning_rate": 9.710256448750126e-05, "loss": 0.0292, "step": 3070 }, { "grad_norm": 0.24568067491054535, "learning_rate": 9.707476572868235e-05, "loss": 0.0281, "step": 3080 }, { "grad_norm": 0.35881665349006653, "learning_rate": 9.704683826925149e-05, "loss": 0.0305, "step": 3090 }, { "grad_norm": 0.25591030716896057, "learning_rate": 9.701878218556129e-05, "loss": 0.0314, "step": 3100 }, { "grad_norm": 0.3215806782245636, "learning_rate": 9.699059755431598e-05, "loss": 0.029, "step": 3110 }, { "grad_norm": 0.19507969915866852, "learning_rate": 9.696228445257132e-05, "loss": 0.0283, "step": 3120 }, { "grad_norm": 0.20025496184825897, "learning_rate": 9.693384295773419e-05, "loss": 0.0302, "step": 3130 }, { "grad_norm": 0.16887402534484863, "learning_rate": 9.690527314756259e-05, "loss": 0.0288, "step": 3140 }, { "grad_norm": 0.2262166291475296, "learning_rate": 9.687657510016527e-05, "loss": 0.0284, "step": 3150 }, { "grad_norm": 0.27111294865608215, "learning_rate": 9.684774889400161e-05, "loss": 0.0266, "step": 3160 }, { "grad_norm": 0.23330195248126984, "learning_rate": 9.681879460788135e-05, "loss": 0.0281, "step": 3170 }, { "grad_norm": 0.2794685661792755, "learning_rate": 9.67897123209644e-05, "loss": 0.0278, "step": 3180 }, { "grad_norm": 0.16981355845928192, "learning_rate": 9.676050211276062e-05, "loss": 0.0291, "step": 3190 }, { "grad_norm": 0.2393995076417923, "learning_rate": 9.673116406312962e-05, "loss": 0.0298, "step": 3200 }, { "grad_norm": 0.23942279815673828, "learning_rate": 9.67016982522805e-05, "loss": 0.0292, "step": 3210 }, { "grad_norm": 0.23963429033756256, "learning_rate": 9.667210476077164e-05, "loss": 0.0255, "step": 3220 }, { "grad_norm": 0.2146851271390915, "learning_rate": 9.664238366951055e-05, "loss": 0.024, "step": 3230 }, { "grad_norm": 0.1981685906648636, "learning_rate": 9.661253505975355e-05, "loss": 0.0281, "step": 3240 }, { "grad_norm": 0.25613826513290405, "learning_rate": 9.658255901310557e-05, "loss": 0.0254, "step": 3250 }, { "grad_norm": 0.26583316922187805, "learning_rate": 9.655245561152e-05, "loss": 0.0298, "step": 3260 }, { "grad_norm": 0.20958495140075684, "learning_rate": 9.65222249372984e-05, "loss": 0.0284, "step": 3270 }, { "grad_norm": 0.24906514585018158, "learning_rate": 9.649186707309026e-05, "loss": 0.0285, "step": 3280 }, { "grad_norm": 0.202833354473114, "learning_rate": 9.646138210189283e-05, "loss": 0.035, "step": 3290 }, { "grad_norm": 0.20827871561050415, "learning_rate": 9.643077010705087e-05, "loss": 0.0266, "step": 3300 }, { "grad_norm": 0.18580417335033417, "learning_rate": 9.640003117225637e-05, "loss": 0.0257, "step": 3310 }, { "grad_norm": 0.29011547565460205, "learning_rate": 9.636916538154846e-05, "loss": 0.0283, "step": 3320 }, { "grad_norm": 0.20532703399658203, "learning_rate": 9.633817281931296e-05, "loss": 0.0242, "step": 3330 }, { "grad_norm": 0.19920717179775238, "learning_rate": 9.630705357028242e-05, "loss": 0.0281, "step": 3340 }, { "grad_norm": 0.31039100885391235, "learning_rate": 9.627580771953563e-05, "loss": 0.0273, "step": 3350 }, { "grad_norm": 0.2606276869773865, "learning_rate": 9.624443535249759e-05, "loss": 0.0278, "step": 3360 }, { "grad_norm": 0.2741427719593048, "learning_rate": 9.621293655493913e-05, "loss": 0.0302, "step": 3370 }, { "grad_norm": 0.24728700518608093, "learning_rate": 9.618131141297675e-05, "loss": 0.0251, "step": 3380 }, { "grad_norm": 0.3269878029823303, "learning_rate": 9.614956001307242e-05, "loss": 0.0241, "step": 3390 }, { "grad_norm": 0.25811758637428284, "learning_rate": 9.611768244203321e-05, "loss": 0.0309, "step": 3400 }, { "grad_norm": 0.22738142311573029, "learning_rate": 9.60856787870112e-05, "loss": 0.0239, "step": 3410 }, { "grad_norm": 0.1896146535873413, "learning_rate": 9.605354913550318e-05, "loss": 0.0245, "step": 3420 }, { "grad_norm": 0.17059610784053802, "learning_rate": 9.602129357535037e-05, "loss": 0.0256, "step": 3430 }, { "grad_norm": 0.24232615530490875, "learning_rate": 9.598891219473825e-05, "loss": 0.0306, "step": 3440 }, { "grad_norm": 0.2908090353012085, "learning_rate": 9.595640508219625e-05, "loss": 0.0284, "step": 3450 }, { "grad_norm": 0.19091495871543884, "learning_rate": 9.592377232659761e-05, "loss": 0.0302, "step": 3460 }, { "grad_norm": 0.16237559914588928, "learning_rate": 9.589101401715904e-05, "loss": 0.0224, "step": 3470 }, { "grad_norm": 0.31245124340057373, "learning_rate": 9.585813024344045e-05, "loss": 0.028, "step": 3480 }, { "grad_norm": 0.18439354002475739, "learning_rate": 9.58251210953449e-05, "loss": 0.0252, "step": 3490 }, { "grad_norm": 0.19568714499473572, "learning_rate": 9.579198666311809e-05, "loss": 0.0252, "step": 3500 }, { "grad_norm": 0.14211125671863556, "learning_rate": 9.575872703734832e-05, "loss": 0.0241, "step": 3510 }, { "grad_norm": 0.21697285771369934, "learning_rate": 9.572534230896611e-05, "loss": 0.0279, "step": 3520 }, { "grad_norm": 0.2325267493724823, "learning_rate": 9.569183256924403e-05, "loss": 0.0277, "step": 3530 }, { "grad_norm": 0.19593337178230286, "learning_rate": 9.565819790979646e-05, "loss": 0.0257, "step": 3540 }, { "grad_norm": 0.21935831010341644, "learning_rate": 9.562443842257925e-05, "loss": 0.0257, "step": 3550 }, { "grad_norm": 0.2298797369003296, "learning_rate": 9.559055419988956e-05, "loss": 0.0267, "step": 3560 }, { "grad_norm": 0.23667609691619873, "learning_rate": 9.555654533436557e-05, "loss": 0.026, "step": 3570 }, { "grad_norm": 0.2320098876953125, "learning_rate": 9.552241191898621e-05, "loss": 0.0278, "step": 3580 }, { "grad_norm": 0.14753927290439606, "learning_rate": 9.548815404707092e-05, "loss": 0.0253, "step": 3590 }, { "grad_norm": 0.2297377735376358, "learning_rate": 9.545377181227942e-05, "loss": 0.0266, "step": 3600 }, { "grad_norm": 0.18869554996490479, "learning_rate": 9.541926530861145e-05, "loss": 0.0256, "step": 3610 }, { "grad_norm": 0.2531162202358246, "learning_rate": 9.538463463040645e-05, "loss": 0.0247, "step": 3620 }, { "grad_norm": 0.2702654004096985, "learning_rate": 9.534987987234337e-05, "loss": 0.0251, "step": 3630 }, { "grad_norm": 0.32327935099601746, "learning_rate": 9.53150011294404e-05, "loss": 0.0256, "step": 3640 }, { "grad_norm": 0.2351316213607788, "learning_rate": 9.527999849705471e-05, "loss": 0.0261, "step": 3650 }, { "grad_norm": 0.17216554284095764, "learning_rate": 9.524487207088213e-05, "loss": 0.0272, "step": 3660 }, { "grad_norm": 0.29451456665992737, "learning_rate": 9.520962194695698e-05, "loss": 0.027, "step": 3670 }, { "grad_norm": 0.2458239048719406, "learning_rate": 9.517424822165175e-05, "loss": 0.0245, "step": 3680 }, { "grad_norm": 0.17441284656524658, "learning_rate": 9.513875099167685e-05, "loss": 0.0254, "step": 3690 }, { "grad_norm": 0.19156545400619507, "learning_rate": 9.510313035408035e-05, "loss": 0.0267, "step": 3700 }, { "grad_norm": 0.17003275454044342, "learning_rate": 9.506738640624775e-05, "loss": 0.0259, "step": 3710 }, { "grad_norm": 0.1798630803823471, "learning_rate": 9.50315192459016e-05, "loss": 0.0262, "step": 3720 }, { "grad_norm": 0.2290538251399994, "learning_rate": 9.499552897110136e-05, "loss": 0.0257, "step": 3730 }, { "grad_norm": 0.23210687935352325, "learning_rate": 9.495941568024304e-05, "loss": 0.0246, "step": 3740 }, { "grad_norm": 0.24333783984184265, "learning_rate": 9.492317947205904e-05, "loss": 0.0265, "step": 3750 }, { "grad_norm": 0.24058283865451813, "learning_rate": 9.488682044561775e-05, "loss": 0.0256, "step": 3760 }, { "grad_norm": 0.26348140835762024, "learning_rate": 9.485033870032335e-05, "loss": 0.0271, "step": 3770 }, { "grad_norm": 0.2988092601299286, "learning_rate": 9.481373433591556e-05, "loss": 0.0245, "step": 3780 }, { "grad_norm": 0.2602173089981079, "learning_rate": 9.47770074524693e-05, "loss": 0.0287, "step": 3790 }, { "grad_norm": 0.2338210642337799, "learning_rate": 9.474015815039446e-05, "loss": 0.0266, "step": 3800 }, { "grad_norm": 0.3367757201194763, "learning_rate": 9.470318653043565e-05, "loss": 0.0237, "step": 3810 }, { "grad_norm": 0.24316680431365967, "learning_rate": 9.466609269367185e-05, "loss": 0.0268, "step": 3820 }, { "grad_norm": 0.2237861603498459, "learning_rate": 9.46288767415162e-05, "loss": 0.027, "step": 3830 }, { "grad_norm": 0.25782522559165955, "learning_rate": 9.459153877571567e-05, "loss": 0.0241, "step": 3840 }, { "grad_norm": 0.20009446144104004, "learning_rate": 9.455407889835087e-05, "loss": 0.0247, "step": 3850 }, { "grad_norm": 0.2721019387245178, "learning_rate": 9.451649721183564e-05, "loss": 0.0252, "step": 3860 }, { "grad_norm": 0.16587068140506744, "learning_rate": 9.447879381891692e-05, "loss": 0.0272, "step": 3870 }, { "grad_norm": 0.2891231179237366, "learning_rate": 9.444096882267428e-05, "loss": 0.0241, "step": 3880 }, { "grad_norm": 0.18716834485530853, "learning_rate": 9.440302232651988e-05, "loss": 0.0241, "step": 3890 }, { "grad_norm": 0.20265820622444153, "learning_rate": 9.436495443419795e-05, "loss": 0.0238, "step": 3900 }, { "grad_norm": 0.2484164834022522, "learning_rate": 9.432676524978466e-05, "loss": 0.0281, "step": 3910 }, { "grad_norm": 0.2169998735189438, "learning_rate": 9.42884548776878e-05, "loss": 0.0297, "step": 3920 }, { "grad_norm": 0.20671480894088745, "learning_rate": 9.425002342264646e-05, "loss": 0.0227, "step": 3930 }, { "grad_norm": 0.32488006353378296, "learning_rate": 9.421147098973077e-05, "loss": 0.0257, "step": 3940 }, { "grad_norm": 0.1924733817577362, "learning_rate": 9.41727976843416e-05, "loss": 0.0237, "step": 3950 }, { "grad_norm": 0.2231021523475647, "learning_rate": 9.413400361221029e-05, "loss": 0.0282, "step": 3960 }, { "grad_norm": 0.19151033461093903, "learning_rate": 9.409508887939835e-05, "loss": 0.0258, "step": 3970 }, { "grad_norm": 0.27332618832588196, "learning_rate": 9.40560535922972e-05, "loss": 0.0306, "step": 3980 }, { "grad_norm": 0.26059091091156006, "learning_rate": 9.40168978576278e-05, "loss": 0.0292, "step": 3990 }, { "grad_norm": 0.1953822821378708, "learning_rate": 9.397762178244043e-05, "loss": 0.0273, "step": 4000 }, { "grad_norm": 0.23925530910491943, "learning_rate": 9.393822547411439e-05, "loss": 0.0279, "step": 4010 }, { "grad_norm": 0.1843254417181015, "learning_rate": 9.389870904035769e-05, "loss": 0.026, "step": 4020 }, { "grad_norm": 0.16704784333705902, "learning_rate": 9.385907258920672e-05, "loss": 0.0238, "step": 4030 }, { "grad_norm": 0.20959529280662537, "learning_rate": 9.381931622902607e-05, "loss": 0.0236, "step": 4040 }, { "grad_norm": 0.1805470883846283, "learning_rate": 9.377944006850807e-05, "loss": 0.0223, "step": 4050 }, { "grad_norm": 0.18840813636779785, "learning_rate": 9.373944421667265e-05, "loss": 0.0248, "step": 4060 }, { "grad_norm": 0.19421975314617157, "learning_rate": 9.369932878286691e-05, "loss": 0.0244, "step": 4070 }, { "grad_norm": 0.19593873620033264, "learning_rate": 9.365909387676494e-05, "loss": 0.0231, "step": 4080 }, { "grad_norm": 0.16126561164855957, "learning_rate": 9.361873960836744e-05, "loss": 0.0267, "step": 4090 }, { "grad_norm": 0.26318052411079407, "learning_rate": 9.357826608800142e-05, "loss": 0.0257, "step": 4100 }, { "grad_norm": 0.2671802043914795, "learning_rate": 9.353767342631994e-05, "loss": 0.0259, "step": 4110 }, { "grad_norm": 0.25529468059539795, "learning_rate": 9.34969617343018e-05, "loss": 0.0241, "step": 4120 }, { "grad_norm": 0.2747957706451416, "learning_rate": 9.345613112325122e-05, "loss": 0.0221, "step": 4130 }, { "grad_norm": 0.19367505609989166, "learning_rate": 9.34151817047975e-05, "loss": 0.0293, "step": 4140 }, { "grad_norm": 0.2029552310705185, "learning_rate": 9.33741135908948e-05, "loss": 0.0228, "step": 4150 }, { "grad_norm": 0.17393606901168823, "learning_rate": 9.33329268938218e-05, "loss": 0.0277, "step": 4160 }, { "grad_norm": 0.19113396108150482, "learning_rate": 9.329162172618132e-05, "loss": 0.0232, "step": 4170 }, { "grad_norm": 0.2463500052690506, "learning_rate": 9.325019820090013e-05, "loss": 0.0257, "step": 4180 }, { "grad_norm": 0.21629436314105988, "learning_rate": 9.320865643122855e-05, "loss": 0.0243, "step": 4190 }, { "grad_norm": 0.1863965392112732, "learning_rate": 9.316699653074023e-05, "loss": 0.0239, "step": 4200 }, { "grad_norm": 0.17694352567195892, "learning_rate": 9.312521861333172e-05, "loss": 0.0234, "step": 4210 }, { "grad_norm": 0.1377595216035843, "learning_rate": 9.308332279322224e-05, "loss": 0.0211, "step": 4220 }, { "grad_norm": 0.17864219844341278, "learning_rate": 9.304130918495338e-05, "loss": 0.0196, "step": 4230 }, { "grad_norm": 0.2446253001689911, "learning_rate": 9.299917790338874e-05, "loss": 0.0257, "step": 4240 }, { "grad_norm": 0.21004924178123474, "learning_rate": 9.295692906371363e-05, "loss": 0.023, "step": 4250 }, { "grad_norm": 0.1981293261051178, "learning_rate": 9.291456278143476e-05, "loss": 0.0231, "step": 4260 }, { "grad_norm": 0.24557314813137054, "learning_rate": 9.287207917237994e-05, "loss": 0.0208, "step": 4270 }, { "grad_norm": 0.21233336627483368, "learning_rate": 9.282947835269773e-05, "loss": 0.0231, "step": 4280 }, { "grad_norm": 0.21565674245357513, "learning_rate": 9.278676043885715e-05, "loss": 0.0238, "step": 4290 }, { "grad_norm": 0.17830128967761993, "learning_rate": 9.274392554764733e-05, "loss": 0.0244, "step": 4300 }, { "grad_norm": 0.14914162456989288, "learning_rate": 9.270097379617723e-05, "loss": 0.0253, "step": 4310 }, { "grad_norm": 0.19047164916992188, "learning_rate": 9.26579053018753e-05, "loss": 0.0225, "step": 4320 }, { "grad_norm": 0.2971799969673157, "learning_rate": 9.261472018248918e-05, "loss": 0.0258, "step": 4330 }, { "grad_norm": 0.3201267123222351, "learning_rate": 9.25714185560853e-05, "loss": 0.0272, "step": 4340 }, { "grad_norm": 0.18034052848815918, "learning_rate": 9.252800054104868e-05, "loss": 0.0239, "step": 4350 }, { "grad_norm": 0.15651677548885345, "learning_rate": 9.248446625608252e-05, "loss": 0.0219, "step": 4360 }, { "grad_norm": 0.1764494925737381, "learning_rate": 9.244081582020789e-05, "loss": 0.0259, "step": 4370 }, { "grad_norm": 0.23399971425533295, "learning_rate": 9.239704935276339e-05, "loss": 0.026, "step": 4380 }, { "grad_norm": 0.18775106966495514, "learning_rate": 9.235316697340489e-05, "loss": 0.024, "step": 4390 }, { "grad_norm": 0.18042977154254913, "learning_rate": 9.230916880210512e-05, "loss": 0.0265, "step": 4400 }, { "grad_norm": 0.19147340953350067, "learning_rate": 9.226505495915342e-05, "loss": 0.0284, "step": 4410 }, { "grad_norm": 0.1806974858045578, "learning_rate": 9.222082556515536e-05, "loss": 0.0217, "step": 4420 }, { "grad_norm": 0.23061969876289368, "learning_rate": 9.217648074103242e-05, "loss": 0.0269, "step": 4430 }, { "grad_norm": 0.20452506840229034, "learning_rate": 9.213202060802161e-05, "loss": 0.0226, "step": 4440 }, { "grad_norm": 0.18311041593551636, "learning_rate": 9.208744528767528e-05, "loss": 0.0242, "step": 4450 }, { "grad_norm": 0.1944350153207779, "learning_rate": 9.204275490186064e-05, "loss": 0.0205, "step": 4460 }, { "grad_norm": 0.15766341984272003, "learning_rate": 9.199794957275949e-05, "loss": 0.0221, "step": 4470 }, { "grad_norm": 0.17332236468791962, "learning_rate": 9.19530294228679e-05, "loss": 0.0253, "step": 4480 }, { "grad_norm": 0.2257440984249115, "learning_rate": 9.190799457499583e-05, "loss": 0.0231, "step": 4490 }, { "grad_norm": 0.22677402198314667, "learning_rate": 9.186284515226686e-05, "loss": 0.0257, "step": 4500 }, { "grad_norm": 0.15556176006793976, "learning_rate": 9.181758127811777e-05, "loss": 0.02, "step": 4510 }, { "grad_norm": 0.14155864715576172, "learning_rate": 9.177220307629825e-05, "loss": 0.0215, "step": 4520 }, { "grad_norm": 0.19456623494625092, "learning_rate": 9.172671067087059e-05, "loss": 0.0238, "step": 4530 }, { "grad_norm": 0.18800154328346252, "learning_rate": 9.16811041862093e-05, "loss": 0.0215, "step": 4540 }, { "grad_norm": 0.21699370443820953, "learning_rate": 9.163538374700076e-05, "loss": 0.0238, "step": 4550 }, { "grad_norm": 0.1724860817193985, "learning_rate": 9.158954947824287e-05, "loss": 0.0245, "step": 4560 }, { "grad_norm": 0.19941389560699463, "learning_rate": 9.154360150524482e-05, "loss": 0.0202, "step": 4570 }, { "grad_norm": 0.1992763876914978, "learning_rate": 9.14975399536266e-05, "loss": 0.027, "step": 4580 }, { "grad_norm": 0.2802664637565613, "learning_rate": 9.14513649493187e-05, "loss": 0.027, "step": 4590 }, { "grad_norm": 0.23534627258777618, "learning_rate": 9.140507661856187e-05, "loss": 0.0234, "step": 4600 }, { "grad_norm": 0.26335158944129944, "learning_rate": 9.135867508790661e-05, "loss": 0.024, "step": 4610 }, { "grad_norm": 0.19839201867580414, "learning_rate": 9.131216048421291e-05, "loss": 0.024, "step": 4620 }, { "grad_norm": 0.19579410552978516, "learning_rate": 9.126553293464998e-05, "loss": 0.0266, "step": 4630 }, { "grad_norm": 0.22774294018745422, "learning_rate": 9.121879256669572e-05, "loss": 0.0228, "step": 4640 }, { "grad_norm": 0.22197744250297546, "learning_rate": 9.117193950813652e-05, "loss": 0.0237, "step": 4650 }, { "grad_norm": 0.18403634428977966, "learning_rate": 9.112497388706685e-05, "loss": 0.0239, "step": 4660 }, { "grad_norm": 0.24239495396614075, "learning_rate": 9.10778958318889e-05, "loss": 0.0262, "step": 4670 }, { "grad_norm": 0.20773662626743317, "learning_rate": 9.103070547131232e-05, "loss": 0.0236, "step": 4680 }, { "grad_norm": 0.2796818017959595, "learning_rate": 9.098340293435375e-05, "loss": 0.0281, "step": 4690 }, { "grad_norm": 0.25900718569755554, "learning_rate": 9.093598835033649e-05, "loss": 0.0233, "step": 4700 }, { "grad_norm": 0.21334485709667206, "learning_rate": 9.088846184889021e-05, "loss": 0.0225, "step": 4710 }, { "grad_norm": 0.20420046150684357, "learning_rate": 9.084082355995057e-05, "loss": 0.0241, "step": 4720 }, { "grad_norm": 0.2521681487560272, "learning_rate": 9.079307361375882e-05, "loss": 0.0231, "step": 4730 }, { "grad_norm": 0.19501960277557373, "learning_rate": 9.074521214086149e-05, "loss": 0.0226, "step": 4740 }, { "grad_norm": 0.1581960916519165, "learning_rate": 9.069723927211001e-05, "loss": 0.0269, "step": 4750 }, { "grad_norm": 0.15163291990756989, "learning_rate": 9.064915513866037e-05, "loss": 0.0257, "step": 4760 }, { "grad_norm": 0.15281733870506287, "learning_rate": 9.060095987197279e-05, "loss": 0.0247, "step": 4770 }, { "grad_norm": 0.17880699038505554, "learning_rate": 9.055265360381126e-05, "loss": 0.0248, "step": 4780 }, { "grad_norm": 0.21005196869373322, "learning_rate": 9.050423646624326e-05, "loss": 0.0263, "step": 4790 }, { "grad_norm": 0.22608084976673126, "learning_rate": 9.045570859163943e-05, "loss": 0.0229, "step": 4800 }, { "grad_norm": 0.19033865630626678, "learning_rate": 9.04070701126731e-05, "loss": 0.0279, "step": 4810 }, { "grad_norm": 0.16103699803352356, "learning_rate": 9.035832116232001e-05, "loss": 0.0232, "step": 4820 }, { "grad_norm": 0.2715504467487335, "learning_rate": 9.030946187385796e-05, "loss": 0.0237, "step": 4830 }, { "grad_norm": 0.20106668770313263, "learning_rate": 9.026049238086635e-05, "loss": 0.0247, "step": 4840 }, { "grad_norm": 0.19311027228832245, "learning_rate": 9.021141281722591e-05, "loss": 0.0202, "step": 4850 }, { "grad_norm": 0.24935133755207062, "learning_rate": 9.01622233171183e-05, "loss": 0.0263, "step": 4860 }, { "grad_norm": 0.19176337122917175, "learning_rate": 9.011292401502574e-05, "loss": 0.0239, "step": 4870 }, { "grad_norm": 0.25212374329566956, "learning_rate": 9.006351504573063e-05, "loss": 0.0231, "step": 4880 }, { "grad_norm": 0.25029537081718445, "learning_rate": 9.001399654431519e-05, "loss": 0.0228, "step": 4890 }, { "grad_norm": 0.17633725702762604, "learning_rate": 8.996436864616116e-05, "loss": 0.0223, "step": 4900 }, { "grad_norm": 0.15019038319587708, "learning_rate": 8.991463148694925e-05, "loss": 0.0216, "step": 4910 }, { "grad_norm": 0.20572742819786072, "learning_rate": 8.986478520265902e-05, "loss": 0.0248, "step": 4920 }, { "grad_norm": 0.21906854212284088, "learning_rate": 8.981482992956827e-05, "loss": 0.0257, "step": 4930 }, { "grad_norm": 0.18399231135845184, "learning_rate": 8.976476580425282e-05, "loss": 0.0222, "step": 4940 }, { "grad_norm": 0.20738139748573303, "learning_rate": 8.971459296358606e-05, "loss": 0.0259, "step": 4950 }, { "grad_norm": 0.1637602001428604, "learning_rate": 8.966431154473864e-05, "loss": 0.025, "step": 4960 }, { "grad_norm": 0.23832815885543823, "learning_rate": 8.961392168517803e-05, "loss": 0.0238, "step": 4970 }, { "grad_norm": 0.29721808433532715, "learning_rate": 8.956342352266821e-05, "loss": 0.0248, "step": 4980 }, { "grad_norm": 0.1656726598739624, "learning_rate": 8.95128171952692e-05, "loss": 0.0227, "step": 4990 }, { "grad_norm": 0.15054400265216827, "learning_rate": 8.946210284133676e-05, "loss": 0.0245, "step": 5000 }, { "grad_norm": 0.16954728960990906, "learning_rate": 8.941128059952201e-05, "loss": 0.0228, "step": 5010 }, { "grad_norm": 0.16612163186073303, "learning_rate": 8.936035060877102e-05, "loss": 0.0242, "step": 5020 }, { "grad_norm": 0.21777351200580597, "learning_rate": 8.930931300832443e-05, "loss": 0.0243, "step": 5030 }, { "grad_norm": 0.20673853158950806, "learning_rate": 8.925816793771711e-05, "loss": 0.0231, "step": 5040 }, { "grad_norm": 0.2513861656188965, "learning_rate": 8.92069155367777e-05, "loss": 0.0225, "step": 5050 }, { "grad_norm": 0.22552914917469025, "learning_rate": 8.915555594562834e-05, "loss": 0.0239, "step": 5060 }, { "grad_norm": 0.1582045555114746, "learning_rate": 8.910408930468416e-05, "loss": 0.0207, "step": 5070 }, { "grad_norm": 0.1962750107049942, "learning_rate": 8.905251575465303e-05, "loss": 0.0211, "step": 5080 }, { "grad_norm": 0.15863190591335297, "learning_rate": 8.900083543653502e-05, "loss": 0.0219, "step": 5090 }, { "grad_norm": 0.25894346833229065, "learning_rate": 8.894904849162218e-05, "loss": 0.0242, "step": 5100 }, { "grad_norm": 0.2870851457118988, "learning_rate": 8.889715506149802e-05, "loss": 0.0253, "step": 5110 }, { "grad_norm": 0.290376216173172, "learning_rate": 8.884515528803722e-05, "loss": 0.0217, "step": 5120 }, { "grad_norm": 0.27522915601730347, "learning_rate": 8.879304931340517e-05, "loss": 0.0209, "step": 5130 }, { "grad_norm": 0.23377522826194763, "learning_rate": 8.874083728005759e-05, "loss": 0.0248, "step": 5140 }, { "grad_norm": 0.22575202584266663, "learning_rate": 8.868851933074021e-05, "loss": 0.024, "step": 5150 }, { "grad_norm": 0.18068405985832214, "learning_rate": 8.863609560848829e-05, "loss": 0.0236, "step": 5160 }, { "grad_norm": 0.27400001883506775, "learning_rate": 8.85835662566263e-05, "loss": 0.0195, "step": 5170 }, { "grad_norm": 0.14330556988716125, "learning_rate": 8.853093141876747e-05, "loss": 0.0192, "step": 5180 }, { "grad_norm": 0.23288457095623016, "learning_rate": 8.847819123881343e-05, "loss": 0.0233, "step": 5190 }, { "grad_norm": 0.15297475457191467, "learning_rate": 8.842534586095383e-05, "loss": 0.0215, "step": 5200 }, { "grad_norm": 0.19600942730903625, "learning_rate": 