{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.7, "eval_steps": 500, "global_step": 70000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.3155679404735565, "learning_rate": 1.8e-07, "loss": 1.1893, "step": 10 }, { "grad_norm": 0.3266640305519104, "learning_rate": 3.8e-07, "loss": 1.1978, "step": 20 }, { "grad_norm": 0.2689058184623718, "learning_rate": 5.8e-07, "loss": 1.1884, "step": 30 }, { "grad_norm": 0.30856195092201233, "learning_rate": 7.8e-07, "loss": 1.1935, "step": 40 }, { "grad_norm": 0.22488512098789215, "learning_rate": 9.8e-07, "loss": 1.1813, "step": 50 }, { "grad_norm": 0.251250296831131, "learning_rate": 1.18e-06, "loss": 1.1807, "step": 60 }, { "grad_norm": 0.19967932999134064, "learning_rate": 1.3800000000000001e-06, "loss": 1.166, "step": 70 }, { "grad_norm": 0.28779885172843933, "learning_rate": 1.5800000000000003e-06, "loss": 1.1599, "step": 80 }, { "grad_norm": 0.3612044155597687, "learning_rate": 1.7800000000000001e-06, "loss": 1.1335, "step": 90 }, { "grad_norm": 1.4725902080535889, "learning_rate": 1.98e-06, "loss": 1.1183, "step": 100 }, { "grad_norm": 2.2247154712677, "learning_rate": 2.1800000000000003e-06, "loss": 1.0902, "step": 110 }, { "grad_norm": 2.9289848804473877, "learning_rate": 2.38e-06, "loss": 1.074, "step": 120 }, { "grad_norm": 2.390022039413452, "learning_rate": 2.5800000000000003e-06, "loss": 1.0613, "step": 130 }, { "grad_norm": 1.4589418172836304, "learning_rate": 2.78e-06, "loss": 1.0507, "step": 140 }, { "grad_norm": 0.6305649280548096, "learning_rate": 2.9800000000000003e-06, "loss": 1.0445, "step": 150 }, { "grad_norm": 1.7754749059677124, "learning_rate": 3.1800000000000005e-06, "loss": 1.046, "step": 160 }, { "grad_norm": 0.7891651391983032, "learning_rate": 3.38e-06, "loss": 1.0343, "step": 170 }, { "grad_norm": 0.6461259126663208, "learning_rate": 3.58e-06, "loss": 1.0308, "step": 180 }, { "grad_norm": 0.6223688721656799, "learning_rate": 3.7800000000000002e-06, "loss": 1.036, "step": 190 }, { "grad_norm": 0.5365666151046753, "learning_rate": 3.98e-06, "loss": 1.0317, "step": 200 }, { "grad_norm": 0.7807283997535706, "learning_rate": 4.18e-06, "loss": 1.0293, "step": 210 }, { "grad_norm": 0.6570593118667603, "learning_rate": 4.38e-06, "loss": 1.0238, "step": 220 }, { "grad_norm": 0.759730339050293, "learning_rate": 4.58e-06, "loss": 1.0052, "step": 230 }, { "grad_norm": 0.8865121603012085, "learning_rate": 4.780000000000001e-06, "loss": 0.9835, "step": 240 }, { "grad_norm": 1.1477457284927368, "learning_rate": 4.98e-06, "loss": 0.9448, "step": 250 }, { "grad_norm": 0.9449464678764343, "learning_rate": 5.18e-06, "loss": 0.897, "step": 260 }, { "grad_norm": 0.9660567045211792, "learning_rate": 5.38e-06, "loss": 0.8511, "step": 270 }, { "grad_norm": 1.5280206203460693, "learning_rate": 5.580000000000001e-06, "loss": 0.8022, "step": 280 }, { "grad_norm": 1.0150240659713745, "learning_rate": 5.78e-06, "loss": 0.7622, "step": 290 }, { "grad_norm": 0.92184978723526, "learning_rate": 5.98e-06, "loss": 0.7186, "step": 300 }, { "grad_norm": 1.1917105913162231, "learning_rate": 6.18e-06, "loss": 0.6762, "step": 310 }, { "grad_norm": 1.667475700378418, "learning_rate": 6.38e-06, "loss": 0.6362, "step": 320 }, { "grad_norm": 1.7184104919433594, "learning_rate": 6.58e-06, "loss": 0.5994, "step": 330 }, { "grad_norm": 1.3695361614227295, "learning_rate": 6.78e-06, "loss": 0.5451, "step": 340 }, { "grad_norm": 1.7706633806228638, "learning_rate": 6.98e-06, "loss": 0.4982, "step": 350 }, { "grad_norm": 1.5146493911743164, "learning_rate": 7.180000000000001e-06, "loss": 0.4617, "step": 360 }, { "grad_norm": 1.7902448177337646, "learning_rate": 7.3800000000000005e-06, "loss": 0.4318, "step": 370 }, { "grad_norm": 1.7373090982437134, "learning_rate": 7.580000000000001e-06, "loss": 0.3983, "step": 380 }, { "grad_norm": 1.50590980052948, "learning_rate": 7.78e-06, "loss": 0.361, "step": 390 }, { "grad_norm": 1.8356488943099976, "learning_rate": 7.98e-06, "loss": 0.3202, "step": 400 }, { "grad_norm": 1.609304666519165, "learning_rate": 8.18e-06, "loss": 0.2813, "step": 410 }, { "grad_norm": 1.9034907817840576, "learning_rate": 8.380000000000001e-06, "loss": 0.2588, "step": 420 }, { "grad_norm": 1.6671228408813477, "learning_rate": 8.580000000000001e-06, "loss": 0.228, "step": 430 }, { "grad_norm": 1.3876415491104126, "learning_rate": 8.78e-06, "loss": 0.2044, "step": 440 }, { "grad_norm": 1.5633909702301025, "learning_rate": 8.98e-06, "loss": 0.1852, "step": 450 }, { "grad_norm": 1.5347857475280762, "learning_rate": 9.180000000000002e-06, "loss": 0.1689, "step": 460 }, { "grad_norm": 1.556295394897461, "learning_rate": 9.38e-06, "loss": 0.1663, "step": 470 }, { "grad_norm": 1.4373313188552856, "learning_rate": 9.58e-06, "loss": 0.1581, "step": 480 }, { "grad_norm": 1.5952221155166626, "learning_rate": 9.78e-06, "loss": 0.1513, "step": 490 }, { "grad_norm": 1.670055627822876, "learning_rate": 9.980000000000001e-06, "loss": 0.1517, "step": 500 }, { "grad_norm": 1.6640011072158813, "learning_rate": 1.018e-05, "loss": 0.1424, "step": 510 }, { "grad_norm": 1.6740251779556274, "learning_rate": 1.038e-05, "loss": 0.13, "step": 520 }, { "grad_norm": 1.6580290794372559, "learning_rate": 1.058e-05, "loss": 0.1257, "step": 530 }, { "grad_norm": 1.904288411140442, "learning_rate": 1.0780000000000002e-05, "loss": 0.1324, "step": 540 }, { "grad_norm": 1.3869507312774658, "learning_rate": 1.098e-05, "loss": 0.1209, "step": 550 }, { "grad_norm": 1.494876742362976, "learning_rate": 1.118e-05, "loss": 0.1171, "step": 560 }, { "grad_norm": 1.437233805656433, "learning_rate": 1.1380000000000001e-05, "loss": 0.1209, "step": 570 }, { "grad_norm": 1.63070809841156, "learning_rate": 1.1580000000000001e-05, "loss": 0.1141, "step": 580 }, { "grad_norm": 2.069399356842041, "learning_rate": 1.178e-05, "loss": 0.1153, "step": 590 }, { "grad_norm": 1.8604623079299927, "learning_rate": 1.198e-05, "loss": 0.1112, "step": 600 }, { "grad_norm": 1.4179965257644653, "learning_rate": 1.2180000000000002e-05, "loss": 0.1078, "step": 610 }, { "grad_norm": 1.7926766872406006, "learning_rate": 1.238e-05, "loss": 0.1052, "step": 620 }, { "grad_norm": 1.6470062732696533, "learning_rate": 1.258e-05, "loss": 0.1026, "step": 630 }, { "grad_norm": 1.4933853149414062, "learning_rate": 1.278e-05, "loss": 0.1036, "step": 640 }, { "grad_norm": 2.120213508605957, "learning_rate": 1.2980000000000001e-05, "loss": 0.0997, "step": 650 }, { "grad_norm": 1.759809970855713, "learning_rate": 1.3180000000000001e-05, "loss": 0.0958, "step": 660 }, { "grad_norm": 1.634535551071167, "learning_rate": 1.338e-05, "loss": 0.1017, "step": 670 }, { "grad_norm": 1.7793388366699219, "learning_rate": 1.358e-05, "loss": 0.0953, "step": 680 }, { "grad_norm": 1.9110687971115112, "learning_rate": 1.3780000000000002e-05, "loss": 0.0913, "step": 690 }, { "grad_norm": 1.758001685142517, "learning_rate": 1.3980000000000002e-05, "loss": 0.091, "step": 700 }, { "grad_norm": 1.8869411945343018, "learning_rate": 1.4180000000000001e-05, "loss": 0.0887, "step": 710 }, { "grad_norm": 1.605458378791809, "learning_rate": 1.4380000000000001e-05, "loss": 0.0887, "step": 720 }, { "grad_norm": 1.7304095029830933, "learning_rate": 1.4580000000000003e-05, "loss": 0.0918, "step": 730 }, { "grad_norm": 1.8110709190368652, "learning_rate": 1.4779999999999999e-05, "loss": 0.0878, "step": 740 }, { "grad_norm": 2.1882247924804688, "learning_rate": 1.4979999999999999e-05, "loss": 0.0871, "step": 750 }, { "grad_norm": 1.6739305257797241, "learning_rate": 1.518e-05, "loss": 0.0839, "step": 760 }, { "grad_norm": 1.602730631828308, "learning_rate": 1.538e-05, "loss": 0.0882, "step": 770 }, { "grad_norm": 1.4840781688690186, "learning_rate": 1.558e-05, "loss": 0.0851, "step": 780 }, { "grad_norm": 1.9383115768432617, "learning_rate": 1.578e-05, "loss": 0.0823, "step": 790 }, { "grad_norm": 1.9728667736053467, "learning_rate": 1.598e-05, "loss": 0.0849, "step": 800 }, { "grad_norm": 1.6790248155593872, "learning_rate": 1.618e-05, "loss": 0.0788, "step": 810 }, { "grad_norm": 1.4615627527236938, "learning_rate": 1.6380000000000002e-05, "loss": 0.0816, "step": 820 }, { "grad_norm": 1.8436310291290283, "learning_rate": 1.658e-05, "loss": 0.0759, "step": 830 }, { "grad_norm": 1.5640043020248413, "learning_rate": 1.6780000000000002e-05, "loss": 0.0787, "step": 840 }, { "grad_norm": 1.8923025131225586, "learning_rate": 1.698e-05, "loss": 0.0891, "step": 850 }, { "grad_norm": 1.6558880805969238, "learning_rate": 1.718e-05, "loss": 0.0783, "step": 860 }, { "grad_norm": 1.7887402772903442, "learning_rate": 1.7380000000000003e-05, "loss": 0.0742, "step": 870 }, { "grad_norm": 1.63361656665802, "learning_rate": 1.758e-05, "loss": 0.0796, "step": 880 }, { "grad_norm": 1.7615007162094116, "learning_rate": 1.7780000000000003e-05, "loss": 0.0754, "step": 890 }, { "grad_norm": 1.9158010482788086, "learning_rate": 1.798e-05, "loss": 0.0756, "step": 900 }, { "grad_norm": 1.8283170461654663, "learning_rate": 1.818e-05, "loss": 0.0733, "step": 910 }, { "grad_norm": 1.8015525341033936, "learning_rate": 1.838e-05, "loss": 0.072, "step": 920 }, { "grad_norm": 1.5374464988708496, "learning_rate": 1.858e-05, "loss": 0.0733, "step": 930 }, { "grad_norm": 1.7557741403579712, "learning_rate": 1.878e-05, "loss": 0.0694, "step": 940 }, { "grad_norm": 1.3546042442321777, "learning_rate": 1.898e-05, "loss": 0.0712, "step": 950 }, { "grad_norm": 1.7642701864242554, "learning_rate": 1.918e-05, "loss": 0.0701, "step": 960 }, { "grad_norm": 1.193820595741272, "learning_rate": 1.938e-05, "loss": 0.0709, "step": 970 }, { "grad_norm": 1.96892511844635, "learning_rate": 1.9580000000000002e-05, "loss": 0.0685, "step": 980 }, { "grad_norm": 2.0533387660980225, "learning_rate": 1.978e-05, "loss": 0.0706, "step": 990 }, { "grad_norm": 1.571601152420044, "learning_rate": 1.9980000000000002e-05, "loss": 0.0679, "step": 1000 }, { "grad_norm": 1.5082200765609741, "learning_rate": 2.0180000000000003e-05, "loss": 0.0684, "step": 1010 }, { "grad_norm": 1.5843249559402466, "learning_rate": 2.038e-05, "loss": 0.0665, "step": 1020 }, { "grad_norm": 1.8364348411560059, "learning_rate": 2.0580000000000003e-05, "loss": 0.0661, "step": 1030 }, { "grad_norm": 1.6813980340957642, "learning_rate": 2.078e-05, "loss": 0.0652, "step": 1040 }, { "grad_norm": 1.7163009643554688, "learning_rate": 2.098e-05, "loss": 0.0666, "step": 1050 }, { "grad_norm": 1.4757578372955322, "learning_rate": 2.118e-05, "loss": 0.0626, "step": 1060 }, { "grad_norm": 1.4533429145812988, "learning_rate": 2.138e-05, "loss": 0.0661, "step": 1070 }, { "grad_norm": 2.0234620571136475, "learning_rate": 2.158e-05, "loss": 0.062, "step": 1080 }, { "grad_norm": 1.5727792978286743, "learning_rate": 2.178e-05, "loss": 0.0624, "step": 1090 }, { "grad_norm": 1.6202192306518555, "learning_rate": 2.198e-05, "loss": 0.0631, "step": 1100 }, { "grad_norm": 1.374827265739441, "learning_rate": 2.218e-05, "loss": 0.0664, "step": 1110 }, { "grad_norm": 1.7237327098846436, "learning_rate": 2.2380000000000003e-05, "loss": 0.0604, "step": 1120 }, { "grad_norm": 1.4992693662643433, "learning_rate": 2.258e-05, "loss": 0.0635, "step": 1130 }, { "grad_norm": 1.5547780990600586, "learning_rate": 2.2780000000000002e-05, "loss": 0.0602, "step": 1140 }, { "grad_norm": 1.2544341087341309, "learning_rate": 2.298e-05, "loss": 0.062, "step": 1150 }, { "grad_norm": 1.4787145853042603, "learning_rate": 2.318e-05, "loss": 0.0626, "step": 1160 }, { "grad_norm": 1.570105791091919, "learning_rate": 2.3380000000000003e-05, "loss": 0.0633, "step": 1170 }, { "grad_norm": 1.2393437623977661, "learning_rate": 2.358e-05, "loss": 0.0633, "step": 1180 }, { "grad_norm": 1.9062597751617432, "learning_rate": 2.3780000000000003e-05, "loss": 0.0606, "step": 1190 }, { "grad_norm": 1.0337529182434082, "learning_rate": 2.398e-05, "loss": 0.0606, "step": 1200 }, { "grad_norm": 1.6141263246536255, "learning_rate": 2.418e-05, "loss": 0.0605, "step": 1210 }, { "grad_norm": 1.4966527223587036, "learning_rate": 2.438e-05, "loss": 0.0587, "step": 1220 }, { "grad_norm": 1.4994752407073975, "learning_rate": 2.4580000000000002e-05, "loss": 0.0582, "step": 1230 }, { "grad_norm": 1.2878470420837402, "learning_rate": 2.478e-05, "loss": 0.0583, "step": 1240 }, { "grad_norm": 1.3733636140823364, "learning_rate": 2.498e-05, "loss": 0.0608, "step": 1250 }, { "grad_norm": 1.3424761295318604, "learning_rate": 2.5180000000000003e-05, "loss": 0.0576, "step": 1260 }, { "grad_norm": 1.4654879570007324, "learning_rate": 2.5380000000000004e-05, "loss": 0.0571, "step": 1270 }, { "grad_norm": 1.5452300310134888, "learning_rate": 2.5580000000000002e-05, "loss": 0.0613, "step": 1280 }, { "grad_norm": 1.3217419385910034, "learning_rate": 2.5779999999999997e-05, "loss": 0.057, "step": 1290 }, { "grad_norm": 1.5481114387512207, "learning_rate": 2.598e-05, "loss": 0.0579, "step": 1300 }, { "grad_norm": 1.346487045288086, "learning_rate": 2.618e-05, "loss": 0.0577, "step": 1310 }, { "grad_norm": 1.518432378768921, "learning_rate": 2.6379999999999998e-05, "loss": 0.0561, "step": 1320 }, { "grad_norm": 1.6993430852890015, "learning_rate": 2.658e-05, "loss": 0.0568, "step": 1330 }, { "grad_norm": 1.3489962816238403, "learning_rate": 2.678e-05, "loss": 0.056, "step": 1340 }, { "grad_norm": 1.6424155235290527, "learning_rate": 2.698e-05, "loss": 0.0561, "step": 1350 }, { "grad_norm": 1.3364284038543701, "learning_rate": 2.718e-05, "loss": 0.0568, "step": 1360 }, { "grad_norm": 1.2581028938293457, "learning_rate": 2.738e-05, "loss": 0.0541, "step": 1370 }, { "grad_norm": 1.318226933479309, "learning_rate": 2.758e-05, "loss": 0.0525, "step": 1380 }, { "grad_norm": 1.5155078172683716, "learning_rate": 2.778e-05, "loss": 0.0541, "step": 1390 }, { "grad_norm": 1.4396296739578247, "learning_rate": 2.798e-05, "loss": 0.0543, "step": 1400 }, { "grad_norm": 1.1951783895492554, "learning_rate": 2.818e-05, "loss": 0.0563, "step": 1410 }, { "grad_norm": 1.120008945465088, "learning_rate": 2.8380000000000003e-05, "loss": 0.0498, "step": 1420 }, { "grad_norm": 1.7680693864822388, "learning_rate": 2.858e-05, "loss": 0.056, "step": 1430 }, { "grad_norm": 1.2508487701416016, "learning_rate": 2.8780000000000002e-05, "loss": 0.0521, "step": 1440 }, { "grad_norm": 1.5528719425201416, "learning_rate": 2.898e-05, "loss": 0.0519, "step": 1450 }, { "grad_norm": 1.5646322965621948, "learning_rate": 2.9180000000000002e-05, "loss": 0.0516, "step": 1460 }, { "grad_norm": 1.3743377923965454, "learning_rate": 2.9380000000000003e-05, "loss": 0.0585, "step": 1470 }, { "grad_norm": 1.4734023809432983, "learning_rate": 2.958e-05, "loss": 0.053, "step": 1480 }, { "grad_norm": 1.4373289346694946, "learning_rate": 2.9780000000000003e-05, "loss": 0.0491, "step": 1490 }, { "grad_norm": 1.3140314817428589, "learning_rate": 2.998e-05, "loss": 0.0498, "step": 1500 }, { "grad_norm": 1.4258960485458374, "learning_rate": 3.0180000000000002e-05, "loss": 0.0493, "step": 1510 }, { "grad_norm": 1.5305309295654297, "learning_rate": 3.0380000000000004e-05, "loss": 0.0514, "step": 1520 }, { "grad_norm": 1.1582916975021362, "learning_rate": 3.058e-05, "loss": 0.0499, "step": 1530 }, { "grad_norm": 1.3921403884887695, "learning_rate": 3.078e-05, "loss": 0.0495, "step": 1540 }, { "grad_norm": 1.1291046142578125, "learning_rate": 3.0980000000000005e-05, "loss": 0.0526, "step": 1550 }, { "grad_norm": 1.053963303565979, "learning_rate": 3.118e-05, "loss": 0.0511, "step": 1560 }, { "grad_norm": 1.43572199344635, "learning_rate": 3.138e-05, "loss": 0.0535, "step": 1570 }, { "grad_norm": 1.339142918586731, "learning_rate": 3.1580000000000006e-05, "loss": 0.0494, "step": 1580 }, { "grad_norm": 1.0345714092254639, "learning_rate": 3.1780000000000004e-05, "loss": 0.0484, "step": 1590 }, { "grad_norm": 1.1243594884872437, "learning_rate": 3.198e-05, "loss": 0.049, "step": 1600 }, { "grad_norm": 0.9229004383087158, "learning_rate": 3.218e-05, "loss": 0.048, "step": 1610 }, { "grad_norm": 1.804132342338562, "learning_rate": 3.238e-05, "loss": 0.0493, "step": 1620 }, { "grad_norm": 1.297217607498169, "learning_rate": 3.2579999999999996e-05, "loss": 0.0492, "step": 1630 }, { "grad_norm": 0.956915020942688, "learning_rate": 3.278e-05, "loss": 0.0491, "step": 1640 }, { "grad_norm": 0.7832722067832947, "learning_rate": 3.298e-05, "loss": 0.0526, "step": 1650 }, { "grad_norm": 1.2906845808029175, "learning_rate": 3.318e-05, "loss": 0.0484, "step": 1660 }, { "grad_norm": 1.2210208177566528, "learning_rate": 3.338e-05, "loss": 0.0475, "step": 1670 }, { "grad_norm": 0.8197080492973328, "learning_rate": 3.358e-05, "loss": 0.0466, "step": 1680 }, { "grad_norm": 1.0556492805480957, "learning_rate": 3.378e-05, "loss": 0.0487, "step": 1690 }, { "grad_norm": 1.2253663539886475, "learning_rate": 3.398e-05, "loss": 0.0479, "step": 1700 }, { "grad_norm": 1.1374088525772095, "learning_rate": 3.418e-05, "loss": 0.0446, "step": 1710 }, { "grad_norm": 1.1550720930099487, "learning_rate": 3.438e-05, "loss": 0.0523, "step": 1720 }, { "grad_norm": 1.221623182296753, "learning_rate": 3.4580000000000004e-05, "loss": 0.0507, "step": 1730 }, { "grad_norm": 1.0140117406845093, "learning_rate": 3.478e-05, "loss": 0.0457, "step": 1740 }, { "grad_norm": 1.1102511882781982, "learning_rate": 3.498e-05, "loss": 0.0465, "step": 1750 }, { "grad_norm": 1.117789626121521, "learning_rate": 3.518e-05, "loss": 0.0469, "step": 1760 }, { "grad_norm": 1.0770186185836792, "learning_rate": 3.5380000000000003e-05, "loss": 0.0475, "step": 1770 }, { "grad_norm": 1.1315999031066895, "learning_rate": 3.558e-05, "loss": 0.0467, "step": 1780 }, { "grad_norm": 1.0082201957702637, "learning_rate": 3.578e-05, "loss": 0.0479, "step": 1790 }, { "grad_norm": 1.2361257076263428, "learning_rate": 3.5980000000000004e-05, "loss": 0.0477, "step": 1800 }, { "grad_norm": 0.9000566005706787, "learning_rate": 3.618e-05, "loss": 0.046, "step": 1810 }, { "grad_norm": 0.9557439088821411, "learning_rate": 3.638e-05, "loss": 0.0445, "step": 1820 }, { "grad_norm": 1.1817715167999268, "learning_rate": 3.6580000000000006e-05, "loss": 0.0458, "step": 1830 }, { "grad_norm": 1.0437337160110474, "learning_rate": 3.6780000000000004e-05, "loss": 0.0453, "step": 1840 }, { "grad_norm": 1.1861902475357056, "learning_rate": 3.698e-05, "loss": 0.0457, "step": 1850 }, { "grad_norm": 0.9860502481460571, "learning_rate": 3.7180000000000007e-05, "loss": 0.0449, "step": 1860 }, { "grad_norm": 0.8514456748962402, "learning_rate": 3.7380000000000005e-05, "loss": 0.045, "step": 1870 }, { "grad_norm": 1.1115258932113647, "learning_rate": 3.758e-05, "loss": 0.0454, "step": 1880 }, { "grad_norm": 1.301672101020813, "learning_rate": 3.778000000000001e-05, "loss": 0.0436, "step": 1890 }, { "grad_norm": 1.0609159469604492, "learning_rate": 3.7980000000000006e-05, "loss": 0.0465, "step": 1900 }, { "grad_norm": 0.9663927555084229, "learning_rate": 3.818e-05, "loss": 0.0432, "step": 1910 }, { "grad_norm": 0.7759345769882202, "learning_rate": 3.838e-05, "loss": 0.0438, "step": 1920 }, { "grad_norm": 1.0238251686096191, "learning_rate": 3.858e-05, "loss": 0.0431, "step": 1930 }, { "grad_norm": 1.2519687414169312, "learning_rate": 3.878e-05, "loss": 0.0452, "step": 1940 }, { "grad_norm": 0.9356254935264587, "learning_rate": 3.898e-05, "loss": 0.044, "step": 1950 }, { "grad_norm": 0.9569411277770996, "learning_rate": 3.918e-05, "loss": 0.0481, "step": 1960 }, { "grad_norm": 1.3545153141021729, "learning_rate": 3.938e-05, "loss": 0.0412, "step": 1970 }, { "grad_norm": 0.7434938549995422, "learning_rate": 3.958e-05, "loss": 0.0415, "step": 1980 }, { "grad_norm": 1.4054657220840454, "learning_rate": 3.978e-05, "loss": 0.0451, "step": 1990 }, { "grad_norm": 1.1333407163619995, "learning_rate": 3.998e-05, "loss": 0.0472, "step": 2000 }, { "grad_norm": 0.995876669883728, "learning_rate": 4.018e-05, "loss": 0.0434, "step": 2010 }, { "grad_norm": 1.0307773351669312, "learning_rate": 4.038e-05, "loss": 0.0407, "step": 2020 }, { "grad_norm": 1.2355519533157349, "learning_rate": 4.058e-05, "loss": 0.0407, "step": 2030 }, { "grad_norm": 0.9694035649299622, "learning_rate": 4.078e-05, "loss": 0.041, "step": 2040 }, { "grad_norm": 0.9844470024108887, "learning_rate": 4.0980000000000004e-05, "loss": 0.0421, "step": 2050 }, { "grad_norm": 0.926734983921051, "learning_rate": 4.118e-05, "loss": 0.0419, "step": 2060 }, { "grad_norm": 0.6805016398429871, "learning_rate": 4.138e-05, "loss": 0.0415, "step": 2070 }, { "grad_norm": 0.9809215068817139, "learning_rate": 4.1580000000000005e-05, "loss": 0.0426, "step": 2080 }, { "grad_norm": 1.0357927083969116, "learning_rate": 4.178e-05, "loss": 0.0408, "step": 2090 }, { "grad_norm": 1.08439040184021, "learning_rate": 4.198e-05, "loss": 0.0405, "step": 2100 }, { "grad_norm": 0.9300412535667419, "learning_rate": 4.2180000000000006e-05, "loss": 0.0397, "step": 2110 }, { "grad_norm": 0.6454827189445496, "learning_rate": 4.2380000000000004e-05, "loss": 0.0402, "step": 2120 }, { "grad_norm": 0.8231477737426758, "learning_rate": 4.258e-05, "loss": 0.0407, "step": 2130 }, { "grad_norm": 0.7849143743515015, "learning_rate": 4.278e-05, "loss": 0.0373, "step": 2140 }, { "grad_norm": 0.7947427034378052, "learning_rate": 4.2980000000000005e-05, "loss": 0.039, "step": 2150 }, { "grad_norm": 0.9772719740867615, "learning_rate": 4.318e-05, "loss": 0.041, "step": 2160 }, { "grad_norm": 0.8835431933403015, "learning_rate": 4.338e-05, "loss": 0.0433, "step": 2170 }, { "grad_norm": 0.9185463190078735, "learning_rate": 4.3580000000000006e-05, "loss": 0.0377, "step": 2180 }, { "grad_norm": 0.8424839973449707, "learning_rate": 4.3780000000000004e-05, "loss": 0.0404, "step": 2190 }, { "grad_norm": 0.7314537167549133, "learning_rate": 4.398e-05, "loss": 0.0388, "step": 2200 }, { "grad_norm": 0.8899115324020386, "learning_rate": 4.418000000000001e-05, "loss": 0.0392, "step": 2210 }, { "grad_norm": 0.98807293176651, "learning_rate": 4.438e-05, "loss": 0.0397, "step": 2220 }, { "grad_norm": 0.8369945883750916, "learning_rate": 4.458e-05, "loss": 0.0385, "step": 2230 }, { "grad_norm": 0.8178835511207581, "learning_rate": 4.478e-05, "loss": 0.0396, "step": 2240 }, { "grad_norm": 0.9960021376609802, "learning_rate": 4.498e-05, "loss": 0.0392, "step": 2250 }, { "grad_norm": 0.9625992178916931, "learning_rate": 4.518e-05, "loss": 0.0375, "step": 2260 }, { "grad_norm": 0.7352570295333862, "learning_rate": 4.538e-05, "loss": 0.038, "step": 2270 }, { "grad_norm": 0.8439925312995911, "learning_rate": 4.558e-05, "loss": 0.0384, "step": 2280 }, { "grad_norm": 0.7421560883522034, "learning_rate": 4.578e-05, "loss": 0.0391, "step": 2290 }, { "grad_norm": 0.7173661589622498, "learning_rate": 4.5980000000000004e-05, "loss": 0.041, "step": 2300 }, { "grad_norm": 0.787682056427002, "learning_rate": 4.618e-05, "loss": 0.0397, "step": 2310 }, { "grad_norm": 0.7403576970100403, "learning_rate": 4.638e-05, "loss": 0.0382, "step": 2320 }, { "grad_norm": 1.0521972179412842, "learning_rate": 4.6580000000000005e-05, "loss": 0.0393, "step": 2330 }, { "grad_norm": 0.7661939859390259, "learning_rate": 4.678e-05, "loss": 0.0367, "step": 2340 }, { "grad_norm": 0.8818099498748779, "learning_rate": 4.698e-05, "loss": 0.0399, "step": 2350 }, { "grad_norm": 0.9937764406204224, "learning_rate": 4.718e-05, "loss": 0.0388, "step": 2360 }, { "grad_norm": 0.887076735496521, "learning_rate": 4.7380000000000004e-05, "loss": 0.0381, "step": 2370 }, { "grad_norm": 0.6500508785247803, "learning_rate": 4.758e-05, "loss": 0.0378, "step": 2380 }, { "grad_norm": 0.7943463921546936, "learning_rate": 4.778e-05, "loss": 0.0384, "step": 2390 }, { "grad_norm": 0.6810513138771057, "learning_rate": 4.7980000000000005e-05, "loss": 0.0379, "step": 2400 }, { "grad_norm": 0.6117453575134277, "learning_rate": 4.818e-05, "loss": 0.0394, "step": 2410 }, { "grad_norm": 0.7734097838401794, "learning_rate": 4.838e-05, "loss": 0.0382, "step": 2420 }, { "grad_norm": 0.7980596423149109, "learning_rate": 4.8580000000000006e-05, "loss": 0.0363, "step": 2430 }, { "grad_norm": 0.7531201243400574, "learning_rate": 4.8780000000000004e-05, "loss": 0.035, "step": 2440 }, { "grad_norm": 0.6339598894119263, "learning_rate": 4.898e-05, "loss": 0.0342, "step": 2450 }, { "grad_norm": 0.8631414771080017, "learning_rate": 4.918000000000001e-05, "loss": 0.0371, "step": 2460 }, { "grad_norm": 0.7743073105812073, "learning_rate": 4.9380000000000005e-05, "loss": 0.0365, "step": 2470 }, { "grad_norm": 0.6172063946723938, "learning_rate": 4.958e-05, "loss": 0.0384, "step": 2480 }, { "grad_norm": 0.8072734475135803, "learning_rate": 4.978e-05, "loss": 0.0356, "step": 2490 }, { "grad_norm": 0.9495946764945984, "learning_rate": 4.9980000000000006e-05, "loss": 0.0366, "step": 2500 }, { "grad_norm": 0.7589433789253235, "learning_rate": 5.0180000000000004e-05, "loss": 0.0376, "step": 2510 }, { "grad_norm": 0.8885795474052429, "learning_rate": 5.038e-05, "loss": 0.0362, "step": 2520 }, { "grad_norm": 0.7948710918426514, "learning_rate": 5.058000000000001e-05, "loss": 0.0343, "step": 2530 }, { "grad_norm": 0.5632930397987366, "learning_rate": 5.0780000000000005e-05, "loss": 0.033, "step": 2540 }, { "grad_norm": 0.5118235349655151, "learning_rate": 5.098e-05, "loss": 0.0334, "step": 2550 }, { "grad_norm": 0.8013012409210205, "learning_rate": 5.118000000000001e-05, "loss": 0.0337, "step": 2560 }, { "grad_norm": 0.49149155616760254, "learning_rate": 5.1380000000000006e-05, "loss": 0.0352, "step": 2570 }, { "grad_norm": 0.801365852355957, "learning_rate": 5.1580000000000004e-05, "loss": 0.0339, "step": 2580 }, { "grad_norm": 0.7871430516242981, "learning_rate": 5.178000000000001e-05, "loss": 0.0319, "step": 2590 }, { "grad_norm": 0.7817251086235046, "learning_rate": 5.198000000000001e-05, "loss": 0.0343, "step": 2600 }, { "grad_norm": 0.7032462954521179, "learning_rate": 5.2180000000000005e-05, "loss": 0.0355, "step": 2610 }, { "grad_norm": 0.6885907649993896, "learning_rate": 5.238000000000001e-05, "loss": 0.0347, "step": 2620 }, { "grad_norm": 0.8060640096664429, "learning_rate": 5.258000000000001e-05, "loss": 0.036, "step": 2630 }, { "grad_norm": 1.0699681043624878, "learning_rate": 5.2780000000000006e-05, "loss": 0.0367, "step": 2640 }, { "grad_norm": 0.667715311050415, "learning_rate": 5.2980000000000004e-05, "loss": 0.0344, "step": 2650 }, { "grad_norm": 0.6152563095092773, "learning_rate": 5.318000000000001e-05, "loss": 0.0332, "step": 2660 }, { "grad_norm": 0.8018913865089417, "learning_rate": 5.338000000000001e-05, "loss": 0.0352, "step": 2670 }, { "grad_norm": 0.7187061905860901, "learning_rate": 5.3580000000000005e-05, "loss": 0.0318, "step": 2680 }, { "grad_norm": 0.7378198504447937, "learning_rate": 5.378e-05, "loss": 0.0327, "step": 2690 }, { "grad_norm": 0.8270434141159058, "learning_rate": 5.3979999999999995e-05, "loss": 0.0337, "step": 2700 }, { "grad_norm": 0.872255802154541, "learning_rate": 5.418e-05, "loss": 0.0347, "step": 2710 }, { "grad_norm": 0.5693914890289307, "learning_rate": 5.438e-05, "loss": 0.0323, "step": 2720 }, { "grad_norm": 0.533828854560852, "learning_rate": 5.4579999999999996e-05, "loss": 0.0329, "step": 2730 }, { "grad_norm": 0.6282193064689636, "learning_rate": 5.478e-05, "loss": 0.0323, "step": 2740 }, { "grad_norm": 0.602408230304718, "learning_rate": 5.498e-05, "loss": 0.0333, "step": 2750 }, { "grad_norm": 0.6772598624229431, "learning_rate": 5.518e-05, "loss": 0.0311, "step": 2760 }, { "grad_norm": 0.8885412216186523, "learning_rate": 5.538e-05, "loss": 0.034, "step": 2770 }, { "grad_norm": 0.5751360058784485, "learning_rate": 5.558e-05, "loss": 0.0352, "step": 2780 }, { "grad_norm": 0.6153193712234497, "learning_rate": 5.578e-05, "loss": 0.0373, "step": 2790 }, { "grad_norm": 0.6047281622886658, "learning_rate": 5.5979999999999996e-05, "loss": 0.0359, "step": 2800 }, { "grad_norm": 0.612216591835022, "learning_rate": 5.618e-05, "loss": 0.0356, "step": 2810 }, { "grad_norm": 0.6737085580825806, "learning_rate": 5.638e-05, "loss": 0.0333, "step": 2820 }, { "grad_norm": 0.7650089263916016, "learning_rate": 5.658e-05, "loss": 0.0337, "step": 2830 }, { "grad_norm": 0.8075273036956787, "learning_rate": 5.678e-05, "loss": 0.0338, "step": 2840 }, { "grad_norm": 0.7780057787895203, "learning_rate": 5.698e-05, "loss": 0.0307, "step": 2850 }, { "grad_norm": 0.7753836512565613, "learning_rate": 5.718e-05, "loss": 0.0319, "step": 2860 }, { "grad_norm": 0.7246688604354858, "learning_rate": 5.738e-05, "loss": 0.0355, "step": 2870 }, { "grad_norm": 0.433624267578125, "learning_rate": 5.758e-05, "loss": 0.0306, "step": 2880 }, { "grad_norm": 0.5517234206199646, "learning_rate": 5.778e-05, "loss": 0.0308, "step": 2890 }, { "grad_norm": 0.4273054003715515, "learning_rate": 5.7980000000000004e-05, "loss": 0.0292, "step": 2900 }, { "grad_norm": 0.6811787486076355, "learning_rate": 5.818e-05, "loss": 0.0318, "step": 2910 }, { "grad_norm": 0.6422380208969116, "learning_rate": 5.838e-05, "loss": 0.0297, "step": 2920 }, { "grad_norm": 0.4313010275363922, "learning_rate": 5.858e-05, "loss": 0.0318, "step": 2930 }, { "grad_norm": 0.5575766563415527, "learning_rate": 5.878e-05, "loss": 0.0301, "step": 2940 }, { "grad_norm": 0.6371253132820129, "learning_rate": 5.898e-05, "loss": 0.0302, "step": 2950 }, { "grad_norm": 0.5276098251342773, "learning_rate": 5.918e-05, "loss": 0.0316, "step": 2960 }, { "grad_norm": 0.6887551546096802, "learning_rate": 5.9380000000000004e-05, "loss": 0.033, "step": 2970 }, { "grad_norm": 0.7494754791259766, "learning_rate": 5.958e-05, "loss": 0.0339, "step": 2980 }, { "grad_norm": 0.8238636255264282, "learning_rate": 5.978e-05, "loss": 0.0338, "step": 2990 }, { "grad_norm": 0.6555530428886414, "learning_rate": 5.9980000000000005e-05, "loss": 0.0336, "step": 3000 }, { "grad_norm": 0.5792906284332275, "learning_rate": 6.018e-05, "loss": 0.034, "step": 3010 }, { "grad_norm": 0.532444953918457, "learning_rate": 6.038e-05, "loss": 0.034, "step": 3020 }, { "grad_norm": 0.6888946294784546, "learning_rate": 6.0580000000000006e-05, "loss": 0.0339, "step": 3030 }, { "grad_norm": 0.5408211946487427, "learning_rate": 6.0780000000000004e-05, "loss": 0.0338, "step": 3040 }, { "grad_norm": 0.6607675552368164, "learning_rate": 6.098e-05, "loss": 0.0326, "step": 3050 }, { "grad_norm": 0.824912965297699, "learning_rate": 6.118000000000001e-05, "loss": 0.0336, "step": 3060 }, { "grad_norm": 0.5933672785758972, "learning_rate": 6.138e-05, "loss": 0.0319, "step": 3070 }, { "grad_norm": 0.4981558322906494, "learning_rate": 6.158e-05, "loss": 0.0301, "step": 3080 }, { "grad_norm": 0.6639956831932068, "learning_rate": 6.178000000000001e-05, "loss": 0.03, "step": 3090 }, { "grad_norm": 0.6950365304946899, "learning_rate": 6.198e-05, "loss": 0.0331, "step": 3100 }, { "grad_norm": 0.6207664608955383, "learning_rate": 6.218e-05, "loss": 0.0305, "step": 3110 }, { "grad_norm": 0.5070662498474121, "learning_rate": 6.238000000000001e-05, "loss": 0.0306, "step": 3120 }, { "grad_norm": 0.3804147243499756, "learning_rate": 6.258e-05, "loss": 0.028, "step": 3130 }, { "grad_norm": 0.5123133659362793, "learning_rate": 6.278e-05, "loss": 0.0269, "step": 3140 }, { "grad_norm": 0.5083014369010925, "learning_rate": 6.298000000000001e-05, "loss": 0.0286, "step": 3150 }, { "grad_norm": 0.5273689031600952, "learning_rate": 6.318e-05, "loss": 0.0285, "step": 3160 }, { "grad_norm": 0.40397027134895325, "learning_rate": 6.338e-05, "loss": 0.0282, "step": 3170 }, { "grad_norm": 0.7971815466880798, "learning_rate": 6.358000000000001e-05, "loss": 0.0315, "step": 3180 }, { "grad_norm": 0.7000440955162048, "learning_rate": 6.378e-05, "loss": 0.0298, "step": 3190 }, { "grad_norm": 0.5718070864677429, "learning_rate": 6.398000000000001e-05, "loss": 0.0324, "step": 3200 }, { "grad_norm": 0.5064878463745117, "learning_rate": 6.418000000000001e-05, "loss": 0.0308, "step": 3210 }, { "grad_norm": 0.49714067578315735, "learning_rate": 6.438e-05, "loss": 0.0277, "step": 3220 }, { "grad_norm": 0.533859372138977, "learning_rate": 6.458000000000001e-05, "loss": 0.0269, "step": 3230 }, { "grad_norm": 0.5595805048942566, "learning_rate": 6.478000000000001e-05, "loss": 0.0276, "step": 3240 }, { "grad_norm": 0.6246131062507629, "learning_rate": 6.498e-05, "loss": 0.0312, "step": 3250 }, { "grad_norm": 0.6171765923500061, "learning_rate": 6.518000000000001e-05, "loss": 0.0313, "step": 3260 }, { "grad_norm": 0.47520002722740173, "learning_rate": 6.538000000000001e-05, "loss": 0.0275, "step": 3270 }, { "grad_norm": 0.6312206983566284, "learning_rate": 6.558e-05, "loss": 0.0299, "step": 3280 }, { "grad_norm": 0.3988887369632721, "learning_rate": 6.578000000000001e-05, "loss": 0.0276, "step": 3290 }, { "grad_norm": 0.616264820098877, "learning_rate": 6.598e-05, "loss": 0.0277, "step": 3300 }, { "grad_norm": 0.7072576880455017, "learning_rate": 6.618e-05, "loss": 0.0321, "step": 3310 }, { "grad_norm": 0.5832338929176331, "learning_rate": 6.638e-05, "loss": 0.0329, "step": 3320 }, { "grad_norm": 0.4970053732395172, "learning_rate": 6.658e-05, "loss": 0.031, "step": 3330 }, { "grad_norm": 0.6191043257713318, "learning_rate": 6.678e-05, "loss": 0.0304, "step": 3340 }, { "grad_norm": 0.44521814584732056, "learning_rate": 6.698e-05, "loss": 0.0282, "step": 3350 }, { "grad_norm": 0.6199343204498291, "learning_rate": 6.718e-05, "loss": 0.0307, "step": 3360 }, { "grad_norm": 0.6452649235725403, "learning_rate": 6.738e-05, "loss": 0.0305, "step": 3370 }, { "grad_norm": 0.527595043182373, "learning_rate": 6.758e-05, "loss": 0.0325, "step": 3380 }, { "grad_norm": 0.5788626670837402, "learning_rate": 6.778e-05, "loss": 0.0312, "step": 3390 }, { "grad_norm": 0.5467917323112488, "learning_rate": 6.798e-05, "loss": 0.0287, "step": 3400 }, { "grad_norm": 0.4993205964565277, "learning_rate": 6.818e-05, "loss": 0.0291, "step": 3410 }, { "grad_norm": 0.44144418835639954, "learning_rate": 6.838e-05, "loss": 0.0273, "step": 3420 }, { "grad_norm": 0.5860689878463745, "learning_rate": 6.858e-05, "loss": 0.0272, "step": 3430 }, { "grad_norm": 0.45036548376083374, "learning_rate": 6.878e-05, "loss": 0.0277, "step": 3440 }, { "grad_norm": 0.5283169150352478, "learning_rate": 6.898e-05, "loss": 0.0273, "step": 3450 }, { "grad_norm": 0.4842246472835541, "learning_rate": 6.918e-05, "loss": 0.0307, "step": 3460 }, { "grad_norm": 0.392135888338089, "learning_rate": 6.938e-05, "loss": 0.0282, "step": 3470 }, { "grad_norm": 0.4261761009693146, "learning_rate": 6.958e-05, "loss": 0.0303, "step": 3480 }, { "grad_norm": 0.43339046835899353, "learning_rate": 6.978e-05, "loss": 0.0283, "step": 3490 }, { "grad_norm": 0.46705928444862366, "learning_rate": 6.998e-05, "loss": 0.0254, "step": 3500 }, { "grad_norm": 0.6546934843063354, "learning_rate": 7.018e-05, "loss": 0.0266, "step": 3510 }, { "grad_norm": 0.47514069080352783, "learning_rate": 7.038e-05, "loss": 0.025, "step": 3520 }, { "grad_norm": 0.4066433906555176, "learning_rate": 7.058e-05, "loss": 0.0255, "step": 3530 }, { "grad_norm": 0.6618328094482422, "learning_rate": 7.078e-05, "loss": 0.0286, "step": 3540 }, { "grad_norm": 0.5109913349151611, "learning_rate": 7.098e-05, "loss": 0.0277, "step": 3550 }, { "grad_norm": 0.4724695086479187, "learning_rate": 7.118e-05, "loss": 0.0272, "step": 3560 }, { "grad_norm": 0.6416662931442261, "learning_rate": 7.138e-05, "loss": 0.0288, "step": 3570 }, { "grad_norm": 0.3398820459842682, "learning_rate": 7.158e-05, "loss": 0.029, "step": 3580 }, { "grad_norm": 0.6120930910110474, "learning_rate": 7.178000000000001e-05, "loss": 0.0316, "step": 3590 }, { "grad_norm": 0.4433113932609558, "learning_rate": 7.198e-05, "loss": 0.0319, "step": 3600 }, { "grad_norm": 0.5791484117507935, "learning_rate": 7.218e-05, "loss": 0.027, "step": 3610 }, { "grad_norm": 0.44602543115615845, "learning_rate": 7.238000000000001e-05, "loss": 0.0305, "step": 3620 }, { "grad_norm": 0.31893324851989746, "learning_rate": 7.258e-05, "loss": 0.0249, "step": 3630 }, { "grad_norm": 0.41330501437187195, "learning_rate": 7.278e-05, "loss": 0.0277, "step": 3640 }, { "grad_norm": 0.5126101970672607, "learning_rate": 7.298000000000001e-05, "loss": 0.0274, "step": 3650 }, { "grad_norm": 0.5136942863464355, "learning_rate": 7.318e-05, "loss": 0.0292, "step": 3660 }, { "grad_norm": 0.46735113859176636, "learning_rate": 7.338e-05, "loss": 0.0267, "step": 3670 }, { "grad_norm": 0.5672528743743896, "learning_rate": 7.358000000000001e-05, "loss": 0.0316, "step": 3680 }, { "grad_norm": 0.4738384783267975, "learning_rate": 7.378e-05, "loss": 0.0301, "step": 3690 }, { "grad_norm": 0.5754106044769287, "learning_rate": 7.398e-05, "loss": 0.0278, "step": 3700 }, { "grad_norm": 0.5415052175521851, "learning_rate": 7.418000000000001e-05, "loss": 0.0262, "step": 3710 }, { "grad_norm": 0.5177817940711975, "learning_rate": 7.438e-05, "loss": 0.0255, "step": 3720 }, { "grad_norm": 0.5231409072875977, "learning_rate": 7.458000000000001e-05, "loss": 0.0262, "step": 3730 }, { "grad_norm": 0.5498318076133728, "learning_rate": 7.478e-05, "loss": 0.0252, "step": 3740 }, { "grad_norm": 0.63883376121521, "learning_rate": 7.498e-05, "loss": 0.0279, "step": 3750 }, { "grad_norm": 0.5600168108940125, "learning_rate": 7.518000000000001e-05, "loss": 0.0287, "step": 3760 }, { "grad_norm": 0.5008676052093506, "learning_rate": 7.538e-05, "loss": 0.0273, "step": 3770 }, { "grad_norm": 0.46899569034576416, "learning_rate": 7.558e-05, "loss": 0.0292, "step": 3780 }, { "grad_norm": 0.4936067759990692, "learning_rate": 7.578000000000001e-05, "loss": 0.0261, "step": 3790 }, { "grad_norm": 0.48622453212738037, "learning_rate": 7.598e-05, "loss": 0.0269, "step": 3800 }, { "grad_norm": 0.5658280849456787, "learning_rate": 7.618e-05, "loss": 0.0263, "step": 3810 }, { "grad_norm": 0.5828018188476562, "learning_rate": 7.638000000000001e-05, "loss": 0.0299, "step": 3820 }, { "grad_norm": 0.4313761293888092, "learning_rate": 7.658e-05, "loss": 0.0292, "step": 3830 }, { "grad_norm": 0.5813295245170593, "learning_rate": 7.678000000000001e-05, "loss": 0.0268, "step": 3840 }, { "grad_norm": 0.47719863057136536, "learning_rate": 7.698000000000001e-05, "loss": 0.0289, "step": 3850 }, { "grad_norm": 0.4087596833705902, "learning_rate": 7.718e-05, "loss": 0.0265, "step": 3860 }, { "grad_norm": 0.4917120933532715, "learning_rate": 7.738000000000001e-05, "loss": 0.0262, "step": 3870 }, { "grad_norm": 0.5257427096366882, "learning_rate": 7.758000000000001e-05, "loss": 0.0263, "step": 3880 }, { "grad_norm": 0.4098742604255676, "learning_rate": 7.778e-05, "loss": 0.0278, "step": 3890 }, { "grad_norm": 0.4655504524707794, "learning_rate": 7.798000000000001e-05, "loss": 0.0239, "step": 3900 }, { "grad_norm": 0.37436965107917786, "learning_rate": 7.818000000000001e-05, "loss": 0.0236, "step": 3910 }, { "grad_norm": 0.4398457407951355, "learning_rate": 7.838e-05, "loss": 0.0252, "step": 3920 }, { "grad_norm": 0.4554412066936493, "learning_rate": 7.858000000000001e-05, "loss": 0.0258, "step": 3930 }, { "grad_norm": 0.4227961003780365, "learning_rate": 7.878e-05, "loss": 0.0253, "step": 3940 }, { "grad_norm": 0.4390091001987457, "learning_rate": 7.897999999999999e-05, "loss": 0.0286, "step": 3950 }, { "grad_norm": 0.5401636958122253, "learning_rate": 7.918e-05, "loss": 0.0271, "step": 3960 }, { "grad_norm": 0.4496190547943115, "learning_rate": 7.938e-05, "loss": 0.0278, "step": 3970 }, { "grad_norm": 0.38724493980407715, "learning_rate": 7.958e-05, "loss": 0.0274, "step": 3980 }, { "grad_norm": 0.5318605899810791, "learning_rate": 7.978e-05, "loss": 0.0292, "step": 3990 }, { "grad_norm": 0.5463786721229553, "learning_rate": 7.998e-05, "loss": 0.0263, "step": 4000 }, { "grad_norm": 0.48062804341316223, "learning_rate": 8.018e-05, "loss": 0.0248, "step": 4010 }, { "grad_norm": 0.5151331424713135, "learning_rate": 8.038e-05, "loss": 0.0264, "step": 4020 }, { "grad_norm": 0.3627775013446808, "learning_rate": 8.058e-05, "loss": 0.026, "step": 4030 }, { "grad_norm": 0.4113871455192566, "learning_rate": 8.078e-05, "loss": 0.0256, "step": 4040 }, { "grad_norm": 0.449267715215683, "learning_rate": 8.098e-05, "loss": 0.0255, "step": 4050 }, { "grad_norm": 0.4297921359539032, "learning_rate": 8.118e-05, "loss": 0.0231, "step": 4060 }, { "grad_norm": 0.41647103428840637, "learning_rate": 8.138e-05, "loss": 0.0239, "step": 4070 }, { "grad_norm": 0.3848278522491455, "learning_rate": 8.158e-05, "loss": 0.0249, "step": 4080 }, { "grad_norm": 0.4057883322238922, "learning_rate": 8.178e-05, "loss": 0.0241, "step": 4090 }, { "grad_norm": 0.4423135221004486, "learning_rate": 8.198e-05, "loss": 0.0241, "step": 4100 }, { "grad_norm": 0.41646623611450195, "learning_rate": 8.218e-05, "loss": 0.0227, "step": 4110 }, { "grad_norm": 0.3494596779346466, "learning_rate": 8.238000000000001e-05, "loss": 0.0233, "step": 4120 }, { "grad_norm": 0.33007413148880005, "learning_rate": 8.258e-05, "loss": 0.0242, "step": 4130 }, { "grad_norm": 0.3187510073184967, "learning_rate": 8.278e-05, "loss": 0.0229, "step": 4140 }, { "grad_norm": 0.49864283204078674, "learning_rate": 8.298000000000001e-05, "loss": 0.0233, "step": 4150 }, { "grad_norm": 0.678604245185852, "learning_rate": 8.318e-05, "loss": 0.0243, "step": 4160 }, { "grad_norm": 0.5188041925430298, "learning_rate": 8.338e-05, "loss": 0.0234, "step": 4170 }, { "grad_norm": 0.47156059741973877, "learning_rate": 8.358e-05, "loss": 0.0251, "step": 4180 }, { "grad_norm": 0.3823423683643341, "learning_rate": 8.378e-05, "loss": 0.0217, "step": 4190 }, { "grad_norm": 0.359088271856308, "learning_rate": 8.398e-05, "loss": 0.0225, "step": 4200 }, { "grad_norm": 0.42152878642082214, "learning_rate": 8.418e-05, "loss": 0.0245, "step": 4210 }, { "grad_norm": 0.501042902469635, "learning_rate": 8.438e-05, "loss": 0.0234, "step": 4220 }, { "grad_norm": 0.37591612339019775, "learning_rate": 8.458e-05, "loss": 0.0277, "step": 4230 }, { "grad_norm": 0.4057016968727112, "learning_rate": 8.478e-05, "loss": 0.0237, "step": 4240 }, { "grad_norm": 0.3802444636821747, "learning_rate": 8.498e-05, "loss": 0.0252, "step": 4250 }, { "grad_norm": 0.3845711052417755, "learning_rate": 8.518000000000001e-05, "loss": 0.0223, "step": 4260 }, { "grad_norm": 0.3715333044528961, "learning_rate": 8.538e-05, "loss": 0.0237, "step": 4270 }, { "grad_norm": 0.46841779351234436, "learning_rate": 8.558e-05, "loss": 0.0236, "step": 4280 }, { "grad_norm": 0.48746412992477417, "learning_rate": 8.578000000000001e-05, "loss": 0.0259, "step": 4290 }, { "grad_norm": 0.42710819840431213, "learning_rate": 8.598e-05, "loss": 0.0228, "step": 4300 }, { "grad_norm": 0.36948928236961365, "learning_rate": 8.618e-05, "loss": 0.0231, "step": 4310 }, { "grad_norm": 0.42444416880607605, "learning_rate": 8.638000000000001e-05, "loss": 0.0253, "step": 4320 }, { "grad_norm": 0.31081005930900574, "learning_rate": 8.658e-05, "loss": 0.0236, "step": 4330 }, { "grad_norm": 0.4188391864299774, "learning_rate": 8.678e-05, "loss": 0.0243, "step": 4340 }, { "grad_norm": 0.38924458622932434, "learning_rate": 8.698000000000001e-05, "loss": 0.0213, "step": 4350 }, { "grad_norm": 0.4063223600387573, "learning_rate": 8.718e-05, "loss": 0.0243, "step": 4360 }, { "grad_norm": 0.4388837516307831, "learning_rate": 8.738000000000001e-05, "loss": 0.0228, "step": 4370 }, { "grad_norm": 0.3955563008785248, "learning_rate": 8.758000000000001e-05, "loss": 0.0217, "step": 4380 }, { "grad_norm": 0.4243549704551697, "learning_rate": 8.778e-05, "loss": 0.0235, "step": 4390 }, { "grad_norm": 0.3617676794528961, "learning_rate": 8.798000000000001e-05, "loss": 0.0217, "step": 4400 }, { "grad_norm": 0.4279113709926605, "learning_rate": 8.818000000000001e-05, "loss": 0.0214, "step": 4410 }, { "grad_norm": 0.3703252077102661, "learning_rate": 8.838e-05, "loss": 0.0231, "step": 4420 }, { "grad_norm": 0.4976182281970978, "learning_rate": 8.858000000000001e-05, "loss": 0.0233, "step": 4430 }, { "grad_norm": 0.44750329852104187, "learning_rate": 8.878000000000001e-05, "loss": 0.0254, "step": 4440 }, { "grad_norm": 0.5396740436553955, "learning_rate": 8.898e-05, "loss": 0.0272, "step": 4450 }, { "grad_norm": 0.4927869737148285, "learning_rate": 8.918000000000001e-05, "loss": 0.0249, "step": 4460 }, { "grad_norm": 0.5400258898735046, "learning_rate": 8.938e-05, "loss": 0.0258, "step": 4470 }, { "grad_norm": 0.38005930185317993, "learning_rate": 8.958e-05, "loss": 0.0239, "step": 4480 }, { "grad_norm": 0.35013386607170105, "learning_rate": 8.978000000000001e-05, "loss": 0.0246, "step": 4490 }, { "grad_norm": 0.4988744854927063, "learning_rate": 8.998e-05, "loss": 0.0257, "step": 4500 }, { "grad_norm": 0.32318535447120667, "learning_rate": 9.018000000000001e-05, "loss": 0.0257, "step": 4510 }, { "grad_norm": 0.3532182574272156, "learning_rate": 9.038000000000001e-05, "loss": 0.0235, "step": 4520 }, { "grad_norm": 0.44985345005989075, "learning_rate": 9.058e-05, "loss": 0.0223, "step": 4530 }, { "grad_norm": 0.3575800061225891, "learning_rate": 9.078000000000001e-05, "loss": 0.0221, "step": 4540 }, { "grad_norm": 0.28355199098587036, "learning_rate": 9.098000000000001e-05, "loss": 0.0233, "step": 4550 }, { "grad_norm": 0.295270711183548, "learning_rate": 9.118e-05, "loss": 0.0239, "step": 4560 }, { "grad_norm": 0.4252861440181732, "learning_rate": 9.138e-05, "loss": 0.0225, "step": 4570 }, { "grad_norm": 0.3083731532096863, "learning_rate": 9.158e-05, "loss": 0.0228, "step": 4580 }, { "grad_norm": 0.4088493287563324, "learning_rate": 9.178e-05, "loss": 0.0225, "step": 4590 }, { "grad_norm": 0.3559637665748596, "learning_rate": 9.198e-05, "loss": 0.0231, "step": 4600 }, { "grad_norm": 0.3971054255962372, "learning_rate": 9.218e-05, "loss": 0.0226, "step": 4610 }, { "grad_norm": 0.3477347195148468, "learning_rate": 9.238e-05, "loss": 0.0213, "step": 4620 }, { "grad_norm": 0.5293851494789124, "learning_rate": 9.258e-05, "loss": 0.0225, "step": 4630 }, { "grad_norm": 0.36530253291130066, "learning_rate": 9.278e-05, "loss": 0.0245, "step": 4640 }, { "grad_norm": 0.39325451850891113, "learning_rate": 9.298e-05, "loss": 0.0224, "step": 4650 }, { "grad_norm": 0.3328976631164551, "learning_rate": 9.318e-05, "loss": 0.0201, "step": 4660 }, { "grad_norm": 0.32924795150756836, "learning_rate": 9.338e-05, "loss": 0.0223, "step": 4670 }, { "grad_norm": 0.3875400722026825, "learning_rate": 9.358e-05, "loss": 0.0224, "step": 4680 }, { "grad_norm": 0.39109042286872864, "learning_rate": 9.378e-05, "loss": 0.0195, "step": 4690 }, { "grad_norm": 0.31067943572998047, "learning_rate": 9.398e-05, "loss": 0.0214, "step": 4700 }, { "grad_norm": 0.37359619140625, "learning_rate": 9.418e-05, "loss": 0.0202, "step": 4710 }, { "grad_norm": 0.29512789845466614, "learning_rate": 9.438e-05, "loss": 0.0203, "step": 4720 }, { "grad_norm": 0.40026965737342834, "learning_rate": 9.458e-05, "loss": 0.0211, "step": 4730 }, { "grad_norm": 0.35310015082359314, "learning_rate": 9.478e-05, "loss": 0.0214, "step": 4740 }, { "grad_norm": 0.26635780930519104, "learning_rate": 9.498e-05, "loss": 0.0184, "step": 4750 }, { "grad_norm": 0.41056162118911743, "learning_rate": 9.518000000000001e-05, "loss": 0.0216, "step": 4760 }, { "grad_norm": 0.3891940414905548, "learning_rate": 9.538e-05, "loss": 0.0203, "step": 4770 }, { "grad_norm": 0.3477391004562378, "learning_rate": 9.558e-05, "loss": 0.0203, "step": 4780 }, { "grad_norm": 0.3986249566078186, "learning_rate": 9.578000000000001e-05, "loss": 0.0219, "step": 4790 }, { "grad_norm": 0.2520107626914978, "learning_rate": 9.598e-05, "loss": 0.0213, "step": 4800 }, { "grad_norm": 0.40769675374031067, "learning_rate": 9.618e-05, "loss": 0.0208, "step": 4810 }, { "grad_norm": 0.37658175826072693, "learning_rate": 9.638000000000001e-05, "loss": 0.0208, "step": 4820 }, { "grad_norm": 0.3347134292125702, "learning_rate": 9.658e-05, "loss": 0.0194, "step": 4830 }, { "grad_norm": 0.3817685842514038, "learning_rate": 9.678e-05, "loss": 0.0187, "step": 4840 }, { "grad_norm": 0.2588033974170685, "learning_rate": 9.698000000000001e-05, "loss": 0.0185, "step": 4850 }, { "grad_norm": 0.45276474952697754, "learning_rate": 9.718e-05, "loss": 0.0205, "step": 4860 }, { "grad_norm": 0.22108136117458344, "learning_rate": 9.738e-05, "loss": 0.0206, "step": 4870 }, { "grad_norm": 0.30137041211128235, "learning_rate": 9.758000000000001e-05, "loss": 0.0205, "step": 4880 }, { "grad_norm": 0.34350067377090454, "learning_rate": 9.778e-05, "loss": 0.022, "step": 4890 }, { "grad_norm": 0.38766682147979736, "learning_rate": 9.798000000000001e-05, "loss": 0.0203, "step": 4900 }, { "grad_norm": 0.3333548307418823, "learning_rate": 9.818000000000001e-05, "loss": 0.0187, "step": 4910 }, { "grad_norm": 0.3721712529659271, "learning_rate": 9.838e-05, "loss": 0.0199, "step": 4920 }, { "grad_norm": 0.32414713501930237, "learning_rate": 9.858000000000001e-05, "loss": 0.0219, "step": 4930 }, { "grad_norm": 0.34805095195770264, "learning_rate": 9.878e-05, "loss": 0.0208, "step": 4940 }, { "grad_norm": 0.2576308250427246, "learning_rate": 9.898e-05, "loss": 0.021, "step": 4950 }, { "grad_norm": 0.32216283679008484, "learning_rate": 9.918000000000001e-05, "loss": 0.021, "step": 4960 }, { "grad_norm": 0.43034347891807556, "learning_rate": 9.938e-05, "loss": 0.0207, "step": 4970 }, { "grad_norm": 0.4210739731788635, "learning_rate": 9.958e-05, "loss": 0.0199, "step": 4980 }, { "grad_norm": 0.3456667959690094, "learning_rate": 9.978000000000001e-05, "loss": 0.0194, "step": 4990 }, { "grad_norm": 0.3130262792110443, "learning_rate": 9.998e-05, "loss": 0.0193, "step": 5000 }, { "grad_norm": 0.44592389464378357, "learning_rate": 9.999999778549045e-05, "loss": 0.0194, "step": 5010 }, { "grad_norm": 0.2718104422092438, "learning_rate": 9.999999013039593e-05, "loss": 0.0196, "step": 5020 }, { "grad_norm": 0.3090195655822754, "learning_rate": 9.999997700737766e-05, "loss": 0.0191, "step": 5030 }, { "grad_norm": 0.36793744564056396, "learning_rate": 9.999995841643709e-05, "loss": 0.0212, "step": 5040 }, { "grad_norm": 0.3405505418777466, "learning_rate": 9.999993435757623e-05, "loss": 0.022, "step": 5050 }, { "grad_norm": 0.29947730898857117, "learning_rate": 9.999990483079773e-05, "loss": 0.0191, "step": 5060 }, { "grad_norm": 0.25067004561424255, "learning_rate": 9.999986983610481e-05, "loss": 0.0173, "step": 5070 }, { "grad_norm": 0.29287782311439514, "learning_rate": 9.99998293735013e-05, "loss": 0.0166, "step": 5080 }, { "grad_norm": 0.2901678681373596, "learning_rate": 9.999978344299161e-05, "loss": 0.0209, "step": 5090 }, { "grad_norm": 0.28681573271751404, "learning_rate": 9.99997320445808e-05, "loss": 0.0184, "step": 5100 }, { "grad_norm": 0.31200337409973145, "learning_rate": 9.999967517827444e-05, "loss": 0.0206, "step": 5110 }, { "grad_norm": 0.3715360164642334, "learning_rate": 9.999961284407879e-05, "loss": 0.0186, "step": 5120 }, { "grad_norm": 0.2892459034919739, "learning_rate": 9.999954504200067e-05, "loss": 0.019, "step": 5130 }, { "grad_norm": 0.32280564308166504, "learning_rate": 9.999947177204744e-05, "loss": 0.0199, "step": 5140 }, { "grad_norm": 0.35019421577453613, "learning_rate": 9.999939303422718e-05, "loss": 0.0203, "step": 5150 }, { "grad_norm": 0.28581634163856506, "learning_rate": 9.999930882854847e-05, "loss": 0.0178, "step": 5160 }, { "grad_norm": 0.34094491600990295, "learning_rate": 9.999921915502051e-05, "loss": 0.0202, "step": 5170 }, { "grad_norm": 0.2573070824146271, "learning_rate": 9.99991240136531e-05, "loss": 0.0192, "step": 5180 }, { "grad_norm": 0.32589611411094666, "learning_rate": 9.999902340445668e-05, "loss": 0.0198, "step": 5190 }, { "grad_norm": 0.37570905685424805, "learning_rate": 9.999891732744224e-05, "loss": 0.0222, "step": 5200 }, { "grad_norm": 0.2900577485561371, "learning_rate": 9.999880578262135e-05, "loss": 0.02, "step": 5210 }, { "grad_norm": 0.2594282627105713, "learning_rate": 9.999868877000624e-05, "loss": 0.0188, "step": 5220 }, { "grad_norm": 0.3787885904312134, "learning_rate": 9.99985662896097e-05, "loss": 0.0186, "step": 5230 }, { "grad_norm": 0.38109833002090454, "learning_rate": 9.999843834144513e-05, "loss": 0.0196, "step": 5240 }, { "grad_norm": 0.3386252820491791, "learning_rate": 9.99983049255265e-05, "loss": 0.0217, "step": 5250 }, { "grad_norm": 0.33503878116607666, "learning_rate": 9.999816604186843e-05, "loss": 0.019, "step": 5260 }, { "grad_norm": 0.34309419989585876, "learning_rate": 9.999802169048609e-05, "loss": 0.0201, "step": 5270 }, { "grad_norm": 0.34431082010269165, "learning_rate": 9.999787187139527e-05, "loss": 0.019, "step": 5280 }, { "grad_norm": 0.2631964683532715, "learning_rate": 9.999771658461234e-05, "loss": 0.0186, "step": 5290 }, { "grad_norm": 0.2983121871948242, "learning_rate": 9.999755583015431e-05, "loss": 0.0175, "step": 5300 }, { "grad_norm": 0.29860273003578186, "learning_rate": 9.999738960803874e-05, "loss": 0.0179, "step": 5310 }, { "grad_norm": 0.42945218086242676, "learning_rate": 9.99972179182838e-05, "loss": 0.0204, "step": 5320 }, { "grad_norm": 0.3074246942996979, "learning_rate": 9.99970407609083e-05, "loss": 0.0183, "step": 5330 }, { "grad_norm": 0.3192512094974518, "learning_rate": 9.999685813593159e-05, "loss": 0.0204, "step": 5340 }, { "grad_norm": 0.2894594967365265, "learning_rate": 9.999667004337362e-05, "loss": 0.0188, "step": 5350 }, { "grad_norm": 0.4129878878593445, "learning_rate": 9.9996476483255e-05, "loss": 0.0191, "step": 5360 }, { "grad_norm": 0.26024365425109863, "learning_rate": 9.999627745559688e-05, "loss": 0.0196, "step": 5370 }, { "grad_norm": 0.23393502831459045, "learning_rate": 9.999607296042101e-05, "loss": 0.0182, "step": 5380 }, { "grad_norm": 0.3440232574939728, "learning_rate": 9.99958629977498e-05, "loss": 0.016, "step": 5390 }, { "grad_norm": 0.28885748982429504, "learning_rate": 9.999564756760615e-05, "loss": 0.0184, "step": 5400 }, { "grad_norm": 0.3361648917198181, "learning_rate": 9.999542667001366e-05, "loss": 0.0189, "step": 5410 }, { "grad_norm": 0.3709474503993988, "learning_rate": 9.999520030499647e-05, "loss": 0.0175, "step": 5420 }, { "grad_norm": 0.36413171887397766, "learning_rate": 9.999496847257936e-05, "loss": 0.0189, "step": 5430 }, { "grad_norm": 0.35868629813194275, "learning_rate": 9.999473117278764e-05, "loss": 0.0208, "step": 5440 }, { "grad_norm": 0.2873101830482483, "learning_rate": 9.999448840564731e-05, "loss": 0.0194, "step": 5450 }, { "grad_norm": 0.22944754362106323, "learning_rate": 9.999424017118488e-05, "loss": 0.0202, "step": 5460 }, { "grad_norm": 0.33538657426834106, "learning_rate": 9.999398646942751e-05, "loss": 0.019, "step": 5470 }, { "grad_norm": 0.25515881180763245, "learning_rate": 9.999372730040296e-05, "loss": 0.0181, "step": 5480 }, { "grad_norm": 0.3444586396217346, "learning_rate": 9.999346266413953e-05, "loss": 0.0191, "step": 5490 }, { "grad_norm": 0.30155321955680847, "learning_rate": 9.99931925606662e-05, "loss": 0.0205, "step": 5500 }, { "grad_norm": 0.35469773411750793, "learning_rate": 9.99929169900125e-05, "loss": 0.0182, "step": 5510 }, { "grad_norm": 0.31493881344795227, "learning_rate": 9.999263595220855e-05, "loss": 0.0184, "step": 5520 }, { "grad_norm": 0.22003278136253357, "learning_rate": 9.99923494472851e-05, "loss": 0.0199, "step": 5530 }, { "grad_norm": 0.3033617436885834, "learning_rate": 9.999205747527348e-05, "loss": 0.0182, "step": 5540 }, { "grad_norm": 0.2545269727706909, "learning_rate": 9.999176003620561e-05, "loss": 0.0176, "step": 5550 }, { "grad_norm": 0.4202108681201935, "learning_rate": 9.999145713011405e-05, "loss": 0.0194, "step": 5560 }, { "grad_norm": 0.2779492735862732, "learning_rate": 9.999114875703186e-05, "loss": 0.0197, "step": 5570 }, { "grad_norm": 0.3652492165565491, "learning_rate": 9.999083491699281e-05, "loss": 0.019, "step": 5580 }, { "grad_norm": 0.3247891664505005, "learning_rate": 9.999051561003123e-05, "loss": 0.018, "step": 5590 }, { "grad_norm": 0.31369680166244507, "learning_rate": 9.999019083618202e-05, "loss": 0.0196, "step": 5600 }, { "grad_norm": 0.2918696403503418, "learning_rate": 9.99898605954807e-05, "loss": 0.0181, "step": 5610 }, { "grad_norm": 0.25099867582321167, "learning_rate": 9.998952488796338e-05, "loss": 0.0199, "step": 5620 }, { "grad_norm": 0.30435910820961, "learning_rate": 9.998918371366676e-05, "loss": 0.0186, "step": 5630 }, { "grad_norm": 0.24554385244846344, "learning_rate": 9.99888370726282e-05, "loss": 0.0175, "step": 5640 }, { "grad_norm": 0.25303885340690613, "learning_rate": 9.998848496488556e-05, "loss": 0.0165, "step": 5650 }, { "grad_norm": 0.22972844541072845, "learning_rate": 9.998812739047736e-05, "loss": 0.0166, "step": 5660 }, { "grad_norm": 0.29519203305244446, "learning_rate": 9.99877643494427e-05, "loss": 0.0168, "step": 5670 }, { "grad_norm": 0.24605463445186615, "learning_rate": 9.998739584182128e-05, "loss": 0.0202, "step": 5680 }, { "grad_norm": 0.3585127592086792, "learning_rate": 9.998702186765342e-05, "loss": 0.0203, "step": 5690 }, { "grad_norm": 0.23532849550247192, "learning_rate": 9.998664242698e-05, "loss": 0.0188, "step": 5700 }, { "grad_norm": 0.24224437773227692, "learning_rate": 9.998625751984251e-05, "loss": 0.0162, "step": 5710 }, { "grad_norm": 0.27233168482780457, "learning_rate": 9.998586714628307e-05, "loss": 0.0164, "step": 5720 }, { "grad_norm": 0.24839676916599274, "learning_rate": 9.998547130634432e-05, "loss": 0.0177, "step": 5730 }, { "grad_norm": 0.29662957787513733, "learning_rate": 9.99850700000696e-05, "loss": 0.0163, "step": 5740 }, { "grad_norm": 0.3004867732524872, "learning_rate": 9.998466322750278e-05, "loss": 0.0163, "step": 5750 }, { "grad_norm": 0.26882851123809814, "learning_rate": 9.998425098868834e-05, "loss": 0.0181, "step": 5760 }, { "grad_norm": 0.36913079023361206, "learning_rate": 9.998383328367136e-05, "loss": 0.0202, "step": 5770 }, { "grad_norm": 0.36860182881355286, "learning_rate": 9.99834101124975e-05, "loss": 0.019, "step": 5780 }, { "grad_norm": 0.27862489223480225, "learning_rate": 9.998298147521309e-05, "loss": 0.0181, "step": 5790 }, { "grad_norm": 0.3180694878101349, "learning_rate": 9.998254737186496e-05, "loss": 0.019, "step": 5800 }, { "grad_norm": 0.29989391565322876, "learning_rate": 9.99821078025006e-05, "loss": 0.0214, "step": 5810 }, { "grad_norm": 0.32176122069358826, "learning_rate": 9.998166276716807e-05, "loss": 0.0216, "step": 5820 }, { "grad_norm": 0.3077927231788635, "learning_rate": 9.998121226591606e-05, "loss": 0.02, "step": 5830 }, { "grad_norm": 0.33672964572906494, "learning_rate": 9.998075629879382e-05, "loss": 0.0179, "step": 5840 }, { "grad_norm": 0.31416746973991394, "learning_rate": 9.99802948658512e-05, "loss": 0.0193, "step": 5850 }, { "grad_norm": 0.22325153648853302, "learning_rate": 9.99798279671387e-05, "loss": 0.0176, "step": 5860 }, { "grad_norm": 0.3424561619758606, "learning_rate": 9.997935560270734e-05, "loss": 0.018, "step": 5870 }, { "grad_norm": 0.2288476675748825, "learning_rate": 9.997887777260879e-05, "loss": 0.0169, "step": 5880 }, { "grad_norm": 0.27838703989982605, "learning_rate": 9.997839447689532e-05, "loss": 0.0153, "step": 5890 }, { "grad_norm": 0.2750995457172394, "learning_rate": 9.997790571561978e-05, "loss": 0.0183, "step": 5900 }, { "grad_norm": 0.3013586103916168, "learning_rate": 9.99774114888356e-05, "loss": 0.0189, "step": 5910 }, { "grad_norm": 0.26110920310020447, "learning_rate": 9.997691179659684e-05, "loss": 0.0182, "step": 5920 }, { "grad_norm": 0.2552572786808014, "learning_rate": 9.997640663895815e-05, "loss": 0.016, "step": 5930 }, { "grad_norm": 0.33938875794410706, "learning_rate": 9.997589601597477e-05, "loss": 0.0183, "step": 5940 }, { "grad_norm": 0.25478458404541016, "learning_rate": 9.997537992770252e-05, "loss": 0.0173, "step": 5950 }, { "grad_norm": 0.255168616771698, "learning_rate": 9.997485837419788e-05, "loss": 0.0171, "step": 5960 }, { "grad_norm": 0.3528013527393341, "learning_rate": 9.997433135551786e-05, "loss": 0.0208, "step": 5970 }, { "grad_norm": 0.3251807391643524, "learning_rate": 9.997379887172009e-05, "loss": 0.0192, "step": 5980 }, { "grad_norm": 0.34731966257095337, "learning_rate": 9.997326092286281e-05, "loss": 0.0174, "step": 5990 }, { "grad_norm": 0.28568610548973083, "learning_rate": 9.997271750900486e-05, "loss": 0.0183, "step": 6000 }, { "grad_norm": 0.3046918213367462, "learning_rate": 9.997216863020565e-05, "loss": 0.018, "step": 6010 }, { "grad_norm": 0.29399338364601135, "learning_rate": 9.99716142865252e-05, "loss": 0.016, "step": 6020 }, { "grad_norm": 0.2881018817424774, "learning_rate": 9.997105447802415e-05, "loss": 0.0175, "step": 6030 }, { "grad_norm": 0.3013134002685547, "learning_rate": 9.997048920476373e-05, "loss": 0.0168, "step": 6040 }, { "grad_norm": 0.32158219814300537, "learning_rate": 9.996991846680572e-05, "loss": 0.0177, "step": 6050 }, { "grad_norm": 0.2294856309890747, "learning_rate": 9.996934226421257e-05, "loss": 0.0172, "step": 6060 }, { "grad_norm": 0.21801072359085083, "learning_rate": 9.996876059704726e-05, "loss": 0.0156, "step": 6070 }, { "grad_norm": 0.2177879363298416, "learning_rate": 9.996817346537343e-05, "loss": 0.0165, "step": 6080 }, { "grad_norm": 0.2978406548500061, "learning_rate": 9.996758086925526e-05, "loss": 0.0172, "step": 6090 }, { "grad_norm": 0.296504944562912, "learning_rate": 9.996698280875759e-05, "loss": 0.0169, "step": 6100 }, { "grad_norm": 0.28709322214126587, "learning_rate": 9.99663792839458e-05, "loss": 0.017, "step": 6110 }, { "grad_norm": 0.24889683723449707, "learning_rate": 9.99657702948859e-05, "loss": 0.0174, "step": 6120 }, { "grad_norm": 0.2850784957408905, "learning_rate": 9.996515584164448e-05, "loss": 0.0154, "step": 6130 }, { "grad_norm": 0.35146382451057434, "learning_rate": 9.996453592428873e-05, "loss": 0.0174, "step": 6140 }, { "grad_norm": 0.3529721200466156, "learning_rate": 9.996391054288646e-05, "loss": 0.0163, "step": 6150 }, { "grad_norm": 0.29828453063964844, "learning_rate": 9.996327969750605e-05, "loss": 0.0164, "step": 6160 }, { "grad_norm": 0.2770377993583679, "learning_rate": 9.996264338821649e-05, "loss": 0.0163, "step": 6170 }, { "grad_norm": 0.277759850025177, "learning_rate": 9.996200161508735e-05, "loss": 0.0174, "step": 6180 }, { "grad_norm": 0.2051776945590973, "learning_rate": 9.996135437818885e-05, "loss": 0.0148, "step": 6190 }, { "grad_norm": 0.32469385862350464, "learning_rate": 9.996070167759175e-05, "loss": 0.0164, "step": 6200 }, { "grad_norm": 0.2890722453594208, "learning_rate": 9.996004351336743e-05, "loss": 0.0169, "step": 6210 }, { "grad_norm": 0.31242281198501587, "learning_rate": 9.995937988558785e-05, "loss": 0.0188, "step": 6220 }, { "grad_norm": 0.25168377161026, "learning_rate": 9.995871079432561e-05, "loss": 0.0164, "step": 6230 }, { "grad_norm": 0.2607565224170685, "learning_rate": 9.995803623965389e-05, "loss": 0.018, "step": 6240 }, { "grad_norm": 0.28490886092185974, "learning_rate": 9.995735622164641e-05, "loss": 0.0176, "step": 6250 }, { "grad_norm": 0.2520447075366974, "learning_rate": 9.995667074037758e-05, "loss": 0.0157, "step": 6260 }, { "grad_norm": 0.27042677998542786, "learning_rate": 9.995597979592232e-05, "loss": 0.0154, "step": 6270 }, { "grad_norm": 0.24708038568496704, "learning_rate": 9.995528338835625e-05, "loss": 0.0176, "step": 6280 }, { "grad_norm": 0.2302970290184021, "learning_rate": 9.995458151775547e-05, "loss": 0.0176, "step": 6290 }, { "grad_norm": 0.24671520292758942, "learning_rate": 9.995387418419677e-05, "loss": 0.017, "step": 6300 }, { "grad_norm": 0.20564372837543488, "learning_rate": 9.99531613877575e-05, "loss": 0.0158, "step": 6310 }, { "grad_norm": 0.17335239052772522, "learning_rate": 9.995244312851559e-05, "loss": 0.0171, "step": 6320 }, { "grad_norm": 0.28108879923820496, "learning_rate": 9.995171940654961e-05, "loss": 0.0152, "step": 6330 }, { "grad_norm": 0.23924003541469574, "learning_rate": 9.995099022193871e-05, "loss": 0.0158, "step": 6340 }, { "grad_norm": 0.20604537427425385, "learning_rate": 9.995025557476261e-05, "loss": 0.017, "step": 6350 }, { "grad_norm": 0.2729603946208954, "learning_rate": 9.994951546510165e-05, "loss": 0.0175, "step": 6360 }, { "grad_norm": 0.2838898003101349, "learning_rate": 9.994876989303679e-05, "loss": 0.0167, "step": 6370 }, { "grad_norm": 0.25916799902915955, "learning_rate": 9.994801885864955e-05, "loss": 0.0169, "step": 6380 }, { "grad_norm": 0.24933689832687378, "learning_rate": 9.994726236202205e-05, "loss": 0.0172, "step": 6390 }, { "grad_norm": 0.28652796149253845, "learning_rate": 9.994650040323704e-05, "loss": 0.0178, "step": 6400 }, { "grad_norm": 0.16983379423618317, "learning_rate": 9.994573298237784e-05, "loss": 0.0157, "step": 6410 }, { "grad_norm": 0.20708167552947998, "learning_rate": 9.994496009952837e-05, "loss": 0.0151, "step": 6420 }, { "grad_norm": 0.20627567172050476, "learning_rate": 9.994418175477316e-05, "loss": 0.018, "step": 6430 }, { "grad_norm": 0.29933494329452515, "learning_rate": 9.994339794819733e-05, "loss": 0.0162, "step": 6440 }, { "grad_norm": 0.3059535026550293, "learning_rate": 9.994260867988658e-05, "loss": 0.0156, "step": 6450 }, { "grad_norm": 0.19936706125736237, "learning_rate": 9.994181394992723e-05, "loss": 0.0154, "step": 6460 }, { "grad_norm": 0.2768833339214325, "learning_rate": 9.994101375840618e-05, "loss": 0.016, "step": 6470 }, { "grad_norm": 0.2515503168106079, "learning_rate": 9.994020810541098e-05, "loss": 0.015, "step": 6480 }, { "grad_norm": 0.28634199500083923, "learning_rate": 9.99393969910297e-05, "loss": 0.0142, "step": 6490 }, { "grad_norm": 0.297982394695282, "learning_rate": 9.993858041535104e-05, "loss": 0.0164, "step": 6500 }, { "grad_norm": 0.30462780594825745, "learning_rate": 9.99377583784643e-05, "loss": 0.0152, "step": 6510 }, { "grad_norm": 0.2751997113227844, "learning_rate": 9.993693088045939e-05, "loss": 0.0151, "step": 6520 }, { "grad_norm": 0.2354867309331894, "learning_rate": 9.99360979214268e-05, "loss": 0.0157, "step": 6530 }, { "grad_norm": 0.3263036608695984, "learning_rate": 9.99352595014576e-05, "loss": 0.0151, "step": 6540 }, { "grad_norm": 0.3929020166397095, "learning_rate": 9.993441562064354e-05, "loss": 0.0163, "step": 6550 }, { "grad_norm": 0.18572616577148438, "learning_rate": 9.993356627907685e-05, "loss": 0.0159, "step": 6560 }, { "grad_norm": 0.1714172065258026, "learning_rate": 9.99327114768504e-05, "loss": 0.014, "step": 6570 }, { "grad_norm": 0.19002367556095123, "learning_rate": 9.99318512140577e-05, "loss": 0.014, "step": 6580 }, { "grad_norm": 0.31241101026535034, "learning_rate": 9.993098549079284e-05, "loss": 0.0147, "step": 6590 }, { "grad_norm": 0.2060771882534027, "learning_rate": 9.993011430715047e-05, "loss": 0.0144, "step": 6600 }, { "grad_norm": 0.2639161944389343, "learning_rate": 9.992923766322586e-05, "loss": 0.0147, "step": 6610 }, { "grad_norm": 0.2585535943508148, "learning_rate": 9.99283555591149e-05, "loss": 0.0147, "step": 6620 }, { "grad_norm": 0.3269290030002594, "learning_rate": 9.992746799491404e-05, "loss": 0.0152, "step": 6630 }, { "grad_norm": 0.25374165177345276, "learning_rate": 9.992657497072033e-05, "loss": 0.0148, "step": 6640 }, { "grad_norm": 0.23564773797988892, "learning_rate": 9.992567648663147e-05, "loss": 0.0162, "step": 6650 }, { "grad_norm": 0.2884715795516968, "learning_rate": 9.992477254274568e-05, "loss": 0.0161, "step": 6660 }, { "grad_norm": 0.2693805396556854, "learning_rate": 9.992386313916183e-05, "loss": 0.0163, "step": 6670 }, { "grad_norm": 0.2807772755622864, "learning_rate": 9.992294827597934e-05, "loss": 0.016, "step": 6680 }, { "grad_norm": 0.28260883688926697, "learning_rate": 9.992202795329831e-05, "loss": 0.0153, "step": 6690 }, { "grad_norm": 0.2773180902004242, "learning_rate": 9.992110217121936e-05, "loss": 0.0152, "step": 6700 }, { "grad_norm": 0.3276716470718384, "learning_rate": 9.992017092984372e-05, "loss": 0.0156, "step": 6710 }, { "grad_norm": 0.28589072823524475, "learning_rate": 9.991923422927326e-05, "loss": 0.0164, "step": 6720 }, { "grad_norm": 0.23119938373565674, "learning_rate": 9.991829206961037e-05, "loss": 0.0152, "step": 6730 }, { "grad_norm": 0.2825298011302948, "learning_rate": 9.991734445095813e-05, "loss": 0.016, "step": 6740 }, { "grad_norm": 0.2252090573310852, "learning_rate": 9.991639137342015e-05, "loss": 0.0156, "step": 6750 }, { "grad_norm": 0.19855956733226776, "learning_rate": 9.991543283710064e-05, "loss": 0.0139, "step": 6760 }, { "grad_norm": 0.21837608516216278, "learning_rate": 9.991446884210445e-05, "loss": 0.0135, "step": 6770 }, { "grad_norm": 0.23084449768066406, "learning_rate": 9.9913499388537e-05, "loss": 0.0149, "step": 6780 }, { "grad_norm": 0.29342517256736755, "learning_rate": 9.99125244765043e-05, "loss": 0.0147, "step": 6790 }, { "grad_norm": 0.24652434885501862, "learning_rate": 9.991154410611296e-05, "loss": 0.0154, "step": 6800 }, { "grad_norm": 0.27148300409317017, "learning_rate": 9.99105582774702e-05, "loss": 0.0163, "step": 6810 }, { "grad_norm": 0.2716660499572754, "learning_rate": 9.990956699068384e-05, "loss": 0.0171, "step": 6820 }, { "grad_norm": 0.23919540643692017, "learning_rate": 9.990857024586224e-05, "loss": 0.0166, "step": 6830 }, { "grad_norm": 0.23781543970108032, "learning_rate": 9.990756804311446e-05, "loss": 0.0156, "step": 6840 }, { "grad_norm": 0.240763321518898, "learning_rate": 9.990656038255006e-05, "loss": 0.0157, "step": 6850 }, { "grad_norm": 0.24816405773162842, "learning_rate": 9.990554726427926e-05, "loss": 0.0148, "step": 6860 }, { "grad_norm": 0.2556403875350952, "learning_rate": 9.990452868841284e-05, "loss": 0.0168, "step": 6870 }, { "grad_norm": 0.3131426274776459, "learning_rate": 9.99035046550622e-05, "loss": 0.0187, "step": 6880 }, { "grad_norm": 0.21381866931915283, "learning_rate": 9.99024751643393e-05, "loss": 0.0138, "step": 6890 }, { "grad_norm": 0.2517854869365692, "learning_rate": 9.990144021635677e-05, "loss": 0.0147, "step": 6900 }, { "grad_norm": 0.3066459894180298, "learning_rate": 9.990039981122775e-05, "loss": 0.0143, "step": 6910 }, { "grad_norm": 0.22187446057796478, "learning_rate": 9.989935394906602e-05, "loss": 0.0172, "step": 6920 }, { "grad_norm": 0.20409820973873138, "learning_rate": 9.989830262998598e-05, "loss": 0.0148, "step": 6930 }, { "grad_norm": 0.20818112790584564, "learning_rate": 9.989724585410259e-05, "loss": 0.0165, "step": 6940 }, { "grad_norm": 0.23473814129829407, "learning_rate": 9.989618362153139e-05, "loss": 0.014, "step": 6950 }, { "grad_norm": 0.2852117419242859, "learning_rate": 9.989511593238859e-05, "loss": 0.0154, "step": 6960 }, { "grad_norm": 0.16650350391864777, "learning_rate": 9.98940427867909e-05, "loss": 0.0133, "step": 6970 }, { "grad_norm": 0.2893199324607849, "learning_rate": 9.989296418485573e-05, "loss": 0.015, "step": 6980 }, { "grad_norm": 0.2345835566520691, "learning_rate": 9.989188012670101e-05, "loss": 0.0152, "step": 6990 }, { "grad_norm": 0.22960469126701355, "learning_rate": 9.989079061244528e-05, "loss": 0.0134, "step": 7000 }, { "grad_norm": 0.2695861756801605, "learning_rate": 9.988969564220769e-05, "loss": 0.0154, "step": 7010 }, { "grad_norm": 0.21393169462680817, "learning_rate": 9.988859521610801e-05, "loss": 0.0136, "step": 7020 }, { "grad_norm": 0.23561535775661469, "learning_rate": 9.988748933426656e-05, "loss": 0.0145, "step": 7030 }, { "grad_norm": 0.2814995348453522, "learning_rate": 9.988637799680428e-05, "loss": 0.0127, "step": 7040 }, { "grad_norm": 0.22369472682476044, "learning_rate": 9.98852612038427e-05, "loss": 0.0146, "step": 7050 }, { "grad_norm": 0.20491239428520203, "learning_rate": 9.988413895550397e-05, "loss": 0.016, "step": 7060 }, { "grad_norm": 0.23080286383628845, "learning_rate": 9.98830112519108e-05, "loss": 0.0144, "step": 7070 }, { "grad_norm": 0.18803003430366516, "learning_rate": 9.98818780931865e-05, "loss": 0.0141, "step": 7080 }, { "grad_norm": 0.249717578291893, "learning_rate": 9.988073947945502e-05, "loss": 0.0145, "step": 7090 }, { "grad_norm": 0.2488114833831787, "learning_rate": 9.987959541084087e-05, "loss": 0.0156, "step": 7100 }, { "grad_norm": 0.2991560697555542, "learning_rate": 9.987844588746915e-05, "loss": 0.016, "step": 7110 }, { "grad_norm": 0.3091375529766083, "learning_rate": 9.987729090946558e-05, "loss": 0.0176, "step": 7120 }, { "grad_norm": 0.3085035979747772, "learning_rate": 9.987613047695647e-05, "loss": 0.0162, "step": 7130 }, { "grad_norm": 0.27475109696388245, "learning_rate": 9.987496459006871e-05, "loss": 0.0159, "step": 7140 }, { "grad_norm": 0.27384576201438904, "learning_rate": 9.987379324892982e-05, "loss": 0.0146, "step": 7150 }, { "grad_norm": 0.2334195077419281, "learning_rate": 9.987261645366788e-05, "loss": 0.0152, "step": 7160 }, { "grad_norm": 0.13924211263656616, "learning_rate": 9.987143420441158e-05, "loss": 0.0138, "step": 7170 }, { "grad_norm": 0.21688255667686462, "learning_rate": 9.987024650129022e-05, "loss": 0.0127, "step": 7180 }, { "grad_norm": 0.22532908618450165, "learning_rate": 9.986905334443368e-05, "loss": 0.013, "step": 7190 }, { "grad_norm": 0.19054026901721954, "learning_rate": 9.986785473397245e-05, "loss": 0.0144, "step": 7200 }, { "grad_norm": 0.22129984200000763, "learning_rate": 9.98666506700376e-05, "loss": 0.0152, "step": 7210 }, { "grad_norm": 0.30045753717422485, "learning_rate": 9.986544115276081e-05, "loss": 0.017, "step": 7220 }, { "grad_norm": 0.22412405908107758, "learning_rate": 9.986422618227433e-05, "loss": 0.0172, "step": 7230 }, { "grad_norm": 0.28978806734085083, "learning_rate": 9.986300575871106e-05, "loss": 0.014, "step": 7240 }, { "grad_norm": 0.21492545306682587, "learning_rate": 9.986177988220444e-05, "loss": 0.0168, "step": 7250 }, { "grad_norm": 0.3402971625328064, "learning_rate": 9.986054855288856e-05, "loss": 0.016, "step": 7260 }, { "grad_norm": 0.31167715787887573, "learning_rate": 9.985931177089802e-05, "loss": 0.0154, "step": 7270 }, { "grad_norm": 0.2402944713830948, "learning_rate": 9.985806953636814e-05, "loss": 0.018, "step": 7280 }, { "grad_norm": 0.23364213109016418, "learning_rate": 9.985682184943471e-05, "loss": 0.0149, "step": 7290 }, { "grad_norm": 0.2359260469675064, "learning_rate": 9.98555687102342e-05, "loss": 0.0158, "step": 7300 }, { "grad_norm": 0.1909271776676178, "learning_rate": 9.985431011890367e-05, "loss": 0.0161, "step": 7310 }, { "grad_norm": 0.2972935140132904, "learning_rate": 9.985304607558075e-05, "loss": 0.0163, "step": 7320 }, { "grad_norm": 0.2266003042459488, "learning_rate": 9.985177658040364e-05, "loss": 0.0149, "step": 7330 }, { "grad_norm": 0.21425704658031464, "learning_rate": 9.985050163351119e-05, "loss": 0.0162, "step": 7340 }, { "grad_norm": 0.26537027955055237, "learning_rate": 9.984922123504286e-05, "loss": 0.0151, "step": 7350 }, { "grad_norm": 0.19015024602413177, "learning_rate": 9.984793538513862e-05, "loss": 0.0135, "step": 7360 }, { "grad_norm": 0.18216770887374878, "learning_rate": 9.984664408393912e-05, "loss": 0.0139, "step": 7370 }, { "grad_norm": 0.19672174751758575, "learning_rate": 9.984534733158556e-05, "loss": 0.0145, "step": 7380 }, { "grad_norm": 0.20898868143558502, "learning_rate": 9.984404512821977e-05, "loss": 0.0137, "step": 7390 }, { "grad_norm": 0.3097553253173828, "learning_rate": 9.984273747398411e-05, "loss": 0.015, "step": 7400 }, { "grad_norm": 0.18451429903507233, "learning_rate": 9.984142436902165e-05, "loss": 0.014, "step": 7410 }, { "grad_norm": 0.20901337265968323, "learning_rate": 9.984010581347596e-05, "loss": 0.0147, "step": 7420 }, { "grad_norm": 0.23984701931476593, "learning_rate": 9.983878180749121e-05, "loss": 0.0134, "step": 7430 }, { "grad_norm": 0.18840496242046356, "learning_rate": 9.983745235121222e-05, "loss": 0.0144, "step": 7440 }, { "grad_norm": 0.28952091932296753, "learning_rate": 9.983611744478438e-05, "loss": 0.0152, "step": 7450 }, { "grad_norm": 0.2176923006772995, "learning_rate": 9.983477708835365e-05, "loss": 0.0153, "step": 7460 }, { "grad_norm": 0.2464076727628708, "learning_rate": 9.983343128206664e-05, "loss": 0.0159, "step": 7470 }, { "grad_norm": 0.18550823628902435, "learning_rate": 9.983208002607049e-05, "loss": 0.0133, "step": 7480 }, { "grad_norm": 0.21315330266952515, "learning_rate": 9.9830723320513e-05, "loss": 0.0144, "step": 7490 }, { "grad_norm": 0.23321138322353363, "learning_rate": 9.982936116554254e-05, "loss": 0.0135, "step": 7500 }, { "grad_norm": 0.23194316029548645, "learning_rate": 9.982799356130803e-05, "loss": 0.0156, "step": 7510 }, { "grad_norm": 0.256669819355011, "learning_rate": 9.982662050795908e-05, "loss": 0.0126, "step": 7520 }, { "grad_norm": 0.17772796750068665, "learning_rate": 9.982524200564583e-05, "loss": 0.0142, "step": 7530 }, { "grad_norm": 0.20441029965877533, "learning_rate": 9.982385805451901e-05, "loss": 0.016, "step": 7540 }, { "grad_norm": 0.2147505283355713, "learning_rate": 9.982246865472998e-05, "loss": 0.0157, "step": 7550 }, { "grad_norm": 0.2367686778306961, "learning_rate": 9.982107380643069e-05, "loss": 0.0144, "step": 7560 }, { "grad_norm": 0.23655197024345398, "learning_rate": 9.981967350977368e-05, "loss": 0.0139, "step": 7570 }, { "grad_norm": 0.23738420009613037, "learning_rate": 9.981826776491208e-05, "loss": 0.0134, "step": 7580 }, { "grad_norm": 0.2075113207101822, "learning_rate": 9.98168565719996e-05, "loss": 0.0132, "step": 7590 }, { "grad_norm": 0.1797456592321396, "learning_rate": 9.98154399311906e-05, "loss": 0.0137, "step": 7600 }, { "grad_norm": 0.2341909259557724, "learning_rate": 9.981401784263997e-05, "loss": 0.013, "step": 7610 }, { "grad_norm": 0.2643430531024933, "learning_rate": 9.981259030650326e-05, "loss": 0.0142, "step": 7620 }, { "grad_norm": 0.1950213760137558, "learning_rate": 9.981115732293655e-05, "loss": 0.0132, "step": 7630 }, { "grad_norm": 0.15880782902240753, "learning_rate": 9.980971889209659e-05, "loss": 0.0147, "step": 7640 }, { "grad_norm": 0.185911163687706, "learning_rate": 9.980827501414064e-05, "loss": 0.0143, "step": 7650 }, { "grad_norm": 0.1371200829744339, "learning_rate": 9.980682568922663e-05, "loss": 0.0108, "step": 7660 }, { "grad_norm": 0.26803019642829895, "learning_rate": 9.980537091751304e-05, "loss": 0.0136, "step": 7670 }, { "grad_norm": 0.18995481729507446, "learning_rate": 9.980391069915897e-05, "loss": 0.0138, "step": 7680 }, { "grad_norm": 0.18292078375816345, "learning_rate": 9.98024450343241e-05, "loss": 0.0143, "step": 7690 }, { "grad_norm": 0.19886238873004913, "learning_rate": 9.980097392316872e-05, "loss": 0.0151, "step": 7700 }, { "grad_norm": 0.29014188051223755, "learning_rate": 9.97994973658537e-05, "loss": 0.0136, "step": 7710 }, { "grad_norm": 0.2310144603252411, "learning_rate": 9.979801536254054e-05, "loss": 0.0138, "step": 7720 }, { "grad_norm": 0.21991699934005737, "learning_rate": 9.979652791339127e-05, "loss": 0.0129, "step": 7730 }, { "grad_norm": 0.20347058773040771, "learning_rate": 9.97950350185686e-05, "loss": 0.0129, "step": 7740 }, { "grad_norm": 0.19902531802654266, "learning_rate": 9.979353667823574e-05, "loss": 0.0145, "step": 7750 }, { "grad_norm": 0.25111398100852966, "learning_rate": 9.979203289255658e-05, "loss": 0.014, "step": 7760 }, { "grad_norm": 0.19249920547008514, "learning_rate": 9.979052366169557e-05, "loss": 0.0124, "step": 7770 }, { "grad_norm": 0.2102239727973938, "learning_rate": 9.978900898581775e-05, "loss": 0.0143, "step": 7780 }, { "grad_norm": 0.24861085414886475, "learning_rate": 9.978748886508875e-05, "loss": 0.0141, "step": 7790 }, { "grad_norm": 0.20044465363025665, "learning_rate": 9.978596329967484e-05, "loss": 0.0134, "step": 7800 }, { "grad_norm": 0.2259143888950348, "learning_rate": 9.978443228974284e-05, "loss": 0.0136, "step": 7810 }, { "grad_norm": 0.16257594525814056, "learning_rate": 9.978289583546015e-05, "loss": 0.0119, "step": 7820 }, { "grad_norm": 0.2844615876674652, "learning_rate": 9.978135393699484e-05, "loss": 0.0155, "step": 7830 }, { "grad_norm": 0.19550983607769012, "learning_rate": 9.977980659451548e-05, "loss": 0.0143, "step": 7840 }, { "grad_norm": 0.2726915776729584, "learning_rate": 9.977825380819135e-05, "loss": 0.0142, "step": 7850 }, { "grad_norm": 0.14635908603668213, "learning_rate": 9.97766955781922e-05, "loss": 0.0137, "step": 7860 }, { "grad_norm": 0.22139497101306915, "learning_rate": 9.977513190468848e-05, "loss": 0.0139, "step": 7870 }, { "grad_norm": 0.16874003410339355, "learning_rate": 9.977356278785116e-05, "loss": 0.0142, "step": 7880 }, { "grad_norm": 0.2500464618206024, "learning_rate": 9.977198822785184e-05, "loss": 0.0132, "step": 7890 }, { "grad_norm": 0.24243265390396118, "learning_rate": 9.977040822486273e-05, "loss": 0.0143, "step": 7900 }, { "grad_norm": 0.21784111857414246, "learning_rate": 9.97688227790566e-05, "loss": 0.0123, "step": 7910 }, { "grad_norm": 0.19784435629844666, "learning_rate": 9.976723189060684e-05, "loss": 0.0129, "step": 7920 }, { "grad_norm": 0.21540510654449463, "learning_rate": 9.976563555968742e-05, "loss": 0.0153, "step": 7930 }, { "grad_norm": 0.26275545358657837, "learning_rate": 9.976403378647292e-05, "loss": 0.0161, "step": 7940 }, { "grad_norm": 0.16742709279060364, "learning_rate": 9.97624265711385e-05, "loss": 0.0159, "step": 7950 }, { "grad_norm": 0.2842450737953186, "learning_rate": 9.976081391385993e-05, "loss": 0.0139, "step": 7960 }, { "grad_norm": 0.2257169932126999, "learning_rate": 9.975919581481356e-05, "loss": 0.0172, "step": 7970 }, { "grad_norm": 0.2789003849029541, "learning_rate": 9.975757227417634e-05, "loss": 0.0158, "step": 7980 }, { "grad_norm": 0.2240723967552185, "learning_rate": 9.975594329212586e-05, "loss": 0.0162, "step": 7990 }, { "grad_norm": 0.2101982980966568, "learning_rate": 9.97543088688402e-05, "loss": 0.0151, "step": 8000 }, { "grad_norm": 0.3019064664840698, "learning_rate": 9.975266900449814e-05, "loss": 0.0147, "step": 8010 }, { "grad_norm": 0.2988121211528778, "learning_rate": 9.975102369927898e-05, "loss": 0.0158, "step": 8020 }, { "grad_norm": 0.240493506193161, "learning_rate": 9.974937295336269e-05, "loss": 0.017, "step": 8030 }, { "grad_norm": 0.26886093616485596, "learning_rate": 9.974771676692975e-05, "loss": 0.0164, "step": 8040 }, { "grad_norm": 0.2602553963661194, "learning_rate": 9.974605514016131e-05, "loss": 0.0147, "step": 8050 }, { "grad_norm": 0.21591413021087646, "learning_rate": 9.974438807323907e-05, "loss": 0.0155, "step": 8060 }, { "grad_norm": 0.2925816476345062, "learning_rate": 9.974271556634535e-05, "loss": 0.0145, "step": 8070 }, { "grad_norm": 0.2540501058101654, "learning_rate": 9.974103761966302e-05, "loss": 0.0145, "step": 8080 }, { "grad_norm": 0.26637202501296997, "learning_rate": 9.973935423337563e-05, "loss": 0.0146, "step": 8090 }, { "grad_norm": 0.18747849762439728, "learning_rate": 9.973766540766722e-05, "loss": 0.0154, "step": 8100 }, { "grad_norm": 0.20696088671684265, "learning_rate": 9.97359711427225e-05, "loss": 0.0142, "step": 8110 }, { "grad_norm": 0.18517832458019257, "learning_rate": 9.973427143872677e-05, "loss": 0.0117, "step": 8120 }, { "grad_norm": 0.201190784573555, "learning_rate": 9.973256629586589e-05, "loss": 0.0142, "step": 8130 }, { "grad_norm": 0.2718018591403961, "learning_rate": 9.973085571432632e-05, "loss": 0.015, "step": 8140 }, { "grad_norm": 0.22241634130477905, "learning_rate": 9.972913969429513e-05, "loss": 0.016, "step": 8150 }, { "grad_norm": 0.23702839016914368, "learning_rate": 9.972741823596e-05, "loss": 0.0128, "step": 8160 }, { "grad_norm": 0.30696234107017517, "learning_rate": 9.972569133950917e-05, "loss": 0.0143, "step": 8170 }, { "grad_norm": 0.18688848614692688, "learning_rate": 9.972395900513151e-05, "loss": 0.0128, "step": 8180 }, { "grad_norm": 0.21005560457706451, "learning_rate": 9.972222123301645e-05, "loss": 0.0124, "step": 8190 }, { "grad_norm": 0.2284628003835678, "learning_rate": 9.972047802335403e-05, "loss": 0.0153, "step": 8200 }, { "grad_norm": 0.19476868212223053, "learning_rate": 9.971872937633488e-05, "loss": 0.0124, "step": 8210 }, { "grad_norm": 0.28272414207458496, "learning_rate": 9.971697529215024e-05, "loss": 0.013, "step": 8220 }, { "grad_norm": 0.2391953319311142, "learning_rate": 9.971521577099192e-05, "loss": 0.0131, "step": 8230 }, { "grad_norm": 0.2026398479938507, "learning_rate": 9.971345081305236e-05, "loss": 0.013, "step": 8240 }, { "grad_norm": 0.17242887616157532, "learning_rate": 9.971168041852456e-05, "loss": 0.0122, "step": 8250 }, { "grad_norm": 0.19436798989772797, "learning_rate": 9.970990458760215e-05, "loss": 0.0124, "step": 8260 }, { "grad_norm": 0.2195732146501541, "learning_rate": 9.970812332047929e-05, "loss": 0.0133, "step": 8270 }, { "grad_norm": 0.2429802119731903, "learning_rate": 9.97063366173508e-05, "loss": 0.0134, "step": 8280 }, { "grad_norm": 0.15038682520389557, "learning_rate": 9.970454447841207e-05, "loss": 0.0137, "step": 8290 }, { "grad_norm": 0.1668236255645752, "learning_rate": 9.970274690385909e-05, "loss": 0.0145, "step": 8300 }, { "grad_norm": 0.21006298065185547, "learning_rate": 9.970094389388844e-05, "loss": 0.0124, "step": 8310 }, { "grad_norm": 0.182036891579628, "learning_rate": 9.969913544869728e-05, "loss": 0.0139, "step": 8320 }, { "grad_norm": 0.2621835172176361, "learning_rate": 9.96973215684834e-05, "loss": 0.0129, "step": 8330 }, { "grad_norm": 0.22434605658054352, "learning_rate": 9.969550225344513e-05, "loss": 0.0131, "step": 8340 }, { "grad_norm": 0.1835152953863144, "learning_rate": 9.969367750378147e-05, "loss": 0.0144, "step": 8350 }, { "grad_norm": 0.1979805827140808, "learning_rate": 9.969184731969194e-05, "loss": 0.0123, "step": 8360 }, { "grad_norm": 0.15206773579120636, "learning_rate": 9.96900117013767e-05, "loss": 0.0116, "step": 8370 }, { "grad_norm": 0.19303056597709656, "learning_rate": 9.96881706490365e-05, "loss": 0.0133, "step": 8380 }, { "grad_norm": 0.2330414354801178, "learning_rate": 9.968632416287265e-05, "loss": 0.0122, "step": 8390 }, { "grad_norm": 0.21931244432926178, "learning_rate": 9.96844722430871e-05, "loss": 0.0127, "step": 8400 }, { "grad_norm": 0.2008669227361679, "learning_rate": 9.968261488988235e-05, "loss": 0.0132, "step": 8410 }, { "grad_norm": 0.19882991909980774, "learning_rate": 9.968075210346155e-05, "loss": 0.0128, "step": 8420 }, { "grad_norm": 0.297805517911911, "learning_rate": 9.967888388402839e-05, "loss": 0.0123, "step": 8430 }, { "grad_norm": 0.15039819478988647, "learning_rate": 9.967701023178717e-05, "loss": 0.0125, "step": 8440 }, { "grad_norm": 0.27227696776390076, "learning_rate": 9.967513114694282e-05, "loss": 0.0135, "step": 8450 }, { "grad_norm": 0.22241003811359406, "learning_rate": 9.967324662970079e-05, "loss": 0.0128, "step": 8460 }, { "grad_norm": 0.22279860079288483, "learning_rate": 9.96713566802672e-05, "loss": 0.0115, "step": 8470 }, { "grad_norm": 0.23938827216625214, "learning_rate": 9.966946129884873e-05, "loss": 0.0117, "step": 8480 }, { "grad_norm": 0.19085048139095306, "learning_rate": 9.966756048565265e-05, "loss": 0.0114, "step": 8490 }, { "grad_norm": 0.2597010135650635, "learning_rate": 9.966565424088681e-05, "loss": 0.0142, "step": 8500 }, { "grad_norm": 0.2213471680879593, "learning_rate": 9.96637425647597e-05, "loss": 0.013, "step": 8510 }, { "grad_norm": 0.24011047184467316, "learning_rate": 9.966182545748038e-05, "loss": 0.0122, "step": 8520 }, { "grad_norm": 0.20737752318382263, "learning_rate": 9.96599029192585e-05, "loss": 0.0137, "step": 8530 }, { "grad_norm": 0.15412428975105286, "learning_rate": 9.965797495030428e-05, "loss": 0.0118, "step": 8540 }, { "grad_norm": 0.19784517586231232, "learning_rate": 9.96560415508286e-05, "loss": 0.012, "step": 8550 }, { "grad_norm": 0.2000436782836914, "learning_rate": 9.965410272104286e-05, "loss": 0.0123, "step": 8560 }, { "grad_norm": 0.1694840043783188, "learning_rate": 9.96521584611591e-05, "loss": 0.0128, "step": 8570 }, { "grad_norm": 0.2771122455596924, "learning_rate": 9.965020877138994e-05, "loss": 0.0131, "step": 8580 }, { "grad_norm": 0.2610730230808258, "learning_rate": 9.964825365194861e-05, "loss": 0.0139, "step": 8590 }, { "grad_norm": 0.14113740622997284, "learning_rate": 9.96462931030489e-05, "loss": 0.0127, "step": 8600 }, { "grad_norm": 0.19901776313781738, "learning_rate": 9.96443271249052e-05, "loss": 0.013, "step": 8610 }, { "grad_norm": 0.20208045840263367, "learning_rate": 9.964235571773255e-05, "loss": 0.0116, "step": 8620 }, { "grad_norm": 0.17395547032356262, "learning_rate": 9.96403788817465e-05, "loss": 0.0126, "step": 8630 }, { "grad_norm": 0.20207159221172333, "learning_rate": 9.963839661716325e-05, "loss": 0.0119, "step": 8640 }, { "grad_norm": 0.27877870202064514, "learning_rate": 9.963640892419958e-05, "loss": 0.0122, "step": 8650 }, { "grad_norm": 0.2515866756439209, "learning_rate": 9.963441580307286e-05, "loss": 0.0124, "step": 8660 }, { "grad_norm": 0.19488373398780823, "learning_rate": 9.963241725400104e-05, "loss": 0.0113, "step": 8670 }, { "grad_norm": 0.19681581854820251, "learning_rate": 9.963041327720271e-05, "loss": 0.0114, "step": 8680 }, { "grad_norm": 0.1937982738018036, "learning_rate": 9.962840387289697e-05, "loss": 0.0121, "step": 8690 }, { "grad_norm": 0.180296391248703, "learning_rate": 9.962638904130363e-05, "loss": 0.0156, "step": 8700 }, { "grad_norm": 0.334004282951355, "learning_rate": 9.962436878264298e-05, "loss": 0.0129, "step": 8710 }, { "grad_norm": 0.18855765461921692, "learning_rate": 9.962234309713598e-05, "loss": 0.0144, "step": 8720 }, { "grad_norm": 0.263766348361969, "learning_rate": 9.962031198500414e-05, "loss": 0.0144, "step": 8730 }, { "grad_norm": 0.29735755920410156, "learning_rate": 9.961827544646958e-05, "loss": 0.0141, "step": 8740 }, { "grad_norm": 0.25716373324394226, "learning_rate": 9.961623348175501e-05, "loss": 0.0139, "step": 8750 }, { "grad_norm": 0.2637555003166199, "learning_rate": 9.961418609108377e-05, "loss": 0.0149, "step": 8760 }, { "grad_norm": 0.24132443964481354, "learning_rate": 9.961213327467971e-05, "loss": 0.0157, "step": 8770 }, { "grad_norm": 0.2007995843887329, "learning_rate": 9.961007503276736e-05, "loss": 0.0144, "step": 8780 }, { "grad_norm": 0.3103058934211731, "learning_rate": 9.960801136557179e-05, "loss": 0.0167, "step": 8790 }, { "grad_norm": 0.26409807801246643, "learning_rate": 9.960594227331866e-05, "loss": 0.0158, "step": 8800 }, { "grad_norm": 0.23932677507400513, "learning_rate": 9.960386775623429e-05, "loss": 0.0139, "step": 8810 }, { "grad_norm": 0.2360755056142807, "learning_rate": 9.96017878145455e-05, "loss": 0.0165, "step": 8820 }, { "grad_norm": 0.2162315845489502, "learning_rate": 9.959970244847977e-05, "loss": 0.0163, "step": 8830 }, { "grad_norm": 0.17307664453983307, "learning_rate": 9.959761165826518e-05, "loss": 0.0145, "step": 8840 }, { "grad_norm": 0.18648959696292877, "learning_rate": 9.959551544413033e-05, "loss": 0.0152, "step": 8850 }, { "grad_norm": 0.2813909947872162, "learning_rate": 9.959341380630448e-05, "loss": 0.0138, "step": 8860 }, { "grad_norm": 0.1492399424314499, "learning_rate": 9.959130674501746e-05, "loss": 0.0137, "step": 8870 }, { "grad_norm": 0.12129876762628555, "learning_rate": 9.958919426049968e-05, "loss": 0.0122, "step": 8880 }, { "grad_norm": 0.24869176745414734, "learning_rate": 9.958707635298219e-05, "loss": 0.0127, "step": 8890 }, { "grad_norm": 0.16108860075473785, "learning_rate": 9.958495302269657e-05, "loss": 0.0123, "step": 8900 }, { "grad_norm": 0.20158176124095917, "learning_rate": 9.958282426987503e-05, "loss": 0.0136, "step": 8910 }, { "grad_norm": 0.1647031456232071, "learning_rate": 9.95806900947504e-05, "loss": 0.0126, "step": 8920 }, { "grad_norm": 0.16095387935638428, "learning_rate": 9.957855049755604e-05, "loss": 0.0125, "step": 8930 }, { "grad_norm": 0.19403661787509918, "learning_rate": 9.957640547852593e-05, "loss": 0.0121, "step": 8940 }, { "grad_norm": 0.21150296926498413, "learning_rate": 9.957425503789466e-05, "loss": 0.0115, "step": 8950 }, { "grad_norm": 0.22667384147644043, "learning_rate": 9.957209917589738e-05, "loss": 0.0132, "step": 8960 }, { "grad_norm": 0.2861906588077545, "learning_rate": 9.956993789276987e-05, "loss": 0.0118, "step": 8970 }, { "grad_norm": 0.2326248288154602, "learning_rate": 9.956777118874847e-05, "loss": 0.0115, "step": 8980 }, { "grad_norm": 0.24861827492713928, "learning_rate": 9.956559906407016e-05, "loss": 0.0119, "step": 8990 }, { "grad_norm": 0.2055436372756958, "learning_rate": 9.956342151897245e-05, "loss": 0.012, "step": 9000 }, { "grad_norm": 0.20783855020999908, "learning_rate": 9.956123855369346e-05, "loss": 0.0122, "step": 9010 }, { "grad_norm": 0.22345402836799622, "learning_rate": 9.955905016847196e-05, "loss": 0.0117, "step": 9020 }, { "grad_norm": 0.209931418299675, "learning_rate": 9.955685636354723e-05, "loss": 0.0114, "step": 9030 }, { "grad_norm": 0.19026412069797516, "learning_rate": 9.95546571391592e-05, "loss": 0.0124, "step": 9040 }, { "grad_norm": 0.1904711127281189, "learning_rate": 9.955245249554837e-05, "loss": 0.013, "step": 9050 }, { "grad_norm": 0.26353418827056885, "learning_rate": 9.955024243295582e-05, "loss": 0.0126, "step": 9060 }, { "grad_norm": 0.15865768492221832, "learning_rate": 9.954802695162328e-05, "loss": 0.0136, "step": 9070 }, { "grad_norm": 0.20243534445762634, "learning_rate": 9.954580605179302e-05, "loss": 0.012, "step": 9080 }, { "grad_norm": 0.2975022792816162, "learning_rate": 9.954357973370788e-05, "loss": 0.0141, "step": 9090 }, { "grad_norm": 0.14293338358402252, "learning_rate": 9.954134799761135e-05, "loss": 0.0115, "step": 9100 }, { "grad_norm": 0.16641628742218018, "learning_rate": 9.953911084374748e-05, "loss": 0.0135, "step": 9110 }, { "grad_norm": 0.18703599274158478, "learning_rate": 9.953686827236093e-05, "loss": 0.0134, "step": 9120 }, { "grad_norm": 0.22326013445854187, "learning_rate": 9.953462028369695e-05, "loss": 0.0128, "step": 9130 }, { "grad_norm": 0.2529016435146332, "learning_rate": 9.953236687800136e-05, "loss": 0.0133, "step": 9140 }, { "grad_norm": 0.1897270381450653, "learning_rate": 9.95301080555206e-05, "loss": 0.0122, "step": 9150 }, { "grad_norm": 0.1873290091753006, "learning_rate": 9.952784381650171e-05, "loss": 0.013, "step": 9160 }, { "grad_norm": 0.2083888053894043, "learning_rate": 9.952557416119226e-05, "loss": 0.0142, "step": 9170 }, { "grad_norm": 0.30247774720191956, "learning_rate": 9.95232990898405e-05, "loss": 0.0139, "step": 9180 }, { "grad_norm": 0.1690492480993271, "learning_rate": 9.95210186026952e-05, "loss": 0.0151, "step": 9190 }, { "grad_norm": 0.2306598573923111, "learning_rate": 9.951873270000576e-05, "loss": 0.0144, "step": 9200 }, { "grad_norm": 0.17201416194438934, "learning_rate": 9.951644138202216e-05, "loss": 0.0153, "step": 9210 }, { "grad_norm": 0.22010384500026703, "learning_rate": 9.951414464899498e-05, "loss": 0.014, "step": 9220 }, { "grad_norm": 0.2255999594926834, "learning_rate": 9.951184250117538e-05, "loss": 0.0137, "step": 9230 }, { "grad_norm": 0.18567241728305817, "learning_rate": 9.950953493881513e-05, "loss": 0.0122, "step": 9240 }, { "grad_norm": 0.25064143538475037, "learning_rate": 9.950722196216658e-05, "loss": 0.0122, "step": 9250 }, { "grad_norm": 0.16418348252773285, "learning_rate": 9.950490357148265e-05, "loss": 0.0136, "step": 9260 }, { "grad_norm": 0.20028679072856903, "learning_rate": 9.950257976701692e-05, "loss": 0.0127, "step": 9270 }, { "grad_norm": 0.2049146145582199, "learning_rate": 9.950025054902348e-05, "loss": 0.0119, "step": 9280 }, { "grad_norm": 0.2126929759979248, "learning_rate": 9.949791591775706e-05, "loss": 0.0122, "step": 9290 }, { "grad_norm": 0.23733720183372498, "learning_rate": 9.949557587347298e-05, "loss": 0.0127, "step": 9300 }, { "grad_norm": 0.15998715162277222, "learning_rate": 9.949323041642713e-05, "loss": 0.0118, "step": 9310 }, { "grad_norm": 0.20829685032367706, "learning_rate": 9.949087954687602e-05, "loss": 0.0119, "step": 9320 }, { "grad_norm": 0.26304030418395996, "learning_rate": 9.948852326507672e-05, "loss": 0.0116, "step": 9330 }, { "grad_norm": 0.39989563822746277, "learning_rate": 9.948616157128694e-05, "loss": 0.0115, "step": 9340 }, { "grad_norm": 0.16542014479637146, "learning_rate": 9.948379446576493e-05, "loss": 0.0148, "step": 9350 }, { "grad_norm": 0.273042768239975, "learning_rate": 9.948142194876952e-05, "loss": 0.0125, "step": 9360 }, { "grad_norm": 0.1552220731973648, "learning_rate": 9.947904402056024e-05, "loss": 0.0128, "step": 9370 }, { "grad_norm": 0.15765748918056488, "learning_rate": 9.947666068139708e-05, "loss": 0.0121, "step": 9380 }, { "grad_norm": 0.2158866822719574, "learning_rate": 9.947427193154071e-05, "loss": 0.0122, "step": 9390 }, { "grad_norm": 0.20932362973690033, "learning_rate": 9.947187777125233e-05, "loss": 0.0128, "step": 9400 }, { "grad_norm": 0.20050828158855438, "learning_rate": 9.946947820079377e-05, "loss": 0.0126, "step": 9410 }, { "grad_norm": 0.16346341371536255, "learning_rate": 9.946707322042747e-05, "loss": 0.0124, "step": 9420 }, { "grad_norm": 0.1952684372663498, "learning_rate": 9.94646628304164e-05, "loss": 0.0118, "step": 9430 }, { "grad_norm": 0.18866761028766632, "learning_rate": 9.946224703102418e-05, "loss": 0.0108, "step": 9440 }, { "grad_norm": 0.18459218740463257, "learning_rate": 9.945982582251498e-05, "loss": 0.0117, "step": 9450 }, { "grad_norm": 0.21241551637649536, "learning_rate": 9.94573992051536e-05, "loss": 0.0123, "step": 9460 }, { "grad_norm": 0.16412173211574554, "learning_rate": 9.94549671792054e-05, "loss": 0.0107, "step": 9470 }, { "grad_norm": 0.1694403439760208, "learning_rate": 9.945252974493635e-05, "loss": 0.0114, "step": 9480 }, { "grad_norm": 0.23947100341320038, "learning_rate": 9.9450086902613e-05, "loss": 0.0133, "step": 9490 }, { "grad_norm": 0.21642908453941345, "learning_rate": 9.944763865250248e-05, "loss": 0.0128, "step": 9500 }, { "grad_norm": 0.2325923889875412, "learning_rate": 9.944518499487254e-05, "loss": 0.0111, "step": 9510 }, { "grad_norm": 0.18427598476409912, "learning_rate": 9.944272592999151e-05, "loss": 0.0135, "step": 9520 }, { "grad_norm": 0.23660695552825928, "learning_rate": 9.94402614581283e-05, "loss": 0.014, "step": 9530 }, { "grad_norm": 0.23011170327663422, "learning_rate": 9.943779157955244e-05, "loss": 0.0135, "step": 9540 }, { "grad_norm": 0.20934879779815674, "learning_rate": 9.943531629453403e-05, "loss": 0.0138, "step": 9550 }, { "grad_norm": 0.16580882668495178, "learning_rate": 9.943283560334375e-05, "loss": 0.0146, "step": 9560 }, { "grad_norm": 0.18453449010849, "learning_rate": 9.943034950625288e-05, "loss": 0.0137, "step": 9570 }, { "grad_norm": 0.16457393765449524, "learning_rate": 9.942785800353332e-05, "loss": 0.0128, "step": 9580 }, { "grad_norm": 0.12824812531471252, "learning_rate": 9.942536109545751e-05, "loss": 0.0112, "step": 9590 }, { "grad_norm": 0.23857146501541138, "learning_rate": 9.942285878229853e-05, "loss": 0.0138, "step": 9600 }, { "grad_norm": 0.2942847013473511, "learning_rate": 9.942035106433001e-05, "loss": 0.0122, "step": 9610 }, { "grad_norm": 0.18199725449085236, "learning_rate": 9.94178379418262e-05, "loss": 0.0121, "step": 9620 }, { "grad_norm": 0.2298555225133896, "learning_rate": 9.941531941506194e-05, "loss": 0.014, "step": 9630 }, { "grad_norm": 0.23166434466838837, "learning_rate": 9.941279548431263e-05, "loss": 0.0128, "step": 9640 }, { "grad_norm": 0.14731933176517487, "learning_rate": 9.941026614985431e-05, "loss": 0.0122, "step": 9650 }, { "grad_norm": 0.17860868573188782, "learning_rate": 9.940773141196357e-05, "loss": 0.0142, "step": 9660 }, { "grad_norm": 0.23296837508678436, "learning_rate": 9.94051912709176e-05, "loss": 0.0128, "step": 9670 }, { "grad_norm": 0.21185709536075592, "learning_rate": 9.940264572699421e-05, "loss": 0.013, "step": 9680 }, { "grad_norm": 0.2198212444782257, "learning_rate": 9.940009478047174e-05, "loss": 0.0141, "step": 9690 }, { "grad_norm": 0.23278194665908813, "learning_rate": 9.939753843162918e-05, "loss": 0.0137, "step": 9700 }, { "grad_norm": 0.19546689093112946, "learning_rate": 9.939497668074609e-05, "loss": 0.0124, "step": 9710 }, { "grad_norm": 0.19590875506401062, "learning_rate": 9.93924095281026e-05, "loss": 0.0132, "step": 9720 }, { "grad_norm": 0.16719414293766022, "learning_rate": 9.938983697397948e-05, "loss": 0.0107, "step": 9730 }, { "grad_norm": 0.20548056066036224, "learning_rate": 9.938725901865805e-05, "loss": 0.0112, "step": 9740 }, { "grad_norm": 0.14978767931461334, "learning_rate": 9.93846756624202e-05, "loss": 0.0102, "step": 9750 }, { "grad_norm": 0.19421014189720154, "learning_rate": 9.938208690554849e-05, "loss": 0.0126, "step": 9760 }, { "grad_norm": 0.17681850492954254, "learning_rate": 9.9379492748326e-05, "loss": 0.0111, "step": 9770 }, { "grad_norm": 0.2683127522468567, "learning_rate": 9.937689319103641e-05, "loss": 0.0132, "step": 9780 }, { "grad_norm": 0.20323263108730316, "learning_rate": 9.937428823396404e-05, "loss": 0.0121, "step": 9790 }, { "grad_norm": 0.23328498005867004, "learning_rate": 9.937167787739372e-05, "loss": 0.0112, "step": 9800 }, { "grad_norm": 0.18614256381988525, "learning_rate": 9.936906212161095e-05, "loss": 0.0115, "step": 9810 }, { "grad_norm": 0.18086983263492584, "learning_rate": 9.936644096690176e-05, "loss": 0.0107, "step": 9820 }, { "grad_norm": 0.22937187552452087, "learning_rate": 9.936381441355282e-05, "loss": 0.0132, "step": 9830 }, { "grad_norm": 0.20154418051242828, "learning_rate": 9.936118246185136e-05, "loss": 0.0125, "step": 9840 }, { "grad_norm": 0.1948123276233673, "learning_rate": 9.935854511208518e-05, "loss": 0.0108, "step": 9850 }, { "grad_norm": 0.1618977040052414, "learning_rate": 9.935590236454272e-05, "loss": 0.0106, "step": 9860 }, { "grad_norm": 0.25239312648773193, "learning_rate": 9.935325421951298e-05, "loss": 0.0123, "step": 9870 }, { "grad_norm": 0.23198235034942627, "learning_rate": 9.935060067728557e-05, "loss": 0.0136, "step": 9880 }, { "grad_norm": 0.21530213952064514, "learning_rate": 9.934794173815067e-05, "loss": 0.0116, "step": 9890 }, { "grad_norm": 0.20845156908035278, "learning_rate": 9.934527740239906e-05, "loss": 0.0125, "step": 9900 }, { "grad_norm": 0.17664584517478943, "learning_rate": 9.934260767032209e-05, "loss": 0.0122, "step": 9910 }, { "grad_norm": 0.1869087815284729, "learning_rate": 9.933993254221172e-05, "loss": 0.0118, "step": 9920 }, { "grad_norm": 0.15793828666210175, "learning_rate": 9.933725201836053e-05, "loss": 0.0112, "step": 9930 }, { "grad_norm": 0.13836295902729034, "learning_rate": 9.933456609906162e-05, "loss": 0.0104, "step": 9940 }, { "grad_norm": 0.1916368007659912, "learning_rate": 9.933187478460875e-05, "loss": 0.0109, "step": 9950 }, { "grad_norm": 0.16216084361076355, "learning_rate": 9.93291780752962e-05, "loss": 0.0129, "step": 9960 }, { "grad_norm": 0.19652359187602997, "learning_rate": 9.932647597141893e-05, "loss": 0.0126, "step": 9970 }, { "grad_norm": 0.1831931620836258, "learning_rate": 9.932376847327239e-05, "loss": 0.0125, "step": 9980 }, { "grad_norm": 0.16473771631717682, "learning_rate": 9.932105558115268e-05, "loss": 0.0101, "step": 9990 }, { "grad_norm": 0.1668979525566101, "learning_rate": 9.931833729535651e-05, "loss": 0.0101, "step": 10000 }, { "grad_norm": 0.19909895956516266, "learning_rate": 9.931561361618111e-05, "loss": 0.0106, "step": 10010 }, { "grad_norm": 0.20508229732513428, "learning_rate": 9.931288454392435e-05, "loss": 0.012, "step": 10020 }, { "grad_norm": 0.1702243536710739, "learning_rate": 9.931015007888467e-05, "loss": 0.0102, "step": 10030 }, { "grad_norm": 0.19457601010799408, "learning_rate": 9.930741022136112e-05, "loss": 0.0111, "step": 10040 }, { "grad_norm": 0.18353278934955597, "learning_rate": 9.930466497165333e-05, "loss": 0.0109, "step": 10050 }, { "grad_norm": 0.20321205258369446, "learning_rate": 9.93019143300615e-05, "loss": 0.0114, "step": 10060 }, { "grad_norm": 0.18413247168064117, "learning_rate": 9.929915829688644e-05, "loss": 0.0137, "step": 10070 }, { "grad_norm": 0.19424480199813843, "learning_rate": 9.929639687242955e-05, "loss": 0.0129, "step": 10080 }, { "grad_norm": 0.1506660133600235, "learning_rate": 9.929363005699281e-05, "loss": 0.0128, "step": 10090 }, { "grad_norm": 0.17955568432807922, "learning_rate": 9.92908578508788e-05, "loss": 0.0118, "step": 10100 }, { "grad_norm": 0.246765598654747, "learning_rate": 9.928808025439069e-05, "loss": 0.0131, "step": 10110 }, { "grad_norm": 0.2552172541618347, "learning_rate": 9.928529726783223e-05, "loss": 0.0123, "step": 10120 }, { "grad_norm": 0.20544452965259552, "learning_rate": 9.928250889150774e-05, "loss": 0.0112, "step": 10130 }, { "grad_norm": 0.18724550306797028, "learning_rate": 9.92797151257222e-05, "loss": 0.0105, "step": 10140 }, { "grad_norm": 0.24017511308193207, "learning_rate": 9.927691597078108e-05, "loss": 0.0126, "step": 10150 }, { "grad_norm": 0.15393739938735962, "learning_rate": 9.927411142699053e-05, "loss": 0.0139, "step": 10160 }, { "grad_norm": 0.18790586292743683, "learning_rate": 9.927130149465725e-05, "loss": 0.0121, "step": 10170 }, { "grad_norm": 0.2207312136888504, "learning_rate": 9.92684861740885e-05, "loss": 0.013, "step": 10180 }, { "grad_norm": 0.2429584562778473, "learning_rate": 9.926566546559217e-05, "loss": 0.0184, "step": 10190 }, { "grad_norm": 0.26546767354011536, "learning_rate": 9.926283936947673e-05, "loss": 0.0156, "step": 10200 }, { "grad_norm": 0.23074433207511902, "learning_rate": 9.926000788605126e-05, "loss": 0.0138, "step": 10210 }, { "grad_norm": 0.22443950176239014, "learning_rate": 9.92571710156254e-05, "loss": 0.0144, "step": 10220 }, { "grad_norm": 0.23210729658603668, "learning_rate": 9.925432875850936e-05, "loss": 0.0137, "step": 10230 }, { "grad_norm": 0.22900648415088654, "learning_rate": 9.925148111501396e-05, "loss": 0.0143, "step": 10240 }, { "grad_norm": 0.19030025601387024, "learning_rate": 9.924862808545066e-05, "loss": 0.0121, "step": 10250 }, { "grad_norm": 0.14732986688613892, "learning_rate": 9.924576967013141e-05, "loss": 0.0116, "step": 10260 }, { "grad_norm": 0.19047176837921143, "learning_rate": 9.924290586936887e-05, "loss": 0.0123, "step": 10270 }, { "grad_norm": 0.1757601797580719, "learning_rate": 9.924003668347614e-05, "loss": 0.0113, "step": 10280 }, { "grad_norm": 0.14995163679122925, "learning_rate": 9.923716211276704e-05, "loss": 0.0126, "step": 10290 }, { "grad_norm": 0.1997346580028534, "learning_rate": 9.923428215755594e-05, "loss": 0.0118, "step": 10300 }, { "grad_norm": 0.15675394237041473, "learning_rate": 9.923139681815775e-05, "loss": 0.0114, "step": 10310 }, { "grad_norm": 0.17617234587669373, "learning_rate": 9.922850609488801e-05, "loss": 0.0117, "step": 10320 }, { "grad_norm": 0.13235239684581757, "learning_rate": 9.922560998806287e-05, "loss": 0.0109, "step": 10330 }, { "grad_norm": 0.17859171330928802, "learning_rate": 9.922270849799905e-05, "loss": 0.0109, "step": 10340 }, { "grad_norm": 0.18889211118221283, "learning_rate": 9.92198016250138e-05, "loss": 0.0121, "step": 10350 }, { "grad_norm": 0.1715780794620514, "learning_rate": 9.921688936942506e-05, "loss": 0.0131, "step": 10360 }, { "grad_norm": 0.22461654245853424, "learning_rate": 9.921397173155129e-05, "loss": 0.012, "step": 10370 }, { "grad_norm": 0.16259844601154327, "learning_rate": 9.921104871171157e-05, "loss": 0.0124, "step": 10380 }, { "grad_norm": 0.12534573674201965, "learning_rate": 9.920812031022554e-05, "loss": 0.0129, "step": 10390 }, { "grad_norm": 0.17756426334381104, "learning_rate": 9.920518652741348e-05, "loss": 0.0114, "step": 10400 }, { "grad_norm": 0.22024551033973694, "learning_rate": 9.920224736359618e-05, "loss": 0.012, "step": 10410 }, { "grad_norm": 0.1923169046640396, "learning_rate": 9.91993028190951e-05, "loss": 0.0114, "step": 10420 }, { "grad_norm": 0.12893813848495483, "learning_rate": 9.919635289423222e-05, "loss": 0.0111, "step": 10430 }, { "grad_norm": 0.1720774918794632, "learning_rate": 9.919339758933015e-05, "loss": 0.011, "step": 10440 }, { "grad_norm": 0.20593856275081635, "learning_rate": 9.919043690471209e-05, "loss": 0.012, "step": 10450 }, { "grad_norm": 0.23754245042800903, "learning_rate": 9.91874708407018e-05, "loss": 0.0118, "step": 10460 }, { "grad_norm": 0.28227922320365906, "learning_rate": 9.918449939762367e-05, "loss": 0.0121, "step": 10470 }, { "grad_norm": 0.19473347067832947, "learning_rate": 9.91815225758026e-05, "loss": 0.0115, "step": 10480 }, { "grad_norm": 0.15916940569877625, "learning_rate": 9.917854037556419e-05, "loss": 0.0129, "step": 10490 }, { "grad_norm": 0.17357237637043, "learning_rate": 9.917555279723454e-05, "loss": 0.0132, "step": 10500 }, { "grad_norm": 0.18053993582725525, "learning_rate": 9.917255984114036e-05, "loss": 0.0126, "step": 10510 }, { "grad_norm": 0.18526530265808105, "learning_rate": 9.916956150760896e-05, "loss": 0.012, "step": 10520 }, { "grad_norm": 0.20805984735488892, "learning_rate": 9.916655779696826e-05, "loss": 0.0103, "step": 10530 }, { "grad_norm": 0.1771669089794159, "learning_rate": 9.916354870954671e-05, "loss": 0.011, "step": 10540 }, { "grad_norm": 0.1950640082359314, "learning_rate": 9.91605342456734e-05, "loss": 0.0107, "step": 10550 }, { "grad_norm": 0.19985724985599518, "learning_rate": 9.915751440567795e-05, "loss": 0.0105, "step": 10560 }, { "grad_norm": 0.13638935983181, "learning_rate": 9.915448918989066e-05, "loss": 0.0117, "step": 10570 }, { "grad_norm": 0.1884709745645523, "learning_rate": 9.915145859864232e-05, "loss": 0.0104, "step": 10580 }, { "grad_norm": 0.1698657125234604, "learning_rate": 9.914842263226437e-05, "loss": 0.0109, "step": 10590 }, { "grad_norm": 0.23908467590808868, "learning_rate": 9.914538129108882e-05, "loss": 0.0111, "step": 10600 }, { "grad_norm": 0.1329643428325653, "learning_rate": 9.914233457544825e-05, "loss": 0.0108, "step": 10610 }, { "grad_norm": 0.1775931566953659, "learning_rate": 9.913928248567586e-05, "loss": 0.0113, "step": 10620 }, { "grad_norm": 0.14128510653972626, "learning_rate": 9.913622502210542e-05, "loss": 0.0097, "step": 10630 }, { "grad_norm": 0.15860438346862793, "learning_rate": 9.913316218507128e-05, "loss": 0.0096, "step": 10640 }, { "grad_norm": 0.17806458473205566, "learning_rate": 9.91300939749084e-05, "loss": 0.0106, "step": 10650 }, { "grad_norm": 0.22654326260089874, "learning_rate": 9.91270203919523e-05, "loss": 0.012, "step": 10660 }, { "grad_norm": 0.19426171481609344, "learning_rate": 9.912394143653912e-05, "loss": 0.0116, "step": 10670 }, { "grad_norm": 0.1763019859790802, "learning_rate": 9.912085710900555e-05, "loss": 0.0127, "step": 10680 }, { "grad_norm": 0.17617563903331757, "learning_rate": 9.911776740968892e-05, "loss": 0.0103, "step": 10690 }, { "grad_norm": 0.24889948964118958, "learning_rate": 9.911467233892709e-05, "loss": 0.0104, "step": 10700 }, { "grad_norm": 0.18424110114574432, "learning_rate": 9.911157189705853e-05, "loss": 0.0111, "step": 10710 }, { "grad_norm": 0.10818704217672348, "learning_rate": 9.910846608442229e-05, "loss": 0.0108, "step": 10720 }, { "grad_norm": 0.19468170404434204, "learning_rate": 9.910535490135805e-05, "loss": 0.0103, "step": 10730 }, { "grad_norm": 0.20660732686519623, "learning_rate": 9.910223834820603e-05, "loss": 0.0104, "step": 10740 }, { "grad_norm": 0.20875300467014313, "learning_rate": 9.909911642530703e-05, "loss": 0.0099, "step": 10750 }, { "grad_norm": 0.18506291508674622, "learning_rate": 9.909598913300249e-05, "loss": 0.0113, "step": 10760 }, { "grad_norm": 0.18285979330539703, "learning_rate": 9.909285647163438e-05, "loss": 0.0106, "step": 10770 }, { "grad_norm": 0.14389081299304962, "learning_rate": 9.908971844154531e-05, "loss": 0.0118, "step": 10780 }, { "grad_norm": 0.1660035252571106, "learning_rate": 9.908657504307843e-05, "loss": 0.0106, "step": 10790 }, { "grad_norm": 0.2080429494380951, "learning_rate": 9.908342627657751e-05, "loss": 0.0101, "step": 10800 }, { "grad_norm": 0.16338184475898743, "learning_rate": 9.908027214238689e-05, "loss": 0.0107, "step": 10810 }, { "grad_norm": 0.196008563041687, "learning_rate": 9.90771126408515e-05, "loss": 0.0103, "step": 10820 }, { "grad_norm": 0.1878090351819992, "learning_rate": 9.907394777231685e-05, "loss": 0.0097, "step": 10830 }, { "grad_norm": 0.1813305765390396, "learning_rate": 9.907077753712905e-05, "loss": 0.0103, "step": 10840 }, { "grad_norm": 0.18464523553848267, "learning_rate": 9.906760193563482e-05, "loss": 0.0106, "step": 10850 }, { "grad_norm": 0.20879048109054565, "learning_rate": 9.906442096818139e-05, "loss": 0.0113, "step": 10860 }, { "grad_norm": 0.13225239515304565, "learning_rate": 9.906123463511665e-05, "loss": 0.01, "step": 10870 }, { "grad_norm": 0.1343221664428711, "learning_rate": 9.905804293678907e-05, "loss": 0.0116, "step": 10880 }, { "grad_norm": 0.16416393220424652, "learning_rate": 9.905484587354766e-05, "loss": 0.0108, "step": 10890 }, { "grad_norm": 0.1957135796546936, "learning_rate": 9.905164344574205e-05, "loss": 0.0092, "step": 10900 }, { "grad_norm": 0.14667874574661255, "learning_rate": 9.904843565372248e-05, "loss": 0.0107, "step": 10910 }, { "grad_norm": 0.15403442084789276, "learning_rate": 9.904522249783972e-05, "loss": 0.0101, "step": 10920 }, { "grad_norm": 0.1865634024143219, "learning_rate": 9.904200397844517e-05, "loss": 0.0117, "step": 10930 }, { "grad_norm": 0.14631448686122894, "learning_rate": 9.903878009589078e-05, "loss": 0.0111, "step": 10940 }, { "grad_norm": 0.18682849407196045, "learning_rate": 9.903555085052915e-05, "loss": 0.0101, "step": 10950 }, { "grad_norm": 0.12840791046619415, "learning_rate": 9.903231624271338e-05, "loss": 0.0098, "step": 10960 }, { "grad_norm": 0.17219194769859314, "learning_rate": 9.902907627279724e-05, "loss": 0.0108, "step": 10970 }, { "grad_norm": 0.21710030734539032, "learning_rate": 9.902583094113504e-05, "loss": 0.0103, "step": 10980 }, { "grad_norm": 0.17704255878925323, "learning_rate": 9.902258024808168e-05, "loss": 0.0105, "step": 10990 }, { "grad_norm": 0.1972360759973526, "learning_rate": 9.901932419399264e-05, "loss": 0.0106, "step": 11000 }, { "grad_norm": 0.16310952603816986, "learning_rate": 9.9016062779224e-05, "loss": 0.0097, "step": 11010 }, { "grad_norm": 0.14685110747814178, "learning_rate": 9.901279600413242e-05, "loss": 0.0101, "step": 11020 }, { "grad_norm": 0.14195112884044647, "learning_rate": 9.900952386907518e-05, "loss": 0.0099, "step": 11030 }, { "grad_norm": 0.20359379053115845, "learning_rate": 9.90062463744101e-05, "loss": 0.0114, "step": 11040 }, { "grad_norm": 0.18331918120384216, "learning_rate": 9.900296352049558e-05, "loss": 0.0112, "step": 11050 }, { "grad_norm": 0.2098132073879242, "learning_rate": 9.899967530769065e-05, "loss": 0.0111, "step": 11060 }, { "grad_norm": 0.22360172867774963, "learning_rate": 9.899638173635489e-05, "loss": 0.0125, "step": 11070 }, { "grad_norm": 0.23242321610450745, "learning_rate": 9.899308280684849e-05, "loss": 0.0125, "step": 11080 }, { "grad_norm": 0.22904212772846222, "learning_rate": 9.898977851953222e-05, "loss": 0.0131, "step": 11090 }, { "grad_norm": 0.22596405446529388, "learning_rate": 9.898646887476741e-05, "loss": 0.0118, "step": 11100 }, { "grad_norm": 0.19737333059310913, "learning_rate": 9.898315387291603e-05, "loss": 0.0127, "step": 11110 }, { "grad_norm": 0.23183348774909973, "learning_rate": 9.89798335143406e-05, "loss": 0.0109, "step": 11120 }, { "grad_norm": 0.23366901278495789, "learning_rate": 9.897650779940419e-05, "loss": 0.011, "step": 11130 }, { "grad_norm": 0.16186729073524475, "learning_rate": 9.897317672847054e-05, "loss": 0.0123, "step": 11140 }, { "grad_norm": 0.15645286440849304, "learning_rate": 9.89698403019039e-05, "loss": 0.0127, "step": 11150 }, { "grad_norm": 0.2518731355667114, "learning_rate": 9.896649852006917e-05, "loss": 0.0123, "step": 11160 }, { "grad_norm": 0.21529148519039154, "learning_rate": 9.896315138333177e-05, "loss": 0.0131, "step": 11170 }, { "grad_norm": 0.23417213559150696, "learning_rate": 9.895979889205774e-05, "loss": 0.015, "step": 11180 }, { "grad_norm": 0.23698244988918304, "learning_rate": 9.895644104661372e-05, "loss": 0.0123, "step": 11190 }, { "grad_norm": 0.20960065722465515, "learning_rate": 9.895307784736691e-05, "loss": 0.0139, "step": 11200 }, { "grad_norm": 0.21935585141181946, "learning_rate": 9.894970929468512e-05, "loss": 0.0134, "step": 11210 }, { "grad_norm": 0.1663726270198822, "learning_rate": 9.89463353889367e-05, "loss": 0.0123, "step": 11220 }, { "grad_norm": 0.1656630039215088, "learning_rate": 9.894295613049065e-05, "loss": 0.0135, "step": 11230 }, { "grad_norm": 0.1418045610189438, "learning_rate": 9.893957151971649e-05, "loss": 0.0115, "step": 11240 }, { "grad_norm": 0.1249738410115242, "learning_rate": 9.893618155698436e-05, "loss": 0.0104, "step": 11250 }, { "grad_norm": 0.24992071092128754, "learning_rate": 9.8932786242665e-05, "loss": 0.0142, "step": 11260 }, { "grad_norm": 0.18503820896148682, "learning_rate": 9.89293855771297e-05, "loss": 0.0127, "step": 11270 }, { "grad_norm": 0.14253538846969604, "learning_rate": 9.892597956075036e-05, "loss": 0.0136, "step": 11280 }, { "grad_norm": 0.21770448982715607, "learning_rate": 9.892256819389947e-05, "loss": 0.0138, "step": 11290 }, { "grad_norm": 0.15497536957263947, "learning_rate": 9.891915147695006e-05, "loss": 0.0118, "step": 11300 }, { "grad_norm": 0.1708633154630661, "learning_rate": 9.891572941027577e-05, "loss": 0.0138, "step": 11310 }, { "grad_norm": 0.1512393355369568, "learning_rate": 9.89123019942509e-05, "loss": 0.0128, "step": 11320 }, { "grad_norm": 0.21423499286174774, "learning_rate": 9.89088692292502e-05, "loss": 0.0135, "step": 11330 }, { "grad_norm": 0.17292281985282898, "learning_rate": 9.89054311156491e-05, "loss": 0.0126, "step": 11340 }, { "grad_norm": 0.17538900673389435, "learning_rate": 9.890198765382357e-05, "loss": 0.0133, "step": 11350 }, { "grad_norm": 0.17221610248088837, "learning_rate": 9.889853884415021e-05, "loss": 0.0132, "step": 11360 }, { "grad_norm": 0.28965359926223755, "learning_rate": 9.889508468700614e-05, "loss": 0.0126, "step": 11370 }, { "grad_norm": 0.16930116713047028, "learning_rate": 9.889162518276915e-05, "loss": 0.0118, "step": 11380 }, { "grad_norm": 0.20894071459770203, "learning_rate": 9.888816033181752e-05, "loss": 0.0128, "step": 11390 }, { "grad_norm": 0.15748149156570435, "learning_rate": 9.888469013453018e-05, "loss": 0.0105, "step": 11400 }, { "grad_norm": 0.18928764760494232, "learning_rate": 9.888121459128663e-05, "loss": 0.0112, "step": 11410 }, { "grad_norm": 0.16739699244499207, "learning_rate": 9.887773370246693e-05, "loss": 0.0097, "step": 11420 }, { "grad_norm": 0.19781507551670074, "learning_rate": 9.887424746845177e-05, "loss": 0.0099, "step": 11430 }, { "grad_norm": 0.19430895149707794, "learning_rate": 9.887075588962239e-05, "loss": 0.0115, "step": 11440 }, { "grad_norm": 0.14696982502937317, "learning_rate": 9.88672589663606e-05, "loss": 0.0105, "step": 11450 }, { "grad_norm": 0.1893145889043808, "learning_rate": 9.886375669904886e-05, "loss": 0.0118, "step": 11460 }, { "grad_norm": 0.28422418236732483, "learning_rate": 9.886024908807014e-05, "loss": 0.0127, "step": 11470 }, { "grad_norm": 0.15173345804214478, "learning_rate": 9.885673613380806e-05, "loss": 0.0114, "step": 11480 }, { "grad_norm": 0.16228167712688446, "learning_rate": 9.885321783664676e-05, "loss": 0.0116, "step": 11490 }, { "grad_norm": 0.22692130506038666, "learning_rate": 9.884969419697101e-05, "loss": 0.0111, "step": 11500 }, { "grad_norm": 0.13843116164207458, "learning_rate": 9.884616521516614e-05, "loss": 0.0103, "step": 11510 }, { "grad_norm": 0.19920578598976135, "learning_rate": 9.88426308916181e-05, "loss": 0.0114, "step": 11520 }, { "grad_norm": 0.16419528424739838, "learning_rate": 9.883909122671335e-05, "loss": 0.0116, "step": 11530 }, { "grad_norm": 0.2714020311832428, "learning_rate": 9.883554622083904e-05, "loss": 0.0118, "step": 11540 }, { "grad_norm": 0.23034809529781342, "learning_rate": 9.88319958743828e-05, "loss": 0.012, "step": 11550 }, { "grad_norm": 0.19811208546161652, "learning_rate": 9.882844018773291e-05, "loss": 0.0123, "step": 11560 }, { "grad_norm": 0.15331265330314636, "learning_rate": 9.882487916127823e-05, "loss": 0.0105, "step": 11570 }, { "grad_norm": 0.2162710577249527, "learning_rate": 9.882131279540815e-05, "loss": 0.0117, "step": 11580 }, { "grad_norm": 0.15201956033706665, "learning_rate": 9.881774109051271e-05, "loss": 0.0101, "step": 11590 }, { "grad_norm": 0.15354777872562408, "learning_rate": 9.881416404698252e-05, "loss": 0.0111, "step": 11600 }, { "grad_norm": 0.12422540038824081, "learning_rate": 9.881058166520873e-05, "loss": 0.0103, "step": 11610 }, { "grad_norm": 0.17200353741645813, "learning_rate": 9.880699394558311e-05, "loss": 0.0103, "step": 11620 }, { "grad_norm": 0.1963767111301422, "learning_rate": 9.880340088849801e-05, "loss": 0.0107, "step": 11630 }, { "grad_norm": 0.2226557433605194, "learning_rate": 9.879980249434637e-05, "loss": 0.0107, "step": 11640 }, { "grad_norm": 0.16473597288131714, "learning_rate": 9.879619876352168e-05, "loss": 0.0104, "step": 11650 }, { "grad_norm": 0.19446615874767303, "learning_rate": 9.879258969641809e-05, "loss": 0.0104, "step": 11660 }, { "grad_norm": 0.22173282504081726, "learning_rate": 9.878897529343023e-05, "loss": 0.0111, "step": 11670 }, { "grad_norm": 0.1610378623008728, "learning_rate": 9.878535555495338e-05, "loss": 0.0103, "step": 11680 }, { "grad_norm": 0.12977071106433868, "learning_rate": 9.87817304813834e-05, "loss": 0.0107, "step": 11690 }, { "grad_norm": 0.19857534766197205, "learning_rate": 9.877810007311671e-05, "loss": 0.01, "step": 11700 }, { "grad_norm": 0.13126525282859802, "learning_rate": 9.877446433055035e-05, "loss": 0.0105, "step": 11710 }, { "grad_norm": 0.1589379608631134, "learning_rate": 9.877082325408191e-05, "loss": 0.0108, "step": 11720 }, { "grad_norm": 0.19083081185817719, "learning_rate": 9.876717684410954e-05, "loss": 0.011, "step": 11730 }, { "grad_norm": 0.12592154741287231, "learning_rate": 9.876352510103204e-05, "loss": 0.0117, "step": 11740 }, { "grad_norm": 0.2031751573085785, "learning_rate": 9.875986802524875e-05, "loss": 0.0101, "step": 11750 }, { "grad_norm": 0.1026829183101654, "learning_rate": 9.87562056171596e-05, "loss": 0.0114, "step": 11760 }, { "grad_norm": 0.22598569095134735, "learning_rate": 9.875253787716511e-05, "loss": 0.0119, "step": 11770 }, { "grad_norm": 0.15054255723953247, "learning_rate": 9.874886480566637e-05, "loss": 0.0111, "step": 11780 }, { "grad_norm": 0.1743924915790558, "learning_rate": 9.874518640306507e-05, "loss": 0.0112, "step": 11790 }, { "grad_norm": 0.1672212779521942, "learning_rate": 9.874150266976347e-05, "loss": 0.0098, "step": 11800 }, { "grad_norm": 0.17648887634277344, "learning_rate": 9.873781360616443e-05, "loss": 0.0101, "step": 11810 }, { "grad_norm": 0.21975623071193695, "learning_rate": 9.873411921267137e-05, "loss": 0.0093, "step": 11820 }, { "grad_norm": 0.1555415391921997, "learning_rate": 9.873041948968829e-05, "loss": 0.0117, "step": 11830 }, { "grad_norm": 0.20095409452915192, "learning_rate": 9.872671443761981e-05, "loss": 0.0101, "step": 11840 }, { "grad_norm": 0.1325235813856125, "learning_rate": 9.872300405687109e-05, "loss": 0.0093, "step": 11850 }, { "grad_norm": 0.18909801542758942, "learning_rate": 9.871928834784792e-05, "loss": 0.0104, "step": 11860 }, { "grad_norm": 0.17653009295463562, "learning_rate": 9.871556731095661e-05, "loss": 0.0099, "step": 11870 }, { "grad_norm": 0.18317903578281403, "learning_rate": 9.871184094660411e-05, "loss": 0.0113, "step": 11880 }, { "grad_norm": 0.19277289509773254, "learning_rate": 9.870810925519791e-05, "loss": 0.0096, "step": 11890 }, { "grad_norm": 0.16105026006698608, "learning_rate": 9.870437223714612e-05, "loss": 0.01, "step": 11900 }, { "grad_norm": 0.18881668150424957, "learning_rate": 9.87006298928574e-05, "loss": 0.0111, "step": 11910 }, { "grad_norm": 0.1417931765317917, "learning_rate": 9.869688222274103e-05, "loss": 0.0103, "step": 11920 }, { "grad_norm": 0.13701803982257843, "learning_rate": 9.869312922720681e-05, "loss": 0.0105, "step": 11930 }, { "grad_norm": 0.24659185111522675, "learning_rate": 9.868937090666521e-05, "loss": 0.0094, "step": 11940 }, { "grad_norm": 0.19053088128566742, "learning_rate": 9.86856072615272e-05, "loss": 0.0095, "step": 11950 }, { "grad_norm": 0.10063048452138901, "learning_rate": 9.868183829220438e-05, "loss": 0.0085, "step": 11960 }, { "grad_norm": 0.14500710368156433, "learning_rate": 9.867806399910893e-05, "loss": 0.0098, "step": 11970 }, { "grad_norm": 0.26124176383018494, "learning_rate": 9.867428438265356e-05, "loss": 0.0107, "step": 11980 }, { "grad_norm": 0.27084094285964966, "learning_rate": 9.867049944325165e-05, "loss": 0.0116, "step": 11990 }, { "grad_norm": 0.21344169974327087, "learning_rate": 9.86667091813171e-05, "loss": 0.0115, "step": 12000 }, { "grad_norm": 0.24062353372573853, "learning_rate": 9.866291359726438e-05, "loss": 0.0113, "step": 12010 }, { "grad_norm": 0.1294729858636856, "learning_rate": 9.865911269150861e-05, "loss": 0.0098, "step": 12020 }, { "grad_norm": 0.13545118272304535, "learning_rate": 9.865530646446544e-05, "loss": 0.0117, "step": 12030 }, { "grad_norm": 0.19611969590187073, "learning_rate": 9.86514949165511e-05, "loss": 0.0114, "step": 12040 }, { "grad_norm": 0.14719629287719727, "learning_rate": 9.864767804818243e-05, "loss": 0.0109, "step": 12050 }, { "grad_norm": 0.11917715519666672, "learning_rate": 9.86438558597768e-05, "loss": 0.0105, "step": 12060 }, { "grad_norm": 0.129703551530838, "learning_rate": 9.864002835175225e-05, "loss": 0.0096, "step": 12070 }, { "grad_norm": 0.22231905162334442, "learning_rate": 9.863619552452734e-05, "loss": 0.0117, "step": 12080 }, { "grad_norm": 0.12463143467903137, "learning_rate": 9.863235737852119e-05, "loss": 0.0104, "step": 12090 }, { "grad_norm": 0.17414261400699615, "learning_rate": 9.862851391415356e-05, "loss": 0.0115, "step": 12100 }, { "grad_norm": 0.21048282086849213, "learning_rate": 9.862466513184477e-05, "loss": 0.0113, "step": 12110 }, { "grad_norm": 0.1541387438774109, "learning_rate": 9.86208110320157e-05, "loss": 0.0097, "step": 12120 }, { "grad_norm": 0.18316125869750977, "learning_rate": 9.861695161508784e-05, "loss": 0.0104, "step": 12130 }, { "grad_norm": 0.14488978683948517, "learning_rate": 9.861308688148324e-05, "loss": 0.0122, "step": 12140 }, { "grad_norm": 0.1834188550710678, "learning_rate": 9.860921683162455e-05, "loss": 0.0132, "step": 12150 }, { "grad_norm": 0.1707218885421753, "learning_rate": 9.860534146593499e-05, "loss": 0.0128, "step": 12160 }, { "grad_norm": 0.22150161862373352, "learning_rate": 9.860146078483836e-05, "loss": 0.0111, "step": 12170 }, { "grad_norm": 0.18248042464256287, "learning_rate": 9.859757478875905e-05, "loss": 0.0116, "step": 12180 }, { "grad_norm": 0.1556466966867447, "learning_rate": 9.859368347812204e-05, "loss": 0.0095, "step": 12190 }, { "grad_norm": 0.15753814578056335, "learning_rate": 9.858978685335285e-05, "loss": 0.0113, "step": 12200 }, { "grad_norm": 0.21182513236999512, "learning_rate": 9.858588491487763e-05, "loss": 0.012, "step": 12210 }, { "grad_norm": 0.17038387060165405, "learning_rate": 9.858197766312308e-05, "loss": 0.012, "step": 12220 }, { "grad_norm": 0.18920394778251648, "learning_rate": 9.857806509851649e-05, "loss": 0.0113, "step": 12230 }, { "grad_norm": 0.17151132225990295, "learning_rate": 9.857414722148574e-05, "loss": 0.0103, "step": 12240 }, { "grad_norm": 0.20143774151802063, "learning_rate": 9.857022403245928e-05, "loss": 0.0098, "step": 12250 }, { "grad_norm": 0.22745758295059204, "learning_rate": 9.856629553186615e-05, "loss": 0.0113, "step": 12260 }, { "grad_norm": 0.17832061648368835, "learning_rate": 9.856236172013595e-05, "loss": 0.0101, "step": 12270 }, { "grad_norm": 0.17095746099948883, "learning_rate": 9.85584225976989e-05, "loss": 0.0107, "step": 12280 }, { "grad_norm": 0.15953487157821655, "learning_rate": 9.855447816498575e-05, "loss": 0.0112, "step": 12290 }, { "grad_norm": 0.17937089502811432, "learning_rate": 9.855052842242787e-05, "loss": 0.0097, "step": 12300 }, { "grad_norm": 0.16301988065242767, "learning_rate": 9.85465733704572e-05, "loss": 0.0099, "step": 12310 }, { "grad_norm": 0.1494085192680359, "learning_rate": 9.854261300950624e-05, "loss": 0.0096, "step": 12320 }, { "grad_norm": 0.13195474445819855, "learning_rate": 9.853864734000813e-05, "loss": 0.0102, "step": 12330 }, { "grad_norm": 0.17785821855068207, "learning_rate": 9.85346763623965e-05, "loss": 0.0085, "step": 12340 }, { "grad_norm": 0.18992233276367188, "learning_rate": 9.853070007710564e-05, "loss": 0.0093, "step": 12350 }, { "grad_norm": 0.13703589141368866, "learning_rate": 9.85267184845704e-05, "loss": 0.0101, "step": 12360 }, { "grad_norm": 0.14515064656734467, "learning_rate": 9.852273158522616e-05, "loss": 0.0097, "step": 12370 }, { "grad_norm": 0.23815323412418365, "learning_rate": 9.851873937950896e-05, "loss": 0.0108, "step": 12380 }, { "grad_norm": 0.16983802616596222, "learning_rate": 9.851474186785537e-05, "loss": 0.0116, "step": 12390 }, { "grad_norm": 0.18209536373615265, "learning_rate": 9.851073905070254e-05, "loss": 0.0105, "step": 12400 }, { "grad_norm": 0.15328504145145416, "learning_rate": 9.850673092848824e-05, "loss": 0.011, "step": 12410 }, { "grad_norm": 0.18319299817085266, "learning_rate": 9.850271750165077e-05, "loss": 0.0105, "step": 12420 }, { "grad_norm": 0.17080922424793243, "learning_rate": 9.849869877062902e-05, "loss": 0.0111, "step": 12430 }, { "grad_norm": 0.19778315722942352, "learning_rate": 9.849467473586252e-05, "loss": 0.0113, "step": 12440 }, { "grad_norm": 0.14647333323955536, "learning_rate": 9.849064539779127e-05, "loss": 0.0102, "step": 12450 }, { "grad_norm": 0.12048648297786713, "learning_rate": 9.848661075685594e-05, "loss": 0.0105, "step": 12460 }, { "grad_norm": 0.1787007451057434, "learning_rate": 9.848257081349778e-05, "loss": 0.0107, "step": 12470 }, { "grad_norm": 0.23585790395736694, "learning_rate": 9.847852556815856e-05, "loss": 0.01, "step": 12480 }, { "grad_norm": 0.16512954235076904, "learning_rate": 9.847447502128067e-05, "loss": 0.0101, "step": 12490 }, { "grad_norm": 0.21901176869869232, "learning_rate": 9.847041917330708e-05, "loss": 0.0113, "step": 12500 }, { "grad_norm": 0.16278110444545746, "learning_rate": 9.846635802468132e-05, "loss": 0.0091, "step": 12510 }, { "grad_norm": 0.1363745629787445, "learning_rate": 9.84622915758475e-05, "loss": 0.0098, "step": 12520 }, { "grad_norm": 0.18625712394714355, "learning_rate": 9.845821982725034e-05, "loss": 0.0091, "step": 12530 }, { "grad_norm": 0.1397726982831955, "learning_rate": 9.845414277933514e-05, "loss": 0.0103, "step": 12540 }, { "grad_norm": 0.18993878364562988, "learning_rate": 9.845006043254771e-05, "loss": 0.0093, "step": 12550 }, { "grad_norm": 0.16614486277103424, "learning_rate": 9.844597278733451e-05, "loss": 0.0102, "step": 12560 }, { "grad_norm": 0.13953807950019836, "learning_rate": 9.844187984414259e-05, "loss": 0.0102, "step": 12570 }, { "grad_norm": 0.20551393926143646, "learning_rate": 9.84377816034195e-05, "loss": 0.0097, "step": 12580 }, { "grad_norm": 0.2014838606119156, "learning_rate": 9.843367806561345e-05, "loss": 0.0097, "step": 12590 }, { "grad_norm": 0.2050715833902359, "learning_rate": 9.842956923117317e-05, "loss": 0.0095, "step": 12600 }, { "grad_norm": 0.1551624983549118, "learning_rate": 9.842545510054802e-05, "loss": 0.0107, "step": 12610 }, { "grad_norm": 0.1767222136259079, "learning_rate": 9.842133567418792e-05, "loss": 0.01, "step": 12620 }, { "grad_norm": 0.17775392532348633, "learning_rate": 9.841721095254333e-05, "loss": 0.0099, "step": 12630 }, { "grad_norm": 0.15825733542442322, "learning_rate": 9.841308093606537e-05, "loss": 0.0102, "step": 12640 }, { "grad_norm": 0.1475563794374466, "learning_rate": 9.840894562520565e-05, "loss": 0.0098, "step": 12650 }, { "grad_norm": 0.16533780097961426, "learning_rate": 9.840480502041642e-05, "loss": 0.0095, "step": 12660 }, { "grad_norm": 0.1274847835302353, "learning_rate": 9.840065912215049e-05, "loss": 0.0088, "step": 12670 }, { "grad_norm": 0.14496341347694397, "learning_rate": 9.839650793086124e-05, "loss": 0.0098, "step": 12680 }, { "grad_norm": 0.12391426414251328, "learning_rate": 9.839235144700265e-05, "loss": 0.0092, "step": 12690 }, { "grad_norm": 0.12241671979427338, "learning_rate": 9.838818967102926e-05, "loss": 0.0097, "step": 12700 }, { "grad_norm": 0.15325984358787537, "learning_rate": 9.83840226033962e-05, "loss": 0.0101, "step": 12710 }, { "grad_norm": 0.18736632168293, "learning_rate": 9.837985024455918e-05, "loss": 0.0096, "step": 12720 }, { "grad_norm": 0.1901477426290512, "learning_rate": 9.837567259497447e-05, "loss": 0.0104, "step": 12730 }, { "grad_norm": 0.15806244313716888, "learning_rate": 9.837148965509894e-05, "loss": 0.0104, "step": 12740 }, { "grad_norm": 0.1355791538953781, "learning_rate": 9.836730142539001e-05, "loss": 0.0098, "step": 12750 }, { "grad_norm": 0.18094614148139954, "learning_rate": 9.836310790630574e-05, "loss": 0.0087, "step": 12760 }, { "grad_norm": 0.16950581967830658, "learning_rate": 9.83589090983047e-05, "loss": 0.0095, "step": 12770 }, { "grad_norm": 0.17193342745304108, "learning_rate": 9.835470500184605e-05, "loss": 0.0097, "step": 12780 }, { "grad_norm": 0.13370144367218018, "learning_rate": 9.835049561738957e-05, "loss": 0.0095, "step": 12790 }, { "grad_norm": 0.12425830960273743, "learning_rate": 9.834628094539558e-05, "loss": 0.0104, "step": 12800 }, { "grad_norm": 0.11738969385623932, "learning_rate": 9.834206098632499e-05, "loss": 0.0081, "step": 12810 }, { "grad_norm": 0.20201066136360168, "learning_rate": 9.833783574063931e-05, "loss": 0.0092, "step": 12820 }, { "grad_norm": 0.18274766206741333, "learning_rate": 9.833360520880058e-05, "loss": 0.0085, "step": 12830 }, { "grad_norm": 0.1387845277786255, "learning_rate": 9.832936939127144e-05, "loss": 0.0095, "step": 12840 }, { "grad_norm": 0.1257607787847519, "learning_rate": 9.832512828851515e-05, "loss": 0.01, "step": 12850 }, { "grad_norm": 0.18539845943450928, "learning_rate": 9.832088190099546e-05, "loss": 0.0099, "step": 12860 }, { "grad_norm": 0.17750316858291626, "learning_rate": 9.831663022917679e-05, "loss": 0.0097, "step": 12870 }, { "grad_norm": 0.17061465978622437, "learning_rate": 9.831237327352407e-05, "loss": 0.0101, "step": 12880 }, { "grad_norm": 0.1341581493616104, "learning_rate": 9.830811103450286e-05, "loss": 0.0095, "step": 12890 }, { "grad_norm": 0.12130798399448395, "learning_rate": 9.830384351257924e-05, "loss": 0.0105, "step": 12900 }, { "grad_norm": 0.23895344138145447, "learning_rate": 9.829957070821993e-05, "loss": 0.0088, "step": 12910 }, { "grad_norm": 0.1477252095937729, "learning_rate": 9.829529262189218e-05, "loss": 0.0095, "step": 12920 }, { "grad_norm": 0.17101863026618958, "learning_rate": 9.829100925406385e-05, "loss": 0.009, "step": 12930 }, { "grad_norm": 0.15765994787216187, "learning_rate": 9.828672060520333e-05, "loss": 0.0097, "step": 12940 }, { "grad_norm": 0.1265387386083603, "learning_rate": 9.828242667577966e-05, "loss": 0.0103, "step": 12950 }, { "grad_norm": 0.1458120048046112, "learning_rate": 9.82781274662624e-05, "loss": 0.0087, "step": 12960 }, { "grad_norm": 0.2058131843805313, "learning_rate": 9.82738229771217e-05, "loss": 0.0097, "step": 12970 }, { "grad_norm": 0.1504838764667511, "learning_rate": 9.826951320882829e-05, "loss": 0.0088, "step": 12980 }, { "grad_norm": 0.1702994406223297, "learning_rate": 9.826519816185351e-05, "loss": 0.0102, "step": 12990 }, { "grad_norm": 0.21548044681549072, "learning_rate": 9.826087783666921e-05, "loss": 0.0101, "step": 13000 }, { "grad_norm": 0.18080350756645203, "learning_rate": 9.825655223374787e-05, "loss": 0.0103, "step": 13010 }, { "grad_norm": 0.21216392517089844, "learning_rate": 9.825222135356253e-05, "loss": 0.0105, "step": 13020 }, { "grad_norm": 0.141102135181427, "learning_rate": 9.82478851965868e-05, "loss": 0.0103, "step": 13030 }, { "grad_norm": 0.1824631243944168, "learning_rate": 9.82435437632949e-05, "loss": 0.0088, "step": 13040 }, { "grad_norm": 0.1304032951593399, "learning_rate": 9.823919705416158e-05, "loss": 0.0085, "step": 13050 }, { "grad_norm": 0.19489416480064392, "learning_rate": 9.82348450696622e-05, "loss": 0.0106, "step": 13060 }, { "grad_norm": 0.17590203881263733, "learning_rate": 9.823048781027268e-05, "loss": 0.0091, "step": 13070 }, { "grad_norm": 0.10798311233520508, "learning_rate": 9.822612527646953e-05, "loss": 0.0094, "step": 13080 }, { "grad_norm": 0.19381070137023926, "learning_rate": 9.822175746872984e-05, "loss": 0.009, "step": 13090 }, { "grad_norm": 0.2237972617149353, "learning_rate": 9.821738438753123e-05, "loss": 0.0094, "step": 13100 }, { "grad_norm": 0.12180545181035995, "learning_rate": 9.821300603335196e-05, "loss": 0.0086, "step": 13110 }, { "grad_norm": 0.1787496954202652, "learning_rate": 9.820862240667085e-05, "loss": 0.0094, "step": 13120 }, { "grad_norm": 0.16212508082389832, "learning_rate": 9.820423350796726e-05, "loss": 0.0113, "step": 13130 }, { "grad_norm": 0.15235988795757294, "learning_rate": 9.819983933772118e-05, "loss": 0.0108, "step": 13140 }, { "grad_norm": 0.1530337631702423, "learning_rate": 9.819543989641314e-05, "loss": 0.009, "step": 13150 }, { "grad_norm": 0.172309011220932, "learning_rate": 9.819103518452423e-05, "loss": 0.0113, "step": 13160 }, { "grad_norm": 0.13816934823989868, "learning_rate": 9.818662520253618e-05, "loss": 0.0122, "step": 13170 }, { "grad_norm": 0.14851471781730652, "learning_rate": 9.818220995093126e-05, "loss": 0.0096, "step": 13180 }, { "grad_norm": 0.16906572878360748, "learning_rate": 9.817778943019228e-05, "loss": 0.0105, "step": 13190 }, { "grad_norm": 0.21002182364463806, "learning_rate": 9.81733636408027e-05, "loss": 0.013, "step": 13200 }, { "grad_norm": 0.2352297008037567, "learning_rate": 9.816893258324649e-05, "loss": 0.0108, "step": 13210 }, { "grad_norm": 0.16701140999794006, "learning_rate": 9.816449625800823e-05, "loss": 0.0111, "step": 13220 }, { "grad_norm": 0.17588646709918976, "learning_rate": 9.816005466557308e-05, "loss": 0.0091, "step": 13230 }, { "grad_norm": 0.2361190766096115, "learning_rate": 9.815560780642674e-05, "loss": 0.0109, "step": 13240 }, { "grad_norm": 0.15047210454940796, "learning_rate": 9.815115568105555e-05, "loss": 0.0123, "step": 13250 }, { "grad_norm": 0.11145389080047607, "learning_rate": 9.814669828994638e-05, "loss": 0.0107, "step": 13260 }, { "grad_norm": 0.1721149981021881, "learning_rate": 9.814223563358665e-05, "loss": 0.0119, "step": 13270 }, { "grad_norm": 0.14389754831790924, "learning_rate": 9.813776771246443e-05, "loss": 0.0107, "step": 13280 }, { "grad_norm": 0.17901675403118134, "learning_rate": 9.813329452706829e-05, "loss": 0.0118, "step": 13290 }, { "grad_norm": 0.13576966524124146, "learning_rate": 9.812881607788744e-05, "loss": 0.0104, "step": 13300 }, { "grad_norm": 0.12194713950157166, "learning_rate": 9.812433236541163e-05, "loss": 0.0093, "step": 13310 }, { "grad_norm": 0.12571930885314941, "learning_rate": 9.811984339013116e-05, "loss": 0.0111, "step": 13320 }, { "grad_norm": 0.16038070619106293, "learning_rate": 9.811534915253698e-05, "loss": 0.0106, "step": 13330 }, { "grad_norm": 0.14913666248321533, "learning_rate": 9.811084965312056e-05, "loss": 0.0099, "step": 13340 }, { "grad_norm": 0.12547247111797333, "learning_rate": 9.810634489237396e-05, "loss": 0.0098, "step": 13350 }, { "grad_norm": 0.1816750019788742, "learning_rate": 9.81018348707898e-05, "loss": 0.0098, "step": 13360 }, { "grad_norm": 0.14139382541179657, "learning_rate": 9.809731958886131e-05, "loss": 0.0114, "step": 13370 }, { "grad_norm": 0.163030743598938, "learning_rate": 9.809279904708224e-05, "loss": 0.01, "step": 13380 }, { "grad_norm": 0.1648268848657608, "learning_rate": 9.808827324594699e-05, "loss": 0.0105, "step": 13390 }, { "grad_norm": 0.18498437106609344, "learning_rate": 9.808374218595046e-05, "loss": 0.0116, "step": 13400 }, { "grad_norm": 0.11605363339185715, "learning_rate": 9.80792058675882e-05, "loss": 0.0091, "step": 13410 }, { "grad_norm": 0.15358592569828033, "learning_rate": 9.807466429135627e-05, "loss": 0.0095, "step": 13420 }, { "grad_norm": 0.12378217279911041, "learning_rate": 9.807011745775132e-05, "loss": 0.0102, "step": 13430 }, { "grad_norm": 0.1646447628736496, "learning_rate": 9.806556536727061e-05, "loss": 0.0098, "step": 13440 }, { "grad_norm": 0.16066132485866547, "learning_rate": 9.806100802041193e-05, "loss": 0.0094, "step": 13450 }, { "grad_norm": 0.11823882162570953, "learning_rate": 9.805644541767368e-05, "loss": 0.0092, "step": 13460 }, { "grad_norm": 0.1925317943096161, "learning_rate": 9.805187755955478e-05, "loss": 0.0091, "step": 13470 }, { "grad_norm": 0.1613841950893402, "learning_rate": 9.804730444655483e-05, "loss": 0.0089, "step": 13480 }, { "grad_norm": 0.168819859623909, "learning_rate": 9.804272607917388e-05, "loss": 0.0093, "step": 13490 }, { "grad_norm": 0.18512262403964996, "learning_rate": 9.803814245791265e-05, "loss": 0.01, "step": 13500 }, { "grad_norm": 0.1908881664276123, "learning_rate": 9.803355358327239e-05, "loss": 0.0092, "step": 13510 }, { "grad_norm": 0.19055049121379852, "learning_rate": 9.802895945575492e-05, "loss": 0.0105, "step": 13520 }, { "grad_norm": 0.19177740812301636, "learning_rate": 9.802436007586266e-05, "loss": 0.0095, "step": 13530 }, { "grad_norm": 0.1673489362001419, "learning_rate": 9.801975544409858e-05, "loss": 0.0102, "step": 13540 }, { "grad_norm": 0.14161883294582367, "learning_rate": 9.801514556096625e-05, "loss": 0.0089, "step": 13550 }, { "grad_norm": 0.16590163111686707, "learning_rate": 9.801053042696977e-05, "loss": 0.0103, "step": 13560 }, { "grad_norm": 0.13264687359333038, "learning_rate": 9.800591004261388e-05, "loss": 0.0111, "step": 13570 }, { "grad_norm": 0.1594018191099167, "learning_rate": 9.800128440840385e-05, "loss": 0.0097, "step": 13580 }, { "grad_norm": 0.23435081541538239, "learning_rate": 9.799665352484552e-05, "loss": 0.0117, "step": 13590 }, { "grad_norm": 0.16799674928188324, "learning_rate": 9.799201739244532e-05, "loss": 0.0105, "step": 13600 }, { "grad_norm": 0.15113669633865356, "learning_rate": 9.798737601171025e-05, "loss": 0.0102, "step": 13610 }, { "grad_norm": 0.13330020010471344, "learning_rate": 9.79827293831479e-05, "loss": 0.0095, "step": 13620 }, { "grad_norm": 0.17103594541549683, "learning_rate": 9.797807750726638e-05, "loss": 0.0104, "step": 13630 }, { "grad_norm": 0.1510397493839264, "learning_rate": 9.797342038457446e-05, "loss": 0.0102, "step": 13640 }, { "grad_norm": 0.17929036915302277, "learning_rate": 9.796875801558141e-05, "loss": 0.0105, "step": 13650 }, { "grad_norm": 0.16108927130699158, "learning_rate": 9.79640904007971e-05, "loss": 0.0099, "step": 13660 }, { "grad_norm": 0.16933351755142212, "learning_rate": 9.795941754073199e-05, "loss": 0.0103, "step": 13670 }, { "grad_norm": 0.2345196008682251, "learning_rate": 9.795473943589705e-05, "loss": 0.0096, "step": 13680 }, { "grad_norm": 0.13000838458538055, "learning_rate": 9.795005608680394e-05, "loss": 0.009, "step": 13690 }, { "grad_norm": 0.13204669952392578, "learning_rate": 9.794536749396477e-05, "loss": 0.0092, "step": 13700 }, { "grad_norm": 0.15213564038276672, "learning_rate": 9.79406736578923e-05, "loss": 0.0092, "step": 13710 }, { "grad_norm": 0.14138229191303253, "learning_rate": 9.793597457909984e-05, "loss": 0.0101, "step": 13720 }, { "grad_norm": 0.17613448202610016, "learning_rate": 9.793127025810127e-05, "loss": 0.0103, "step": 13730 }, { "grad_norm": 0.19608139991760254, "learning_rate": 9.792656069541104e-05, "loss": 0.0091, "step": 13740 }, { "grad_norm": 0.12538453936576843, "learning_rate": 9.79218458915442e-05, "loss": 0.009, "step": 13750 }, { "grad_norm": 0.16326072812080383, "learning_rate": 9.791712584701634e-05, "loss": 0.0099, "step": 13760 }, { "grad_norm": 0.18242037296295166, "learning_rate": 9.791240056234364e-05, "loss": 0.0087, "step": 13770 }, { "grad_norm": 0.14457343518733978, "learning_rate": 9.790767003804283e-05, "loss": 0.0095, "step": 13780 }, { "grad_norm": 0.11081008613109589, "learning_rate": 9.790293427463126e-05, "loss": 0.0087, "step": 13790 }, { "grad_norm": 0.13121449947357178, "learning_rate": 9.789819327262684e-05, "loss": 0.0107, "step": 13800 }, { "grad_norm": 0.26572495698928833, "learning_rate": 9.7893447032548e-05, "loss": 0.012, "step": 13810 }, { "grad_norm": 0.1735915243625641, "learning_rate": 9.78886955549138e-05, "loss": 0.0125, "step": 13820 }, { "grad_norm": 0.2050962895154953, "learning_rate": 9.788393884024387e-05, "loss": 0.011, "step": 13830 }, { "grad_norm": 0.14995962381362915, "learning_rate": 9.787917688905836e-05, "loss": 0.0102, "step": 13840 }, { "grad_norm": 0.17690148949623108, "learning_rate": 9.787440970187807e-05, "loss": 0.0117, "step": 13850 }, { "grad_norm": 0.17090697586536407, "learning_rate": 9.786963727922429e-05, "loss": 0.0133, "step": 13860 }, { "grad_norm": 0.22454339265823364, "learning_rate": 9.786485962161897e-05, "loss": 0.0128, "step": 13870 }, { "grad_norm": 0.22768168151378632, "learning_rate": 9.786007672958455e-05, "loss": 0.0127, "step": 13880 }, { "grad_norm": 0.1582845151424408, "learning_rate": 9.78552886036441e-05, "loss": 0.0115, "step": 13890 }, { "grad_norm": 0.2090630680322647, "learning_rate": 9.785049524432124e-05, "loss": 0.0172, "step": 13900 }, { "grad_norm": 0.35438272356987, "learning_rate": 9.784569665214016e-05, "loss": 0.0202, "step": 13910 }, { "grad_norm": 0.3897165358066559, "learning_rate": 9.784089282762563e-05, "loss": 0.018, "step": 13920 }, { "grad_norm": 0.2198696881532669, "learning_rate": 9.7836083771303e-05, "loss": 0.0175, "step": 13930 }, { "grad_norm": 0.18584170937538147, "learning_rate": 9.783126948369817e-05, "loss": 0.0149, "step": 13940 }, { "grad_norm": 0.1782980114221573, "learning_rate": 9.78264499653376e-05, "loss": 0.0146, "step": 13950 }, { "grad_norm": 0.18526828289031982, "learning_rate": 9.782162521674838e-05, "loss": 0.0147, "step": 13960 }, { "grad_norm": 0.26658374071121216, "learning_rate": 9.781679523845812e-05, "loss": 0.0135, "step": 13970 }, { "grad_norm": 0.15909349918365479, "learning_rate": 9.781196003099502e-05, "loss": 0.0134, "step": 13980 }, { "grad_norm": 0.17502310872077942, "learning_rate": 9.780711959488786e-05, "loss": 0.0125, "step": 13990 }, { "grad_norm": 0.250114381313324, "learning_rate": 9.780227393066599e-05, "loss": 0.0135, "step": 14000 }, { "grad_norm": 0.13250787556171417, "learning_rate": 9.77974230388593e-05, "loss": 0.0125, "step": 14010 }, { "grad_norm": 0.19933801889419556, "learning_rate": 9.779256691999829e-05, "loss": 0.0108, "step": 14020 }, { "grad_norm": 0.20301322638988495, "learning_rate": 9.778770557461403e-05, "loss": 0.011, "step": 14030 }, { "grad_norm": 0.18720121681690216, "learning_rate": 9.778283900323812e-05, "loss": 0.0098, "step": 14040 }, { "grad_norm": 0.18269823491573334, "learning_rate": 9.777796720640277e-05, "loss": 0.0101, "step": 14050 }, { "grad_norm": 0.16834691166877747, "learning_rate": 9.777309018464078e-05, "loss": 0.0121, "step": 14060 }, { "grad_norm": 0.13687504827976227, "learning_rate": 9.776820793848547e-05, "loss": 0.0117, "step": 14070 }, { "grad_norm": 0.178886279463768, "learning_rate": 9.776332046847075e-05, "loss": 0.0113, "step": 14080 }, { "grad_norm": 0.1559576541185379, "learning_rate": 9.775842777513111e-05, "loss": 0.0115, "step": 14090 }, { "grad_norm": 0.1640985757112503, "learning_rate": 9.775352985900163e-05, "loss": 0.0096, "step": 14100 }, { "grad_norm": 0.16396772861480713, "learning_rate": 9.774862672061791e-05, "loss": 0.0094, "step": 14110 }, { "grad_norm": 0.15081949532032013, "learning_rate": 9.774371836051616e-05, "loss": 0.0101, "step": 14120 }, { "grad_norm": 0.1590190976858139, "learning_rate": 9.773880477923315e-05, "loss": 0.0121, "step": 14130 }, { "grad_norm": 0.12910987436771393, "learning_rate": 9.773388597730623e-05, "loss": 0.0098, "step": 14140 }, { "grad_norm": 0.1930747926235199, "learning_rate": 9.77289619552733e-05, "loss": 0.0094, "step": 14150 }, { "grad_norm": 0.18852578103542328, "learning_rate": 9.772403271367285e-05, "loss": 0.0105, "step": 14160 }, { "grad_norm": 0.21052230894565582, "learning_rate": 9.771909825304396e-05, "loss": 0.0109, "step": 14170 }, { "grad_norm": 0.23927152156829834, "learning_rate": 9.771415857392619e-05, "loss": 0.0106, "step": 14180 }, { "grad_norm": 0.15031719207763672, "learning_rate": 9.770921367685978e-05, "loss": 0.0098, "step": 14190 }, { "grad_norm": 0.17688626050949097, "learning_rate": 9.770426356238551e-05, "loss": 0.01, "step": 14200 }, { "grad_norm": 0.12519872188568115, "learning_rate": 9.769930823104469e-05, "loss": 0.0092, "step": 14210 }, { "grad_norm": 0.12781663239002228, "learning_rate": 9.769434768337926e-05, "loss": 0.0081, "step": 14220 }, { "grad_norm": 0.10604963451623917, "learning_rate": 9.768938191993164e-05, "loss": 0.0083, "step": 14230 }, { "grad_norm": 0.17337797582149506, "learning_rate": 9.768441094124494e-05, "loss": 0.0089, "step": 14240 }, { "grad_norm": 0.18179622292518616, "learning_rate": 9.767943474786275e-05, "loss": 0.0096, "step": 14250 }, { "grad_norm": 0.11641573160886765, "learning_rate": 9.767445334032923e-05, "loss": 0.0083, "step": 14260 }, { "grad_norm": 0.1507706195116043, "learning_rate": 9.766946671918919e-05, "loss": 0.0089, "step": 14270 }, { "grad_norm": 0.1610371619462967, "learning_rate": 9.766447488498796e-05, "loss": 0.0085, "step": 14280 }, { "grad_norm": 0.12487764656543732, "learning_rate": 9.765947783827139e-05, "loss": 0.0112, "step": 14290 }, { "grad_norm": 0.12120288610458374, "learning_rate": 9.765447557958599e-05, "loss": 0.009, "step": 14300 }, { "grad_norm": 0.18515288829803467, "learning_rate": 9.764946810947879e-05, "loss": 0.0088, "step": 14310 }, { "grad_norm": 0.16903503239154816, "learning_rate": 9.764445542849738e-05, "loss": 0.009, "step": 14320 }, { "grad_norm": 0.17042836546897888, "learning_rate": 9.763943753718998e-05, "loss": 0.0084, "step": 14330 }, { "grad_norm": 0.15164689719676971, "learning_rate": 9.76344144361053e-05, "loss": 0.0089, "step": 14340 }, { "grad_norm": 0.1933702826499939, "learning_rate": 9.762938612579269e-05, "loss": 0.0092, "step": 14350 }, { "grad_norm": 0.16523882746696472, "learning_rate": 9.762435260680202e-05, "loss": 0.0088, "step": 14360 }, { "grad_norm": 0.1663448065519333, "learning_rate": 9.761931387968373e-05, "loss": 0.0081, "step": 14370 }, { "grad_norm": 0.21285012364387512, "learning_rate": 9.76142699449889e-05, "loss": 0.0101, "step": 14380 }, { "grad_norm": 0.13495787978172302, "learning_rate": 9.760922080326908e-05, "loss": 0.0095, "step": 14390 }, { "grad_norm": 0.1496119201183319, "learning_rate": 9.760416645507644e-05, "loss": 0.009, "step": 14400 }, { "grad_norm": 0.12283606082201004, "learning_rate": 9.759910690096375e-05, "loss": 0.0103, "step": 14410 }, { "grad_norm": 0.15622852742671967, "learning_rate": 9.759404214148429e-05, "loss": 0.0089, "step": 14420 }, { "grad_norm": 0.11271950602531433, "learning_rate": 9.758897217719191e-05, "loss": 0.0101, "step": 14430 }, { "grad_norm": 0.14500001072883606, "learning_rate": 9.758389700864113e-05, "loss": 0.0105, "step": 14440 }, { "grad_norm": 0.1930793970823288, "learning_rate": 9.757881663638688e-05, "loss": 0.009, "step": 14450 }, { "grad_norm": 0.15629969537258148, "learning_rate": 9.757373106098478e-05, "loss": 0.0098, "step": 14460 }, { "grad_norm": 0.12328136712312698, "learning_rate": 9.756864028299097e-05, "loss": 0.0083, "step": 14470 }, { "grad_norm": 0.16879154741764069, "learning_rate": 9.75635443029622e-05, "loss": 0.0101, "step": 14480 }, { "grad_norm": 0.13210870325565338, "learning_rate": 9.755844312145572e-05, "loss": 0.0098, "step": 14490 }, { "grad_norm": 0.17498517036437988, "learning_rate": 9.755333673902941e-05, "loss": 0.0093, "step": 14500 }, { "grad_norm": 0.15562115609645844, "learning_rate": 9.75482251562417e-05, "loss": 0.0095, "step": 14510 }, { "grad_norm": 0.12519364058971405, "learning_rate": 9.754310837365155e-05, "loss": 0.0091, "step": 14520 }, { "grad_norm": 0.12951503694057465, "learning_rate": 9.753798639181856e-05, "loss": 0.0097, "step": 14530 }, { "grad_norm": 0.1496262550354004, "learning_rate": 9.753285921130286e-05, "loss": 0.01, "step": 14540 }, { "grad_norm": 0.15929169952869415, "learning_rate": 9.752772683266512e-05, "loss": 0.0086, "step": 14550 }, { "grad_norm": 0.11689289659261703, "learning_rate": 9.752258925646665e-05, "loss": 0.0095, "step": 14560 }, { "grad_norm": 0.1881639063358307, "learning_rate": 9.751744648326926e-05, "loss": 0.0099, "step": 14570 }, { "grad_norm": 0.1771891564130783, "learning_rate": 9.751229851363536e-05, "loss": 0.0103, "step": 14580 }, { "grad_norm": 0.15451954305171967, "learning_rate": 9.750714534812793e-05, "loss": 0.011, "step": 14590 }, { "grad_norm": 0.24291524291038513, "learning_rate": 9.750198698731053e-05, "loss": 0.0113, "step": 14600 }, { "grad_norm": 0.1391625553369522, "learning_rate": 9.749682343174722e-05, "loss": 0.0112, "step": 14610 }, { "grad_norm": 0.16151639819145203, "learning_rate": 9.749165468200272e-05, "loss": 0.0117, "step": 14620 }, { "grad_norm": 0.1602884829044342, "learning_rate": 9.748648073864229e-05, "loss": 0.0097, "step": 14630 }, { "grad_norm": 0.19042545557022095, "learning_rate": 9.748130160223168e-05, "loss": 0.0113, "step": 14640 }, { "grad_norm": 0.1366593837738037, "learning_rate": 9.747611727333734e-05, "loss": 0.0105, "step": 14650 }, { "grad_norm": 0.17811763286590576, "learning_rate": 9.74709277525262e-05, "loss": 0.0105, "step": 14660 }, { "grad_norm": 0.1352313607931137, "learning_rate": 9.746573304036576e-05, "loss": 0.0101, "step": 14670 }, { "grad_norm": 0.13962538540363312, "learning_rate": 9.746053313742412e-05, "loss": 0.0106, "step": 14680 }, { "grad_norm": 0.15706521272659302, "learning_rate": 9.745532804426994e-05, "loss": 0.0086, "step": 14690 }, { "grad_norm": 0.15619374811649323, "learning_rate": 9.745011776147242e-05, "loss": 0.0094, "step": 14700 }, { "grad_norm": 0.25205039978027344, "learning_rate": 9.744490228960138e-05, "loss": 0.0094, "step": 14710 }, { "grad_norm": 0.146835058927536, "learning_rate": 9.743968162922713e-05, "loss": 0.0096, "step": 14720 }, { "grad_norm": 0.16598552465438843, "learning_rate": 9.743445578092064e-05, "loss": 0.0095, "step": 14730 }, { "grad_norm": 0.12118010967969894, "learning_rate": 9.742922474525338e-05, "loss": 0.0093, "step": 14740 }, { "grad_norm": 0.11344394087791443, "learning_rate": 9.742398852279741e-05, "loss": 0.01, "step": 14750 }, { "grad_norm": 0.13545474410057068, "learning_rate": 9.741874711412535e-05, "loss": 0.0104, "step": 14760 }, { "grad_norm": 0.17336834967136383, "learning_rate": 9.741350051981042e-05, "loss": 0.0101, "step": 14770 }, { "grad_norm": 0.18828441202640533, "learning_rate": 9.740824874042633e-05, "loss": 0.0104, "step": 14780 }, { "grad_norm": 0.09131170809268951, "learning_rate": 9.740299177654746e-05, "loss": 0.0121, "step": 14790 }, { "grad_norm": 0.20090971887111664, "learning_rate": 9.739772962874867e-05, "loss": 0.0097, "step": 14800 }, { "grad_norm": 0.19312052428722382, "learning_rate": 9.739246229760541e-05, "loss": 0.0105, "step": 14810 }, { "grad_norm": 0.1810598522424698, "learning_rate": 9.738718978369376e-05, "loss": 0.011, "step": 14820 }, { "grad_norm": 0.16586369276046753, "learning_rate": 9.738191208759025e-05, "loss": 0.0095, "step": 14830 }, { "grad_norm": 0.1328834444284439, "learning_rate": 9.73766292098721e-05, "loss": 0.0088, "step": 14840 }, { "grad_norm": 0.15557913482189178, "learning_rate": 9.737134115111699e-05, "loss": 0.0098, "step": 14850 }, { "grad_norm": 0.1076507419347763, "learning_rate": 9.736604791190323e-05, "loss": 0.0084, "step": 14860 }, { "grad_norm": 0.144586980342865, "learning_rate": 9.73607494928097e-05, "loss": 0.0081, "step": 14870 }, { "grad_norm": 0.13278982043266296, "learning_rate": 9.735544589441581e-05, "loss": 0.0096, "step": 14880 }, { "grad_norm": 0.1857997626066208, "learning_rate": 9.735013711730154e-05, "loss": 0.0103, "step": 14890 }, { "grad_norm": 0.19870349764823914, "learning_rate": 9.734482316204747e-05, "loss": 0.0124, "step": 14900 }, { "grad_norm": 0.14579765498638153, "learning_rate": 9.733950402923473e-05, "loss": 0.0115, "step": 14910 }, { "grad_norm": 0.22341366112232208, "learning_rate": 9.7334179719445e-05, "loss": 0.0114, "step": 14920 }, { "grad_norm": 0.15315556526184082, "learning_rate": 9.732885023326053e-05, "loss": 0.0129, "step": 14930 }, { "grad_norm": 0.2140553742647171, "learning_rate": 9.732351557126418e-05, "loss": 0.0121, "step": 14940 }, { "grad_norm": 0.12669210135936737, "learning_rate": 9.731817573403929e-05, "loss": 0.0117, "step": 14950 }, { "grad_norm": 0.1878410130739212, "learning_rate": 9.731283072216985e-05, "loss": 0.0108, "step": 14960 }, { "grad_norm": 0.15959353744983673, "learning_rate": 9.730748053624039e-05, "loss": 0.011, "step": 14970 }, { "grad_norm": 0.1475571244955063, "learning_rate": 9.730212517683598e-05, "loss": 0.0106, "step": 14980 }, { "grad_norm": 0.12117626518011093, "learning_rate": 9.729676464454228e-05, "loss": 0.0096, "step": 14990 }, { "grad_norm": 0.1927078813314438, "learning_rate": 9.72913989399455e-05, "loss": 0.0083, "step": 15000 }, { "grad_norm": 0.1536884307861328, "learning_rate": 9.728602806363242e-05, "loss": 0.0101, "step": 15010 }, { "grad_norm": 0.16110248863697052, "learning_rate": 9.728065201619043e-05, "loss": 0.0093, "step": 15020 }, { "grad_norm": 0.17774982750415802, "learning_rate": 9.727527079820742e-05, "loss": 0.0088, "step": 15030 }, { "grad_norm": 0.09792646020650864, "learning_rate": 9.726988441027186e-05, "loss": 0.0099, "step": 15040 }, { "grad_norm": 0.13053561747074127, "learning_rate": 9.726449285297281e-05, "loss": 0.0111, "step": 15050 }, { "grad_norm": 0.1848902404308319, "learning_rate": 9.72590961268999e-05, "loss": 0.0086, "step": 15060 }, { "grad_norm": 0.13079141080379486, "learning_rate": 9.725369423264328e-05, "loss": 0.0095, "step": 15070 }, { "grad_norm": 0.11994373053312302, "learning_rate": 9.72482871707937e-05, "loss": 0.0094, "step": 15080 }, { "grad_norm": 0.12968090176582336, "learning_rate": 9.724287494194247e-05, "loss": 0.0099, "step": 15090 }, { "grad_norm": 0.16312195360660553, "learning_rate": 9.723745754668147e-05, "loss": 0.0088, "step": 15100 }, { "grad_norm": 0.237204447388649, "learning_rate": 9.723203498560313e-05, "loss": 0.0096, "step": 15110 }, { "grad_norm": 0.22082659602165222, "learning_rate": 9.722660725930046e-05, "loss": 0.0099, "step": 15120 }, { "grad_norm": 0.1180269792675972, "learning_rate": 9.722117436836702e-05, "loss": 0.0091, "step": 15130 }, { "grad_norm": 0.18763722479343414, "learning_rate": 9.721573631339696e-05, "loss": 0.0102, "step": 15140 }, { "grad_norm": 0.13907280564308167, "learning_rate": 9.721029309498494e-05, "loss": 0.0093, "step": 15150 }, { "grad_norm": 0.155034139752388, "learning_rate": 9.720484471372627e-05, "loss": 0.0088, "step": 15160 }, { "grad_norm": 0.11994747072458267, "learning_rate": 9.719939117021673e-05, "loss": 0.0103, "step": 15170 }, { "grad_norm": 0.11956417560577393, "learning_rate": 9.719393246505275e-05, "loss": 0.0096, "step": 15180 }, { "grad_norm": 0.1359158456325531, "learning_rate": 9.718846859883128e-05, "loss": 0.0085, "step": 15190 }, { "grad_norm": 0.15035292506217957, "learning_rate": 9.718299957214982e-05, "loss": 0.0099, "step": 15200 }, { "grad_norm": 0.12015806138515472, "learning_rate": 9.717752538560646e-05, "loss": 0.0102, "step": 15210 }, { "grad_norm": 0.15236236155033112, "learning_rate": 9.717204603979986e-05, "loss": 0.0078, "step": 15220 }, { "grad_norm": 0.11112502962350845, "learning_rate": 9.716656153532922e-05, "loss": 0.0093, "step": 15230 }, { "grad_norm": 0.1340203732252121, "learning_rate": 9.716107187279434e-05, "loss": 0.0084, "step": 15240 }, { "grad_norm": 0.11084307730197906, "learning_rate": 9.715557705279555e-05, "loss": 0.0081, "step": 15250 }, { "grad_norm": 0.10279291868209839, "learning_rate": 9.715007707593372e-05, "loss": 0.0086, "step": 15260 }, { "grad_norm": 0.17452166974544525, "learning_rate": 9.714457194281036e-05, "loss": 0.0087, "step": 15270 }, { "grad_norm": 0.18772412836551666, "learning_rate": 9.713906165402751e-05, "loss": 0.0091, "step": 15280 }, { "grad_norm": 0.1460839956998825, "learning_rate": 9.713354621018774e-05, "loss": 0.0082, "step": 15290 }, { "grad_norm": 0.13885924220085144, "learning_rate": 9.712802561189422e-05, "loss": 0.0089, "step": 15300 }, { "grad_norm": 0.14832627773284912, "learning_rate": 9.712249985975069e-05, "loss": 0.0088, "step": 15310 }, { "grad_norm": 0.12966632843017578, "learning_rate": 9.71169689543614e-05, "loss": 0.0081, "step": 15320 }, { "grad_norm": 0.16043566167354584, "learning_rate": 9.711143289633123e-05, "loss": 0.0072, "step": 15330 }, { "grad_norm": 0.23976311087608337, "learning_rate": 9.710589168626561e-05, "loss": 0.0093, "step": 15340 }, { "grad_norm": 0.1313503086566925, "learning_rate": 9.710034532477048e-05, "loss": 0.0078, "step": 15350 }, { "grad_norm": 0.168138325214386, "learning_rate": 9.709479381245239e-05, "loss": 0.0088, "step": 15360 }, { "grad_norm": 0.1469086855649948, "learning_rate": 9.708923714991847e-05, "loss": 0.009, "step": 15370 }, { "grad_norm": 0.16199912130832672, "learning_rate": 9.708367533777638e-05, "loss": 0.0085, "step": 15380 }, { "grad_norm": 0.14881719648838043, "learning_rate": 9.707810837663431e-05, "loss": 0.0077, "step": 15390 }, { "grad_norm": 0.15654951333999634, "learning_rate": 9.707253626710113e-05, "loss": 0.0076, "step": 15400 }, { "grad_norm": 0.16288532316684723, "learning_rate": 9.706695900978613e-05, "loss": 0.008, "step": 15410 }, { "grad_norm": 0.09667052328586578, "learning_rate": 9.706137660529926e-05, "loss": 0.0081, "step": 15420 }, { "grad_norm": 0.12815465033054352, "learning_rate": 9.705578905425101e-05, "loss": 0.0076, "step": 15430 }, { "grad_norm": 0.1594310998916626, "learning_rate": 9.705019635725241e-05, "loss": 0.0075, "step": 15440 }, { "grad_norm": 0.19017194211483002, "learning_rate": 9.704459851491508e-05, "loss": 0.0102, "step": 15450 }, { "grad_norm": 0.14119230210781097, "learning_rate": 9.703899552785118e-05, "loss": 0.0077, "step": 15460 }, { "grad_norm": 0.11176075786352158, "learning_rate": 9.703338739667346e-05, "loss": 0.0085, "step": 15470 }, { "grad_norm": 0.16187205910682678, "learning_rate": 9.70277741219952e-05, "loss": 0.0104, "step": 15480 }, { "grad_norm": 0.1631164401769638, "learning_rate": 9.702215570443027e-05, "loss": 0.008, "step": 15490 }, { "grad_norm": 0.18849453330039978, "learning_rate": 9.701653214459309e-05, "loss": 0.0094, "step": 15500 }, { "grad_norm": 0.18478335440158844, "learning_rate": 9.701090344309865e-05, "loss": 0.0094, "step": 15510 }, { "grad_norm": 0.16853518784046173, "learning_rate": 9.700526960056247e-05, "loss": 0.0086, "step": 15520 }, { "grad_norm": 0.12473087757825851, "learning_rate": 9.699963061760068e-05, "loss": 0.0091, "step": 15530 }, { "grad_norm": 0.17572163045406342, "learning_rate": 9.699398649482997e-05, "loss": 0.0084, "step": 15540 }, { "grad_norm": 0.1257840096950531, "learning_rate": 9.698833723286753e-05, "loss": 0.0088, "step": 15550 }, { "grad_norm": 0.16610826551914215, "learning_rate": 9.698268283233118e-05, "loss": 0.0094, "step": 15560 }, { "grad_norm": 0.16215738654136658, "learning_rate": 9.697702329383929e-05, "loss": 0.0084, "step": 15570 }, { "grad_norm": 0.1206171065568924, "learning_rate": 9.697135861801074e-05, "loss": 0.0083, "step": 15580 }, { "grad_norm": 0.1089465469121933, "learning_rate": 9.696568880546505e-05, "loss": 0.008, "step": 15590 }, { "grad_norm": 0.15214058756828308, "learning_rate": 9.696001385682223e-05, "loss": 0.0087, "step": 15600 }, { "grad_norm": 0.13754406571388245, "learning_rate": 9.695433377270291e-05, "loss": 0.0082, "step": 15610 }, { "grad_norm": 0.1623402237892151, "learning_rate": 9.694864855372824e-05, "loss": 0.0073, "step": 15620 }, { "grad_norm": 0.14267703890800476, "learning_rate": 9.694295820051995e-05, "loss": 0.0086, "step": 15630 }, { "grad_norm": 0.13015910983085632, "learning_rate": 9.693726271370032e-05, "loss": 0.0083, "step": 15640 }, { "grad_norm": 0.11317679286003113, "learning_rate": 9.693156209389221e-05, "loss": 0.0082, "step": 15650 }, { "grad_norm": 0.11205277591943741, "learning_rate": 9.692585634171905e-05, "loss": 0.0082, "step": 15660 }, { "grad_norm": 0.13214315474033356, "learning_rate": 9.692014545780476e-05, "loss": 0.0094, "step": 15670 }, { "grad_norm": 0.1307613104581833, "learning_rate": 9.691442944277393e-05, "loss": 0.0087, "step": 15680 }, { "grad_norm": 0.1682303547859192, "learning_rate": 9.690870829725162e-05, "loss": 0.0095, "step": 15690 }, { "grad_norm": 0.12976764142513275, "learning_rate": 9.69029820218635e-05, "loss": 0.0095, "step": 15700 }, { "grad_norm": 0.18410001695156097, "learning_rate": 9.689725061723579e-05, "loss": 0.0091, "step": 15710 }, { "grad_norm": 0.12066303193569183, "learning_rate": 9.689151408399527e-05, "loss": 0.0086, "step": 15720 }, { "grad_norm": 0.10504818707704544, "learning_rate": 9.688577242276924e-05, "loss": 0.0081, "step": 15730 }, { "grad_norm": 0.17436476051807404, "learning_rate": 9.688002563418566e-05, "loss": 0.0084, "step": 15740 }, { "grad_norm": 0.11943629384040833, "learning_rate": 9.687427371887293e-05, "loss": 0.0079, "step": 15750 }, { "grad_norm": 0.1460217535495758, "learning_rate": 9.686851667746012e-05, "loss": 0.0085, "step": 15760 }, { "grad_norm": 0.13753214478492737, "learning_rate": 9.686275451057677e-05, "loss": 0.0086, "step": 15770 }, { "grad_norm": 0.10622606426477432, "learning_rate": 9.685698721885308e-05, "loss": 0.0085, "step": 15780 }, { "grad_norm": 0.09835547208786011, "learning_rate": 9.68512148029197e-05, "loss": 0.0081, "step": 15790 }, { "grad_norm": 0.16895493865013123, "learning_rate": 9.684543726340791e-05, "loss": 0.008, "step": 15800 }, { "grad_norm": 0.19696831703186035, "learning_rate": 9.683965460094952e-05, "loss": 0.0086, "step": 15810 }, { "grad_norm": 0.14765746891498566, "learning_rate": 9.683386681617694e-05, "loss": 0.0088, "step": 15820 }, { "grad_norm": 0.11167865246534348, "learning_rate": 9.68280739097231e-05, "loss": 0.0085, "step": 15830 }, { "grad_norm": 0.18117870390415192, "learning_rate": 9.682227588222148e-05, "loss": 0.0087, "step": 15840 }, { "grad_norm": 0.18765728175640106, "learning_rate": 9.681647273430618e-05, "loss": 0.0089, "step": 15850 }, { "grad_norm": 0.21520328521728516, "learning_rate": 9.681066446661182e-05, "loss": 0.0088, "step": 15860 }, { "grad_norm": 0.1160750761628151, "learning_rate": 9.680485107977357e-05, "loss": 0.0091, "step": 15870 }, { "grad_norm": 0.14784879982471466, "learning_rate": 9.679903257442716e-05, "loss": 0.0082, "step": 15880 }, { "grad_norm": 0.11901433020830154, "learning_rate": 9.679320895120891e-05, "loss": 0.0096, "step": 15890 }, { "grad_norm": 0.19426314532756805, "learning_rate": 9.67873802107557e-05, "loss": 0.0099, "step": 15900 }, { "grad_norm": 0.14233087003231049, "learning_rate": 9.67815463537049e-05, "loss": 0.0091, "step": 15910 }, { "grad_norm": 0.2760486900806427, "learning_rate": 9.677570738069457e-05, "loss": 0.0091, "step": 15920 }, { "grad_norm": 0.13759692013263702, "learning_rate": 9.676986329236318e-05, "loss": 0.0104, "step": 15930 }, { "grad_norm": 0.1962614506483078, "learning_rate": 9.676401408934987e-05, "loss": 0.0095, "step": 15940 }, { "grad_norm": 0.2106517106294632, "learning_rate": 9.675815977229428e-05, "loss": 0.0107, "step": 15950 }, { "grad_norm": 0.13647185266017914, "learning_rate": 9.675230034183664e-05, "loss": 0.0086, "step": 15960 }, { "grad_norm": 0.14619435369968414, "learning_rate": 9.674643579861773e-05, "loss": 0.0092, "step": 15970 }, { "grad_norm": 0.13297314941883087, "learning_rate": 9.674056614327886e-05, "loss": 0.0101, "step": 15980 }, { "grad_norm": 0.2219514697790146, "learning_rate": 9.673469137646198e-05, "loss": 0.0104, "step": 15990 }, { "grad_norm": 0.13986901938915253, "learning_rate": 9.67288114988095e-05, "loss": 0.0096, "step": 16000 }, { "grad_norm": 0.14518018066883087, "learning_rate": 9.672292651096447e-05, "loss": 0.0102, "step": 16010 }, { "grad_norm": 0.1595945805311203, "learning_rate": 9.671703641357042e-05, "loss": 0.0106, "step": 16020 }, { "grad_norm": 0.18492230772972107, "learning_rate": 9.67111412072715e-05, "loss": 0.0098, "step": 16030 }, { "grad_norm": 0.11406111717224121, "learning_rate": 9.670524089271242e-05, "loss": 0.0081, "step": 16040 }, { "grad_norm": 0.1532430201768875, "learning_rate": 9.669933547053842e-05, "loss": 0.0087, "step": 16050 }, { "grad_norm": 0.13056355714797974, "learning_rate": 9.669342494139531e-05, "loss": 0.0094, "step": 16060 }, { "grad_norm": 0.13280391693115234, "learning_rate": 9.668750930592943e-05, "loss": 0.0086, "step": 16070 }, { "grad_norm": 0.18399770557880402, "learning_rate": 9.668158856478775e-05, "loss": 0.0093, "step": 16080 }, { "grad_norm": 0.14535851776599884, "learning_rate": 9.66756627186177e-05, "loss": 0.0087, "step": 16090 }, { "grad_norm": 0.14812809228897095, "learning_rate": 9.666973176806737e-05, "loss": 0.0077, "step": 16100 }, { "grad_norm": 0.1657167375087738, "learning_rate": 9.666379571378534e-05, "loss": 0.0082, "step": 16110 }, { "grad_norm": 0.10698068141937256, "learning_rate": 9.665785455642076e-05, "loss": 0.009, "step": 16120 }, { "grad_norm": 0.14949443936347961, "learning_rate": 9.665190829662337e-05, "loss": 0.0082, "step": 16130 }, { "grad_norm": 0.1420239806175232, "learning_rate": 9.664595693504342e-05, "loss": 0.0077, "step": 16140 }, { "grad_norm": 0.1577903926372528, "learning_rate": 9.664000047233175e-05, "loss": 0.0086, "step": 16150 }, { "grad_norm": 0.13950298726558685, "learning_rate": 9.663403890913976e-05, "loss": 0.0087, "step": 16160 }, { "grad_norm": 0.14673326909542084, "learning_rate": 9.662807224611938e-05, "loss": 0.0084, "step": 16170 }, { "grad_norm": 0.1690548062324524, "learning_rate": 9.662210048392311e-05, "loss": 0.0084, "step": 16180 }, { "grad_norm": 0.14276528358459473, "learning_rate": 9.661612362320405e-05, "loss": 0.0087, "step": 16190 }, { "grad_norm": 0.18794994056224823, "learning_rate": 9.661014166461579e-05, "loss": 0.0085, "step": 16200 }, { "grad_norm": 0.1308334469795227, "learning_rate": 9.66041546088125e-05, "loss": 0.0103, "step": 16210 }, { "grad_norm": 0.18026717007160187, "learning_rate": 9.659816245644895e-05, "loss": 0.0092, "step": 16220 }, { "grad_norm": 0.12346335500478745, "learning_rate": 9.65921652081804e-05, "loss": 0.0081, "step": 16230 }, { "grad_norm": 0.13057999312877655, "learning_rate": 9.658616286466271e-05, "loss": 0.0083, "step": 16240 }, { "grad_norm": 0.1253214180469513, "learning_rate": 9.65801554265523e-05, "loss": 0.0087, "step": 16250 }, { "grad_norm": 0.17602768540382385, "learning_rate": 9.657414289450612e-05, "loss": 0.0095, "step": 16260 }, { "grad_norm": 0.1146782711148262, "learning_rate": 9.656812526918171e-05, "loss": 0.0091, "step": 16270 }, { "grad_norm": 0.14861957728862762, "learning_rate": 9.656210255123712e-05, "loss": 0.0095, "step": 16280 }, { "grad_norm": 0.1398642212152481, "learning_rate": 9.6556074741331e-05, "loss": 0.0102, "step": 16290 }, { "grad_norm": 0.1696161925792694, "learning_rate": 9.655004184012256e-05, "loss": 0.0092, "step": 16300 }, { "grad_norm": 0.18090544641017914, "learning_rate": 9.654400384827152e-05, "loss": 0.0106, "step": 16310 }, { "grad_norm": 0.13669951260089874, "learning_rate": 9.653796076643818e-05, "loss": 0.0087, "step": 16320 }, { "grad_norm": 0.23128637671470642, "learning_rate": 9.653191259528344e-05, "loss": 0.0086, "step": 16330 }, { "grad_norm": 0.14365597069263458, "learning_rate": 9.65258593354687e-05, "loss": 0.0095, "step": 16340 }, { "grad_norm": 0.14441800117492676, "learning_rate": 9.651980098765591e-05, "loss": 0.0093, "step": 16350 }, { "grad_norm": 0.16584444046020508, "learning_rate": 9.651373755250765e-05, "loss": 0.0088, "step": 16360 }, { "grad_norm": 0.14958423376083374, "learning_rate": 9.650766903068697e-05, "loss": 0.0093, "step": 16370 }, { "grad_norm": 0.18733011186122894, "learning_rate": 9.650159542285753e-05, "loss": 0.0096, "step": 16380 }, { "grad_norm": 0.1290605068206787, "learning_rate": 9.649551672968353e-05, "loss": 0.0095, "step": 16390 }, { "grad_norm": 0.13142475485801697, "learning_rate": 9.648943295182973e-05, "loss": 0.0086, "step": 16400 }, { "grad_norm": 0.11837010830640793, "learning_rate": 9.648334408996144e-05, "loss": 0.008, "step": 16410 }, { "grad_norm": 0.15234489738941193, "learning_rate": 9.647725014474452e-05, "loss": 0.0079, "step": 16420 }, { "grad_norm": 0.18504564464092255, "learning_rate": 9.64711511168454e-05, "loss": 0.0098, "step": 16430 }, { "grad_norm": 0.09921599179506302, "learning_rate": 9.646504700693108e-05, "loss": 0.0087, "step": 16440 }, { "grad_norm": 0.10065943747758865, "learning_rate": 9.645893781566907e-05, "loss": 0.0092, "step": 16450 }, { "grad_norm": 0.14401578903198242, "learning_rate": 9.645282354372744e-05, "loss": 0.0077, "step": 16460 }, { "grad_norm": 0.18240498006343842, "learning_rate": 9.644670419177491e-05, "loss": 0.0082, "step": 16470 }, { "grad_norm": 0.19778503477573395, "learning_rate": 9.644057976048062e-05, "loss": 0.0089, "step": 16480 }, { "grad_norm": 0.12732528150081635, "learning_rate": 9.643445025051435e-05, "loss": 0.0081, "step": 16490 }, { "grad_norm": 0.14288808405399323, "learning_rate": 9.642831566254641e-05, "loss": 0.0094, "step": 16500 }, { "grad_norm": 0.14506471157073975, "learning_rate": 9.642217599724769e-05, "loss": 0.0084, "step": 16510 }, { "grad_norm": 0.13887986540794373, "learning_rate": 9.64160312552896e-05, "loss": 0.0078, "step": 16520 }, { "grad_norm": 0.1282307207584381, "learning_rate": 9.64098814373441e-05, "loss": 0.0077, "step": 16530 }, { "grad_norm": 0.17354996502399445, "learning_rate": 9.640372654408374e-05, "loss": 0.0078, "step": 16540 }, { "grad_norm": 0.18547335267066956, "learning_rate": 9.639756657618162e-05, "loss": 0.0085, "step": 16550 }, { "grad_norm": 0.17696918547153473, "learning_rate": 9.639140153431138e-05, "loss": 0.0077, "step": 16560 }, { "grad_norm": 0.11762721836566925, "learning_rate": 9.638523141914721e-05, "loss": 0.0077, "step": 16570 }, { "grad_norm": 0.12804628908634186, "learning_rate": 9.637905623136388e-05, "loss": 0.008, "step": 16580 }, { "grad_norm": 0.14917032420635223, "learning_rate": 9.637287597163669e-05, "loss": 0.0077, "step": 16590 }, { "grad_norm": 0.13127820193767548, "learning_rate": 9.63666906406415e-05, "loss": 0.008, "step": 16600 }, { "grad_norm": 0.12708322703838348, "learning_rate": 9.636050023905473e-05, "loss": 0.0085, "step": 16610 }, { "grad_norm": 0.14612579345703125, "learning_rate": 9.635430476755336e-05, "loss": 0.0083, "step": 16620 }, { "grad_norm": 0.15731488168239594, "learning_rate": 9.63481042268149e-05, "loss": 0.0088, "step": 16630 }, { "grad_norm": 0.15990252792835236, "learning_rate": 9.634189861751745e-05, "loss": 0.008, "step": 16640 }, { "grad_norm": 0.13100504875183105, "learning_rate": 9.633568794033967e-05, "loss": 0.0087, "step": 16650 }, { "grad_norm": 0.13180622458457947, "learning_rate": 9.63294721959607e-05, "loss": 0.0084, "step": 16660 }, { "grad_norm": 0.10911405831575394, "learning_rate": 9.63232513850603e-05, "loss": 0.0081, "step": 16670 }, { "grad_norm": 0.1874103993177414, "learning_rate": 9.631702550831878e-05, "loss": 0.0102, "step": 16680 }, { "grad_norm": 0.13526560366153717, "learning_rate": 9.631079456641698e-05, "loss": 0.0079, "step": 16690 }, { "grad_norm": 0.13185325264930725, "learning_rate": 9.630455856003632e-05, "loss": 0.0082, "step": 16700 }, { "grad_norm": 0.14947271347045898, "learning_rate": 9.629831748985876e-05, "loss": 0.0085, "step": 16710 }, { "grad_norm": 0.1127859354019165, "learning_rate": 9.629207135656679e-05, "loss": 0.0095, "step": 16720 }, { "grad_norm": 0.13694919645786285, "learning_rate": 9.628582016084353e-05, "loss": 0.0092, "step": 16730 }, { "grad_norm": 0.15017183125019073, "learning_rate": 9.627956390337254e-05, "loss": 0.0084, "step": 16740 }, { "grad_norm": 0.10783182829618454, "learning_rate": 9.627330258483802e-05, "loss": 0.0088, "step": 16750 }, { "grad_norm": 0.17369382083415985, "learning_rate": 9.62670362059247e-05, "loss": 0.008, "step": 16760 }, { "grad_norm": 0.14854776859283447, "learning_rate": 9.626076476731786e-05, "loss": 0.0084, "step": 16770 }, { "grad_norm": 0.12433286756277084, "learning_rate": 9.625448826970336e-05, "loss": 0.0086, "step": 16780 }, { "grad_norm": 0.14961978793144226, "learning_rate": 9.624820671376755e-05, "loss": 0.0085, "step": 16790 }, { "grad_norm": 0.19650812447071075, "learning_rate": 9.62419201001974e-05, "loss": 0.008, "step": 16800 }, { "grad_norm": 0.0931919515132904, "learning_rate": 9.623562842968037e-05, "loss": 0.0087, "step": 16810 }, { "grad_norm": 0.18411828577518463, "learning_rate": 9.622933170290454e-05, "loss": 0.0095, "step": 16820 }, { "grad_norm": 0.1544245034456253, "learning_rate": 9.622302992055849e-05, "loss": 0.0087, "step": 16830 }, { "grad_norm": 0.13990575075149536, "learning_rate": 9.62167230833314e-05, "loss": 0.0079, "step": 16840 }, { "grad_norm": 0.17957831919193268, "learning_rate": 9.621041119191295e-05, "loss": 0.0099, "step": 16850 }, { "grad_norm": 0.13972057402133942, "learning_rate": 9.620409424699342e-05, "loss": 0.0087, "step": 16860 }, { "grad_norm": 0.11968889832496643, "learning_rate": 9.619777224926359e-05, "loss": 0.0092, "step": 16870 }, { "grad_norm": 0.13547194004058838, "learning_rate": 9.619144519941485e-05, "loss": 0.0085, "step": 16880 }, { "grad_norm": 0.13576626777648926, "learning_rate": 9.618511309813912e-05, "loss": 0.0083, "step": 16890 }, { "grad_norm": 0.16598115861415863, "learning_rate": 9.617877594612886e-05, "loss": 0.0092, "step": 16900 }, { "grad_norm": 0.17874354124069214, "learning_rate": 9.617243374407707e-05, "loss": 0.0083, "step": 16910 }, { "grad_norm": 0.21124538779258728, "learning_rate": 9.616608649267736e-05, "loss": 0.0092, "step": 16920 }, { "grad_norm": 0.18933427333831787, "learning_rate": 9.615973419262385e-05, "loss": 0.0088, "step": 16930 }, { "grad_norm": 0.17442715167999268, "learning_rate": 9.615337684461119e-05, "loss": 0.0099, "step": 16940 }, { "grad_norm": 0.11790895462036133, "learning_rate": 9.614701444933465e-05, "loss": 0.0104, "step": 16950 }, { "grad_norm": 0.19579152762889862, "learning_rate": 9.614064700748997e-05, "loss": 0.0076, "step": 16960 }, { "grad_norm": 0.10169141739606857, "learning_rate": 9.613427451977352e-05, "loss": 0.0087, "step": 16970 }, { "grad_norm": 0.14535510540008545, "learning_rate": 9.612789698688216e-05, "loss": 0.0084, "step": 16980 }, { "grad_norm": 0.20116011798381805, "learning_rate": 9.612151440951334e-05, "loss": 0.0085, "step": 16990 }, { "grad_norm": 0.1875116378068924, "learning_rate": 9.611512678836506e-05, "loss": 0.0119, "step": 17000 }, { "grad_norm": 0.1436138153076172, "learning_rate": 9.610873412413584e-05, "loss": 0.0115, "step": 17010 }, { "grad_norm": 0.1553129255771637, "learning_rate": 9.610233641752476e-05, "loss": 0.0125, "step": 17020 }, { "grad_norm": 0.2234368473291397, "learning_rate": 9.609593366923151e-05, "loss": 0.0148, "step": 17030 }, { "grad_norm": 0.20610277354717255, "learning_rate": 9.608952587995625e-05, "loss": 0.0106, "step": 17040 }, { "grad_norm": 0.1383116990327835, "learning_rate": 9.608311305039972e-05, "loss": 0.011, "step": 17050 }, { "grad_norm": 0.13257378339767456, "learning_rate": 9.607669518126326e-05, "loss": 0.0097, "step": 17060 }, { "grad_norm": 0.1979174017906189, "learning_rate": 9.607027227324866e-05, "loss": 0.0101, "step": 17070 }, { "grad_norm": 0.16721951961517334, "learning_rate": 9.606384432705837e-05, "loss": 0.0104, "step": 17080 }, { "grad_norm": 0.10622473806142807, "learning_rate": 9.60574113433953e-05, "loss": 0.0118, "step": 17090 }, { "grad_norm": 0.13881585001945496, "learning_rate": 9.6050973322963e-05, "loss": 0.0109, "step": 17100 }, { "grad_norm": 0.16729675233364105, "learning_rate": 9.604453026646547e-05, "loss": 0.0101, "step": 17110 }, { "grad_norm": 0.144328773021698, "learning_rate": 9.603808217460735e-05, "loss": 0.0106, "step": 17120 }, { "grad_norm": 0.1263548880815506, "learning_rate": 9.603162904809377e-05, "loss": 0.0098, "step": 17130 }, { "grad_norm": 0.12659984827041626, "learning_rate": 9.602517088763045e-05, "loss": 0.0094, "step": 17140 }, { "grad_norm": 0.1286853402853012, "learning_rate": 9.601870769392365e-05, "loss": 0.0095, "step": 17150 }, { "grad_norm": 0.190130814909935, "learning_rate": 9.601223946768017e-05, "loss": 0.0105, "step": 17160 }, { "grad_norm": 0.11230356991291046, "learning_rate": 9.600576620960734e-05, "loss": 0.0112, "step": 17170 }, { "grad_norm": 0.12336166203022003, "learning_rate": 9.599928792041308e-05, "loss": 0.0104, "step": 17180 }, { "grad_norm": 0.15633924305438995, "learning_rate": 9.599280460080587e-05, "loss": 0.009, "step": 17190 }, { "grad_norm": 0.13788850605487823, "learning_rate": 9.59863162514947e-05, "loss": 0.0094, "step": 17200 }, { "grad_norm": 0.18093745410442352, "learning_rate": 9.597982287318911e-05, "loss": 0.0103, "step": 17210 }, { "grad_norm": 0.21981565654277802, "learning_rate": 9.597332446659923e-05, "loss": 0.0103, "step": 17220 }, { "grad_norm": 0.20832610130310059, "learning_rate": 9.59668210324357e-05, "loss": 0.0088, "step": 17230 }, { "grad_norm": 0.1556137353181839, "learning_rate": 9.596031257140974e-05, "loss": 0.0087, "step": 17240 }, { "grad_norm": 0.1532529890537262, "learning_rate": 9.59537990842331e-05, "loss": 0.0098, "step": 17250 }, { "grad_norm": 0.14965464174747467, "learning_rate": 9.594728057161806e-05, "loss": 0.0082, "step": 17260 }, { "grad_norm": 0.20257078111171722, "learning_rate": 9.594075703427752e-05, "loss": 0.0103, "step": 17270 }, { "grad_norm": 0.13908398151397705, "learning_rate": 9.593422847292486e-05, "loss": 0.0105, "step": 17280 }, { "grad_norm": 0.15795791149139404, "learning_rate": 9.592769488827402e-05, "loss": 0.0081, "step": 17290 }, { "grad_norm": 0.1696564257144928, "learning_rate": 9.592115628103952e-05, "loss": 0.0104, "step": 17300 }, { "grad_norm": 0.13265201449394226, "learning_rate": 9.591461265193643e-05, "loss": 0.0087, "step": 17310 }, { "grad_norm": 0.15682794153690338, "learning_rate": 9.590806400168032e-05, "loss": 0.0086, "step": 17320 }, { "grad_norm": 0.19386973977088928, "learning_rate": 9.590151033098735e-05, "loss": 0.0081, "step": 17330 }, { "grad_norm": 0.12574484944343567, "learning_rate": 9.589495164057423e-05, "loss": 0.008, "step": 17340 }, { "grad_norm": 0.17612886428833008, "learning_rate": 9.58883879311582e-05, "loss": 0.0078, "step": 17350 }, { "grad_norm": 0.1314210593700409, "learning_rate": 9.588181920345705e-05, "loss": 0.0081, "step": 17360 }, { "grad_norm": 0.16158924996852875, "learning_rate": 9.587524545818913e-05, "loss": 0.0093, "step": 17370 }, { "grad_norm": 0.1251295804977417, "learning_rate": 9.586866669607335e-05, "loss": 0.011, "step": 17380 }, { "grad_norm": 0.165981262922287, "learning_rate": 9.586208291782915e-05, "loss": 0.0087, "step": 17390 }, { "grad_norm": 0.16581323742866516, "learning_rate": 9.58554941241765e-05, "loss": 0.009, "step": 17400 }, { "grad_norm": 0.17422664165496826, "learning_rate": 9.584890031583596e-05, "loss": 0.0095, "step": 17410 }, { "grad_norm": 0.19156305491924286, "learning_rate": 9.584230149352861e-05, "loss": 0.0086, "step": 17420 }, { "grad_norm": 0.16522657871246338, "learning_rate": 9.58356976579761e-05, "loss": 0.0096, "step": 17430 }, { "grad_norm": 0.12279955297708511, "learning_rate": 9.58290888099006e-05, "loss": 0.0083, "step": 17440 }, { "grad_norm": 0.1415499448776245, "learning_rate": 9.582247495002486e-05, "loss": 0.0079, "step": 17450 }, { "grad_norm": 0.11972229927778244, "learning_rate": 9.581585607907214e-05, "loss": 0.0083, "step": 17460 }, { "grad_norm": 0.15567201375961304, "learning_rate": 9.580923219776628e-05, "loss": 0.0079, "step": 17470 }, { "grad_norm": 0.1349046677350998, "learning_rate": 9.580260330683167e-05, "loss": 0.0075, "step": 17480 }, { "grad_norm": 0.08903279155492783, "learning_rate": 9.579596940699322e-05, "loss": 0.0082, "step": 17490 }, { "grad_norm": 0.09584199637174606, "learning_rate": 9.578933049897643e-05, "loss": 0.0079, "step": 17500 }, { "grad_norm": 0.10311345756053925, "learning_rate": 9.578268658350728e-05, "loss": 0.0081, "step": 17510 }, { "grad_norm": 0.11942526698112488, "learning_rate": 9.577603766131235e-05, "loss": 0.0077, "step": 17520 }, { "grad_norm": 0.10032171756029129, "learning_rate": 9.576938373311878e-05, "loss": 0.0078, "step": 17530 }, { "grad_norm": 0.09155547618865967, "learning_rate": 9.576272479965421e-05, "loss": 0.0075, "step": 17540 }, { "grad_norm": 0.1415141373872757, "learning_rate": 9.575606086164687e-05, "loss": 0.0081, "step": 17550 }, { "grad_norm": 0.11687972396612167, "learning_rate": 9.57493919198255e-05, "loss": 0.008, "step": 17560 }, { "grad_norm": 0.1382148563861847, "learning_rate": 9.57427179749194e-05, "loss": 0.0067, "step": 17570 }, { "grad_norm": 0.14569509029388428, "learning_rate": 9.573603902765846e-05, "loss": 0.0078, "step": 17580 }, { "grad_norm": 0.20885629951953888, "learning_rate": 9.572935507877304e-05, "loss": 0.0073, "step": 17590 }, { "grad_norm": 0.13863115012645721, "learning_rate": 9.57226661289941e-05, "loss": 0.0074, "step": 17600 }, { "grad_norm": 0.11283013224601746, "learning_rate": 9.571597217905315e-05, "loss": 0.0088, "step": 17610 }, { "grad_norm": 0.14000505208969116, "learning_rate": 9.57092732296822e-05, "loss": 0.0083, "step": 17620 }, { "grad_norm": 0.09848646074533463, "learning_rate": 9.570256928161385e-05, "loss": 0.0069, "step": 17630 }, { "grad_norm": 0.1901836395263672, "learning_rate": 9.569586033558126e-05, "loss": 0.0085, "step": 17640 }, { "grad_norm": 0.16183380782604218, "learning_rate": 9.568914639231807e-05, "loss": 0.0087, "step": 17650 }, { "grad_norm": 0.14897534251213074, "learning_rate": 9.568242745255852e-05, "loss": 0.0078, "step": 17660 }, { "grad_norm": 0.11720766872167587, "learning_rate": 9.567570351703739e-05, "loss": 0.0084, "step": 17670 }, { "grad_norm": 0.10565494000911713, "learning_rate": 9.566897458649001e-05, "loss": 0.0081, "step": 17680 }, { "grad_norm": 0.161061629652977, "learning_rate": 9.566224066165221e-05, "loss": 0.0088, "step": 17690 }, { "grad_norm": 0.12943968176841736, "learning_rate": 9.565550174326043e-05, "loss": 0.0083, "step": 17700 }, { "grad_norm": 0.15196731686592102, "learning_rate": 9.564875783205162e-05, "loss": 0.008, "step": 17710 }, { "grad_norm": 0.13108262419700623, "learning_rate": 9.564200892876328e-05, "loss": 0.0079, "step": 17720 }, { "grad_norm": 0.11986333876848221, "learning_rate": 9.563525503413348e-05, "loss": 0.0083, "step": 17730 }, { "grad_norm": 0.10671839863061905, "learning_rate": 9.562849614890079e-05, "loss": 0.0081, "step": 17740 }, { "grad_norm": 0.12295147776603699, "learning_rate": 9.562173227380436e-05, "loss": 0.0079, "step": 17750 }, { "grad_norm": 0.20699912309646606, "learning_rate": 9.561496340958389e-05, "loss": 0.0089, "step": 17760 }, { "grad_norm": 0.13143236935138702, "learning_rate": 9.560818955697959e-05, "loss": 0.0084, "step": 17770 }, { "grad_norm": 0.13395945727825165, "learning_rate": 9.560141071673228e-05, "loss": 0.0084, "step": 17780 }, { "grad_norm": 0.1595979481935501, "learning_rate": 9.559462688958323e-05, "loss": 0.0079, "step": 17790 }, { "grad_norm": 0.1323358118534088, "learning_rate": 9.558783807627434e-05, "loss": 0.0085, "step": 17800 }, { "grad_norm": 0.17435172200202942, "learning_rate": 9.558104427754801e-05, "loss": 0.0082, "step": 17810 }, { "grad_norm": 0.14117100834846497, "learning_rate": 9.557424549414722e-05, "loss": 0.0076, "step": 17820 }, { "grad_norm": 0.14478932321071625, "learning_rate": 9.556744172681546e-05, "loss": 0.0083, "step": 17830 }, { "grad_norm": 0.13266777992248535, "learning_rate": 9.556063297629677e-05, "loss": 0.0075, "step": 17840 }, { "grad_norm": 0.1307954490184784, "learning_rate": 9.555381924333578e-05, "loss": 0.0072, "step": 17850 }, { "grad_norm": 0.10788778960704803, "learning_rate": 9.554700052867758e-05, "loss": 0.0083, "step": 17860 }, { "grad_norm": 0.11088696867227554, "learning_rate": 9.554017683306789e-05, "loss": 0.0072, "step": 17870 }, { "grad_norm": 0.12104590237140656, "learning_rate": 9.553334815725294e-05, "loss": 0.0073, "step": 17880 }, { "grad_norm": 0.10516233742237091, "learning_rate": 9.552651450197949e-05, "loss": 0.0071, "step": 17890 }, { "grad_norm": 0.12131718546152115, "learning_rate": 9.551967586799486e-05, "loss": 0.0074, "step": 17900 }, { "grad_norm": 0.11782762408256531, "learning_rate": 9.551283225604692e-05, "loss": 0.0087, "step": 17910 }, { "grad_norm": 0.13809649646282196, "learning_rate": 9.550598366688406e-05, "loss": 0.0071, "step": 17920 }, { "grad_norm": 0.11245551705360413, "learning_rate": 9.549913010125526e-05, "loss": 0.008, "step": 17930 }, { "grad_norm": 0.1071302518248558, "learning_rate": 9.549227155990999e-05, "loss": 0.0077, "step": 17940 }, { "grad_norm": 0.16632018983364105, "learning_rate": 9.548540804359828e-05, "loss": 0.0075, "step": 17950 }, { "grad_norm": 0.15231011807918549, "learning_rate": 9.547853955307077e-05, "loss": 0.0084, "step": 17960 }, { "grad_norm": 0.14982442557811737, "learning_rate": 9.547166608907853e-05, "loss": 0.0076, "step": 17970 }, { "grad_norm": 0.14228397607803345, "learning_rate": 9.546478765237326e-05, "loss": 0.0092, "step": 17980 }, { "grad_norm": 0.10763382166624069, "learning_rate": 9.545790424370715e-05, "loss": 0.0095, "step": 17990 }, { "grad_norm": 0.14387892186641693, "learning_rate": 9.5451015863833e-05, "loss": 0.0088, "step": 18000 }, { "grad_norm": 0.10834018141031265, "learning_rate": 9.544412251350408e-05, "loss": 0.0081, "step": 18010 }, { "grad_norm": 0.10636366158723831, "learning_rate": 9.543722419347422e-05, "loss": 0.008, "step": 18020 }, { "grad_norm": 0.10577370226383209, "learning_rate": 9.543032090449788e-05, "loss": 0.0075, "step": 18030 }, { "grad_norm": 0.13433629274368286, "learning_rate": 9.542341264732992e-05, "loss": 0.0084, "step": 18040 }, { "grad_norm": 0.10946722328662872, "learning_rate": 9.541649942272585e-05, "loss": 0.0078, "step": 18050 }, { "grad_norm": 0.1151462271809578, "learning_rate": 9.54095812314417e-05, "loss": 0.0078, "step": 18060 }, { "grad_norm": 0.14610089361667633, "learning_rate": 9.540265807423401e-05, "loss": 0.0073, "step": 18070 }, { "grad_norm": 0.162100687623024, "learning_rate": 9.53957299518599e-05, "loss": 0.0079, "step": 18080 }, { "grad_norm": 0.147406667470932, "learning_rate": 9.5388796865077e-05, "loss": 0.0088, "step": 18090 }, { "grad_norm": 0.15306971967220306, "learning_rate": 9.538185881464353e-05, "loss": 0.0085, "step": 18100 }, { "grad_norm": 0.15018479526042938, "learning_rate": 9.537491580131821e-05, "loss": 0.0082, "step": 18110 }, { "grad_norm": 0.2152872234582901, "learning_rate": 9.53679678258603e-05, "loss": 0.0097, "step": 18120 }, { "grad_norm": 0.10214269906282425, "learning_rate": 9.536101488902966e-05, "loss": 0.0077, "step": 18130 }, { "grad_norm": 0.13888737559318542, "learning_rate": 9.535405699158663e-05, "loss": 0.0091, "step": 18140 }, { "grad_norm": 0.15565426647663116, "learning_rate": 9.53470941342921e-05, "loss": 0.009, "step": 18150 }, { "grad_norm": 0.11785149574279785, "learning_rate": 9.534012631790756e-05, "loss": 0.0079, "step": 18160 }, { "grad_norm": 0.1612296998500824, "learning_rate": 9.533315354319494e-05, "loss": 0.0085, "step": 18170 }, { "grad_norm": 0.1531142145395279, "learning_rate": 9.532617581091682e-05, "loss": 0.0082, "step": 18180 }, { "grad_norm": 0.14113469421863556, "learning_rate": 9.531919312183629e-05, "loss": 0.009, "step": 18190 }, { "grad_norm": 0.1613248735666275, "learning_rate": 9.531220547671688e-05, "loss": 0.0095, "step": 18200 }, { "grad_norm": 0.1429944783449173, "learning_rate": 9.530521287632285e-05, "loss": 0.0081, "step": 18210 }, { "grad_norm": 0.12825196981430054, "learning_rate": 9.529821532141884e-05, "loss": 0.0079, "step": 18220 }, { "grad_norm": 0.14123208820819855, "learning_rate": 9.52912128127701e-05, "loss": 0.0083, "step": 18230 }, { "grad_norm": 0.1358853131532669, "learning_rate": 9.528420535114244e-05, "loss": 0.0073, "step": 18240 }, { "grad_norm": 0.12271984666585922, "learning_rate": 9.527719293730215e-05, "loss": 0.0089, "step": 18250 }, { "grad_norm": 0.13495783507823944, "learning_rate": 9.527017557201611e-05, "loss": 0.0074, "step": 18260 }, { "grad_norm": 0.10595802962779999, "learning_rate": 9.526315325605176e-05, "loss": 0.0083, "step": 18270 }, { "grad_norm": 0.12103544175624847, "learning_rate": 9.525612599017699e-05, "loss": 0.0077, "step": 18280 }, { "grad_norm": 0.14095279574394226, "learning_rate": 9.524909377516033e-05, "loss": 0.008, "step": 18290 }, { "grad_norm": 0.11993099749088287, "learning_rate": 9.524205661177081e-05, "loss": 0.0083, "step": 18300 }, { "grad_norm": 0.17859284579753876, "learning_rate": 9.523501450077801e-05, "loss": 0.0079, "step": 18310 }, { "grad_norm": 0.19286778569221497, "learning_rate": 9.522796744295202e-05, "loss": 0.01, "step": 18320 }, { "grad_norm": 0.14496344327926636, "learning_rate": 9.522091543906352e-05, "loss": 0.0084, "step": 18330 }, { "grad_norm": 0.11554134637117386, "learning_rate": 9.521385848988369e-05, "loss": 0.0074, "step": 18340 }, { "grad_norm": 0.12152841687202454, "learning_rate": 9.520679659618428e-05, "loss": 0.0073, "step": 18350 }, { "grad_norm": 0.16104364395141602, "learning_rate": 9.519972975873754e-05, "loss": 0.0085, "step": 18360 }, { "grad_norm": 0.11819438636302948, "learning_rate": 9.519265797831633e-05, "loss": 0.0078, "step": 18370 }, { "grad_norm": 0.13795767724514008, "learning_rate": 9.518558125569399e-05, "loss": 0.0077, "step": 18380 }, { "grad_norm": 0.11696916073560715, "learning_rate": 9.517849959164442e-05, "loss": 0.0087, "step": 18390 }, { "grad_norm": 0.12309619039297104, "learning_rate": 9.517141298694205e-05, "loss": 0.0075, "step": 18400 }, { "grad_norm": 0.10748564451932907, "learning_rate": 9.516432144236188e-05, "loss": 0.0087, "step": 18410 }, { "grad_norm": 0.11189955472946167, "learning_rate": 9.515722495867941e-05, "loss": 0.0076, "step": 18420 }, { "grad_norm": 0.15328404307365417, "learning_rate": 9.515012353667072e-05, "loss": 0.0078, "step": 18430 }, { "grad_norm": 0.17668676376342773, "learning_rate": 9.51430171771124e-05, "loss": 0.0081, "step": 18440 }, { "grad_norm": 0.10807269811630249, "learning_rate": 9.513590588078159e-05, "loss": 0.0087, "step": 18450 }, { "grad_norm": 0.13017554581165314, "learning_rate": 9.512878964845597e-05, "loss": 0.0073, "step": 18460 }, { "grad_norm": 0.14128464460372925, "learning_rate": 9.512166848091377e-05, "loss": 0.008, "step": 18470 }, { "grad_norm": 0.14262138307094574, "learning_rate": 9.511454237893376e-05, "loss": 0.0073, "step": 18480 }, { "grad_norm": 0.13909529149532318, "learning_rate": 9.51074113432952e-05, "loss": 0.0078, "step": 18490 }, { "grad_norm": 0.11648419499397278, "learning_rate": 9.510027537477797e-05, "loss": 0.0071, "step": 18500 }, { "grad_norm": 0.09038618206977844, "learning_rate": 9.509313447416242e-05, "loss": 0.0069, "step": 18510 }, { "grad_norm": 0.14551541209220886, "learning_rate": 9.508598864222949e-05, "loss": 0.0089, "step": 18520 }, { "grad_norm": 0.14265955984592438, "learning_rate": 9.507883787976062e-05, "loss": 0.0082, "step": 18530 }, { "grad_norm": 0.12185318768024445, "learning_rate": 9.507168218753781e-05, "loss": 0.0082, "step": 18540 }, { "grad_norm": 0.20257717370986938, "learning_rate": 9.506452156634362e-05, "loss": 0.0078, "step": 18550 }, { "grad_norm": 0.14230522513389587, "learning_rate": 9.505735601696109e-05, "loss": 0.0079, "step": 18560 }, { "grad_norm": 0.12090729922056198, "learning_rate": 9.505018554017385e-05, "loss": 0.008, "step": 18570 }, { "grad_norm": 0.1072193905711174, "learning_rate": 9.504301013676604e-05, "loss": 0.0074, "step": 18580 }, { "grad_norm": 0.12205422669649124, "learning_rate": 9.503582980752238e-05, "loss": 0.0074, "step": 18590 }, { "grad_norm": 0.12186075001955032, "learning_rate": 9.502864455322809e-05, "loss": 0.0094, "step": 18600 }, { "grad_norm": 0.16582202911376953, "learning_rate": 9.502145437466891e-05, "loss": 0.0085, "step": 18610 }, { "grad_norm": 0.13094475865364075, "learning_rate": 9.501425927263116e-05, "loss": 0.0079, "step": 18620 }, { "grad_norm": 0.12739649415016174, "learning_rate": 9.500705924790172e-05, "loss": 0.0077, "step": 18630 }, { "grad_norm": 0.10221894830465317, "learning_rate": 9.499985430126794e-05, "loss": 0.0085, "step": 18640 }, { "grad_norm": 0.120255246758461, "learning_rate": 9.499264443351775e-05, "loss": 0.0086, "step": 18650 }, { "grad_norm": 0.11179458349943161, "learning_rate": 9.498542964543961e-05, "loss": 0.0073, "step": 18660 }, { "grad_norm": 0.09365399926900864, "learning_rate": 9.497820993782252e-05, "loss": 0.0072, "step": 18670 }, { "grad_norm": 0.11356845498085022, "learning_rate": 9.497098531145601e-05, "loss": 0.0075, "step": 18680 }, { "grad_norm": 0.11992473900318146, "learning_rate": 9.496375576713017e-05, "loss": 0.0075, "step": 18690 }, { "grad_norm": 0.13774560391902924, "learning_rate": 9.49565213056356e-05, "loss": 0.0078, "step": 18700 }, { "grad_norm": 0.1559152901172638, "learning_rate": 9.494928192776342e-05, "loss": 0.008, "step": 18710 }, { "grad_norm": 0.16370989382266998, "learning_rate": 9.494203763430538e-05, "loss": 0.0069, "step": 18720 }, { "grad_norm": 0.124579519033432, "learning_rate": 9.493478842605366e-05, "loss": 0.0083, "step": 18730 }, { "grad_norm": 0.23522509634494781, "learning_rate": 9.492753430380105e-05, "loss": 0.0074, "step": 18740 }, { "grad_norm": 0.14816977083683014, "learning_rate": 9.492027526834083e-05, "loss": 0.0081, "step": 18750 }, { "grad_norm": 0.13290058076381683, "learning_rate": 9.491301132046684e-05, "loss": 0.0102, "step": 18760 }, { "grad_norm": 0.1436096727848053, "learning_rate": 9.490574246097345e-05, "loss": 0.0083, "step": 18770 }, { "grad_norm": 0.15113712847232819, "learning_rate": 9.48984686906556e-05, "loss": 0.0074, "step": 18780 }, { "grad_norm": 0.09335353225469589, "learning_rate": 9.489119001030871e-05, "loss": 0.008, "step": 18790 }, { "grad_norm": 0.17178189754486084, "learning_rate": 9.488390642072878e-05, "loss": 0.0105, "step": 18800 }, { "grad_norm": 0.09991805255413055, "learning_rate": 9.48766179227123e-05, "loss": 0.0085, "step": 18810 }, { "grad_norm": 0.15453042089939117, "learning_rate": 9.486932451705636e-05, "loss": 0.0088, "step": 18820 }, { "grad_norm": 0.15653884410858154, "learning_rate": 9.486202620455857e-05, "loss": 0.0079, "step": 18830 }, { "grad_norm": 0.13794510066509247, "learning_rate": 9.485472298601704e-05, "loss": 0.0074, "step": 18840 }, { "grad_norm": 0.10253830999135971, "learning_rate": 9.484741486223043e-05, "loss": 0.0084, "step": 18850 }, { "grad_norm": 0.14169436693191528, "learning_rate": 9.484010183399797e-05, "loss": 0.0084, "step": 18860 }, { "grad_norm": 0.17206957936286926, "learning_rate": 9.483278390211938e-05, "loss": 0.0093, "step": 18870 }, { "grad_norm": 0.14943358302116394, "learning_rate": 9.482546106739496e-05, "loss": 0.0083, "step": 18880 }, { "grad_norm": 0.09754665940999985, "learning_rate": 9.48181333306255e-05, "loss": 0.008, "step": 18890 }, { "grad_norm": 0.2063722461462021, "learning_rate": 9.481080069261237e-05, "loss": 0.008, "step": 18900 }, { "grad_norm": 0.14797940850257874, "learning_rate": 9.480346315415745e-05, "loss": 0.0083, "step": 18910 }, { "grad_norm": 0.1426113396883011, "learning_rate": 9.479612071606314e-05, "loss": 0.0085, "step": 18920 }, { "grad_norm": 0.15551571547985077, "learning_rate": 9.478877337913244e-05, "loss": 0.0083, "step": 18930 }, { "grad_norm": 0.09916377812623978, "learning_rate": 9.478142114416881e-05, "loss": 0.0089, "step": 18940 }, { "grad_norm": 0.14179790019989014, "learning_rate": 9.47740640119763e-05, "loss": 0.0071, "step": 18950 }, { "grad_norm": 0.11608278006315231, "learning_rate": 9.476670198335947e-05, "loss": 0.0076, "step": 18960 }, { "grad_norm": 0.13273201882839203, "learning_rate": 9.47593350591234e-05, "loss": 0.0081, "step": 18970 }, { "grad_norm": 0.10801196098327637, "learning_rate": 9.475196324007376e-05, "loss": 0.0079, "step": 18980 }, { "grad_norm": 0.13308371603488922, "learning_rate": 9.474458652701669e-05, "loss": 0.0083, "step": 18990 }, { "grad_norm": 0.10856131464242935, "learning_rate": 9.473720492075892e-05, "loss": 0.007, "step": 19000 }, { "grad_norm": 0.11607111990451813, "learning_rate": 9.472981842210768e-05, "loss": 0.0069, "step": 19010 }, { "grad_norm": 0.09356235712766647, "learning_rate": 9.472242703187074e-05, "loss": 0.0072, "step": 19020 }, { "grad_norm": 0.2380795031785965, "learning_rate": 9.471503075085643e-05, "loss": 0.0108, "step": 19030 }, { "grad_norm": 0.13703584671020508, "learning_rate": 9.470762957987359e-05, "loss": 0.0117, "step": 19040 }, { "grad_norm": 0.19109879434108734, "learning_rate": 9.470022351973158e-05, "loss": 0.0091, "step": 19050 }, { "grad_norm": 0.17585784196853638, "learning_rate": 9.469281257124034e-05, "loss": 0.0095, "step": 19060 }, { "grad_norm": 0.15536583960056305, "learning_rate": 9.46853967352103e-05, "loss": 0.0086, "step": 19070 }, { "grad_norm": 0.11134232580661774, "learning_rate": 9.467797601245246e-05, "loss": 0.0097, "step": 19080 }, { "grad_norm": 0.1551210582256317, "learning_rate": 9.467055040377834e-05, "loss": 0.0087, "step": 19090 }, { "grad_norm": 0.16228829324245453, "learning_rate": 9.466311990999999e-05, "loss": 0.0089, "step": 19100 }, { "grad_norm": 0.15209978818893433, "learning_rate": 9.465568453193e-05, "loss": 0.0086, "step": 19110 }, { "grad_norm": 0.1070411279797554, "learning_rate": 9.464824427038148e-05, "loss": 0.0079, "step": 19120 }, { "grad_norm": 0.23042872548103333, "learning_rate": 9.46407991261681e-05, "loss": 0.0086, "step": 19130 }, { "grad_norm": 0.17658883333206177, "learning_rate": 9.463334910010404e-05, "loss": 0.0099, "step": 19140 }, { "grad_norm": 0.19542217254638672, "learning_rate": 9.462589419300403e-05, "loss": 0.0112, "step": 19150 }, { "grad_norm": 0.1093064472079277, "learning_rate": 9.461843440568333e-05, "loss": 0.0105, "step": 19160 }, { "grad_norm": 0.175808385014534, "learning_rate": 9.461096973895773e-05, "loss": 0.009, "step": 19170 }, { "grad_norm": 0.12857168912887573, "learning_rate": 9.460350019364355e-05, "loss": 0.0077, "step": 19180 }, { "grad_norm": 0.08682992309331894, "learning_rate": 9.459602577055764e-05, "loss": 0.0071, "step": 19190 }, { "grad_norm": 0.17149613797664642, "learning_rate": 9.45885464705174e-05, "loss": 0.0084, "step": 19200 }, { "grad_norm": 0.17529410123825073, "learning_rate": 9.458106229434076e-05, "loss": 0.0086, "step": 19210 }, { "grad_norm": 0.1685234159231186, "learning_rate": 9.457357324284617e-05, "loss": 0.0096, "step": 19220 }, { "grad_norm": 0.090998575091362, "learning_rate": 9.456607931685262e-05, "loss": 0.0099, "step": 19230 }, { "grad_norm": 0.1380227506160736, "learning_rate": 9.455858051717965e-05, "loss": 0.0088, "step": 19240 }, { "grad_norm": 0.13827139139175415, "learning_rate": 9.45510768446473e-05, "loss": 0.0091, "step": 19250 }, { "grad_norm": 0.13042640686035156, "learning_rate": 9.454356830007618e-05, "loss": 0.0086, "step": 19260 }, { "grad_norm": 0.2090277373790741, "learning_rate": 9.45360548842874e-05, "loss": 0.0084, "step": 19270 }, { "grad_norm": 0.14055883884429932, "learning_rate": 9.452853659810261e-05, "loss": 0.0086, "step": 19280 }, { "grad_norm": 0.188359335064888, "learning_rate": 9.452101344234401e-05, "loss": 0.0084, "step": 19290 }, { "grad_norm": 0.17135590314865112, "learning_rate": 9.451348541783431e-05, "loss": 0.0085, "step": 19300 }, { "grad_norm": 0.10133565962314606, "learning_rate": 9.450595252539678e-05, "loss": 0.0086, "step": 19310 }, { "grad_norm": 0.15330299735069275, "learning_rate": 9.449841476585518e-05, "loss": 0.0075, "step": 19320 }, { "grad_norm": 0.09569143503904343, "learning_rate": 9.449087214003384e-05, "loss": 0.0076, "step": 19330 }, { "grad_norm": 0.1111653596162796, "learning_rate": 9.448332464875765e-05, "loss": 0.0079, "step": 19340 }, { "grad_norm": 0.11380594968795776, "learning_rate": 9.447577229285192e-05, "loss": 0.0079, "step": 19350 }, { "grad_norm": 0.1652001440525055, "learning_rate": 9.446821507314261e-05, "loss": 0.0065, "step": 19360 }, { "grad_norm": 0.11607524752616882, "learning_rate": 9.446065299045617e-05, "loss": 0.0083, "step": 19370 }, { "grad_norm": 0.13438302278518677, "learning_rate": 9.445308604561955e-05, "loss": 0.0088, "step": 19380 }, { "grad_norm": 0.1531267911195755, "learning_rate": 9.444551423946028e-05, "loss": 0.0071, "step": 19390 }, { "grad_norm": 0.12372606247663498, "learning_rate": 9.443793757280638e-05, "loss": 0.0069, "step": 19400 }, { "grad_norm": 0.12561491131782532, "learning_rate": 9.443035604648646e-05, "loss": 0.008, "step": 19410 }, { "grad_norm": 0.09942672401666641, "learning_rate": 9.44227696613296e-05, "loss": 0.0088, "step": 19420 }, { "grad_norm": 0.1050134152173996, "learning_rate": 9.441517841816542e-05, "loss": 0.0074, "step": 19430 }, { "grad_norm": 0.1193767562508583, "learning_rate": 9.440758231782413e-05, "loss": 0.0087, "step": 19440 }, { "grad_norm": 0.1424209326505661, "learning_rate": 9.439998136113639e-05, "loss": 0.0073, "step": 19450 }, { "grad_norm": 0.15197592973709106, "learning_rate": 9.439237554893344e-05, "loss": 0.0072, "step": 19460 }, { "grad_norm": 0.11494890600442886, "learning_rate": 9.438476488204705e-05, "loss": 0.0074, "step": 19470 }, { "grad_norm": 0.11468345671892166, "learning_rate": 9.43771493613095e-05, "loss": 0.0072, "step": 19480 }, { "grad_norm": 0.16724982857704163, "learning_rate": 9.436952898755362e-05, "loss": 0.0062, "step": 19490 }, { "grad_norm": 0.13402940332889557, "learning_rate": 9.436190376161276e-05, "loss": 0.0085, "step": 19500 }, { "grad_norm": 0.10340897738933563, "learning_rate": 9.43542736843208e-05, "loss": 0.0072, "step": 19510 }, { "grad_norm": 0.11101121455430984, "learning_rate": 9.434663875651216e-05, "loss": 0.0076, "step": 19520 }, { "grad_norm": 0.10719065368175507, "learning_rate": 9.433899897902177e-05, "loss": 0.0068, "step": 19530 }, { "grad_norm": 0.10222405195236206, "learning_rate": 9.433135435268511e-05, "loss": 0.0071, "step": 19540 }, { "grad_norm": 0.15366028249263763, "learning_rate": 9.432370487833819e-05, "loss": 0.0072, "step": 19550 }, { "grad_norm": 0.1478523313999176, "learning_rate": 9.431605055681756e-05, "loss": 0.007, "step": 19560 }, { "grad_norm": 0.1477160006761551, "learning_rate": 9.430839138896026e-05, "loss": 0.0071, "step": 19570 }, { "grad_norm": 0.1598418653011322, "learning_rate": 9.43007273756039e-05, "loss": 0.0078, "step": 19580 }, { "grad_norm": 0.11625345796346664, "learning_rate": 9.429305851758658e-05, "loss": 0.0079, "step": 19590 }, { "grad_norm": 0.10930033028125763, "learning_rate": 9.428538481574699e-05, "loss": 0.0068, "step": 19600 }, { "grad_norm": 0.10472577065229416, "learning_rate": 9.42777062709243e-05, "loss": 0.0083, "step": 19610 }, { "grad_norm": 0.18260663747787476, "learning_rate": 9.427002288395821e-05, "loss": 0.007, "step": 19620 }, { "grad_norm": 0.09345921128988266, "learning_rate": 9.426233465568898e-05, "loss": 0.0073, "step": 19630 }, { "grad_norm": 0.10851339995861053, "learning_rate": 9.42546415869574e-05, "loss": 0.0074, "step": 19640 }, { "grad_norm": 0.13508546352386475, "learning_rate": 9.424694367860473e-05, "loss": 0.0063, "step": 19650 }, { "grad_norm": 0.10466153919696808, "learning_rate": 9.423924093147284e-05, "loss": 0.0092, "step": 19660 }, { "grad_norm": 0.11782397329807281, "learning_rate": 9.423153334640407e-05, "loss": 0.008, "step": 19670 }, { "grad_norm": 0.10263889282941818, "learning_rate": 9.42238209242413e-05, "loss": 0.008, "step": 19680 }, { "grad_norm": 0.14214946329593658, "learning_rate": 9.421610366582798e-05, "loss": 0.0088, "step": 19690 }, { "grad_norm": 0.11207079142332077, "learning_rate": 9.420838157200803e-05, "loss": 0.0077, "step": 19700 }, { "grad_norm": 0.11347932368516922, "learning_rate": 9.420065464362594e-05, "loss": 0.0085, "step": 19710 }, { "grad_norm": 0.1530204713344574, "learning_rate": 9.419292288152673e-05, "loss": 0.0072, "step": 19720 }, { "grad_norm": 0.13823983073234558, "learning_rate": 9.418518628655588e-05, "loss": 0.0081, "step": 19730 }, { "grad_norm": 0.12390706688165665, "learning_rate": 9.417744485955951e-05, "loss": 0.0066, "step": 19740 }, { "grad_norm": 0.09203900396823883, "learning_rate": 9.41696986013842e-05, "loss": 0.0061, "step": 19750 }, { "grad_norm": 0.1112990453839302, "learning_rate": 9.416194751287705e-05, "loss": 0.0063, "step": 19760 }, { "grad_norm": 0.0897761657834053, "learning_rate": 9.415419159488572e-05, "loss": 0.0073, "step": 19770 }, { "grad_norm": 0.13800916075706482, "learning_rate": 9.414643084825837e-05, "loss": 0.0081, "step": 19780 }, { "grad_norm": 0.13831020891666412, "learning_rate": 9.413866527384372e-05, "loss": 0.0088, "step": 19790 }, { "grad_norm": 0.13664378225803375, "learning_rate": 9.4130894872491e-05, "loss": 0.0074, "step": 19800 }, { "grad_norm": 0.11432941257953644, "learning_rate": 9.412311964504998e-05, "loss": 0.0083, "step": 19810 }, { "grad_norm": 0.1291920244693756, "learning_rate": 9.411533959237091e-05, "loss": 0.0072, "step": 19820 }, { "grad_norm": 0.1186920627951622, "learning_rate": 9.410755471530464e-05, "loss": 0.0088, "step": 19830 }, { "grad_norm": 0.12768784165382385, "learning_rate": 9.40997650147025e-05, "loss": 0.0084, "step": 19840 }, { "grad_norm": 0.16056318581104279, "learning_rate": 9.409197049141637e-05, "loss": 0.0094, "step": 19850 }, { "grad_norm": 0.14612562954425812, "learning_rate": 9.408417114629863e-05, "loss": 0.0078, "step": 19860 }, { "grad_norm": 0.13058456778526306, "learning_rate": 9.40763669802022e-05, "loss": 0.0076, "step": 19870 }, { "grad_norm": 0.09132331609725952, "learning_rate": 9.406855799398056e-05, "loss": 0.007, "step": 19880 }, { "grad_norm": 0.14750918745994568, "learning_rate": 9.406074418848767e-05, "loss": 0.0081, "step": 19890 }, { "grad_norm": 0.17224977910518646, "learning_rate": 9.405292556457805e-05, "loss": 0.0092, "step": 19900 }, { "grad_norm": 0.1265760213136673, "learning_rate": 9.404510212310671e-05, "loss": 0.0097, "step": 19910 }, { "grad_norm": 0.1347476691007614, "learning_rate": 9.403727386492924e-05, "loss": 0.0077, "step": 19920 }, { "grad_norm": 0.11661148071289062, "learning_rate": 9.40294407909017e-05, "loss": 0.0074, "step": 19930 }, { "grad_norm": 0.11266865581274033, "learning_rate": 9.40216029018807e-05, "loss": 0.0075, "step": 19940 }, { "grad_norm": 0.14582586288452148, "learning_rate": 9.401376019872338e-05, "loss": 0.0072, "step": 19950 }, { "grad_norm": 0.08987518399953842, "learning_rate": 9.400591268228746e-05, "loss": 0.0069, "step": 19960 }, { "grad_norm": 0.12319789081811905, "learning_rate": 9.399806035343106e-05, "loss": 0.0062, "step": 19970 }, { "grad_norm": 0.11836206167936325, "learning_rate": 9.399020321301294e-05, "loss": 0.0064, "step": 19980 }, { "grad_norm": 0.10243524610996246, "learning_rate": 9.398234126189234e-05, "loss": 0.0076, "step": 19990 }, { "grad_norm": 0.1502894163131714, "learning_rate": 9.397447450092902e-05, "loss": 0.008, "step": 20000 }, { "grad_norm": 0.14128822088241577, "learning_rate": 9.39666029309833e-05, "loss": 0.0075, "step": 20010 }, { "grad_norm": 0.14136959612369537, "learning_rate": 9.395872655291596e-05, "loss": 0.0083, "step": 20020 }, { "grad_norm": 0.14235550165176392, "learning_rate": 9.395084536758838e-05, "loss": 0.0075, "step": 20030 }, { "grad_norm": 0.24401704967021942, "learning_rate": 9.394295937586243e-05, "loss": 0.01, "step": 20040 }, { "grad_norm": 0.16308464109897614, "learning_rate": 9.393506857860052e-05, "loss": 0.0091, "step": 20050 }, { "grad_norm": 0.10798946022987366, "learning_rate": 9.392717297666555e-05, "loss": 0.008, "step": 20060 }, { "grad_norm": 0.08800308406352997, "learning_rate": 9.391927257092101e-05, "loss": 0.0076, "step": 20070 }, { "grad_norm": 0.1264796108007431, "learning_rate": 9.391136736223085e-05, "loss": 0.0083, "step": 20080 }, { "grad_norm": 0.10535486042499542, "learning_rate": 9.390345735145956e-05, "loss": 0.0073, "step": 20090 }, { "grad_norm": 0.10749930888414383, "learning_rate": 9.389554253947219e-05, "loss": 0.0074, "step": 20100 }, { "grad_norm": 0.11634404212236404, "learning_rate": 9.388762292713428e-05, "loss": 0.0069, "step": 20110 }, { "grad_norm": 0.14203289151191711, "learning_rate": 9.38796985153119e-05, "loss": 0.0077, "step": 20120 }, { "grad_norm": 0.13517846167087555, "learning_rate": 9.387176930487169e-05, "loss": 0.0083, "step": 20130 }, { "grad_norm": 0.1330045908689499, "learning_rate": 9.386383529668072e-05, "loss": 0.0071, "step": 20140 }, { "grad_norm": 0.16529181599617004, "learning_rate": 9.385589649160669e-05, "loss": 0.0092, "step": 20150 }, { "grad_norm": 0.11112305521965027, "learning_rate": 9.384795289051775e-05, "loss": 0.008, "step": 20160 }, { "grad_norm": 0.15982893109321594, "learning_rate": 9.384000449428261e-05, "loss": 0.0075, "step": 20170 }, { "grad_norm": 0.17696599662303925, "learning_rate": 9.383205130377048e-05, "loss": 0.0083, "step": 20180 }, { "grad_norm": 0.1189337819814682, "learning_rate": 9.382409331985114e-05, "loss": 0.0076, "step": 20190 }, { "grad_norm": 0.15061672031879425, "learning_rate": 9.381613054339482e-05, "loss": 0.0074, "step": 20200 }, { "grad_norm": 0.12089713662862778, "learning_rate": 9.380816297527235e-05, "loss": 0.0076, "step": 20210 }, { "grad_norm": 0.10467564314603806, "learning_rate": 9.380019061635506e-05, "loss": 0.0077, "step": 20220 }, { "grad_norm": 0.1414906531572342, "learning_rate": 9.379221346751474e-05, "loss": 0.0062, "step": 20230 }, { "grad_norm": 0.20147131383419037, "learning_rate": 9.378423152962382e-05, "loss": 0.0073, "step": 20240 }, { "grad_norm": 0.20441798865795135, "learning_rate": 9.377624480355517e-05, "loss": 0.0074, "step": 20250 }, { "grad_norm": 0.11190611869096756, "learning_rate": 9.376825329018219e-05, "loss": 0.0072, "step": 20260 }, { "grad_norm": 0.13959245383739471, "learning_rate": 9.376025699037884e-05, "loss": 0.0075, "step": 20270 }, { "grad_norm": 0.12817493081092834, "learning_rate": 9.37522559050196e-05, "loss": 0.0065, "step": 20280 }, { "grad_norm": 0.13122199475765228, "learning_rate": 9.37442500349794e-05, "loss": 0.0073, "step": 20290 }, { "grad_norm": 0.1477041095495224, "learning_rate": 9.373623938113381e-05, "loss": 0.0072, "step": 20300 }, { "grad_norm": 0.11673248559236526, "learning_rate": 9.372822394435883e-05, "loss": 0.0079, "step": 20310 }, { "grad_norm": 0.09457863122224808, "learning_rate": 9.372020372553102e-05, "loss": 0.0066, "step": 20320 }, { "grad_norm": 0.0935785174369812, "learning_rate": 9.371217872552746e-05, "loss": 0.0064, "step": 20330 }, { "grad_norm": 0.10988721996545792, "learning_rate": 9.370414894522576e-05, "loss": 0.0066, "step": 20340 }, { "grad_norm": 0.10432510823011398, "learning_rate": 9.369611438550406e-05, "loss": 0.0068, "step": 20350 }, { "grad_norm": 0.12482372671365738, "learning_rate": 9.368807504724095e-05, "loss": 0.0076, "step": 20360 }, { "grad_norm": 0.12161969393491745, "learning_rate": 9.368003093131565e-05, "loss": 0.0082, "step": 20370 }, { "grad_norm": 0.13112609088420868, "learning_rate": 9.367198203860785e-05, "loss": 0.0077, "step": 20380 }, { "grad_norm": 0.14587005972862244, "learning_rate": 9.366392836999774e-05, "loss": 0.0075, "step": 20390 }, { "grad_norm": 0.14353275299072266, "learning_rate": 9.365586992636607e-05, "loss": 0.0072, "step": 20400 }, { "grad_norm": 0.1274472326040268, "learning_rate": 9.364780670859412e-05, "loss": 0.0075, "step": 20410 }, { "grad_norm": 0.125466451048851, "learning_rate": 9.363973871756364e-05, "loss": 0.0078, "step": 20420 }, { "grad_norm": 0.10203810781240463, "learning_rate": 9.363166595415696e-05, "loss": 0.007, "step": 20430 }, { "grad_norm": 0.11355553567409515, "learning_rate": 9.362358841925686e-05, "loss": 0.0083, "step": 20440 }, { "grad_norm": 0.14952437579631805, "learning_rate": 9.361550611374674e-05, "loss": 0.0071, "step": 20450 }, { "grad_norm": 0.13082866370677948, "learning_rate": 9.360741903851043e-05, "loss": 0.0067, "step": 20460 }, { "grad_norm": 0.08836882561445236, "learning_rate": 9.359932719443236e-05, "loss": 0.0075, "step": 20470 }, { "grad_norm": 0.09794595092535019, "learning_rate": 9.35912305823974e-05, "loss": 0.0078, "step": 20480 }, { "grad_norm": 0.1548624485731125, "learning_rate": 9.358312920329101e-05, "loss": 0.0079, "step": 20490 }, { "grad_norm": 0.14379706978797913, "learning_rate": 9.357502305799914e-05, "loss": 0.0071, "step": 20500 }, { "grad_norm": 0.12134488672018051, "learning_rate": 9.356691214740824e-05, "loss": 0.0071, "step": 20510 }, { "grad_norm": 0.08011563867330551, "learning_rate": 9.355879647240535e-05, "loss": 0.0076, "step": 20520 }, { "grad_norm": 0.07489188015460968, "learning_rate": 9.355067603387798e-05, "loss": 