8.837239542966593e-05, "loss": 0.0228, "step": 5210 }, { "grad_norm": 0.15891338884830475, "learning_rate": 8.831934008971417e-05, "loss": 0.0219, "step": 5220 }, { "grad_norm": 0.1806361824274063, "learning_rate": 8.826617998614982e-05, "loss": 0.0205, "step": 5230 }, { "grad_norm": 0.1526317149400711, "learning_rate": 8.821291526431056e-05, "loss": 0.0225, "step": 5240 }, { "grad_norm": 0.13692589104175568, "learning_rate": 8.815954606982015e-05, "loss": 0.0207, "step": 5250 }, { "grad_norm": 0.23868529498577118, "learning_rate": 8.810607254858789e-05, "loss": 0.0246, "step": 5260 }, { "grad_norm": 0.16436894237995148, "learning_rate": 8.805249484680838e-05, "loss": 0.0222, "step": 5270 }, { "grad_norm": 0.1534588783979416, "learning_rate": 8.799881311096096e-05, "loss": 0.0204, "step": 5280 }, { "grad_norm": 0.2169351726770401, "learning_rate": 8.794502748780949e-05, "loss": 0.0194, "step": 5290 }, { "grad_norm": 0.2143266350030899, "learning_rate": 8.78911381244018e-05, "loss": 0.021, "step": 5300 }, { "grad_norm": 0.27708718180656433, "learning_rate": 8.783714516806933e-05, "loss": 0.0204, "step": 5310 }, { "grad_norm": 0.1811063289642334, "learning_rate": 8.77830487664268e-05, "loss": 0.0212, "step": 5320 }, { "grad_norm": 0.1971118003129959, "learning_rate": 8.772884906737167e-05, "loss": 0.02, "step": 5330 }, { "grad_norm": 0.21687600016593933, "learning_rate": 8.767454621908387e-05, "loss": 0.0225, "step": 5340 }, { "grad_norm": 0.17817649245262146, "learning_rate": 8.76201403700253e-05, "loss": 0.019, "step": 5350 }, { "grad_norm": 0.1756051480770111, "learning_rate": 8.756563166893949e-05, "loss": 0.0211, "step": 5360 }, { "grad_norm": 0.23976975679397583, "learning_rate": 8.751102026485113e-05, "loss": 0.0213, "step": 5370 }, { "grad_norm": 0.24947892129421234, "learning_rate": 8.745630630706571e-05, "loss": 0.0197, "step": 5380 }, { "grad_norm": 0.24512451887130737, "learning_rate": 8.740148994516912e-05, "loss": 0.0197, "step": 5390 }, { "grad_norm": 0.2781716287136078, "learning_rate": 8.73465713290272e-05, "loss": 0.0228, "step": 5400 }, { "grad_norm": 0.1621004194021225, "learning_rate": 8.729155060878533e-05, "loss": 0.0221, "step": 5410 }, { "grad_norm": 0.19999860227108002, "learning_rate": 8.723642793486809e-05, "loss": 0.0221, "step": 5420 }, { "grad_norm": 0.15398412942886353, "learning_rate": 8.718120345797873e-05, "loss": 0.0208, "step": 5430 }, { "grad_norm": 0.18921099603176117, "learning_rate": 8.712587732909889e-05, "loss": 0.0198, "step": 5440 }, { "grad_norm": 0.12318053841590881, "learning_rate": 8.707044969948806e-05, "loss": 0.0197, "step": 5450 }, { "grad_norm": 0.16622845828533173, "learning_rate": 8.701492072068329e-05, "loss": 0.0181, "step": 5460 }, { "grad_norm": 0.2303917109966278, "learning_rate": 8.695929054449869e-05, "loss": 0.0204, "step": 5470 }, { "grad_norm": 0.24900710582733154, "learning_rate": 8.690355932302501e-05, "loss": 0.0215, "step": 5480 }, { "grad_norm": 0.15708234906196594, "learning_rate": 8.684772720862931e-05, "loss": 0.0205, "step": 5490 }, { "grad_norm": 0.15909716486930847, "learning_rate": 8.679179435395446e-05, "loss": 0.0247, "step": 5500 }, { "grad_norm": 0.19485701620578766, "learning_rate": 8.673576091191874e-05, "loss": 0.0206, "step": 5510 }, { "grad_norm": 0.12717552483081818, "learning_rate": 8.667962703571541e-05, "loss": 0.0215, "step": 5520 }, { "grad_norm": 0.1715942621231079, "learning_rate": 8.662339287881238e-05, "loss": 0.0204, "step": 5530 }, { "grad_norm": 0.1936575174331665, "learning_rate": 8.656705859495169e-05, "loss": 0.0225, "step": 5540 }, { "grad_norm": 0.31662046909332275, "learning_rate": 8.651062433814912e-05, "loss": 0.024, "step": 5550 }, { "grad_norm": 0.22982411086559296, "learning_rate": 8.645409026269375e-05, "loss": 0.0252, "step": 5560 }, { "grad_norm": 0.23461347818374634, "learning_rate": 8.639745652314759e-05, "loss": 0.025, "step": 5570 }, { "grad_norm": 0.17542695999145508, "learning_rate": 8.634072327434515e-05, "loss": 0.0272, "step": 5580 }, { "grad_norm": 0.1950596570968628, "learning_rate": 8.628389067139294e-05, "loss": 0.022, "step": 5590 }, { "grad_norm": 0.14540629088878632, "learning_rate": 8.622695886966911e-05, "loss": 0.0259, "step": 5600 }, { "grad_norm": 0.21625404059886932, "learning_rate": 8.616992802482308e-05, "loss": 0.0237, "step": 5610 }, { "grad_norm": 0.2495943009853363, "learning_rate": 8.611279829277496e-05, "loss": 0.0228, "step": 5620 }, { "grad_norm": 0.20330318808555603, "learning_rate": 8.605556982971528e-05, "loss": 0.0235, "step": 5630 }, { "grad_norm": 0.2065277397632599, "learning_rate": 8.599824279210447e-05, "loss": 0.0193, "step": 5640 }, { "grad_norm": 0.1795617640018463, "learning_rate": 8.594081733667243e-05, "loss": 0.0218, "step": 5650 }, { "grad_norm": 0.13556398451328278, "learning_rate": 8.58832936204182e-05, "loss": 0.0189, "step": 5660 }, { "grad_norm": 0.14024759829044342, "learning_rate": 8.582567180060942e-05, "loss": 0.0241, "step": 5670 }, { "grad_norm": 0.19141484797000885, "learning_rate": 8.576795203478194e-05, "loss": 0.0197, "step": 5680 }, { "grad_norm": 0.21098577976226807, "learning_rate": 8.571013448073939e-05, "loss": 0.0236, "step": 5690 }, { "grad_norm": 0.12799809873104095, "learning_rate": 8.565221929655275e-05, "loss": 0.019, "step": 5700 }, { "grad_norm": 0.18170933425426483, "learning_rate": 8.559420664055992e-05, "loss": 0.0221, "step": 5710 }, { "grad_norm": 0.28051143884658813, "learning_rate": 8.553609667136532e-05, "loss": 0.021, "step": 5720 }, { "grad_norm": 0.1817805916070938, "learning_rate": 8.547788954783936e-05, "loss": 0.0258, "step": 5730 }, { "grad_norm": 0.2778211534023285, "learning_rate": 8.541958542911808e-05, "loss": 0.0232, "step": 5740 }, { "grad_norm": 0.18251830339431763, "learning_rate": 8.536118447460275e-05, "loss": 0.0204, "step": 5750 }, { "grad_norm": 0.16428755223751068, "learning_rate": 8.530268684395932e-05, "loss": 0.0198, "step": 5760 }, { "grad_norm": 0.15734358131885529, "learning_rate": 8.524409269711807e-05, "loss": 0.0207, "step": 5770 }, { "grad_norm": 0.16812390089035034, "learning_rate": 8.51854021942732e-05, "loss": 0.0213, "step": 5780 }, { "grad_norm": 0.21087315678596497, "learning_rate": 8.512661549588227e-05, "loss": 0.0184, "step": 5790 }, { "grad_norm": 0.1807655692100525, "learning_rate": 8.506773276266588e-05, "loss": 0.0197, "step": 5800 }, { "grad_norm": 0.145188108086586, "learning_rate": 8.500875415560721e-05, "loss": 0.0203, "step": 5810 }, { "grad_norm": 0.14397341012954712, "learning_rate": 8.494967983595144e-05, "loss": 0.0206, "step": 5820 }, { "grad_norm": 0.13907107710838318, "learning_rate": 8.489050996520558e-05, "loss": 0.0176, "step": 5830 }, { "grad_norm": 0.1721039116382599, "learning_rate": 8.483124470513775e-05, "loss": 0.0204, "step": 5840 }, { "grad_norm": 0.16737200319766998, "learning_rate": 8.477188421777692e-05, "loss": 0.0224, "step": 5850 }, { "grad_norm": 0.16352951526641846, "learning_rate": 8.47124286654124e-05, "loss": 0.0226, "step": 5860 }, { "grad_norm": 0.12920720875263214, "learning_rate": 8.465287821059341e-05, "loss": 0.0187, "step": 5870 }, { "grad_norm": 0.19436359405517578, "learning_rate": 8.45932330161286e-05, "loss": 0.0217, "step": 5880 }, { "grad_norm": 0.17122848331928253, "learning_rate": 8.453349324508567e-05, "loss": 0.017, "step": 5890 }, { "grad_norm": 0.12285362929105759, "learning_rate": 8.447365906079088e-05, "loss": 0.0195, "step": 5900 }, { "grad_norm": 0.13264913856983185, "learning_rate": 8.441373062682856e-05, "loss": 0.0198, "step": 5910 }, { "grad_norm": 0.12016724050045013, "learning_rate": 8.43537081070408e-05, "loss": 0.0175, "step": 5920 }, { "grad_norm": 0.13568389415740967, "learning_rate": 8.429359166552689e-05, "loss": 0.0213, "step": 5930 }, { "grad_norm": 0.19307930767536163, "learning_rate": 8.423338146664284e-05, "loss": 0.0207, "step": 5940 }, { "grad_norm": 0.16518783569335938, "learning_rate": 8.417307767500107e-05, "loss": 0.0217, "step": 5950 }, { "grad_norm": 0.15711931884288788, "learning_rate": 8.411268045546983e-05, "loss": 0.0212, "step": 5960 }, { "grad_norm": 0.21010319888591766, "learning_rate": 8.405218997317281e-05, "loss": 0.0187, "step": 5970 }, { "grad_norm": 0.1782633364200592, "learning_rate": 8.399160639348869e-05, "loss": 0.0194, "step": 5980 }, { "grad_norm": 0.1871320903301239, "learning_rate": 8.393092988205065e-05, "loss": 0.0204, "step": 5990 }, { "grad_norm": 0.17632342875003815, "learning_rate": 8.387016060474597e-05, "loss": 0.0181, "step": 6000 }, { "grad_norm": 0.1523282676935196, "learning_rate": 8.380929872771551e-05, "loss": 0.0218, "step": 6010 }, { "grad_norm": 0.13645745813846588, "learning_rate": 8.374834441735335e-05, "loss": 0.02, "step": 6020 }, { "grad_norm": 0.18748165667057037, "learning_rate": 8.368729784030622e-05, "loss": 0.0229, "step": 6030 }, { "grad_norm": 0.15729068219661713, "learning_rate": 8.362615916347315e-05, "loss": 0.0185, "step": 6040 }, { "grad_norm": 0.16364344954490662, "learning_rate": 8.356492855400493e-05, "loss": 0.0215, "step": 6050 }, { "grad_norm": 0.22612427175045013, "learning_rate": 8.350360617930371e-05, "loss": 0.0212, "step": 6060 }, { "grad_norm": 0.14270199835300446, "learning_rate": 8.344219220702255e-05, "loss": 0.0191, "step": 6070 }, { "grad_norm": 0.15274250507354736, "learning_rate": 8.338068680506485e-05, "loss": 0.0242, "step": 6080 }, { "grad_norm": 0.22516117990016937, "learning_rate": 8.33190901415841e-05, "loss": 0.0215, "step": 6090 }, { "grad_norm": 0.2206544131040573, "learning_rate": 8.325740238498317e-05, "loss": 0.0212, "step": 6100 }, { "grad_norm": 0.2182512730360031, "learning_rate": 8.319562370391406e-05, "loss": 0.023, "step": 6110 }, { "grad_norm": 0.21175847947597504, "learning_rate": 8.31337542672773e-05, "loss": 0.0215, "step": 6120 }, { "grad_norm": 0.15002600848674774, "learning_rate": 8.307179424422158e-05, "loss": 0.0194, "step": 6130 }, { "grad_norm": 0.1514759659767151, "learning_rate": 8.300974380414327e-05, "loss": 0.0206, "step": 6140 }, { "grad_norm": 0.1608678251504898, "learning_rate": 8.294760311668586e-05, "loss": 0.0172, "step": 6150 }, { "grad_norm": 0.16048648953437805, "learning_rate": 8.288537235173961e-05, "loss": 0.018, "step": 6160 }, { "grad_norm": 0.14986516535282135, "learning_rate": 8.282305167944108e-05, "loss": 0.0176, "step": 6170 }, { "grad_norm": 0.24611257016658783, "learning_rate": 8.276064127017262e-05, "loss": 0.0191, "step": 6180 }, { "grad_norm": 0.30079731345176697, "learning_rate": 8.269814129456189e-05, "loss": 0.0212, "step": 6190 }, { "grad_norm": 0.2160930335521698, "learning_rate": 8.263555192348143e-05, "loss": 0.02, "step": 6200 }, { "grad_norm": 0.18474191427230835, "learning_rate": 8.257287332804819e-05, "loss": 0.0208, "step": 6210 }, { "grad_norm": 0.1945219486951828, "learning_rate": 8.251010567962307e-05, "loss": 0.0201, "step": 6220 }, { "grad_norm": 0.13645923137664795, "learning_rate": 8.244724914981041e-05, "loss": 0.0201, "step": 6230 }, { "grad_norm": 0.1840837150812149, "learning_rate": 8.238430391045757e-05, "loss": 0.0198, "step": 6240 }, { "grad_norm": 0.13867904245853424, "learning_rate": 8.232127013365445e-05, "loss": 0.0175, "step": 6250 }, { "grad_norm": 0.14002598822116852, "learning_rate": 8.225814799173295e-05, "loss": 0.0169, "step": 6260 }, { "grad_norm": 0.22629636526107788, "learning_rate": 8.219493765726663e-05, "loss": 0.0184, "step": 6270 }, { "grad_norm": 0.18000899255275726, "learning_rate": 8.21316393030701e-05, "loss": 0.0182, "step": 6280 }, { "grad_norm": 0.13968732953071594, "learning_rate": 8.206825310219865e-05, "loss": 0.0212, "step": 6290 }, { "grad_norm": 0.1700585037469864, "learning_rate": 8.200477922794776e-05, "loss": 0.0175, "step": 6300 }, { "grad_norm": 0.12798351049423218, "learning_rate": 8.194121785385256e-05, "loss": 0.0172, "step": 6310 }, { "grad_norm": 0.14947959780693054, "learning_rate": 8.187756915368741e-05, "loss": 0.0197, "step": 6320 }, { "grad_norm": 0.16527236998081207, "learning_rate": 8.181383330146544e-05, "loss": 0.0186, "step": 6330 }, { "grad_norm": 0.13126209378242493, "learning_rate": 8.175001047143804e-05, "loss": 0.0204, "step": 6340 }, { "grad_norm": 0.17379534244537354, "learning_rate": 8.168610083809438e-05, "loss": 0.0244, "step": 6350 }, { "grad_norm": 0.197370707988739, "learning_rate": 8.162210457616095e-05, "loss": 0.0193, "step": 6360 }, { "grad_norm": 0.2126130908727646, "learning_rate": 8.155802186060109e-05, "loss": 0.0211, "step": 6370 }, { "grad_norm": 0.1839226484298706, "learning_rate": 8.149385286661453e-05, "loss": 0.0192, "step": 6380 }, { "grad_norm": 0.23453545570373535, "learning_rate": 8.14295977696368e-05, "loss": 0.0209, "step": 6390 }, { "grad_norm": 0.1851550191640854, "learning_rate": 8.13652567453389e-05, "loss": 0.0231, "step": 6400 }, { "grad_norm": 0.16166961193084717, "learning_rate": 8.130082996962676e-05, "loss": 0.0228, "step": 6410 }, { "grad_norm": 0.13346044719219208, "learning_rate": 8.123631761864068e-05, "loss": 0.0206, "step": 6420 }, { "grad_norm": 0.1878715306520462, "learning_rate": 8.1171719868755e-05, "loss": 0.0178, "step": 6430 }, { "grad_norm": 0.20252929627895355, "learning_rate": 8.110703689657748e-05, "loss": 0.0224, "step": 6440 }, { "grad_norm": 0.16666507720947266, "learning_rate": 8.104226887894892e-05, "loss": 0.0178, "step": 6450 }, { "grad_norm": 0.16517062485218048, "learning_rate": 8.097741599294257e-05, "loss": 0.0185, "step": 6460 }, { "grad_norm": 0.13401320576667786, "learning_rate": 8.091247841586378e-05, "loss": 0.0164, "step": 6470 }, { "grad_norm": 0.15741941332817078, "learning_rate": 8.084745632524939e-05, "loss": 0.0183, "step": 6480 }, { "grad_norm": 0.154473677277565, "learning_rate": 8.07823498988673e-05, "loss": 0.02, "step": 6490 }, { "grad_norm": 0.15709365904331207, "learning_rate": 8.071715931471602e-05, "loss": 0.0207, "step": 6500 }, { "grad_norm": 0.25021007657051086, "learning_rate": 8.06518847510241e-05, "loss": 0.024, "step": 6510 }, { "grad_norm": 0.20203040540218353, "learning_rate": 8.058652638624971e-05, "loss": 0.019, "step": 6520 }, { "grad_norm": 0.1349378228187561, "learning_rate": 8.052108439908013e-05, "loss": 0.0212, "step": 6530 }, { "grad_norm": 0.2142387330532074, "learning_rate": 8.045555896843125e-05, "loss": 0.0151, "step": 6540 }, { "grad_norm": 0.18499287962913513, "learning_rate": 8.03899502734471e-05, "loss": 0.0193, "step": 6550 }, { "grad_norm": 0.29113802313804626, "learning_rate": 8.032425849349931e-05, "loss": 0.0191, "step": 6560 }, { "grad_norm": 0.19822870194911957, "learning_rate": 8.025848380818674e-05, "loss": 0.0199, "step": 6570 }, { "grad_norm": 0.12532080709934235, "learning_rate": 8.019262639733487e-05, "loss": 0.0162, "step": 6580 }, { "grad_norm": 0.1719275861978531, "learning_rate": 8.012668644099531e-05, "loss": 0.0189, "step": 6590 }, { "grad_norm": 0.15529029071331024, "learning_rate": 8.006066411944542e-05, "loss": 0.019, "step": 6600 }, { "grad_norm": 0.14322994649410248, "learning_rate": 7.999455961318769e-05, "loss": 0.02, "step": 6610 }, { "grad_norm": 0.13898257911205292, "learning_rate": 7.992837310294932e-05, "loss": 0.0212, "step": 6620 }, { "grad_norm": 0.16213196516036987, "learning_rate": 7.986210476968167e-05, "loss": 0.0215, "step": 6630 }, { "grad_norm": 0.1574016511440277, "learning_rate": 7.97957547945599e-05, "loss": 0.0213, "step": 6640 }, { "grad_norm": 0.19076316058635712, "learning_rate": 7.972932335898226e-05, "loss": 0.0229, "step": 6650 }, { "grad_norm": 0.15973800420761108, "learning_rate": 7.966281064456975e-05, "loss": 0.0179, "step": 6660 }, { "grad_norm": 0.13582278788089752, "learning_rate": 7.959621683316563e-05, "loss": 0.0193, "step": 6670 }, { "grad_norm": 0.19925376772880554, "learning_rate": 7.952954210683481e-05, "loss": 0.0171, "step": 6680 }, { "grad_norm": 0.18712837994098663, "learning_rate": 7.946278664786345e-05, "loss": 0.0182, "step": 6690 }, { "grad_norm": 0.17281898856163025, "learning_rate": 7.939595063875842e-05, "loss": 0.0186, "step": 6700 }, { "grad_norm": 0.2677015960216522, "learning_rate": 7.932903426224683e-05, "loss": 0.0204, "step": 6710 }, { "grad_norm": 0.21418102085590363, "learning_rate": 7.926203770127552e-05, "loss": 0.0238, "step": 6720 }, { "grad_norm": 0.21113035082817078, "learning_rate": 7.919496113901046e-05, "loss": 0.0209, "step": 6730 }, { "grad_norm": 0.17297054827213287, "learning_rate": 7.912780475883649e-05, "loss": 0.0185, "step": 6740 }, { "grad_norm": 0.12711337208747864, "learning_rate": 7.906056874435652e-05, "loss": 0.0259, "step": 6750 }, { "grad_norm": 0.19699472188949585, "learning_rate": 7.899325327939131e-05, "loss": 0.0209, "step": 6760 }, { "grad_norm": 0.21699859201908112, "learning_rate": 7.892585854797872e-05, "loss": 0.021, "step": 6770 }, { "grad_norm": 0.1799279898405075, "learning_rate": 7.88583847343734e-05, "loss": 0.0211, "step": 6780 }, { "grad_norm": 0.14707234501838684, "learning_rate": 7.879083202304616e-05, "loss": 0.0168, "step": 6790 }, { "grad_norm": 0.16201511025428772, "learning_rate": 7.872320059868355e-05, "loss": 0.0186, "step": 6800 }, { "grad_norm": 0.16492746770381927, "learning_rate": 7.865549064618729e-05, "loss": 0.0196, "step": 6810 }, { "grad_norm": 0.17611610889434814, "learning_rate": 7.858770235067381e-05, "loss": 0.0203, "step": 6820 }, { "grad_norm": 0.14879129827022552, "learning_rate": 7.851983589747374e-05, "loss": 0.0191, "step": 6830 }, { "grad_norm": 0.19194968044757843, "learning_rate": 7.845189147213133e-05, "loss": 0.0186, "step": 6840 }, { "grad_norm": 0.15415453910827637, "learning_rate": 7.838386926040407e-05, "loss": 0.0183, "step": 6850 }, { "grad_norm": 0.16589607298374176, "learning_rate": 7.83157694482621e-05, "loss": 0.0178, "step": 6860 }, { "grad_norm": 0.19144010543823242, "learning_rate": 7.824759222188768e-05, "loss": 0.017, "step": 6870 }, { "grad_norm": 0.1125003769993782, "learning_rate": 7.817933776767478e-05, "loss": 0.0191, "step": 6880 }, { "grad_norm": 0.15303091704845428, "learning_rate": 7.811100627222842e-05, "loss": 0.017, "step": 6890 }, { "grad_norm": 0.18834351003170013, "learning_rate": 7.804259792236435e-05, "loss": 0.02, "step": 6900 }, { "grad_norm": 0.18068701028823853, "learning_rate": 7.797411290510835e-05, "loss": 0.0209, "step": 6910 }, { "grad_norm": 0.12338202446699142, "learning_rate": 7.790555140769586e-05, "loss": 0.0199, "step": 6920 }, { "grad_norm": 0.20274800062179565, "learning_rate": 7.78369136175714e-05, "loss": 0.02, "step": 6930 }, { "grad_norm": 0.17414060235023499, "learning_rate": 7.776819972238806e-05, "loss": 0.0152, "step": 6940 }, { "grad_norm": 0.19039417803287506, "learning_rate": 7.7699409910007e-05, "loss": 0.0191, "step": 6950 }, { "grad_norm": 0.2114764153957367, "learning_rate": 7.763054436849694e-05, "loss": 0.0181, "step": 6960 }, { "grad_norm": 0.15872064232826233, "learning_rate": 7.756160328613364e-05, "loss": 0.017, "step": 6970 }, { "grad_norm": 0.12650226056575775, "learning_rate": 7.749258685139942e-05, "loss": 0.0186, "step": 6980 }, { "grad_norm": 0.1296301931142807, "learning_rate": 7.742349525298253e-05, "loss": 0.0164, "step": 6990 }, { "grad_norm": 0.14347945153713226, "learning_rate": 7.735432867977679e-05, "loss": 0.0164, "step": 7000 }, { "grad_norm": 0.10224805027246475, "learning_rate": 7.728508732088096e-05, "loss": 0.0187, "step": 7010 }, { "grad_norm": 0.131644606590271, "learning_rate": 7.721577136559825e-05, "loss": 0.0177, "step": 7020 }, { "grad_norm": 0.20682132244110107, "learning_rate": 7.714638100343588e-05, "loss": 0.0167, "step": 7030 }, { "grad_norm": 0.17725659906864166, "learning_rate": 7.707691642410444e-05, "loss": 0.0171, "step": 7040 }, { "grad_norm": 0.15362116694450378, "learning_rate": 7.70073778175174e-05, "loss": 0.0173, "step": 7050 }, { "grad_norm": 0.17513050138950348, "learning_rate": 7.69377653737907e-05, "loss": 0.0146, "step": 7060 }, { "grad_norm": 0.1590604931116104, "learning_rate": 7.686807928324209e-05, "loss": 0.0218, "step": 7070 }, { "grad_norm": 0.15646639466285706, "learning_rate": 7.679831973639065e-05, "loss": 0.0169, "step": 7080 }, { "grad_norm": 0.21683405339717865, "learning_rate": 7.672848692395637e-05, "loss": 0.0187, "step": 7090 }, { "grad_norm": 0.21950334310531616, "learning_rate": 7.665858103685944e-05, "loss": 0.0176, "step": 7100 }, { "grad_norm": 0.12550988793373108, "learning_rate": 7.658860226621991e-05, "loss": 0.0189, "step": 7110 }, { "grad_norm": 0.1424347162246704, "learning_rate": 7.651855080335708e-05, "loss": 0.0174, "step": 7120 }, { "grad_norm": 0.1414279341697693, "learning_rate": 7.644842683978896e-05, "loss": 0.0206, "step": 7130 }, { "grad_norm": 0.19176964461803436, "learning_rate": 7.63782305672318e-05, "loss": 0.0173, "step": 7140 }, { "grad_norm": 0.17511357367038727, "learning_rate": 7.63079621775995e-05, "loss": 0.0157, "step": 7150 }, { "grad_norm": 0.172024667263031, "learning_rate": 7.623762186300319e-05, "loss": 0.019, "step": 7160 }, { "grad_norm": 0.15785501897335052, "learning_rate": 7.616720981575057e-05, "loss": 0.0204, "step": 7170 }, { "grad_norm": 0.15673957765102386, "learning_rate": 7.609672622834552e-05, "loss": 0.0183, "step": 7180 }, { "grad_norm": 0.12619225680828094, "learning_rate": 7.602617129348747e-05, "loss": 0.0199, "step": 7190 }, { "grad_norm": 0.16102445125579834, "learning_rate": 7.595554520407088e-05, "loss": 0.0176, "step": 7200 }, { "grad_norm": 0.244862899184227, "learning_rate": 7.588484815318484e-05, "loss": 0.0215, "step": 7210 }, { "grad_norm": 0.15186691284179688, "learning_rate": 7.581408033411234e-05, "loss": 0.0153, "step": 7220 }, { "grad_norm": 0.13785800337791443, "learning_rate": 7.574324194032995e-05, "loss": 0.0209, "step": 7230 }, { "grad_norm": 0.18052054941654205, "learning_rate": 7.567233316550705e-05, "loss": 0.0213, "step": 7240 }, { "grad_norm": 0.07835818827152252, "learning_rate": 7.560135420350562e-05, "loss": 0.0153, "step": 7250 }, { "grad_norm": 0.13483449816703796, "learning_rate": 7.553030524837935e-05, "loss": 0.0162, "step": 7260 }, { "grad_norm": 0.20061151683330536, "learning_rate": 7.545918649437341e-05, "loss": 0.0191, "step": 7270 }, { "grad_norm": 0.16476666927337646, "learning_rate": 7.538799813592377e-05, "loss": 0.0171, "step": 7280 }, { "grad_norm": 0.10083651542663574, "learning_rate": 7.531674036765662e-05, "loss": 0.0174, "step": 7290 }, { "grad_norm": 0.13007965683937073, "learning_rate": 7.524541338438807e-05, "loss": 0.0164, "step": 7300 }, { "grad_norm": 0.12419658899307251, "learning_rate": 7.517401738112328e-05, "loss": 0.0174, "step": 7310 }, { "grad_norm": 0.18935181200504303, "learning_rate": 7.510255255305628e-05, "loss": 0.0222, "step": 7320 }, { "grad_norm": 0.15515828132629395, "learning_rate": 7.503101909556911e-05, "loss": 0.0168, "step": 7330 }, { "grad_norm": 0.11114658415317535, "learning_rate": 7.495941720423154e-05, "loss": 0.0157, "step": 7340 }, { "grad_norm": 0.13707908987998962, "learning_rate": 