0.0073, "step": 20530 }, { "grad_norm": 0.10285583883523941, "learning_rate": 9.354255083271412e-05, "loss": 0.0064, "step": 20540 }, { "grad_norm": 0.09083209931850433, "learning_rate": 9.353442086980239e-05, "loss": 0.007, "step": 20550 }, { "grad_norm": 0.13494722545146942, "learning_rate": 9.352628614603185e-05, "loss": 0.0095, "step": 20560 }, { "grad_norm": 0.130925714969635, "learning_rate": 9.351814666229209e-05, "loss": 0.0074, "step": 20570 }, { "grad_norm": 0.1225721463561058, "learning_rate": 9.351000241947324e-05, "loss": 0.0069, "step": 20580 }, { "grad_norm": 0.1392023265361786, "learning_rate": 9.350185341846594e-05, "loss": 0.0081, "step": 20590 }, { "grad_norm": 0.1362086832523346, "learning_rate": 9.349369966016134e-05, "loss": 0.0072, "step": 20600 }, { "grad_norm": 0.11304633319377899, "learning_rate": 9.348554114545117e-05, "loss": 0.0065, "step": 20610 }, { "grad_norm": 0.16017937660217285, "learning_rate": 9.347737787522758e-05, "loss": 0.007, "step": 20620 }, { "grad_norm": 0.12239007651805878, "learning_rate": 9.346920985038332e-05, "loss": 0.0075, "step": 20630 }, { "grad_norm": 0.15071211755275726, "learning_rate": 9.346103707181162e-05, "loss": 0.008, "step": 20640 }, { "grad_norm": 0.16227471828460693, "learning_rate": 9.345285954040626e-05, "loss": 0.0082, "step": 20650 }, { "grad_norm": 0.15819577872753143, "learning_rate": 9.34446772570615e-05, "loss": 0.0083, "step": 20660 }, { "grad_norm": 0.1341995745897293, "learning_rate": 9.343649022267214e-05, "loss": 0.0078, "step": 20670 }, { "grad_norm": 0.14424344897270203, "learning_rate": 9.342829843813353e-05, "loss": 0.0072, "step": 20680 }, { "grad_norm": 0.08249339461326599, "learning_rate": 9.342010190434149e-05, "loss": 0.0077, "step": 20690 }, { "grad_norm": 0.11829278618097305, "learning_rate": 9.34119006221924e-05, "loss": 0.0081, "step": 20700 }, { "grad_norm": 0.12789520621299744, "learning_rate": 9.340369459258313e-05, "loss": 0.0067, "step": 20710 }, { "grad_norm": 0.11479116976261139, "learning_rate": 9.339548381641106e-05, "loss": 0.007, "step": 20720 }, { "grad_norm": 0.14643655717372894, "learning_rate": 9.338726829457413e-05, "loss": 0.0066, "step": 20730 }, { "grad_norm": 0.10496355593204498, "learning_rate": 9.337904802797078e-05, "loss": 0.0083, "step": 20740 }, { "grad_norm": 0.13081300258636475, "learning_rate": 9.337082301749993e-05, "loss": 0.0075, "step": 20750 }, { "grad_norm": 0.1083369255065918, "learning_rate": 9.336259326406109e-05, "loss": 0.0074, "step": 20760 }, { "grad_norm": 0.09888697415590286, "learning_rate": 9.335435876855427e-05, "loss": 0.0067, "step": 20770 }, { "grad_norm": 0.14218077063560486, "learning_rate": 9.334611953187994e-05, "loss": 0.0067, "step": 20780 }, { "grad_norm": 0.07730615139007568, "learning_rate": 9.333787555493914e-05, "loss": 0.0069, "step": 20790 }, { "grad_norm": 0.10251544415950775, "learning_rate": 9.332962683863345e-05, "loss": 0.0068, "step": 20800 }, { "grad_norm": 0.1422746777534485, "learning_rate": 9.332137338386489e-05, "loss": 0.0073, "step": 20810 }, { "grad_norm": 0.1402711123228073, "learning_rate": 9.33131151915361e-05, "loss": 0.0078, "step": 20820 }, { "grad_norm": 0.1542763113975525, "learning_rate": 9.330485226255012e-05, "loss": 0.0075, "step": 20830 }, { "grad_norm": 0.11941017210483551, "learning_rate": 9.329658459781061e-05, "loss": 0.0073, "step": 20840 }, { "grad_norm": 0.12566395103931427, "learning_rate": 9.328831219822172e-05, "loss": 0.0074, "step": 20850 }, { "grad_norm": 0.17685844004154205, "learning_rate": 9.328003506468808e-05, "loss": 0.0064, "step": 20860 }, { "grad_norm": 0.07667388767004013, "learning_rate": 9.327175319811488e-05, "loss": 0.0071, "step": 20870 }, { "grad_norm": 0.14419938623905182, "learning_rate": 9.326346659940781e-05, "loss": 0.0069, "step": 20880 }, { "grad_norm": 0.14471180737018585, "learning_rate": 9.325517526947308e-05, "loss": 0.0083, "step": 20890 }, { "grad_norm": 0.2358313500881195, "learning_rate": 9.32468792092174e-05, "loss": 0.0069, "step": 20900 }, { "grad_norm": 0.1578315794467926, "learning_rate": 9.323857841954803e-05, "loss": 0.0067, "step": 20910 }, { "grad_norm": 0.1503359079360962, "learning_rate": 9.323027290137276e-05, "loss": 0.0086, "step": 20920 }, { "grad_norm": 0.12434159964323044, "learning_rate": 9.322196265559981e-05, "loss": 0.0088, "step": 20930 }, { "grad_norm": 0.09981242567300797, "learning_rate": 9.321364768313803e-05, "loss": 0.0077, "step": 20940 }, { "grad_norm": 0.18727363646030426, "learning_rate": 9.32053279848967e-05, "loss": 0.0091, "step": 20950 }, { "grad_norm": 0.09554339945316315, "learning_rate": 9.319700356178567e-05, "loss": 0.0073, "step": 20960 }, { "grad_norm": 0.13987280428409576, "learning_rate": 9.318867441471527e-05, "loss": 0.0082, "step": 20970 }, { "grad_norm": 0.16900160908699036, "learning_rate": 9.318034054459637e-05, "loss": 0.0076, "step": 20980 }, { "grad_norm": 0.12430787086486816, "learning_rate": 9.317200195234034e-05, "loss": 0.0083, "step": 20990 }, { "grad_norm": 0.14867861568927765, "learning_rate": 9.316365863885909e-05, "loss": 0.0062, "step": 21000 }, { "grad_norm": 0.1343913972377777, "learning_rate": 9.315531060506502e-05, "loss": 0.0074, "step": 21010 }, { "grad_norm": 0.14653584361076355, "learning_rate": 9.314695785187108e-05, "loss": 0.0074, "step": 21020 }, { "grad_norm": 0.12844029068946838, "learning_rate": 9.313860038019069e-05, "loss": 0.0091, "step": 21030 }, { "grad_norm": 0.10761989653110504, "learning_rate": 9.313023819093782e-05, "loss": 0.0098, "step": 21040 }, { "grad_norm": 0.13045068085193634, "learning_rate": 9.312187128502695e-05, "loss": 0.007, "step": 21050 }, { "grad_norm": 0.08210019022226334, "learning_rate": 9.311349966337307e-05, "loss": 0.0073, "step": 21060 }, { "grad_norm": 0.1156170442700386, "learning_rate": 9.310512332689169e-05, "loss": 0.0072, "step": 21070 }, { "grad_norm": 0.1811123639345169, "learning_rate": 9.309674227649883e-05, "loss": 0.0071, "step": 21080 }, { "grad_norm": 0.14447921514511108, "learning_rate": 9.308835651311103e-05, "loss": 0.0072, "step": 21090 }, { "grad_norm": 0.18697214126586914, "learning_rate": 9.307996603764533e-05, "loss": 0.0082, "step": 21100 }, { "grad_norm": 0.08340372890233994, "learning_rate": 9.307157085101932e-05, "loss": 0.0072, "step": 21110 }, { "grad_norm": 0.12368609756231308, "learning_rate": 9.306317095415109e-05, "loss": 0.0066, "step": 21120 }, { "grad_norm": 0.15306681394577026, "learning_rate": 9.305476634795922e-05, "loss": 0.0086, "step": 21130 }, { "grad_norm": 0.148601233959198, "learning_rate": 9.304635703336284e-05, "loss": 0.0082, "step": 21140 }, { "grad_norm": 0.11589232832193375, "learning_rate": 9.303794301128157e-05, "loss": 0.0078, "step": 21150 }, { "grad_norm": 0.11021063476800919, "learning_rate": 9.302952428263555e-05, "loss": 0.008, "step": 21160 }, { "grad_norm": 0.1503002643585205, "learning_rate": 9.302110084834545e-05, "loss": 0.0076, "step": 21170 }, { "grad_norm": 0.08188364654779434, "learning_rate": 9.301267270933245e-05, "loss": 0.0069, "step": 21180 }, { "grad_norm": 0.17055700719356537, "learning_rate": 9.300423986651823e-05, "loss": 0.0091, "step": 21190 }, { "grad_norm": 0.12039832770824432, "learning_rate": 9.299580232082501e-05, "loss": 0.0076, "step": 21200 }, { "grad_norm": 0.1330139935016632, "learning_rate": 9.298736007317547e-05, "loss": 0.009, "step": 21210 }, { "grad_norm": 0.1539038121700287, "learning_rate": 9.297891312449288e-05, "loss": 0.0125, "step": 21220 }, { "grad_norm": 0.20794636011123657, "learning_rate": 9.297046147570094e-05, "loss": 0.0095, "step": 21230 }, { "grad_norm": 0.09998442977666855, "learning_rate": 9.296200512772396e-05, "loss": 0.0093, "step": 21240 }, { "grad_norm": 0.10732529312372208, "learning_rate": 9.295354408148668e-05, "loss": 0.0092, "step": 21250 }, { "grad_norm": 0.12620142102241516, "learning_rate": 9.294507833791441e-05, "loss": 0.0082, "step": 21260 }, { "grad_norm": 0.15387976169586182, "learning_rate": 9.293660789793295e-05, "loss": 0.0091, "step": 21270 }, { "grad_norm": 0.10075311362743378, "learning_rate": 9.292813276246858e-05, "loss": 0.0082, "step": 21280 }, { "grad_norm": 0.11194463074207306, "learning_rate": 9.291965293244816e-05, "loss": 0.0073, "step": 21290 }, { "grad_norm": 0.13380561769008636, "learning_rate": 9.291116840879904e-05, "loss": 0.0076, "step": 21300 }, { "grad_norm": 0.09116257727146149, "learning_rate": 9.290267919244904e-05, "loss": 0.0072, "step": 21310 }, { "grad_norm": 0.15758153796195984, "learning_rate": 9.289418528432655e-05, "loss": 0.0083, "step": 21320 }, { "grad_norm": 0.2175915539264679, "learning_rate": 9.288568668536045e-05, "loss": 0.0079, "step": 21330 }, { "grad_norm": 0.19600960612297058, "learning_rate": 9.287718339648013e-05, "loss": 0.0096, "step": 21340 }, { "grad_norm": 0.1651330590248108, "learning_rate": 9.28686754186155e-05, "loss": 0.0122, "step": 21350 }, { "grad_norm": 0.16764573752880096, "learning_rate": 9.286016275269698e-05, "loss": 0.0102, "step": 21360 }, { "grad_norm": 0.15889011323451996, "learning_rate": 9.285164539965551e-05, "loss": 0.0096, "step": 21370 }, { "grad_norm": 0.11980371922254562, "learning_rate": 9.284312336042251e-05, "loss": 0.0089, "step": 21380 }, { "grad_norm": 0.16444161534309387, "learning_rate": 9.283459663592996e-05, "loss": 0.0098, "step": 21390 }, { "grad_norm": 0.19708524644374847, "learning_rate": 9.282606522711033e-05, "loss": 0.0083, "step": 21400 }, { "grad_norm": 0.15990087389945984, "learning_rate": 9.281752913489657e-05, "loss": 0.0089, "step": 21410 }, { "grad_norm": 0.15081170201301575, "learning_rate": 9.280898836022222e-05, "loss": 0.0096, "step": 21420 }, { "grad_norm": 0.10112395137548447, "learning_rate": 9.280044290402126e-05, "loss": 0.0085, "step": 21430 }, { "grad_norm": 0.12567442655563354, "learning_rate": 9.279189276722821e-05, "loss": 0.0077, "step": 21440 }, { "grad_norm": 0.16901089251041412, "learning_rate": 9.278333795077812e-05, "loss": 0.0091, "step": 21450 }, { "grad_norm": 0.12467736750841141, "learning_rate": 9.27747784556065e-05, "loss": 0.0088, "step": 21460 }, { "grad_norm": 0.12175596505403519, "learning_rate": 9.276621428264942e-05, "loss": 0.0068, "step": 21470 }, { "grad_norm": 0.09606867283582687, "learning_rate": 9.275764543284345e-05, "loss": 0.0072, "step": 21480 }, { "grad_norm": 0.09096445888280869, "learning_rate": 9.274907190712566e-05, "loss": 0.0073, "step": 21490 }, { "grad_norm": 0.12170467525720596, "learning_rate": 9.274049370643363e-05, "loss": 0.0078, "step": 21500 }, { "grad_norm": 0.13982516527175903, "learning_rate": 9.273191083170547e-05, "loss": 0.0076, "step": 21510 }, { "grad_norm": 0.15513375401496887, "learning_rate": 9.27233232838798e-05, "loss": 0.0079, "step": 21520 }, { "grad_norm": 0.12050507217645645, "learning_rate": 9.27147310638957e-05, "loss": 0.0071, "step": 21530 }, { "grad_norm": 0.1490347534418106, "learning_rate": 9.270613417269286e-05, "loss": 0.0074, "step": 21540 }, { "grad_norm": 0.1498323380947113, "learning_rate": 9.269753261121138e-05, "loss": 0.0076, "step": 21550 }, { "grad_norm": 0.10802365839481354, "learning_rate": 9.268892638039194e-05, "loss": 0.0064, "step": 21560 }, { "grad_norm": 0.143743634223938, "learning_rate": 9.268031548117569e-05, "loss": 0.0073, "step": 21570 }, { "grad_norm": 0.13859491050243378, "learning_rate": 9.26716999145043e-05, "loss": 0.0079, "step": 21580 }, { "grad_norm": 0.12622381746768951, "learning_rate": 9.266307968131998e-05, "loss": 0.0084, "step": 21590 }, { "grad_norm": 0.11902432143688202, "learning_rate": 9.26544547825654e-05, "loss": 0.0077, "step": 21600 }, { "grad_norm": 0.1010633260011673, "learning_rate": 9.264582521918376e-05, "loss": 0.0081, "step": 21610 }, { "grad_norm": 0.167796790599823, "learning_rate": 9.263719099211881e-05, "loss": 0.0091, "step": 21620 }, { "grad_norm": 0.1863236278295517, "learning_rate": 9.262855210231476e-05, "loss": 0.0078, "step": 21630 }, { "grad_norm": 0.12193603068590164, "learning_rate": 9.261990855071633e-05, "loss": 0.0107, "step": 21640 }, { "grad_norm": 0.11495644599199295, "learning_rate": 9.261126033826878e-05, "loss": 0.0086, "step": 21650 }, { "grad_norm": 0.10192105174064636, "learning_rate": 9.260260746591786e-05, "loss": 0.0085, "step": 21660 }, { "grad_norm": 0.10683012753725052, "learning_rate": 9.259394993460985e-05, "loss": 0.0079, "step": 21670 }, { "grad_norm": 0.15190573036670685, "learning_rate": 9.258528774529151e-05, "loss": 0.0083, "step": 21680 }, { "grad_norm": 0.11167111992835999, "learning_rate": 9.257662089891013e-05, "loss": 0.0089, "step": 21690 }, { "grad_norm": 0.10680805891752243, "learning_rate": 9.25679493964135e-05, "loss": 0.0071, "step": 21700 }, { "grad_norm": 0.09811244159936905, "learning_rate": 9.255927323874994e-05, "loss": 0.0072, "step": 21710 }, { "grad_norm": 0.14014041423797607, "learning_rate": 9.255059242686822e-05, "loss": 0.0063, "step": 21720 }, { "grad_norm": 0.10722103714942932, "learning_rate": 9.254190696171769e-05, "loss": 0.0064, "step": 21730 }, { "grad_norm": 0.1119316816329956, "learning_rate": 9.25332168442482e-05, "loss": 0.007, "step": 21740 }, { "grad_norm": 0.10332874953746796, "learning_rate": 9.252452207541004e-05, "loss": 0.0071, "step": 21750 }, { "grad_norm": 0.10138724744319916, "learning_rate": 9.251582265615409e-05, "loss": 0.0069, "step": 21760 }, { "grad_norm": 0.12085055559873581, "learning_rate": 9.250711858743169e-05, "loss": 0.0069, "step": 21770 }, { "grad_norm": 0.12247183918952942, "learning_rate": 9.24984098701947e-05, "loss": 0.0091, "step": 21780 }, { "grad_norm": 0.10022421181201935, "learning_rate": 9.248969650539552e-05, "loss": 0.007, "step": 21790 }, { "grad_norm": 0.12927724421024323, "learning_rate": 9.2480978493987e-05, "loss": 0.0079, "step": 21800 }, { "grad_norm": 0.17625375092029572, "learning_rate": 9.247225583692256e-05, "loss": 0.0077, "step": 21810 }, { "grad_norm": 0.12050217390060425, "learning_rate": 9.246352853515607e-05, "loss": 0.0067, "step": 21820 }, { "grad_norm": 0.0969705581665039, "learning_rate": 9.245479658964194e-05, "loss": 0.0071, "step": 21830 }, { "grad_norm": 0.08563584089279175, "learning_rate": 9.244606000133507e-05, "loss": 0.0061, "step": 21840 }, { "grad_norm": 0.12096358835697174, "learning_rate": 9.24373187711909e-05, "loss": 0.007, "step": 21850 }, { "grad_norm": 0.08822088688611984, "learning_rate": 9.242857290016537e-05, "loss": 0.007, "step": 21860 }, { "grad_norm": 0.0846429318189621, "learning_rate": 9.241982238921488e-05, "loss": 0.0064, "step": 21870 }, { "grad_norm": 0.07728340476751328, "learning_rate": 9.24110672392964e-05, "loss": 0.0071, "step": 21880 }, { "grad_norm": 0.11628582328557968, "learning_rate": 9.240230745136737e-05, "loss": 0.0067, "step": 21890 }, { "grad_norm": 0.10900221019983292, "learning_rate": 9.239354302638575e-05, "loss": 0.0069, "step": 21900 }, { "grad_norm": 0.12021118402481079, "learning_rate": 9.238477396531e-05, "loss": 0.0066, "step": 21910 }, { "grad_norm": 0.12571614980697632, "learning_rate": 9.23760002690991e-05, "loss": 0.007, "step": 21920 }, { "grad_norm": 0.15028522908687592, "learning_rate": 9.236722193871252e-05, "loss": 0.0071, "step": 21930 }, { "grad_norm": 0.10998917371034622, "learning_rate": 9.235843897511023e-05, "loss": 0.0069, "step": 21940 }, { "grad_norm": 0.12430863827466965, "learning_rate": 9.234965137925276e-05, "loss": 0.0074, "step": 21950 }, { "grad_norm": 0.10918103903532028, "learning_rate": 9.234085915210108e-05, "loss": 0.0073, "step": 21960 }, { "grad_norm": 0.13923048973083496, "learning_rate": 9.23320622946167e-05, "loss": 0.0069, "step": 21970 }, { "grad_norm": 0.1024337112903595, "learning_rate": 9.232326080776163e-05, "loss": 0.0065, "step": 21980 }, { "grad_norm": 0.1542663723230362, "learning_rate": 9.23144546924984e-05, "loss": 0.0078, "step": 21990 }, { "grad_norm": 0.09658179432153702, "learning_rate": 9.230564394979e-05, "loss": 0.0073, "step": 22000 }, { "grad_norm": 0.1438750922679901, "learning_rate": 9.22968285806e-05, "loss": 0.0062, "step": 22010 }, { "grad_norm": 0.11057587713003159, "learning_rate": 9.228800858589242e-05, "loss": 0.007, "step": 22020 }, { "grad_norm": 0.08854331821203232, "learning_rate": 9.227918396663179e-05, "loss": 0.0073, "step": 22030 }, { "grad_norm": 0.11618909984827042, "learning_rate": 9.227035472378319e-05, "loss": 0.0069, "step": 22040 }, { "grad_norm": 0.1549559235572815, "learning_rate": 9.226152085831213e-05, "loss": 0.0072, "step": 22050 }, { "grad_norm": 0.09236685931682587, "learning_rate": 9.22526823711847e-05, "loss": 0.0079, "step": 22060 }, { "grad_norm": 0.11880312860012054, "learning_rate": 9.224383926336745e-05, "loss": 0.0067, "step": 22070 }, { "grad_norm": 0.14421910047531128, "learning_rate": 9.223499153582744e-05, "loss": 0.009, "step": 22080 }, { "grad_norm": 0.11966728419065475, "learning_rate": 9.222613918953226e-05, "loss": 0.007, "step": 22090 }, { "grad_norm": 0.15311677753925323, "learning_rate": 9.221728222544999e-05, "loss": 0.0063, "step": 22100 }, { "grad_norm": 0.13177117705345154, "learning_rate": 9.22084206445492e-05, "loss": 0.0076, "step": 22110 }, { "grad_norm": 0.12515699863433838, "learning_rate": 9.2199554447799e-05, "loss": 0.0084, "step": 22120 }, { "grad_norm": 0.13192059099674225, "learning_rate": 9.219068363616897e-05, "loss": 0.0073, "step": 22130 }, { "grad_norm": 0.17965424060821533, "learning_rate": 9.218180821062919e-05, "loss": 0.0092, "step": 22140 }, { "grad_norm": 0.17254379391670227, "learning_rate": 9.21729281721503e-05, "loss": 0.0084, "step": 22150 }, { "grad_norm": 0.1830468773841858, "learning_rate": 9.216404352170339e-05, "loss": 0.0085, "step": 22160 }, { "grad_norm": 0.1275540143251419, "learning_rate": 9.215515426026007e-05, "loss": 0.008, "step": 22170 }, { "grad_norm": 0.12389069050550461, "learning_rate": 9.214626038879246e-05, "loss": 0.0077, "step": 22180 }, { "grad_norm": 0.07980254292488098, "learning_rate": 9.21373619082732e-05, "loss": 0.0076, "step": 22190 }, { "grad_norm": 0.08347667753696442, "learning_rate": 9.212845881967535e-05, "loss": 0.0069, "step": 22200 }, { "grad_norm": 0.11371288448572159, "learning_rate": 9.211955112397262e-05, "loss": 0.0074, "step": 22210 }, { "grad_norm": 0.10790517926216125, "learning_rate": 9.211063882213909e-05, "loss": 0.0072, "step": 22220 }, { "grad_norm": 0.15985897183418274, "learning_rate": 9.210172191514942e-05, "loss": 0.0071, "step": 22230 }, { "grad_norm": 0.11719472706317902, "learning_rate": 9.209280040397874e-05, "loss": 0.0074, "step": 22240 }, { "grad_norm": 0.15046648681163788, "learning_rate": 9.208387428960268e-05, "loss": 0.0074, "step": 22250 }, { "grad_norm": 0.09171905368566513, "learning_rate": 9.20749435729974e-05, "loss": 0.0064, "step": 22260 }, { "grad_norm": 0.10746438801288605, "learning_rate": 9.206600825513957e-05, "loss": 0.0083, "step": 22270 }, { "grad_norm": 0.10047658532857895, "learning_rate": 9.20570683370063e-05, "loss": 0.0075, "step": 22280 }, { "grad_norm": 0.14921371638774872, "learning_rate": 9.204812381957528e-05, "loss": 0.0083, "step": 22290 }, { "grad_norm": 0.10902568697929382, "learning_rate": 9.203917470382465e-05, "loss": 0.0062, "step": 22300 }, { "grad_norm": 0.12171252816915512, "learning_rate": 9.203022099073309e-05, "loss": 0.0066, "step": 22310 }, { "grad_norm": 0.11005423963069916, "learning_rate": 9.202126268127976e-05, "loss": 0.0067, "step": 22320 }, { "grad_norm": 0.11299777030944824, "learning_rate": 9.20122997764443e-05, "loss": 0.0073, "step": 22330 }, { "grad_norm": 0.11440033465623856, "learning_rate": 9.200333227720692e-05, "loss": 0.0084, "step": 22340 }, { "grad_norm": 0.18772441148757935, "learning_rate": 9.199436018454826e-05, "loss": 0.0082, "step": 22350 }, { "grad_norm": 0.12343790382146835, "learning_rate": 9.198538349944952e-05, "loss": 0.008, "step": 22360 }, { "grad_norm": 0.17618004977703094, "learning_rate": 9.197640222289234e-05, "loss": 0.0073, "step": 22370 }, { "grad_norm": 0.12779688835144043, "learning_rate": 9.196741635585895e-05, "loss": 0.0066, "step": 22380 }, { "grad_norm": 0.14755311608314514, "learning_rate": 9.195842589933199e-05, "loss": 0.0067, "step": 22390 }, { "grad_norm": 0.15725693106651306, "learning_rate": 9.194943085429466e-05, "loss": 0.0066, "step": 22400 }, { "grad_norm": 0.12302956730127335, "learning_rate": 9.194043122173065e-05, "loss": 0.0081, "step": 22410 }, { "grad_norm": 0.15167434513568878, "learning_rate": 9.193142700262413e-05, "loss": 0.008, "step": 22420 }, { "grad_norm": 0.1544055938720703, "learning_rate": 9.192241819795979e-05, "loss": 0.0081, "step": 22430 }, { "grad_norm": 0.10676046460866928, "learning_rate": 9.191340480872284e-05, "loss": 0.0069, "step": 22440 }, { "grad_norm": 0.13349667191505432, "learning_rate": 9.190438683589895e-05, "loss": 0.0083, "step": 22450 }, { "grad_norm": 0.0899505689740181, "learning_rate": 9.189536428047432e-05, "loss": 0.0083, "step": 22460 }, { "grad_norm": 0.11108796298503876, "learning_rate": 9.188633714343564e-05, "loss": 0.0074, "step": 22470 }, { "grad_norm": 0.08933752775192261, "learning_rate": 9.18773054257701e-05, "loss": 0.0061, "step": 22480 }, { "grad_norm": 0.13111068308353424, "learning_rate": 9.18682691284654e-05, "loss": 0.0064, "step": 22490 }, { "grad_norm": 0.11225742846727371, "learning_rate": 9.185922825250974e-05, "loss": 0.0081, "step": 22500 }, { "grad_norm": 0.17263875901699066, "learning_rate": 9.185018279889181e-05, "loss": 0.0087, "step": 22510 }, { "grad_norm": 0.19701725244522095, "learning_rate": 9.184113276860082e-05, "loss": 0.0088, "step": 22520 }, { "grad_norm": 0.11526130139827728, "learning_rate": 9.183207816262645e-05, "loss": 0.0074, "step": 22530 }, { "grad_norm": 0.16894254088401794, "learning_rate": 9.182301898195891e-05, "loss": 0.0076, "step": 22540 }, { "grad_norm": 0.12613053619861603, "learning_rate": 9.181395522758889e-05, "loss": 0.008, "step": 22550 }, { "grad_norm": 0.17644020915031433, "learning_rate": 9.180488690050759e-05, "loss": 0.0078, "step": 22560 }, { "grad_norm": 0.0950874388217926, "learning_rate": 9.179581400170671e-05, "loss": 0.007, "step": 22570 }, { "grad_norm": 0.08248233050107956, "learning_rate": 9.178673653217845e-05, "loss": 0.0067, "step": 22580 }, { "grad_norm": 0.09248244017362595, "learning_rate": 9.177765449291551e-05, "loss": 0.006, "step": 22590 }, { "grad_norm": 0.09137780964374542, "learning_rate": 9.176856788491109e-05, "loss": 0.007, "step": 22600 }, { "grad_norm": 0.115350641310215, "learning_rate": 9.175947670915887e-05, "loss": 0.0074, "step": 22610 }, { "grad_norm": 0.08769883215427399, "learning_rate": 9.175038096665309e-05, "loss": 0.0063, "step": 22620 }, { "grad_norm": 0.09524790197610855, "learning_rate": 9.17412806583884e-05, "loss": 0.0069, "step": 22630 }, { "grad_norm": 0.1070304811000824, "learning_rate": 9.173217578536002e-05, "loss": 0.0071, "step": 22640 }, { "grad_norm": 0.11737461388111115, "learning_rate": 9.172306634856362e-05, "loss": 0.0079, "step": 22650 }, { "grad_norm": 0.13098250329494476, "learning_rate": 9.171395234899545e-05, "loss": 0.0079, "step": 22660 }, { "grad_norm": 0.10379409044981003, "learning_rate": 9.170483378765214e-05, "loss": 0.0065, "step": 22670 }, { "grad_norm": 0.106168732047081, "learning_rate": 9.169571066553091e-05, "loss": 0.0065, "step": 22680 }, { "grad_norm": 0.16248711943626404, "learning_rate": 9.168658298362946e-05, "loss": 0.007, "step": 22690 }, { "grad_norm": 0.10392972826957703, "learning_rate": 9.167745074294598e-05, "loss": 0.0071, "step": 22700 }, { "grad_norm": 0.1301575005054474, "learning_rate": 9.166831394447913e-05, "loss": 0.0072, "step": 22710 }, { "grad_norm": 0.1426442265510559, "learning_rate": 9.165917258922812e-05, "loss": 0.0065, "step": 22720 }, { "grad_norm": 0.130841925740242, "learning_rate": 9.165002667819262e-05, "loss": 0.0095, "step": 22730 }, { "grad_norm": 0.14491434395313263, "learning_rate": 9.164087621237282e-05, "loss": 0.0082, "step": 22740 }, { "grad_norm": 0.15692660212516785, "learning_rate": 9.163172119276942e-05, "loss": 0.0079, "step": 22750 }, { "grad_norm": 0.1431681364774704, "learning_rate": 9.162256162038358e-05, "loss": 0.0089, "step": 22760 }, { "grad_norm": 0.13509796559810638, "learning_rate": 9.161339749621698e-05, "loss": 0.0076, "step": 22770 }, { "grad_norm": 0.0977037250995636, "learning_rate": 9.160422882127177e-05, "loss": 0.0084, "step": 22780 }, { "grad_norm": 0.15118128061294556, "learning_rate": 9.159505559655069e-05, "loss": 0.0073, "step": 22790 }, { "grad_norm": 0.08908198028802872, "learning_rate": 9.158587782305684e-05, "loss": 0.007, "step": 22800 }, { "grad_norm": 0.13224388659000397, "learning_rate": 9.157669550179391e-05, "loss": 0.0062, "step": 22810 }, { "grad_norm": 0.10909077525138855, "learning_rate": 9.156750863376609e-05, "loss": 0.008, "step": 22820 }, { "grad_norm": 0.11895590275526047, "learning_rate": 9.155831721997801e-05, "loss": 0.0069, "step": 22830 }, { "grad_norm": 0.12746064364910126, "learning_rate": 9.154912126143484e-05, "loss": 0.007, "step": 22840 }, { "grad_norm": 0.10475040227174759, "learning_rate": 9.153992075914224e-05, "loss": 0.0066, "step": 22850 }, { "grad_norm": 0.12790298461914062, "learning_rate": 9.153071571410635e-05, "loss": 0.0075, "step": 22860 }, { "grad_norm": 0.0928729996085167, "learning_rate": 9.152150612733384e-05, "loss": 0.0065, "step": 22870 }, { "grad_norm": 0.11377652734518051, "learning_rate": 9.151229199983184e-05, "loss": 0.0074, "step": 22880 }, { "grad_norm": 0.10768118500709534, "learning_rate": 9.150307333260802e-05, "loss": 0.0067, "step": 22890 }, { "grad_norm": 0.08738138526678085, "learning_rate": 9.149385012667048e-05, "loss": 0.0059, "step": 22900 }, { "grad_norm": 0.09861995279788971, "learning_rate": 9.148462238302788e-05, "loss": 0.006, "step": 22910 }, { "grad_norm": 0.09332608431577682, "learning_rate": 9.147539010268936e-05, "loss": 0.006, "step": 22920 }, { "grad_norm": 0.10893181711435318, "learning_rate": 9.14661532866645e-05, "loss": 0.006, "step": 22930 }, { "grad_norm": 0.09917427599430084, "learning_rate": 9.145691193596348e-05, "loss": 0.0063, "step": 22940 }, { "grad_norm": 0.10189864784479141, "learning_rate": 9.144766605159691e-05, "loss": 0.0068, "step": 22950 }, { "grad_norm": 0.15254373848438263, "learning_rate": 9.14384156345759e-05, "loss": 0.0071, "step": 22960 }, { "grad_norm": 0.14047075808048248, "learning_rate": 9.142916068591204e-05, "loss": 0.0066, "step": 22970 }, { "grad_norm": 0.14248870313167572, "learning_rate": 9.141990120661746e-05, "loss": 0.0066, "step": 22980 }, { "grad_norm": 0.11602047830820084, "learning_rate": 9.141063719770475e-05, "loss": 0.0085, "step": 22990 }, { "grad_norm": 0.08380257338285446, "learning_rate": 9.140136866018704e-05, "loss": 0.0075, "step": 23000 }, { "grad_norm": 0.1255900263786316, "learning_rate": 9.139209559507788e-05, "loss": 0.0074, "step": 23010 }, { "grad_norm": 0.11844684928655624, "learning_rate": 9.13828180033914e-05, "loss": 0.0077, "step": 23020 }, { "grad_norm": 0.1432913988828659, "learning_rate": 9.137353588614212e-05, "loss": 0.0068, "step": 23030 }, { "grad_norm": 0.09890887886285782, "learning_rate": 9.136424924434519e-05, "loss": 0.0062, "step": 23040 }, { "grad_norm": 0.0999009981751442, "learning_rate": 9.135495807901615e-05, "loss": 0.0054, "step": 23050 }, { "grad_norm": 0.12397343665361404, "learning_rate": 9.134566239117108e-05, "loss": 0.0065, "step": 23060 }, { "grad_norm": 0.07421932369470596, "learning_rate": 9.13363621818265e-05, "loss": 0.0053, "step": 23070 }, { "grad_norm": 0.1740255355834961, "learning_rate": 9.132705745199953e-05, "loss": 0.007, "step": 23080 }, { "grad_norm": 0.13903063535690308, "learning_rate": 9.131774820270768e-05, "loss": 0.0073, "step": 23090 }, { "grad_norm": 0.12797456979751587, "learning_rate": 9.130843443496901e-05, "loss": 0.0057, "step": 23100 }, { "grad_norm": 0.13183525204658508, "learning_rate": 9.129911614980206e-05, "loss": 0.0061, "step": 23110 }, { "grad_norm": 0.11272799968719482, "learning_rate": 9.128979334822584e-05, "loss": 0.0074, "step": 23120 }, { "grad_norm": 0.1229560524225235, "learning_rate": 9.128046603125992e-05, "loss": 0.0074, "step": 23130 }, { "grad_norm": 0.10849584639072418, "learning_rate": 9.12711341999243e-05, "loss": 0.007, "step": 23140 }, { "grad_norm": 0.09614299237728119, "learning_rate": 9.12617978552395e-05, "loss": 0.0085, "step": 23150 }, { "grad_norm": 0.1408141702413559, "learning_rate": 9.12524569982265e-05, "loss": 0.007, "step": 23160 }, { "grad_norm": 0.15771862864494324, "learning_rate": 9.124311162990684e-05, "loss": 0.0068, "step": 23170 }, { "grad_norm": 0.1299286037683487, "learning_rate": 9.12337617513025e-05, "loss": 0.0069, "step": 23180 }, { "grad_norm": 0.1407076120376587, "learning_rate": 9.122440736343596e-05, "loss": 0.0067, "step": 23190 }, { "grad_norm": 0.12512576580047607, "learning_rate": 9.12150484673302e-05, "loss": 0.0071, "step": 23200 }, { "grad_norm": 0.12159707397222519, "learning_rate": 9.120568506400873e-05, "loss": 0.0074, "step": 23210 }, { "grad_norm": 0.11863623559474945, "learning_rate": 9.119631715449548e-05, "loss": 0.0066, "step": 23220 }, { "grad_norm": 0.11075308173894882, "learning_rate": 9.118694473981493e-05, "loss": 0.0062, "step": 23230 }, { "grad_norm": 0.1518683284521103, "learning_rate": 9.117756782099203e-05, "loss": 0.0093, "step": 23240 }, { "grad_norm": 0.11812041699886322, "learning_rate": 9.11681863990522e-05, "loss": 0.0074, "step": 23250 }, { "grad_norm": 0.09439845383167267, "learning_rate": 9.115880047502142e-05, "loss": 0.0067, "step": 23260 }, { "grad_norm": 0.17323681712150574, "learning_rate": 9.114941004992609e-05, "loss": 0.0084, "step": 23270 }, { "grad_norm": 0.1338960975408554, "learning_rate": 9.114001512479317e-05, "loss": 0.0079, "step": 23280 }, { "grad_norm": 0.088929183781147, "learning_rate": 9.113061570065003e-05, "loss": 0.006, "step": 23290 }, { "grad_norm": 0.10525830835103989, "learning_rate": 9.112121177852459e-05, "loss": 0.0072, "step": 23300 }, { "grad_norm": 0.10192026942968369, "learning_rate": 9.111180335944527e-05, "loss": 0.0075, "step": 23310 }, { "grad_norm": 0.119313083589077, "learning_rate": 9.110239044444093e-05, "loss": 0.0076, "step": 23320 }, { "grad_norm": 0.09529414027929306, "learning_rate": 9.109297303454099e-05, "loss": 0.0077, "step": 23330 }, { "grad_norm": 0.13637691736221313, "learning_rate": 9.108355113077526e-05, "loss": 0.0073, "step": 23340 }, { "grad_norm": 0.08384083956480026, "learning_rate": 9.107412473417419e-05, "loss": 0.0065, "step": 23350 }, { "grad_norm": 0.1269790083169937, "learning_rate": 9.106469384576858e-05, "loss": 0.0062, "step": 23360 }, { "grad_norm": 0.11024155467748642, "learning_rate": 9.105525846658978e-05, "loss": 0.0069, "step": 23370 }, { "grad_norm": 0.10974305123090744, "learning_rate": 9.104581859766965e-05, "loss": 0.0075, "step": 23380 }, { "grad_norm": 0.16609296202659607, "learning_rate": 9.10363742400405e-05, "loss": 0.0075, "step": 23390 }, { "grad_norm": 0.10582710802555084, "learning_rate": 9.102692539473518e-05, "loss": 0.0067, "step": 23400 }, { "grad_norm": 0.13826614618301392, "learning_rate": 9.101747206278697e-05, "loss": 0.0077, "step": 23410 }, { "grad_norm": 0.11676791310310364, "learning_rate": 9.100801424522968e-05, "loss": 0.006, "step": 23420 }, { "grad_norm": 0.09515200555324554, "learning_rate": 9.099855194309762e-05, "loss": 0.0067, "step": 23430 }, { "grad_norm": 0.11036756634712219, "learning_rate": 9.098908515742554e-05, "loss": 0.0068, "step": 23440 }, { "grad_norm": 0.10277653485536575, "learning_rate": 9.097961388924873e-05, "loss": 0.0069, "step": 23450 }, { "grad_norm": 0.08470502495765686, "learning_rate": 9.097013813960298e-05, "loss": 0.0083, "step": 23460 }, { "grad_norm": 0.11853808164596558, "learning_rate": 9.09606579095245e-05, "loss": 0.0077, "step": 23470 }, { "grad_norm": 0.08988603949546814, "learning_rate": 9.095117320005008e-05, "loss": 0.0064, "step": 23480 }, { "grad_norm": 0.09022940695285797, "learning_rate": 9.094168401221691e-05, "loss": 0.0061, "step": 23490 }, { "grad_norm": 0.12855350971221924, "learning_rate": 9.093219034706273e-05, "loss": 0.0074, "step": 23500 }, { "grad_norm": 0.11169834434986115, "learning_rate": 9.092269220562577e-05, "loss": 0.0067, "step": 23510 }, { "grad_norm": 0.0992548018693924, "learning_rate": 9.09131895889447e-05, "loss": 0.0059, "step": 23520 }, { "grad_norm": 0.0980331227183342, "learning_rate": 9.090368249805873e-05, "loss": 0.0063, "step": 23530 }, { "grad_norm": 0.09956244379281998, "learning_rate": 9.089417093400754e-05, "loss": 0.0072, "step": 23540 }, { "grad_norm": 0.0947335734963417, "learning_rate": 9.088465489783131e-05, "loss": 0.0063, "step": 23550 }, { "grad_norm": 0.11138373613357544, "learning_rate": 9.087513439057068e-05, "loss": 0.0061, "step": 23560 }, { "grad_norm": 0.09076327085494995, "learning_rate": 9.08656094132668e-05, "loss": 0.0064, "step": 23570 }, { "grad_norm": 0.0986577719449997, "learning_rate": 9.085607996696134e-05, "loss": 0.0066, "step": 23580 }, { "grad_norm": 0.12018120288848877, "learning_rate": 9.084654605269639e-05, "loss": 0.0082, "step": 23590 }, { "grad_norm": 0.1052122414112091, "learning_rate": 9.083700767151457e-05, "loss": 0.0065, "step": 23600 }, { "grad_norm": 0.10445987433195114, "learning_rate": 9.082746482445898e-05, "loss": 0.0066, "step": 23610 }, { "grad_norm": 0.09348554164171219, "learning_rate": 9.081791751257325e-05, "loss": 0.0065, "step": 23620 }, { "grad_norm": 0.09763622283935547, "learning_rate": 9.080836573690142e-05, "loss": 0.0058, "step": 23630 }, { "grad_norm": 0.13004514575004578, "learning_rate": 9.079880949848805e-05, "loss": 0.0065, "step": 23640 }, { "grad_norm": 0.11860872060060501, "learning_rate": 9.078924879837822e-05, "loss": 0.0058, "step": 23650 }, { "grad_norm": 0.09075774252414703, "learning_rate": 9.077968363761747e-05, "loss": 0.0062, "step": 23660 }, { "grad_norm": 0.157266303896904, "learning_rate": 9.077011401725182e-05, "loss": 0.0067, "step": 23670 }, { "grad_norm": 0.09187464416027069, "learning_rate": 9.07605399383278e-05, "loss": 0.0067, "step": 23680 }, { "grad_norm": 0.09520301967859268, "learning_rate": 9.075096140189243e-05, "loss": 0.0077, "step": 23690 }, { "grad_norm": 0.07428234070539474, "learning_rate": 9.074137840899318e-05, "loss": 0.0075, "step": 23700 }, { "grad_norm": 0.14643673598766327, "learning_rate": 9.073179096067804e-05, "loss": 0.0066, "step": 23710 }, { "grad_norm": 0.08619088679552078, "learning_rate": 9.072219905799549e-05, "loss": 0.0081, "step": 23720 }, { "grad_norm": 0.09385840594768524, "learning_rate": 9.071260270199447e-05, "loss": 0.0058, "step": 23730 }, { "grad_norm": 0.1035328060388565, "learning_rate": 9.070300189372441e-05, "loss": 0.0064, "step": 23740 }, { "grad_norm": 0.09159410744905472, "learning_rate": 9.069339663423528e-05, "loss": 0.0073, "step": 23750 }, { "grad_norm": 0.10310669988393784, "learning_rate": 9.068378692457747e-05, "loss": 0.0072, "step": 23760 }, { "grad_norm": 0.14768463373184204, "learning_rate": 9.067417276580189e-05, "loss": 0.0064, "step": 23770 }, { "grad_norm": 0.117985300719738, "learning_rate": 9.066455415895993e-05, "loss": 0.007, "step": 23780 }, { "grad_norm": 0.1368062049150467, "learning_rate": 9.065493110510346e-05, "loss": 0.0068, "step": 23790 }, { "grad_norm": 0.09995105862617493, "learning_rate": 9.064530360528484e-05, "loss": 0.0076, "step": 23800 }, { "grad_norm": 0.0907561406493187, "learning_rate": 9.063567166055695e-05, "loss": 0.0068, "step": 23810 }, { "grad_norm": 0.10150761157274246, "learning_rate": 9.062603527197308e-05, "loss": 0.0073, "step": 23820 }, { "grad_norm": 0.13623347878456116, "learning_rate": 9.06163944405871e-05, "loss": 0.0067, "step": 23830 }, { "grad_norm": 0.11535169184207916, "learning_rate": 9.060674916745327e-05, "loss": 0.0067, "step": 23840 }, { "grad_norm": 0.08998505026102066, "learning_rate": 9.05970994536264e-05, "loss": 0.0071, "step": 23850 }, { "grad_norm": 0.07520020753145218, "learning_rate": 9.05874453001618e-05, "loss": 0.0059, "step": 23860 }, { "grad_norm": 0.10855483263731003, "learning_rate": 9.057778670811517e-05, "loss": 0.0066, "step": 23870 }, { "grad_norm": 0.10259216278791428, "learning_rate": 9.056812367854281e-05, "loss": 0.0058, "step": 23880 }, { "grad_norm": 0.15739743411540985, "learning_rate": 9.055845621250143e-05, "loss": 0.0069, "step": 23890 }, { "grad_norm": 0.10929834842681885, "learning_rate": 9.054878431104825e-05, "loss": 0.0061, "step": 23900 }, { "grad_norm": 0.08816229552030563, "learning_rate": 9.0539107975241e-05, "loss": 0.0067, "step": 23910 }, { "grad_norm": 0.09351688623428345, "learning_rate": 9.052942720613784e-05, "loss": 0.0062, "step": 23920 }, { "grad_norm": 0.1172691360116005, "learning_rate": 9.051974200479745e-05, "loss": 0.0061, "step": 23930 }, { "grad_norm": 0.12958358228206635, "learning_rate": 9.051005237227901e-05, "loss": 0.0068, "step": 23940 }, { "grad_norm": 0.11235389113426208, "learning_rate": 9.050035830964215e-05, "loss": 0.0083, "step": 23950 }, { "grad_norm": 0.13869141042232513, "learning_rate": 9.049065981794698e-05, "loss": 0.0074, "step": 23960 }, { "grad_norm": 0.12185212224721909, "learning_rate": 9.048095689825414e-05, "loss": 0.0074, "step": 23970 }, { "grad_norm": 0.10850608348846436, "learning_rate": 9.047124955162472e-05, "loss": 0.0069, "step": 23980 }, { "grad_norm": 0.08826526999473572, "learning_rate": 9.046153777912028e-05, "loss": 0.007, "step": 23990 }, { "grad_norm": 0.10154137760400772, "learning_rate": 9.045182158180292e-05, "loss": 0.0065, "step": 24000 }, { "grad_norm": 0.1111886277794838, "learning_rate": 9.044210096073516e-05, "loss": 0.0061, "step": 24010 }, { "grad_norm": 0.0887298583984375, "learning_rate": 9.043237591698004e-05, "loss": 0.0066, "step": 24020 }, { "grad_norm": 0.1372874677181244, "learning_rate": 9.04226464516011e-05, "loss": 0.0067, "step": 24030 }, { "grad_norm": 0.09512130171060562, "learning_rate": 9.041291256566229e-05, "loss": 0.0063, "step": 24040 }, { "grad_norm": 0.11972249299287796, "learning_rate": 9.040317426022814e-05, "loss": 0.0062, "step": 24050 }, { "grad_norm": 0.10972216725349426, "learning_rate": 9.03934315363636e-05, "loss": 0.0069, "step": 24060 }, { "grad_norm": 0.09815719723701477, "learning_rate": 9.038368439513409e-05, "loss": 0.0056, "step": 24070 }, { "grad_norm": 0.12933555245399475, "learning_rate": 9.03739328376056e-05, "loss": 0.0068, "step": 24080 }, { "grad_norm": 0.1303568035364151, "learning_rate": 9.036417686484451e-05, "loss": 0.0071, "step": 24090 }, { "grad_norm": 0.11229238659143448, "learning_rate": 9.035441647791773e-05, "loss": 0.0057, "step": 24100 }, { "grad_norm": 0.11040564626455307, "learning_rate": 9.034465167789263e-05, "loss": 0.0067, "step": 24110 }, { "grad_norm": 0.09260862320661545, "learning_rate": 9.033488246583706e-05, "loss": 0.0067, "step": 24120 }, { "grad_norm": 0.10488137602806091, "learning_rate": 9.032510884281941e-05, "loss": 0.0062, "step": 24130 }, { "grad_norm": 0.10317833721637726, "learning_rate": 9.031533080990848e-05, "loss": 0.0061, "step": 24140 }, { "grad_norm": 0.1054992601275444, "learning_rate": 9.030554836817358e-05, "loss": 0.0077, "step": 24150 }, { "grad_norm": 0.10657479614019394, "learning_rate": 9.029576151868451e-05, "loss": 0.0079, "step": 24160 }, { "grad_norm": 0.11488797515630722, "learning_rate": 9.028597026251155e-05, "loss": 0.006, "step": 24170 }, { "grad_norm": 0.13083332777023315, "learning_rate": 9.027617460072547e-05, "loss": 0.0066, "step": 24180 }, { "grad_norm": 0.11438363045454025, "learning_rate": 9.026637453439745e-05, "loss": 0.0065, "step": 24190 }, { "grad_norm": 0.10324472188949585, "learning_rate": 9.025657006459927e-05, "loss": 0.0057, "step": 24200 }, { "grad_norm": 0.08049779385328293, "learning_rate": 9.024676119240311e-05, "loss": 0.0062, "step": 24210 }, { "grad_norm": 0.1053127720952034, "learning_rate": 9.023694791888166e-05, "loss": 0.0065, "step": 24220 }, { "grad_norm": 0.07887762039899826, "learning_rate": 9.022713024510808e-05, "loss": 0.0065, "step": 24230 }, { "grad_norm": 0.09604945778846741, "learning_rate": 9.021730817215601e-05, "loss": 0.0068, "step": 24240 }, { "grad_norm": 0.10868532210588455, "learning_rate": 9.02074817010996e-05, "loss": 0.0064, "step": 24250 }, { "grad_norm": 0.09751076996326447, "learning_rate": 9.019765083301342e-05, "loss": 0.0067, "step": 24260 }, { "grad_norm": 0.13270136713981628, "learning_rate": 9.01878155689726e-05, "loss": 0.006, "step": 24270 }, { "grad_norm": 0.13621090352535248, "learning_rate": 9.017797591005268e-05, "loss": 0.0077, "step": 24280 }, { "grad_norm": 0.10125140100717545, "learning_rate": 9.016813185732972e-05, "loss": 0.0076, "step": 24290 }, { "grad_norm": 0.1421031355857849, "learning_rate": 9.015828341188027e-05, "loss": 0.0068, "step": 24300 }, { "grad_norm": 0.11532287299633026, "learning_rate": 9.01484305747813e-05, "loss": 0.0071, "step": 24310 }, { "grad_norm": 0.10252141952514648, "learning_rate": 9.013857334711033e-05, "loss": 0.007, "step": 24320 }, { "grad_norm": 0.12299595773220062, "learning_rate": 9.012871172994534e-05, "loss": 0.0076, "step": 24330 }, { "grad_norm": 0.09878963232040405, "learning_rate": 9.011884572436476e-05, "loss": 0.0062, "step": 24340 }, { "grad_norm": 0.09365390986204147, "learning_rate": 9.010897533144754e-05, "loss": 0.0068, "step": 24350 }, { "grad_norm": 0.08743923902511597, "learning_rate": 9.009910055227306e-05, "loss": 0.0058, "step": 24360 }, { "grad_norm": 0.09643775224685669, "learning_rate": 9.008922138792124e-05, "loss": 0.0067, "step": 24370 }, { "grad_norm": 0.10695799440145493, "learning_rate": 9.007933783947244e-05, "loss": 0.0072, "step": 24380 }, { "grad_norm": 0.10349708050489426, "learning_rate": 9.006944990800752e-05, "loss": 0.0073, "step": 24390 }, { "grad_norm": 0.12549401819705963, "learning_rate": 9.005955759460779e-05, "loss": 0.0079, "step": 24400 }, { "grad_norm": 0.0740833729505539, "learning_rate": 9.004966090035508e-05, "loss": 0.0062, "step": 24410 }, { "grad_norm": 0.1393670290708542, "learning_rate": 9.003975982633166e-05, "loss": 0.0066, "step": 24420 }, { "grad_norm": 0.09412693977355957, "learning_rate": 9.00298543736203e-05, "loss": 0.0065, "step": 24430 }, { "grad_norm": 0.10268135368824005, "learning_rate": 9.001994454330427e-05, "loss": 0.0063, "step": 24440 }, { "grad_norm": 0.10294836014509201, "learning_rate": 9.001003033646727e-05, "loss": 0.0058, "step": 24450 }, { "grad_norm": 0.09452217072248459, "learning_rate": 9.00001117541935e-05, "loss": 0.0056, "step": 24460 }, { "grad_norm": 0.12412182241678238, "learning_rate": 8.999018879756764e-05, "loss": 0.0069, "step": 24470 }, { "grad_norm": 0.12474483251571655, "learning_rate": 8.998026146767487e-05, "loss": 0.0065, "step": 24480 }, { "grad_norm": 0.11420055478811264, "learning_rate": 8.99703297656008e-05, "loss": 0.0067, "step": 24490 }, { "grad_norm": 0.11171423643827438, "learning_rate": 8.996039369243156e-05, "loss": 0.0061, "step": 24500 }, { "grad_norm": 0.1049572303891182, "learning_rate": 8.995045324925378e-05, "loss": 0.0089, "step": 24510 }, { "grad_norm": 0.09792402386665344, "learning_rate": 8.994050843715448e-05, "loss": 0.007, "step": 24520 }, { "grad_norm": 0.0895262360572815, "learning_rate": 8.993055925722121e-05, "loss": 0.0079, "step": 24530 }, { "grad_norm": 0.12993763387203217, "learning_rate": 8.992060571054202e-05, "loss": 0.0064, "step": 24540 }, { "grad_norm": 0.10516185313463211, "learning_rate": 8.991064779820542e-05, "loss": 0.0068, "step": 24550 }, { "grad_norm": 0.12086090445518494, "learning_rate": 8.990068552130036e-05, "loss": 0.0074, "step": 24560 }, { "grad_norm": 0.09154237061738968, "learning_rate": 8.989071888091634e-05, "loss": 0.0057, "step": 24570 }, { "grad_norm": 0.09265456348657608, "learning_rate": 8.988074787814329e-05, "loss": 0.0068, "step": 24580 }, { "grad_norm": 0.11065397411584854, "learning_rate": 8.987077251407158e-05, "loss": 0.0064, "step": 24590 }, { "grad_norm": 0.16600961983203888, "learning_rate": 8.986079278979216e-05, "loss": 0.0082, "step": 24600 }, { "grad_norm": 0.1176387220621109, "learning_rate": 8.985080870639635e-05, "loss": 0.007, "step": 24610 }, { "grad_norm": 0.12224656343460083, "learning_rate": 8.984082026497603e-05, "loss": 0.0075, "step": 24620 }, { "grad_norm": 0.07794006913900375, "learning_rate": 8.98308274666235e-05, "loss": 0.0073, "step": 24630 }, { "grad_norm": 0.10484036803245544, "learning_rate": 8.982083031243155e-05, "loss": 0.0076, "step": 24640 }, { "grad_norm": 0.16162404417991638, "learning_rate": 8.98108288034935e-05, "loss": 0.0081, "step": 24650 }, { "grad_norm": 0.13695299625396729, "learning_rate": 8.980082294090305e-05, "loss": 0.0063, "step": 24660 }, { "grad_norm": 0.1369638442993164, "learning_rate": 8.979081272575443e-05, "loss": 0.0072, "step": 24670 }, { "grad_norm": 0.12875837087631226, "learning_rate": 8.978079815914236e-05, "loss": 0.0069, "step": 24680 }, { "grad_norm": 0.08301596343517303, "learning_rate": 8.977077924216202e-05, "loss": 0.0064, "step": 24690 }, { "grad_norm": 0.07878248393535614, "learning_rate": 8.976075597590905e-05, "loss": 0.0055, "step": 24700 }, { "grad_norm": 0.15108251571655273, "learning_rate": 8.975072836147958e-05, "loss": 0.0069, "step": 24710 }, { "grad_norm": 0.10995043069124222, "learning_rate": 8.974069639997025e-05, "loss": 0.0065, "step": 24720 }, { "grad_norm": 0.119132399559021, "learning_rate": 8.973066009247808e-05, "loss": 0.0064, "step": 24730 }, { "grad_norm": 0.11225631833076477, "learning_rate": 8.972061944010066e-05, "loss": 0.0069, "step": 24740 }, { "grad_norm": 0.1401589810848236, "learning_rate": 8.971057444393603e-05, "loss": 0.0067, "step": 24750 }, { "grad_norm": 0.06195145100355148, "learning_rate": 8.970052510508268e-05, "loss": 0.0059, "step": 24760 }, { "grad_norm": 0.10443951934576035, "learning_rate": 8.969047142463959e-05, "loss": 0.0063, "step": 24770 }, { "grad_norm": 0.11340977251529694, "learning_rate": 8.968041340370621e-05, "loss": 0.0073, "step": 24780 }, { "grad_norm": 0.16958165168762207, "learning_rate": 8.96703510433825e-05, "loss": 0.0066, "step": 24790 }, { "grad_norm": 0.10762352496385574, "learning_rate": 8.966028434476883e-05, "loss": 0.0074, "step": 24800 }, { "grad_norm": 0.08650325983762741, "learning_rate": 8.96502133089661e-05, "loss": 0.006, "step": 24810 }, { "grad_norm": 0.08300147950649261, "learning_rate": 8.964013793707564e-05, "loss": 0.0071, "step": 24820 }, { "grad_norm": 0.10650700330734253, "learning_rate": 8.963005823019932e-05, "loss": 0.0063, "step": 24830 }, { "grad_norm": 0.10866150259971619, "learning_rate": 8.961997418943939e-05, "loss": 0.0062, "step": 24840 }, { "grad_norm": 0.08098237216472626, "learning_rate": 8.960988581589865e-05, "loss": 0.0066, "step": 24850 }, { "grad_norm": 0.11361490935087204, "learning_rate": 8.959979311068037e-05, "loss": 0.0063, "step": 24860 }, { "grad_norm": 0.13125339150428772, "learning_rate": 8.958969607488823e-05, "loss": 0.0064, "step": 24870 }, { "grad_norm": 0.134102463722229, "learning_rate": 8.957959470962647e-05, "loss": 0.0077, "step": 24880 }, { "grad_norm": 0.12128689140081406, "learning_rate": 8.956948901599971e-05, "loss": 0.0073, "step": 24890 }, { "grad_norm": 0.11381101608276367, "learning_rate": 8.955937899511315e-05, "loss": 0.0072, "step": 24900 }, { "grad_norm": 0.19227701425552368, "learning_rate": 8.954926464807238e-05, "loss": 0.0076, "step": 24910 }, { "grad_norm": 0.12237026542425156, "learning_rate": 8.953914597598347e-05, "loss": 0.0088, "step": 24920 }, { "grad_norm": 0.1411585509777069, "learning_rate": 8.952902297995303e-05, "loss": 0.0081, "step": 24930 }, { "grad_norm": 0.1511821746826172, "learning_rate": 8.951889566108804e-05, "loss": 0.0073, "step": 24940 }, { "grad_norm": 0.11689449846744537, "learning_rate": 8.950876402049606e-05, "loss": 0.0069, "step": 24950 }, { "grad_norm": 0.16557149589061737, "learning_rate": 8.949862805928504e-05, "loss": 0.0075, "step": 24960 }, { "grad_norm": 0.1140487864613533, "learning_rate": 8.948848777856343e-05, "loss": 0.007, "step": 24970 }, { "grad_norm": 0.09559417515993118, "learning_rate": 8.947834317944017e-05, "loss": 0.0071, "step": 24980 }, { "grad_norm": 0.13384687900543213, "learning_rate": 8.946819426302466e-05, "loss": 0.0065, "step": 24990 }, { "grad_norm": 0.10114815086126328, "learning_rate": 8.945804103042676e-05, "loss": 0.0066, "step": 25000 }, { "grad_norm": 0.12227018922567368, "learning_rate": 8.944788348275681e-05, "loss": 0.0081, "step": 25010 }, { "grad_norm": 0.1455608606338501, "learning_rate": 8.943772162112565e-05, "loss": 0.0081, "step": 25020 }, { "grad_norm": 0.11716248840093613, "learning_rate": 8.942755544664454e-05, "loss": 0.0073, "step": 25030 }, { "grad_norm": 0.11819440871477127, "learning_rate": 8.941738496042525e-05, "loss": 0.008, "step": 25040 }, { "grad_norm": 0.14078065752983093, "learning_rate": 8.940721016357999e-05, "loss": 0.0064, "step": 25050 }, { "grad_norm": 0.11504477262496948, "learning_rate": 8.939703105722148e-05, "loss": 0.0066, "step": 25060 }, { "grad_norm": 0.10261978954076767, "learning_rate": 8.93868476424629e-05, "loss": 0.0071, "step": 25070 }, { "grad_norm": 0.12925985455513, "learning_rate": 8.937665992041786e-05, "loss": 0.0067, "step": 25080 }, { "grad_norm": 0.12968221306800842, "learning_rate": 8.93664678922005e-05, "loss": 0.0071, "step": 25090 }, { "grad_norm": 0.10309441387653351, "learning_rate": 8.93562715589254e-05, "loss": 0.0079, "step": 25100 }, { "grad_norm": 0.08896569907665253, "learning_rate": 8.934607092170762e-05, "loss": 0.0065, "step": 25110 }, { "grad_norm": 0.10577993839979172, "learning_rate": 8.933586598166266e-05, "loss": 0.0059, "step": 25120 }, { "grad_norm": 0.1151190847158432, "learning_rate": 8.932565673990655e-05, "loss": 0.0056, "step": 25130 }, { "grad_norm": 0.15604600310325623, "learning_rate": 8.931544319755574e-05, "loss": 0.0069, "step": 25140 }, { "grad_norm": 0.10805365443229675, "learning_rate": 8.930522535572718e-05, "loss": 0.0066, "step": 25150 }, { "grad_norm": 0.10848548263311386, "learning_rate": 8.929500321553826e-05, "loss": 0.0065, "step": 25160 }, { "grad_norm": 0.10301098227500916, "learning_rate": 8.928477677810686e-05, "loss": 0.0064, "step": 25170 }, { "grad_norm": 0.09841988235712051, "learning_rate": 8.927454604455137e-05, "loss": 0.0057, "step": 25180 }, { "grad_norm": 0.11175355315208435, "learning_rate": 8.926431101599053e-05, "loss": 0.0065, "step": 25190 }, { "grad_norm": 0.10857278853654861, "learning_rate": 8.925407169354369e-05, "loss": 0.0057, "step": 25200 }, { "grad_norm": 0.08917788416147232, "learning_rate": 8.92438280783306e-05, "loss": 0.0062, "step": 25210 }, { "grad_norm": 0.11649098992347717, "learning_rate": 8.923358017147146e-05, "loss": 0.0053, "step": 25220 }, { "grad_norm": 0.1000128835439682, "learning_rate": 8.922332797408697e-05, "loss": 0.0071, "step": 25230 }, { "grad_norm": 0.10320153087377548, "learning_rate": 8.921307148729831e-05, "loss": 0.0077, "step": 25240 }, { "grad_norm": 0.15707281231880188, "learning_rate": 8.920281071222712e-05, "loss": 0.0056, "step": 25250 }, { "grad_norm": 0.12230616807937622, "learning_rate": 8.919254564999548e-05, "loss": 0.0067, "step": 25260 }, { "grad_norm": 0.12775924801826477, "learning_rate": 8.918227630172598e-05, "loss": 0.0073, "step": 25270 }, { "grad_norm": 0.11678695678710938, "learning_rate": 8.917200266854165e-05, "loss": 0.0071, "step": 25280 }, { "grad_norm": 0.10851678997278214, "learning_rate": 8.9161724751566e-05, "loss": 0.0074, "step": 25290 }, { "grad_norm": 0.109674833714962, "learning_rate": 8.915144255192302e-05, "loss": 0.0064, "step": 25300 }, { "grad_norm": 0.10149412602186203, "learning_rate": 8.914115607073714e-05, "loss": 0.0076, "step": 25310 }, { "grad_norm": 0.19781675934791565, "learning_rate": 8.913086530913327e-05, "loss": 0.0081, "step": 25320 }, { "grad_norm": 0.10556675493717194, "learning_rate": 8.912057026823681e-05, "loss": 0.0079, "step": 25330 }, { "grad_norm": 0.09285827726125717, "learning_rate": 8.91102709491736e-05, "loss": 0.007, "step": 25340 }, { "grad_norm": 0.17506727576255798, "learning_rate": 8.909996735306996e-05, "loss": 0.0077, "step": 25350 }, { "grad_norm": 0.12720952928066254, "learning_rate": 8.908965948105268e-05, "loss": 0.0081, "step": 25360 }, { "grad_norm": 0.13399538397789001, "learning_rate": 8.907934733424901e-05, "loss": 0.0065, "step": 25370 }, { "grad_norm": 0.08755545318126678, "learning_rate": 8.906903091378666e-05, "loss": 0.007, "step": 25380 }, { "grad_norm": 0.08468316495418549, "learning_rate": 8.905871022079384e-05, "loss": 0.0078, "step": 25390 }, { "grad_norm": 0.09701883792877197, "learning_rate": 8.90483852563992e-05, "loss": 0.0074, "step": 25400 }, { "grad_norm": 0.18075832724571228, "learning_rate": 8.903805602173185e-05, "loss": 0.0076, "step": 25410 }, { "grad_norm": 0.11301559209823608, "learning_rate": 8.902772251792137e-05, "loss": 0.0069, "step": 25420 }, { "grad_norm": 0.11083913594484329, "learning_rate": 8.901738474609786e-05, "loss": 0.0071, "step": 25430 }, { "grad_norm": 0.10452710092067719, "learning_rate": 8.900704270739179e-05, "loss": 0.0063, "step": 25440 }, { "grad_norm": 0.12024609744548798, "learning_rate": 8.89966964029342e-05, "loss": 0.0065, "step": 25450 }, { "grad_norm": 0.1378166675567627, "learning_rate": 8.898634583385652e-05, "loss": 0.0067, "step": 25460 }, { "grad_norm": 0.0919070914387703, "learning_rate": 8.897599100129065e-05, "loss": 0.0071, "step": 25470 }, { "grad_norm": 0.1102418527007103, "learning_rate": 8.896563190636903e-05, "loss": 0.0065, "step": 25480 }, { "grad_norm": 0.11953972280025482, "learning_rate": 8.895526855022448e-05, "loss": 0.0066, "step": 25490 }, { "grad_norm": 0.10683997720479965, "learning_rate": 8.894490093399033e-05, "loss": 0.0063, "step": 25500 }, { "grad_norm": 0.11681689321994781, "learning_rate": 8.893452905880035e-05, "loss": 0.0051, "step": 25510 }, { "grad_norm": 0.12602102756500244, "learning_rate": 8.892415292578883e-05, "loss": 0.0061, "step": 25520 }, { "grad_norm": 0.1405046582221985, "learning_rate": 8.891377253609046e-05, "loss": 0.0068, "step": 25530 }, { "grad_norm": 0.13013502955436707, "learning_rate": 8.890338789084043e-05, "loss": 0.0061, "step": 25540 }, { "grad_norm": 0.11442366987466812, "learning_rate": 8.88929989911744e-05, "loss": 0.0066, "step": 25550 }, { "grad_norm": 0.0837593525648117, "learning_rate": 8.888260583822847e-05, "loss": 0.0061, "step": 25560 }, { "grad_norm": 0.08243399113416672, "learning_rate": 8.887220843313921e-05, "loss": 0.0063, "step": 25570 }, { "grad_norm": 0.12714233994483948, "learning_rate": 8.88618067770437e-05, "loss": 0.0063, "step": 25580 }, { "grad_norm": 0.09384936094284058, "learning_rate": 8.885140087107942e-05, "loss": 0.0075, "step": 25590 }, { "grad_norm": 0.11568498611450195, "learning_rate": 8.884099071638436e-05, "loss": 0.0057, "step": 25600 }, { "grad_norm": 0.09827302396297455, "learning_rate": 8.883057631409695e-05, "loss": 0.0068, "step": 25610 }, { "grad_norm": 0.11023246496915817, "learning_rate": 8.882015766535608e-05, "loss": 0.0068, "step": 25620 }, { "grad_norm": 0.14440996944904327, "learning_rate": 8.880973477130115e-05, "loss": 0.0068, "step": 25630 }, { "grad_norm": 0.12586893141269684, "learning_rate": 8.879930763307197e-05, "loss": 0.0054, "step": 25640 }, { "grad_norm": 0.12762540578842163, "learning_rate": 8.878887625180884e-05, "loss": 0.0056, "step": 25650 }, { "grad_norm": 0.1465461403131485, "learning_rate": 8.877844062865253e-05, "loss": 0.0059, "step": 25660 }, { "grad_norm": 0.10809203237295151, "learning_rate": 8.876800076474424e-05, "loss": 0.0058, "step": 25670 }, { "grad_norm": 0.07980774343013763, "learning_rate": 8.875755666122568e-05, "loss": 0.006, "step": 25680 }, { "grad_norm": 0.12565498054027557, "learning_rate": 8.8747108319239e-05, "loss": 0.0072, "step": 25690 }, { "grad_norm": 0.10269460082054138, "learning_rate": 8.87366557399268e-05, "loss": 0.0057, "step": 25700 }, { "grad_norm": 0.11454799026250839, "learning_rate": 8.872619892443217e-05, "loss": 0.0061, "step": 25710 }, { "grad_norm": 0.11384212225675583, "learning_rate": 8.871573787389865e-05, "loss": 0.0062, "step": 25720 }, { "grad_norm": 0.13303224742412567, "learning_rate": 8.870527258947024e-05, "loss": 0.0065, "step": 25730 }, { "grad_norm": 0.07437195628881454, "learning_rate": 8.869480307229143e-05, "loss": 0.0059, "step": 25740 }, { "grad_norm": 0.12279506027698517, "learning_rate": 8.868432932350712e-05, "loss": 0.0063, "step": 25750 }, { "grad_norm": 0.10207542032003403, "learning_rate": 8.867385134426272e-05, "loss": 0.0059, "step": 25760 }, { "grad_norm": 0.14173631370067596, "learning_rate": 8.866336913570407e-05, "loss": 0.0073, "step": 25770 }, { "grad_norm": 0.09506621211767197, "learning_rate": 8.865288269897751e-05, "loss": 0.0063, "step": 25780 }, { "grad_norm": 0.12843981385231018, "learning_rate": 8.864239203522981e-05, "loss": 0.006, "step": 25790 }, { "grad_norm": 0.0882379561662674, "learning_rate": 8.863189714560822e-05, "loss": 0.0063, "step": 25800 }, { "grad_norm": 0.11049462854862213, "learning_rate": 8.862139803126043e-05, "loss": 0.0079, "step": 25810 }, { "grad_norm": 0.14067256450653076, "learning_rate": 8.861089469333463e-05, "loss": 0.0072, "step": 25820 }, { "grad_norm": 0.17603954672813416, "learning_rate": 8.860038713297944e-05, "loss": 0.0063, "step": 25830 }, { "grad_norm": 0.09813807159662247, "learning_rate": 8.858987535134394e-05, "loss": 0.0063, "step": 25840 }, { "grad_norm": 0.08289328217506409, "learning_rate": 8.857935934957769e-05, "loss": 0.0056, "step": 25850 }, { "grad_norm": 0.08709117770195007, "learning_rate": 8.856883912883071e-05, "loss": 0.0069, "step": 25860 }, { "grad_norm": 0.11318980157375336, "learning_rate": 8.855831469025346e-05, "loss": 0.0067, "step": 25870 }, { "grad_norm": 0.09904018044471741, "learning_rate": 8.854778603499689e-05, "loss": 0.0061, "step": 25880 }, { "grad_norm": 0.09716953337192535, "learning_rate": 8.85372531642124e-05, "loss": 0.0059, "step": 25890 }, { "grad_norm": 0.14358316361904144, "learning_rate": 8.852671607905185e-05, "loss": 0.0064, "step": 25900 }, { "grad_norm": 0.14665335416793823, "learning_rate": 8.851617478066754e-05, "loss": 0.006, "step": 25910 }, { "grad_norm": 0.08904802054166794, "learning_rate": 8.850562927021227e-05, "loss": 0.0059, "step": 25920 }, { "grad_norm": 0.0782405361533165, "learning_rate": 8.849507954883928e-05, "loss": 0.0062, "step": 25930 }, { "grad_norm": 0.08761126548051834, "learning_rate": 8.848452561770226e-05, "loss": 0.007, "step": 25940 }, { "grad_norm": 0.10349107533693314, "learning_rate": 8.847396747795538e-05, "loss": 0.0055, "step": 25950 }, { "grad_norm": 0.11198128014802933, "learning_rate": 8.846340513075327e-05, "loss": 0.0057, "step": 25960 }, { "grad_norm": 0.1223902702331543, "learning_rate": 8.845283857725099e-05, "loss": 0.0065, "step": 25970 }, { "grad_norm": 0.08662503212690353, "learning_rate": 8.844226781860409e-05, "loss": 0.0071, "step": 25980 }, { "grad_norm": 0.07568006962537766, "learning_rate": 8.84316928559686e-05, "loss": 0.0063, "step": 25990 }, { "grad_norm": 0.0996391773223877, "learning_rate": 8.842111369050094e-05, "loss": 0.0065, "step": 26000 }, { "grad_norm": 0.11017477512359619, "learning_rate": 8.841053032335808e-05, "loss": 0.0055, "step": 26010 }, { "grad_norm": 0.10955488681793213, "learning_rate": 8.839994275569735e-05, "loss": 0.0064, "step": 26020 }, { "grad_norm": 0.12047277390956879, "learning_rate": 8.838935098867662e-05, "loss": 0.006, "step": 26030 }, { "grad_norm": 0.09404421597719193, "learning_rate": 8.837875502345418e-05, "loss": 0.0053, "step": 26040 }, { "grad_norm": 0.06906820088624954, "learning_rate": 8.83681548611888e-05, "loss": 0.0061, "step": 26050 }, { "grad_norm": 0.08079200983047485, "learning_rate": 8.835755050303969e-05, "loss": 0.0056, "step": 26060 }, { "grad_norm": 0.0974162369966507, "learning_rate": 8.834694195016653e-05, "loss": 0.0062, "step": 26070 }, { "grad_norm": 0.07733652740716934, "learning_rate": 8.833632920372942e-05, "loss": 0.0052, "step": 26080 }, { "grad_norm": 0.114463672041893, "learning_rate": 8.832571226488903e-05, "loss": 0.0059, "step": 26090 }, { "grad_norm": 0.18202079832553864, "learning_rate": 8.831509113480634e-05, "loss": 0.006, "step": 26100 }, { "grad_norm": 0.10211455076932907, "learning_rate": 8.83044658146429e-05, "loss": 0.0061, "step": 26110 }, { "grad_norm": 0.12416371703147888, "learning_rate": 8.829383630556067e-05, "loss": 0.0067, "step": 26120 }, { "grad_norm": 0.11887634545564651, "learning_rate": 8.828320260872207e-05, "loss": 0.0061, "step": 26130 }, { "grad_norm": 0.0777919664978981, "learning_rate": 8.827256472529e-05, "loss": 0.0061, "step": 26140 }, { "grad_norm": 0.10191082209348679, "learning_rate": 8.826192265642778e-05, "loss": 0.0074, "step": 26150 }, { "grad_norm": 0.11168745905160904, "learning_rate": 8.825127640329923e-05, "loss": 0.0075, "step": 26160 }, { "grad_norm": 0.10802078247070312, "learning_rate": 8.824062596706861e-05, "loss": 0.0079, "step": 26170 }, { "grad_norm": 0.1461343914270401, "learning_rate": 8.822997134890062e-05, "loss": 0.0087, "step": 26180 }, { "grad_norm": 0.16335806250572205, "learning_rate": 8.821931254996044e-05, "loss": 0.0092, "step": 26190 }, { "grad_norm": 0.1583220511674881, "learning_rate": 8.82086495714137e-05, "loss": 0.0106, "step": 26200 }, { "grad_norm": 0.12419626861810684, "learning_rate": 8.81979824144265e-05, "loss": 0.0092, "step": 26210 }, { "grad_norm": 0.09327526390552521, "learning_rate": 8.818731108016536e-05, "loss": 0.0096, "step": 26220 }, { "grad_norm": 0.11668036878108978, "learning_rate": 8.81766355697973e-05, "loss": 0.0081, "step": 26230 }, { "grad_norm": 0.1179707869887352, "learning_rate": 8.816595588448977e-05, "loss": 0.0072, "step": 26240 }, { "grad_norm": 0.11143364757299423, "learning_rate": 8.81552720254107e-05, "loss": 0.0071, "step": 26250 }, { "grad_norm": 0.09653468430042267, "learning_rate": 8.814458399372842e-05, "loss": 0.0072, "step": 26260 }, { "grad_norm": 0.10178063809871674, "learning_rate": 8.813389179061181e-05, "loss": 0.0068, "step": 26270 }, { "grad_norm": 0.13764643669128418, "learning_rate": 8.812319541723012e-05, "loss": 0.006, "step": 26280 }, { "grad_norm": 0.12100439518690109, "learning_rate": 8.811249487475309e-05, "loss": 0.0075, "step": 26290 }, { "grad_norm": 0.11945345252752304, "learning_rate": 8.810179016435092e-05, "loss": 0.0075, "step": 26300 }, { "grad_norm": 0.11109413206577301, "learning_rate": 8.809108128719428e-05, "loss": 0.0071, "step": 26310 }, { "grad_norm": 0.13916774094104767, "learning_rate": 8.808036824445424e-05, "loss": 0.007, "step": 26320 }, { "grad_norm": 0.15079525113105774, "learning_rate": 8.806965103730238e-05, "loss": 0.0073, "step": 26330 }, { "grad_norm": 0.11706850677728653, "learning_rate": 8.805892966691074e-05, "loss": 0.0084, "step": 26340 }, { "grad_norm": 0.13104653358459473, "learning_rate": 8.804820413445175e-05, "loss": 0.0072, "step": 26350 }, { "grad_norm": 0.10353278368711472, "learning_rate": 8.803747444109837e-05, "loss": 0.0066, "step": 26360 }, { "grad_norm": 0.09432612359523773, "learning_rate": 8.802674058802399e-05, "loss": 0.0067, "step": 26370 }, { "grad_norm": 0.10414237529039383, "learning_rate": 8.801600257640241e-05, "loss": 0.0066, "step": 26380 }, { "grad_norm": 0.11128375679254532, "learning_rate": 8.800526040740795e-05, "loss": 0.0062, "step": 26390 }, { "grad_norm": 0.08407063037157059, "learning_rate": 8.799451408221535e-05, "loss": 0.0065, "step": 26400 }, { "grad_norm": 0.10197140276432037, "learning_rate": 8.798376360199982e-05, "loss": 0.0061, "step": 26410 }, { "grad_norm": 0.1364297717809677, "learning_rate": 8.797300896793701e-05, "loss": 0.0061, "step": 26420 }, { "grad_norm": 0.10187841951847076, "learning_rate": 8.796225018120302e-05, "loss": 0.0065, "step": 26430 }, { "grad_norm": 0.09591113775968552, "learning_rate": 8.795148724297444e-05, "loss": 0.006, "step": 26440 }, { "grad_norm": 0.12004302442073822, "learning_rate": 8.794072015442825e-05, "loss": 0.0068, "step": 26450 }, { "grad_norm": 0.07819569855928421, "learning_rate": 8.792994891674198e-05, "loss": 0.0055, "step": 26460 }, { "grad_norm": 0.096390500664711, "learning_rate": 8.79191735310935e-05, "loss": 0.0068, "step": 26470 }, { "grad_norm": 0.11793889105319977, "learning_rate": 8.790839399866122e-05, "loss": 0.0061, "step": 26480 }, { "grad_norm": 0.10490217804908752, "learning_rate": 8.789761032062397e-05, "loss": 0.0056, "step": 26490 }, { "grad_norm": 0.10951423645019531, "learning_rate": 8.788682249816103e-05, "loss": 0.0059, "step": 26500 }, { "grad_norm": 0.09631205350160599, "learning_rate": 8.787603053245215e-05, "loss": 0.0057, "step": 26510 }, { "grad_norm": 0.09937945753335953, "learning_rate": 8.78652344246775e-05, "loss": 0.0073, "step": 26520 }, { "grad_norm": 0.10089948773384094, "learning_rate": 8.785443417601776e-05, "loss": 0.0062, "step": 26530 }, { "grad_norm": 0.0883505642414093, "learning_rate": 8.784362978765401e-05, "loss": 0.0059, "step": 26540 }, { "grad_norm": 0.11638437956571579, "learning_rate": 8.783282126076779e-05, "loss": 0.006, "step": 26550 }, { "grad_norm": 0.15784427523612976, "learning_rate": 8.782200859654112e-05, "loss": 0.0075, "step": 26560 }, { "grad_norm": 0.12587010860443115, "learning_rate": 8.781119179615646e-05, "loss": 0.0068, "step": 26570 }, { "grad_norm": 0.0718662366271019, "learning_rate": 8.780037086079674e-05, "loss": 0.0048, "step": 26580 }, { "grad_norm": 0.09406300634145737, "learning_rate": 8.778954579164527e-05, "loss": 0.0059, "step": 26590 }, { "grad_norm": 0.09817912429571152, "learning_rate": 8.777871658988588e-05, "loss": 0.0064, "step": 26600 }, { "grad_norm": 0.09075459092855453, "learning_rate": 8.776788325670285e-05, "loss": 0.0066, "step": 26610 }, { "grad_norm": 0.10419051349163055, "learning_rate": 8.775704579328089e-05, "loss": 0.0069, "step": 26620 }, { "grad_norm": 0.0883369892835617, "learning_rate": 8.774620420080517e-05, "loss": 0.0066, "step": 26630 }, { "grad_norm": 0.1700088530778885, "learning_rate": 8.773535848046131e-05, "loss": 0.0072, "step": 26640 }, { "grad_norm": 0.10101724416017532, "learning_rate": 8.772450863343538e-05, "loss": 0.0069, "step": 26650 }, { "grad_norm": 0.11736587435007095, "learning_rate": 8.77136546609139e-05, "loss": 0.0091, "step": 26660 }, { "grad_norm": 0.12927687168121338, "learning_rate": 8.770279656408385e-05, "loss": 0.0069, "step": 26670 }, { "grad_norm": 0.1327967345714569, "learning_rate": 8.769193434413265e-05, "loss": 0.0066, "step": 26680 }, { "grad_norm": 0.08802218735218048, "learning_rate": 8.76810680022482e-05, "loss": 0.006, "step": 26690 }, { "grad_norm": 0.12104912102222443, "learning_rate": 8.767019753961878e-05, "loss": 0.0059, "step": 26700 }, { "grad_norm": 0.11718250811100006, "learning_rate": 8.765932295743321e-05, "loss": 0.0061, "step": 26710 }, { "grad_norm": 0.10577742010354996, "learning_rate": 8.764844425688068e-05, "loss": 0.006, "step": 26720 }, { "grad_norm": 0.11023180186748505, "learning_rate": 8.763756143915092e-05, "loss": 0.0057, "step": 26730 }, { "grad_norm": 0.09182602167129517, "learning_rate": 8.7626674505434e-05, "loss": 0.0058, "step": 26740 }, { "grad_norm": 0.11413022875785828, "learning_rate": 8.761578345692053e-05, "loss": 0.0059, "step": 26750 }, { "grad_norm": 0.0906199961900711, "learning_rate": 8.760488829480156e-05, "loss": 0.0074, "step": 26760 }, { "grad_norm": 0.07892689853906631, "learning_rate": 8.759398902026854e-05, "loss": 0.0054, "step": 26770 }, { "grad_norm": 0.11169344931840897, "learning_rate": 8.758308563451339e-05, "loss": 0.0063, "step": 26780 }, { "grad_norm": 0.11916762590408325, "learning_rate": 8.75721781387285e-05, "loss": 0.006, "step": 26790 }, { "grad_norm": 0.09043529629707336, "learning_rate": 8.75612665341067e-05, "loss": 0.0054, "step": 26800 }, { "grad_norm": 0.09947314113378525, "learning_rate": 8.755035082184126e-05, "loss": 0.0062, "step": 26810 }, { "grad_norm": 0.10611942410469055, "learning_rate": 8.753943100312592e-05, "loss": 0.0058, "step": 26820 }, { "grad_norm": 0.1164524108171463, "learning_rate": 8.752850707915484e-05, "loss": 0.0062, "step": 26830 }, { "grad_norm": 0.0923718512058258, "learning_rate": 8.751757905112264e-05, "loss": 0.007, "step": 26840 }, { "grad_norm": 0.14707216620445251, "learning_rate": 8.75066469202244e-05, "loss": 0.006, "step": 26850 }, { "grad_norm": 0.06261006742715836, "learning_rate": 8.749571068765567e-05, "loss": 0.006, "step": 26860 }, { "grad_norm": 0.10854039341211319, "learning_rate": 8.748477035461238e-05, "loss": 0.0058, "step": 26870 }, { "grad_norm": 0.08556681871414185, "learning_rate": 8.747382592229095e-05, "loss": 0.0059, "step": 26880 }, { "grad_norm": 0.0881209671497345, "learning_rate": 8.746287739188828e-05, "loss": 0.0062, "step": 26890 }, { "grad_norm": 0.1027696505188942, "learning_rate": 8.745192476460165e-05, "loss": 0.0056, "step": 26900 }, { "grad_norm": 0.08729997277259827, "learning_rate": 8.744096804162882e-05, "loss": 0.0057, "step": 26910 }, { "grad_norm": 0.09055516123771667, "learning_rate": 8.743000722416804e-05, "loss": 0.0053, "step": 26920 }, { "grad_norm": 0.1180436834692955, "learning_rate": 8.741904231341793e-05, "loss": 0.006, "step": 26930 }, { "grad_norm": 0.1072467565536499, "learning_rate": 8.740807331057762e-05, "loss": 0.0063, "step": 26940 }, { "grad_norm": 0.0927969217300415, "learning_rate": 8.739710021684667e-05, "loss": 0.0066, "step": 26950 }, { "grad_norm": 0.09811876714229584, "learning_rate": 8.738612303342503e-05, "loss": 0.007, "step": 26960 }, { "grad_norm": 0.14697138965129852, "learning_rate": 8.73751417615132e-05, "loss": 0.0067, "step": 26970 }, { "grad_norm": 0.11022821068763733, "learning_rate": 8.736415640231208e-05, "loss": 0.0059, "step": 26980 }, { "grad_norm": 0.14335069060325623, "learning_rate": 8.735316695702297e-05, "loss": 0.0077, "step": 26990 }, { "grad_norm": 0.08716314285993576, "learning_rate": 8.734217342684769e-05, "loss": 0.0065, "step": 27000 }, { "grad_norm": 0.13428272306919098, "learning_rate": 8.733117581298847e-05, "loss": 0.0066, "step": 27010 }, { "grad_norm": 0.1102469190955162, "learning_rate": 8.732017411664796e-05, "loss": 0.0067, "step": 27020 }, { "grad_norm": 0.12032242864370346, "learning_rate": 8.730916833902936e-05, "loss": 0.0057, "step": 27030 }, { "grad_norm": 0.06934598833322525, "learning_rate": 8.729815848133618e-05, "loss": 0.006, "step": 27040 }, { "grad_norm": 0.10018966346979141, "learning_rate": 8.728714454477247e-05, "loss": 0.0066, "step": 27050 }, { "grad_norm": 0.12365851551294327, "learning_rate": 8.727612653054269e-05, "loss": 0.0066, "step": 27060 }, { "grad_norm": 0.1665346324443817, "learning_rate": 8.726510443985176e-05, "loss": 0.0067, "step": 27070 }, { "grad_norm": 0.1007365733385086, "learning_rate": 8.725407827390503e-05, "loss": 0.006, "step": 27080 }, { "grad_norm": 0.11207126080989838, "learning_rate": 8.724304803390833e-05, "loss": 0.0056, "step": 27090 }, { "grad_norm": 0.11289842426776886, "learning_rate": 8.723201372106788e-05, "loss": 0.0064, "step": 27100 }, { "grad_norm": 0.0752837210893631, "learning_rate": 8.722097533659038e-05, "loss": 0.0055, "step": 27110 }, { "grad_norm": 0.1164155974984169, "learning_rate": 8.720993288168299e-05, "loss": 0.006, "step": 27120 }, { "grad_norm": 0.11470785737037659, "learning_rate": 8.719888635755327e-05, "loss": 0.006, "step": 27130 }, { "grad_norm": 0.09024831652641296, "learning_rate": 8.718783576540928e-05, "loss": 0.0051, "step": 27140 }, { "grad_norm": 0.09016584604978561, "learning_rate": 8.717678110645948e-05, "loss": 0.0058, "step": 27150 }, { "grad_norm": 0.09964700043201447, "learning_rate": 8.716572238191279e-05, "loss": 0.0054, "step": 27160 }, { "grad_norm": 0.10261007398366928, "learning_rate": 8.715465959297857e-05, "loss": 0.0061, "step": 27170 }, { "grad_norm": 0.08435088396072388, "learning_rate": 8.714359274086665e-05, "loss": 0.0061, "step": 27180 }, { "grad_norm": 0.06736753135919571, "learning_rate": 8.713252182678726e-05, "loss": 0.0057, "step": 27190 }, { "grad_norm": 0.09874417632818222, "learning_rate": 8.712144685195112e-05, "loss": 0.0063, "step": 27200 }, { "grad_norm": 0.10455476492643356, "learning_rate": 8.711036781756936e-05, "loss": 0.0057, "step": 27210 }, { "grad_norm": 0.11614499241113663, "learning_rate": 8.709928472485357e-05, "loss": 0.0061, "step": 27220 }, { "grad_norm": 0.0886281207203865, "learning_rate": 8.708819757501579e-05, "loss": 0.0061, "step": 27230 }, { "grad_norm": 0.0866553857922554, "learning_rate": 8.707710636926846e-05, "loss": 0.0065, "step": 27240 }, { "grad_norm": 0.07744000107049942, "learning_rate": 8.706601110882455e-05, "loss": 0.0058, "step": 27250 }, { "grad_norm": 0.11335629224777222, "learning_rate": 8.705491179489738e-05, "loss": 0.0066, "step": 27260 }, { "grad_norm": 0.09637326747179031, "learning_rate": 8.704380842870077e-05, "loss": 0.0057, "step": 27270 }, { "grad_norm": 0.12729282677173615, "learning_rate": 8.703270101144895e-05, "loss": 0.0059, "step": 27280 }, { "grad_norm": 0.10709984600543976, "learning_rate": 8.702158954435664e-05, "loss": 0.0055, "step": 27290 }, { "grad_norm": 0.07704029232263565, "learning_rate": 8.701047402863896e-05, "loss": 0.0057, "step": 27300 }, { "grad_norm": 0.09921188652515411, "learning_rate": 8.699935446551148e-05, "loss": 0.0063, "step": 27310 }, { "grad_norm": 0.09673301875591278, "learning_rate": 8.698823085619022e-05, "loss": 0.0057, "step": 27320 }, { "grad_norm": 0.08491779118776321, "learning_rate": 8.697710320189166e-05, "loss": 0.0062, "step": 27330 }, { "grad_norm": 0.08424355089664459, "learning_rate": 8.696597150383268e-05, "loss": 0.0063, "step": 27340 }, { "grad_norm": 0.11665070801973343, "learning_rate": 8.695483576323063e-05, "loss": 0.0051, "step": 27350 }, { "grad_norm": 0.08038531988859177, "learning_rate": 8.69436959813033e-05, "loss": 0.0059, "step": 27360 }, { "grad_norm": 0.08169329911470413, "learning_rate": 8.693255215926892e-05, "loss": 0.0066, "step": 27370 }, { "grad_norm": 0.07348084449768066, "learning_rate": 8.692140429834617e-05, "loss": 0.0068, "step": 27380 }, { "grad_norm": 0.132622629404068, "learning_rate": 8.691025239975415e-05, "loss": 0.0063, "step": 27390 }, { "grad_norm": 0.15778884291648865, "learning_rate": 8.689909646471243e-05, "loss": 0.0055, "step": 27400 }, { "grad_norm": 0.0767272338271141, "learning_rate": 8.688793649444099e-05, "loss": 0.0062, "step": 27410 }, { "grad_norm": 0.1108819916844368, "learning_rate": 8.687677249016029e-05, "loss": 0.0062, "step": 27420 }, { "grad_norm": 0.061103932559490204, "learning_rate": 8.686560445309118e-05, "loss": 0.0056, "step": 27430 }, { "grad_norm": 0.1153397187590599, "learning_rate": 8.685443238445499e-05, "loss": 0.0059, "step": 27440 }, { "grad_norm": 0.1139264777302742, "learning_rate": 8.68432562854735e-05, "loss": 0.0069, "step": 27450 }, { "grad_norm": 0.15249361097812653, "learning_rate": 8.683207615736887e-05, "loss": 0.0069, "step": 27460 }, { "grad_norm": 0.12926709651947021, "learning_rate": 8.682089200136379e-05, "loss": 0.006, "step": 27470 }, { "grad_norm": 0.10542689263820648, "learning_rate": 8.680970381868132e-05, "loss": 0.0065, "step": 27480 }, { "grad_norm": 0.1250101625919342, "learning_rate": 8.679851161054498e-05, "loss": 0.0061, "step": 27490 }, { "grad_norm": 0.06979770213365555, "learning_rate": 8.678731537817873e-05, "loss": 0.0059, "step": 27500 }, { "grad_norm": 0.1135697215795517, "learning_rate": 8.677611512280697e-05, "loss": 0.0064, "step": 27510 }, { "grad_norm": 0.08212529122829437, "learning_rate": 8.676491084565457e-05, "loss": 0.0063, "step": 27520 }, { "grad_norm": 0.1027425080537796, "learning_rate": 8.675370254794678e-05, "loss": 0.0052, "step": 27530 }, { "grad_norm": 0.10440777242183685, "learning_rate": 8.674249023090935e-05, "loss": 0.0059, "step": 27540 }, { "grad_norm": 0.0938803181052208, "learning_rate": 8.673127389576843e-05, "loss": 0.0068, "step": 27550 }, { "grad_norm": 0.09319949150085449, "learning_rate": 8.67200535437506e-05, "loss": 0.0064, "step": 27560 }, { "grad_norm": 0.0929044857621193, "learning_rate": 8.670882917608296e-05, "loss": 0.0055, "step": 27570 }, { "grad_norm": 0.08510461449623108, "learning_rate": 8.669760079399292e-05, "loss": 0.0059, "step": 27580 }, { "grad_norm": 0.09912239015102386, "learning_rate": 8.668636839870845e-05, "loss": 0.0063, "step": 27590 }, { "grad_norm": 0.08132375776767731, "learning_rate": 8.667513199145789e-05, "loss": 0.0071, "step": 27600 }, { "grad_norm": 0.11504913866519928, "learning_rate": 8.666389157347002e-05, "loss": 0.0059, "step": 27610 }, { "grad_norm": 0.14850859344005585, "learning_rate": 8.66526471459741e-05, "loss": 0.006, "step": 27620 }, { "grad_norm": 0.13384394347667694, "learning_rate": 8.66413987101998e-05, "loss": 0.0063, "step": 27630 }, { "grad_norm": 0.1012975201010704, "learning_rate": 8.663014626737723e-05, "loss": 0.0062, "step": 27640 }, { "grad_norm": 0.10902118682861328, "learning_rate": 8.661888981873691e-05, "loss": 0.006, "step": 27650 }, { "grad_norm": 0.18244321644306183, "learning_rate": 8.660762936550988e-05, "loss": 0.0067, "step": 27660 }, { "grad_norm": 0.1310531347990036, "learning_rate": 8.659636490892753e-05, "loss": 0.0069, "step": 27670 }, { "grad_norm": 0.1418515294790268, "learning_rate": 8.658509645022174e-05, "loss": 0.0082, "step": 27680 }, { "grad_norm": 0.18618836998939514, "learning_rate": 8.657382399062481e-05, "loss": 0.0086, "step": 27690 }, { "grad_norm": 0.0913124606013298, "learning_rate": 8.656254753136946e-05, "loss": 0.0087, "step": 27700 }, { "grad_norm": 0.13957913219928741, "learning_rate": 8.655126707368891e-05, "loss": 0.0081, "step": 27710 }, { "grad_norm": 0.1306142807006836, "learning_rate": 8.653998261881672e-05, "loss": 0.0067, "step": 27720 }, { "grad_norm": 0.17913420498371124, "learning_rate": 8.652869416798699e-05, "loss": 0.0075, "step": 27730 }, { "grad_norm": 0.13395823538303375, "learning_rate": 8.651740172243417e-05, "loss": 0.0074, "step": 27740 }, { "grad_norm": 0.11319456994533539, "learning_rate": 8.65061052833932e-05, "loss": 0.0074, "step": 27750 }, { "grad_norm": 0.12650653719902039, "learning_rate": 8.649480485209945e-05, "loss": 0.0064, "step": 27760 }, { "grad_norm": 0.07633552700281143, "learning_rate": 8.64835004297887e-05, "loss": 0.0066, "step": 27770 }, { "grad_norm": 0.0972241684794426, "learning_rate": 8.64721920176972e-05, "loss": 0.0058, "step": 27780 }, { "grad_norm": 0.11673504114151001, "learning_rate": 8.646087961706164e-05, "loss": 0.0071, "step": 27790 }, { "grad_norm": 0.13756752014160156, "learning_rate": 8.644956322911908e-05, "loss": 0.0066, "step": 27800 }, { "grad_norm": 0.1748843938112259, "learning_rate": 8.643824285510709e-05, "loss": 0.0085, "step": 27810 }, { "grad_norm": 0.1253460943698883, "learning_rate": 8.642691849626364e-05, "loss": 0.0076, "step": 27820 }, { "grad_norm": 0.10133055597543716, "learning_rate": 8.641559015382717e-05, "loss": 0.0072, "step": 27830 }, { "grad_norm": 0.10457088053226471, "learning_rate": 8.640425782903649e-05, "loss": 0.0082, "step": 27840 }, { "grad_norm": 0.10824693739414215, "learning_rate": 8.639292152313091e-05, "loss": 0.0072, "step": 27850 }, { "grad_norm": 0.11409828811883926, "learning_rate": 8.638158123735015e-05, "loss": 0.0062, "step": 27860 }, { "grad_norm": 0.11964543908834457, "learning_rate": 8.637023697293436e-05, "loss": 0.0084, "step": 27870 }, { "grad_norm": 0.0851965919137001, "learning_rate": 8.635888873112414e-05, "loss": 0.007, "step": 27880 }, { "grad_norm": 0.09278776496648788, "learning_rate": 8.634753651316052e-05, "loss": 0.007, "step": 27890 }, { "grad_norm": 0.08639003336429596, "learning_rate": 8.633618032028496e-05, "loss": 0.0065, "step": 27900 }, { "grad_norm": 0.11549056321382523, "learning_rate": 8.632482015373934e-05, "loss": 0.0058, "step": 27910 }, { "grad_norm": 0.11005351692438126, "learning_rate": 8.6313456014766e-05, "loss": 0.0067, "step": 27920 }, { "grad_norm": 0.10021322965621948, "learning_rate": 8.630208790460771e-05, "loss": 0.006, "step": 27930 }, { "grad_norm": 0.10744816809892654, "learning_rate": 8.629071582450768e-05, "loss": 0.0066, "step": 27940 }, { "grad_norm": 0.07141602039337158, "learning_rate": 8.62793397757095e-05, "loss": 0.0056, "step": 27950 }, { "grad_norm": 0.10395091027021408, "learning_rate": 8.626795975945729e-05, "loss": 0.0065, "step": 27960 }, { "grad_norm": 0.08862226456403732, "learning_rate": 8.625657577699551e-05, "loss": 0.0057, "step": 27970 }, { "grad_norm": 0.11853686720132828, "learning_rate": 8.624518782956914e-05, "loss": 0.0067, "step": 27980 }, { "grad_norm": 0.08585062623023987, "learning_rate": 8.62337959184235e-05, "loss": 0.0057, "step": 27990 }, { "grad_norm": 0.15406323969364166, "learning_rate": 8.622240004480441e-05, "loss": 0.0067, "step": 28000 }, { "grad_norm": 0.09502924978733063, "learning_rate": 8.621100020995814e-05, "loss": 0.0064, "step": 28010 }, { "grad_norm": 0.13196615874767303, "learning_rate": 8.619959641513132e-05, "loss": 0.0074, "step": 28020 }, { "grad_norm": 0.08685281127691269, "learning_rate": 8.618818866157105e-05, "loss": 0.0066, "step": 28030 }, { "grad_norm": 0.06732708215713501, "learning_rate": 8.617677695052487e-05, "loss": 0.006, "step": 28040 }, { "grad_norm": 0.09672953933477402, "learning_rate": 8.616536128324078e-05, "loss": 0.007, "step": 28050 }, { "grad_norm": 0.1337481290102005, "learning_rate": 8.615394166096712e-05, "loss": 0.0065, "step": 28060 }, { "grad_norm": 0.12155748158693314, "learning_rate": 8.614251808495279e-05, "loss": 0.0075, "step": 28070 }, { "grad_norm": 0.07688301056623459, "learning_rate": 8.6131090556447e-05, "loss": 0.0059, "step": 28080 }, { "grad_norm": 0.08716367930173874, "learning_rate": 8.611965907669947e-05, "loss": 0.0058, "step": 28090 }, { "grad_norm": 0.09675691276788712, "learning_rate": 8.610822364696034e-05, "loss": 0.0056, "step": 28100 }, { "grad_norm": 0.1138535812497139, "learning_rate": 8.609678426848015e-05, "loss": 0.0071, "step": 28110 }, { "grad_norm": 0.09736268222332001, "learning_rate": 8.60853409425099e-05, "loss": 0.0056, "step": 28120 }, { "grad_norm": 0.08309423178434372, "learning_rate": 8.607389367030104e-05, "loss": 0.0064, "step": 28130 }, { "grad_norm": 0.0809098482131958, "learning_rate": 8.606244245310538e-05, "loss": 0.0061, "step": 28140 }, { "grad_norm": 0.08751780539751053, "learning_rate": 8.605098729217525e-05, "loss": 0.0068, "step": 28150 }, { "grad_norm": 0.11656979471445084, "learning_rate": 8.603952818876335e-05, "loss": 0.0083, "step": 28160 }, { "grad_norm": 0.07153504341840744, "learning_rate": 8.602806514412281e-05, "loss": 0.006, "step": 28170 }, { "grad_norm": 0.09929590672254562, "learning_rate": 8.601659815950726e-05, "loss": 0.0069, "step": 28180 }, { "grad_norm": 0.09413973242044449, "learning_rate": 8.600512723617067e-05, "loss": 0.0057, "step": 28190 }, { "grad_norm": 0.07580946385860443, "learning_rate": 8.59936523753675e-05, "loss": 0.0069, "step": 28200 }, { "grad_norm": 0.11112987995147705, "learning_rate": 8.598217357835264e-05, "loss": 0.0065, "step": 28210 }, { "grad_norm": 0.09023784101009369, "learning_rate": 8.597069084638135e-05, "loss": 0.0063, "step": 28220 }, { "grad_norm": 0.07521194964647293, "learning_rate": 8.595920418070939e-05, "loss": 0.0063, "step": 28230 }, { "grad_norm": 0.07902568578720093, "learning_rate": 8.594771358259295e-05, "loss": 0.0052, "step": 28240 }, { "grad_norm": 0.11448334902524948, "learning_rate": 8.593621905328858e-05, "loss": 0.0057, "step": 28250 }, { "grad_norm": 0.07914076000452042, "learning_rate": 8.592472059405333e-05, "loss": 0.0057, "step": 28260 }, { "grad_norm": 0.10140808671712875, "learning_rate": 8.591321820614464e-05, "loss": 0.0068, "step": 28270 }, { "grad_norm": 0.12611298263072968, "learning_rate": 8.590171189082041e-05, "loss": 0.0057, "step": 28280 }, { "grad_norm": 0.07250553369522095, "learning_rate": 8.589020164933894e-05, "loss": 0.0056, "step": 28290 }, { "grad_norm": 0.1218738779425621, "learning_rate": 8.587868748295898e-05, "loss": 0.0057, "step": 28300 }, { "grad_norm": 0.1502341777086258, "learning_rate": 8.586716939293971e-05, "loss": 0.0063, "step": 28310 }, { "grad_norm": 0.10085375607013702, "learning_rate": 8.58556473805407e-05, "loss": 0.0062, "step": 28320 }, { "grad_norm": 0.10666307061910629, "learning_rate": 8.584412144702202e-05, "loss": 0.0065, "step": 28330 }, { "grad_norm": 0.10938619077205658, "learning_rate": 8.58325915936441e-05, "loss": 0.0056, "step": 28340 }, { "grad_norm": 0.07654164731502533, "learning_rate": 8.582105782166783e-05, "loss": 0.0046, "step": 28350 }, { "grad_norm": 0.10009705275297165, "learning_rate": 8.580952013235455e-05, "loss": 0.0057, "step": 28360 }, { "grad_norm": 0.1013428196310997, "learning_rate": 8.579797852696596e-05, "loss": 0.0054, "step": 28370 }, { "grad_norm": 0.07380576431751251, "learning_rate": 8.578643300676428e-05, "loss": 0.0054, "step": 28380 }, { "grad_norm": 0.11159330606460571, "learning_rate": 8.577488357301209e-05, "loss": 0.0055, "step": 28390 }, { "grad_norm": 0.06544379889965057, "learning_rate": 8.576333022697242e-05, "loss": 0.0058, "step": 28400 }, { "grad_norm": 0.12553369998931885, "learning_rate": 8.575177296990873e-05, "loss": 0.0056, "step": 28410 }, { "grad_norm": 0.10948129743337631, "learning_rate": 8.574021180308489e-05, "loss": 0.0062, "step": 28420 }, { "grad_norm": 0.10296601057052612, "learning_rate": 8.572864672776523e-05, "loss": 0.0055, "step": 28430 }, { "grad_norm": 0.12356085330247879, "learning_rate": 8.571707774521447e-05, "loss": 0.0066, "step": 28440 }, { "grad_norm": 0.11520729959011078, "learning_rate": 8.57055048566978e-05, "loss": 0.0058, "step": 28450 }, { "grad_norm": 0.11612621694803238, "learning_rate": 8.569392806348078e-05, "loss": 0.0054, "step": 28460 }, { "grad_norm": 0.09453852474689484, "learning_rate": 8.568234736682947e-05, "loss": 0.0055, "step": 28470 }, { "grad_norm": 0.12113562226295471, "learning_rate": 8.567076276801029e-05, "loss": 0.0058, "step": 28480 }, { "grad_norm": 0.1022225171327591, "learning_rate": 8.565917426829013e-05, "loss": 0.0059, "step": 28490 }, { "grad_norm": 0.09585157781839371, "learning_rate": 8.564758186893628e-05, "loss": 0.0056, "step": 28500 }, { "grad_norm": 0.07514737546443939, "learning_rate": 8.563598557121649e-05, "loss": 0.0049, "step": 28510 }, { "grad_norm": 0.10731843113899231, "learning_rate": 8.562438537639888e-05, "loss": 0.0061, "step": 28520 }, { "grad_norm": 0.08305666595697403, "learning_rate": 8.561278128575206e-05, "loss": 0.0051, "step": 28530 }, { "grad_norm": 0.0933310016989708, "learning_rate": 8.5601173300545e-05, "loss": 0.0061, "step": 28540 }, { "grad_norm": 0.09388124197721481, "learning_rate": 8.558956142204717e-05, "loss": 0.0059, "step": 28550 }, { "grad_norm": 0.062093112617731094, "learning_rate": 8.55779456515284e-05, "loss": 0.0055, "step": 28560 }, { "grad_norm": 0.09971511363983154, "learning_rate": 8.556632599025898e-05, "loss": 0.0049, "step": 28570 }, { "grad_norm": 0.1567281186580658, "learning_rate": 8.555470243950964e-05, "loss": 0.0059, "step": 28580 }, { "grad_norm": 0.15510733425617218, "learning_rate": 8.554307500055148e-05, "loss": 0.0059, "step": 28590 }, { "grad_norm": 0.09043712168931961, "learning_rate": 8.553144367465609e-05, "loss": 0.0057, "step": 28600 }, { "grad_norm": 0.06490062922239304, "learning_rate": 8.551980846309544e-05, "loss": 0.0057, "step": 28610 }, { "grad_norm": 0.13738518953323364, "learning_rate": 8.550816936714193e-05, "loss": 0.0074, "step": 28620 }, { "grad_norm": 0.09885197132825851, "learning_rate": 8.549652638806841e-05, "loss": 0.0061, "step": 28630 }, { "grad_norm": 0.0964825227856636, "learning_rate": 8.548487952714812e-05, "loss": 0.0062, "step": 28640 }, { "grad_norm": 0.11868531256914139, "learning_rate": 8.547322878565478e-05, "loss": 0.0053, "step": 28650 }, { "grad_norm": 0.10541929304599762, "learning_rate": 8.546157416486245e-05, "loss": 0.0062, "step": 28660 }, { "grad_norm": 0.08670956641435623, "learning_rate": 8.54499156660457e-05, "loss": 0.0058, "step": 28670 }, { "grad_norm": 0.16221210360527039, "learning_rate": 8.543825329047947e-05, "loss": 0.0067, "step": 28680 }, { "grad_norm": 0.1663188636302948, "learning_rate": 8.542658703943913e-05, "loss": 0.0069, "step": 28690 }, { "grad_norm": 0.08263424783945084, "learning_rate": 8.541491691420051e-05, "loss": 0.0066, "step": 28700 }, { "grad_norm": 0.11513595283031464, "learning_rate": 8.54032429160398e-05, "loss": 0.006, "step": 28710 }, { "grad_norm": 0.10432489961385727, "learning_rate": 8.539156504623369e-05, "loss": 0.006, "step": 28720 }, { "grad_norm": 0.07770485430955887, "learning_rate": 8.537988330605923e-05, "loss": 0.006, "step": 28730 }, { "grad_norm": 0.0874713584780693, "learning_rate": 8.536819769679393e-05, "loss": 0.0055, "step": 28740 }, { "grad_norm": 0.08388703316450119, "learning_rate": 8.53565082197157e-05, "loss": 0.0048, "step": 28750 }, { "grad_norm": 0.11222291737794876, "learning_rate": 8.534481487610289e-05, "loss": 0.0054, "step": 28760 }, { "grad_norm": 0.09871142357587814, "learning_rate": 8.533311766723428e-05, "loss": 0.0056, "step": 28770 }, { "grad_norm": 0.11976365000009537, "learning_rate": 8.532141659438901e-05, "loss": 0.006, "step": 28780 }, { "grad_norm": 0.1173616573214531, "learning_rate": 8.530971165884675e-05, "loss": 0.005, "step": 28790 }, { "grad_norm": 0.09266577661037445, "learning_rate": 8.529800286188752e-05, "loss": 0.0053, "step": 28800 }, { "grad_norm": 0.11572317034006119, "learning_rate": 8.528629020479175e-05, "loss": 0.0056, "step": 28810 }, { "grad_norm": 0.07998374849557877, "learning_rate": 8.527457368884033e-05, "loss": 0.0051, "step": 28820 }, { "grad_norm": 0.09478239715099335, "learning_rate": 8.526285331531458e-05, "loss": 0.005, "step": 28830 }, { "grad_norm": 0.11682485789060593, "learning_rate": 8.525112908549621e-05, "loss": 0.0057, "step": 28840 }, { "grad_norm": 0.1258527934551239, "learning_rate": 8.523940100066735e-05, "loss": 0.006, "step": 28850 }, { "grad_norm": 0.12257412075996399, "learning_rate": 8.52276690621106e-05, "loss": 0.0053, "step": 28860 }, { "grad_norm": 0.1030609980225563, "learning_rate": 8.521593327110889e-05, "loss": 0.0059, "step": 28870 }, { "grad_norm": 0.088913694024086, "learning_rate": 8.520419362894569e-05, "loss": 0.0061, "step": 28880 }, { "grad_norm": 0.09325308352708817, "learning_rate": 8.51924501369048e-05, "loss": 0.0052, "step": 28890 }, { "grad_norm": 0.11024152487516403, "learning_rate": 8.518070279627047e-05, "loss": 0.0067, "step": 28900 }, { "grad_norm": 0.08939896523952484, "learning_rate": 8.516895160832737e-05, "loss": 0.0056, "step": 28910 }, { "grad_norm": 0.1120666116476059, "learning_rate": 8.515719657436061e-05, "loss": 0.0051, "step": 28920 }, { "grad_norm": 0.07703156024217606, "learning_rate": 8.514543769565568e-05, "loss": 0.0056, "step": 28930 }, { "grad_norm": 0.07550881057977676, "learning_rate": 8.513367497349853e-05, "loss": 0.0043, "step": 28940 }, { "grad_norm": 0.0953260213136673, "learning_rate": 8.51219084091755e-05, "loss": 0.0057, "step": 28950 }, { "grad_norm": 0.09316819161176682, "learning_rate": 8.511013800397338e-05, "loss": 0.0053, "step": 28960 }, { "grad_norm": 0.09663723409175873, "learning_rate": 8.509836375917937e-05, "loss": 0.0054, "step": 28970 }, { "grad_norm": 0.08961985260248184, "learning_rate": 8.508658567608104e-05, "loss": 0.0055, "step": 28980 }, { "grad_norm": 0.11175328493118286, "learning_rate": 8.507480375596647e-05, "loss": 0.0057, "step": 28990 }, { "grad_norm": 0.08921270072460175, "learning_rate": 8.506301800012408e-05, "loss": 0.0054, "step": 29000 }, { "grad_norm": 0.1300271451473236, "learning_rate": 8.505122840984278e-05, "loss": 0.0048, "step": 29010 }, { "grad_norm": 0.09002631902694702, "learning_rate": 8.503943498641182e-05, "loss": 0.0052, "step": 29020 }, { "grad_norm": 0.08989264816045761, "learning_rate": 8.502763773112095e-05, "loss": 0.006, "step": 29030 }, { "grad_norm": 0.09946487098932266, "learning_rate": 8.501583664526026e-05, "loss": 0.0057, "step": 29040 }, { "grad_norm": 0.11045008152723312, "learning_rate": 8.500403173012032e-05, "loss": 0.0065, "step": 29050 }, { "grad_norm": 0.10969376564025879, "learning_rate": 8.499222298699211e-05, "loss": 0.0049, "step": 29060 }, { "grad_norm": 0.09687976539134979, "learning_rate": 8.498041041716701e-05, "loss": 0.0052, "step": 29070 }, { "grad_norm": 0.0974084883928299, "learning_rate": 8.496859402193681e-05, "loss": 0.0055, "step": 29080 }, { "grad_norm": 0.08818894624710083, "learning_rate": 8.495677380259374e-05, "loss": 0.0053, "step": 29090 }, { "grad_norm": 0.12123794853687286, "learning_rate": 8.494494976043045e-05, "loss": 0.0072, "step": 29100 }, { "grad_norm": 0.09700310975313187, "learning_rate": 8.493312189673998e-05, "loss": 0.0068, "step": 29110 }, { "grad_norm": 0.10594078153371811, "learning_rate": 8.492129021281584e-05, "loss": 0.0057, "step": 29120 }, { "grad_norm": 0.07142796367406845, "learning_rate": 8.490945470995188e-05, "loss": 0.0063, "step": 29130 }, { "grad_norm": 0.07015196233987808, "learning_rate": 8.489761538944247e-05, "loss": 0.0048, "step": 29140 }, { "grad_norm": 0.11729316413402557, "learning_rate": 8.48857722525823e-05, "loss": 0.0068, "step": 29150 }, { "grad_norm": 0.10624965280294418, "learning_rate": 8.487392530066652e-05, "loss": 0.0062, "step": 29160 }, { "grad_norm": 0.09089779853820801, "learning_rate": 8.486207453499069e-05, "loss": 0.0058, "step": 29170 }, { "grad_norm": 0.14053720235824585, "learning_rate": 8.485021995685082e-05, "loss": 0.0067, "step": 29180 }, { "grad_norm": 0.11927703768014908, "learning_rate": 8.483836156754328e-05, "loss": 0.0058, "step": 29190 }, { "grad_norm": 0.1229027733206749, "learning_rate": 8.482649936836491e-05, "loss": 0.006, "step": 29200 }, { "grad_norm": 0.08223932981491089, "learning_rate": 8.481463336061293e-05, "loss": 0.0061, "step": 29210 }, { "grad_norm": 0.08727114647626877, "learning_rate": 8.480276354558496e-05, "loss": 0.0064, "step": 29220 }, { "grad_norm": 0.07423357665538788, "learning_rate": 8.479088992457913e-05, "loss": 0.0064, "step": 29230 }, { "grad_norm": 0.10576703399419785, "learning_rate": 8.477901249889387e-05, "loss": 0.0063, "step": 29240 }, { "grad_norm": 0.10300680994987488, "learning_rate": 8.47671312698281e-05, "loss": 0.0056, "step": 29250 }, { "grad_norm": 0.09716696292161942, "learning_rate": 8.475524623868112e-05, "loss": 0.0057, "step": 29260 }, { "grad_norm": 0.10422216355800629, "learning_rate": 8.474335740675266e-05, "loss": 0.0076, "step": 29270 }, { "grad_norm": 0.11066757142543793, "learning_rate": 8.473146477534289e-05, "loss": 0.0065, "step": 29280 }, { "grad_norm": 0.09086096286773682, "learning_rate": 8.471956834575232e-05, "loss": 0.0064, "step": 29290 }, { "grad_norm": 0.07759246230125427, "learning_rate": 8.470766811928197e-05, "loss": 0.0052, "step": 29300 }, { "grad_norm": 0.08801230043172836, "learning_rate": 8.469576409723323e-05, "loss": 0.0055, "step": 29310 }, { "grad_norm": 0.11852844059467316, "learning_rate": 8.468385628090788e-05, "loss": 0.0053, "step": 29320 }, { "grad_norm": 0.09871581941843033, "learning_rate": 8.467194467160815e-05, "loss": 0.005, "step": 29330 }, { "grad_norm": 0.07729718834161758, "learning_rate": 8.466002927063667e-05, "loss": 0.0067, "step": 29340 }, { "grad_norm": 0.06498250365257263, "learning_rate": 8.464811007929651e-05, "loss": 0.0056, "step": 29350 }, { "grad_norm": 0.0716731920838356, "learning_rate": 8.463618709889114e-05, "loss": 0.0059, "step": 29360 }, { "grad_norm": 0.12371302396059036, "learning_rate": 8.462426033072442e-05, "loss": 0.0063, "step": 29370 }, { "grad_norm": 0.11194377392530441, "learning_rate": 8.461232977610061e-05, "loss": 0.0065, "step": 29380 }, { "grad_norm": 0.08947941660881042, "learning_rate": 8.46003954363245e-05, "loss": 0.0065, "step": 29390 }, { "grad_norm": 0.08835162967443466, "learning_rate": 8.458845731270115e-05, "loss": 0.0073, "step": 29400 }, { "grad_norm": 0.11069284379482269, "learning_rate": 8.45765154065361e-05, "loss": 0.0064, "step": 29410 }, { "grad_norm": 0.08394458889961243, "learning_rate": 8.456456971913532e-05, "loss": 0.0056, "step": 29420 }, { "grad_norm": 0.08954472094774246, "learning_rate": 8.455262025180517e-05, "loss": 0.0053, "step": 29430 }, { "grad_norm": 0.0774906650185585, "learning_rate": 8.454066700585242e-05, "loss": 0.0052, "step": 29440 }, { "grad_norm": 0.09083514660596848, "learning_rate": 8.452870998258423e-05, "loss": 0.0059, "step": 29450 }, { "grad_norm": 0.0871434360742569, "learning_rate": 8.451674918330825e-05, "loss": 0.0054, "step": 29460 }, { "grad_norm": 0.11456456780433655, "learning_rate": 8.450478460933246e-05, "loss": 0.0069, "step": 29470 }, { "grad_norm": 0.08671864867210388, "learning_rate": 8.449281626196532e-05, "loss": 0.0056, "step": 29480 }, { "grad_norm": 0.08169959485530853, "learning_rate": 8.448084414251564e-05, "loss": 0.0046, "step": 29490 }, { "grad_norm": 0.09646974503993988, "learning_rate": 8.446886825229271e-05, "loss": 0.005, "step": 29500 }, { "grad_norm": 0.06900402158498764, "learning_rate": 8.445688859260615e-05, "loss": 0.0055, "step": 29510 }, { "grad_norm": 0.09483013302087784, "learning_rate": 8.444490516476606e-05, "loss": 0.006, "step": 29520 }, { "grad_norm": 0.09175171703100204, "learning_rate": 8.443291797008293e-05, "loss": 0.0057, "step": 29530 }, { "grad_norm": 0.10430698096752167, "learning_rate": 8.442092700986765e-05, "loss": 0.0062, "step": 29540 }, { "grad_norm": 0.13500678539276123, "learning_rate": 8.440893228543156e-05, "loss": 0.0055, "step": 29550 }, { "grad_norm": 0.12402060627937317, "learning_rate": 8.439693379808638e-05, "loss": 0.0052, "step": 29560 }, { "grad_norm": 0.10101013630628586, "learning_rate": 8.43849315491442e-05, "loss": 0.0057, "step": 29570 }, { "grad_norm": 0.09830433130264282, "learning_rate": 8.437292553991763e-05, "loss": 0.0053, "step": 29580 }, { "grad_norm": 0.09238659590482712, "learning_rate": 8.436091577171959e-05, "loss": 0.0058, "step": 29590 }, { "grad_norm": 0.14607931673526764, "learning_rate": 8.434890224586347e-05, "loss": 0.0075, "step": 29600 }, { "grad_norm": 0.10638131946325302, "learning_rate": 8.433688496366303e-05, "loss": 0.0057, "step": 29610 }, { "grad_norm": 0.10382265597581863, "learning_rate": 8.432486392643248e-05, "loss": 0.0048, "step": 29620 }, { "grad_norm": 0.0820130854845047, "learning_rate": 8.431283913548643e-05, "loss": 0.0052, "step": 29630 }, { "grad_norm": 0.09211981296539307, "learning_rate": 8.430081059213985e-05, "loss": 0.0059, "step": 29640 }, { "grad_norm": 0.0826278030872345, "learning_rate": 8.428877829770823e-05, "loss": 0.0058, "step": 29650 }, { "grad_norm": 0.08662501722574234, "learning_rate": 8.427674225350735e-05, "loss": 0.0061, "step": 29660 }, { "grad_norm": 0.12003705650568008, "learning_rate": 8.426470246085347e-05, "loss": 0.0077, "step": 29670 }, { "grad_norm": 0.12380493432283401, "learning_rate": 8.425265892106324e-05, "loss": 0.0076, "step": 29680 }, { "grad_norm": 0.14345820248126984, "learning_rate": 8.424061163545374e-05, "loss": 0.0056, "step": 29690 }, { "grad_norm": 0.09130923449993134, "learning_rate": 8.422856060534243e-05, "loss": 0.0066, "step": 29700 }, { "grad_norm": 0.09144331514835358, "learning_rate": 8.421650583204718e-05, "loss": 0.007, "step": 29710 }, { "grad_norm": 0.08926298469305038, "learning_rate": 8.420444731688633e-05, "loss": 0.0054, "step": 29720 }, { "grad_norm": 0.08942573517560959, "learning_rate": 8.419238506117852e-05, "loss": 0.0052, "step": 29730 }, { "grad_norm": 0.11312582343816757, "learning_rate": 8.418031906624289e-05, "loss": 0.007, "step": 29740 }, { "grad_norm": 0.11529707908630371, "learning_rate": 8.416824933339898e-05, "loss": 0.006, "step": 29750 }, { "grad_norm": 0.09976647049188614, "learning_rate": 8.415617586396667e-05, "loss": 0.0059, "step": 29760 }, { "grad_norm": 0.0940675139427185, "learning_rate": 8.414409865926632e-05, "loss": 0.0064, "step": 29770 }, { "grad_norm": 0.1063423752784729, "learning_rate": 8.413201772061867e-05, "loss": 0.0061, "step": 29780 }, { "grad_norm": 0.08743254095315933, "learning_rate": 8.411993304934488e-05, "loss": 0.0065, "step": 29790 }, { "grad_norm": 0.09828414767980576, "learning_rate": 8.410784464676654e-05, "loss": 0.0054, "step": 29800 }, { "grad_norm": 0.0804239958524704, "learning_rate": 8.409575251420556e-05, "loss": 0.0061, "step": 29810 }, { "grad_norm": 0.10883566737174988, "learning_rate": 8.408365665298435e-05, "loss": 0.0056, "step": 29820 }, { "grad_norm": 0.1345990002155304, "learning_rate": 8.40715570644257e-05, "loss": 0.0064, "step": 29830 }, { "grad_norm": 0.13782954216003418, "learning_rate": 8.40594537498528e-05, "loss": 0.0065, "step": 29840 }, { "grad_norm": 0.0836268737912178, "learning_rate": 8.404734671058924e-05, "loss": 0.0054, "step": 29850 }, { "grad_norm": 0.15829366445541382, "learning_rate": 8.403523594795902e-05, "loss": 0.0058, "step": 29860 }, { "grad_norm": 0.10213883966207504, "learning_rate": 8.402312146328659e-05, "loss": 0.0055, "step": 29870 }, { "grad_norm": 0.12132079154253006, "learning_rate": 8.401100325789675e-05, "loss": 0.0058, "step": 29880 }, { "grad_norm": 0.09110510349273682, "learning_rate": 8.399888133311472e-05, "loss": 0.0061, "step": 29890 }, { "grad_norm": 0.08809953927993774, "learning_rate": 8.398675569026613e-05, "loss": 0.0053, "step": 29900 }, { "grad_norm": 0.15753644704818726, "learning_rate": 8.397462633067705e-05, "loss": 0.0065, "step": 29910 }, { "grad_norm": 0.1214558333158493, "learning_rate": 8.396249325567392e-05, "loss": 0.0086, "step": 29920 }, { "grad_norm": 0.14369778335094452, "learning_rate": 8.395035646658357e-05, "loss": 0.0102, "step": 29930 }, { "grad_norm": 0.10801433026790619, "learning_rate": 8.39382159647333e-05, "loss": 0.008, "step": 29940 }, { "grad_norm": 0.09128688275814056, "learning_rate": 8.392607175145075e-05, "loss": 0.007, "step": 29950 }, { "grad_norm": 0.1094067394733429, "learning_rate": 8.3913923828064e-05, "loss": 0.0069, "step": 29960 }, { "grad_norm": 0.07877183705568314, "learning_rate": 8.390177219590152e-05, "loss": 0.0064, "step": 29970 }, { "grad_norm": 0.08262689411640167, "learning_rate": 8.388961685629222e-05, "loss": 0.0058, "step": 29980 }, { "grad_norm": 0.12799449265003204, "learning_rate": 8.387745781056536e-05, "loss": 0.0064, "step": 29990 }, { "grad_norm": 0.09439747035503387, "learning_rate": 8.386529506005065e-05, "loss": 0.0059, "step": 30000 }, { "grad_norm": 0.08868951350450516, "learning_rate": 8.38531286060782e-05, "loss": 0.0064, "step": 30010 }, { "grad_norm": 0.09727774560451508, "learning_rate": 8.384095844997849e-05, "loss": 0.0059, "step": 30020 }, { "grad_norm": 0.09796050935983658, "learning_rate": 8.382878459308245e-05, "loss": 0.0051, "step": 30030 }, { "grad_norm": 0.13780397176742554, "learning_rate": 8.381660703672138e-05, "loss": 0.0065, "step": 30040 }, { "grad_norm": 0.0856325775384903, "learning_rate": 8.380442578222702e-05, "loss": 0.0049, "step": 30050 }, { "grad_norm": 0.10721193253993988, "learning_rate": 8.379224083093146e-05, "loss": 0.0055, "step": 30060 }, { "grad_norm": 0.09944172948598862, "learning_rate": 8.378005218416727e-05, "loss": 0.0058, "step": 30070 }, { "grad_norm": 0.11275215446949005, "learning_rate": 8.376785984326735e-05, "loss": 0.0051, "step": 30080 }, { "grad_norm": 0.06691024452447891, "learning_rate": 8.375566380956506e-05, "loss": 0.0051, "step": 30090 }, { "grad_norm": 0.08783702552318573, "learning_rate": 8.374346408439411e-05, "loss": 0.0055, "step": 30100 }, { "grad_norm": 0.0928642526268959, "learning_rate": 8.373126066908868e-05, "loss": 0.006, "step": 30110 }, { "grad_norm": 0.11541660130023956, "learning_rate": 8.371905356498326e-05, "loss": 0.0061, "step": 30120 }, { "grad_norm": 0.13510939478874207, "learning_rate": 8.370684277341288e-05, "loss": 0.006, "step": 30130 }, { "grad_norm": 0.10280255228281021, "learning_rate": 8.369462829571282e-05, "loss": 0.0056, "step": 30140 }, { "grad_norm": 0.093548983335495, "learning_rate": 8.36824101332189e-05, "loss": 0.005, "step": 30150 }, { "grad_norm": 0.11446633189916611, "learning_rate": 8.367018828726721e-05, "loss": 0.005, "step": 30160 }, { "grad_norm": 0.09583988785743713, "learning_rate": 8.365796275919438e-05, "loss": 0.0051, "step": 30170 }, { "grad_norm": 0.08436553925275803, "learning_rate": 8.364573355033734e-05, "loss": 0.0057, "step": 30180 }, { "grad_norm": 0.09928467869758606, "learning_rate": 8.363350066203346e-05, "loss": 0.007, "step": 30190 }, { "grad_norm": 0.08227257430553436, "learning_rate": 8.362126409562053e-05, "loss": 0.0059, "step": 30200 }, { "grad_norm": 0.11380435526371002, "learning_rate": 8.360902385243667e-05, "loss": 0.0054, "step": 30210 }, { "grad_norm": 0.08153247088193893, "learning_rate": 8.359677993382052e-05, "loss": 0.0079, "step": 30220 }, { "grad_norm": 0.08872176706790924, "learning_rate": 8.358453234111103e-05, "loss": 0.006, "step": 30230 }, { "grad_norm": 0.10647431761026382, "learning_rate": 8.357228107564756e-05, "loss": 0.007, "step": 30240 }, { "grad_norm": 0.10558200627565384, "learning_rate": 8.356002613876993e-05, "loss": 0.008, "step": 30250 }, { "grad_norm": 0.10816218703985214, "learning_rate": 8.35477675318183e-05, "loss": 0.0078, "step": 30260 }, { "grad_norm": 0.09868203103542328, "learning_rate": 8.353550525613323e-05, "loss": 0.0072, "step": 30270 }, { "grad_norm": 0.12305299937725067, "learning_rate": 8.352323931305572e-05, "loss": 0.0069, "step": 30280 }, { "grad_norm": 0.13453467190265656, "learning_rate": 8.351096970392717e-05, "loss": 0.008, "step": 30290 }, { "grad_norm": 0.10563497990369797, "learning_rate": 8.349869643008937e-05, "loss": 0.0059, "step": 30300 }, { "grad_norm": 0.10512752830982208, "learning_rate": 8.348641949288449e-05, "loss": 0.0061, "step": 30310 }, { "grad_norm": 0.12241173535585403, "learning_rate": 8.34741388936551e-05, "loss": 0.0066, "step": 30320 }, { "grad_norm": 0.09501872956752777, "learning_rate": 8.346185463374423e-05, "loss": 0.0063, "step": 30330 }, { "grad_norm": 0.0968979224562645, "learning_rate": 8.344956671449524e-05, "loss": 0.0059, "step": 30340 }, { "grad_norm": 0.10092738270759583, "learning_rate": 8.343727513725192e-05, "loss": 0.0051, "step": 30350 }, { "grad_norm": 0.09775170683860779, "learning_rate": 8.342497990335847e-05, "loss": 0.006, "step": 30360 }, { "grad_norm": 0.10693950206041336, "learning_rate": 8.341268101415946e-05, "loss": 0.0062, "step": 30370 }, { "grad_norm": 0.09467554092407227, "learning_rate": 8.34003784709999e-05, "loss": 0.0066, "step": 30380 }, { "grad_norm": 0.08296223729848862, "learning_rate": 8.338807227522516e-05, "loss": 0.0058, "step": 30390 }, { "grad_norm": 0.08403544872999191, "learning_rate": 8.337576242818103e-05, "loss": 0.0061, "step": 30400 }, { "grad_norm": 0.10806674510240555, "learning_rate": 8.336344893121372e-05, "loss": 0.0063, "step": 30410 }, { "grad_norm": 0.08667417615652084, "learning_rate": 8.335113178566977e-05, "loss": 0.006, "step": 30420 }, { "grad_norm": 0.08223841339349747, "learning_rate": 8.33388109928962e-05, "loss": 0.0064, "step": 30430 }, { "grad_norm": 0.12215513736009598, "learning_rate": 8.33264865542404e-05, "loss": 0.0056, "step": 30440 }, { "grad_norm": 0.0831843912601471, "learning_rate": 8.331415847105013e-05, "loss": 0.0063, "step": 30450 }, { "grad_norm": 0.09793562442064285, "learning_rate": 8.330182674467357e-05, "loss": 0.0054, "step": 30460 }, { "grad_norm": 0.11479897797107697, "learning_rate": 8.32894913764593e-05, "loss": 0.0052, "step": 30470 }, { "grad_norm": 0.09185143560171127, "learning_rate": 8.327715236775633e-05, "loss": 0.0067, "step": 30480 }, { "grad_norm": 0.08347487449645996, "learning_rate": 8.326480971991398e-05, "loss": 0.0065, "step": 30490 }, { "grad_norm": 0.09325137734413147, "learning_rate": 8.325246343428206e-05, "loss": 0.0052, "step": 30500 }, { "grad_norm": 0.06785605102777481, "learning_rate": 8.324011351221072e-05, "loss": 0.0058, "step": 30510 }, { "grad_norm": 0.09381444752216339, "learning_rate": 8.322775995505057e-05, "loss": 0.0064, "step": 30520 }, { "grad_norm": 0.11949396133422852, "learning_rate": 8.321540276415254e-05, "loss": 0.0061, "step": 30530 }, { "grad_norm": 0.11312435567378998, "learning_rate": 8.320304194086798e-05, "loss": 0.0054, "step": 