7.488774707480042e-05, "loss": 0.0161, "step": 7350 }, { "grad_norm": 0.14754724502563477, "learning_rate": 7.481600890321911e-05, "loss": 0.0236, "step": 7360 }, { "grad_norm": 0.20627304911613464, "learning_rate": 7.474420288561708e-05, "loss": 0.0217, "step": 7370 }, { "grad_norm": 0.14729149639606476, "learning_rate": 7.467232921830921e-05, "loss": 0.0185, "step": 7380 }, { "grad_norm": 0.14813154935836792, "learning_rate": 7.460038809779537e-05, "loss": 0.0204, "step": 7390 }, { "grad_norm": 0.12695573270320892, "learning_rate": 7.452837972075983e-05, "loss": 0.0163, "step": 7400 }, { "grad_norm": 0.13950158655643463, "learning_rate": 7.445630428407074e-05, "loss": 0.0168, "step": 7410 }, { "grad_norm": 0.14056864380836487, "learning_rate": 7.43841619847796e-05, "loss": 0.0171, "step": 7420 }, { "grad_norm": 0.17867505550384521, "learning_rate": 7.431195302012072e-05, "loss": 0.0167, "step": 7430 }, { "grad_norm": 0.10542461276054382, "learning_rate": 7.423967758751061e-05, "loss": 0.0171, "step": 7440 }, { "grad_norm": 0.12210749834775925, "learning_rate": 7.416733588454758e-05, "loss": 0.016, "step": 7450 }, { "grad_norm": 0.21978414058685303, "learning_rate": 7.409492810901106e-05, "loss": 0.0185, "step": 7460 }, { "grad_norm": 0.11824891716241837, "learning_rate": 7.402245445886116e-05, "loss": 0.0165, "step": 7470 }, { "grad_norm": 0.16502009332180023, "learning_rate": 7.394991513223806e-05, "loss": 0.0161, "step": 7480 }, { "grad_norm": 0.10794845968484879, "learning_rate": 7.38773103274615e-05, "loss": 0.0146, "step": 7490 }, { "grad_norm": 0.1266213208436966, "learning_rate": 7.380464024303028e-05, "loss": 0.0189, "step": 7500 }, { "grad_norm": 0.169349804520607, "learning_rate": 7.373190507762162e-05, "loss": 0.0165, "step": 7510 }, { "grad_norm": 0.22548776865005493, "learning_rate": 7.365910503009066e-05, "loss": 0.0198, "step": 7520 }, { "grad_norm": 0.16755923628807068, "learning_rate": 7.358624029946996e-05, "loss": 0.0167, "step": 7530 }, { "grad_norm": 0.20859819650650024, "learning_rate": 7.351331108496893e-05, "loss": 0.0175, "step": 7540 }, { "grad_norm": 0.16441601514816284, "learning_rate": 7.344031758597325e-05, "loss": 0.0174, "step": 7550 }, { "grad_norm": 0.13209064304828644, "learning_rate": 7.336726000204435e-05, "loss": 0.0208, "step": 7560 }, { "grad_norm": 0.159614697098732, "learning_rate": 7.32941385329189e-05, "loss": 0.0155, "step": 7570 }, { "grad_norm": 0.11548754572868347, "learning_rate": 7.322095337850816e-05, "loss": 0.0143, "step": 7580 }, { "grad_norm": 0.12122035026550293, "learning_rate": 7.314770473889758e-05, "loss": 0.0158, "step": 7590 }, { "grad_norm": 0.19841183722019196, "learning_rate": 7.307439281434615e-05, "loss": 0.0188, "step": 7600 }, { "grad_norm": 0.12525832653045654, "learning_rate": 7.300101780528585e-05, "loss": 0.0158, "step": 7610 }, { "grad_norm": 0.17717115581035614, "learning_rate": 7.292757991232117e-05, "loss": 0.0187, "step": 7620 }, { "grad_norm": 0.16742336750030518, "learning_rate": 7.285407933622848e-05, "loss": 0.0197, "step": 7630 }, { "grad_norm": 0.1493634581565857, "learning_rate": 7.278051627795557e-05, "loss": 0.0187, "step": 7640 }, { "grad_norm": 0.1942657232284546, "learning_rate": 7.270689093862105e-05, "loss": 0.0178, "step": 7650 }, { "grad_norm": 0.19402702152729034, "learning_rate": 7.263320351951374e-05, "loss": 0.0211, "step": 7660 }, { "grad_norm": 0.15949857234954834, "learning_rate": 7.255945422209227e-05, "loss": 0.02, "step": 7670 }, { "grad_norm": 0.18142332136631012, "learning_rate": 7.248564324798437e-05, "loss": 0.0163, "step": 7680 }, { "grad_norm": 0.12319275736808777, "learning_rate": 7.241177079898644e-05, "loss": 0.0152, "step": 7690 }, { "grad_norm": 0.20039460062980652, "learning_rate": 7.233783707706295e-05, "loss": 0.0208, "step": 7700 }, { "grad_norm": 0.17680075764656067, "learning_rate": 7.226384228434586e-05, "loss": 0.0181, "step": 7710 }, { "grad_norm": 0.15215392410755157, "learning_rate": 7.21897866231341e-05, "loss": 0.0156, "step": 7720 }, { "grad_norm": 0.1541053056716919, "learning_rate": 7.211567029589303e-05, "loss": 0.0169, "step": 7730 }, { "grad_norm": 0.1532987505197525, "learning_rate": 7.204149350525387e-05, "loss": 0.0155, "step": 7740 }, { "grad_norm": 0.1315370351076126, "learning_rate": 7.196725645401309e-05, "loss": 0.0179, "step": 7750 }, { "grad_norm": 0.1956956535577774, "learning_rate": 7.1892959345132e-05, "loss": 0.0165, "step": 7760 }, { "grad_norm": 0.18244892358779907, "learning_rate": 7.181860238173605e-05, "loss": 0.021, "step": 7770 }, { "grad_norm": 0.10399963706731796, "learning_rate": 7.174418576711432e-05, "loss": 0.0153, "step": 7780 }, { "grad_norm": 0.16426914930343628, "learning_rate": 7.1669709704719e-05, "loss": 0.0137, "step": 7790 }, { "grad_norm": 0.13457612693309784, "learning_rate": 7.159517439816481e-05, "loss": 0.0165, "step": 7800 }, { "grad_norm": 0.1164279654622078, "learning_rate": 7.152058005122842e-05, "loss": 0.0154, "step": 7810 }, { "grad_norm": 0.14225716888904572, "learning_rate": 7.144592686784793e-05, "loss": 0.0141, "step": 7820 }, { "grad_norm": 0.11547508835792542, "learning_rate": 7.137121505212229e-05, "loss": 0.0181, "step": 7830 }, { "grad_norm": 0.15677455067634583, "learning_rate": 7.129644480831077e-05, "loss": 0.0151, "step": 7840 }, { "grad_norm": 0.24100778996944427, "learning_rate": 7.122161634083234e-05, "loss": 0.0168, "step": 7850 }, { "grad_norm": 0.15490888059139252, "learning_rate": 7.114672985426516e-05, "loss": 0.019, "step": 7860 }, { "grad_norm": 0.26544031500816345, "learning_rate": 7.107178555334606e-05, "loss": 0.0174, "step": 7870 }, { "grad_norm": 0.1521638184785843, "learning_rate": 7.099678364296989e-05, "loss": 0.0155, "step": 7880 }, { "grad_norm": 0.16411571204662323, "learning_rate": 7.0921724328189e-05, "loss": 0.0172, "step": 7890 }, { "grad_norm": 0.15566524863243103, "learning_rate": 7.084660781421268e-05, "loss": 0.0167, "step": 7900 }, { "grad_norm": 0.15498609840869904, "learning_rate": 7.077143430640662e-05, "loss": 0.0181, "step": 7910 }, { "grad_norm": 0.1291123628616333, "learning_rate": 7.069620401029232e-05, "loss": 0.0161, "step": 7920 }, { "grad_norm": 0.15748730301856995, "learning_rate": 7.062091713154655e-05, "loss": 0.0162, "step": 7930 }, { "grad_norm": 0.16603681445121765, "learning_rate": 7.054557387600075e-05, "loss": 0.0178, "step": 7940 }, { "grad_norm": 0.16718164086341858, "learning_rate": 7.04701744496405e-05, "loss": 0.0162, "step": 7950 }, { "grad_norm": 0.1429743617773056, "learning_rate": 7.039471905860495e-05, "loss": 0.0178, "step": 7960 }, { "grad_norm": 0.17504434287548065, "learning_rate": 7.031920790918628e-05, "loss": 0.0189, "step": 7970 }, { "grad_norm": 0.1412392109632492, "learning_rate": 7.024364120782906e-05, "loss": 0.0192, "step": 7980 }, { "grad_norm": 0.12423244118690491, "learning_rate": 7.016801916112978e-05, "loss": 0.0176, "step": 7990 }, { "grad_norm": 0.20896317064762115, "learning_rate": 7.009234197583623e-05, "loss": 0.015, "step": 8000 }, { "grad_norm": 0.15057845413684845, "learning_rate": 7.001660985884692e-05, "loss": 0.0128, "step": 8010 }, { "grad_norm": 0.161129429936409, "learning_rate": 6.994082301721063e-05, "loss": 0.0147, "step": 8020 }, { "grad_norm": 0.10862451791763306, "learning_rate": 6.986498165812563e-05, "loss": 0.0149, "step": 8030 }, { "grad_norm": 0.12585623562335968, "learning_rate": 6.978908598893932e-05, "loss": 0.0172, "step": 8040 }, { "grad_norm": 0.22039827704429626, "learning_rate": 6.971313621714756e-05, "loss": 0.0164, "step": 8050 }, { "grad_norm": 0.17915983498096466, "learning_rate": 6.96371325503941e-05, "loss": 0.0159, "step": 8060 }, { "grad_norm": 0.14381596446037292, "learning_rate": 6.956107519647014e-05, "loss": 0.0148, "step": 8070 }, { "grad_norm": 0.14740067720413208, "learning_rate": 6.94849643633135e-05, "loss": 0.0167, "step": 8080 }, { "grad_norm": 0.2096027433872223, "learning_rate": 6.940880025900834e-05, "loss": 0.0175, "step": 8090 }, { "grad_norm": 0.11374073475599289, "learning_rate": 6.933258309178438e-05, "loss": 0.0183, "step": 8100 }, { "grad_norm": 0.15331900119781494, "learning_rate": 6.925631307001646e-05, "loss": 0.0152, "step": 8110 }, { "grad_norm": 0.1773197501897812, "learning_rate": 6.91799904022239e-05, "loss": 0.0182, "step": 8120 }, { "grad_norm": 0.16684658825397491, "learning_rate": 6.910361529706997e-05, "loss": 0.0165, "step": 8130 }, { "grad_norm": 0.155971497297287, "learning_rate": 6.902718796336131e-05, "loss": 0.0153, "step": 8140 }, { "grad_norm": 0.15327471494674683, "learning_rate": 6.895070861004729e-05, "loss": 0.017, "step": 8150 }, { "grad_norm": 0.12731382250785828, "learning_rate": 6.887417744621956e-05, "loss": 0.0178, "step": 8160 }, { "grad_norm": 0.09598815441131592, "learning_rate": 6.87975946811114e-05, "loss": 0.016, "step": 8170 }, { "grad_norm": 0.1547369658946991, "learning_rate": 6.872096052409718e-05, "loss": 0.0172, "step": 8180 }, { "grad_norm": 0.1362910121679306, "learning_rate": 6.864427518469174e-05, "loss": 0.0178, "step": 8190 }, { "grad_norm": 0.19546937942504883, "learning_rate": 6.856753887254986e-05, "loss": 0.0165, "step": 8200 }, { "grad_norm": 0.15911701321601868, "learning_rate": 6.849075179746572e-05, "loss": 0.0159, "step": 8210 }, { "grad_norm": 0.2192545235157013, "learning_rate": 6.841391416937221e-05, "loss": 0.0128, "step": 8220 }, { "grad_norm": 0.21248213946819305, "learning_rate": 6.833702619834053e-05, "loss": 0.0154, "step": 8230 }, { "grad_norm": 0.13653047382831573, "learning_rate": 6.82600880945794e-05, "loss": 0.0155, "step": 8240 }, { "grad_norm": 0.1513722687959671, "learning_rate": 6.818310006843468e-05, "loss": 0.0175, "step": 8250 }, { "grad_norm": 0.1396065652370453, "learning_rate": 6.810606233038868e-05, "loss": 0.0152, "step": 8260 }, { "grad_norm": 0.11816348880529404, "learning_rate": 6.802897509105966e-05, "loss": 0.0145, "step": 8270 }, { "grad_norm": 0.16390925645828247, "learning_rate": 6.79518385612012e-05, "loss": 0.0162, "step": 8280 }, { "grad_norm": 0.17345911264419556, "learning_rate": 6.787465295170157e-05, "loss": 0.0177, "step": 8290 }, { "grad_norm": 0.15061445534229279, "learning_rate": 6.779741847358332e-05, "loss": 0.0161, "step": 8300 }, { "grad_norm": 0.16690070927143097, "learning_rate": 6.772013533800256e-05, "loss": 0.0184, "step": 8310 }, { "grad_norm": 0.1651339828968048, "learning_rate": 6.764280375624843e-05, "loss": 0.0149, "step": 8320 }, { "grad_norm": 0.09905768185853958, "learning_rate": 6.756542393974252e-05, "loss": 0.0165, "step": 8330 }, { "grad_norm": 0.16429457068443298, "learning_rate": 6.748799610003828e-05, "loss": 0.0165, "step": 8340 }, { "grad_norm": 0.14917202293872833, "learning_rate": 6.741052044882048e-05, "loss": 0.0175, "step": 8350 }, { "grad_norm": 0.14229144155979156, "learning_rate": 6.73329971979046e-05, "loss": 0.0164, "step": 8360 }, { "grad_norm": 0.12179477512836456, "learning_rate": 6.725542655923625e-05, "loss": 0.0126, "step": 8370 }, { "grad_norm": 0.18678165972232819, "learning_rate": 6.717780874489057e-05, "loss": 0.0152, "step": 8380 }, { "grad_norm": 0.1588679701089859, "learning_rate": 6.710014396707172e-05, "loss": 0.0168, "step": 8390 }, { "grad_norm": 0.11333762109279633, "learning_rate": 6.702243243811221e-05, "loss": 0.0161, "step": 8400 }, { "grad_norm": 0.11029520630836487, "learning_rate": 6.694467437047244e-05, "loss": 0.0176, "step": 8410 }, { "grad_norm": 0.14706335961818695, "learning_rate": 6.686686997673997e-05, "loss": 0.0155, "step": 8420 }, { "grad_norm": 0.13338269293308258, "learning_rate": 6.678901946962903e-05, "loss": 0.0131, "step": 8430 }, { "grad_norm": 0.14797654747962952, "learning_rate": 6.671112306197996e-05, "loss": 0.0164, "step": 8440 }, { "grad_norm": 0.22411692142486572, "learning_rate": 6.663318096675854e-05, "loss": 0.0218, "step": 8450 }, { "grad_norm": 0.12939627468585968, "learning_rate": 6.655519339705552e-05, "loss": 0.0157, "step": 8460 }, { "grad_norm": 0.15027618408203125, "learning_rate": 6.647716056608588e-05, "loss": 0.0151, "step": 8470 }, { "grad_norm": 0.17822995781898499, "learning_rate": 6.639908268718843e-05, "loss": 0.0158, "step": 8480 }, { "grad_norm": 0.13639095425605774, "learning_rate": 6.632095997382514e-05, "loss": 0.016, "step": 8490 }, { "grad_norm": 0.1883944720029831, "learning_rate": 6.624279263958047e-05, "loss": 0.0186, "step": 8500 }, { "grad_norm": 0.18717031180858612, "learning_rate": 6.616458089816097e-05, "loss": 0.0151, "step": 8510 }, { "grad_norm": 0.1482684463262558, "learning_rate": 6.608632496339454e-05, "loss": 0.015, "step": 8520 }, { "grad_norm": 0.1221562847495079, "learning_rate": 6.600802504922988e-05, "loss": 0.0139, "step": 8530 }, { "grad_norm": 0.18009908497333527, "learning_rate": 6.592968136973604e-05, "loss": 0.0185, "step": 8540 }, { "grad_norm": 0.12899819016456604, "learning_rate": 6.585129413910159e-05, "loss": 0.0142, "step": 8550 }, { "grad_norm": 0.17109514772891998, "learning_rate": 6.577286357163424e-05, "loss": 0.0163, "step": 8560 }, { "grad_norm": 0.12011495232582092, "learning_rate": 6.569438988176018e-05, "loss": 0.0217, "step": 8570 }, { "grad_norm": 0.13880039751529694, "learning_rate": 6.561587328402347e-05, "loss": 0.0151, "step": 8580 }, { "grad_norm": 0.20035898685455322, "learning_rate": 6.553731399308549e-05, "loss": 0.0148, "step": 8590 }, { "grad_norm": 0.16783171892166138, "learning_rate": 6.545871222372436e-05, "loss": 0.0136, "step": 8600 }, { "grad_norm": 0.1264665722846985, "learning_rate": 6.538006819083426e-05, "loss": 0.0168, "step": 8610 }, { "grad_norm": 0.1363346129655838, "learning_rate": 6.530138210942505e-05, "loss": 0.014, "step": 8620 }, { "grad_norm": 0.15918788313865662, "learning_rate": 6.522265419462141e-05, "loss": 0.0172, "step": 8630 }, { "grad_norm": 0.1680510938167572, "learning_rate": 6.514388466166248e-05, "loss": 0.0174, "step": 8640 }, { "grad_norm": 0.1941153109073639, "learning_rate": 6.506507372590119e-05, "loss": 0.0169, "step": 8650 }, { "grad_norm": 0.23341475427150726, "learning_rate": 6.498622160280355e-05, "loss": 0.015, "step": 8660 }, { "grad_norm": 0.13256250321865082, "learning_rate": 6.490732850794832e-05, "loss": 0.0172, "step": 8670 }, { "grad_norm": 0.18063566088676453, "learning_rate": 6.482839465702616e-05, "loss": 0.0175, "step": 8680 }, { "grad_norm": 0.13047568500041962, "learning_rate": 6.474942026583923e-05, "loss": 0.0139, "step": 8690 }, { "grad_norm": 0.14261992275714874, "learning_rate": 6.467040555030052e-05, "loss": 0.0131, "step": 8700 }, { "grad_norm": 0.18131746351718903, "learning_rate": 6.459135072643321e-05, "loss": 0.0161, "step": 8710 }, { "grad_norm": 0.1412636637687683, "learning_rate": 6.451225601037019e-05, "loss": 0.0157, "step": 8720 }, { "grad_norm": 0.1656758040189743, "learning_rate": 6.443312161835338e-05, "loss": 0.0149, "step": 8730 }, { "grad_norm": 0.13524064421653748, "learning_rate": 6.43539477667332e-05, "loss": 0.0159, "step": 8740 }, { "grad_norm": 0.12824025750160217, "learning_rate": 6.427473467196793e-05, "loss": 0.0126, "step": 8750 }, { "grad_norm": 0.12482555955648422, "learning_rate": 6.419548255062315e-05, "loss": 0.014, "step": 8760 }, { "grad_norm": 0.1854359358549118, "learning_rate": 6.411619161937112e-05, "loss": 0.0176, "step": 8770 }, { "grad_norm": 0.2023647278547287, "learning_rate": 6.403686209499022e-05, "loss": 0.0173, "step": 8780 }, { "grad_norm": 0.21108640730381012, "learning_rate": 6.395749419436437e-05, "loss": 0.0166, "step": 8790 }, { "grad_norm": 0.1438765674829483, "learning_rate": 6.387808813448234e-05, "loss": 0.0122, "step": 8800 }, { "grad_norm": 0.09879841655492783, "learning_rate": 6.37986441324373e-05, "loss": 0.0122, "step": 8810 }, { "grad_norm": 0.1581912636756897, "learning_rate": 6.37191624054261e-05, "loss": 0.0152, "step": 8820 }, { "grad_norm": 0.21173761785030365, "learning_rate": 6.363964317074872e-05, "loss": 0.0148, "step": 8830 }, { "grad_norm": 0.22477108240127563, "learning_rate": 6.356008664580776e-05, "loss": 0.0195, "step": 8840 }, { "grad_norm": 0.12742134928703308, "learning_rate": 6.348049304810771e-05, "loss": 0.0164, "step": 8850 }, { "grad_norm": 0.10113922506570816, "learning_rate": 6.340086259525442e-05, "loss": 0.0149, "step": 8860 }, { "grad_norm": 0.12814676761627197, "learning_rate": 6.332119550495448e-05, "loss": 0.016, "step": 8870 }, { "grad_norm": 0.16140882670879364, "learning_rate": 6.324149199501473e-05, "loss": 0.0153, "step": 8880 }, { "grad_norm": 0.19577400386333466, "learning_rate": 6.316175228334146e-05, "loss": 0.0189, "step": 8890 }, { "grad_norm": 0.2105592042207718, "learning_rate": 6.308197658794003e-05, "loss": 0.0174, "step": 8900 }, { "grad_norm": 0.2152465283870697, "learning_rate": 6.300216512691417e-05, "loss": 0.0151, "step": 8910 }, { "grad_norm": 0.11741747707128525, "learning_rate": 6.292231811846532e-05, "loss": 0.0155, "step": 8920 }, { "grad_norm": 0.15391118824481964, "learning_rate": 6.284243578089217e-05, "loss": 0.018, "step": 8930 }, { "grad_norm": 0.15297570824623108, "learning_rate": 6.276251833258999e-05, "loss": 0.0147, "step": 8940 }, { "grad_norm": 0.1840389221906662, "learning_rate": 6.268256599205003e-05, "loss": 0.0139, "step": 8950 }, { "grad_norm": 0.13984337449073792, "learning_rate": 6.260257897785892e-05, "loss": 0.0162, "step": 8960 }, { "grad_norm": 0.1481754332780838, "learning_rate": 6.252255750869811e-05, "loss": 0.0169, "step": 8970 }, { "grad_norm": 0.1514410674571991, "learning_rate": 6.244250180334325e-05, "loss": 0.015, "step": 8980 }, { "grad_norm": 0.09923096746206284, "learning_rate": 6.236241208066356e-05, "loss": 0.0174, "step": 8990 }, { "grad_norm": 0.16595590114593506, "learning_rate": 6.228228855962133e-05, "loss": 0.017, "step": 9000 }, { "grad_norm": 0.13913275301456451, "learning_rate": 6.220213145927115e-05, "loss": 0.0178, "step": 9010 }, { "grad_norm": 0.1251859962940216, "learning_rate": 6.212194099875951e-05, "loss": 0.0182, "step": 9020 }, { "grad_norm": 0.13626161217689514, "learning_rate": 6.204171739732405e-05, "loss": 0.017, "step": 9030 }, { "grad_norm": 0.1914566159248352, "learning_rate": 6.196146087429303e-05, "loss": 0.0144, "step": 9040 }, { "grad_norm": 0.16061173379421234, "learning_rate": 6.188117164908474e-05, "loss": 0.0141, "step": 9050 }, { "grad_norm": 0.16121314465999603, "learning_rate": 6.180084994120684e-05, "loss": 0.0156, "step": 9060 }, { "grad_norm": 0.1587548404932022, "learning_rate": 6.17204959702558e-05, "loss": 0.0141, "step": 9070 }, { "grad_norm": 0.13000820577144623, "learning_rate": 6.164010995591635e-05, "loss": 0.0145, "step": 9080 }, { "grad_norm": 0.13612189888954163, "learning_rate": 6.155969211796076e-05, "loss": 0.0156, "step": 9090 }, { "grad_norm": 0.11115013808012009, "learning_rate": 6.147924267624829e-05, "loss": 0.0144, "step": 9100 }, { "grad_norm": 0.17304812371730804, "learning_rate": 6.13987618507247e-05, "loss": 0.0129, "step": 9110 }, { "grad_norm": 0.14822974801063538, "learning_rate": 6.131824986142147e-05, "loss": 0.0152, "step": 9120 }, { "grad_norm": 0.1734798699617386, "learning_rate": 6.123770692845529e-05, "loss": 0.015, "step": 9130 }, { "grad_norm": 0.14160504937171936, "learning_rate": 6.11571332720275e-05, "loss": 0.0134, "step": 9140 }, { "grad_norm": 0.13432805240154266, "learning_rate": 6.107652911242336e-05, "loss": 0.0166, "step": 9150 }, { "grad_norm": 0.14062321186065674, "learning_rate": 6.0995894670011586e-05, "loss": 0.013, "step": 9160 }, { "grad_norm": 0.14371217787265778, "learning_rate": 6.091523016524368e-05, "loss": 0.0146, "step": 9170 }, { "grad_norm": 0.1384711116552353, "learning_rate": 6.083453581865328e-05, "loss": 0.0148, "step": 9180 }, { "grad_norm": 0.15511776506900787, "learning_rate": 6.075381185085568e-05, "loss": 0.0158, "step": 9190 }, { "grad_norm": 0.14515270292758942, "learning_rate": 6.067305848254709e-05, "loss": 0.0143, "step": 9200 }, { "grad_norm": 0.15997879207134247, "learning_rate": 6.059227593450418e-05, "loss": 0.0155, "step": 9210 }, { "grad_norm": 0.13221235573291779, "learning_rate": 6.051146442758333e-05, "loss": 0.0153, "step": 9220 }, { "grad_norm": 0.1969967782497406, "learning_rate": 6.043062418272012e-05, "loss": 0.0163, "step": 9230 }, { "grad_norm": 0.1575111448764801, "learning_rate": 6.0349755420928666e-05, "loss": 0.0119, "step": 9240 }, { "grad_norm": 0.12794987857341766, "learning_rate": 6.0268858363301105e-05, "loss": 0.0157, "step": 9250 }, { "grad_norm": 0.152072936296463, "learning_rate": 6.018793323100689e-05, "loss": 0.0164, "step": 9260 }, { "grad_norm": 0.17848001420497894, "learning_rate": 6.0106980245292255e-05, "loss": 0.0178, "step": 9270 }, { "grad_norm": 0.16517706215381622, "learning_rate": 6.002599962747957e-05, "loss": 0.0166, "step": 9280 }, { "grad_norm": 0.11146853864192963, "learning_rate": 5.994499159896673e-05, "loss": 0.014, "step": 9290 }, { "grad_norm": 0.13090497255325317, "learning_rate": 5.9863956381226607e-05, "loss": 0.0157, "step": 9300 }, { "grad_norm": 0.14349132776260376, "learning_rate": 5.9782894195806394e-05, "loss": 0.0146, "step": 9310 }, { "grad_norm": 0.10222000628709793, "learning_rate": 5.9701805264327004e-05, "loss": 0.0132, "step": 9320 }, { "grad_norm": 0.10786471515893936, "learning_rate": 5.96206898084825e-05, "loss": 0.0139, "step": 9330 }, { "grad_norm": 0.12517696619033813, "learning_rate": 5.953954805003942e-05, "loss": 0.0187, "step": 9340 }, { "grad_norm": 0.20478032529354095, "learning_rate": 5.945838021083623e-05, "loss": 0.017, "step": 9350 }, { "grad_norm": 0.1768694519996643, "learning_rate": 5.9377186512782714e-05, "loss": 0.0144, "step": 9360 }, { "grad_norm": 0.10167831927537918, "learning_rate": 5.929596717785935e-05, "loss": 0.0132, "step": 9370 }, { "grad_norm": 0.09763530641794205, "learning_rate": 5.921472242811668e-05, "loss": 0.0121, "step": 9380 }, { "grad_norm": 0.13721071183681488, "learning_rate": 5.913345248567475e-05, "loss": 0.0134, "step": 9390 }, { "grad_norm": 0.11736586689949036, "learning_rate": 5.905215757272248e-05, "loss": 0.0151, "step": 9400 }, { "grad_norm": 0.15586896240711212, "learning_rate": 5.897083791151706e-05, "loss": 0.016, "step": 9410 }, { "grad_norm": 0.1565990447998047, "learning_rate": 5.888949372438336e-05, "loss": 0.0133, "step": 9420 }, { "grad_norm": 0.18745926022529602, "learning_rate": 5.8808125233713255e-05, "loss": 0.0141, "step": 9430 }, { "grad_norm": 0.22998793423175812, "learning_rate": 5.872673266196509e-05, "loss": 0.0153, "step": 9440 }, { "grad_norm": 0.17663384974002838, "learning_rate": 5.864531623166305e-05, "loss": 0.0142, "step": 9450 }, { "grad_norm": 0.13807439804077148, "learning_rate": 5.856387616539656e-05, "loss": 0.0148, "step": 9460 }, { "grad_norm": 0.10099150985479355, "learning_rate": 5.848241268581967e-05, "loss": 0.0141, "step": 9470 }, { "grad_norm": 0.17072321474552155, "learning_rate": 5.840092601565037e-05, "loss": 0.0129, "step": 9480 }, { "grad_norm": 