30540 }, { "grad_norm": 0.10845266282558441, "learning_rate": 8.31906774865487e-05, "loss": 0.0052, "step": 30550 }, { "grad_norm": 0.09447989612817764, "learning_rate": 8.317830940254682e-05, "loss": 0.0055, "step": 30560 }, { "grad_norm": 0.08361494541168213, "learning_rate": 8.316593769021491e-05, "loss": 0.0054, "step": 30570 }, { "grad_norm": 0.1152341440320015, "learning_rate": 8.315356235090592e-05, "loss": 0.0062, "step": 30580 }, { "grad_norm": 0.10370723903179169, "learning_rate": 8.314118338597319e-05, "loss": 0.007, "step": 30590 }, { "grad_norm": 0.13103023171424866, "learning_rate": 8.312880079677048e-05, "loss": 0.0051, "step": 30600 }, { "grad_norm": 0.1178681030869484, "learning_rate": 8.311641458465191e-05, "loss": 0.0066, "step": 30610 }, { "grad_norm": 0.08102329820394516, "learning_rate": 8.310402475097205e-05, "loss": 0.0056, "step": 30620 }, { "grad_norm": 0.0974535346031189, "learning_rate": 8.309163129708581e-05, "loss": 0.0064, "step": 30630 }, { "grad_norm": 0.09683409333229065, "learning_rate": 8.307923422434852e-05, "loss": 0.0061, "step": 30640 }, { "grad_norm": 0.11830434948205948, "learning_rate": 8.30668335341159e-05, "loss": 0.0069, "step": 30650 }, { "grad_norm": 0.07919679582118988, "learning_rate": 8.305442922774408e-05, "loss": 0.0049, "step": 30660 }, { "grad_norm": 0.09526797384023666, "learning_rate": 8.304202130658959e-05, "loss": 0.0051, "step": 30670 }, { "grad_norm": 0.1267169713973999, "learning_rate": 8.302960977200931e-05, "loss": 0.0058, "step": 30680 }, { "grad_norm": 0.07485530525445938, "learning_rate": 8.301719462536058e-05, "loss": 0.0048, "step": 30690 }, { "grad_norm": 0.08271685242652893, "learning_rate": 8.300477586800108e-05, "loss": 0.0053, "step": 30700 }, { "grad_norm": 0.09794146567583084, "learning_rate": 8.299235350128892e-05, "loss": 0.0052, "step": 30710 }, { "grad_norm": 0.08565759658813477, "learning_rate": 8.297992752658258e-05, "loss": 0.005, "step": 30720 }, { "grad_norm": 0.11992228031158447, "learning_rate": 8.296749794524094e-05, "loss": 0.0063, "step": 30730 }, { "grad_norm": 0.09715692698955536, "learning_rate": 8.295506475862332e-05, "loss": 0.0065, "step": 30740 }, { "grad_norm": 0.10107166320085526, "learning_rate": 8.294262796808933e-05, "loss": 0.0056, "step": 30750 }, { "grad_norm": 0.07967779785394669, "learning_rate": 8.293018757499909e-05, "loss": 0.0048, "step": 30760 }, { "grad_norm": 0.08607984334230423, "learning_rate": 8.291774358071305e-05, "loss": 0.0056, "step": 30770 }, { "grad_norm": 0.09299922734498978, "learning_rate": 8.290529598659205e-05, "loss": 0.005, "step": 30780 }, { "grad_norm": 0.10662306100130081, "learning_rate": 8.289284479399738e-05, "loss": 0.0057, "step": 30790 }, { "grad_norm": 0.08406666666269302, "learning_rate": 8.288039000429064e-05, "loss": 0.0052, "step": 30800 }, { "grad_norm": 0.11884917318820953, "learning_rate": 8.286793161883388e-05, "loss": 0.0051, "step": 30810 }, { "grad_norm": 0.08079204708337784, "learning_rate": 8.285546963898954e-05, "loss": 0.0049, "step": 30820 }, { "grad_norm": 0.09076650440692902, "learning_rate": 8.284300406612044e-05, "loss": 0.0049, "step": 30830 }, { "grad_norm": 0.15072473883628845, "learning_rate": 8.283053490158978e-05, "loss": 0.0069, "step": 30840 }, { "grad_norm": 0.12701785564422607, "learning_rate": 8.28180621467612e-05, "loss": 0.0051, "step": 30850 }, { "grad_norm": 0.07158643007278442, "learning_rate": 8.280558580299867e-05, "loss": 0.0051, "step": 30860 }, { "grad_norm": 0.1427103877067566, "learning_rate": 8.279310587166661e-05, "loss": 0.006, "step": 30870 }, { "grad_norm": 0.08842150121927261, "learning_rate": 8.278062235412978e-05, "loss": 0.0061, "step": 30880 }, { "grad_norm": 0.09296612441539764, "learning_rate": 8.276813525175339e-05, "loss": 0.0049, "step": 30890 }, { "grad_norm": 0.06533409655094147, "learning_rate": 8.2755644565903e-05, "loss": 0.006, "step": 30900 }, { "grad_norm": 0.08412248641252518, "learning_rate": 8.274315029794454e-05, "loss": 0.007, "step": 30910 }, { "grad_norm": 0.10624503344297409, "learning_rate": 8.273065244924443e-05, "loss": 0.0068, "step": 30920 }, { "grad_norm": 0.09419957548379898, "learning_rate": 8.271815102116936e-05, "loss": 0.0056, "step": 30930 }, { "grad_norm": 0.0738157108426094, "learning_rate": 8.270564601508648e-05, "loss": 0.0059, "step": 30940 }, { "grad_norm": 0.09784716367721558, "learning_rate": 8.269313743236333e-05, "loss": 0.0052, "step": 30950 }, { "grad_norm": 0.09481006860733032, "learning_rate": 8.268062527436783e-05, "loss": 0.005, "step": 30960 }, { "grad_norm": 0.06652582436800003, "learning_rate": 8.266810954246828e-05, "loss": 0.0056, "step": 30970 }, { "grad_norm": 0.08741310238838196, "learning_rate": 8.265559023803338e-05, "loss": 0.0053, "step": 30980 }, { "grad_norm": 0.1270357370376587, "learning_rate": 8.264306736243223e-05, "loss": 0.0045, "step": 30990 }, { "grad_norm": 0.11951111257076263, "learning_rate": 8.263054091703432e-05, "loss": 0.005, "step": 31000 }, { "grad_norm": 0.10675991326570511, "learning_rate": 8.26180109032095e-05, "loss": 0.0049, "step": 31010 }, { "grad_norm": 0.10039909183979034, "learning_rate": 8.260547732232807e-05, "loss": 0.0051, "step": 31020 }, { "grad_norm": 0.1389952301979065, "learning_rate": 8.259294017576063e-05, "loss": 0.0057, "step": 31030 }, { "grad_norm": 0.09163794666528702, "learning_rate": 8.258039946487828e-05, "loss": 0.0061, "step": 31040 }, { "grad_norm": 0.09392856806516647, "learning_rate": 8.256785519105241e-05, "loss": 0.0051, "step": 31050 }, { "grad_norm": 0.08757194876670837, "learning_rate": 8.255530735565488e-05, "loss": 0.0049, "step": 31060 }, { "grad_norm": 0.08330706506967545, "learning_rate": 8.254275596005787e-05, "loss": 0.0047, "step": 31070 }, { "grad_norm": 0.0888349711894989, "learning_rate": 8.2530201005634e-05, "loss": 0.0055, "step": 31080 }, { "grad_norm": 0.06173813343048096, "learning_rate": 8.251764249375626e-05, "loss": 0.0047, "step": 31090 }, { "grad_norm": 0.0919293463230133, "learning_rate": 8.250508042579803e-05, "loss": 0.0056, "step": 31100 }, { "grad_norm": 0.10426491498947144, "learning_rate": 8.249251480313307e-05, "loss": 0.0055, "step": 31110 }, { "grad_norm": 0.08341965079307556, "learning_rate": 8.247994562713555e-05, "loss": 0.0064, "step": 31120 }, { "grad_norm": 0.09159167855978012, "learning_rate": 8.246737289917998e-05, "loss": 0.0053, "step": 31130 }, { "grad_norm": 0.08143151551485062, "learning_rate": 8.245479662064135e-05, "loss": 0.0057, "step": 31140 }, { "grad_norm": 0.0805826336145401, "learning_rate": 8.244221679289496e-05, "loss": 0.006, "step": 31150 }, { "grad_norm": 0.09991243481636047, "learning_rate": 8.242963341731652e-05, "loss": 0.0052, "step": 31160 }, { "grad_norm": 0.08321838825941086, "learning_rate": 8.24170464952821e-05, "loss": 0.0067, "step": 31170 }, { "grad_norm": 0.09227551519870758, "learning_rate": 8.240445602816824e-05, "loss": 0.0056, "step": 31180 }, { "grad_norm": 0.1021609976887703, "learning_rate": 8.239186201735179e-05, "loss": 0.0055, "step": 31190 }, { "grad_norm": 0.09902515262365341, "learning_rate": 8.237926446420998e-05, "loss": 0.0059, "step": 31200 }, { "grad_norm": 0.0865212008357048, "learning_rate": 8.236666337012052e-05, "loss": 0.005, "step": 31210 }, { "grad_norm": 0.10658304393291473, "learning_rate": 8.23540587364614e-05, "loss": 0.0066, "step": 31220 }, { "grad_norm": 0.08041328191757202, "learning_rate": 8.234145056461107e-05, "loss": 0.0068, "step": 31230 }, { "grad_norm": 0.08447081595659256, "learning_rate": 8.232883885594832e-05, "loss": 0.0043, "step": 31240 }, { "grad_norm": 0.11240639537572861, "learning_rate": 8.231622361185236e-05, "loss": 0.0062, "step": 31250 }, { "grad_norm": 0.1196788102388382, "learning_rate": 8.230360483370278e-05, "loss": 0.0058, "step": 31260 }, { "grad_norm": 0.10604944825172424, "learning_rate": 8.229098252287953e-05, "loss": 0.0058, "step": 31270 }, { "grad_norm": 0.09977346658706665, "learning_rate": 8.2278356680763e-05, "loss": 0.006, "step": 31280 }, { "grad_norm": 0.10047324746847153, "learning_rate": 8.22657273087339e-05, "loss": 0.0056, "step": 31290 }, { "grad_norm": 0.10267072916030884, "learning_rate": 8.225309440817336e-05, "loss": 0.0057, "step": 31300 }, { "grad_norm": 0.11306992173194885, "learning_rate": 8.224045798046293e-05, "loss": 0.006, "step": 31310 }, { "grad_norm": 0.1427885740995407, "learning_rate": 8.22278180269845e-05, "loss": 0.0058, "step": 31320 }, { "grad_norm": 0.07858502864837646, "learning_rate": 8.221517454912031e-05, "loss": 0.0051, "step": 31330 }, { "grad_norm": 0.1292828768491745, "learning_rate": 8.220252754825308e-05, "loss": 0.0066, "step": 31340 }, { "grad_norm": 0.08356514573097229, "learning_rate": 8.218987702576586e-05, "loss": 0.0054, "step": 31350 }, { "grad_norm": 0.06659851968288422, "learning_rate": 8.217722298304207e-05, "loss": 0.0072, "step": 31360 }, { "grad_norm": 0.08057847619056702, "learning_rate": 8.216456542146557e-05, "loss": 0.005, "step": 31370 }, { "grad_norm": 0.08351925760507584, "learning_rate": 8.215190434242055e-05, "loss": 0.0068, "step": 31380 }, { "grad_norm": 0.13769645988941193, "learning_rate": 8.213923974729161e-05, "loss": 0.0058, "step": 31390 }, { "grad_norm": 0.11544682085514069, "learning_rate": 8.212657163746373e-05, "loss": 0.0063, "step": 31400 }, { "grad_norm": 0.08674126863479614, "learning_rate": 8.211390001432227e-05, "loss": 0.0053, "step": 31410 }, { "grad_norm": 0.11023790389299393, "learning_rate": 8.210122487925297e-05, "loss": 0.005, "step": 31420 }, { "grad_norm": 0.08292404562234879, "learning_rate": 8.208854623364202e-05, "loss": 0.0047, "step": 31430 }, { "grad_norm": 0.0711560845375061, "learning_rate": 8.207586407887589e-05, "loss": 0.006, "step": 31440 }, { "grad_norm": 0.10772416740655899, "learning_rate": 8.206317841634148e-05, "loss": 0.0048, "step": 31450 }, { "grad_norm": 0.0939640924334526, "learning_rate": 8.205048924742609e-05, "loss": 0.0056, "step": 31460 }, { "grad_norm": 0.08229053020477295, "learning_rate": 8.203779657351738e-05, "loss": 0.0049, "step": 31470 }, { "grad_norm": 0.11326109617948532, "learning_rate": 8.20251003960034e-05, "loss": 0.0057, "step": 31480 }, { "grad_norm": 0.0817723423242569, "learning_rate": 8.201240071627258e-05, "loss": 0.0058, "step": 31490 }, { "grad_norm": 0.09315987676382065, "learning_rate": 8.199969753571377e-05, "loss": 0.0067, "step": 31500 }, { "grad_norm": 0.0974363163113594, "learning_rate": 8.198699085571615e-05, "loss": 0.0063, "step": 31510 }, { "grad_norm": 0.10656995326280594, "learning_rate": 8.197428067766928e-05, "loss": 0.0067, "step": 31520 }, { "grad_norm": 0.07493121922016144, "learning_rate": 8.196156700296316e-05, "loss": 0.0053, "step": 31530 }, { "grad_norm": 0.10444459319114685, "learning_rate": 8.194884983298814e-05, "loss": 0.005, "step": 31540 }, { "grad_norm": 0.09736833721399307, "learning_rate": 8.193612916913491e-05, "loss": 0.0054, "step": 31550 }, { "grad_norm": 0.12677054107189178, "learning_rate": 8.192340501279463e-05, "loss": 0.005, "step": 31560 }, { "grad_norm": 0.13077394664287567, "learning_rate": 8.191067736535876e-05, "loss": 0.0065, "step": 31570 }, { "grad_norm": 0.10504917055368423, "learning_rate": 8.18979462282192e-05, "loss": 0.0058, "step": 31580 }, { "grad_norm": 0.10521326214075089, "learning_rate": 8.188521160276819e-05, "loss": 0.0056, "step": 31590 }, { "grad_norm": 0.08223911374807358, "learning_rate": 8.187247349039837e-05, "loss": 0.0046, "step": 31600 }, { "grad_norm": 0.07185538858175278, "learning_rate": 8.185973189250278e-05, "loss": 0.0055, "step": 31610 }, { "grad_norm": 0.10139212012290955, "learning_rate": 8.184698681047482e-05, "loss": 0.0056, "step": 31620 }, { "grad_norm": 0.0886465311050415, "learning_rate": 8.183423824570827e-05, "loss": 0.0063, "step": 31630 }, { "grad_norm": 0.11529909819364548, "learning_rate": 8.182148619959725e-05, "loss": 0.0051, "step": 31640 }, { "grad_norm": 0.10046155005693436, "learning_rate": 8.180873067353636e-05, "loss": 0.0053, "step": 31650 }, { "grad_norm": 0.0807328000664711, "learning_rate": 8.179597166892052e-05, "loss": 0.0048, "step": 31660 }, { "grad_norm": 0.09970559179782867, "learning_rate": 8.178320918714501e-05, "loss": 0.0045, "step": 31670 }, { "grad_norm": 0.0976487323641777, "learning_rate": 8.177044322960554e-05, "loss": 0.006, "step": 31680 }, { "grad_norm": 0.0838102400302887, "learning_rate": 8.175767379769816e-05, "loss": 0.0055, "step": 31690 }, { "grad_norm": 0.07190094143152237, "learning_rate": 8.174490089281932e-05, "loss": 0.0064, "step": 31700 }, { "grad_norm": 0.11861678957939148, "learning_rate": 8.173212451636584e-05, "loss": 0.0056, "step": 31710 }, { "grad_norm": 0.08666251599788666, "learning_rate": 8.171934466973493e-05, "loss": 0.0052, "step": 31720 }, { "grad_norm": 0.10141096264123917, "learning_rate": 8.170656135432418e-05, "loss": 0.0052, "step": 31730 }, { "grad_norm": 0.07993163913488388, "learning_rate": 8.169377457153155e-05, "loss": 0.0048, "step": 31740 }, { "grad_norm": 0.0652061477303505, "learning_rate": 8.168098432275539e-05, "loss": 0.0049, "step": 31750 }, { "grad_norm": 0.0835782065987587, "learning_rate": 8.166819060939442e-05, "loss": 0.0049, "step": 31760 }, { "grad_norm": 0.1053629070520401, "learning_rate": 8.165539343284772e-05, "loss": 0.0051, "step": 31770 }, { "grad_norm": 0.06873492151498795, "learning_rate": 8.16425927945148e-05, "loss": 0.0051, "step": 31780 }, { "grad_norm": 0.075914166867733, "learning_rate": 8.162978869579551e-05, "loss": 0.0055, "step": 31790 }, { "grad_norm": 0.10206092894077301, "learning_rate": 8.161698113809007e-05, "loss": 0.0052, "step": 31800 }, { "grad_norm": 0.12067896127700806, "learning_rate": 8.160417012279911e-05, "loss": 0.0052, "step": 31810 }, { "grad_norm": 0.10553796589374542, "learning_rate": 8.159135565132363e-05, "loss": 0.005, "step": 31820 }, { "grad_norm": 0.08267194032669067, "learning_rate": 8.157853772506498e-05, "loss": 0.0064, "step": 31830 }, { "grad_norm": 0.0769134908914566, "learning_rate": 8.156571634542494e-05, "loss": 0.0055, "step": 31840 }, { "grad_norm": 0.12121061980724335, "learning_rate": 8.15528915138056e-05, "loss": 0.0056, "step": 31850 }, { "grad_norm": 0.08604548126459122, "learning_rate": 8.154006323160949e-05, "loss": 0.0059, "step": 31860 }, { "grad_norm": 0.0696331188082695, "learning_rate": 8.152723150023949e-05, "loss": 0.0055, "step": 31870 }, { "grad_norm": 0.11785139888525009, "learning_rate": 8.151439632109886e-05, "loss": 0.0059, "step": 31880 }, { "grad_norm": 0.07167310267686844, "learning_rate": 8.150155769559122e-05, "loss": 0.0056, "step": 31890 }, { "grad_norm": 0.1113371029496193, "learning_rate": 8.148871562512058e-05, "loss": 0.0057, "step": 31900 }, { "grad_norm": 0.10897180438041687, "learning_rate": 8.147587011109136e-05, "loss": 0.0062, "step": 31910 }, { "grad_norm": 0.11235637962818146, "learning_rate": 8.14630211549083e-05, "loss": 0.0054, "step": 31920 }, { "grad_norm": 0.08398091793060303, "learning_rate": 8.145016875797655e-05, "loss": 0.0061, "step": 31930 }, { "grad_norm": 0.08499784767627716, "learning_rate": 8.143731292170164e-05, "loss": 0.0053, "step": 31940 }, { "grad_norm": 0.1091485247015953, "learning_rate": 8.142445364748944e-05, "loss": 0.0058, "step": 31950 }, { "grad_norm": 0.09344194829463959, "learning_rate": 8.141159093674624e-05, "loss": 0.0069, "step": 31960 }, { "grad_norm": 0.07022622227668762, "learning_rate": 8.139872479087869e-05, "loss": 0.0055, "step": 31970 }, { "grad_norm": 0.09343837201595306, "learning_rate": 8.13858552112938e-05, "loss": 0.0051, "step": 31980 }, { "grad_norm": 0.10780066251754761, "learning_rate": 8.137298219939895e-05, "loss": 0.0064, "step": 31990 }, { "grad_norm": 0.14006218314170837, "learning_rate": 8.136010575660196e-05, "loss": 0.0057, "step": 32000 }, { "grad_norm": 0.11706402897834778, "learning_rate": 8.134722588431095e-05, "loss": 0.0061, "step": 32010 }, { "grad_norm": 0.09378752112388611, "learning_rate": 8.133434258393444e-05, "loss": 0.0058, "step": 32020 }, { "grad_norm": 0.07981916517019272, "learning_rate": 8.132145585688134e-05, "loss": 0.0059, "step": 32030 }, { "grad_norm": 0.11570147424936295, "learning_rate": 8.130856570456093e-05, "loss": 0.0059, "step": 32040 }, { "grad_norm": 0.09242437779903412, "learning_rate": 8.129567212838283e-05, "loss": 0.0056, "step": 32050 }, { "grad_norm": 0.10704270750284195, "learning_rate": 8.128277512975708e-05, "loss": 0.0058, "step": 32060 }, { "grad_norm": 0.07415404915809631, "learning_rate": 8.126987471009408e-05, "loss": 0.0054, "step": 32070 }, { "grad_norm": 0.12409878522157669, "learning_rate": 8.125697087080459e-05, "loss": 0.0074, "step": 32080 }, { "grad_norm": 0.10762187093496323, "learning_rate": 8.124406361329976e-05, "loss": 0.006, "step": 32090 }, { "grad_norm": 0.1048855260014534, "learning_rate": 8.123115293899111e-05, "loss": 0.0052, "step": 32100 }, { "grad_norm": 0.08724203705787659, "learning_rate": 8.121823884929055e-05, "loss": 0.005, "step": 32110 }, { "grad_norm": 0.12592226266860962, "learning_rate": 8.12053213456103e-05, "loss": 0.0057, "step": 32120 }, { "grad_norm": 0.10993462800979614, "learning_rate": 8.119240042936303e-05, "loss": 0.0052, "step": 32130 }, { "grad_norm": 0.07904543727636337, "learning_rate": 8.117947610196175e-05, "loss": 0.0056, "step": 32140 }, { "grad_norm": 0.07981979101896286, "learning_rate": 8.116654836481982e-05, "loss": 0.0056, "step": 32150 }, { "grad_norm": 0.1075981929898262, "learning_rate": 8.115361721935106e-05, "loss": 0.0052, "step": 32160 }, { "grad_norm": 0.08726360648870468, "learning_rate": 8.114068266696953e-05, "loss": 0.0049, "step": 32170 }, { "grad_norm": 0.08225808292627335, "learning_rate": 8.112774470908978e-05, "loss": 0.0058, "step": 32180 }, { "grad_norm": 0.08133094012737274, "learning_rate": 8.111480334712665e-05, "loss": 0.0068, "step": 32190 }, { "grad_norm": 0.11642366647720337, "learning_rate": 8.110185858249542e-05, "loss": 0.0057, "step": 32200 }, { "grad_norm": 0.09619107842445374, "learning_rate": 8.108891041661168e-05, "loss": 0.0054, "step": 32210 }, { "grad_norm": 0.08391758054494858, "learning_rate": 8.107595885089146e-05, "loss": 0.0059, "step": 32220 }, { "grad_norm": 0.08471697568893433, "learning_rate": 8.106300388675108e-05, "loss": 0.0053, "step": 32230 }, { "grad_norm": 0.07648996263742447, "learning_rate": 8.105004552560731e-05, "loss": 0.0056, "step": 32240 }, { "grad_norm": 0.09224235266447067, "learning_rate": 8.103708376887724e-05, "loss": 0.0053, "step": 32250 }, { "grad_norm": 0.09119857102632523, "learning_rate": 8.102411861797836e-05, "loss": 0.0051, "step": 32260 }, { "grad_norm": 0.10301055014133453, "learning_rate": 8.101115007432851e-05, "loss": 0.0056, "step": 32270 }, { "grad_norm": 0.10869032144546509, "learning_rate": 8.09981781393459e-05, "loss": 0.0056, "step": 32280 }, { "grad_norm": 0.07786466181278229, "learning_rate": 8.098520281444915e-05, "loss": 0.0053, "step": 32290 }, { "grad_norm": 0.07805036008358002, "learning_rate": 8.09722241010572e-05, "loss": 0.0064, "step": 32300 }, { "grad_norm": 0.07172340154647827, "learning_rate": 8.095924200058939e-05, "loss": 0.0051, "step": 32310 }, { "grad_norm": 0.08278172463178635, "learning_rate": 8.094625651446541e-05, "loss": 0.0051, "step": 32320 }, { "grad_norm": 0.12496905773878098, "learning_rate": 8.093326764410536e-05, "loss": 0.0057, "step": 32330 }, { "grad_norm": 0.0927143543958664, "learning_rate": 8.092027539092966e-05, "loss": 0.0067, "step": 32340 }, { "grad_norm": 0.08108580857515335, "learning_rate": 8.090727975635913e-05, "loss": 0.006, "step": 32350 }, { "grad_norm": 0.08911644667387009, "learning_rate": 8.089428074181497e-05, "loss": 0.0056, "step": 32360 }, { "grad_norm": 0.09357605874538422, "learning_rate": 8.088127834871871e-05, "loss": 0.0051, "step": 32370 }, { "grad_norm": 0.08911113440990448, "learning_rate": 8.086827257849226e-05, "loss": 0.0057, "step": 32380 }, { "grad_norm": 0.11638546735048294, "learning_rate": 8.085526343255795e-05, "loss": 0.0053, "step": 32390 }, { "grad_norm": 0.11041335761547089, "learning_rate": 8.084225091233842e-05, "loss": 0.0067, "step": 32400 }, { "grad_norm": 0.0855010524392128, "learning_rate": 8.082923501925668e-05, "loss": 0.0057, "step": 32410 }, { "grad_norm": 0.10945077240467072, "learning_rate": 8.081621575473617e-05, "loss": 0.0064, "step": 32420 }, { "grad_norm": 0.10483470559120178, "learning_rate": 8.080319312020064e-05, "loss": 0.0059, "step": 32430 }, { "grad_norm": 0.08180274069309235, "learning_rate": 8.079016711707421e-05, "loss": 0.0057, "step": 32440 }, { "grad_norm": 0.09013842791318893, "learning_rate": 8.077713774678139e-05, "loss": 0.0047, "step": 32450 }, { "grad_norm": 0.08951867371797562, "learning_rate": 8.076410501074707e-05, "loss": 0.0044, "step": 32460 }, { "grad_norm": 0.07890770584344864, "learning_rate": 8.075106891039647e-05, "loss": 0.0046, "step": 32470 }, { "grad_norm": 0.09624994546175003, "learning_rate": 8.073802944715523e-05, "loss": 0.0049, "step": 32480 }, { "grad_norm": 0.08491140604019165, "learning_rate": 8.072498662244929e-05, "loss": 0.006, "step": 32490 }, { "grad_norm": 0.06579598784446716, "learning_rate": 8.0711940437705e-05, "loss": 0.0062, "step": 32500 }, { "grad_norm": 0.0734674334526062, "learning_rate": 8.06988908943491e-05, "loss": 0.0057, "step": 32510 }, { "grad_norm": 0.10613779723644257, "learning_rate": 8.068583799380863e-05, "loss": 0.0052, "step": 32520 }, { "grad_norm": 0.11243196576833725, "learning_rate": 8.067278173751104e-05, "loss": 0.0053, "step": 32530 }, { "grad_norm": 0.08869917690753937, "learning_rate": 8.065972212688417e-05, "loss": 0.0055, "step": 32540 }, { "grad_norm": 0.10461145639419556, "learning_rate": 8.064665916335618e-05, "loss": 0.0053, "step": 32550 }, { "grad_norm": 0.09441179037094116, "learning_rate": 8.063359284835564e-05, "loss": 0.0042, "step": 32560 }, { "grad_norm": 0.10402523726224899, "learning_rate": 8.062052318331142e-05, "loss": 0.0057, "step": 32570 }, { "grad_norm": 0.09279567748308182, "learning_rate": 8.060745016965283e-05, "loss": 0.005, "step": 32580 }, { "grad_norm": 0.07654614001512527, "learning_rate": 8.059437380880952e-05, "loss": 0.0048, "step": 32590 }, { "grad_norm": 0.09217886626720428, "learning_rate": 8.058129410221146e-05, "loss": 0.0061, "step": 32600 }, { "grad_norm": 0.06840866804122925, "learning_rate": 8.056821105128908e-05, "loss": 0.006, "step": 32610 }, { "grad_norm": 0.11095651239156723, "learning_rate": 8.05551246574731e-05, "loss": 0.0058, "step": 32620 }, { "grad_norm": 0.10324158519506454, "learning_rate": 8.05420349221946e-05, "loss": 0.0063, "step": 32630 }, { "grad_norm": 0.0777294933795929, "learning_rate": 8.05289418468851e-05, "loss": 0.0048, "step": 32640 }, { "grad_norm": 0.10762995481491089, "learning_rate": 8.051584543297642e-05, "loss": 0.006, "step": 32650 }, { "grad_norm": 0.09799942374229431, "learning_rate": 8.050274568190074e-05, "loss": 0.0062, "step": 32660 }, { "grad_norm": 0.08702124655246735, "learning_rate": 8.048964259509067e-05, "loss": 0.0049, "step": 32670 }, { "grad_norm": 0.07766599953174591, "learning_rate": 8.047653617397914e-05, "loss": 0.0053, "step": 32680 }, { "grad_norm": 0.10062870383262634, "learning_rate": 8.046342641999941e-05, "loss": 0.0051, "step": 32690 }, { "grad_norm": 0.08460000902414322, "learning_rate": 8.045031333458517e-05, "loss": 0.0048, "step": 32700 }, { "grad_norm": 0.1193973645567894, "learning_rate": 8.043719691917047e-05, "loss": 0.0059, "step": 32710 }, { "grad_norm": 0.10495792329311371, "learning_rate": 8.042407717518966e-05, "loss": 0.0054, "step": 32720 }, { "grad_norm": 0.11512342840433121, "learning_rate": 8.041095410407751e-05, "loss": 0.0059, "step": 32730 }, { "grad_norm": 0.08799021691083908, "learning_rate": 8.039782770726913e-05, "loss": 0.006, "step": 32740 }, { "grad_norm": 0.11190519481897354, "learning_rate": 8.038469798620004e-05, "loss": 0.0056, "step": 32750 }, { "grad_norm": 0.06397456675767899, "learning_rate": 8.037156494230604e-05, "loss": 0.0044, "step": 32760 }, { "grad_norm": 0.07972460240125656, "learning_rate": 8.035842857702338e-05, "loss": 0.0047, "step": 32770 }, { "grad_norm": 0.07716546952724457, "learning_rate": 8.03452888917886e-05, "loss": 0.0039, "step": 32780 }, { "grad_norm": 0.0741053894162178, "learning_rate": 8.033214588803866e-05, "loss": 0.0044, "step": 32790 }, { "grad_norm": 0.07593229413032532, "learning_rate": 8.031899956721083e-05, "loss": 0.0048, "step": 32800 }, { "grad_norm": 0.10173163563013077, "learning_rate": 8.030584993074282e-05, "loss": 0.0048, "step": 32810 }, { "grad_norm": 0.08689805865287781, "learning_rate": 8.02926969800726e-05, "loss": 0.0066, "step": 32820 }, { "grad_norm": 0.13672296702861786, "learning_rate": 8.027954071663859e-05, "loss": 0.0047, "step": 32830 }, { "grad_norm": 0.09783239662647247, "learning_rate": 8.026638114187954e-05, "loss": 0.005, "step": 32840 }, { "grad_norm": 0.06996440142393112, "learning_rate": 8.025321825723456e-05, "loss": 0.0055, "step": 32850 }, { "grad_norm": 0.0984201729297638, "learning_rate": 8.02400520641431e-05, "loss": 0.0058, "step": 32860 }, { "grad_norm": 0.1017015278339386, "learning_rate": 8.022688256404501e-05, "loss": 0.0046, "step": 32870 }, { "grad_norm": 0.10950662195682526, "learning_rate": 8.02137097583805e-05, "loss": 0.0057, "step": 32880 }, { "grad_norm": 0.08126933872699738, "learning_rate": 8.02005336485901e-05, "loss": 0.0062, "step": 32890 }, { "grad_norm": 0.10147615522146225, "learning_rate": 8.018735423611476e-05, "loss": 0.0056, "step": 32900 }, { "grad_norm": 0.113735631108284, "learning_rate": 8.017417152239574e-05, "loss": 0.0051, "step": 32910 }, { "grad_norm": 0.12537160515785217, "learning_rate": 8.01609855088747e-05, "loss": 0.0057, "step": 32920 }, { "grad_norm": 0.08231750130653381, "learning_rate": 8.014779619699362e-05, "loss": 0.0044, "step": 32930 }, { "grad_norm": 0.069846972823143, "learning_rate": 8.013460358819489e-05, "loss": 0.005, "step": 32940 }, { "grad_norm": 0.07761014252901077, "learning_rate": 8.01214076839212e-05, "loss": 0.0056, "step": 32950 }, { "grad_norm": 0.11930274963378906, "learning_rate": 8.010820848561565e-05, "loss": 0.0051, "step": 32960 }, { "grad_norm": 0.08494653552770615, "learning_rate": 8.009500599472171e-05, "loss": 0.0053, "step": 32970 }, { "grad_norm": 0.09235383570194244, "learning_rate": 8.008180021268314e-05, "loss": 0.0048, "step": 32980 }, { "grad_norm": 0.08010843396186829, "learning_rate": 8.006859114094414e-05, "loss": 0.0054, "step": 32990 }, { "grad_norm": 0.08126207441091537, "learning_rate": 8.005537878094921e-05, "loss": 0.0062, "step": 33000 }, { "grad_norm": 0.06801445037126541, "learning_rate": 8.004216313414323e-05, "loss": 0.0044, "step": 33010 }, { "grad_norm": 0.07911136746406555, "learning_rate": 8.002894420197149e-05, "loss": 0.0054, "step": 33020 }, { "grad_norm": 0.09876085817813873, "learning_rate": 8.001572198587954e-05, "loss": 0.0055, "step": 33030 }, { "grad_norm": 0.09128837287425995, "learning_rate": 8.000249648731338e-05, "loss": 0.0048, "step": 33040 }, { "grad_norm": 0.12406846880912781, "learning_rate": 7.998926770771928e-05, "loss": 0.005, "step": 33050 }, { "grad_norm": 0.11624976247549057, "learning_rate": 7.997603564854397e-05, "loss": 0.0057, "step": 33060 }, { "grad_norm": 0.1268363744020462, "learning_rate": 7.996280031123448e-05, "loss": 0.0055, "step": 33070 }, { "grad_norm": 0.10332013666629791, "learning_rate": 7.994956169723818e-05, "loss": 0.005, "step": 33080 }, { "grad_norm": 0.13112927973270416, "learning_rate": 7.993631980800285e-05, "loss": 0.005, "step": 33090 }, { "grad_norm": 0.17249251902103424, "learning_rate": 7.992307464497659e-05, "loss": 0.006, "step": 33100 }, { "grad_norm": 0.09942138195037842, "learning_rate": 7.990982620960787e-05, "loss": 0.0066, "step": 33110 }, { "grad_norm": 0.08512724936008453, "learning_rate": 7.989657450334554e-05, "loss": 0.0058, "step": 33120 }, { "grad_norm": 0.09701719135046005, "learning_rate": 7.988331952763877e-05, "loss": 0.007, "step": 33130 }, { "grad_norm": 0.11406667530536652, "learning_rate": 7.987006128393709e-05, "loss": 0.007, "step": 33140 }, { "grad_norm": 0.11177383363246918, "learning_rate": 7.985679977369043e-05, "loss": 0.0069, "step": 33150 }, { "grad_norm": 0.11733313649892807, "learning_rate": 7.984353499834902e-05, "loss": 0.0068, "step": 33160 }, { "grad_norm": 0.11775416135787964, "learning_rate": 7.983026695936351e-05, "loss": 0.0066, "step": 33170 }, { "grad_norm": 0.12714774906635284, "learning_rate": 7.981699565818486e-05, "loss": 0.0059, "step": 33180 }, { "grad_norm": 0.07789045572280884, "learning_rate": 7.980372109626437e-05, "loss": 0.0052, "step": 33190 }, { "grad_norm": 0.12992742657661438, "learning_rate": 7.979044327505375e-05, "loss": 0.0066, "step": 33200 }, { "grad_norm": 0.07233305275440216, "learning_rate": 7.977716219600506e-05, "loss": 0.0057, "step": 33210 }, { "grad_norm": 0.0742095336318016, "learning_rate": 7.97638778605707e-05, "loss": 0.0055, "step": 33220 }, { "grad_norm": 0.09641699492931366, "learning_rate": 7.975059027020338e-05, "loss": 0.0055, "step": 33230 }, { "grad_norm": 0.10218707472085953, "learning_rate": 7.973729942635623e-05, "loss": 0.0055, "step": 33240 }, { "grad_norm": 0.08579685539007187, "learning_rate": 7.972400533048273e-05, "loss": 0.0048, "step": 33250 }, { "grad_norm": 0.10144844651222229, "learning_rate": 7.97107079840367e-05, "loss": 0.0057, "step": 33260 }, { "grad_norm": 0.08264244347810745, "learning_rate": 7.969740738847231e-05, "loss": 0.0051, "step": 33270 }, { "grad_norm": 0.08131501823663712, "learning_rate": 7.968410354524411e-05, "loss": 0.0056, "step": 33280 }, { "grad_norm": 0.10856900364160538, "learning_rate": 7.967079645580697e-05, "loss": 0.0055, "step": 33290 }, { "grad_norm": 0.08846329152584076, "learning_rate": 7.965748612161612e-05, "loss": 0.0054, "step": 33300 }, { "grad_norm": 0.1063232421875, "learning_rate": 7.96441725441272e-05, "loss": 0.0052, "step": 33310 }, { "grad_norm": 0.11413400620222092, "learning_rate": 7.963085572479614e-05, "loss": 0.0058, "step": 33320 }, { "grad_norm": 0.11171624064445496, "learning_rate": 7.961753566507924e-05, "loss": 0.006, "step": 33330 }, { "grad_norm": 0.06526708602905273, "learning_rate": 7.960421236643316e-05, "loss": 0.0045, "step": 33340 }, { "grad_norm": 0.09978209435939789, "learning_rate": 7.959088583031496e-05, "loss": 0.0049, "step": 33350 }, { "grad_norm": 0.10266142338514328, "learning_rate": 7.957755605818194e-05, "loss": 0.0051, "step": 33360 }, { "grad_norm": 0.10707021504640579, "learning_rate": 7.956422305149185e-05, "loss": 0.0051, "step": 33370 }, { "grad_norm": 0.08297715336084366, "learning_rate": 7.95508868117028e-05, "loss": 0.0049, "step": 33380 }, { "grad_norm": 0.077968530356884, "learning_rate": 7.953754734027318e-05, "loss": 0.0044, "step": 33390 }, { "grad_norm": 0.06895224750041962, "learning_rate": 7.952420463866182e-05, "loss": 0.0048, "step": 33400 }, { "grad_norm": 0.0927564725279808, "learning_rate": 7.951085870832782e-05, "loss": 0.0051, "step": 33410 }, { "grad_norm": 0.0803578794002533, "learning_rate": 7.949750955073067e-05, "loss": 0.0055, "step": 33420 }, { "grad_norm": 0.09690950065851212, "learning_rate": 7.948415716733022e-05, "loss": 0.0048, "step": 33430 }, { "grad_norm": 0.0750134214758873, "learning_rate": 7.947080155958669e-05, "loss": 0.005, "step": 33440 }, { "grad_norm": 0.07643909007310867, "learning_rate": 7.94574427289606e-05, "loss": 0.0051, "step": 33450 }, { "grad_norm": 0.07138728350400925, "learning_rate": 7.944408067691284e-05, "loss": 0.0048, "step": 33460 }, { "grad_norm": 0.09680534154176712, "learning_rate": 7.943071540490473e-05, "loss": 0.0055, "step": 33470 }, { "grad_norm": 0.0985930934548378, "learning_rate": 7.94173469143978e-05, "loss": 0.0044, "step": 33480 }, { "grad_norm": 0.09563606232404709, "learning_rate": 7.940397520685406e-05, "loss": 0.0054, "step": 33490 }, { "grad_norm": 0.07804621011018753, "learning_rate": 7.939060028373577e-05, "loss": 0.0053, "step": 33500 }, { "grad_norm": 0.11788435280323029, "learning_rate": 7.937722214650565e-05, "loss": 0.0043, "step": 33510 }, { "grad_norm": 0.10362286865711212, "learning_rate": 7.936384079662666e-05, "loss": 0.0044, "step": 33520 }, { "grad_norm": 0.07242528349161148, "learning_rate": 7.93504562355622e-05, "loss": 0.0043, "step": 33530 }, { "grad_norm": 0.07850563526153564, "learning_rate": 7.933706846477599e-05, "loss": 0.0055, "step": 33540 }, { "grad_norm": 0.11009027063846588, "learning_rate": 7.932367748573206e-05, "loss": 0.0049, "step": 33550 }, { "grad_norm": 0.12608841061592102, "learning_rate": 7.931028329989485e-05, "loss": 0.0052, "step": 33560 }, { "grad_norm": 0.093488909304142, "learning_rate": 7.929688590872913e-05, "loss": 0.0054, "step": 33570 }, { "grad_norm": 0.1138579323887825, "learning_rate": 7.928348531370003e-05, "loss": 0.0052, "step": 33580 }, { "grad_norm": 0.09873701632022858, "learning_rate": 7.927008151627297e-05, "loss": 0.0064, "step": 33590 }, { "grad_norm": 0.1069081574678421, "learning_rate": 7.925667451791383e-05, "loss": 0.0054, "step": 33600 }, { "grad_norm": 0.09774434566497803, "learning_rate": 7.924326432008874e-05, "loss": 0.0055, "step": 33610 }, { "grad_norm": 0.08233524113893509, "learning_rate": 7.922985092426422e-05, "loss": 0.0051, "step": 33620 }, { "grad_norm": 0.0760449767112732, "learning_rate": 7.921643433190717e-05, "loss": 0.005, "step": 33630 }, { "grad_norm": 0.08362706750631332, "learning_rate": 7.920301454448478e-05, "loss": 0.0054, "step": 33640 }, { "grad_norm": 0.07627914845943451, "learning_rate": 7.918959156346461e-05, "loss": 0.0052, "step": 33650 }, { "grad_norm": 0.07849734276533127, "learning_rate": 7.91761653903146e-05, "loss": 0.0057, "step": 33660 }, { "grad_norm": 0.09742715954780579, "learning_rate": 7.916273602650302e-05, "loss": 0.0047, "step": 33670 }, { "grad_norm": 0.08701206743717194, "learning_rate": 7.914930347349847e-05, "loss": 0.0049, "step": 33680 }, { "grad_norm": 0.1071370393037796, "learning_rate": 7.913586773276992e-05, "loss": 0.0044, "step": 33690 }, { "grad_norm": 0.08208425343036652, "learning_rate": 7.912242880578667e-05, "loss": 0.005, "step": 33700 }, { "grad_norm": 0.14241035282611847, "learning_rate": 7.910898669401839e-05, "loss": 0.0052, "step": 33710 }, { "grad_norm": 0.11530792713165283, "learning_rate": 7.909554139893511e-05, "loss": 0.0061, "step": 33720 }, { "grad_norm": 0.1318555772304535, "learning_rate": 7.908209292200715e-05, "loss": 0.0058, "step": 33730 }, { "grad_norm": 0.07852119952440262, "learning_rate": 7.906864126470523e-05, "loss": 0.0049, "step": 33740 }, { "grad_norm": 0.10375451296567917, "learning_rate": 7.905518642850041e-05, "loss": 0.0051, "step": 33750 }, { "grad_norm": 0.08406810462474823, "learning_rate": 7.904172841486409e-05, "loss": 0.006, "step": 33760 }, { "grad_norm": 0.07751700282096863, "learning_rate": 7.902826722526801e-05, "loss": 0.005, "step": 33770 }, { "grad_norm": 0.07758137583732605, "learning_rate": 7.901480286118427e-05, "loss": 0.0048, "step": 33780 }, { "grad_norm": 0.07707015424966812, "learning_rate": 7.900133532408531e-05, "loss": 0.005, "step": 33790 }, { "grad_norm": 0.09886583685874939, "learning_rate": 7.898786461544395e-05, "loss": 0.0046, "step": 33800 }, { "grad_norm": 0.10419037938117981, "learning_rate": 7.897439073673325e-05, "loss": 0.0057, "step": 33810 }, { "grad_norm": 0.08908398449420929, "learning_rate": 7.896091368942677e-05, "loss": 0.0044, "step": 33820 }, { "grad_norm": 0.08506602048873901, "learning_rate": 7.894743347499832e-05, "loss": 0.0051, "step": 33830 }, { "grad_norm": 0.10615329444408417, "learning_rate": 7.893395009492203e-05, "loss": 0.005, "step": 33840 }, { "grad_norm": 0.11256647109985352, "learning_rate": 7.892046355067248e-05, "loss": 0.0047, "step": 33850 }, { "grad_norm": 0.10169462859630585, "learning_rate": 7.890697384372451e-05, "loss": 0.005, "step": 33860 }, { "grad_norm": 0.0963849127292633, "learning_rate": 7.889348097555336e-05, "loss": 0.0053, "step": 33870 }, { "grad_norm": 0.0843062549829483, "learning_rate": 7.887998494763455e-05, "loss": 0.0056, "step": 33880 }, { "grad_norm": 0.11260859668254852, "learning_rate": 7.886648576144404e-05, "loss": 0.005, "step": 33890 }, { "grad_norm": 0.07353870570659637, "learning_rate": 7.885298341845802e-05, "loss": 0.0054, "step": 33900 }, { "grad_norm": 0.09196636080741882, "learning_rate": 7.883947792015311e-05, "loss": 0.0053, "step": 33910 }, { "grad_norm": 0.08460047841072083, "learning_rate": 7.882596926800628e-05, "loss": 0.0054, "step": 33920 }, { "grad_norm": 0.09826676547527313, "learning_rate": 7.881245746349477e-05, "loss": 0.0052, "step": 33930 }, { "grad_norm": 0.1207214966416359, "learning_rate": 7.879894250809623e-05, "loss": 0.0049, "step": 33940 }, { "grad_norm": 0.07677648216485977, "learning_rate": 7.878542440328865e-05, "loss": 0.0061, "step": 33950 }, { "grad_norm": 0.08732382953166962, "learning_rate": 7.877190315055031e-05, "loss": 0.005, "step": 33960 }, { "grad_norm": 0.07382843643426895, "learning_rate": 7.875837875135991e-05, "loss": 0.0069, "step": 33970 }, { "grad_norm": 0.0849551185965538, "learning_rate": 7.874485120719646e-05, "loss": 0.0055, "step": 33980 }, { "grad_norm": 0.07383653521537781, "learning_rate": 7.873132051953928e-05, "loss": 0.0049, "step": 33990 }, { "grad_norm": 0.11196839064359665, "learning_rate": 7.87177866898681e-05, "loss": 0.0055, "step": 34000 }, { "grad_norm": 0.09014622867107391, "learning_rate": 7.870424971966294e-05, "loss": 0.005, "step": 34010 }, { "grad_norm": 0.08475399017333984, "learning_rate": 7.869070961040419e-05, "loss": 0.0044, "step": 34020 }, { "grad_norm": 0.07296386361122131, "learning_rate": 7.867716636357257e-05, "loss": 0.0057, "step": 34030 }, { "grad_norm": 0.09054713696241379, "learning_rate": 7.866361998064915e-05, "loss": 0.0049, "step": 34040 }, { "grad_norm": 0.08828567713499069, "learning_rate": 7.865007046311534e-05, "loss": 0.0061, "step": 34050 }, { "grad_norm": 0.08501854538917542, "learning_rate": 7.86365178124529e-05, "loss": 0.0054, "step": 34060 }, { "grad_norm": 0.06646312028169632, "learning_rate": 7.862296203014394e-05, "loss": 0.005, "step": 34070 }, { "grad_norm": 0.07494023442268372, "learning_rate": 7.860940311767088e-05, "loss": 0.0052, "step": 34080 }, { "grad_norm": 0.09098447114229202, "learning_rate": 7.85958410765165e-05, "loss": 0.0068, "step": 34090 }, { "grad_norm": 0.07727976888418198, "learning_rate": 7.858227590816394e-05, "loss": 0.0059, "step": 34100 }, { "grad_norm": 0.06810140609741211, "learning_rate": 7.856870761409664e-05, "loss": 0.005, "step": 34110 }, { "grad_norm": 0.08170007914304733, "learning_rate": 7.855513619579846e-05, "loss": 0.0049, "step": 34120 }, { "grad_norm": 0.08037027716636658, "learning_rate": 7.85415616547535e-05, "loss": 0.0041, "step": 34130 }, { "grad_norm": 0.10327374190092087, "learning_rate": 7.852798399244627e-05, "loss": 0.0049, "step": 34140 }, { "grad_norm": 0.11664839088916779, "learning_rate": 7.851440321036161e-05, "loss": 0.0053, "step": 34150 }, { "grad_norm": 0.10039547830820084, "learning_rate": 7.850081930998468e-05, "loss": 0.0055, "step": 34160 }, { "grad_norm": 0.09502936899662018, "learning_rate": 7.8487232292801e-05, "loss": 0.0052, "step": 34170 }, { "grad_norm": 0.09254265576601028, "learning_rate": 7.847364216029642e-05, "loss": 0.0053, "step": 34180 }, { "grad_norm": 0.0704173818230629, "learning_rate": 7.846004891395716e-05, "loss": 0.0053, "step": 34190 }, { "grad_norm": 0.08707914501428604, "learning_rate": 7.844645255526972e-05, "loss": 0.0049, "step": 34200 }, { "grad_norm": 0.10465798527002335, "learning_rate": 7.843285308572101e-05, "loss": 0.0059, "step": 34210 }, { "grad_norm": 0.10942985862493515, "learning_rate": 7.841925050679823e-05, "loss": 0.0053, "step": 34220 }, { "grad_norm": 0.0988374650478363, "learning_rate": 7.840564481998895e-05, "loss": 0.0051, "step": 34230 }, { "grad_norm": 0.0627787783741951, "learning_rate": 7.839203602678105e-05, "loss": 0.0044, "step": 34240 }, { "grad_norm": 0.09173806756734848, "learning_rate": 7.837842412866279e-05, "loss": 0.0043, "step": 34250 }, { "grad_norm": 0.08611553907394409, "learning_rate": 7.836480912712272e-05, "loss": 0.0051, "step": 34260 }, { "grad_norm": 0.08048788458108902, "learning_rate": 7.835119102364976e-05, "loss": 0.0052, "step": 34270 }, { "grad_norm": 0.09985944628715515, "learning_rate": 7.83375698197332e-05, "loss": 0.0045, "step": 34280 }, { "grad_norm": 0.12055724114179611, "learning_rate": 7.83239455168626e-05, "loss": 0.0044, "step": 34290 }, { "grad_norm": 0.09428654611110687, "learning_rate": 7.83103181165279e-05, "loss": 0.0057, "step": 34300 }, { "grad_norm": 0.07962041348218918, "learning_rate": 7.829668762021937e-05, "loss": 0.0059, "step": 34310 }, { "grad_norm": 0.06791242957115173, "learning_rate": 7.828305402942764e-05, "loss": 0.0045, "step": 34320 }, { "grad_norm": 0.11398741602897644, "learning_rate": 7.826941734564363e-05, "loss": 0.0056, "step": 34330 }, { "grad_norm": 0.07860609889030457, "learning_rate": 7.825577757035865e-05, "loss": 0.0046, "step": 34340 }, { "grad_norm": 0.1034916564822197, "learning_rate": 7.824213470506431e-05, "loss": 0.0048, "step": 34350 }, { "grad_norm": 0.08735482394695282, "learning_rate": 7.822848875125257e-05, "loss": 0.0053, "step": 34360 }, { "grad_norm": 0.09033478796482086, "learning_rate": 7.821483971041576e-05, "loss": 0.0053, "step": 34370 }, { "grad_norm": 0.08611531555652618, "learning_rate": 7.820118758404649e-05, "loss": 0.0052, "step": 34380 }, { "grad_norm": 0.08462691307067871, "learning_rate": 7.818753237363776e-05, "loss": 0.005, "step": 34390 }, { "grad_norm": 0.0807817354798317, "learning_rate": 7.817387408068286e-05, "loss": 0.006, "step": 34400 }, { "grad_norm": 0.09041306376457214, "learning_rate": 7.816021270667544e-05, "loss": 0.0052, "step": 34410 }, { "grad_norm": 0.08840849995613098, "learning_rate": 7.81465482531095e-05, "loss": 0.0051, "step": 34420 }, { "grad_norm": 0.09162286669015884, "learning_rate": 7.813288072147938e-05, "loss": 0.006, "step": 34430 }, { "grad_norm": 0.07741491496562958, "learning_rate": 7.811921011327972e-05, "loss": 0.0055, "step": 34440 }, { "grad_norm": 0.08624675869941711, "learning_rate": 7.810553643000549e-05, "loss": 0.0043, "step": 34450 }, { "grad_norm": 0.08413751423358917, "learning_rate": 7.809185967315206e-05, "loss": 0.0049, "step": 34460 }, { "grad_norm": 0.07473386824131012, "learning_rate": 7.80781798442151e-05, "loss": 0.0051, "step": 34470 }, { "grad_norm": 0.10417567938566208, "learning_rate": 7.806449694469061e-05, "loss": 0.0057, "step": 34480 }, { "grad_norm": 0.10141193866729736, "learning_rate": 7.805081097607492e-05, "loss": 0.005, "step": 34490 }, { "grad_norm": 0.0887540727853775, "learning_rate": 7.803712193986474e-05, "loss": 0.0061, "step": 34500 }, { "grad_norm": 0.07812866568565369, "learning_rate": 7.802342983755702e-05, "loss": 0.0057, "step": 34510 }, { "grad_norm": 0.08885367214679718, "learning_rate": 7.800973467064918e-05, "loss": 0.0058, "step": 34520 }, { "grad_norm": 0.10222206264734268, "learning_rate": 7.799603644063884e-05, "loss": 0.0063, "step": 34530 }, { "grad_norm": 0.08902285248041153, "learning_rate": 7.798233514902405e-05, "loss": 0.0046, "step": 34540 }, { "grad_norm": 0.07594326138496399, "learning_rate": 7.796863079730318e-05, "loss": 0.0046, "step": 34550 }, { "grad_norm": 0.14360038936138153, "learning_rate": 7.795492338697488e-05, "loss": 0.0051, "step": 34560 }, { "grad_norm": 0.10264497995376587, "learning_rate": 7.794121291953819e-05, "loss": 0.0067, "step": 34570 }, { "grad_norm": 0.11090137809515, "learning_rate": 7.792749939649246e-05, "loss": 0.0068, "step": 34580 }, { "grad_norm": 0.09629756212234497, "learning_rate": 7.79137828193374e-05, "loss": 0.0056, "step": 34590 }, { "grad_norm": 0.08969628810882568, "learning_rate": 7.790006318957301e-05, "loss": 0.0053, "step": 34600 }, { "grad_norm": 0.0839901864528656, "learning_rate": 7.788634050869965e-05, "loss": 0.0057, "step": 34610 }, { "grad_norm": 0.06514941155910492, "learning_rate": 7.787261477821803e-05, "loss": 0.005, "step": 34620 }, { "grad_norm": 0.07417032122612, "learning_rate": 7.785888599962916e-05, "loss": 0.0052, "step": 34630 }, { "grad_norm": 0.10330855846405029, "learning_rate": 7.784515417443439e-05, "loss": 0.0046, "step": 34640 }, { "grad_norm": 0.09951271861791611, "learning_rate": 7.783141930413545e-05, "loss": 0.0051, "step": 34650 }, { "grad_norm": 0.08558173477649689, "learning_rate": 7.78176813902343e-05, "loss": 0.0047, "step": 34660 }, { "grad_norm": 0.09763579815626144, "learning_rate": 7.780394043423336e-05, "loss": 0.0064, "step": 34670 }, { "grad_norm": 0.07827607542276382, "learning_rate": 7.77901964376353e-05, "loss": 0.0053, "step": 34680 }, { "grad_norm": 0.08593764156103134, "learning_rate": 7.777644940194316e-05, "loss": 0.0054, "step": 34690 }, { "grad_norm": 0.07697346061468124, "learning_rate": 7.776269932866023e-05, "loss": 0.0047, "step": 34700 }, { "grad_norm": 0.06318904459476471, "learning_rate": 7.774894621929026e-05, "loss": 0.0046, "step": 34710 }, { "grad_norm": 0.107724629342556, "learning_rate": 7.773519007533725e-05, "loss": 0.0051, "step": 34720 }, { "grad_norm": 0.07321009039878845, "learning_rate": 7.772143089830556e-05, "loss": 0.0055, "step": 34730 }, { "grad_norm": 0.07225440442562103, "learning_rate": 7.770766868969985e-05, "loss": 0.005, "step": 34740 }, { "grad_norm": 0.09977243095636368, "learning_rate": 7.769390345102518e-05, "loss": 0.0047, "step": 34750 }, { "grad_norm": 0.05830943584442139, "learning_rate": 7.768013518378683e-05, "loss": 0.0055, "step": 34760 }, { "grad_norm": 0.0856640562415123, "learning_rate": 7.766636388949053e-05, "loss": 0.0054, "step": 34770 }, { "grad_norm": 0.07178892195224762, "learning_rate": 7.765258956964229e-05, "loss": 0.0056, "step": 34780 }, { "grad_norm": 0.1023157611489296, "learning_rate": 7.76388122257484e-05, "loss": 0.0052, "step": 34790 }, { "grad_norm": 0.09533276408910751, "learning_rate": 7.762503185931558e-05, "loss": 0.005, "step": 34800 }, { "grad_norm": 0.10884730517864227, "learning_rate": 7.76112484718508e-05, "loss": 0.0058, "step": 34810 }, { "grad_norm": 0.10399377346038818, "learning_rate": 7.75974620648614e-05, "loss": 0.005, "step": 34820 }, { "grad_norm": 0.09824783354997635, "learning_rate": 7.758367263985503e-05, "loss": 0.0054, "step": 34830 }, { "grad_norm": 0.09909436106681824, "learning_rate": 7.75698801983397e-05, "loss": 0.0044, "step": 34840 }, { "grad_norm": 0.105819471180439, "learning_rate": 7.755608474182372e-05, "loss": 0.0057, "step": 34850 }, { "grad_norm": 0.0750407949090004, "learning_rate": 7.754228627181574e-05, "loss": 0.0058, "step": 34860 }, { "grad_norm": 0.09121038764715195, "learning_rate": 7.752848478982476e-05, "loss": 0.0053, "step": 34870 }, { "grad_norm": 0.08819735795259476, "learning_rate": 7.751468029736006e-05, "loss": 0.0052, "step": 34880 }, { "grad_norm": 0.0723751038312912, "learning_rate": 7.750087279593129e-05, "loss": 0.0041, "step": 34890 }, { "grad_norm": 0.07938013970851898, "learning_rate": 7.748706228704843e-05, "loss": 0.0048, "step": 34900 }, { "grad_norm": 0.06690184026956558, "learning_rate": 7.747324877222176e-05, "loss": 0.0049, "step": 34910 }, { "grad_norm": 0.1582815945148468, "learning_rate": 7.745943225296188e-05, "loss": 0.0055, "step": 34920 }, { "grad_norm": 0.10211540758609772, "learning_rate": 7.744561273077981e-05, "loss": 0.0055, "step": 34930 }, { "grad_norm": 0.10617775470018387, "learning_rate": 7.743179020718678e-05, "loss": 0.0059, "step": 34940 }, { "grad_norm": 0.10786308348178864, "learning_rate": 7.741796468369443e-05, "loss": 0.005, "step": 34950 }, { "grad_norm": 0.07887759804725647, "learning_rate": 7.740413616181466e-05, "loss": 0.0054, "step": 34960 }, { "grad_norm": 0.08621536195278168, "learning_rate": 7.739030464305978e-05, "loss": 0.0053, "step": 34970 }, { "grad_norm": 0.10111987590789795, "learning_rate": 7.737647012894235e-05, "loss": 0.0049, "step": 34980 }, { "grad_norm": 0.0741083174943924, "learning_rate": 7.736263262097532e-05, "loss": 0.005, "step": 34990 }, { "grad_norm": 0.06255330890417099, "learning_rate": 7.734879212067192e-05, "loss": 0.0048, "step": 35000 }, { "grad_norm": 0.09806478023529053, "learning_rate": 7.733494862954573e-05, "loss": 0.0044, "step": 35010 }, { "grad_norm": 0.0785619392991066, "learning_rate": 7.732110214911066e-05, "loss": 0.0053, "step": 35020 }, { "grad_norm": 0.07978715002536774, "learning_rate": 7.730725268088094e-05, "loss": 0.0047, "step": 35030 }, { "grad_norm": 0.07777281105518341, "learning_rate": 7.729340022637112e-05, "loss": 0.005, "step": 35040 }, { "grad_norm": 0.07697845995426178, "learning_rate": 7.727954478709607e-05, "loss": 0.0049, "step": 35050 }, { "grad_norm": 0.09056830406188965, "learning_rate": 7.726568636457103e-05, "loss": 0.0044, "step": 35060 }, { "grad_norm": 0.06107325851917267, "learning_rate": 7.725182496031153e-05, "loss": 0.0043, "step": 35070 }, { "grad_norm": 0.09841863065958023, "learning_rate": 7.723796057583341e-05, "loss": 0.0056, "step": 35080 }, { "grad_norm": 0.1045713871717453, "learning_rate": 7.722409321265287e-05, "loss": 0.0054, "step": 35090 }, { "grad_norm": 0.07526394724845886, "learning_rate": 7.721022287228645e-05, "loss": 0.0042, "step": 35100 }, { "grad_norm": 0.0810876414179802, "learning_rate": 7.719634955625094e-05, "loss": 0.0059, "step": 35110 }, { "grad_norm": 0.10351298004388809, "learning_rate": 7.718247326606355e-05, "loss": 0.0069, "step": 35120 }, { "grad_norm": 0.1795121431350708, "learning_rate": 7.716859400324175e-05, "loss": 0.005, "step": 35130 }, { "grad_norm": 0.08183453977108002, "learning_rate": 7.715471176930335e-05, "loss": 0.0066, "step": 35140 }, { "grad_norm": 0.10198360681533813, "learning_rate": 7.714082656576651e-05, "loss": 0.0056, "step": 35150 }, { "grad_norm": 0.11061017215251923, "learning_rate": 7.712693839414968e-05, "loss": 0.0061, "step": 35160 }, { "grad_norm": 0.0819166824221611, "learning_rate": 7.711304725597164e-05, "loss": 0.0056, "step": 35170 }, { "grad_norm": 0.0814613625407219, "learning_rate": 7.709915315275151e-05, "loss": 0.0053, "step": 35180 }, { "grad_norm": 0.093715138733387, "learning_rate": 7.708525608600876e-05, "loss": 0.0052, "step": 35190 }, { "grad_norm": 0.07040296494960785, "learning_rate": 7.707135605726311e-05, "loss": 0.0055, "step": 35200 }, { "grad_norm": 0.10513614863157272, "learning_rate": 7.705745306803466e-05, "loss": 0.0045, "step": 35210 }, { "grad_norm": 0.09913155436515808, "learning_rate": 7.704354711984383e-05, "loss": 0.0063, "step": 35220 }, { "grad_norm": 0.09151507914066315, "learning_rate": 7.702963821421133e-05, "loss": 0.0058, "step": 35230 }, { "grad_norm": 0.07403352856636047, "learning_rate": 7.701572635265826e-05, "loss": 0.0062, "step": 35240 }, { "grad_norm": 0.08204447478055954, "learning_rate": 7.700181153670596e-05, "loss": 0.0047, "step": 35250 }, { "grad_norm": 0.07901014387607574, "learning_rate": 7.698789376787616e-05, "loss": 0.0051, "step": 35260 }, { "grad_norm": 0.0930560901761055, "learning_rate": 7.697397304769085e-05, "loss": 0.007, "step": 35270 }, { "grad_norm": 0.09904618561267853, "learning_rate": 7.696004937767241e-05, "loss": 0.0061, "step": 35280 }, { "grad_norm": 0.09918916970491409, "learning_rate": 7.694612275934352e-05, "loss": 0.005, "step": 35290 }, { "grad_norm": 0.08005815744400024, "learning_rate": 7.693219319422714e-05, "loss": 0.0058, "step": 35300 }, { "grad_norm": 0.0944032296538353, "learning_rate": 7.69182606838466e-05, "loss": 0.0056, "step": 35310 }, { "grad_norm": 0.11157611012458801, "learning_rate": 7.690432522972558e-05, "loss": 0.0049, "step": 35320 }, { "grad_norm": 0.08750426024198532, "learning_rate": 7.689038683338796e-05, "loss": 0.0063, "step": 35330 }, { "grad_norm": 0.07040144503116608, "learning_rate": 7.687644549635808e-05, "loss": 0.005, "step": 35340 }, { "grad_norm": 0.08900479972362518, "learning_rate": 7.686250122016053e-05, "loss": 0.0045, "step": 35350 }, { "grad_norm": 0.07508295029401779, "learning_rate": 7.684855400632022e-05, "loss": 0.0064, "step": 35360 }, { "grad_norm": 0.11803514510393143, "learning_rate": 7.683460385636243e-05, "loss": 0.0056, "step": 35370 }, { "grad_norm": 0.10466475039720535, "learning_rate": 7.68206507718127e-05, "loss": 0.0049, "step": 35380 }, { "grad_norm": 0.07951617240905762, "learning_rate": 7.680669475419693e-05, "loss": 0.0059, "step": 35390 }, { "grad_norm": 0.09851931780576706, "learning_rate": 7.679273580504132e-05, "loss": 0.0058, "step": 35400 }, { "grad_norm": 0.08396068215370178, "learning_rate": 7.67787739258724e-05, "loss": 0.0049, "step": 35410 }, { "grad_norm": 0.10487242043018341, "learning_rate": 7.676480911821705e-05, "loss": 0.0049, "step": 35420 }, { "grad_norm": 0.07434017211198807, "learning_rate": 7.675084138360238e-05, "loss": 0.0049, "step": 35430 }, { "grad_norm": 0.1403835117816925, "learning_rate": 7.673687072355592e-05, "loss": 0.0059, "step": 35440 }, { "grad_norm": 0.10487737506628036, "learning_rate": 7.67228971396055e-05, "loss": 0.0046, "step": 35450 }, { "grad_norm": 0.10497952252626419, "learning_rate": 7.670892063327922e-05, "loss": 0.0056, "step": 35460 }, { "grad_norm": 0.11490616202354431, "learning_rate": 7.669494120610552e-05, "loss": 0.0051, "step": 35470 }, { "grad_norm": 0.08787227421998978, "learning_rate": 7.66809588596132e-05, "loss": 0.0053, "step": 35480 }, { "grad_norm": 0.08480498194694519, "learning_rate": 7.666697359533132e-05, "loss": 0.005, "step": 35490 }, { "grad_norm": 0.09260202199220657, "learning_rate": 7.665298541478932e-05, "loss": 0.0048, "step": 35500 }, { "grad_norm": 0.07801716774702072, "learning_rate": 7.66389943195169e-05, "loss": 0.0048, "step": 35510 }, { "grad_norm": 0.10963737219572067, "learning_rate": 7.66250003110441e-05, "loss": 0.0047, "step": 35520 }, { "grad_norm": 0.07206618040800095, "learning_rate": 7.66110033909013e-05, "loss": 0.0054, "step": 35530 }, { "grad_norm": 0.08209228515625, "learning_rate": 7.659700356061918e-05, "loss": 0.0054, "step": 35540 }, { "grad_norm": 0.084974005818367, "learning_rate": 