0.16802653670310974, "learning_rate": 5.8319416377670144e-05, "loss": 0.0135, "step": 9490 }, { "grad_norm": 0.08352859318256378, "learning_rate": 5.82378839947232e-05, "loss": 0.0099, "step": 9500 }, { "grad_norm": 0.15814140439033508, "learning_rate": 5.815632908971599e-05, "loss": 0.0139, "step": 9510 }, { "grad_norm": 0.14148813486099243, "learning_rate": 5.80747518856165e-05, "loss": 0.0144, "step": 9520 }, { "grad_norm": 0.1144755482673645, "learning_rate": 5.799315260545367e-05, "loss": 0.0121, "step": 9530 }, { "grad_norm": 0.07684751600027084, "learning_rate": 5.791153147231686e-05, "loss": 0.0143, "step": 9540 }, { "grad_norm": 0.1394650638103485, "learning_rate": 5.782988870935509e-05, "loss": 0.0123, "step": 9550 }, { "grad_norm": 0.16465698182582855, "learning_rate": 5.774822453977657e-05, "loss": 0.0124, "step": 9560 }, { "grad_norm": 0.10952208191156387, "learning_rate": 5.7666539186848036e-05, "loss": 0.0137, "step": 9570 }, { "grad_norm": 0.11911491304636002, "learning_rate": 5.758483287389411e-05, "loss": 0.0151, "step": 9580 }, { "grad_norm": 0.1302008032798767, "learning_rate": 5.7503105824296735e-05, "loss": 0.0136, "step": 9590 }, { "grad_norm": 0.17709599435329437, "learning_rate": 5.742135826149453e-05, "loss": 0.0138, "step": 9600 }, { "grad_norm": 0.10012666136026382, "learning_rate": 5.7339590408982223e-05, "loss": 0.0117, "step": 9610 }, { "grad_norm": 0.20848406851291656, "learning_rate": 5.725780249031e-05, "loss": 0.0153, "step": 9620 }, { "grad_norm": 0.1413688361644745, "learning_rate": 5.717599472908292e-05, "loss": 0.0148, "step": 9630 }, { "grad_norm": 0.12486383318901062, "learning_rate": 5.7094167348960237e-05, "loss": 0.0145, "step": 9640 }, { "grad_norm": 0.09949912875890732, "learning_rate": 5.7012320573654945e-05, "loss": 0.0132, "step": 9650 }, { "grad_norm": 0.09780202805995941, "learning_rate": 5.693045462693295e-05, "loss": 0.0116, "step": 9660 }, { "grad_norm": 0.1419958770275116, "learning_rate": 5.684856973261266e-05, "loss": 0.0139, "step": 9670 }, { "grad_norm": 0.1398886889219284, "learning_rate": 5.6766666114564215e-05, "loss": 0.0139, "step": 9680 }, { "grad_norm": 0.10168332606554031, "learning_rate": 5.668474399670899e-05, "loss": 0.0133, "step": 9690 }, { "grad_norm": 0.1551263928413391, "learning_rate": 5.660280360301896e-05, "loss": 0.0132, "step": 9700 }, { "grad_norm": 0.12198057025671005, "learning_rate": 5.652084515751599e-05, "loss": 0.0138, "step": 9710 }, { "grad_norm": 0.1873226910829544, "learning_rate": 5.643886888427137e-05, "loss": 0.0132, "step": 9720 }, { "grad_norm": 0.08701168745756149, "learning_rate": 5.6356875007405074e-05, "loss": 0.0119, "step": 9730 }, { "grad_norm": 0.15093523263931274, "learning_rate": 5.627486375108525e-05, "loss": 0.0173, "step": 9740 }, { "grad_norm": 0.1273067444562912, "learning_rate": 5.619283533952754e-05, "loss": 0.0158, "step": 9750 }, { "grad_norm": 0.13635140657424927, "learning_rate": 5.6110789996994474e-05, "loss": 0.0128, "step": 9760 }, { "grad_norm": 0.1678747981786728, "learning_rate": 5.602872794779491e-05, "loss": 0.0143, "step": 9770 }, { "grad_norm": 0.11044449359178543, "learning_rate": 5.594664941628334e-05, "loss": 0.0164, "step": 9780 }, { "grad_norm": 0.11335909366607666, "learning_rate": 5.5864554626859324e-05, "loss": 0.0156, "step": 9790 }, { "grad_norm": 0.11399553716182709, "learning_rate": 5.578244380396691e-05, "loss": 0.0116, "step": 9800 }, { "grad_norm": 0.1027611717581749, "learning_rate": 5.570031717209394e-05, "loss": 0.0111, "step": 9810 }, { "grad_norm": 0.13834890723228455, "learning_rate": 5.561817495577147e-05, "loss": 0.0113, "step": 9820 }, { "grad_norm": 0.12169817090034485, "learning_rate": 5.5536017379573215e-05, "loss": 0.0125, "step": 9830 }, { "grad_norm": 0.11151054501533508, "learning_rate": 5.545384466811483e-05, "loss": 0.0125, "step": 9840 }, { "grad_norm": 0.11269786208868027, "learning_rate": 5.5371657046053384e-05, "loss": 0.0156, "step": 9850 }, { "grad_norm": 0.12539169192314148, "learning_rate": 5.528945473808669e-05, "loss": 0.0124, "step": 9860 }, { "grad_norm": 0.16476993262767792, "learning_rate": 5.520723796895272e-05, "loss": 0.014, "step": 9870 }, { "grad_norm": 0.1306542456150055, "learning_rate": 5.512500696342897e-05, "loss": 0.015, "step": 9880 }, { "grad_norm": 0.09963291883468628, "learning_rate": 5.504276194633188e-05, "loss": 0.0149, "step": 9890 }, { "grad_norm": 0.12878122925758362, "learning_rate": 5.49605031425162e-05, "loss": 0.0137, "step": 9900 }, { "grad_norm": 0.18913251161575317, "learning_rate": 5.487823077687434e-05, "loss": 0.0131, "step": 9910 }, { "grad_norm": 0.11837610602378845, "learning_rate": 5.4795945074335806e-05, "loss": 0.017, "step": 9920 }, { "grad_norm": 0.16047100722789764, "learning_rate": 5.471364625986657e-05, "loss": 0.018, "step": 9930 }, { "grad_norm": 0.10407508164644241, "learning_rate": 5.463133455846845e-05, "loss": 0.0135, "step": 9940 }, { "grad_norm": 0.11982892453670502, "learning_rate": 5.4549010195178505e-05, "loss": 0.0144, "step": 9950 }, { "grad_norm": 0.12296675145626068, "learning_rate": 5.446667339506838e-05, "loss": 0.0115, "step": 9960 }, { "grad_norm": 0.08932709693908691, "learning_rate": 5.4384324383243756e-05, "loss": 0.0118, "step": 9970 }, { "grad_norm": 0.11902914196252823, "learning_rate": 5.430196338484368e-05, "loss": 0.0131, "step": 9980 }, { "grad_norm": 0.17003805935382843, "learning_rate": 5.4219590625039975e-05, "loss": 0.0129, "step": 9990 }, { "grad_norm": 0.10413390398025513, "learning_rate": 5.413720632903664e-05, "loss": 0.0123, "step": 10000 }, { "grad_norm": 0.08527959138154984, "learning_rate": 5.405481072206917e-05, "loss": 0.0121, "step": 10010 }, { "grad_norm": 0.11480070650577545, "learning_rate": 5.397240402940402e-05, "loss": 0.0137, "step": 10020 }, { "grad_norm": 0.10560019314289093, "learning_rate": 5.388998647633794e-05, "loss": 0.0141, "step": 10030 }, { "grad_norm": 0.1376556158065796, "learning_rate": 5.380755828819737e-05, "loss": 0.0123, "step": 10040 }, { "grad_norm": 0.16762304306030273, "learning_rate": 5.3725119690337846e-05, "loss": 0.0125, "step": 10050 }, { "grad_norm": 0.0971708744764328, "learning_rate": 5.3642670908143324e-05, "loss": 0.0111, "step": 10060 }, { "grad_norm": 0.0989919900894165, "learning_rate": 5.356021216702562e-05, "loss": 0.0119, "step": 10070 }, { "grad_norm": 0.1705482453107834, "learning_rate": 5.347774369242381e-05, "loss": 0.0157, "step": 10080 }, { "grad_norm": 0.15000873804092407, "learning_rate": 5.3395265709803545e-05, "loss": 0.0138, "step": 10090 }, { "grad_norm": 0.15792034566402435, "learning_rate": 5.331277844465647e-05, "loss": 0.0151, "step": 10100 }, { "grad_norm": 0.14917050302028656, "learning_rate": 5.323028212249963e-05, "loss": 0.0131, "step": 10110 }, { "grad_norm": 0.16132472455501556, "learning_rate": 5.314777696887481e-05, "loss": 0.0118, "step": 10120 }, { "grad_norm": 0.1786492019891739, "learning_rate": 5.306526320934796e-05, "loss": 0.012, "step": 10130 }, { "grad_norm": 0.12978942692279816, "learning_rate": 5.298274106950854e-05, "loss": 0.0128, "step": 10140 }, { "grad_norm": 0.11711091548204422, "learning_rate": 5.290021077496893e-05, "loss": 0.0108, "step": 10150 }, { "grad_norm": 0.13735301792621613, "learning_rate": 5.2817672551363816e-05, "loss": 0.011, "step": 10160 }, { "grad_norm": 0.12286277115345001, "learning_rate": 5.273512662434952e-05, "loss": 0.0165, "step": 10170 }, { "grad_norm": 0.1795782595872879, "learning_rate": 5.265257321960349e-05, "loss": 0.0114, "step": 10180 }, { "grad_norm": 0.11815380305051804, "learning_rate": 5.257001256282357e-05, "loss": 0.0097, "step": 10190 }, { "grad_norm": 0.11677169054746628, "learning_rate": 5.248744487972742e-05, "loss": 0.012, "step": 10200 }, { "grad_norm": 0.09602950513362885, "learning_rate": 5.240487039605196e-05, "loss": 0.0127, "step": 10210 }, { "grad_norm": 0.15039651095867157, "learning_rate": 5.232228933755267e-05, "loss": 0.0134, "step": 10220 }, { "grad_norm": 0.10250694304704666, "learning_rate": 5.2239701930003006e-05, "loss": 0.0108, "step": 10230 }, { "grad_norm": 0.09431096166372299, "learning_rate": 5.215710839919379e-05, "loss": 0.0114, "step": 10240 }, { "grad_norm": 0.1180989071726799, "learning_rate": 5.207450897093257e-05, "loss": 0.0107, "step": 10250 }, { "grad_norm": 0.14753028750419617, "learning_rate": 5.1991903871043046e-05, "loss": 0.0134, "step": 10260 }, { "grad_norm": 0.15140195190906525, "learning_rate": 5.190929332536439e-05, "loss": 0.0122, "step": 10270 }, { "grad_norm": 0.12583374977111816, "learning_rate": 5.182667755975071e-05, "loss": 0.0121, "step": 10280 }, { "grad_norm": 0.13606469333171844, "learning_rate": 5.1744056800070315e-05, "loss": 0.0138, "step": 10290 }, { "grad_norm": 0.15014155209064484, "learning_rate": 5.166143127220524e-05, "loss": 0.0141, "step": 10300 }, { "grad_norm": 0.10969877988100052, "learning_rate": 5.1578801202050485e-05, "loss": 0.0123, "step": 10310 }, { "grad_norm": 0.1625463217496872, "learning_rate": 5.149616681551355e-05, "loss": 0.0104, "step": 10320 }, { "grad_norm": 0.11795984953641891, "learning_rate": 5.141352833851367e-05, "loss": 0.014, "step": 10330 }, { "grad_norm": 0.1256106048822403, "learning_rate": 5.1330885996981285e-05, "loss": 0.0121, "step": 10340 }, { "grad_norm": 0.14026691019535065, "learning_rate": 5.124824001685741e-05, "loss": 0.0123, "step": 10350 }, { "grad_norm": 0.11885134875774384, "learning_rate": 5.116559062409298e-05, "loss": 0.0114, "step": 10360 }, { "grad_norm": 0.12769021093845367, "learning_rate": 5.10829380446483e-05, "loss": 0.0125, "step": 10370 }, { "grad_norm": 0.0956525206565857, "learning_rate": 5.100028250449235e-05, "loss": 0.011, "step": 10380 }, { "grad_norm": 0.1474621444940567, "learning_rate": 5.0917624229602234e-05, "loss": 0.0131, "step": 10390 }, { "grad_norm": 0.1842966079711914, "learning_rate": 5.0834963445962524e-05, "loss": 0.0143, "step": 10400 }, { "grad_norm": 0.111672043800354, "learning_rate": 5.075230037956461e-05, "loss": 0.0117, "step": 10410 }, { "grad_norm": 0.15885749459266663, "learning_rate": 5.0669635256406213e-05, "loss": 0.0125, "step": 10420 }, { "grad_norm": 0.10357831418514252, "learning_rate": 5.058696830249058e-05, "loss": 0.011, "step": 10430 }, { "grad_norm": 0.1389058530330658, "learning_rate": 5.050429974382602e-05, "loss": 0.0106, "step": 10440 }, { "grad_norm": 0.15014374256134033, "learning_rate": 5.042162980642523e-05, "loss": 0.0113, "step": 10450 }, { "grad_norm": 0.1283929944038391, "learning_rate": 5.033895871630462e-05, "loss": 0.0122, "step": 10460 }, { "grad_norm": 0.1389484852552414, "learning_rate": 5.025628669948386e-05, "loss": 0.0121, "step": 10470 }, { "grad_norm": 0.1330685317516327, "learning_rate": 5.017361398198502e-05, "loss": 0.0131, "step": 10480 }, { "grad_norm": 0.1292942464351654, "learning_rate": 5.009094078983221e-05, "loss": 0.0133, "step": 10490 }, { "grad_norm": 0.1259763240814209, "learning_rate": 5.000826734905073e-05, "loss": 0.0169, "step": 10500 }, { "grad_norm": 0.11874383687973022, "learning_rate": 4.9925593885666645e-05, "loss": 0.0091, "step": 10510 }, { "grad_norm": 0.15004810690879822, "learning_rate": 4.984292062570602e-05, "loss": 0.0109, "step": 10520 }, { "grad_norm": 0.11827261000871658, "learning_rate": 4.976024779519442e-05, "loss": 0.0129, "step": 10530 }, { "grad_norm": 0.08720264583826065, "learning_rate": 4.9677575620156194e-05, "loss": 0.0131, "step": 10540 }, { "grad_norm": 0.19930215179920197, "learning_rate": 4.959490432661391e-05, "loss": 0.0135, "step": 10550 }, { "grad_norm": 0.143882617354393, "learning_rate": 4.9512234140587726e-05, "loss": 0.0115, "step": 10560 }, { "grad_norm": 0.11557275056838989, "learning_rate": 4.942956528809477e-05, "loss": 0.0126, "step": 10570 }, { "grad_norm": 0.10790234804153442, "learning_rate": 4.934689799514854e-05, "loss": 0.0102, "step": 10580 }, { "grad_norm": 0.13516627252101898, "learning_rate": 4.926423248775827e-05, "loss": 0.012, "step": 10590 }, { "grad_norm": 0.1811671257019043, "learning_rate": 4.918156899192826e-05, "loss": 0.0125, "step": 10600 }, { "grad_norm": 0.1128087192773819, "learning_rate": 4.909890773365738e-05, "loss": 0.0109, "step": 10610 }, { "grad_norm": 0.14425809681415558, "learning_rate": 4.9016248938938344e-05, "loss": 0.0148, "step": 10620 }, { "grad_norm": 0.15613360702991486, "learning_rate": 4.8933592833757156e-05, "loss": 0.013, "step": 10630 }, { "grad_norm": 0.2057190239429474, "learning_rate": 4.8850939644092435e-05, "loss": 0.0153, "step": 10640 }, { "grad_norm": 0.1101556122303009, "learning_rate": 4.876828959591485e-05, "loss": 0.0149, "step": 10650 }, { "grad_norm": 0.11830054968595505, "learning_rate": 4.8685642915186474e-05, "loss": 0.0105, "step": 10660 }, { "grad_norm": 0.19624674320220947, "learning_rate": 4.860299982786018e-05, "loss": 0.0115, "step": 10670 }, { "grad_norm": 0.21109960973262787, "learning_rate": 4.852036055987901e-05, "loss": 0.0137, "step": 10680 }, { "grad_norm": 0.1270499974489212, "learning_rate": 4.843772533717558e-05, "loss": 0.0126, "step": 10690 }, { "grad_norm": 0.09953948110342026, "learning_rate": 4.835509438567142e-05, "loss": 0.0116, "step": 10700 }, { "grad_norm": 0.11002638936042786, "learning_rate": 4.827246793127639e-05, "loss": 0.0121, "step": 10710 }, { "grad_norm": 0.1646372675895691, "learning_rate": 4.818984619988807e-05, "loss": 0.0129, "step": 10720 }, { "grad_norm": 0.09237392991781235, "learning_rate": 4.810722941739115e-05, "loss": 0.0136, "step": 10730 }, { "grad_norm": 0.12287517637014389, "learning_rate": 4.8024617809656684e-05, "loss": 0.0139, "step": 10740 }, { "grad_norm": 0.13128209114074707, "learning_rate": 4.794201160254171e-05, "loss": 0.0115, "step": 10750 }, { "grad_norm": 0.11448313295841217, "learning_rate": 4.785941102188844e-05, "loss": 0.0126, "step": 10760 }, { "grad_norm": 0.2137468457221985, "learning_rate": 4.7776816293523686e-05, "loss": 0.0114, "step": 10770 }, { "grad_norm": 0.09405191987752914, "learning_rate": 4.769422764325832e-05, "loss": 0.009, "step": 10780 }, { "grad_norm": 0.1201275885105133, "learning_rate": 4.76116452968865e-05, "loss": 0.0113, "step": 10790 }, { "grad_norm": 0.14066974818706512, "learning_rate": 4.752906948018525e-05, "loss": 0.0124, "step": 10800 }, { "grad_norm": 0.11676714569330215, "learning_rate": 4.7446500418913684e-05, "loss": 0.013, "step": 10810 }, { "grad_norm": 0.12355909496545792, "learning_rate": 4.736393833881247e-05, "loss": 0.011, "step": 10820 }, { "grad_norm": 0.09907860308885574, "learning_rate": 4.7281383465603194e-05, "loss": 0.0101, "step": 10830 }, { "grad_norm": 0.12503182888031006, "learning_rate": 4.71988360249877e-05, "loss": 0.0103, "step": 10840 }, { "grad_norm": 0.13462644815444946, "learning_rate": 4.7116296242647554e-05, "loss": 0.0133, "step": 10850 }, { "grad_norm": 0.11425887793302536, "learning_rate": 4.703376434424336e-05, "loss": 0.0107, "step": 10860 }, { "grad_norm": 0.13051722943782806, "learning_rate": 4.695124055541421e-05, "loss": 0.0117, "step": 10870 }, { "grad_norm": 0.13383111357688904, "learning_rate": 4.6868725101776934e-05, "loss": 0.0124, "step": 10880 }, { "grad_norm": 0.16682562232017517, "learning_rate": 4.678621820892567e-05, "loss": 0.0107, "step": 10890 }, { "grad_norm": 0.12564709782600403, "learning_rate": 4.670372010243111e-05, "loss": 0.0111, "step": 10900 }, { "grad_norm": 0.1308935284614563, "learning_rate": 4.662123100783992e-05, "loss": 0.0119, "step": 10910 }, { "grad_norm": 0.1478988230228424, "learning_rate": 4.653875115067415e-05, "loss": 0.0122, "step": 10920 }, { "grad_norm": 0.11631995439529419, "learning_rate": 4.6456280756430545e-05, "loss": 0.0125, "step": 10930 }, { "grad_norm": 0.15251383185386658, "learning_rate": 4.637382005058004e-05, "loss": 0.0117, "step": 10940 }, { "grad_norm": 0.16367854177951813, "learning_rate": 4.629136925856705e-05, "loss": 0.011, "step": 10950 }, { "grad_norm": 0.11026017367839813, "learning_rate": 4.6208928605808895e-05, "loss": 0.0103, "step": 10960 }, { "grad_norm": 0.1456904113292694, "learning_rate": 4.612649831769519e-05, "loss": 0.0138, "step": 10970 }, { "grad_norm": 0.1261056810617447, "learning_rate": 4.604407861958715e-05, "loss": 0.0112, "step": 10980 }, { "grad_norm": 0.1651112288236618, "learning_rate": 4.5961669736817114e-05, "loss": 0.0141, "step": 10990 }, { "grad_norm": 0.17818231880664825, "learning_rate": 4.5879271894687814e-05, "loss": 0.0109, "step": 11000 }, { "grad_norm": 0.1038869246840477, "learning_rate": 4.5796885318471826e-05, "loss": 0.0111, "step": 11010 }, { "grad_norm": 0.10409891605377197, "learning_rate": 4.571451023341086e-05, "loss": 0.0107, "step": 11020 }, { "grad_norm": 0.05496879294514656, "learning_rate": 4.563214686471527e-05, "loss": 0.0097, "step": 11030 }, { "grad_norm": 0.1322827935218811, "learning_rate": 4.5549795437563365e-05, "loss": 0.0107, "step": 11040 }, { "grad_norm": 0.12714450061321259, "learning_rate": 4.546745617710081e-05, "loss": 0.012, "step": 11050 }, { "grad_norm": 0.09634961932897568, "learning_rate": 4.5385129308440014e-05, "loss": 0.0131, "step": 11060 }, { "grad_norm": 0.08260700851678848, "learning_rate": 4.530281505665944e-05, "loss": 0.0098, "step": 11070 }, { "grad_norm": 0.09752333909273148, "learning_rate": 4.5220513646803134e-05, "loss": 0.0113, "step": 11080 }, { "grad_norm": 0.10862159729003906, "learning_rate": 4.513822530388003e-05, "loss": 0.0101, "step": 11090 }, { "grad_norm": 0.10010521113872528, "learning_rate": 4.5055950252863296e-05, "loss": 0.0112, "step": 11100 }, { "grad_norm": 0.14806623756885529, "learning_rate": 4.4973688718689803e-05, "loss": 0.0113, "step": 11110 }, { "grad_norm": 0.1333647519350052, "learning_rate": 4.4891440926259406e-05, "loss": 0.0121, "step": 11120 }, { "grad_norm": 0.15246905386447906, "learning_rate": 4.480920710043443e-05, "loss": 0.0122, "step": 11130 }, { "grad_norm": 0.09965476393699646, "learning_rate": 4.4726987466039044e-05, "loss": 0.0109, "step": 11140 }, { "grad_norm": 0.09680090099573135, "learning_rate": 4.46447822478586e-05, "loss": 0.0098, "step": 11150 }, { "grad_norm": 0.13520585000514984, "learning_rate": 4.4562591670638974e-05, "loss": 0.0121, "step": 11160 }, { "grad_norm": 0.14275327324867249, "learning_rate": 4.4480415959086105e-05, "loss": 0.0116, "step": 11170 }, { "grad_norm": 0.16175375878810883, "learning_rate": 4.439825533786522e-05, "loss": 0.0102, "step": 11180 }, { "grad_norm": 0.08839321881532669, "learning_rate": 4.431611003160035e-05, "loss": 0.0125, "step": 11190 }, { "grad_norm": 0.18216873705387115, "learning_rate": 4.4233980264873636e-05, "loss": 0.0132, "step": 11200 }, { "grad_norm": 0.14739346504211426, "learning_rate": 4.4151866262224684e-05, "loss": 0.0127, "step": 11210 }, { "grad_norm": 0.10047344863414764, "learning_rate": 4.406976824815006e-05, "loss": 0.0102, "step": 11220 }, { "grad_norm": 0.12273157387971878, "learning_rate": 4.3987686447102595e-05, "loss": 0.0165, "step": 11230 }, { "grad_norm": 0.13684073090553284, "learning_rate": 4.3905621083490804e-05, "loss": 0.0109, "step": 11240 }, { "grad_norm": 0.14631929993629456, "learning_rate": 4.3823572381678286e-05, "loss": 0.012, "step": 11250 }, { "grad_norm": 0.1356762945652008, "learning_rate": 4.374154056598301e-05, "loss": 0.0116, "step": 11260 }, { "grad_norm": 0.08261608332395554, "learning_rate": 4.3659525860676845e-05, "loss": 0.0109, "step": 11270 }, { "grad_norm": 0.14000627398490906, "learning_rate": 4.3577528489984854e-05, "loss": 0.0093, "step": 11280 }, { "grad_norm": 0.09692571312189102, "learning_rate": 4.349554867808476e-05, "loss": 0.0093, "step": 11290 }, { "grad_norm": 0.11413992196321487, "learning_rate": 4.34135866491062e-05, "loss": 0.0101, "step": 11300 }, { "grad_norm": 0.11533535271883011, "learning_rate": 4.333164262713022e-05, "loss": 0.0116, "step": 11310 }, { "grad_norm": 0.12516534328460693, "learning_rate": 4.324971683618868e-05, "loss": 0.0102, "step": 11320 }, { "grad_norm": 0.10780619084835052, "learning_rate": 4.316780950026354e-05, "loss": 0.011, "step": 11330 }, { "grad_norm": 0.14423823356628418, "learning_rate": 4.308592084328637e-05, "loss": 0.0113, "step": 11340 }, { "grad_norm": 0.11786215007305145, "learning_rate": 4.3004051089137576e-05, "loss": 0.0144, "step": 11350 }, { "grad_norm": 0.11672234535217285, "learning_rate": 4.292220046164597e-05, "loss": 0.0122, "step": 11360 }, { "grad_norm": 0.10574919730424881, "learning_rate": 4.2840369184588035e-05, "loss": 0.0134, "step": 11370 }, { "grad_norm": 0.1400466412305832, "learning_rate": 4.2758557481687345e-05, "loss": 0.0115, "step": 11380 }, { "grad_norm": 0.09881485998630524, "learning_rate": 4.267676557661403e-05, "loss": 0.0115, "step": 11390 }, { "grad_norm": 0.13815532624721527, "learning_rate": 4.2594993692983955e-05, "loss": 0.0142, "step": 11400 }, { "grad_norm": 0.09254142642021179, "learning_rate": 4.251324205435837e-05, "loss": 0.0126, "step": 11410 }, { "grad_norm": 0.12065224349498749, "learning_rate": 4.243151088424312e-05, "loss": 0.0101, "step": 11420 }, { "grad_norm": 0.11732470244169235, "learning_rate": 4.234980040608813e-05, "loss": 0.0107, "step": 11430 }, { "grad_norm": 0.1047016903758049, "learning_rate": 4.22681108432867e-05, "loss": 0.0097, "step": 11440 }, { "grad_norm": 0.11532041430473328, "learning_rate": 4.2186442419174984e-05, "loss": 0.011, "step": 11450 }, { "grad_norm": 0.14401431381702423, "learning_rate": 4.210479535703133e-05, "loss": 0.0117, "step": 11460 }, { "grad_norm": 0.10373225808143616, "learning_rate": 4.202316988007567e-05, "loss": 0.011, "step": 11470 }, { "grad_norm": 0.1150376945734024, "learning_rate": 4.194156621146901e-05, "loss": 0.0112, "step": 11480 }, { "grad_norm": 0.13139379024505615, "learning_rate": 4.1859984574312596e-05, "loss": 0.0105, "step": 11490 }, { "grad_norm": 0.11595311760902405, "learning_rate": 4.177842519164752e-05, "loss": 0.0106, "step": 11500 }, { "grad_norm": 0.11034642159938812, "learning_rate": 4.169688828645404e-05, "loss": 0.0115, "step": 11510 }, { "grad_norm": 0.12707245349884033, "learning_rate": 4.161537408165092e-05, "loss": 0.0119, "step": 11520 }, { "grad_norm": 0.12673160433769226, "learning_rate": 4.1533882800094924e-05, "loss": 0.0131, "step": 11530 }, { "grad_norm": 0.0931282490491867, "learning_rate": 4.145241466458005e-05, "loss": 0.0132, "step": 11540 }, { "grad_norm": 0.11265479028224945, "learning_rate": 4.13709698978371e-05, "loss": 0.0125, "step": 11550 }, { "grad_norm": 0.162606880068779, "learning_rate": 4.1289548722532944e-05, "loss": 0.0119, "step": 11560 }, { "grad_norm": 0.13939565420150757, "learning_rate": 4.120815136126999e-05, "loss": 0.0102, "step": 11570 }, { "grad_norm": 0.12251929938793182, "learning_rate": 4.112677803658548e-05, "loss": 0.0133, "step": 11580 }, { "grad_norm": 0.1161055639386177, "learning_rate": 4.1045428970951e-05, "loss": 0.0113, "step": 11590 }, { "grad_norm": 0.11543667316436768, "learning_rate": 4.0964104386771785e-05, "loss": 0.0105, "step": 11600 }, { "grad_norm": 0.12905564904212952, "learning_rate": 4.0882804506386144e-05, "loss": 0.0133, "step": 11610 }, { "grad_norm": 0.14158843457698822, "learning_rate": 4.080152955206485e-05, "loss": 0.0152, "step": 11620 }, { "grad_norm": 0.17138244211673737, "learning_rate": 4.0720279746010505e-05, "loss": 0.0141, "step": 11630 }, { "grad_norm": 0.11947762966156006, "learning_rate": 4.063905531035699e-05, "loss": 0.0104, "step": 11640 }, { "grad_norm": 0.12572987377643585, "learning_rate": 4.055785646716882e-05, "loss": 0.0119, "step": 11650 }, { "grad_norm": 0.11250671744346619, "learning_rate": 4.047668343844051e-05, "loss": 0.0112, "step": 11660 }, { "grad_norm": 0.13423985242843628, "learning_rate": 4.039553644609604e-05, "loss": 0.0119, "step": 11670 }, { "grad_norm": 0.11347131431102753, "learning_rate": 4.0314415711988176e-05, "loss": 0.0095, "step": 11680 }, { "grad_norm": 0.14366436004638672, "learning_rate": 4.023332145789792e-05, "loss": 0.0099, "step": 11690 }, { "grad_norm": 0.11896447837352753, "learning_rate": 4.015225390553385e-05, "loss": 0.0127, "step": 11700 }, { "grad_norm": 0.18747885525226593, "learning_rate": 4.007121327653158e-05, "loss": 0.0101, "step": 11710 }, { "grad_norm": 0.12582837045192719, "learning_rate": 3.9990199792453064e-05, "loss": 0.0103, "step": 11720 }, { "grad_norm": 0.10064931958913803, "learning_rate": 3.9909213674786103e-05, "loss": 0.012, "step": 11730 }, { "grad_norm": 0.13999265432357788, "learning_rate": 3.982825514494363e-05, "loss": 0.0118, "step": 11740 }, { "grad_norm": 0.10877490788698196, "learning_rate": 3.974732442426319e-05, "loss": 0.0113, "step": 11750 }, { "grad_norm": 0.11761355400085449, "learning_rate": 3.966642173400629e-05, "loss": 0.0115, "step": 11760 }, { "grad_norm": 0.11996804922819138, "learning_rate": 3.9585547295357764e-05, "loss": 0.0115, "step": 11770 }, { "grad_norm": 0.08869068324565887, "learning_rate": 3.950470132942526e-05, "loss": 0.0087, "step": 11780 }, { "grad_norm": 0.12849071621894836, "learning_rate": 3.942388405723856e-05, "loss": 0.0113, "step": 11790 }, { "grad_norm": 0.15727119147777557, "learning_rate": 3.9343095699749e-05, "loss": 0.0119, "step": 11800 }, { "grad_norm": 0.14103196561336517, "learning_rate": 3.9262336477828874e-05, "loss": 0.0122, "step": 11810 }, { "grad_norm": 0.16075658798217773, "learning_rate": 3.9181606612270794e-05, "loss": 0.0121, "step": 11820 }, { "grad_norm": 0.08164077997207642, "learning_rate": 3.910090632378713e-05, "loss": 0.0117, "step": 11830 }, { "grad_norm": 0.15798579156398773, "learning_rate": 3.90202358330094e-05, "loss": 0.0112, "step": 11840 }, { "grad_norm": 0.13368584215641022, "learning_rate": 3.8939595360487656e-05, "loss": 0.0107, "step": 11850 }, { "grad_norm": 0.11600898951292038, "learning_rate": 3.885898512668984e-05, "loss": 0.0089, "step": 11860 }, { "grad_norm": 0.0996062308549881, "learning_rate": 3.877840535200127e-05, "loss": 0.0122, "step": 11870 }, { "grad_norm": 0.10543225705623627, "learning_rate": 3.869785625672397e-05, "loss": 0.0121, "step": 11880 }, { "grad_norm": 0.12774696946144104, "learning_rate": 3.8617338061076094e-05, "loss": 0.0112, "step": 11890 }, { "grad_norm": 0.11575794219970703, "learning_rate": 3.853685098519132e-05, "loss": 0.0114, "step": 11900 }, { "grad_norm": 0.10708192735910416, "learning_rate": 3.845639524911823e-05, "loss": 0.0112, "step": 11910 }, { "grad_norm": 0.1381128877401352, "learning_rate": 3.837597107281974e-05, "loss": 0.0117, "step": 11920 }, { "grad_norm": 0.11683300882577896, "learning_rate": 3.829557867617247e-05, "loss": 0.0138, "step": 11930 }, { "grad_norm": 0.15777988731861115, "learning_rate": 3.821521827896618e-05, "loss": 0.0152, "step": 11940 }, { "grad_norm": 0.11536698043346405, "learning_rate": 3.81348901009031e-05, "loss": 0.0116, "step": 11950 }, { "grad_norm": 0.08959528803825378, "learning_rate": 3.805459436159741e-05, "loss": 0.0125, "step": 11960 }, { "grad_norm": 0.08846994489431381, "learning_rate": 3.797433128057461e-05, "loss": 0.0098, "step": 11970 }, { "grad_norm": 0.11954237520694733, "learning_rate": 3.789410107727089e-05, "loss": 0.0112, "step": 11980 }, { "grad_norm": 0.1141890436410904, "learning_rate": 3.781390397103257e-05, "loss": 0.0094, "step": 11990 }, { "grad_norm": 0.11511945724487305, "learning_rate": 3.7733740181115455e-05, "loss": 0.0131, "step": 12000 }, { "grad_norm": 0.121440589427948, "learning_rate": 3.7653609926684306e-05, "loss": 0.0134, "step": 12010 }, { "grad_norm": 0.093757763504982, "learning_rate": 3.757351342681217e-05, "loss": 0.01, "step": 12020 }, { "grad_norm": 0.13121575117111206, "learning_rate": 3.749345090047982e-05, "loss": 0.0101, "step": 12030 }, { "grad_norm": 0.06732475757598877, "learning_rate": 3.741342256657515e-05, "loss": 0.0096, "step": 12040 }, { "grad_norm": 0.080066978931427, "learning_rate": 3.7333428643892567e-05, "loss": 0.0098, "step": 12050 }, { "grad_norm": 0.17739364504814148, "learning_rate": 3.725346935113239e-05, "loss": 0.0102, "step": 12060 }, { "grad_norm": 0.15182557702064514, "learning_rate": 3.717354490690029e-05, "loss": 0.0088, "step": 12070 }, { "grad_norm": 0.10810228437185287, "learning_rate": 3.709365552970664e-05, "loss": 0.0094, "step": 12080 }, { "grad_norm": 0.0786350890994072, "learning_rate": 3.7013801437965945e-05, "loss": 0.0095, "step": 12090 }, { "grad_norm": 0.10364408791065216, "learning_rate": 3.693398284999623e-05, "loss": 0.0115, "step": 12100 }, { "grad_norm": 0.1293162703514099, "learning_rate": 3.6854199984018484e-05, "loss": 0.01, "step": 12110 }, { "grad_norm": 0.1895095854997635, "learning_rate": 3.677445305815601e-05, "loss": 0.0112, "step": 12120 }, { "grad_norm": 0.09717131406068802, "learning_rate": 3.669474229043387e-05, "loss": 0.0101, "step": 12130 }, { "grad_norm": 0.11882030963897705, "learning_rate": 3.6615067898778235e-05, "loss": 0.0099, "step": 12140 }, { "grad_norm": 0.14968472719192505, "learning_rate": 3.6535430101015866e-05, "loss": 0.0095, "step": 12150 }, { "grad_norm": 0.11034590750932693, "learning_rate": 3.645582911487345e-05, "loss": 0.0103, "step": 12160 }, { "grad_norm": 0.11888076364994049, "learning_rate": 3.637626515797706e-05, "loss": 0.0088, "step": 12170 }, { "grad_norm": 0.14004234969615936, "learning_rate": 3.629673844785152e-05, "loss": 0.0116, "step": 12180 }, { "grad_norm": 0.11913584917783737, "learning_rate": 3.621724920191979e-05, "loss": 0.009, "step": 12190 }, { "grad_norm": 0.08658146858215332, "learning_rate": 3.6137797637502444e-05, "loss": 0.0115, "step": 12200 }, { "grad_norm": 0.18775847554206848, "learning_rate": 3.6058383971817035e-05, "loss": 0.0122, "step": 12210 }, { "grad_norm": 0.09732454270124435, "learning_rate": 3.59790084219775e-05, "loss": 0.0077, "step": 12220 }, { "grad_norm": 0.07154260575771332, "learning_rate": 3.589967120499353e-05, "loss": 0.0101, "step": 12230 }, { "grad_norm": 0.13687583804130554, "learning_rate": 3.5820372537770075e-05, "loss": 0.0108, "step": 12240 }, { "grad_norm": 0.07548828423023224, "learning_rate": 3.5741112637106655e-05, "loss": 0.0078, "step": 12250 }, { "grad_norm": 0.11854778975248337, "learning_rate": 3.5661891719696804e-05, "loss": 0.0092, "step": 12260 }, { "grad_norm": 0.1722133457660675, "learning_rate": 3.5582710002127504e-05, "loss": 0.0125, "step": 12270 }, { "grad_norm": 0.06886011362075806, "learning_rate": 3.550356770087853e-05, "loss": 0.008, "step": 12280 }, { "grad_norm": 0.0962163656949997, "learning_rate": 3.5424465032321914e-05, "loss": 0.0069, "step": 12290 }, { "grad_norm": 0.09341788291931152, "learning_rate": 3.5345402212721335e-05, "loss": 0.0132, "step": 12300 }, { "grad_norm": 0.16465502977371216, "learning_rate": 3.526637945823152e-05, "loss": 0.0125, "step": 12310 }, { "grad_norm": 0.1150645911693573, "learning_rate": 3.518739698489767e-05, "loss": 0.0091, "step": 12320 }, { "grad_norm": 0.10335718095302582, "learning_rate": 3.510845500865485e-05, "loss": 0.0101, "step": 12330 }, { "grad_norm": 0.1562296450138092, "learning_rate": 3.502955374532739e-05, "loss": 0.0107, "step": 12340 }, { "grad_norm": 0.12020246684551239, "learning_rate": 3.495069341062836e-05, "loss": 0.0117, "step": 12350 }, { "grad_norm": 0.10088761150836945, "learning_rate": 3.4871874220158896e-05, "loss": 0.0091, "step": 12360 }, { "grad_norm": 0.09826041013002396, "learning_rate": 3.479309638940762e-05, "loss": 0.0089, "step": 12370 }, { "grad_norm": 0.07779894769191742, "learning_rate": 3.4714360133750146e-05, "loss": 0.0089, "step": 12380 }, { "grad_norm": 0.09035561978816986, "learning_rate": 3.463566566844839e-05, "loss": 0.0092, "step": 12390 }, { "grad_norm": 0.14518152177333832, "learning_rate": 3.4557013208650016e-05, "loss": 0.0097, "step": 12400 }, { "grad_norm": 0.11916865408420563, "learning_rate": 3.4478402969387857e-05, "loss": 0.0086, "step": 12410 }, { "grad_norm": 0.07842212170362473, "learning_rate": 3.4399835165579266e-05, "loss": 0.0116, "step": 12420 }, { "grad_norm": 0.11167407780885696, "learning_rate": 3.4321310012025645e-05, "loss": 0.0131, "step": 12430 }, { "grad_norm": 0.09849042445421219, "learning_rate": 3.424282772341176e-05, "loss": 0.0085, "step": 12440 }, { "grad_norm": 0.14067302644252777, "learning_rate": 3.416438851430519e-05, "loss": 0.0078, "step": 12450 }, { "grad_norm": 0.09264612197875977, "learning_rate": 3.408599259915577e-05, "loss": 0.0084, "step": 12460 }, { "grad_norm": 0.12152516096830368, "learning_rate": 3.400764019229487e-05, "loss": 0.0119, "step": 12470 }, { "grad_norm": 0.10386856645345688, "learning_rate": 3.3929331507935035e-05, "loss": 0.0082, "step": 12480 }, { "grad_norm": 0.09282225370407104, "learning_rate": 3.3851066760169196e-05, "loss": 0.0129, "step": 12490 }, { "grad_norm": 0.11006038635969162, "learning_rate": 3.377284616297021e-05, "loss": 0.0089, "step": 12500 }, { "grad_norm": 0.1531502902507782, "learning_rate": 3.3694669930190166e-05, "loss": 0.0122, "step": 12510 }, { "grad_norm": 0.16491395235061646, "learning_rate": 3.36165382755599e-05, "loss": 0.0076, "step": 12520 }, { "grad_norm": 0.16175046563148499, "learning_rate": 3.35384514126884e-05, "loss": 0.0103, "step": 12530 }, { "grad_norm": 0.11994735151529312, "learning_rate": 3.3460409555062154e-05, "loss": 0.0077, "step": 12540 }, { "grad_norm": 0.18081294000148773, "learning_rate": 3.3382412916044645e-05, "loss": 0.0098, "step": 12550 }, { "grad_norm": 0.11653721332550049, "learning_rate": 3.330446170887566e-05, "loss": 0.0119, "step": 12560 }, { "grad_norm": 0.17638236284255981, "learning_rate": 3.3226556146670834e-05, "loss": 0.0095, "step": 12570 }, { "grad_norm": 0.10515972226858139, "learning_rate": 3.314869644242102e-05, "loss": 0.0113, "step": 12580 }, { "grad_norm": 0.08748297393321991, "learning_rate": 3.3070882808991674e-05, "loss": 0.0092, "step": 12590 }, { "grad_norm": 0.10160213708877563, "learning_rate": 3.2993115459122305e-05, "loss": 0.0094, "step": 12600 }, { "grad_norm": 0.09474991261959076, "learning_rate": 3.2915394605425835e-05, "loss": 0.0074, "step": 12610 }, { "grad_norm": 0.10275330394506454, "learning_rate": 3.283772046038816e-05, "loss": 0.0083, "step": 12620 }, { "grad_norm": 0.1062268614768982, "learning_rate": 3.276009323636739e-05, "loss": 0.0098, "step": 12630 }, { "grad_norm": 0.1300700306892395, "learning_rate": 3.268251314559344e-05, "loss": 0.01, "step": 12640 }, { "grad_norm": 0.08797676861286163, "learning_rate": 3.2604980400167254e-05, "loss": 0.0086, "step": 12650 }, { "grad_norm": 0.10255326330661774, "learning_rate": 3.252749521206042e-05, "loss": 0.0098, "step": 12660 }, { "grad_norm": 0.09169215708971024, "learning_rate": 3.2450057793114494e-05, "loss": 0.009, "step": 12670 }, { "grad_norm": 0.09809188544750214, "learning_rate": 3.2372668355040435e-05, "loss": 0.0094, "step": 12680 }, { "grad_norm": 0.1251409649848938, "learning_rate": 3.2295327109418005e-05, "loss": 0.0127, "step": 12690 }, { "grad_norm": 0.09874588251113892, "learning_rate": 3.221803426769518e-05, "loss": 0.0078, "step": 12700 }, { "grad_norm": 0.1396438181400299, "learning_rate": 3.214079004118768e-05, "loss": 0.01, "step": 12710 }, { "grad_norm": 0.10899023711681366, "learning_rate": 3.2063594641078234e-05, "loss": 0.0122, "step": 12720 }, { "grad_norm": 0.14221332967281342, "learning_rate": 3.198644827841616e-05, "loss": 0.0121, "step": 12730 }, { "grad_norm": 0.10441400110721588, "learning_rate": 3.1909351164116654e-05, "loss": 0.0083, "step": 12740 }, { "grad_norm": 0.09342914819717407, "learning_rate": 3.183230350896026e-05, "loss": 0.0096, "step": 12750 }, { "grad_norm": 0.09429465234279633, "learning_rate": 3.1755305523592337e-05, "loss": 0.0108, "step": 12760 }, { "grad_norm": 0.3735600411891937, "learning_rate": 3.167835741852245e-05, "loss": 0.0119, "step": 12770 }, { "grad_norm": 0.11392326653003693, "learning_rate": 3.160145940412378e-05, "loss": 0.0104, "step": 12780 }, { "grad_norm": 0.10871228575706482, "learning_rate": 3.1524611690632545e-05, "loss": 0.011, "step": 12790 }, { "grad_norm": 0.11359618604183197, "learning_rate": 3.144781448814746e-05, "loss": 0.0108, "step": 12800 }, { "grad_norm": 0.09116849303245544, "learning_rate": 3.1371068006629145e-05, "loss": 0.0094, "step": 12810 }, { "grad_norm": 0.13012342154979706, "learning_rate": 3.129437245589956e-05, "loss": 0.0113, "step": 12820 }, { "grad_norm": 0.09033514559268951, "learning_rate": 3.121772804564143e-05, "loss": 0.009, "step": 12830 }, { "grad_norm": 0.12001784890890121, "learning_rate": 3.11411349853976e-05, "loss": 0.0094, "step": 12840 }, { "grad_norm": 0.12029877305030823, "learning_rate": 3.10645934845706e-05, "loss": 0.0099, "step": 12850 }, { "grad_norm": 0.1052558571100235, "learning_rate": 3.098810375242196e-05, "loss": 0.0092, "step": 12860 }, { "grad_norm": 0.1081351637840271, "learning_rate": 3.0911665998071704e-05, "loss": 0.0097, "step": 12870 }, { "grad_norm": 0.08347229659557343, "learning_rate": 3.083528043049774e-05, "loss": 0.0083, "step": 12880 }, { "grad_norm": 0.1347927451133728, "learning_rate": 3.0758947258535255e-05, "loss": 0.0101, "step": 12890 }, { "grad_norm": 0.11204929649829865, "learning_rate": 3.068266669087625e-05, "loss": 0.0077, "step": 12900 }, { "grad_norm": 0.08387834578752518, "learning_rate": 3.060643893606887e-05, "loss": 0.0071, "step": 12910 }, { "grad_norm": 0.08360913395881653, "learning_rate": 3.053026420251693e-05, "loss": 0.0103, "step": 12920 }, { "grad_norm": 0.09632116556167603, "learning_rate": 3.0454142698479183e-05, "loss": 0.0114, "step": 12930 }, { "grad_norm": 0.09380485862493515, "learning_rate": 3.0378074632068954e-05, "loss": 0.0088, "step": 12940 }, { "grad_norm": 0.09574024379253387, "learning_rate": 3.0302060211253408e-05, "loss": 0.0095, "step": 12950 }, { "grad_norm": 0.2156762331724167, "learning_rate": 3.0226099643853073e-05, "loss": 0.0077, "step": 12960 }, { "grad_norm": 0.0779445618391037, "learning_rate": 3.0150193137541283e-05, "loss": 0.0082, "step": 12970 }, { "grad_norm": 0.08626121282577515, "learning_rate": 3.0074340899843467e-05, "loss": 0.009, "step": 12980 }, { "grad_norm": 0.09372642636299133, "learning_rate": 2.999854313813677e-05, "loss": 0.0095, "step": 12990 }, { "grad_norm": 0.15425384044647217, "learning_rate": 2.9922800059649382e-05, "loss": 0.0107, "step": 13000 }, { "grad_norm": 0.124394990503788, "learning_rate": 2.9847111871459976e-05, "loss": 0.0091, "step": 13010 }, { "grad_norm": 0.10490158945322037, "learning_rate": 2.977147878049721e-05, "loss": 0.0104, "step": 13020 }, { "grad_norm": 0.0882851704955101, "learning_rate": 2.9695900993539006e-05, "loss": 0.0098, "step": 13030 }, { "grad_norm": 0.09919983893632889, "learning_rate": 2.9620378717212183e-05, "loss": 0.0086, "step": 13040 }, { "grad_norm": 0.08302854001522064, "learning_rate": 2.9544912157991745e-05, "loss": 0.0109, "step": 13050 }, { "grad_norm": 0.12182004004716873, "learning_rate": 2.9469501522200405e-05, "loss": 0.0094, "step": 13060 }, { "grad_norm": 0.08473879843950272, "learning_rate": 2.9394147016007946e-05, "loss": 0.0077, "step": 13070 }, { "grad_norm": 0.19055606424808502, "learning_rate": 2.9318848845430702e-05, "loss": 0.0106, "step": 13080 }, { "grad_norm": 0.10733821243047714, "learning_rate": 2.9243607216331013e-05, "loss": 0.0106, "step": 13090 }, { "grad_norm": 0.20864446461200714, "learning_rate": 2.916842233441661e-05, "loss": 0.0101, "step": 13100 }, { "grad_norm": 0.11460768431425095, "learning_rate": 2.90932944052401e-05, "loss": 0.0099, "step": 13110 }, { "grad_norm": 0.09573196619749069, "learning_rate": 2.9018223634198354e-05, "loss": 0.01, "step": 13120 }, { "grad_norm": 0.08526074886322021, "learning_rate": 2.8943210226532025e-05, "loss": 0.0095, "step": 13130 }, { "grad_norm": 0.09101131558418274, "learning_rate": 2.8868254387324857e-05, "loss": 0.0078, "step": 13140 }, { "grad_norm": 0.11679825186729431, "learning_rate": 2.8793356321503306e-05, "loss": 0.009, "step": 13150 }, { "grad_norm": 0.07848545908927917, "learning_rate": 2.87185162338358e-05, "loss": 0.0074, "step": 13160 }, { "grad_norm": 0.08281484246253967, "learning_rate": 2.8643734328932253e-05, "loss": 0.009, "step": 13170 }, { "grad_norm": 0.07693672925233841, "learning_rate": 2.856901081124359e-05, "loss": 0.009, "step": 13180 }, { "grad_norm": 0.1631159782409668, "learning_rate": 2.8494345885061002e-05, "loss": 0.01, "step": 13190 }, { "grad_norm": 0.10857614874839783, "learning_rate": 2.8419739754515616e-05, "loss": 0.0063, "step": 13200 }, { "grad_norm": 0.15411965548992157, "learning_rate": 2.8345192623577666e-05, "loss": 0.0084, "step": 13210 }, { "grad_norm": 0.09344217926263809, "learning_rate": 2.8270704696056193e-05, "loss": 0.0081, "step": 13220 }, { "grad_norm": 0.09495799243450165, "learning_rate": 2.8196276175598367e-05, "loss": 0.0097, "step": 13230 }, { "grad_norm": 0.10831760615110397, "learning_rate": 2.8121907265688884e-05, "loss": 0.0072, "step": 13240 }, { "grad_norm": 0.06629833579063416, "learning_rate": 2.804759816964957e-05, "loss": 0.0085, "step": 13250 }, { "grad_norm": 0.07908910512924194, "learning_rate": 2.797334909063857e-05, "loss": 0.0098, "step": 13260 }, { "grad_norm": 0.06599361449480057, "learning_rate": 2.7899160231650056e-05, "loss": 0.0096, "step": 13270 }, { "grad_norm": 0.09079551696777344, "learning_rate": 2.7825031795513585e-05, "loss": 0.0077, "step": 13280 }, { "grad_norm": 0.09553603082895279, "learning_rate": 2.775096398489341e-05, "loss": 0.0107, "step": 13290 }, { "grad_norm": 0.12538474798202515, "learning_rate": 2.7676957002288163e-05, "loss": 0.0104, "step": 13300 }, { "grad_norm": 0.12555016577243805, "learning_rate": 2.760301105003003e-05, "loss": 0.0084, "step": 13310 }, { "grad_norm": 0.14150360226631165, "learning_rate": 2.752912633028446e-05, "loss": 0.0132, "step": 13320 }, { "grad_norm": 0.12843383848667145, "learning_rate": 2.7455303045049474e-05, "loss": 0.0113, "step": 13330 }, { "grad_norm": 0.1276029348373413, "learning_rate": 2.7381541396155098e-05, "loss": 0.0082, "step": 13340 }, { "grad_norm": 0.11665509641170502, "learning_rate": 2.730784158526286e-05, "loss": 0.0085, "step": 13350 }, { "grad_norm": 0.16628283262252808, "learning_rate": 2.723420381386521e-05, "loss": 0.0109, "step": 13360 }, { "grad_norm": 0.09441974759101868, "learning_rate": 2.7160628283285018e-05, "loss": 0.0087, "step": 13370 }, { "grad_norm": 0.10430309921503067, "learning_rate": 2.7087115194675007e-05, "loss": 0.008, "step": 13380 }, { "grad_norm": 0.11628322303295135, "learning_rate": 2.701366474901712e-05, "loss": 0.0079, "step": 13390 }, { "grad_norm": 0.10232027620077133, "learning_rate": 2.6940277147122085e-05, "loss": 0.0092, "step": 13400 }, { "grad_norm": 0.12912708520889282, "learning_rate": 2.686695258962878e-05, "loss": 0.009, "step": 13410 }, { "grad_norm": 0.08580455929040909, "learning_rate": 2.679369127700375e-05, "loss": 0.0099, "step": 13420 }, { "grad_norm": 0.1467631608247757, "learning_rate": 2.672049340954067e-05, "loss": 0.0114, "step": 13430 }, { "grad_norm": 0.1057540625333786, "learning_rate": 2.6647359187359676e-05, "loss": 0.0102, "step": 13440 }, { "grad_norm": 0.12082751095294952, "learning_rate": 2.6574288810406946e-05, "loss": 0.008, "step": 13450 }, { "grad_norm": 0.11722404509782791, "learning_rate": 2.6501282478454083e-05, "loss": 0.0084, "step": 13460 }, { "grad_norm": 0.11409171670675278, "learning_rate": 2.6428340391097618e-05, "loss": 0.0113, "step": 13470 }, { "grad_norm": 0.08317497372627258, "learning_rate": 2.6355462747758485e-05, "loss": 0.0097, "step": 13480 }, { "grad_norm": 0.08932055532932281, "learning_rate": 2.6282649747681304e-05, "loss": 0.0076, "step": 13490 }, { "grad_norm": 0.0890669971704483, "learning_rate": 2.620990158993406e-05, "loss": 0.0086, "step": 13500 }, { "grad_norm": 0.11296665668487549, "learning_rate": 2.6137218473407477e-05, "loss": 0.0091, "step": 13510 }, { "grad_norm": 0.16163450479507446, "learning_rate": 2.606460059681436e-05, "loss": 0.0072, "step": 13520 }, { "grad_norm": 0.1007835641503334, "learning_rate": 2.599204815868928e-05, "loss": 0.0075, "step": 13530 }, { "grad_norm": 0.12690286338329315, "learning_rate": 2.5919561357387756e-05, "loss": 0.0091, "step": 13540 }, { "grad_norm": 0.08488639444112778, "learning_rate": 2.5847140391085972e-05, "loss": 0.0118, "step": 13550 }, { "grad_norm": 0.11314214766025543, "learning_rate": 2.5774785457780103e-05, "loss": 0.008, "step": 13560 }, { "grad_norm": 0.1324324756860733, "learning_rate": 2.5702496755285753e-05, "loss": 0.0104, "step": 13570 }, { "grad_norm": 0.12775099277496338, "learning_rate": 2.5630274481237483e-05, "loss": 0.0086, "step": 13580 }, { "grad_norm": 0.12558342516422272, "learning_rate": 2.5558118833088197e-05, "loss": 0.0123, "step": 13590 }, { "grad_norm": 0.07108517736196518, "learning_rate": 2.548603000810872e-05, "loss": 0.0095, "step": 13600 }, { "grad_norm": 0.12291180342435837, "learning_rate": 2.5414008203387152e-05, "loss": 0.0076, "step": 13610 }, { "grad_norm": 0.07368358969688416, "learning_rate": 2.534205361582834e-05, "loss": 0.0069, "step": 13620 }, { "grad_norm": 0.10483675450086594, "learning_rate": 2.527016644215338e-05, "loss": 0.0097, "step": 13630 }, { "grad_norm": 0.09949961304664612, "learning_rate": 2.519834687889905e-05, "loss": 0.0089, "step": 13640 }, { "grad_norm": 0.11482270061969757, "learning_rate": 2.5126595122417295e-05, "loss": 0.0102, "step": 13650 }, { "grad_norm": 0.08015670627355576, "learning_rate": 2.5054911368874713e-05, "loss": 0.0082, "step": 13660 }, { "grad_norm": 0.10067854821681976, "learning_rate": 2.4983295814251916e-05, "loss": 0.0081, "step": 13670 }, { "grad_norm": 0.14506493508815765, "learning_rate": 2.4911748654343105e-05, "loss": 0.0107, "step": 13680 }, { "grad_norm": 0.10510905086994171, "learning_rate": 2.4840270084755463e-05, "loss": 0.0091, "step": 13690 }, { "grad_norm": 0.09186901897192001, "learning_rate": 2.4768860300908685e-05, "loss": 0.0091, "step": 13700 }, { "grad_norm": 0.08813094347715378, "learning_rate": 2.469751949803443e-05, "loss": 0.0094, "step": 13710 }, { "grad_norm": 0.0709739699959755, "learning_rate": 2.4626247871175666e-05, "loss": 0.0069, "step": 13720 }, { "grad_norm": 0.08740218728780746, "learning_rate": 2.4555045615186346e-05, "loss": 0.0109, "step": 13730 }, { "grad_norm": 0.10533691197633743, "learning_rate": 2.4483912924730677e-05, "loss": 0.0078, "step": 13740 }, { "grad_norm": 0.13945093750953674, "learning_rate": 2.4412849994282742e-05, "loss": 0.011, "step": 13750 }, { "grad_norm": 0.12541213631629944, "learning_rate": 2.434185701812592e-05, "loss": 0.0087, "step": 13760 }, { "grad_norm": 0.08571644872426987, "learning_rate": 2.4270934190352218e-05, "loss": 0.0064, "step": 13770 }, { "grad_norm": 0.08555253595113754, "learning_rate": 2.4200081704861998e-05, "loss": 0.007, "step": 13780 }, { "grad_norm": 0.06634468585252762, "learning_rate": 2.412929975536321e-05, "loss": 0.0079, "step": 13790 }, { "grad_norm": 0.10664807260036469, "learning_rate": 2.4058588535371017e-05, "loss": 0.0072, "step": 13800 }, { "grad_norm": 0.10927361249923706, "learning_rate": 2.3987948238207243e-05, "loss": 0.0078, "step": 13810 }, { "grad_norm": 0.10451497882604599, "learning_rate": 2.3917379056999678e-05, "loss": 0.0087, "step": 13820 }, { "grad_norm": 0.1398306041955948, "learning_rate": 2.3846881184681824e-05, "loss": 0.0092, "step": 13830 }, { "grad_norm": 0.13651806116104126, "learning_rate": 2.377645481399214e-05, "loss": 0.0089, "step": 13840 }, { "grad_norm": 0.08929361402988434, "learning_rate": 2.3706100137473667e-05, "loss": 0.0071, "step": 13850 }, { "grad_norm": 0.08114518225193024, "learning_rate": 2.3635817347473394e-05, "loss": 0.0069, "step": 13860 }, { "grad_norm": 0.0828942134976387, "learning_rate": 2.3565606636141757e-05, "loss": 0.0098, "step": 13870 }, { "grad_norm": 0.13303370773792267, "learning_rate": 2.3495468195432203e-05, "loss": 0.0086, "step": 13880 }, { "grad_norm": 0.09444092214107513, "learning_rate": 2.3425402217100507e-05, "loss": 0.0082, "step": 13890 }, { "grad_norm": 0.08021314442157745, "learning_rate": 2.3355408892704424e-05, "loss": 0.007, "step": 13900 }, { "grad_norm": 0.12002775073051453, "learning_rate": 2.3285488413603003e-05, "loss": 0.0099, "step": 13910 }, { "grad_norm": 0.09221331775188446, "learning_rate": 2.321564097095615e-05, "loss": 0.0076, "step": 13920 }, { "grad_norm": 0.08415579795837402, "learning_rate": 2.3145866755724142e-05, "loss": 0.0068, "step": 13930 }, { "grad_norm": 0.1021721288561821, "learning_rate": 2.307616595866699e-05, "loss": 0.0072, "step": 13940 }, { "grad_norm": 0.09167303889989853, "learning_rate": 2.3006538770344032e-05, "loss": 0.0084, "step": 13950 }, { "grad_norm": 0.08627080917358398, "learning_rate": 2.293698538111334e-05, "loss": 0.0082, "step": 13960 }, { "grad_norm": 0.12373010069131851, "learning_rate": 2.28675059811312e-05, "loss": 0.0096, "step": 13970 }, { "grad_norm": 0.11642234027385712, "learning_rate": 2.279810076035167e-05, "loss": 0.008, "step": 13980 }, { "grad_norm": 0.22530335187911987, "learning_rate": 2.272876990852596e-05, "loss": 0.0075, "step": 13990 }, { "grad_norm": 0.08689238876104355, "learning_rate": 2.265951361520195e-05, "loss": 0.006, "step": 14000 }, { "grad_norm": 0.14653277397155762, "learning_rate": 2.2590332069723748e-05, "loss": 0.0087, "step": 14010 }, { "grad_norm": 0.11313546448945999, "learning_rate": 2.2521225461231004e-05, "loss": 0.0098, "step": 14020 }, { "grad_norm": 0.08564630895853043, "learning_rate": 2.2452193978658597e-05, "loss": 0.0071, "step": 14030 }, { "grad_norm": 0.06725704669952393, "learning_rate": 2.238323781073594e-05, "loss": 0.0061, "step": 14040 }, { "grad_norm": 0.09111366420984268, "learning_rate": 2.2314357145986552e-05, "loss": 0.0096, "step": 14050 }, { "grad_norm": 0.09331098943948746, "learning_rate": 2.224555217272757e-05, "loss": 0.0072, "step": 14060 }, { "grad_norm": 0.06564987450838089, "learning_rate": 2.2176823079069127e-05, "loss": 0.0058, "step": 14070 }, { "grad_norm": 0.0880548506975174, "learning_rate": 2.210817005291398e-05, "loss": 0.0059, "step": 14080 }, { "grad_norm": 0.09144202619791031, "learning_rate": 2.203959328195686e-05, "loss": 0.0068, "step": 14090 }, { "grad_norm": 0.11451786011457443, "learning_rate": 2.1971092953684026e-05, "loss": 0.0072, "step": 14100 }, { "grad_norm": 0.1506766378879547, "learning_rate": 2.1902669255372788e-05, "loss": 0.008, "step": 14110 }, { "grad_norm": 0.06353414803743362, "learning_rate": 2.1834322374090897e-05, "loss": 0.0086, "step": 14120 }, { "grad_norm": 0.11372896283864975, "learning_rate": 2.1766052496696153e-05, "loss": 0.0075, "step": 14130 }, { "grad_norm": 0.12518884241580963, "learning_rate": 2.169785980983577e-05, "loss": 0.0077, "step": 14140 }, { "grad_norm": 0.08899479359388351, "learning_rate": 2.162974449994593e-05, "loss": 0.0062, "step": 14150 }, { "grad_norm": 0.1005115807056427, "learning_rate": 2.1561706753251337e-05, "loss": 0.0081, "step": 14160 }, { "grad_norm": 0.13001303374767303, "learning_rate": 2.1493746755764544e-05, "loss": 0.0088, "step": 14170 }, { "grad_norm": 0.10268504917621613, "learning_rate": 2.1425864693285635e-05, "loss": 0.0073, "step": 14180 }, { "grad_norm": 0.1029808521270752, "learning_rate": 2.1358060751401547e-05, "loss": 0.0089, "step": 14190 }, { "grad_norm": 0.10161738842725754, "learning_rate": 2.129033511548566e-05, "loss": 0.0077, "step": 14200 }, { "grad_norm": 0.0979568213224411, "learning_rate": 2.1222687970697315e-05, "loss": 0.0094, "step": 14210 }, { "grad_norm": 0.07767055928707123, "learning_rate": 2.1155119501981173e-05, "loss": 0.0083, "step": 14220 }, { "grad_norm": 0.12480639666318893, "learning_rate": 2.1087629894066895e-05, "loss": 0.0089, "step": 14230 }, { "grad_norm": 0.1004820168018341, "learning_rate": 2.1020219331468473e-05, "loss": 0.0084, "step": 14240 }, { "grad_norm": 0.09391839802265167, "learning_rate": 2.095288799848379e-05, "loss": 0.0086, "step": 14250 }, { "grad_norm": 0.18741102516651154, "learning_rate": 2.088563607919417e-05, "loss": 0.0072, "step": 14260 }, { "grad_norm": 0.14175111055374146, "learning_rate": 2.0818463757463786e-05, "loss": 0.009, "step": 14270 }, { "grad_norm": 0.10307285189628601, "learning_rate": 2.0751371216939175e-05, "loss": 0.0072, "step": 14280 }, { "grad_norm": 0.09157941490411758, "learning_rate": 2.068435864104882e-05, "loss": 0.0069, "step": 14290 }, { "grad_norm": 0.12749050557613373, "learning_rate": 2.0617426213002506e-05, "loss": 0.0075, "step": 14300 }, { "grad_norm": 0.10182447731494904, "learning_rate": 2.055057411579097e-05, "loss": 0.01, "step": 14310 }, { "grad_norm": 0.22169116139411926, "learning_rate": 2.0483802532185286e-05, "loss": 0.0089, "step": 14320 }, { "grad_norm": 0.2199435830116272, "learning_rate": 2.041711164473638e-05, "loss": 0.0063, "step": 14330 }, { "grad_norm": 0.0738394632935524, "learning_rate": 2.0350501635774637e-05, "loss": 0.0089, "step": 14340 }, { "grad_norm": 0.12519657611846924, "learning_rate": 2.0283972687409247e-05, "loss": 0.007, "step": 14350 }, { "grad_norm": 0.1399499922990799, "learning_rate": 2.021752498152784e-05, "loss": 0.007, "step": 14360 }, { "grad_norm": 0.09296436607837677, "learning_rate": 2.015115869979589e-05, "loss": 0.006, "step": 14370 }, { "grad_norm": 0.10002819448709488, "learning_rate": 2.0084874023656265e-05, "loss": 0.0076, "step": 14380 }, { "grad_norm": 0.10664018988609314, "learning_rate": 2.001867113432877e-05, "loss": 0.0087, "step": 14390 }, { "grad_norm": 0.08933674544095993, "learning_rate": 1.995255021280954e-05, "loss": 0.0071, "step": 14400 }, { "grad_norm": 0.12043409794569016, "learning_rate": 1.9886511439870688e-05, "loss": 0.0064, "step": 14410 }, { "grad_norm": 0.08205411583185196, "learning_rate": 1.9820554996059675e-05, "loss": 0.0057, "step": 14420 }, { "grad_norm": 0.08462446182966232, "learning_rate": 1.9754681061698893e-05, "loss": 0.0078, "step": 14430 }, { "grad_norm": 0.12133453786373138, "learning_rate": 1.9688889816885185e-05, "loss": 0.0084, "step": 14440 }, { "grad_norm": 0.10380852967500687, "learning_rate": 1.962318144148928e-05, "loss": 0.0072, "step": 14450 }, { "grad_norm": 0.09530024230480194, "learning_rate": 1.955755611515539e-05, "loss": 0.0092, "step": 14460 }, { "grad_norm": 0.10805962234735489, "learning_rate": 1.9492014017300642e-05, "loss": 0.0076, "step": 14470 }, { "grad_norm": 0.09695396572351456, "learning_rate": 1.942655532711461e-05, "loss": 0.0062, "step": 14480 }, { "grad_norm": 0.0773329809308052, "learning_rate": 1.9361180223558882e-05, "loss": 0.0071, "step": 14490 }, { "grad_norm": 0.08303295075893402, "learning_rate": 1.929588888536647e-05, "loss": 0.0101, "step": 14500 }, { "grad_norm": 0.11311710625886917, "learning_rate": 1.9230681491041425e-05, "loss": 0.0084, "step": 14510 }, { "grad_norm": 0.11405989527702332, "learning_rate": 1.9165558218858264e-05, "loss": 0.0069, "step": 14520 }, { "grad_norm": 0.08472662419080734, "learning_rate": 1.9100519246861505e-05, "loss": 0.0069, "step": 14530 }, { "grad_norm": 0.07290489971637726, "learning_rate": 1.9035564752865248e-05, "loss": 0.0069, "step": 14540 }, { "grad_norm": 0.08588680624961853, "learning_rate": 1.897069491445258e-05, "loss": 0.0088, "step": 14550 }, { "grad_norm": 0.07295157015323639, "learning_rate": 1.890590990897515e-05, "loss": 0.0086, "step": 14560 }, { "grad_norm": 0.0657152533531189, "learning_rate": 1.884120991355272e-05, "loss": 0.0057, "step": 14570 }, { "grad_norm": 0.09512811154127121, "learning_rate": 1.8776595105072576e-05, "loss": 0.0059, "step": 14580 }, { "grad_norm": 0.06638731062412262, "learning_rate": 1.8712065660189166e-05, "loss": 0.007, "step": 14590 }, { "grad_norm": 0.06674745678901672, "learning_rate": 1.8647621755323513e-05, "loss": 0.0075, "step": 14600 }, { "grad_norm": 0.06829272210597992, "learning_rate": 1.858326356666278e-05, "loss": 0.009, "step": 14610 }, { "grad_norm": 0.0793204978108406, "learning_rate": 1.851899127015983e-05, "loss": 0.0061, "step": 14620 }, { "grad_norm": 0.11140291392803192, "learning_rate": 1.8454805041532626e-05, "loss": 0.008, "step": 14630 }, { "grad_norm": 0.08972975611686707, "learning_rate": 1.8390705056263906e-05, "loss": 0.0071, "step": 14640 }, { "grad_norm": 0.11833430081605911, "learning_rate": 1.832669148960057e-05, "loss": 0.007, "step": 14650 }, { "grad_norm": 0.09125308692455292, "learning_rate": 1.8262764516553233e-05, "loss": 0.0067, "step": 14660 }, { "grad_norm": 0.12853814661502838, "learning_rate": 1.8198924311895843e-05, "loss": 0.0079, "step": 14670 }, { "grad_norm": 0.089320607483387, "learning_rate": 1.813517105016505e-05, "loss": 0.007, "step": 14680 }, { "grad_norm": 0.09511534869670868, "learning_rate": 1.8071504905659888e-05, "loss": 0.006, "step": 14690 }, { "grad_norm": 0.0785360038280487, "learning_rate": 1.800792605244109e-05, "loss": 0.0082, "step": 14700 }, { "grad_norm": 0.08107185363769531, "learning_rate": 1.7944434664330844e-05, "loss": 0.0056, "step": 14710 }, { "grad_norm": 0.08964671194553375, "learning_rate": 1.7881030914912212e-05, "loss": 0.0064, "step": 14720 }, { "grad_norm": 0.1088772639632225, "learning_rate": 1.7817714977528577e-05, "loss": 0.0077, "step": 14730 }, { "grad_norm": 0.07961886376142502, "learning_rate": 1.7754487025283332e-05, "loss": 0.0068, "step": 14740 }, { "grad_norm": 0.09294617921113968, "learning_rate": 1.7691347231039275e-05, "loss": 0.0058, "step": 14750 }, { "grad_norm": 0.12177979946136475, "learning_rate": 1.7628295767418164e-05, "loss": 0.0098, "step": 14760 }, { "grad_norm": 0.11283393949270248, "learning_rate": 1.7565332806800333e-05, "loss": 0.0069, "step": 14770 }, { "grad_norm": 0.14111700654029846, "learning_rate": 1.750245852132408e-05, "loss": 0.0071, "step": 14780 }, { "grad_norm": 0.11223061382770538, "learning_rate": 1.7439673082885323e-05, "loss": 0.0068, "step": 14790 }, { "grad_norm": 0.08355996012687683, "learning_rate": 1.7376976663137047e-05, "loss": 0.0085, "step": 14800 }, { "grad_norm": 0.10193824768066406, "learning_rate": 1.7314369433488853e-05, "loss": 0.0062, "step": 14810 }, { "grad_norm": 0.09209271520376205, "learning_rate": 1.7251851565106548e-05, "loss": 0.0057, "step": 14820 }, { "grad_norm": 0.09654773771762848, "learning_rate": 1.7189423228911574e-05, "loss": 0.0077, "step": 14830 }, { "grad_norm": 0.07318886369466782, "learning_rate": 1.7127084595580606e-05, "loss": 0.0069, "step": 14840 }, { "grad_norm": 0.06927516311407089, "learning_rate": 1.706483583554513e-05, "loss": 0.0054, "step": 14850 }, { "grad_norm": 0.0898497998714447, "learning_rate": 1.700267711899083e-05, "loss": 0.0074, "step": 14860 }, { "grad_norm": 0.10982313752174377, "learning_rate": 1.69406086158573e-05, "loss": 0.0078, "step": 14870 }, { "grad_norm": 0.07620345801115036, "learning_rate": 1.6878630495837455e-05, "loss": 0.0075, "step": 14880 }, { "grad_norm": 0.0814691036939621, "learning_rate": 1.681674292837707e-05, "loss": 0.0074, "step": 14890 }, { "grad_norm": 0.11770932376384735, "learning_rate": 1.6754946082674444e-05, "loss": 0.0081, "step": 14900 }, { "grad_norm": 0.06912347674369812, "learning_rate": 1.6693240127679748e-05, "loss": 0.0085, "step": 14910 }, { "grad_norm": 0.07074784487485886, "learning_rate": 1.663162523209475e-05, "loss": 0.0076, "step": 14920 }, { "grad_norm": 0.11345496028661728, "learning_rate": 1.6570101564372193e-05, "loss": 0.0077, "step": 14930 }, { "grad_norm": 0.09759191423654556, "learning_rate": 1.650866929271543e-05, "loss": 0.0079, "step": 14940 }, { "grad_norm": 0.13991639018058777, "learning_rate": 1.644732858507797e-05, "loss": 0.0073, "step": 14950 }, { "grad_norm": 0.10301797837018967, "learning_rate": 1.6386079609162943e-05, "loss": 0.0066, "step": 14960 }, { "grad_norm": 0.11252191662788391, "learning_rate": 1.6324922532422742e-05, "loss": 0.0077, "step": 14970 }, { "grad_norm": 0.11729101091623306, "learning_rate": 1.6263857522058434e-05, "loss": 0.0073, "step": 14980 }, { "grad_norm": 0.14178204536437988, "learning_rate": 1.6202884745019443e-05, "loss": 0.0067, "step": 14990 }, { "grad_norm": 0.09560935944318771, "learning_rate": 1.614200436800304e-05, "loss": 0.0076, "step": 15000 }, { "grad_norm": 0.111383356153965, "learning_rate": 1.6081216557453814e-05, "loss": 0.0093, "step": 15010 }, { "grad_norm": 0.10216289758682251, "learning_rate": 1.6020521479563367e-05, "loss": 0.0092, "step": 15020 }, { "grad_norm": 0.1042502298951149, "learning_rate": 1.5959919300269654e-05, "loss": 0.0057, "step": 15030 }, { "grad_norm": 0.07481419295072556, "learning_rate": 1.5899410185256764e-05, "loss": 0.0058, "step": 15040 }, { "grad_norm": 0.08798514306545258, "learning_rate": 1.583899429995431e-05, "loss": 0.0085, "step": 15050 }, { "grad_norm": 0.12407084554433823, "learning_rate": 1.5778671809536993e-05, "loss": 0.0074, "step": 15060 }, { "grad_norm": 0.11018476635217667, "learning_rate": 1.5718442878924246e-05, "loss": 0.0093, "step": 15070 }, { "grad_norm": 0.12427444010972977, "learning_rate": 1.5658307672779593e-05, "loss": 0.0089, "step": 15080 }, { "grad_norm": 0.16910147666931152, "learning_rate": 1.5598266355510427e-05, "loss": 0.0096, "step": 15090 }, { "grad_norm": 0.09085175395011902, "learning_rate": 1.553831909126744e-05, "loss": 0.0076, "step": 15100 }, { "grad_norm": 0.2008506804704666, "learning_rate": 1.5478466043944135e-05, "loss": 0.0084, "step": 15110 }, { "grad_norm": 0.08312593400478363, "learning_rate": 1.5418707377176468e-05, "loss": 0.008, "step": 15120 }, { "grad_norm": 0.05535462498664856, "learning_rate": 1.535904325434233e-05, "loss": 0.0052, "step": 15130 }, { "grad_norm": 0.3113829791545868, "learning_rate": 1.529947383856118e-05, "loss": 0.0067, "step": 15140 }, { "grad_norm": 0.0743807703256607, "learning_rate": 1.5239999292693524e-05, "loss": 0.006, "step": 15150 }, { "grad_norm": 0.15367582440376282, "learning_rate": 1.5180619779340505e-05, "loss": 0.0075, "step": 15160 }, { "grad_norm": 0.09368933737277985, "learning_rate": 1.5121335460843428e-05, "loss": 0.0073, "step": 15170 }, { "grad_norm": 0.07674387842416763, "learning_rate": 1.5062146499283347e-05, "loss": 0.0057, "step": 15180 }, { "grad_norm": 0.16500553488731384, "learning_rate": 1.5003053056480643e-05, "loss": 0.0093, "step": 15190 }, { "grad_norm": 0.06905755400657654, "learning_rate": 1.4944055293994551e-05, "loss": 0.0073, "step": 15200 }, { "grad_norm": 0.08445880562067032, "learning_rate": 1.4885153373122656e-05, "loss": 0.0062, "step": 15210 }, { "grad_norm": 0.07760252803564072, "learning_rate": 1.482634745490059e-05, "loss": 0.0063, "step": 15220 }, { "grad_norm": 0.08332260698080063, "learning_rate": 1.4767637700101466e-05, "loss": 0.0078, "step": 15230 }, { "grad_norm": 0.14454983174800873, "learning_rate": 1.4709024269235528e-05, "loss": 0.007, "step": 15240 }, { "grad_norm": 0.08025256544351578, "learning_rate": 1.4650507322549684e-05, "loss": 0.0063, "step": 15250 }, { "grad_norm": 0.10077039152383804, "learning_rate": 1.4592087020026972e-05, "loss": 0.0063, "step": 15260 }, { "grad_norm": 0.09173734486103058, "learning_rate": 1.4533763521386318e-05, "loss": 0.0082, "step": 15270 }, { "grad_norm": 0.12206606566905975, "learning_rate": 1.44755369860819e-05, "loss": 0.008, "step": 15280 }, { "grad_norm": 0.09893602132797241, "learning_rate": 1.441740757330287e-05, "loss": 0.0079, "step": 15290 }, { "grad_norm": 0.08898919075727463, "learning_rate": 1.4359375441972844e-05, "loss": 0.0086, "step": 15300 }, { "grad_norm": 0.07001844793558121, "learning_rate": 1.4301440750749395e-05, "loss": 0.0058, "step": 15310 }, { "grad_norm": 0.06008646637201309, "learning_rate": 1.4243603658023808e-05, "loss": 0.0042, "step": 15320 }, { "grad_norm": 0.06566991657018661, "learning_rate": 1.4185864321920444e-05, "loss": 0.0065, "step": 15330 }, { "grad_norm": 0.1540137529373169, "learning_rate": 1.4128222900296485e-05, "loss": 0.0061, "step": 15340 }, { "grad_norm": 0.08324749022722244, "learning_rate": 1.407067955074135e-05, "loss": 0.0064, "step": 15350 }, { "grad_norm": 0.06702782213687897, "learning_rate": 1.4013234430576356e-05, "loss": 0.0082, "step": 15360 }, { "grad_norm": 0.09729523211717606, "learning_rate": 1.3955887696854286e-05, "loss": 0.0086, "step": 15370 }, { "grad_norm": 0.09266554564237595, "learning_rate": 1.38986395063589e-05, "loss": 0.0061, "step": 15380 }, { "grad_norm": 0.10898573696613312, "learning_rate": 1.3841490015604597e-05, "loss": 0.008, "step": 15390 }, { "grad_norm": 0.07453274726867676, "learning_rate": 1.3784439380835879e-05, "loss": 0.0078, "step": 15400 }, { "grad_norm": 0.09289222210645676, "learning_rate": 1.3727487758026986e-05, "loss": 0.0047, "step": 15410 }, { "grad_norm": 0.11144881695508957, "learning_rate": 1.3670635302881525e-05, "loss": 0.0092, "step": 15420 }, { "grad_norm": 0.1033562496304512, "learning_rate": 1.3613882170831888e-05, "loss": 0.0065, "step": 15430 }, { "grad_norm": 0.1621370017528534, "learning_rate": 1.355722851703901e-05, "loss": 0.0087, "step": 15440 }, { "grad_norm": 0.0954129546880722, "learning_rate": 1.3500674496391814e-05, "loss": 0.0051, "step": 15450 }, { "grad_norm": 0.07662945240736008, "learning_rate": 1.3444220263506795e-05, "loss": 0.0055, "step": 15460 }, { "grad_norm": 0.04798706993460655, "learning_rate": 1.3387865972727714e-05, "loss": 0.0047, "step": 15470 }, { "grad_norm": 0.09507722407579422, "learning_rate": 1.3331611778125036e-05, "loss": 0.0086, "step": 15480 }, { "grad_norm": 0.11110657453536987, "learning_rate": 1.3275457833495564e-05, "loss": 0.0085, "step": 15490 }, { "grad_norm": 0.08958341181278229, "learning_rate": 1.3219404292362065e-05, "loss": 0.0089, "step": 15500 }, { "grad_norm": 0.09792760014533997, "learning_rate": 1.3163451307972751e-05, "loss": 0.0063, "step": 15510 }, { "grad_norm": 0.09412389248609543, "learning_rate": 1.3107599033300977e-05, "loss": 0.0081, "step": 15520 }, { "grad_norm": 0.10493183881044388, "learning_rate": 1.305184762104471e-05, "loss": 0.0062, "step": 15530 }, { "grad_norm": 0.07390110194683075, "learning_rate": 1.2996197223626178e-05, "loss": 0.0067, "step": 15540 }, { "grad_norm": 0.06120933219790459, "learning_rate": 1.2940647993191457e-05, "loss": 0.0073, "step": 15550 }, { "grad_norm": 0.09218470752239227, "learning_rate": 1.2885200081610005e-05, "loss": 0.0085, "step": 15560 }, { "grad_norm": 0.11428312212228775, "learning_rate": 1.2829853640474316e-05, "loss": 0.0072, "step": 15570 }, { "grad_norm": 0.060681622475385666, "learning_rate": 1.2774608821099438e-05, "loss": 0.0081, "step": 15580 }, { "grad_norm": 0.0773649513721466, "learning_rate": 1.2719465774522577e-05, "loss": 0.0051, "step": 15590 }, { "grad_norm": 0.06299198418855667, "learning_rate": 1.2664424651502755e-05, "loss": 0.0065, "step": 15600 }, { "grad_norm": 0.1236218512058258, "learning_rate": 1.260948560252026e-05, "loss": 0.0058, "step": 15610 }, { "grad_norm": 0.10900051891803741, "learning_rate": 1.2554648777776396e-05, "loss": 0.006, "step": 15620 }, { "grad_norm": 0.09777753800153732, "learning_rate": 1.2499914327192919e-05, "loss": 0.0056, "step": 15630 }, { "grad_norm": 0.09088607877492905, "learning_rate": 1.2445282400411722e-05, "loss": 0.0056, "step": 15640 }, { "grad_norm": 0.10962533950805664, "learning_rate": 1.2390753146794437e-05, "loss": 0.0063, "step": 15650 }, { "grad_norm": 0.159529909491539, "learning_rate": 1.2336326715421925e-05, "loss": 0.0066, "step": 15660 }, { "grad_norm": 0.07237947732210159, "learning_rate": 1.2282003255094005e-05, "loss": 0.0058, "step": 15670 }, { "grad_norm": 0.0412864089012146, "learning_rate": 1.2227782914328928e-05, "loss": 0.0065, "step": 15680 }, { "grad_norm": 0.06408300995826721, "learning_rate": 1.2173665841363018e-05, "loss": 0.0062, "step": 15690 }, { "grad_norm": 0.08847568184137344, "learning_rate": 1.211965218415032e-05, "loss": 0.0048, "step": 15700 }, { "grad_norm": 0.07362963259220123, "learning_rate": 1.2065742090362082e-05, "loss": 0.0071, "step": 15710 }, { "grad_norm": 0.06620814651250839, "learning_rate": 1.2011935707386457e-05, "loss": 0.0058, "step": 15720 }, { "grad_norm": 0.06448313593864441, "learning_rate": 1.1958233182328044e-05, "loss": 0.0069, "step": 15730 }, { "grad_norm": 0.06796927005052567, "learning_rate": 1.1904634662007474e-05, "loss": 0.0064, "step": 15740 }, { "grad_norm": 0.10287852585315704, "learning_rate": 1.1851140292961088e-05, "loss": 0.0077, "step": 15750 }, { "grad_norm": 0.13440538942813873, "learning_rate": 1.1797750221440424e-05, "loss": 0.0062, "step": 15760 }, { "grad_norm": 0.059975624084472656, "learning_rate": 1.1744464593411897e-05, "loss": 0.0077, "step": 15770 }, { "grad_norm": 0.0797332301735878, "learning_rate": 1.1691283554556399e-05, "loss": 0.0058, "step": 15780 }, { "grad_norm": 0.08228960633277893, "learning_rate": 1.1638207250268834e-05, "loss": 0.0079, "step": 15790 }, { "grad_norm": 0.06919419020414352, "learning_rate": 1.158523582565782e-05, "loss": 0.0089, "step": 15800 }, { "grad_norm": 0.05788502097129822, "learning_rate": 1.1532369425545192e-05, "loss": 0.0061, "step": 15810 }, { "grad_norm": 0.12093570083379745, "learning_rate": 1.1479608194465662e-05, "loss": 0.0074, "step": 15820 }, { "grad_norm": 0.1303563416004181, "learning_rate": 1.1426952276666442e-05, "loss": 0.0074, "step": 15830 }, { "grad_norm": 0.10586512088775635, "learning_rate": 1.1374401816106778e-05, "loss": 0.0068, "step": 15840 }, { "grad_norm": 0.08656790107488632, "learning_rate": 1.1321956956457646e-05, "loss": 0.0068, "step": 15850 }, { "grad_norm": 0.14733947813510895, "learning_rate": 1.1269617841101277e-05, "loss": 0.009, "step": 15860 }, { "grad_norm": 0.08039134740829468, "learning_rate": 1.1217384613130804e-05, "loss": 0.0059, "step": 15870 }, { "grad_norm": 0.07003935426473618, "learning_rate": 1.11652574153499e-05, "loss": 0.0078, "step": 15880 }, { "grad_norm": 0.2308933138847351, "learning_rate": 1.1113236390272303e-05, "loss": 0.0079, "step": 15890 }, { "grad_norm": 0.11959139257669449, "learning_rate": 1.106132168012155e-05, "loss": 0.0068, "step": 15900 }, { "grad_norm": 0.1086306944489479, "learning_rate": 1.1009513426830448e-05, "loss": 0.0059, "step": 15910 }, { "grad_norm": 0.10157997906208038, "learning_rate": 1.0957811772040777e-05, "loss": 0.0064, "step": 15920 }, { "grad_norm": 0.10171937197446823, "learning_rate": 1.0906216857102913e-05, "loss": 0.0061, "step": 15930 }, { "grad_norm": 0.14361846446990967, "learning_rate": 1.0854728823075355e-05, "loss": 0.0054, "step": 15940 }, { "grad_norm": 0.11922954022884369, "learning_rate": 1.0803347810724452e-05, "loss": 0.0069, "step": 15950 }, { "grad_norm": 0.06636133790016174, "learning_rate": 1.0752073960523911e-05, "loss": 0.005, "step": 15960 }, { "grad_norm": 0.09184479713439941, "learning_rate": 1.070090741265447e-05, "loss": 0.0075, "step": 15970 }, { "grad_norm": 0.14686807990074158, "learning_rate": 1.0649848307003547e-05, "loss": 0.0068, "step": 15980 }, { "grad_norm": 0.08753345161676407, "learning_rate": 1.0598896783164757e-05, "loss": 0.0075, "step": 15990 }, { "grad_norm": 0.07070576399564743, "learning_rate": 1.0548052980437645e-05, "loss": 0.006, "step": 16000 }, { "grad_norm": 0.09851239621639252, "learning_rate": 1.049731703782722e-05, "loss": 0.0051, "step": 16010 }, { "grad_norm": 0.09283073991537094, "learning_rate": 1.0446689094043587e-05, "loss": 0.0067, "step": 16020 }, { "grad_norm": 0.09354682266712189, "learning_rate": 1.039616928750165e-05, "loss": 0.0063, "step": 16030 }, { "grad_norm": 0.06721899658441544, "learning_rate": 1.0345757756320612e-05, "loss": 0.0057, "step": 16040 }, { "grad_norm": 0.09103065729141235, "learning_rate": 1.0295454638323666e-05, "loss": 0.0063, "step": 16050 }, { "grad_norm": 0.12501469254493713, "learning_rate": 1.0245260071037632e-05, "loss": 0.0077, "step": 16060 }, { "grad_norm": 0.08893635123968124, "learning_rate": 1.0195174191692518e-05, "loss": 0.0046, "step": 16070 }, { "grad_norm": 0.06310978531837463, "learning_rate": 1.014519713722124e-05, "loss": 0.0064, "step": 16080 }, { "grad_norm": 0.09981871396303177, "learning_rate": 1.0095329044259132e-05, "loss": 0.008, "step": 16090 }, { "grad_norm": 0.06274831295013428, "learning_rate": 1.004557004914365e-05, "loss": 0.0042, "step": 16100 }, { "grad_norm": 0.06786265969276428, "learning_rate": 9.995920287914007e-06, "loss": 0.0056, "step": 16110 }, { "grad_norm": 0.09739448130130768, "learning_rate": 9.946379896310737e-06, "loss": 0.0067, "step": 16120 }, { "grad_norm": 0.0846710279583931, "learning_rate": 9.896949009775396e-06, "loss": 0.0063, "step": 16130 }, { "grad_norm": 0.10824328660964966, "learning_rate": 9.847627763450134e-06, "loss": 0.0056, "step": 16140 }, { "grad_norm": 0.07965362071990967, "learning_rate": 9.798416292177337e-06, "loss": 0.0059, "step": 16150 }, { "grad_norm": 0.0817146971821785, "learning_rate": 9.74931473049932e-06, "loss": 0.006, "step": 16160 }, { "grad_norm": 0.12738506495952606, "learning_rate": 9.700323212657847e-06, "loss": 0.0057, "step": 16170 }, { "grad_norm": 0.09187974780797958, "learning_rate": 9.65144187259388e-06, "loss": 0.0054, "step": 16180 }, { "grad_norm": 0.13302300870418549, "learning_rate": 9.602670843947132e-06, "loss": 0.0065, "step": 16190 }, { "grad_norm": 0.06065499410033226, "learning_rate": 9.554010260055713e-06, "loss": 0.0076, "step": 16200 }, { "grad_norm": 0.09114003926515579, "learning_rate": 9.505460253955834e-06, "loss": 0.0061, "step": 16210 }, { "grad_norm": 0.10234687477350235, "learning_rate": 9.457020958381324e-06, "loss": 0.0071, "step": 16220 }, { "grad_norm": 0.11170613020658493, "learning_rate": 9.408692505763395e-06, "loss": 0.0057, "step": 16230 }, { "grad_norm": 0.06377864629030228, "learning_rate": 9.360475028230181e-06, "loss": 0.0048, "step": 16240 }, { "grad_norm": 0.08116354793310165, "learning_rate": 9.312368657606412e-06, "loss": 0.0076, "step": 16250 }, { "grad_norm": 0.0872003436088562, "learning_rate": 9.264373525413096e-06, "loss": 0.0063, "step": 16260 }, { "grad_norm": 0.18986864387989044, "learning_rate": 9.216489762867058e-06, "loss": 0.0073, "step": 16270 }, { "grad_norm": 0.17253999412059784, "learning_rate": 9.168717500880708e-06, "loss": 0.0066, "step": 16280 }, { "grad_norm": 0.07282308489084244, "learning_rate": 9.121056870061574e-06, "loss": 0.0099, "step": 16290 }, { "grad_norm": 0.05808091536164284, "learning_rate": 9.073508000711983e-06, "loss": 0.0067, "step": 16300 }, { "grad_norm": 0.11794750392436981, "learning_rate": 9.026071022828758e-06, "loss": 0.0057, "step": 16310 }, { "grad_norm": 0.07870656996965408, "learning_rate": 8.978746066102771e-06, "loss": 0.0063, "step": 16320 }, { "grad_norm": 0.055361539125442505, "learning_rate": 8.931533259918634e-06, "loss": 0.007, "step": 16330 }, { "grad_norm": 0.10337762534618378, "learning_rate": 8.884432733354382e-06, "loss": 0.0065, "step": 16340 }, { "grad_norm": 0.0819077268242836, "learning_rate": 8.837444615181029e-06, "loss": 0.0047, "step": 16350 }, { "grad_norm": 0.10389268398284912, "learning_rate": 8.790569033862323e-06, "loss": 0.0058, "step": 16360 }, { "grad_norm": 0.08228936046361923, "learning_rate": 8.7438061175543e-06, "loss": 0.0051, "step": 16370 }, { "grad_norm": 0.08764046430587769, "learning_rate": 8.697155994104978e-06, "loss": 0.0057, "step": 16380 }, { "grad_norm": 0.1107165589928627, "learning_rate": 8.650618791054033e-06, "loss": 0.0088, "step": 16390 }, { "grad_norm": 0.10121536254882812, "learning_rate": 8.604194635632373e-06, "loss": 0.0059, "step": 16400 }, { "grad_norm": 0.06671851128339767, "learning_rate": 8.557883654761906e-06, "loss": 0.0052, "step": 16410 }, { "grad_norm": 0.06821329891681671, "learning_rate": 8.511685975055061e-06, "loss": 0.0059, "step": 16420 }, { "grad_norm": 0.08398965001106262, "learning_rate": 8.46560172281452e-06, "loss": 0.0068, "step": 16430 }, { "grad_norm": 0.1557084321975708, "learning_rate": 8.419631024032893e-06, "loss": 0.0052, "step": 16440 }, { "grad_norm": 0.050409622490406036, "learning_rate": 8.373774004392293e-06, "loss": 0.0066, "step": 16450 }, { "grad_norm": 0.060342758893966675, "learning_rate": 8.32803078926409e-06, "loss": 0.0049, "step": 16460 }, { "grad_norm": 0.08255565166473389, "learning_rate": 8.282401503708454e-06, "loss": 0.0058, "step": 16470 }, { "grad_norm": 0.07480427622795105, "learning_rate": 8.23688627247412e-06, "loss": 0.0071, "step": 16480 }, { "grad_norm": 0.09599100053310394, "learning_rate": 8.191485219998007e-06, "loss": 0.007, "step": 16490 }, { "grad_norm": 0.061760883778333664, "learning_rate": 8.146198470404843e-06, "loss": 0.004, "step": 16500 }, { "grad_norm": 0.09875854849815369, "learning_rate": 8.101026147506897e-06, "loss": 0.0059, "step": 16510 }, { "grad_norm": 0.060174472630023956, "learning_rate": 8.05596837480353e-06, "loss": 0.0046, "step": 16520 }, { "grad_norm": 0.13009759783744812, "learning_rate": 8.011025275480998e-06, "loss": 0.0056, "step": 16530 }, { "grad_norm": 0.05840182676911354, "learning_rate": 7.966196972412027e-06, "loss": 0.0044, "step": 16540 }, { "grad_norm": 0.03272266313433647, "learning_rate": 7.92148358815547e-06, "loss": 0.0055, "step": 16550 }, { "grad_norm": 0.04531165584921837, "learning_rate": 7.87688524495604e-06, "loss": 0.0048, "step": 16560 }, { "grad_norm": 0.07629932463169098, "learning_rate": 7.83240206474386e-06, "loss": 0.0048, "step": 16570 }, { "grad_norm": 0.08712384849786758, "learning_rate": 7.788034169134272e-06, "loss": 0.0069, "step": 16580 }, { "grad_norm": 0.06564653664827347, "learning_rate": 7.743781679427414e-06, "loss": 0.0063, "step": 16590 }, { "grad_norm": 0.10010945051908493, "learning_rate": 7.699644716607895e-06, "loss": 0.0083, "step": 16600 }, { "grad_norm": 0.05900498107075691, "learning_rate": 7.655623401344486e-06, "loss": 0.0061, "step": 16610 }, { "grad_norm": 0.08640993386507034, "learning_rate": 7.611717853989775e-06, "loss": 0.0049, "step": 16620 }, { "grad_norm": 0.09353779256343842, "learning_rate": 7.567928194579854e-06, "loss": 0.0059, "step": 16630 }, { "grad_norm": 0.08959268778562546, "learning_rate": 7.524254542833997e-06, "loss": 0.0075, "step": 16640 }, { "grad_norm": 0.10222340375185013, "learning_rate": 7.480697018154286e-06, "loss": 0.005, "step": 16650 }, { "grad_norm": 0.04498433321714401, "learning_rate": 7.437255739625332e-06, "loss": 0.0058, "step": 16660 }, { "grad_norm": 0.10478605329990387, "learning_rate": 7.393930826013923e-06, "loss": 0.0064, "step": 16670 }, { "grad_norm": 0.050683025270700455, "learning_rate": 7.350722395768722e-06, "loss": 0.0056, "step": 16680 }, { "grad_norm": 0.11258295178413391, "learning_rate": 7.307630567019963e-06, "loss": 0.0101, "step": 16690 }, { "grad_norm": 0.15317095816135406, "learning_rate": 7.264655457579e-06, "loss": 0.0059, "step": 16700 }, { "grad_norm": 0.09390193223953247, "learning_rate": 7.221797184938184e-06, "loss": 0.0053, "step": 16710 }, { "grad_norm": 0.08691168576478958, "learning_rate": 7.179055866270373e-06, "loss": 0.0057, "step": 16720 }, { "grad_norm": 0.08521914482116699, "learning_rate": 7.136431618428707e-06, "loss": 0.0052, "step": 16730 }, { "grad_norm": 0.09342680126428604, "learning_rate": 7.09392455794628e-06, "loss": 0.005, "step": 16740 }, { "grad_norm": 0.17857925593852997, "learning_rate": 7.051534801035725e-06, "loss": 0.008, "step": 16750 }, { "grad_norm": 0.06931804120540619, "learning_rate": 7.00926246358905e-06, "loss": 0.005, "step": 16760 }, { "grad_norm": 0.06058470532298088, "learning_rate": 6.967107661177191e-06, "loss": 0.0063, "step": 16770 }, { "grad_norm": 0.08824970573186874, "learning_rate": 6.925070509049786e-06, "loss": 0.0052, "step": 16780 }, { "grad_norm": 0.12389223277568817, "learning_rate": 6.883151122134812e-06, "loss": 0.0053, "step": 16790 }, { "grad_norm": 0.09244880825281143, "learning_rate": 6.8413496150382394e-06, "loss": 0.0065, "step": 16800 }, { "grad_norm": 0.086562879383564, "learning_rate": 6.7996661020438165e-06, "loss": 0.0081, "step": 16810 }, { "grad_norm": 0.0657518059015274, "learning_rate": 6.758100697112662e-06, "loss": 0.0072, "step": 16820 }, { "grad_norm": 0.09960281103849411, "learning_rate": 6.716653513883026e-06, "loss": 0.0072, "step": 16830 }, { "grad_norm": 0.08126631379127502, "learning_rate": 6.675324665669913e-06, "loss": 0.0052, "step": 16840 }, { "grad_norm": 0.0641186386346817, "learning_rate": 6.634114265464803e-06, "loss": 0.0054, "step": 16850 }, { "grad_norm": 0.07775498926639557, "learning_rate": 6.59302242593538e-06, "loss": 0.0049, "step": 16860 }, { "grad_norm": 0.07605016231536865, "learning_rate": 6.552049259425141e-06, "loss": 0.007, "step": 16870 }, { "grad_norm": 0.21370579302310944, "learning_rate": 6.511194877953181e-06, "loss": 0.0052, "step": 16880 }, { "grad_norm": 0.04260757565498352, "learning_rate": 6.470459393213813e-06, "loss": 0.0057, "step": 16890 }, { "grad_norm": 0.11486883461475372, "learning_rate": 6.429842916576279e-06, "loss": 0.0076, "step": 16900 }, { "grad_norm": 0.09330947697162628, "learning_rate": 6.389345559084503e-06, "loss": 0.0047, "step": 16910 }, { "grad_norm": 0.10489291697740555, "learning_rate": 6.348967431456682e-06, "loss": 0.0054, "step": 16920 }, { "grad_norm": 0.05917362868785858, "learning_rate": 6.30870864408511e-06, "loss": 0.0053, "step": 16930 }, { "grad_norm": 0.13839983940124512, "learning_rate": 6.268569307035754e-06, "loss": 0.0043, "step": 16940 }, { "grad_norm": 0.10065950453281403, "learning_rate": 6.228549530048022e-06, "loss": 0.0052, "step": 16950 }, { "grad_norm": 0.05466950312256813, "learning_rate": 6.1886494225344814e-06, "loss": 0.0054, "step": 16960 }, { "grad_norm": 0.05870116874575615, "learning_rate": 6.148869093580479e-06, "loss": 0.004, "step": 16970 }, { "grad_norm": 0.11038341373205185, "learning_rate": 6.109208651943921e-06, "loss": 0.0061, "step": 16980 }, { "grad_norm": 0.06627320498228073, "learning_rate": 6.069668206054946e-06, "loss": 0.0078, "step": 16990 }, { "grad_norm": 0.08225332945585251, "learning_rate": 6.0302478640156145e-06, "loss": 0.0066, "step": 17000 }, { "grad_norm": 0.08503419160842896, "learning_rate": 5.990947733599644e-06, "loss": 0.0064, "step": 17010 }, { "grad_norm": 0.0990796685218811, "learning_rate": 5.951767922252105e-06, "loss": 0.0065, "step": 17020 }, { "grad_norm": 0.07395527511835098, "learning_rate": 5.912708537089068e-06, "loss": 0.0045, "step": 17030 }, { "grad_norm": 0.05186518281698227, "learning_rate": 5.873769684897434e-06, "loss": 0.0069, "step": 17040 }, { "grad_norm": 0.05340685695409775, "learning_rate": 5.834951472134514e-06, "loss": 0.0047, "step": 17050 }, { "grad_norm": 0.04467863589525223, "learning_rate": 5.796254004927832e-06, "loss": 0.007, "step": 17060 }, { "grad_norm": 0.07906480133533478, "learning_rate": 5.757677389074806e-06, "loss": 0.0045, "step": 17070 }, { "grad_norm": 0.09099625796079636, "learning_rate": 5.719221730042385e-06, "loss": 0.006, "step": 17080 }, { "grad_norm": 0.0703350305557251, "learning_rate": 5.680887132966911e-06, "loss": 0.0054, "step": 17090 }, { "grad_norm": 0.108420230448246, "learning_rate": 5.642673702653683e-06, "loss": 0.0057, "step": 17100 }, { "grad_norm": 0.08459882438182831, "learning_rate": 5.604581543576781e-06, "loss": 0.0067, "step": 17110 }, { "grad_norm": 0.09169035404920578, "learning_rate": 5.566610759878704e-06, "loss": 0.0043, "step": 17120 }, { "grad_norm": 0.07617104053497314, "learning_rate": 5.528761455370119e-06, "loss": 0.0053, "step": 17130 }, { "grad_norm": 0.09099805355072021, "learning_rate": 5.491033733529594e-06, "loss": 0.0067, "step": 17140 }, { "grad_norm": 0.07508064061403275, "learning_rate": 5.453427697503255e-06, "loss": 0.0051, "step": 17150 }, { "grad_norm": 0.034046635031700134, "learning_rate": 5.415943450104599e-06, "loss": 0.0062, "step": 17160 }, { "grad_norm": 0.09095154702663422, "learning_rate": 5.378581093814111e-06, "loss": 0.0058, "step": 17170 }, { "grad_norm": 0.2392134964466095, "learning_rate": 5.3413407307790375e-06, "loss": 0.0063, "step": 17180 }, { "grad_norm": 0.07625044882297516, "learning_rate": 5.30422246281313e-06, "loss": 0.0053, "step": 17190 }, { "grad_norm": 0.09091664105653763, "learning_rate": 5.267226391396296e-06, "loss": 0.0051, "step": 17200 }, { "grad_norm": 0.10593272000551224, "learning_rate": 5.2303526176744e-06, "loss": 0.0055, "step": 17210 }, { "grad_norm": 0.12199395895004272, "learning_rate": 5.193601242458929e-06, "loss": 0.0083, "step": 17220 }, { "grad_norm": 0.11696618050336838, "learning_rate": 5.156972366226714e-06, "loss": 0.0055, "step": 17230 }, { "grad_norm": 0.05498863756656647, "learning_rate": 5.120466089119735e-06, "loss": 0.0046, "step": 17240 }, { "grad_norm": 0.12094159424304962, "learning_rate": 5.084082510944749e-06, "loss": 0.006, "step": 17250 }, { "grad_norm": 0.11300729960203171, "learning_rate": 5.047821731173058e-06, "loss": 0.0052, "step": 17260 }, { "grad_norm": 0.08559301495552063, "learning_rate": 5.011683848940274e-06, "loss": 0.0041, "step": 17270 }, { "grad_norm": 0.0989033430814743, "learning_rate": 4.975668963045954e-06, "loss": 0.0052, "step": 17280 }, { "grad_norm": 0.08544724434614182, "learning_rate": 4.9397771719534525e-06, "loss": 0.0042, "step": 17290 }, { "grad_norm": 0.05603135749697685, "learning_rate": 4.904008573789548e-06, "loss": 0.0076, "step": 17300 }, { "grad_norm": 0.1568598598241806, "learning_rate": 4.8683632663442005e-06, "loss": 0.0067, "step": 17310 }, { "grad_norm": 0.0529240220785141, "learning_rate": 4.832841347070343e-06, "loss": 0.0053, "step": 17320 }, { "grad_norm": 0.09001423418521881, "learning_rate": 4.797442913083539e-06, "loss": 0.0066, "step": 17330 }, { "grad_norm": 0.11325569450855255, "learning_rate": 4.7621680611617596e-06, "loss": 0.0044, "step": 17340 }, { "grad_norm": 0.10633333027362823, "learning_rate": 4.727016887745095e-06, "loss": 0.0051, "step": 17350 }, { "grad_norm": 0.08478104323148727, "learning_rate": 4.691989488935511e-06, "loss": 0.0049, "step": 17360 }, { "grad_norm": 0.10347475856542587, "learning_rate": 4.657085960496588e-06, "loss": 0.0052, "step": 17370 }, { "grad_norm": 0.07681947946548462, "learning_rate": 4.6223063978532265e-06, "loss": 0.0054, "step": 17380 }, { "grad_norm": 0.10335340350866318, "learning_rate": 4.587650896091439e-06, "loss": 0.0054, "step": 17390 }, { "grad_norm": 0.08291575312614441, "learning_rate": 4.553119549958035e-06, "loss": 0.0087, "step": 17400 }, { "grad_norm": 0.11735987663269043, "learning_rate": 4.518712453860385e-06, "loss": 0.0049, "step": 17410 }, { "grad_norm": 0.06377241760492325, "learning_rate": 4.484429701866205e-06, "loss": 0.0046, "step": 17420 }, { "grad_norm": 0.07619430124759674, "learning_rate": 4.4502713877031975e-06, "loss": 0.0052, "step": 17430 }, { "grad_norm": 0.08911009877920151, "learning_rate": 4.416237604758911e-06, "loss": 0.0041, "step": 17440 }, { "grad_norm": 0.07895364612340927, "learning_rate": 4.3823284460804025e-06, "loss": 0.004, "step": 17450 }, { "grad_norm": 0.1730530560016632, "learning_rate": 4.348544004374011e-06, "loss": 0.0073, "step": 17460 }, { "grad_norm": 0.06624027341604233, "learning_rate": 4.314884372005123e-06, "loss": 0.0057, "step": 17470 }, { "grad_norm": 0.15018588304519653, "learning_rate": 4.281349640997867e-06, "loss": 0.0047, "step": 17480 }, { "grad_norm": 0.056739870458841324, "learning_rate": 4.247939903034942e-06, "loss": 0.0065, "step": 17490 }, { "grad_norm": 0.058833058923482895, "learning_rate": 4.214655249457284e-06, "loss": 0.0046, "step": 17500 }, { "grad_norm": 0.06925462931394577, "learning_rate": 4.181495771263855e-06, "loss": 0.0042, "step": 17510 }, { "grad_norm": 0.05602520331740379, "learning_rate": 4.148461559111427e-06, "loss": 0.0062, "step": 17520 }, { "grad_norm": 0.07457327097654343, "learning_rate": 4.115552703314252e-06, "loss": 0.0078, "step": 17530 }, { "grad_norm": 0.08467822521924973, "learning_rate": 4.082769293843886e-06, "loss": 0.0059, "step": 17540 }, { "grad_norm": 0.09878405928611755, "learning_rate": 4.050111420328939e-06, "loss": 0.0056, "step": 17550 }, { "grad_norm": 0.05782167240977287, "learning_rate": 4.017579172054764e-06, "loss": 0.0065, "step": 17560 }, { "grad_norm": 0.0647156611084938, "learning_rate": 3.985172637963308e-06, "loss": 0.0075, "step": 17570 }, { "grad_norm": 0.06802534312009811, "learning_rate": 3.952891906652784e-06, "loss": 0.0058, "step": 17580 }, { "grad_norm": 0.1012989953160286, "learning_rate": 3.920737066377478e-06, "loss": 0.0061, "step": 17590 }, { "grad_norm": 0.1153709664940834, "learning_rate": 3.888708205047509e-06, "loss": 0.0063, "step": 17600 }, { "grad_norm": 0.06168808043003082, "learning_rate": 3.856805410228542e-06, "loss": 0.0046, "step": 17610 }, { "grad_norm": 0.0732150673866272, "learning_rate": 3.82502876914162e-06, "loss": 0.0038, "step": 17620 }, { "grad_norm": 0.09910286217927933, "learning_rate": 3.7933783686628586e-06, "loss": 0.0072, "step": 17630 }, { "grad_norm": 0.13740767538547516, "learning_rate": 3.7618542953232306e-06, "loss": 0.0066, "step": 17640 }, { "grad_norm": 0.0758737325668335, "learning_rate": 3.7304566353083658e-06, "loss": 0.0063, "step": 17650 }, { "grad_norm": 0.12121056020259857, "learning_rate": 3.6991854744582555e-06, "loss": 0.0073, "step": 17660 }, { "grad_norm": 0.051570530980825424, "learning_rate": 3.6680408982670777e-06, "loss": 0.0041, "step": 17670 }, { "grad_norm": 0.10041578859090805, "learning_rate": 3.637022991882899e-06, "loss": 0.005, "step": 17680 }, { "grad_norm": 0.10393265634775162, "learning_rate": 3.606131840107485e-06, "loss": 0.0087, "step": 17690 }, { "grad_norm": 0.0896223932504654, "learning_rate": 3.575367527396084e-06, "loss": 0.0046, "step": 17700 }, { "grad_norm": 0.05148059502243996, "learning_rate": 3.5447301378571386e-06, "loss": 0.004, "step": 17710 }, { "grad_norm": 0.1565588414669037, "learning_rate": 3.514219755252113e-06, "loss": 0.0042, "step": 17720 }, { "grad_norm": 0.06882771849632263, "learning_rate": 3.4838364629952213e-06, "loss": 0.0045, "step": 17730 }, { "grad_norm": 0.06501244753599167, "learning_rate": 3.4535803441532123e-06, "loss": 0.0047, "step": 17740 }, { "grad_norm": 0.06643030047416687, "learning_rate": 3.4234514814451836e-06, "loss": 0.0046, "step": 17750 }, { "grad_norm": 0.06516750901937485, "learning_rate": 3.393449957242273e-06, "loss": 0.0043, "step": 17760 }, { "grad_norm": 0.04781252518296242, "learning_rate": 3.363575853567524e-06, "loss": 0.0042, "step": 17770 }, { "grad_norm": 0.11086055636405945, "learning_rate": 3.3338292520955826e-06, "loss": 0.0047, "step": 17780 }, { "grad_norm": 0.08373504132032394, "learning_rate": 3.304210234152516e-06, "loss": 0.0055, "step": 17790 }, { "grad_norm": 0.08310943096876144, "learning_rate": 3.2747188807155993e-06, "loss": 0.0049, "step": 17800 }, { "grad_norm": 0.062434121966362, "learning_rate": 3.2453552724130643e-06, "loss": 0.0059, "step": 17810 }, { "grad_norm": 0.05270116776227951, "learning_rate": 3.216119489523889e-06, "loss": 0.0058, "step": 17820 }, { "grad_norm": 0.0889195054769516, "learning_rate": 3.1870116119775917e-06, "loss": 0.0044, "step": 17830 }, { "grad_norm": 0.06541448086500168, "learning_rate": 3.158031719353999e-06, "loss": 0.0053, "step": 17840 }, { "grad_norm": 0.06153072789311409, "learning_rate": 3.1291798908830273e-06, "loss": 0.0058, "step": 17850 }, { "grad_norm": 0.0906241238117218, "learning_rate": 3.1004562054444853e-06, "loss": 0.005, "step": 17860 }, { "grad_norm": 0.08475061506032944, "learning_rate": 3.071860741567806e-06, "loss": 0.0061, "step": 17870 }, { "grad_norm": 0.11429177224636078, "learning_rate": 3.04339357743193e-06, "loss": 0.0051, "step": 17880 }, { "grad_norm": 0.06172405555844307, "learning_rate": 3.0150547908649628e-06, "loss": 0.0049, "step": 17890 }, { "grad_norm": 0.09239153563976288, "learning_rate": 2.9868444593440957e-06, "loss": 0.0048, "step": 17900 }, { "grad_norm": 0.04949268698692322, "learning_rate": 2.9587626599952846e-06, "loss": 0.0045, "step": 17910 }, { "grad_norm": 0.1046813353896141, "learning_rate": 2.930809469593082e-06, "loss": 0.0061, "step": 17920 }, { "grad_norm": 0.043978720903396606, "learning_rate": 2.9029849645604733e-06, "loss": 0.0049, "step": 17930 }, { "grad_norm": 0.15293167531490326, "learning_rate": 2.8752892209685632e-06, "loss": 0.0049, "step": 17940 }, { "grad_norm": 0.054882146418094635, "learning_rate": 2.847722314536483e-06, "loss": 0.0044, "step": 17950 }, { "grad_norm": 0.08179155737161636, "learning_rate": 2.820284320631078e-06, "loss": 0.0043, "step": 17960 }, { "grad_norm": 0.06589469313621521, "learning_rate": 2.792975314266788e-06, "loss": 0.0064, "step": 17970 }, { "grad_norm": 0.05115808546543121, "learning_rate": 2.7657953701054007e-06, "loss": 0.0037, "step": 17980 }, { "grad_norm": 0.05994449555873871, "learning_rate": 2.7387445624558306e-06, "loss": 0.0046, "step": 17990 }, { "grad_norm": 0.06756210327148438, "learning_rate": 2.7118229652739747e-06, "loss": 0.0053, "step": 18000 }, { "grad_norm": 0.10849850624799728, "learning_rate": 2.6850306521624236e-06, "loss": 0.0055, "step": 18010 }, { "grad_norm": 0.08543810248374939, "learning_rate": 2.6583676963703507e-06, "loss": 0.0048, "step": 18020 }, { "grad_norm": 0.1348736435174942, "learning_rate": 2.631834170793268e-06, "loss": 0.0047, "step": 18030 }, { "grad_norm": 0.06921602785587311, "learning_rate": 2.6054301479728036e-06, "loss": 0.0056, "step": 18040 }, { "grad_norm": 0.061177805066108704, "learning_rate": 2.579155700096575e-06, "loss": 0.0065, "step": 18050 }, { "grad_norm": 0.06774687021970749, "learning_rate": 2.5530108989978873e-06, "loss": 0.0043, "step": 18060 }, { "grad_norm": 0.0930180773139, "learning_rate": 2.5269958161556416e-06, "loss": 0.0053, "step": 18070 }, { "grad_norm": 0.07851269841194153, "learning_rate": 2.5011105226940888e-06, "loss": 0.0049, "step": 18080 }, { "grad_norm": 0.08097855001688004, "learning_rate": 2.4753550893826248e-06, "loss": 0.0059, "step": 18090 }, { "grad_norm": 0.06241879239678383, "learning_rate": 2.4497295866356296e-06, "loss": 0.0044, "step": 18100 }, { "grad_norm": 0.08909155428409576, "learning_rate": 2.424234084512228e-06, "loss": 0.0049, "step": 18110 }, { "grad_norm": 0.07696700841188431, "learning_rate": 2.3988686527161687e-06, "loss": 0.0049, "step": 18120 }, { "grad_norm": 0.07498790323734283, "learning_rate": 2.373633360595573e-06, "loss": 0.0045, "step": 18130 }, { "grad_norm": 0.0639505535364151, "learning_rate": 2.3485282771427585e-06, "loss": 0.0071, "step": 18140 }, { "grad_norm": 0.09106449037790298, "learning_rate": 2.3235534709940665e-06, "loss": 0.0035, "step": 18150 }, { "grad_norm": 0.08326848596334457, "learning_rate": 2.2987090104296617e-06, "loss": 0.0044, "step": 18160 }, { "grad_norm": 0.060185596346855164, "learning_rate": 2.273994963373355e-06, "loss": 0.0087, "step": 18170 }, { "grad_norm": 0.07413960248231888, "learning_rate": 2.249411397392409e-06, "loss": 0.0056, "step": 18180 }, { "grad_norm": 0.07386936247348785, "learning_rate": 2.2249583796973506e-06, "loss": 0.005, "step": 18190 }, { "grad_norm": 0.030898235738277435, "learning_rate": 2.200635977141796e-06, "loss": 0.0041, "step": 18200 }, { "grad_norm": 0.03212233632802963, "learning_rate": 2.17644425622226e-06, "loss": 0.0058, "step": 18210 }, { "grad_norm": 0.11856678128242493, "learning_rate": 2.152383283077991e-06, "loss": 0.0067, "step": 18220 }, { "grad_norm": 0.052465345710515976, "learning_rate": 2.128453123490781e-06, "loss": 0.0056, "step": 18230 }, { "grad_norm": 0.07447845488786697, "learning_rate": 2.1046538428847462e-06, "loss": 0.0051, "step": 18240 }, { "grad_norm": 0.06449835747480392, "learning_rate": 2.0809855063262273e-06, "loss": 0.0046, "step": 18250 }, { "grad_norm": 0.1206793561577797, "learning_rate": 2.057448178523558e-06, "loss": 0.0045, "step": 18260 }, { "grad_norm": 0.06678812950849533, "learning_rate": 2.034041923826885e-06, "loss": 0.005, "step": 18270 }, { "grad_norm": 0.06261760741472244, "learning_rate": 2.0107668062280204e-06, "loss": 0.0056, "step": 18280 }, { "grad_norm": 0.06050224229693413, "learning_rate": 1.9876228893602357e-06, "loss": 0.0051, "step": 18290 }, { "grad_norm": 0.05431816726922989, "learning_rate": 1.9646102364981266e-06, "loss": 0.0051, "step": 18300 }, { "grad_norm": 0.07837079465389252, "learning_rate": 1.9417289105574053e-06, "loss": 0.0055, "step": 18310 }, { "grad_norm": 0.07131019979715347, "learning_rate": 1.9189789740947427e-06, "loss": 0.0049, "step": 18320 }, { "grad_norm": 0.12421789765357971, "learning_rate": 1.896360489307597e-06, "loss": 0.0072, "step": 18330 }, { "grad_norm": 0.042724527418613434, "learning_rate": 1.8738735180340362e-06, "loss": 0.0051, "step": 18340 }, { "grad_norm": 0.07402852922677994, "learning_rate": 1.8515181217525824e-06, "loss": 0.0043, "step": 18350 }, { "grad_norm": 0.030661897733807564, "learning_rate": 1.8292943615820457e-06, "loss": 0.0044, "step": 18360 }, { "grad_norm": 0.06247671693563461, "learning_rate": 1.8072022982813296e-06, "loss": 0.0047, "step": 18370 }, { "grad_norm": 0.0922677144408226, "learning_rate": 1.7852419922492925e-06, "loss": 0.0062, "step": 18380 }, { "grad_norm": 0.08214414119720459, "learning_rate": 1.763413503524569e-06, "loss": 0.0054, "step": 18390 }, { "grad_norm": 0.05245650187134743, "learning_rate": 1.7417168917854165e-06, "loss": 0.0055, "step": 18400 }, { "grad_norm": 0.1169767752289772, "learning_rate": 1.720152216349552e-06, "loss": 0.0081, "step": 18410 }, { "grad_norm": 0.06361883878707886, "learning_rate": 1.6987195361739595e-06, "loss": 0.0045, "step": 18420 }, { "grad_norm": 0.10117273777723312, "learning_rate": 1.6774189098547832e-06, "loss": 0.0072, "step": 18430 }, { "grad_norm": 0.06884409487247467, "learning_rate": 1.6562503956271069e-06, "loss": 0.0041, "step": 18440 }, { "grad_norm": 0.11984467506408691, "learning_rate": 1.6352140513648417e-06, "loss": 0.0056, "step": 18450 }, { "grad_norm": 0.03949007764458656, "learning_rate": 1.6143099345805712e-06, "loss": 0.0056, "step": 18460 }, { "grad_norm": 0.04537283629179001, "learning_rate": 1.5935381024253293e-06, "loss": 0.0053, "step": 18470 }, { "grad_norm": 0.12034434080123901, "learning_rate": 1.572898611688517e-06, "loss": 0.0049, "step": 18480 }, { "grad_norm": 0.07979085296392441, "learning_rate": 1.5523915187977133e-06, "loss": 0.0053, "step": 18490 }, { "grad_norm": 0.04655562713742256, "learning_rate": 1.532016879818532e-06, "loss": 0.0056, "step": 18500 }, { "grad_norm": 0.10737938433885574, "learning_rate": 1.51177475045447e-06, "loss": 0.0059, "step": 18510 }, { "grad_norm": 0.06764158606529236, "learning_rate": 1.4916651860467035e-06, "loss": 0.0074, "step": 18520 }, { "grad_norm": 0.1326180398464203, "learning_rate": 1.471688241574043e-06, "loss": 0.0041, "step": 18530 }, { "grad_norm": 0.08991081267595291, "learning_rate": 1.451843971652672e-06, "loss": 0.0066, "step": 18540 }, { "grad_norm": 0.08501219749450684, "learning_rate": 1.432132430536076e-06, "loss": 0.0047, "step": 18550 }, { "grad_norm": 0.09303219616413116, "learning_rate": 1.412553672114869e-06, "loss": 0.0052, "step": 18560 }, { "grad_norm": 0.05906869098544121, "learning_rate": 1.3931077499166056e-06, "loss": 0.0069, "step": 18570 }, { "grad_norm": 0.03835143521428108, "learning_rate": 1.3737947171057085e-06, "loss": 0.004, "step": 18580 }, { "grad_norm": 0.10100048035383224, "learning_rate": 1.3546146264832582e-06, "loss": 0.0054, "step": 18590 }, { "grad_norm": 0.08225246518850327, "learning_rate": 1.3355675304869086e-06, "loss": 0.0046, "step": 18600 }, { "grad_norm": 0.04323491454124451, "learning_rate": 1.3166534811906827e-06, "loss": 0.0051, "step": 18610 }, { "grad_norm": 0.059321362525224686, "learning_rate": 1.2978725303048666e-06, "loss": 0.0039, "step": 18620 }, { "grad_norm": 0.07290496677160263, "learning_rate": 1.2792247291758762e-06, "loss": 0.005, "step": 18630 }, { "grad_norm": 0.0604105070233345, "learning_rate": 1.2607101287860635e-06, "loss": 0.005, "step": 18640 }, { "grad_norm": 0.0800408273935318, "learning_rate": 1.2423287797536654e-06, "loss": 0.0041, "step": 18650 }, { "grad_norm": 0.06597708910703659, "learning_rate": 1.2240807323325776e-06, "loss": 0.0044, "step": 18660 }, { "grad_norm": 0.14294591546058655, "learning_rate": 1.205966036412254e-06, "loss": 0.0032, "step": 18670 }, { "grad_norm": 0.07966846227645874, "learning_rate": 1.1879847415175949e-06, "loss": 0.0061, "step": 18680 }, { "grad_norm": 0.07043656706809998, "learning_rate": 1.1701368968087712e-06, "loss": 0.0055, "step": 18690 }, { "grad_norm": 0.06817292422056198, "learning_rate": 1.1524225510811116e-06, "loss": 0.0044, "step": 18700 }, { "grad_norm": 0.04660884663462639, "learning_rate": 1.1348417527649535e-06, "loss": 0.0039, "step": 18710 }, { "grad_norm": 0.10866593569517136, "learning_rate": 1.1173945499255268e-06, "loss": 0.0049, "step": 18720 }, { "grad_norm": 0.06027756631374359, "learning_rate": 1.1000809902628307e-06, "loss": 0.0033, "step": 18730 }, { "grad_norm": 0.053037941455841064, "learning_rate": 1.082901121111468e-06, "loss": 0.0069, "step": 18740 }, { "grad_norm": 0.206001877784729, "learning_rate": 1.0658549894405456e-06, "loss": 0.0058, "step": 18750 }, { "grad_norm": 0.06229070574045181, "learning_rate": 1.0489426418535342e-06, "loss": 0.0044, "step": 18760 }, { "grad_norm": 0.07082782685756683, "learning_rate": 1.0321641245881474e-06, "loss": 0.0056, "step": 18770 }, { "grad_norm": 0.4507696032524109, "learning_rate": 1.015519483516214e-06, "loss": 0.0055, "step": 18780 }, { "grad_norm": 0.05996058136224747, "learning_rate": 9.990087641435443e-07, "loss": 0.0052, "step": 18790 }, { "grad_norm": 0.17796622216701508, "learning_rate": 9.826320116098132e-07, "loss": 0.0067, "step": 18800 }, { "grad_norm": 0.09721294045448303, "learning_rate": 9.663892706884447e-07, "loss": 0.0055, "step": 18810 }, { "grad_norm": 0.09979432076215744, "learning_rate": 9.502805857864616e-07, "loss": 0.0051, "step": 18820 }, { "grad_norm": 0.09904192388057709, "learning_rate": 9.34306000944396e-07, "loss": 0.0045, "step": 18830 }, { "grad_norm": 0.057067129760980606, "learning_rate": 9.184655598361624e-07, "loss": 0.0053, "step": 18840 }, { "grad_norm": 0.15480734407901764, "learning_rate": 9.027593057689076e-07, "loss": 0.0062, "step": 18850 }, { "grad_norm": 0.04189983010292053, "learning_rate": 8.871872816829441e-07, "loss": 0.0047, "step": 18860 }, { "grad_norm": 0.07434888929128647, "learning_rate": 8.717495301515777e-07, "loss": 0.0037, "step": 18870 }, { "grad_norm": 0.041023120284080505, "learning_rate": 8.564460933810415e-07, "loss": 0.0044, "step": 18880 }, { "grad_norm": 0.15709863603115082, "learning_rate": 8.412770132103453e-07, "loss": 0.0045, "step": 18890 }, { "grad_norm": 0.04923916980624199, "learning_rate": 8.262423311111711e-07, "loss": 0.005, "step": 18900 }, { "grad_norm": 0.07005514204502106, "learning_rate": 8.113420881877665e-07, "loss": 0.0039, "step": 18910 }, { "grad_norm": 0.08594105392694473, "learning_rate": 7.965763251768288e-07, "loss": 0.0067, "step": 18920 }, { "grad_norm": 0.06526818871498108, "learning_rate": 7.819450824473995e-07, "loss": 0.005, "step": 18930 }, { "grad_norm": 0.048307787626981735, "learning_rate": 7.674484000007198e-07, "loss": 0.0039, "step": 18940 }, { "grad_norm": 0.058567676693201065, "learning_rate": 7.530863174701752e-07, "loss": 0.0059, "step": 18950 }, { "grad_norm": 0.05375925451517105, "learning_rate": 7.38858874121151e-07, "loss": 0.0048, "step": 18960 }, { "grad_norm": 0.05018064007163048, "learning_rate": 7.247661088509328e-07, "loss": 0.0069, "step": 18970 }, { "grad_norm": 0.07494211941957474, "learning_rate": 7.108080601886002e-07, "loss": 0.0059, "step": 18980 }, { "grad_norm": 0.1557677835226059, "learning_rate": 6.969847662949336e-07, "loss": 0.0062, "step": 18990 }, { "grad_norm": 0.04898068308830261, "learning_rate": 6.832962649622798e-07, "loss": 0.0047, "step": 19000 }, { "grad_norm": 0.064796581864357, "learning_rate": 6.697425936144863e-07, "loss": 0.0037, "step": 19010 }, { "grad_norm": 0.12094861268997192, "learning_rate": 6.563237893067731e-07, "loss": 0.0042, "step": 19020 }, { "grad_norm": 0.05841851979494095, "learning_rate": 6.430398887256328e-07, "loss": 0.0068, "step": 19030 }, { "grad_norm": 0.09714256972074509, "learning_rate": 6.298909281887478e-07, "loss": 0.0065, "step": 19040 }, { "grad_norm": 0.05553006753325462, "learning_rate": 6.168769436448673e-07, "loss": 0.005, "step": 19050 }, { "grad_norm": 0.10708530247211456, "learning_rate": 6.03997970673742e-07, "loss": 0.0082, "step": 19060 }, { "grad_norm": 0.09929335117340088, "learning_rate": 5.912540444859782e-07, "loss": 0.0056, "step": 19070 }, { "grad_norm": 0.08023998141288757, "learning_rate": 5.786451999229837e-07, "loss": 0.0056, "step": 19080 }, { "grad_norm": 0.08746291697025299, "learning_rate": 5.661714714568722e-07, "loss": 0.0055, "step": 19090 }, { "grad_norm": 0.09090830385684967, "learning_rate": 5.538328931903259e-07, "loss": 0.0066, "step": 19100 }, { "grad_norm": 0.09133350849151611, "learning_rate": 5.416294988565551e-07, "loss": 0.0082, "step": 19110 }, { "grad_norm": 0.06570694595575333, "learning_rate": 5.29561321819172e-07, "loss": 0.0054, "step": 19120 }, { "grad_norm": 0.09256274998188019, "learning_rate": 5.176283950721061e-07, "loss": 0.0053, "step": 19130 }, { "grad_norm": 0.09102249890565872, "learning_rate": 5.058307512395332e-07, "loss": 0.0061, "step": 19140 }, { "grad_norm": 0.062302254140377045, "learning_rate": 4.941684225757526e-07, "loss": 0.0062, "step": 19150 }, { "grad_norm": 0.04880385845899582, "learning_rate": 4.826414409651314e-07, "loss": 0.0054, "step": 19160 }, { "grad_norm": 0.055859196931123734, "learning_rate": 4.712498379219943e-07, "loss": 0.0038, "step": 19170 }, { "grad_norm": 0.07357624173164368, "learning_rate": 4.599936445905506e-07, "loss": 0.005, "step": 19180 }, { "grad_norm": 0.07378475368022919, "learning_rate": 4.4887289174480594e-07, "loss": 0.0067, "step": 19190 }, { "grad_norm": 0.08428577333688736, "learning_rate": 4.378876097884621e-07, "loss": 0.0068, "step": 19200 }, { "grad_norm": 0.0869697630405426, "learning_rate": 4.2703782875487264e-07, "loss": 0.0049, "step": 19210 }, { "grad_norm": 0.05357169359922409, "learning_rate": 4.163235783069208e-07, "loss": 0.0053, "step": 19220 }, { "grad_norm": 0.07352638989686966, "learning_rate": 4.057448877369585e-07, "loss": 0.0043, "step": 19230 }, { "grad_norm": 0.12299855053424835, "learning_rate": 3.9530178596672295e-07, "loss": 0.0053, "step": 19240 }, { "grad_norm": 0.19235359132289886, "learning_rate": 3.849943015472479e-07, "loss": 0.0069, "step": 19250 }, { "grad_norm": 0.05793895944952965, "learning_rate": 3.748224626588137e-07, "loss": 0.0061, "step": 19260 }, { "grad_norm": 0.08344893902540207, "learning_rate": 3.647862971108307e-07, "loss": 0.0037, "step": 19270 }, { "grad_norm": 0.13513457775115967, "learning_rate": 3.5488583234179473e-07, "loss": 0.006, "step": 19280 }, { "grad_norm": 0.04682665318250656, "learning_rate": 3.4512109541920413e-07, "loss": 0.0044, "step": 19290 }, { "grad_norm": 0.07172989845275879, "learning_rate": 3.354921130394706e-07, "loss": 0.006, "step": 19300 }, { "grad_norm": 0.046024806797504425, "learning_rate": 3.259989115278639e-07, "loss": 0.005, "step": 19310 }, { "grad_norm": 0.142673060297966, "learning_rate": 3.1664151683843403e-07, "loss": 0.0065, "step": 19320 }, { "grad_norm": 0.03470339998602867, "learning_rate": 3.074199545539447e-07, "loss": 0.0044, "step": 19330 }, { "grad_norm": 0.09273118525743484, "learning_rate": 2.983342498857955e-07, "loss": 0.0049, "step": 19340 }, { "grad_norm": 0.030313219875097275, "learning_rate": 2.893844276739499e-07, "loss": 0.0043, "step": 19350 }, { "grad_norm": 0.07048440724611282, "learning_rate": 2.8057051238688514e-07, "loss": 0.0041, "step": 19360 }, { "grad_norm": 0.1686461865901947, "learning_rate": 2.71892528121509e-07, "loss": 0.0092, "step": 19370 }, { "grad_norm": 0.10760743916034698, "learning_rate": 2.633504986030988e-07, "loss": 0.005, "step": 19380 }, { "grad_norm": 0.06864163279533386, "learning_rate": 2.549444471852347e-07, "loss": 0.0069, "step": 19390 }, { "grad_norm": 0.05440777912735939, "learning_rate": 2.4667439684974423e-07, "loss": 0.0063, "step": 19400 }, { "grad_norm": 0.06063523516058922, "learning_rate": 2.3854037020662467e-07, "loss": 0.0065, "step": 19410 }, { "grad_norm": 0.04062748700380325, "learning_rate": 2.3054238949399288e-07, "loss": 0.0053, "step": 19420 }, { "grad_norm": 0.07250646501779556, "learning_rate": 2.2268047657802993e-07, "loss": 0.0048, "step": 19430 }, { "grad_norm": 0.04116007313132286, "learning_rate": 2.149546529529034e-07, "loss": 0.0063, "step": 19440 }, { "grad_norm": 0.03562156856060028, "learning_rate": 2.0736493974071736e-07, "loss": 0.0034, "step": 19450 }, { "grad_norm": 0.08485918492078781, "learning_rate": 1.9991135769145686e-07, "loss": 0.0046, "step": 19460 }, { "grad_norm": 0.0834183320403099, "learning_rate": 1.9259392718293245e-07, "loss": 0.004, "step": 19470 }, { "grad_norm": 0.05375256389379501, "learning_rate": 1.8541266822072467e-07, "loss": 0.0044, "step": 19480 }, { "grad_norm": 0.03167695179581642, "learning_rate": 1.7836760043811184e-07, "loss": 0.0037, "step": 19490 }, { "grad_norm": 0.05684887245297432, "learning_rate": 1.7145874309604792e-07, "loss": 0.0036, "step": 19500 }, { "grad_norm": 0.08162637799978256, "learning_rate": 1.6468611508308474e-07, "loss": 0.0081, "step": 19510 }, { "grad_norm": 0.11516433954238892, "learning_rate": 1.5804973491532204e-07, "loss": 0.0043, "step": 19520 }, { "grad_norm": 0.13682876527309418, "learning_rate": 1.5154962073637424e-07, "loss": 0.0059, "step": 19530 }, { "grad_norm": 0.13256461918354034, "learning_rate": 1.4518579031730372e-07, "loss": 0.0036, "step": 19540 }, { "grad_norm": 0.07801523059606552, "learning_rate": 1.389582610565876e-07, "loss": 0.0053, "step": 19550 }, { "grad_norm": 0.11882422864437103, "learning_rate": 1.3286704998003995e-07, "loss": 0.0057, "step": 19560 }, { "grad_norm": 0.06176578253507614, "learning_rate": 1.2691217374080632e-07, "loss": 0.0059, "step": 19570 }, { "grad_norm": 0.08417752385139465, "learning_rate": 1.2109364861929705e-07, "loss": 0.0045, "step": 19580 }, { "grad_norm": 0.07603209465742111, "learning_rate": 1.1541149052312628e-07, "loss": 0.0037, "step": 19590 }, { "grad_norm": 0.0634196549654007, "learning_rate": 1.0986571498710074e-07, "loss": 0.0049, "step": 19600 }, { "grad_norm": 0.03994888812303543, "learning_rate": 1.0445633717316438e-07, "loss": 0.0054, "step": 19610 }, { "grad_norm": 0.055595844984054565, "learning_rate": 9.918337187034277e-08, "loss": 0.0041, "step": 19620 }, { "grad_norm": 0.08552831411361694, "learning_rate": 9.404683349472643e-08, "loss": 0.0059, "step": 19630 }, { "grad_norm": 0.06240672618150711, "learning_rate": 8.904673608940983e-08, "loss": 0.0036, "step": 19640 }, { "grad_norm": 0.06001579761505127, "learning_rate": 8.418309332447471e-08, "loss": 0.0056, "step": 19650 }, { "grad_norm": 0.07581871747970581, "learning_rate": 7.945591849692902e-08, "loss": 0.0074, "step": 19660 }, { "grad_norm": 0.11336157470941544, "learning_rate": 7.486522453069578e-08, "loss": 0.0069, "step": 19670 }, { "grad_norm": 0.060028690844774246, "learning_rate": 7.041102397655208e-08, "loss": 0.004, "step": 19680 }, { "grad_norm": 0.12431247532367706, "learning_rate": 6.609332901210685e-08, "loss": 0.0046, "step": 19690 }, { "grad_norm": 0.08701920509338379, "learning_rate": 6.191215144178419e-08, "loss": 0.0034, "step": 19700 }, { "grad_norm": 0.06494241207838058, "learning_rate": 5.786750269675678e-08, "loss": 0.0041, "step": 19710 }, { "grad_norm": 0.06485304981470108, "learning_rate": 5.395939383494031e-08, "loss": 0.0046, "step": 19720 }, { "grad_norm": 0.03299275040626526, "learning_rate": 5.018783554095463e-08, "loss": 0.0043, "step": 19730 }, { "grad_norm": 0.06459636241197586, "learning_rate": 4.655283812610156e-08, "loss": 0.0047, "step": 19740 }, { "grad_norm": 0.10682912915945053, "learning_rate": 4.305441152831491e-08, "loss": 0.0054, "step": 19750 }, { "grad_norm": 0.05271412432193756, "learning_rate": 3.9692565312171584e-08, "loss": 0.0078, "step": 19760 }, { "grad_norm": 0.07198057323694229, "learning_rate": 3.6467308668824975e-08, "loss": 0.0047, "step": 19770 }, { "grad_norm": 0.07922971248626709, "learning_rate": 3.3378650416004964e-08, "loss": 0.0041, "step": 19780 }, { "grad_norm": 0.053983330726623535, "learning_rate": 3.042659899797906e-08, "loss": 0.0064, "step": 19790 }, { "grad_norm": 0.11476776003837585, "learning_rate": 2.76111624855524e-08, "loss": 0.0059, "step": 19800 }, { "grad_norm": 0.13085563480854034, "learning_rate": 2.4932348576017784e-08, "loss": 0.0057, "step": 19810 }, { "grad_norm": 0.06645885854959488, "learning_rate": 2.239016459314458e-08, "loss": 0.0043, "step": 19820 }, { "grad_norm": 0.162321075797081, "learning_rate": 1.9984617487173174e-08, "loss": 0.0057, "step": 19830 }, { "grad_norm": 0.0987459272146225, "learning_rate": 1.7715713834776105e-08, "loss": 0.0053, "step": 19840 }, { "grad_norm": 0.0820586159825325, "learning_rate": 1.5583459839046964e-08, "loss": 0.0046, "step": 19850 }, { "grad_norm": 0.11219776421785355, "learning_rate": 1.3587861329489304e-08, "loss": 0.0052, "step": 19860 }, { "grad_norm": 0.05586789548397064, "learning_rate": 1.1728923761994415e-08, "loss": 0.0051, "step": 19870 }, { "grad_norm": 0.08609634637832642, "learning_rate": 1.0006652218819135e-08, "loss": 0.0051, "step": 19880 }, { "grad_norm": 0.08983197063207626, "learning_rate": 8.421051408596947e-09, "loss": 0.0049, "step": 19890 }, { "grad_norm": 0.04703143984079361, "learning_rate": 6.972125666299123e-09, "loss": 0.0052, "step": 19900 }, { "grad_norm": 0.05714048817753792, "learning_rate": 5.659878953229169e-09, "loss": 0.0067, "step": 19910 }, { "grad_norm": 0.06288735568523407, "learning_rate": 4.48431485701728e-09, "loss": 0.005, "step": 19920 }, { "grad_norm": 0.10251980274915695, "learning_rate": 3.4454365916203322e-09, "loss": 0.0081, "step": 19930 }, { "grad_norm": 0.02477150410413742, "learning_rate": 2.5432469972830332e-09, "loss": 0.0043, "step": 19940 }, { "grad_norm": 0.05386173352599144, "learning_rate": 1.7777485405601203e-09, "loss": 0.0048, "step": 19950 }, { "grad_norm": 0.12039000540971756, "learning_rate": 1.1489433142941597e-09, "loss": 0.0086, "step": 19960 }, { "grad_norm": 0.12772920727729797, "learning_rate": 6.568330376210963e-10, "loss": 0.0041, "step": 19970 }, { "grad_norm": 0.06290634721517563, "learning_rate": 3.0141905594249787e-10, "loss": 0.005, "step": 19980 }, { "grad_norm": 0.08486431837081909, "learning_rate": 8.270234094776008e-11, "loss": 0.0048, "step": 19990 }, { "grad_norm": 0.06710675358772278, "learning_rate": 6.834906085551041e-13, "loss": 0.0056, "step": 20000 } ], "logging_steps": 10, "max_steps": 20000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 2000, "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": 32, "trial_name": null, "trial_params": null }