7.658300082172875e-05, "loss": 0.0064, "step": 35550 }, { "grad_norm": 0.08191252499818802, "learning_rate": 7.65689951757613e-05, "loss": 0.0051, "step": 35560 }, { "grad_norm": 0.07481682300567627, "learning_rate": 7.655498662424849e-05, "loss": 0.0052, "step": 35570 }, { "grad_norm": 0.08546600490808487, "learning_rate": 7.654097516872227e-05, "loss": 0.0044, "step": 35580 }, { "grad_norm": 0.07259590178728104, "learning_rate": 7.65269608107149e-05, "loss": 0.0057, "step": 35590 }, { "grad_norm": 0.06739213317632675, "learning_rate": 7.6512943551759e-05, "loss": 0.005, "step": 35600 }, { "grad_norm": 0.09639006853103638, "learning_rate": 7.649892339338743e-05, "loss": 0.0044, "step": 35610 }, { "grad_norm": 0.09978696703910828, "learning_rate": 7.648490033713344e-05, "loss": 0.0051, "step": 35620 }, { "grad_norm": 0.06863708049058914, "learning_rate": 7.647087438453058e-05, "loss": 0.0049, "step": 35630 }, { "grad_norm": 0.08843039721250534, "learning_rate": 7.64568455371127e-05, "loss": 0.0045, "step": 35640 }, { "grad_norm": 0.07814983278512955, "learning_rate": 7.644281379641396e-05, "loss": 0.0056, "step": 35650 }, { "grad_norm": 0.10639283806085587, "learning_rate": 7.642877916396887e-05, "loss": 0.0051, "step": 35660 }, { "grad_norm": 0.09244870394468307, "learning_rate": 7.641474164131221e-05, "loss": 0.0044, "step": 35670 }, { "grad_norm": 0.10251431167125702, "learning_rate": 7.640070122997913e-05, "loss": 0.0053, "step": 35680 }, { "grad_norm": 0.11560862511396408, "learning_rate": 7.638665793150505e-05, "loss": 0.0063, "step": 35690 }, { "grad_norm": 0.08279648423194885, "learning_rate": 7.637261174742574e-05, "loss": 0.0054, "step": 35700 }, { "grad_norm": 0.0823393389582634, "learning_rate": 7.635856267927725e-05, "loss": 0.0052, "step": 35710 }, { "grad_norm": 0.09452959895133972, "learning_rate": 7.634451072859597e-05, "loss": 0.0051, "step": 35720 }, { "grad_norm": 0.11700112372636795, "learning_rate": 7.633045589691863e-05, "loss": 0.0059, "step": 35730 }, { "grad_norm": 0.08586615324020386, "learning_rate": 7.63163981857822e-05, "loss": 0.0054, "step": 35740 }, { "grad_norm": 0.06298106908798218, "learning_rate": 7.630233759672403e-05, "loss": 0.0055, "step": 35750 }, { "grad_norm": 0.19043929874897003, "learning_rate": 7.628827413128177e-05, "loss": 0.007, "step": 35760 }, { "grad_norm": 0.07196421176195145, "learning_rate": 7.627420779099338e-05, "loss": 0.0054, "step": 35770 }, { "grad_norm": 0.1089271605014801, "learning_rate": 7.626013857739711e-05, "loss": 0.0053, "step": 35780 }, { "grad_norm": 0.09010601788759232, "learning_rate": 7.624606649203158e-05, "loss": 0.005, "step": 35790 }, { "grad_norm": 0.08751551061868668, "learning_rate": 7.623199153643569e-05, "loss": 0.0044, "step": 35800 }, { "grad_norm": 0.06888308376073837, "learning_rate": 7.621791371214863e-05, "loss": 0.0043, "step": 35810 }, { "grad_norm": 0.2106829732656479, "learning_rate": 7.620383302070995e-05, "loss": 0.0054, "step": 35820 }, { "grad_norm": 0.084720179438591, "learning_rate": 7.61897494636595e-05, "loss": 0.0053, "step": 35830 }, { "grad_norm": 0.11423788219690323, "learning_rate": 7.617566304253739e-05, "loss": 0.0053, "step": 35840 }, { "grad_norm": 0.06815073639154434, "learning_rate": 7.616157375888416e-05, "loss": 0.0048, "step": 35850 }, { "grad_norm": 0.09678176790475845, "learning_rate": 7.614748161424053e-05, "loss": 0.0055, "step": 35860 }, { "grad_norm": 0.09208299964666367, "learning_rate": 7.613338661014763e-05, "loss": 0.0058, "step": 35870 }, { "grad_norm": 0.09311780333518982, "learning_rate": 7.611928874814686e-05, "loss": 0.0042, "step": 35880 }, { "grad_norm": 0.09469356387853622, "learning_rate": 7.610518802977996e-05, "loss": 0.0054, "step": 35890 }, { "grad_norm": 0.12650620937347412, "learning_rate": 7.609108445658893e-05, "loss": 0.0048, "step": 35900 }, { "grad_norm": 0.09988320618867874, "learning_rate": 7.607697803011612e-05, "loss": 0.0044, "step": 35910 }, { "grad_norm": 0.11301302909851074, "learning_rate": 7.606286875190421e-05, "loss": 0.0057, "step": 35920 }, { "grad_norm": 0.09468971192836761, "learning_rate": 7.604875662349617e-05, "loss": 0.0083, "step": 35930 }, { "grad_norm": 0.11875271797180176, "learning_rate": 7.603464164643525e-05, "loss": 0.0071, "step": 35940 }, { "grad_norm": 0.11382992565631866, "learning_rate": 7.602052382226507e-05, "loss": 0.0059, "step": 35950 }, { "grad_norm": 0.09144569933414459, "learning_rate": 7.600640315252954e-05, "loss": 0.006, "step": 35960 }, { "grad_norm": 0.08179853111505508, "learning_rate": 7.599227963877284e-05, "loss": 0.0051, "step": 35970 }, { "grad_norm": 0.09012430161237717, "learning_rate": 7.597815328253954e-05, "loss": 0.0055, "step": 35980 }, { "grad_norm": 0.07716470956802368, "learning_rate": 7.596402408537443e-05, "loss": 0.0061, "step": 35990 }, { "grad_norm": 0.10422058403491974, "learning_rate": 7.59498920488227e-05, "loss": 0.0061, "step": 36000 }, { "grad_norm": 0.10989837348461151, "learning_rate": 7.593575717442979e-05, "loss": 0.0061, "step": 36010 }, { "grad_norm": 0.10818437486886978, "learning_rate": 7.592161946374147e-05, "loss": 0.005, "step": 36020 }, { "grad_norm": 0.08913133293390274, "learning_rate": 7.590747891830381e-05, "loss": 0.0054, "step": 36030 }, { "grad_norm": 0.06836891174316406, "learning_rate": 7.58933355396632e-05, "loss": 0.0058, "step": 36040 }, { "grad_norm": 0.1343550980091095, "learning_rate": 7.587918932936636e-05, "loss": 0.0075, "step": 36050 }, { "grad_norm": 0.10773255676031113, "learning_rate": 7.586504028896028e-05, "loss": 0.0077, "step": 36060 }, { "grad_norm": 0.16348567605018616, "learning_rate": 7.585088841999228e-05, "loss": 0.0069, "step": 36070 }, { "grad_norm": 0.11289560049772263, "learning_rate": 7.583673372400999e-05, "loss": 0.0066, "step": 36080 }, { "grad_norm": 0.13084085285663605, "learning_rate": 7.582257620256134e-05, "loss": 0.006, "step": 36090 }, { "grad_norm": 0.13854679465293884, "learning_rate": 7.580841585719458e-05, "loss": 0.0056, "step": 36100 }, { "grad_norm": 0.08947505801916122, "learning_rate": 7.579425268945825e-05, "loss": 0.0061, "step": 36110 }, { "grad_norm": 0.10640763491392136, "learning_rate": 7.578008670090127e-05, "loss": 0.006, "step": 36120 }, { "grad_norm": 0.09740900993347168, "learning_rate": 7.576591789307272e-05, "loss": 0.0071, "step": 36130 }, { "grad_norm": 0.08335921168327332, "learning_rate": 7.575174626752216e-05, "loss": 0.006, "step": 36140 }, { "grad_norm": 0.13269203901290894, "learning_rate": 7.573757182579934e-05, "loss": 0.005, "step": 36150 }, { "grad_norm": 0.08726518601179123, "learning_rate": 7.572339456945435e-05, "loss": 0.0061, "step": 36160 }, { "grad_norm": 0.0697655901312828, "learning_rate": 7.570921450003762e-05, "loss": 0.0056, "step": 36170 }, { "grad_norm": 0.11156445741653442, "learning_rate": 7.569503161909983e-05, "loss": 0.0062, "step": 36180 }, { "grad_norm": 0.09893647581338882, "learning_rate": 7.568084592819202e-05, "loss": 0.006, "step": 36190 }, { "grad_norm": 0.07211235165596008, "learning_rate": 7.566665742886551e-05, "loss": 0.006, "step": 36200 }, { "grad_norm": 0.11352059990167618, "learning_rate": 7.565246612267194e-05, "loss": 0.0054, "step": 36210 }, { "grad_norm": 0.08867021650075912, "learning_rate": 7.563827201116325e-05, "loss": 0.0058, "step": 36220 }, { "grad_norm": 0.09853993356227875, "learning_rate": 7.562407509589164e-05, "loss": 0.0058, "step": 36230 }, { "grad_norm": 0.0766797885298729, "learning_rate": 7.560987537840973e-05, "loss": 0.005, "step": 36240 }, { "grad_norm": 0.07420503348112106, "learning_rate": 7.559567286027036e-05, "loss": 0.0057, "step": 36250 }, { "grad_norm": 0.06713148206472397, "learning_rate": 7.558146754302668e-05, "loss": 0.0059, "step": 36260 }, { "grad_norm": 0.06483450531959534, "learning_rate": 7.556725942823217e-05, "loss": 0.0054, "step": 36270 }, { "grad_norm": 0.12176014482975006, "learning_rate": 7.555304851744061e-05, "loss": 0.0056, "step": 36280 }, { "grad_norm": 0.09091707319021225, "learning_rate": 7.553883481220608e-05, "loss": 0.0052, "step": 36290 }, { "grad_norm": 0.11436028778553009, "learning_rate": 7.552461831408298e-05, "loss": 0.0061, "step": 36300 }, { "grad_norm": 0.09938210248947144, "learning_rate": 7.551039902462599e-05, "loss": 0.005, "step": 36310 }, { "grad_norm": 0.12627631425857544, "learning_rate": 7.549617694539011e-05, "loss": 0.0067, "step": 36320 }, { "grad_norm": 0.09738034754991531, "learning_rate": 7.548195207793066e-05, "loss": 0.0057, "step": 36330 }, { "grad_norm": 0.09668127447366714, "learning_rate": 7.546772442380323e-05, "loss": 0.005, "step": 36340 }, { "grad_norm": 0.07687513530254364, "learning_rate": 7.545349398456376e-05, "loss": 0.0049, "step": 36350 }, { "grad_norm": 0.08742531388998032, "learning_rate": 7.543926076176845e-05, "loss": 0.0046, "step": 36360 }, { "grad_norm": 0.07795562595129013, "learning_rate": 7.542502475697385e-05, "loss": 0.0047, "step": 36370 }, { "grad_norm": 0.059372082352638245, "learning_rate": 7.541078597173675e-05, "loss": 0.0047, "step": 36380 }, { "grad_norm": 0.07219819724559784, "learning_rate": 7.539654440761431e-05, "loss": 0.0052, "step": 36390 }, { "grad_norm": 0.12513287365436554, "learning_rate": 7.538230006616395e-05, "loss": 0.006, "step": 36400 }, { "grad_norm": 0.08285967260599136, "learning_rate": 7.536805294894342e-05, "loss": 0.0047, "step": 36410 }, { "grad_norm": 0.07844068109989166, "learning_rate": 7.535380305751076e-05, "loss": 0.0044, "step": 36420 }, { "grad_norm": 0.08270315080881119, "learning_rate": 7.533955039342431e-05, "loss": 0.0047, "step": 36430 }, { "grad_norm": 0.08727811276912689, "learning_rate": 7.532529495824274e-05, "loss": 0.0046, "step": 36440 }, { "grad_norm": 0.048749133944511414, "learning_rate": 7.531103675352497e-05, "loss": 0.0046, "step": 36450 }, { "grad_norm": 0.09396115690469742, "learning_rate": 7.529677578083029e-05, "loss": 0.0054, "step": 36460 }, { "grad_norm": 0.1320187896490097, "learning_rate": 7.528251204171823e-05, "loss": 0.0045, "step": 36470 }, { "grad_norm": 0.07342706620693207, "learning_rate": 7.526824553774867e-05, "loss": 0.0051, "step": 36480 }, { "grad_norm": 0.08572118729352951, "learning_rate": 7.525397627048178e-05, "loss": 0.0046, "step": 36490 }, { "grad_norm": 0.09532064944505692, "learning_rate": 7.523970424147802e-05, "loss": 0.0049, "step": 36500 }, { "grad_norm": 0.07826346158981323, "learning_rate": 7.522542945229813e-05, "loss": 0.0045, "step": 36510 }, { "grad_norm": 0.07431714981794357, "learning_rate": 7.521115190450321e-05, "loss": 0.0048, "step": 36520 }, { "grad_norm": 0.10802292823791504, "learning_rate": 7.519687159965462e-05, "loss": 0.0051, "step": 36530 }, { "grad_norm": 0.09783908724784851, "learning_rate": 7.518258853931403e-05, "loss": 0.0054, "step": 36540 }, { "grad_norm": 0.0802980586886406, "learning_rate": 7.516830272504342e-05, "loss": 0.0046, "step": 36550 }, { "grad_norm": 0.10032613575458527, "learning_rate": 7.515401415840509e-05, "loss": 0.0071, "step": 36560 }, { "grad_norm": 0.1128876805305481, "learning_rate": 7.513972284096155e-05, "loss": 0.0059, "step": 36570 }, { "grad_norm": 0.07027676701545715, "learning_rate": 7.512542877427574e-05, "loss": 0.0046, "step": 36580 }, { "grad_norm": 0.08206039667129517, "learning_rate": 7.51111319599108e-05, "loss": 0.0054, "step": 36590 }, { "grad_norm": 0.05704151466488838, "learning_rate": 7.509683239943024e-05, "loss": 0.0053, "step": 36600 }, { "grad_norm": 0.12475931644439697, "learning_rate": 7.50825300943978e-05, "loss": 0.0064, "step": 36610 }, { "grad_norm": 0.05985477939248085, "learning_rate": 7.50682250463776e-05, "loss": 0.0051, "step": 36620 }, { "grad_norm": 0.10182714462280273, "learning_rate": 7.5053917256934e-05, "loss": 0.0047, "step": 36630 }, { "grad_norm": 0.0719185322523117, "learning_rate": 7.503960672763166e-05, "loss": 0.0055, "step": 36640 }, { "grad_norm": 0.08227360248565674, "learning_rate": 7.502529346003559e-05, "loss": 0.0041, "step": 36650 }, { "grad_norm": 0.07071094959974289, "learning_rate": 7.501097745571106e-05, "loss": 0.0044, "step": 36660 }, { "grad_norm": 0.10038372874259949, "learning_rate": 7.499665871622361e-05, "loss": 0.0045, "step": 36670 }, { "grad_norm": 0.08551924675703049, "learning_rate": 7.498233724313917e-05, "loss": 0.0045, "step": 36680 }, { "grad_norm": 0.09404309839010239, "learning_rate": 7.496801303802387e-05, "loss": 0.0049, "step": 36690 }, { "grad_norm": 0.07478833943605423, "learning_rate": 7.495368610244423e-05, "loss": 0.0057, "step": 36700 }, { "grad_norm": 0.08350277692079544, "learning_rate": 7.493935643796697e-05, "loss": 0.0043, "step": 36710 }, { "grad_norm": 0.10390989482402802, "learning_rate": 7.49250240461592e-05, "loss": 0.0056, "step": 36720 }, { "grad_norm": 0.09319161623716354, "learning_rate": 7.491068892858826e-05, "loss": 0.0056, "step": 36730 }, { "grad_norm": 0.10754012316465378, "learning_rate": 7.489635108682184e-05, "loss": 0.0046, "step": 36740 }, { "grad_norm": 0.10665606707334518, "learning_rate": 7.48820105224279e-05, "loss": 0.0053, "step": 36750 }, { "grad_norm": 0.08326774835586548, "learning_rate": 7.48676672369747e-05, "loss": 0.0059, "step": 36760 }, { "grad_norm": 0.08673085272312164, "learning_rate": 7.485332123203079e-05, "loss": 0.0047, "step": 36770 }, { "grad_norm": 0.09548291563987732, "learning_rate": 7.483897250916504e-05, "loss": 0.004, "step": 36780 }, { "grad_norm": 0.07987768948078156, "learning_rate": 7.48246210699466e-05, "loss": 0.0049, "step": 36790 }, { "grad_norm": 0.10402616858482361, "learning_rate": 7.481026691594492e-05, "loss": 0.0062, "step": 36800 }, { "grad_norm": 0.10468266904354095, "learning_rate": 7.479591004872975e-05, "loss": 0.0056, "step": 36810 }, { "grad_norm": 0.07696793973445892, "learning_rate": 7.478155046987114e-05, "loss": 0.0059, "step": 36820 }, { "grad_norm": 0.05261504277586937, "learning_rate": 7.476718818093944e-05, "loss": 0.0049, "step": 36830 }, { "grad_norm": 0.08487553894519806, "learning_rate": 7.475282318350527e-05, "loss": 0.0051, "step": 36840 }, { "grad_norm": 0.07028214633464813, "learning_rate": 7.473845547913957e-05, "loss": 0.0041, "step": 36850 }, { "grad_norm": 0.08976711332798004, "learning_rate": 7.472408506941357e-05, "loss": 0.0043, "step": 36860 }, { "grad_norm": 0.0805719792842865, "learning_rate": 7.470971195589881e-05, "loss": 0.0052, "step": 36870 }, { "grad_norm": 0.06665300577878952, "learning_rate": 7.46953361401671e-05, "loss": 0.0042, "step": 36880 }, { "grad_norm": 0.08066613972187042, "learning_rate": 7.468095762379055e-05, "loss": 0.0048, "step": 36890 }, { "grad_norm": 0.06795849651098251, "learning_rate": 7.466657640834158e-05, "loss": 0.0049, "step": 36900 }, { "grad_norm": 0.07369440048933029, "learning_rate": 7.46521924953929e-05, "loss": 0.0039, "step": 36910 }, { "grad_norm": 0.09592470526695251, "learning_rate": 7.463780588651752e-05, "loss": 0.0049, "step": 36920 }, { "grad_norm": 0.052282411605119705, "learning_rate": 7.462341658328873e-05, "loss": 0.004, "step": 36930 }, { "grad_norm": 0.08089721202850342, "learning_rate": 7.460902458728012e-05, "loss": 0.004, "step": 36940 }, { "grad_norm": 0.08333155512809753, "learning_rate": 7.459462990006558e-05, "loss": 0.006, "step": 36950 }, { "grad_norm": 0.047424398362636566, "learning_rate": 7.458023252321929e-05, "loss": 0.0042, "step": 36960 }, { "grad_norm": 0.0747370570898056, "learning_rate": 7.456583245831573e-05, "loss": 0.0041, "step": 36970 }, { "grad_norm": 0.0988735556602478, "learning_rate": 7.455142970692968e-05, "loss": 0.0047, "step": 36980 }, { "grad_norm": 0.06385873258113861, "learning_rate": 7.453702427063615e-05, "loss": 0.0058, "step": 36990 }, { "grad_norm": 0.07944260537624359, "learning_rate": 7.452261615101057e-05, "loss": 0.0051, "step": 37000 }, { "grad_norm": 0.06591238081455231, "learning_rate": 7.450820534962856e-05, "loss": 0.0052, "step": 37010 }, { "grad_norm": 0.07320266216993332, "learning_rate": 7.449379186806604e-05, "loss": 0.0043, "step": 37020 }, { "grad_norm": 0.11244235932826996, "learning_rate": 7.44793757078993e-05, "loss": 0.0058, "step": 37030 }, { "grad_norm": 0.08692619204521179, "learning_rate": 7.44649568707048e-05, "loss": 0.0049, "step": 37040 }, { "grad_norm": 0.07493840903043747, "learning_rate": 7.445053535805942e-05, "loss": 0.0041, "step": 37050 }, { "grad_norm": 0.07236506789922714, "learning_rate": 7.443611117154024e-05, "loss": 0.005, "step": 37060 }, { "grad_norm": 0.08026251941919327, "learning_rate": 7.442168431272469e-05, "loss": 0.0043, "step": 37070 }, { "grad_norm": 0.07906165719032288, "learning_rate": 7.440725478319044e-05, "loss": 0.0061, "step": 37080 }, { "grad_norm": 0.06648439168930054, "learning_rate": 7.439282258451553e-05, "loss": 0.0047, "step": 37090 }, { "grad_norm": 0.1078457161784172, "learning_rate": 7.43783877182782e-05, "loss": 0.0048, "step": 37100 }, { "grad_norm": 0.0789954885840416, "learning_rate": 7.436395018605707e-05, "loss": 0.0048, "step": 37110 }, { "grad_norm": 0.06919383257627487, "learning_rate": 7.434950998943094e-05, "loss": 0.0057, "step": 37120 }, { "grad_norm": 0.06335172057151794, "learning_rate": 7.433506712997902e-05, "loss": 0.004, "step": 37130 }, { "grad_norm": 0.08512979000806808, "learning_rate": 7.432062160928077e-05, "loss": 0.0047, "step": 37140 }, { "grad_norm": 0.08110515028238297, "learning_rate": 7.430617342891588e-05, "loss": 0.0046, "step": 37150 }, { "grad_norm": 0.07614468038082123, "learning_rate": 7.429172259046442e-05, "loss": 0.0041, "step": 37160 }, { "grad_norm": 0.09315472841262817, "learning_rate": 7.42772690955067e-05, "loss": 0.0047, "step": 37170 }, { "grad_norm": 0.08981503546237946, "learning_rate": 7.426281294562332e-05, "loss": 0.0048, "step": 37180 }, { "grad_norm": 0.07736874371767044, "learning_rate": 7.424835414239519e-05, "loss": 0.0044, "step": 37190 }, { "grad_norm": 0.11939584463834763, "learning_rate": 7.423389268740352e-05, "loss": 0.0051, "step": 37200 }, { "grad_norm": 0.0644964650273323, "learning_rate": 7.42194285822298e-05, "loss": 0.005, "step": 37210 }, { "grad_norm": 0.07918615639209747, "learning_rate": 7.420496182845576e-05, "loss": 0.0043, "step": 37220 }, { "grad_norm": 0.07517626881599426, "learning_rate": 7.419049242766348e-05, "loss": 0.0042, "step": 37230 }, { "grad_norm": 0.07255571335554123, "learning_rate": 7.417602038143533e-05, "loss": 0.0043, "step": 37240 }, { "grad_norm": 0.08706789463758469, "learning_rate": 7.416154569135393e-05, "loss": 0.0045, "step": 37250 }, { "grad_norm": 0.13086651265621185, "learning_rate": 7.414706835900222e-05, "loss": 0.0054, "step": 37260 }, { "grad_norm": 0.08753592520952225, "learning_rate": 7.413258838596345e-05, "loss": 0.0049, "step": 37270 }, { "grad_norm": 0.1168084517121315, "learning_rate": 7.411810577382106e-05, "loss": 0.0055, "step": 37280 }, { "grad_norm": 0.07944998145103455, "learning_rate": 7.410362052415889e-05, "loss": 0.0046, "step": 37290 }, { "grad_norm": 0.0948445051908493, "learning_rate": 7.408913263856102e-05, "loss": 0.005, "step": 37300 }, { "grad_norm": 0.07682090997695923, "learning_rate": 7.407464211861183e-05, "loss": 0.0049, "step": 37310 }, { "grad_norm": 0.06354981660842896, "learning_rate": 7.406014896589597e-05, "loss": 0.0036, "step": 37320 }, { "grad_norm": 0.08589698374271393, "learning_rate": 7.404565318199838e-05, "loss": 0.0051, "step": 37330 }, { "grad_norm": 0.07840652763843536, "learning_rate": 7.403115476850432e-05, "loss": 0.0046, "step": 37340 }, { "grad_norm": 0.0870472639799118, "learning_rate": 7.401665372699932e-05, "loss": 0.0057, "step": 37350 }, { "grad_norm": 0.10658686608076096, "learning_rate": 7.400215005906916e-05, "loss": 0.0049, "step": 37360 }, { "grad_norm": 0.07971178740262985, "learning_rate": 7.398764376629996e-05, "loss": 0.0055, "step": 37370 }, { "grad_norm": 0.08684606105089188, "learning_rate": 7.39731348502781e-05, "loss": 0.0042, "step": 37380 }, { "grad_norm": 0.08061421662569046, "learning_rate": 7.395862331259028e-05, "loss": 0.0049, "step": 37390 }, { "grad_norm": 0.07180366665124893, "learning_rate": 7.394410915482341e-05, "loss": 0.0042, "step": 37400 }, { "grad_norm": 0.1156533882021904, "learning_rate": 7.392959237856476e-05, "loss": 0.0038, "step": 37410 }, { "grad_norm": 0.06756961345672607, "learning_rate": 7.391507298540187e-05, "loss": 0.0048, "step": 37420 }, { "grad_norm": 0.08909817785024643, "learning_rate": 7.390055097692257e-05, "loss": 0.0039, "step": 37430 }, { "grad_norm": 0.10328825563192368, "learning_rate": 7.388602635471494e-05, "loss": 0.0039, "step": 37440 }, { "grad_norm": 0.07408862560987473, "learning_rate": 7.387149912036738e-05, "loss": 0.0048, "step": 37450 }, { "grad_norm": 0.08345276862382889, "learning_rate": 7.385696927546854e-05, "loss": 0.0049, "step": 37460 }, { "grad_norm": 0.10440807789564133, "learning_rate": 7.384243682160743e-05, "loss": 0.0052, "step": 37470 }, { "grad_norm": 0.0780286192893982, "learning_rate": 7.382790176037329e-05, "loss": 0.0041, "step": 37480 }, { "grad_norm": 0.08368676900863647, "learning_rate": 7.381336409335562e-05, "loss": 0.0048, "step": 37490 }, { "grad_norm": 0.07139823585748672, "learning_rate": 7.379882382214426e-05, "loss": 0.0046, "step": 37500 }, { "grad_norm": 0.07796372473239899, "learning_rate": 7.378428094832931e-05, "loss": 0.0042, "step": 37510 }, { "grad_norm": 0.07060053199529648, "learning_rate": 7.376973547350114e-05, "loss": 0.0042, "step": 37520 }, { "grad_norm": 0.08591002970933914, "learning_rate": 7.375518739925046e-05, "loss": 0.0041, "step": 37530 }, { "grad_norm": 0.08363981544971466, "learning_rate": 7.374063672716818e-05, "loss": 0.0043, "step": 37540 }, { "grad_norm": 0.07302320748567581, "learning_rate": 7.372608345884558e-05, "loss": 0.0053, "step": 37550 }, { "grad_norm": 0.06702364236116409, "learning_rate": 7.371152759587415e-05, "loss": 0.0053, "step": 37560 }, { "grad_norm": 0.06203567609190941, "learning_rate": 7.369696913984572e-05, "loss": 0.0036, "step": 37570 }, { "grad_norm": 0.0759623721241951, "learning_rate": 7.368240809235237e-05, "loss": 0.0045, "step": 37580 }, { "grad_norm": 0.09971918910741806, "learning_rate": 7.366784445498649e-05, "loss": 0.0046, "step": 37590 }, { "grad_norm": 0.08755284547805786, "learning_rate": 7.36532782293407e-05, "loss": 0.0038, "step": 37600 }, { "grad_norm": 0.08570840954780579, "learning_rate": 7.363870941700797e-05, "loss": 0.0043, "step": 37610 }, { "grad_norm": 0.09931222349405289, "learning_rate": 7.362413801958152e-05, "loss": 0.0043, "step": 37620 }, { "grad_norm": 0.0819491297006607, "learning_rate": 7.360956403865486e-05, "loss": 0.0047, "step": 37630 }, { "grad_norm": 0.09456393122673035, "learning_rate": 7.359498747582177e-05, "loss": 0.0044, "step": 37640 }, { "grad_norm": 0.071699358522892, "learning_rate": 7.358040833267634e-05, "loss": 0.0044, "step": 37650 }, { "grad_norm": 0.08480697870254517, "learning_rate": 7.356582661081288e-05, "loss": 0.005, "step": 37660 }, { "grad_norm": 0.050781700760126114, "learning_rate": 7.355124231182607e-05, "loss": 0.0041, "step": 37670 }, { "grad_norm": 0.052999347448349, "learning_rate": 7.353665543731083e-05, "loss": 0.0042, "step": 37680 }, { "grad_norm": 0.08323159068822861, "learning_rate": 7.352206598886229e-05, "loss": 0.0048, "step": 37690 }, { "grad_norm": 0.08822634816169739, "learning_rate": 7.350747396807601e-05, "loss": 0.0046, "step": 37700 }, { "grad_norm": 0.07263915985822678, "learning_rate": 7.349287937654772e-05, "loss": 0.0045, "step": 37710 }, { "grad_norm": 0.08310758322477341, "learning_rate": 7.347828221587345e-05, "loss": 0.0048, "step": 37720 }, { "grad_norm": 0.1043071374297142, "learning_rate": 7.346368248764955e-05, "loss": 0.0044, "step": 37730 }, { "grad_norm": 0.07431557774543762, "learning_rate": 7.34490801934726e-05, "loss": 0.0043, "step": 37740 }, { "grad_norm": 0.10198839753866196, "learning_rate": 7.343447533493947e-05, "loss": 0.0061, "step": 37750 }, { "grad_norm": 0.07866591960191727, "learning_rate": 7.341986791364739e-05, "loss": 0.0052, "step": 37760 }, { "grad_norm": 0.06675819307565689, "learning_rate": 7.340525793119374e-05, "loss": 0.0047, "step": 37770 }, { "grad_norm": 0.0818929672241211, "learning_rate": 7.33906453891763e-05, "loss": 0.0038, "step": 37780 }, { "grad_norm": 0.0750902071595192, "learning_rate": 7.337603028919303e-05, "loss": 0.0047, "step": 37790 }, { "grad_norm": 0.07601331919431686, "learning_rate": 7.336141263284225e-05, "loss": 0.0041, "step": 37800 }, { "grad_norm": 0.09781038761138916, "learning_rate": 7.334679242172249e-05, "loss": 0.0051, "step": 37810 }, { "grad_norm": 0.07410521805286407, "learning_rate": 7.33321696574326e-05, "loss": 0.0045, "step": 37820 }, { "grad_norm": 0.07231982052326202, "learning_rate": 7.331754434157174e-05, "loss": 0.0052, "step": 37830 }, { "grad_norm": 0.0976276844739914, "learning_rate": 7.330291647573928e-05, "loss": 0.005, "step": 37840 }, { "grad_norm": 0.12121003866195679, "learning_rate": 7.32882860615349e-05, "loss": 0.0044, "step": 37850 }, { "grad_norm": 0.10502840578556061, "learning_rate": 7.327365310055858e-05, "loss": 0.0044, "step": 37860 }, { "grad_norm": 0.0871213898062706, "learning_rate": 7.325901759441055e-05, "loss": 0.0043, "step": 37870 }, { "grad_norm": 0.09560661017894745, "learning_rate": 7.324437954469132e-05, "loss": 0.0048, "step": 37880 }, { "grad_norm": 0.07902438193559647, "learning_rate": 7.32297389530017e-05, "loss": 0.0046, "step": 37890 }, { "grad_norm": 0.09304910898208618, "learning_rate": 7.321509582094274e-05, "loss": 0.0047, "step": 37900 }, { "grad_norm": 0.07509811967611313, "learning_rate": 7.320045015011582e-05, "loss": 0.0049, "step": 37910 }, { "grad_norm": 0.08348529785871506, "learning_rate": 7.318580194212253e-05, "loss": 0.0047, "step": 37920 }, { "grad_norm": 0.0944826602935791, "learning_rate": 7.317115119856483e-05, "loss": 0.0048, "step": 37930 }, { "grad_norm": 0.06586436182260513, "learning_rate": 7.315649792104485e-05, "loss": 0.0038, "step": 37940 }, { "grad_norm": 0.1132541373372078, "learning_rate": 7.314184211116507e-05, "loss": 0.0041, "step": 37950 }, { "grad_norm": 0.08458216488361359, "learning_rate": 7.312718377052823e-05, "loss": 0.0041, "step": 37960 }, { "grad_norm": 0.07441788166761398, "learning_rate": 7.311252290073737e-05, "loss": 0.0045, "step": 37970 }, { "grad_norm": 0.10011522471904755, "learning_rate": 7.309785950339572e-05, "loss": 0.0051, "step": 37980 }, { "grad_norm": 0.0899055227637291, "learning_rate": 7.308319358010692e-05, "loss": 0.0059, "step": 37990 }, { "grad_norm": 0.09138716757297516, "learning_rate": 7.306852513247474e-05, "loss": 0.006, "step": 38000 }, { "grad_norm": 0.09346889704465866, "learning_rate": 7.305385416210336e-05, "loss": 0.0049, "step": 38010 }, { "grad_norm": 0.11825516074895859, "learning_rate": 7.303918067059711e-05, "loss": 0.005, "step": 38020 }, { "grad_norm": 0.07543154805898666, "learning_rate": 7.302450465956075e-05, "loss": 0.0054, "step": 38030 }, { "grad_norm": 0.07622579485177994, "learning_rate": 7.300982613059914e-05, "loss": 0.0045, "step": 38040 }, { "grad_norm": 0.09168945997953415, "learning_rate": 7.299514508531757e-05, "loss": 0.0064, "step": 38050 }, { "grad_norm": 0.06512840837240219, "learning_rate": 7.298046152532148e-05, "loss": 0.0051, "step": 38060 }, { "grad_norm": 0.07904282957315445, "learning_rate": 7.296577545221668e-05, "loss": 0.0055, "step": 38070 }, { "grad_norm": 0.0691300556063652, "learning_rate": 7.29510868676092e-05, "loss": 0.005, "step": 38080 }, { "grad_norm": 0.060271911323070526, "learning_rate": 7.293639577310538e-05, "loss": 0.0044, "step": 38090 }, { "grad_norm": 0.07542581856250763, "learning_rate": 7.292170217031179e-05, "loss": 0.0053, "step": 38100 }, { "grad_norm": 0.07469379156827927, "learning_rate": 7.290700606083532e-05, "loss": 0.0044, "step": 38110 }, { "grad_norm": 0.08369069546461105, "learning_rate": 7.28923074462831e-05, "loss": 0.0049, "step": 38120 }, { "grad_norm": 0.07661259919404984, "learning_rate": 7.287760632826255e-05, "loss": 0.0049, "step": 38130 }, { "grad_norm": 0.07077974081039429, "learning_rate": 7.286290270838139e-05, "loss": 0.0045, "step": 38140 }, { "grad_norm": 0.08816853165626526, "learning_rate": 7.284819658824756e-05, "loss": 0.0045, "step": 38150 }, { "grad_norm": 0.10261167585849762, "learning_rate": 7.283348796946929e-05, "loss": 0.004, "step": 38160 }, { "grad_norm": 0.08031416684389114, "learning_rate": 7.28187768536551e-05, "loss": 0.0045, "step": 38170 }, { "grad_norm": 0.08130433410406113, "learning_rate": 7.280406324241379e-05, "loss": 0.0043, "step": 38180 }, { "grad_norm": 0.06179811805486679, "learning_rate": 7.278934713735441e-05, "loss": 0.0038, "step": 38190 }, { "grad_norm": 0.09138771146535873, "learning_rate": 7.277462854008629e-05, "loss": 0.0053, "step": 38200 }, { "grad_norm": 0.0939398854970932, "learning_rate": 7.275990745221904e-05, "loss": 0.0044, "step": 38210 }, { "grad_norm": 0.09995082020759583, "learning_rate": 7.274518387536252e-05, "loss": 0.0054, "step": 38220 }, { "grad_norm": 0.09910576045513153, "learning_rate": 7.273045781112689e-05, "loss": 0.0044, "step": 38230 }, { "grad_norm": 0.08745802193880081, "learning_rate": 7.271572926112257e-05, "loss": 0.0043, "step": 38240 }, { "grad_norm": 0.10470639169216156, "learning_rate": 7.270099822696024e-05, "loss": 0.006, "step": 38250 }, { "grad_norm": 0.07518558949232101, "learning_rate": 7.26862647102509e-05, "loss": 0.0056, "step": 38260 }, { "grad_norm": 0.0737374797463417, "learning_rate": 7.267152871260573e-05, "loss": 0.0055, "step": 38270 }, { "grad_norm": 0.0832301452755928, "learning_rate": 7.265679023563629e-05, "loss": 0.005, "step": 38280 }, { "grad_norm": 0.08875308930873871, "learning_rate": 7.264204928095433e-05, "loss": 0.0047, "step": 38290 }, { "grad_norm": 0.09386991709470749, "learning_rate": 7.262730585017188e-05, "loss": 0.0056, "step": 38300 }, { "grad_norm": 0.09849142283201218, "learning_rate": 7.261255994490131e-05, "loss": 0.0047, "step": 38310 }, { "grad_norm": 0.0999075248837471, "learning_rate": 7.259781156675517e-05, "loss": 0.0048, "step": 38320 }, { "grad_norm": 0.07590833306312561, "learning_rate": 7.258306071734633e-05, "loss": 0.0044, "step": 38330 }, { "grad_norm": 0.07967250049114227, "learning_rate": 7.256830739828792e-05, "loss": 0.0047, "step": 38340 }, { "grad_norm": 0.08689969778060913, "learning_rate": 7.255355161119336e-05, "loss": 0.0042, "step": 38350 }, { "grad_norm": 0.10697705298662186, "learning_rate": 7.253879335767628e-05, "loss": 0.0044, "step": 38360 }, { "grad_norm": 0.08790308237075806, "learning_rate": 7.252403263935067e-05, "loss": 0.0049, "step": 38370 }, { "grad_norm": 0.08677897602319717, "learning_rate": 7.25092694578307e-05, "loss": 0.0044, "step": 38380 }, { "grad_norm": 0.08613357692956924, "learning_rate": 7.249450381473087e-05, "loss": 0.0048, "step": 38390 }, { "grad_norm": 0.09982344508171082, "learning_rate": 7.247973571166593e-05, "loss": 0.0052, "step": 38400 }, { "grad_norm": 0.06349657475948334, "learning_rate": 7.246496515025089e-05, "loss": 0.0045, "step": 38410 }, { "grad_norm": 0.10225106030702591, "learning_rate": 7.245019213210101e-05, "loss": 0.0043, "step": 38420 }, { "grad_norm": 0.0733862966299057, "learning_rate": 7.24354166588319e-05, "loss": 0.0048, "step": 38430 }, { "grad_norm": 0.09511798620223999, "learning_rate": 7.242063873205935e-05, "loss": 0.0046, "step": 38440 }, { "grad_norm": 0.07314044237136841, "learning_rate": 7.240585835339946e-05, "loss": 0.0052, "step": 38450 }, { "grad_norm": 0.0773957148194313, "learning_rate": 7.239107552446857e-05, "loss": 0.0058, "step": 38460 }, { "grad_norm": 0.09081606566905975, "learning_rate": 7.237629024688334e-05, "loss": 0.0047, "step": 38470 }, { "grad_norm": 0.09366349130868912, "learning_rate": 7.236150252226064e-05, "loss": 0.0047, "step": 38480 }, { "grad_norm": 0.132660374045372, "learning_rate": 7.234671235221765e-05, "loss": 0.0056, "step": 38490 }, { "grad_norm": 0.09819842875003815, "learning_rate": 7.233191973837179e-05, "loss": 0.0056, "step": 38500 }, { "grad_norm": 0.08517108112573624, "learning_rate": 7.231712468234078e-05, "loss": 0.0047, "step": 38510 }, { "grad_norm": 0.0876772552728653, "learning_rate": 7.230232718574254e-05, "loss": 0.006, "step": 38520 }, { "grad_norm": 0.08597612380981445, "learning_rate": 7.228752725019535e-05, "loss": 0.0044, "step": 38530 }, { "grad_norm": 0.09499289840459824, "learning_rate": 7.227272487731769e-05, "loss": 0.0053, "step": 38540 }, { "grad_norm": 0.07722820341587067, "learning_rate": 7.225792006872831e-05, "loss": 0.0048, "step": 38550 }, { "grad_norm": 0.06079525128006935, "learning_rate": 7.224311282604628e-05, "loss": 0.005, "step": 38560 }, { "grad_norm": 0.07789815217256546, "learning_rate": 7.222830315089085e-05, "loss": 0.005, "step": 38570 }, { "grad_norm": 0.07582005113363266, "learning_rate": 7.22134910448816e-05, "loss": 0.0041, "step": 38580 }, { "grad_norm": 0.05648438259959221, "learning_rate": 7.219867650963839e-05, "loss": 0.0044, "step": 38590 }, { "grad_norm": 0.08427536487579346, "learning_rate": 7.21838595467813e-05, "loss": 0.0048, "step": 38600 }, { "grad_norm": 0.0865463986992836, "learning_rate": 7.216904015793067e-05, "loss": 0.0055, "step": 38610 }, { "grad_norm": 0.06912471354007721, "learning_rate": 7.215421834470713e-05, "loss": 0.0055, "step": 38620 }, { "grad_norm": 0.10597191005945206, "learning_rate": 7.213939410873158e-05, "loss": 0.0041, "step": 38630 }, { "grad_norm": 0.12978948652744293, "learning_rate": 7.212456745162519e-05, "loss": 0.0055, "step": 38640 }, { "grad_norm": 0.10128751397132874, "learning_rate": 7.210973837500937e-05, "loss": 0.0041, "step": 38650 }, { "grad_norm": 0.1320563703775406, "learning_rate": 7.209490688050578e-05, "loss": 0.0048, "step": 38660 }, { "grad_norm": 0.07012726366519928, "learning_rate": 7.208007296973641e-05, "loss": 0.0043, "step": 38670 }, { "grad_norm": 0.0883343294262886, "learning_rate": 7.206523664432345e-05, "loss": 0.0045, "step": 38680 }, { "grad_norm": 0.07722649723291397, "learning_rate": 7.205039790588939e-05, "loss": 0.0051, "step": 38690 }, { "grad_norm": 0.10905741900205612, "learning_rate": 7.203555675605697e-05, "loss": 0.0052, "step": 38700 }, { "grad_norm": 0.106388621032238, "learning_rate": 7.202071319644917e-05, "loss": 0.0054, "step": 38710 }, { "grad_norm": 0.12289871275424957, "learning_rate": 7.200586722868932e-05, "loss": 0.0041, "step": 38720 }, { "grad_norm": 0.08351334929466248, "learning_rate": 7.199101885440088e-05, "loss": 0.0043, "step": 38730 }, { "grad_norm": 0.07302531599998474, "learning_rate": 7.197616807520771e-05, "loss": 0.0047, "step": 38740 }, { "grad_norm": 0.10124434530735016, "learning_rate": 7.196131489273381e-05, "loss": 0.0049, "step": 38750 }, { "grad_norm": 0.0899650976061821, "learning_rate": 7.194645930860355e-05, "loss": 0.0052, "step": 38760 }, { "grad_norm": 0.07327993959188461, "learning_rate": 7.19316013244415e-05, "loss": 0.0048, "step": 38770 }, { "grad_norm": 0.12617945671081543, "learning_rate": 7.191674094187248e-05, "loss": 0.0057, "step": 38780 }, { "grad_norm": 0.08854375034570694, "learning_rate": 7.190187816252165e-05, "loss": 0.0053, "step": 38790 }, { "grad_norm": 0.09064297378063202, "learning_rate": 7.188701298801435e-05, "loss": 0.0046, "step": 38800 }, { "grad_norm": 0.09162948280572891, "learning_rate": 7.18721454199762e-05, "loss": 0.0041, "step": 38810 }, { "grad_norm": 0.09302850812673569, "learning_rate": 7.185727546003312e-05, "loss": 0.005, "step": 38820 }, { "grad_norm": 0.07495076954364777, "learning_rate": 7.184240310981126e-05, "loss": 0.0059, "step": 38830 }, { "grad_norm": 0.08477535843849182, "learning_rate": 7.182752837093704e-05, "loss": 0.0044, "step": 38840 }, { "grad_norm": 0.060470275580883026, "learning_rate": 7.181265124503711e-05, "loss": 0.0046, "step": 38850 }, { "grad_norm": 0.06482609361410141, "learning_rate": 7.179777173373847e-05, "loss": 0.0048, "step": 38860 }, { "grad_norm": 0.08988083899021149, "learning_rate": 7.178288983866826e-05, "loss": 0.0039, "step": 38870 }, { "grad_norm": 0.054382503032684326, "learning_rate": 7.176800556145397e-05, "loss": 0.004, "step": 38880 }, { "grad_norm": 0.09573578834533691, "learning_rate": 7.175311890372334e-05, "loss": 0.0043, "step": 38890 }, { "grad_norm": 0.06498923152685165, "learning_rate": 7.17382298671043e-05, "loss": 0.0053, "step": 38900 }, { "grad_norm": 0.07588472962379456, "learning_rate": 7.172333845322515e-05, "loss": 0.0047, "step": 38910 }, { "grad_norm": 0.08555347472429276, "learning_rate": 7.170844466371436e-05, "loss": 0.0044, "step": 38920 }, { "grad_norm": 0.0930897518992424, "learning_rate": 7.16935485002007e-05, "loss": 0.0044, "step": 38930 }, { "grad_norm": 0.08022043853998184, "learning_rate": 7.167864996431319e-05, "loss": 0.0042, "step": 38940 }, { "grad_norm": 0.09456346184015274, "learning_rate": 7.166374905768111e-05, "loss": 0.004, "step": 38950 }, { "grad_norm": 0.09164950251579285, "learning_rate": 7.164884578193401e-05, "loss": 0.0045, "step": 38960 }, { "grad_norm": 0.0704352855682373, "learning_rate": 7.163394013870168e-05, "loss": 0.005, "step": 38970 }, { "grad_norm": 0.07977528870105743, "learning_rate": 7.161903212961419e-05, "loss": 0.0052, "step": 38980 }, { "grad_norm": 0.0737413689494133, "learning_rate": 7.160412175630183e-05, "loss": 0.0054, "step": 38990 }, { "grad_norm": 0.11902008205652237, "learning_rate": 7.158920902039521e-05, "loss": 0.005, "step": 39000 }, { "grad_norm": 0.11836770921945572, "learning_rate": 7.157429392352514e-05, "loss": 0.0068, "step": 39010 }, { "grad_norm": 0.0971200168132782, "learning_rate": 7.155937646732275e-05, "loss": 0.005, "step": 39020 }, { "grad_norm": 0.152443066239357, "learning_rate": 7.154445665341933e-05, "loss": 0.0046, "step": 39030 }, { "grad_norm": 0.08915545791387558, "learning_rate": 7.152953448344654e-05, "loss": 0.0041, "step": 39040 }, { "grad_norm": 0.052090276032686234, "learning_rate": 7.151460995903624e-05, "loss": 0.0038, "step": 39050 }, { "grad_norm": 0.07348229736089706, "learning_rate": 7.149968308182052e-05, "loss": 0.0046, "step": 39060 }, { "grad_norm": 0.0819687694311142, "learning_rate": 7.14847538534318e-05, "loss": 0.0058, "step": 39070 }, { "grad_norm": 0.05649803951382637, "learning_rate": 7.14698222755027e-05, "loss": 0.0054, "step": 39080 }, { "grad_norm": 0.07278842478990555, "learning_rate": 7.14548883496661e-05, "loss": 0.0047, "step": 39090 }, { "grad_norm": 0.0670071542263031, "learning_rate": 7.143995207755517e-05, "loss": 0.0049, "step": 39100 }, { "grad_norm": 0.07995501160621643, "learning_rate": 7.142501346080333e-05, "loss": 0.0051, "step": 39110 }, { "grad_norm": 0.08871123939752579, "learning_rate": 7.141007250104421e-05, "loss": 0.0063, "step": 39120 }, { "grad_norm": 0.09436548501253128, "learning_rate": 7.139512919991176e-05, "loss": 0.0051, "step": 39130 }, { "grad_norm": 0.06516768783330917, "learning_rate": 7.138018355904014e-05, "loss": 0.005, "step": 39140 }, { "grad_norm": 0.11300558596849442, "learning_rate": 7.13652355800638e-05, "loss": 0.0045, "step": 39150 }, { "grad_norm": 0.07764315605163574, "learning_rate": 7.135028526461743e-05, "loss": 0.005, "step": 39160 }, { "grad_norm": 0.10049328953027725, "learning_rate": 7.133533261433594e-05, "loss": 0.0043, "step": 39170 }, { "grad_norm": 0.11217191070318222, "learning_rate": 7.132037763085457e-05, "loss": 0.0046, "step": 39180 }, { "grad_norm": 0.07334350049495697, "learning_rate": 7.130542031580875e-05, "loss": 0.0042, "step": 39190 }, { "grad_norm": 0.07362296432256699, "learning_rate": 7.12904606708342e-05, "loss": 0.0044, "step": 39200 }, { "grad_norm": 0.07154306769371033, "learning_rate": 7.127549869756687e-05, "loss": 0.0044, "step": 39210 }, { "grad_norm": 0.049095381051301956, "learning_rate": 7.126053439764299e-05, "loss": 0.0039, "step": 39220 }, { "grad_norm": 0.08101421594619751, "learning_rate": 7.124556777269904e-05, "loss": 0.0043, "step": 39230 }, { "grad_norm": 0.09385495632886887, "learning_rate": 7.123059882437174e-05, "loss": 0.0037, "step": 39240 }, { "grad_norm": 0.06875644624233246, "learning_rate": 7.121562755429807e-05, "loss": 0.0045, "step": 39250 }, { "grad_norm": 0.06506960839033127, "learning_rate": 7.120065396411528e-05, "loss": 0.004, "step": 39260 }, { "grad_norm": 0.06746066361665726, "learning_rate": 7.118567805546084e-05, "loss": 0.005, "step": 39270 }, { "grad_norm": 0.05517992749810219, "learning_rate": 7.117069982997248e-05, "loss": 0.0044, "step": 39280 }, { "grad_norm": 0.07175634056329727, "learning_rate": 7.115571928928825e-05, "loss": 0.0046, "step": 39290 }, { "grad_norm": 0.10016866773366928, "learning_rate": 7.114073643504635e-05, "loss": 0.0055, "step": 39300 }, { "grad_norm": 0.06713707000017166, "learning_rate": 7.11257512688853e-05, "loss": 0.0049, "step": 39310 }, { "grad_norm": 0.08550155162811279, "learning_rate": 7.111076379244384e-05, "loss": 0.004, "step": 39320 }, { "grad_norm": 0.07638116180896759, "learning_rate": 7.109577400736101e-05, "loss": 0.0045, "step": 39330 }, { "grad_norm": 0.06322094798088074, "learning_rate": 7.108078191527605e-05, "loss": 0.0048, "step": 39340 }, { "grad_norm": 0.07931078225374222, "learning_rate": 7.106578751782847e-05, "loss": 0.0047, "step": 39350 }, { "grad_norm": 0.06314974278211594, "learning_rate": 7.105079081665803e-05, "loss": 0.0042, "step": 39360 }, { "grad_norm": 0.08078794926404953, "learning_rate": 7.103579181340476e-05, "loss": 0.0044, "step": 39370 }, { "grad_norm": 0.07013750821352005, "learning_rate": 7.102079050970893e-05, "loss": 0.0045, "step": 39380 }, { "grad_norm": 0.08756130188703537, "learning_rate": 7.100578690721105e-05, "loss": 0.0041, "step": 39390 }, { "grad_norm": 0.04761231690645218, "learning_rate": 7.09907810075519e-05, "loss": 0.0041, "step": 39400 }, { "grad_norm": 0.09001222997903824, "learning_rate": 7.097577281237249e-05, "loss": 0.0052, "step": 39410 }, { "grad_norm": 0.06547056883573532, "learning_rate": 7.09607623233141e-05, "loss": 0.0052, "step": 39420 }, { "grad_norm": 0.12689535319805145, "learning_rate": 7.094574954201827e-05, "loss": 0.0046, "step": 39430 }, { "grad_norm": 0.08731696754693985, "learning_rate": 7.093073447012675e-05, "loss": 0.0063, "step": 39440 }, { "grad_norm": 0.09869043529033661, "learning_rate": 7.09157171092816e-05, "loss": 0.0055, "step": 39450 }, { "grad_norm": 0.09395278245210648, "learning_rate": 7.090069746112504e-05, "loss": 0.0051, "step": 39460 }, { "grad_norm": 0.05883113667368889, "learning_rate": 7.088567552729965e-05, "loss": 0.0048, "step": 39470 }, { "grad_norm": 0.09104781597852707, "learning_rate": 7.087065130944818e-05, "loss": 0.0053, "step": 39480 }, { "grad_norm": 0.07684928923845291, "learning_rate": 7.085562480921366e-05, "loss": 0.0049, "step": 39490 }, { "grad_norm": 0.08985322713851929, "learning_rate": 7.084059602823937e-05, "loss": 0.0066, "step": 39500 }, { "grad_norm": 0.08208915591239929, "learning_rate": 7.082556496816882e-05, "loss": 0.0049, "step": 39510 }, { "grad_norm": 0.08299332857131958, "learning_rate": 7.081053163064582e-05, "loss": 0.0042, "step": 39520 }, { "grad_norm": 0.08516845852136612, "learning_rate": 7.079549601731434e-05, "loss": 0.0052, "step": 39530 }, { "grad_norm": 0.11323387175798416, "learning_rate": 7.07804581298187e-05, "loss": 0.0054, "step": 39540 }, { "grad_norm": 0.11524172872304916, "learning_rate": 7.07654179698034e-05, "loss": 0.0062, "step": 39550 }, { "grad_norm": 0.10735184699296951, "learning_rate": 7.075037553891321e-05, "loss": 0.0061, "step": 39560 }, { "grad_norm": 0.09842605143785477, "learning_rate": 7.073533083879315e-05, "loss": 0.0065, "step": 39570 }, { "grad_norm": 0.09668498486280441, "learning_rate": 7.072028387108849e-05, "loss": 0.0079, "step": 39580 }, { "grad_norm": 0.07165498286485672, "learning_rate": 7.070523463744472e-05, "loss": 0.0063, "step": 39590 }, { "grad_norm": 0.1076110452413559, "learning_rate": 7.069018313950763e-05, "loss": 0.0061, "step": 39600 }, { "grad_norm": 0.09294050186872482, "learning_rate": 7.067512937892322e-05, "loss": 0.0049, "step": 39610 }, { "grad_norm": 0.1497235894203186, "learning_rate": 7.066007335733775e-05, "loss": 0.0048, "step": 39620 }, { "grad_norm": 0.10247289389371872, "learning_rate": 7.064501507639772e-05, "loss": 0.0053, "step": 39630 }, { "grad_norm": 0.0986848697066307, "learning_rate": 7.062995453774987e-05, "loss": 0.0052, "step": 39640 }, { "grad_norm": 0.10730491578578949, "learning_rate": 7.061489174304121e-05, "loss": 0.0051, "step": 39650 }, { "grad_norm": 0.11269311606884003, "learning_rate": 7.0599826693919e-05, "loss": 0.0049, "step": 39660 }, { "grad_norm": 0.0889938473701477, "learning_rate": 7.05847593920307e-05, "loss": 0.0051, "step": 39670 }, { "grad_norm": 0.1320664882659912, "learning_rate": 7.056968983902406e-05, "loss": 0.0053, "step": 39680 }, { "grad_norm": 0.10340393334627151, "learning_rate": 7.055461803654706e-05, "loss": 0.0063, "step": 39690 }, { "grad_norm": 0.09576692432165146, "learning_rate": 7.053954398624794e-05, "loss": 0.005, "step": 39700 }, { "grad_norm": 0.08370181173086166, "learning_rate": 7.052446768977518e-05, "loss": 0.0041, "step": 39710 }, { "grad_norm": 0.06745652109384537, "learning_rate": 7.050938914877748e-05, "loss": 0.0045, "step": 39720 }, { "grad_norm": 0.07719653844833374, "learning_rate": 7.049430836490382e-05, "loss": 0.0042, "step": 39730 }, { "grad_norm": 0.06366168707609177, "learning_rate": 7.047922533980342e-05, "loss": 0.004, "step": 39740 }, { "grad_norm": 0.06538037955760956, "learning_rate": 7.046414007512571e-05, "loss": 0.0044, "step": 39750 }, { "grad_norm": 0.0930742472410202, "learning_rate": 7.044905257252042e-05, "loss": 0.0047, "step": 39760 }, { "grad_norm": 0.18682706356048584, "learning_rate": 7.04339628336375e-05, "loss": 0.0053, "step": 39770 }, { "grad_norm": 0.06335584819316864, "learning_rate": 7.041887086012711e-05, "loss": 0.0055, "step": 39780 }, { "grad_norm": 0.08498547226190567, "learning_rate": 7.040377665363969e-05, "loss": 0.0055, "step": 39790 }, { "grad_norm": 0.10889891535043716, "learning_rate": 7.038868021582594e-05, "loss": 0.006, "step": 39800 }, { "grad_norm": 0.09223891794681549, "learning_rate": 7.037358154833679e-05, "loss": 0.0058, "step": 39810 }, { "grad_norm": 0.09265341609716415, "learning_rate": 7.035848065282339e-05, "loss": 0.0057, "step": 39820 }, { "grad_norm": 0.11500921100378036, "learning_rate": 7.034337753093714e-05, "loss": 0.0057, "step": 39830 }, { "grad_norm": 0.11461472511291504, "learning_rate": 7.032827218432972e-05, "loss": 0.0054, "step": 39840 }, { "grad_norm": 0.09022970497608185, "learning_rate": 7.031316461465302e-05, "loss": 0.0057, "step": 39850 }, { "grad_norm": 0.08822586387395859, "learning_rate": 7.029805482355915e-05, "loss": 0.0055, "step": 39860 }, { "grad_norm": 0.0675630122423172, "learning_rate": 7.028294281270055e-05, "loss": 0.0063, "step": 39870 }, { "grad_norm": 0.08569236099720001, "learning_rate": 7.026782858372982e-05, "loss": 0.0047, "step": 39880 }, { "grad_norm": 0.09572268277406693, "learning_rate": 7.025271213829982e-05, "loss": 0.0052, "step": 39890 }, { "grad_norm": 0.13768413662910461, "learning_rate": 7.023759347806366e-05, "loss": 0.0049, "step": 39900 }, { "grad_norm": 0.09733065962791443, "learning_rate": 7.022247260467472e-05, "loss": 0.0056, "step": 39910 }, { "grad_norm": 0.07114840298891068, "learning_rate": 7.020734951978658e-05, "loss": 0.0043, "step": 39920 }, { "grad_norm": 0.08135844022035599, "learning_rate": 7.019222422505307e-05, "loss": 0.0071, "step": 39930 }, { "grad_norm": 0.09248718619346619, "learning_rate": 7.017709672212827e-05, "loss": 0.0055, "step": 39940 }, { "grad_norm": 0.09234029799699783, "learning_rate": 7.016196701266652e-05, "loss": 0.0056, "step": 39950 }, { "grad_norm": 0.0894460454583168, "learning_rate": 7.014683509832238e-05, "loss": 0.0046, "step": 39960 }, { "grad_norm": 0.06506828218698502, "learning_rate": 7.013170098075063e-05, "loss": 0.0067, "step": 39970 }, { "grad_norm": 0.06561338901519775, "learning_rate": 7.011656466160632e-05, "loss": 0.0044, "step": 39980 }, { "grad_norm": 0.06402655690908432, "learning_rate": 7.010142614254475e-05, "loss": 0.0049, "step": 39990 }, { "grad_norm": 0.07015358656644821, "learning_rate": 7.008628542522147e-05, "loss": 0.0049, "step": 40000 }, { "grad_norm": 0.07756106555461884, "learning_rate": 7.007114251129218e-05, "loss": 0.0048, "step": 40010 }, { "grad_norm": 0.07981322705745697, "learning_rate": 7.005599740241293e-05, "loss": 0.0049, "step": 40020 }, { "grad_norm": 0.09230601787567139, "learning_rate": 7.004085010023996e-05, "loss": 0.0042, "step": 40030 }, { "grad_norm": 0.057835355401039124, "learning_rate": 7.002570060642976e-05, "loss": 0.004, "step": 40040 }, { "grad_norm": 0.06412016600370407, "learning_rate": 7.001054892263903e-05, "loss": 0.0047, "step": 40050 }, { "grad_norm": 0.08437038213014603, "learning_rate": 6.999539505052477e-05, "loss": 0.0042, "step": 40060 }, { "grad_norm": 0.08389189839363098, "learning_rate": 6.998023899174418e-05, "loss": 0.0043, "step": 40070 }, { "grad_norm": 0.08027121424674988, "learning_rate": 6.996508074795467e-05, "loss": 0.0047, "step": 40080 }, { "grad_norm": 0.07157783955335617, "learning_rate": 6.994992032081396e-05, "loss": 0.0042, "step": 40090 }, { "grad_norm": 0.0935213565826416, "learning_rate": 6.993475771197995e-05, "loss": 0.0054, "step": 40100 }, { "grad_norm": 0.08687661588191986, "learning_rate": 6.991959292311082e-05, "loss": 0.0048, "step": 40110 }, { "grad_norm": 0.09000471979379654, "learning_rate": 6.990442595586495e-05, "loss": 0.0043, "step": 40120 }, { "grad_norm": 0.09251353144645691, "learning_rate": 6.988925681190098e-05, "loss": 0.0045, "step": 40130 }, { "grad_norm": 0.06335105001926422, "learning_rate": 6.987408549287778e-05, "loss": 0.0037, "step": 40140 }, { "grad_norm": 0.08662662655115128, "learning_rate": 6.985891200045449e-05, "loss": 0.0049, "step": 40150 }, { "grad_norm": 0.06618640571832657, "learning_rate": 6.984373633629045e-05, "loss": 0.0039, "step": 40160 }, { "grad_norm": 0.07845745235681534, "learning_rate": 6.98285585020452e-05, "loss": 0.0052, "step": 40170 }, { "grad_norm": 0.09140586107969284, "learning_rate": 6.981337849937864e-05, "loss": 0.0041, "step": 40180 }, { "grad_norm": 0.07981383800506592, "learning_rate": 6.979819632995078e-05, "loss": 0.0052, "step": 40190 }, { "grad_norm": 0.0951743870973587, "learning_rate": 6.978301199542193e-05, "loss": 0.0042, "step": 40200 }, { "grad_norm": 0.07405754178762436, "learning_rate": 6.976782549745263e-05, "loss": 0.0049, "step": 40210 }, { "grad_norm": 0.09251152724027634, "learning_rate": 6.975263683770367e-05, "loss": 0.0053, "step": 40220 }, { "grad_norm": 0.0663965567946434, "learning_rate": 6.973744601783604e-05, "loss": 0.0046, "step": 40230 }, { "grad_norm": 0.07773632556200027, "learning_rate": 6.972225303951097e-05, "loss": 0.005, "step": 40240 }, { "grad_norm": 0.07507554441690445, "learning_rate": 6.970705790438998e-05, "loss": 0.0043, "step": 40250 }, { "grad_norm": 0.09648009389638901, "learning_rate": 6.969186061413477e-05, "loss": 0.0042, "step": 40260 }, { "grad_norm": 0.0758160725235939, "learning_rate": 6.967666117040727e-05, "loss": 0.0039, "step": 40270 }, { "grad_norm": 0.0810534656047821, "learning_rate": 6.966145957486972e-05, "loss": 0.0039, "step": 40280 }, { "grad_norm": 0.07139324396848679, "learning_rate": 6.964625582918449e-05, "loss": 0.0044, "step": 40290 }, { "grad_norm": 0.08678939938545227, "learning_rate": 6.963104993501425e-05, "loss": 0.0048, "step": 40300 }, { "grad_norm": 0.07258744537830353, "learning_rate": 6.961584189402192e-05, "loss": 0.0045, "step": 40310 }, { "grad_norm": 0.09950601309537888, "learning_rate": 6.96006317078706e-05, "loss": 0.0048, "step": 40320 }, { "grad_norm": 0.08335830271244049, "learning_rate": 6.958541937822367e-05, "loss": 0.0048, "step": 40330 }, { "grad_norm": 0.06841398030519485, "learning_rate": 6.957020490674473e-05, "loss": 0.0044, "step": 40340 }, { "grad_norm": 0.054441358894109726, "learning_rate": 6.95549882950976e-05, "loss": 0.0047, "step": 40350 }, { "grad_norm": 0.07959412038326263, "learning_rate": 6.953976954494635e-05, "loss": 0.0051, "step": 40360 }, { "grad_norm": 0.08095230162143707, "learning_rate": 6.952454865795528e-05, "loss": 0.005, "step": 40370 }, { "grad_norm": 0.10473547130823135, "learning_rate": 6.950932563578892e-05, "loss": 0.004, "step": 40380 }, { "grad_norm": 0.09063456207513809, "learning_rate": 6.949410048011206e-05, "loss": 0.0046, "step": 40390 }, { "grad_norm": 0.07740643620491028, "learning_rate": 6.947887319258966e-05, "loss": 0.0044, "step": 40400 }, { "grad_norm": 0.08492928743362427, "learning_rate": 6.946364377488696e-05, "loss": 0.0052, "step": 40410 }, { "grad_norm": 0.07997629046440125, "learning_rate": 6.944841222866947e-05, "loss": 0.0051, "step": 40420 }, { "grad_norm": 0.058388352394104004, "learning_rate": 6.943317855560284e-05, "loss": 0.0051, "step": 40430 }, { "grad_norm": 0.08236440271139145, "learning_rate": 6.941794275735302e-05, "loss": 0.005, "step": 40440 }, { "grad_norm": 0.0869407206773758, "learning_rate": 6.94027048355862e-05, "loss": 0.0049, "step": 40450 }, { "grad_norm": 0.09299766272306442, "learning_rate": 6.938746479196871e-05, "loss": 0.006, "step": 40460 }, { "grad_norm": 0.07496507465839386, "learning_rate": 6.937222262816724e-05, "loss": 0.0052, "step": 40470 }, { "grad_norm": 0.059770260006189346, "learning_rate": 6.935697834584865e-05, "loss": 0.004, "step": 40480 }, { "grad_norm": 0.07009512931108475, "learning_rate": 6.934173194667998e-05, "loss": 0.004, "step": 40490 }, { "grad_norm": 0.06882408261299133, "learning_rate": 6.93264834323286e-05, "loss": 0.0043, "step": 40500 }, { "grad_norm": 0.06011546775698662, "learning_rate": 6.931123280446206e-05, "loss": 0.0045, "step": 40510 }, { "grad_norm": 0.07721716165542603, "learning_rate": 6.929598006474811e-05, "loss": 0.004, "step": 40520 }, { "grad_norm": 0.07986035197973251, "learning_rate": 6.92807252148548e-05, "loss": 0.0047, "step": 40530 }, { "grad_norm": 0.07008884102106094, "learning_rate": 6.926546825645037e-05, "loss": 0.0042, "step": 40540 }, { "grad_norm": 0.06729087978601456, "learning_rate": 6.92502091912033e-05, "loss": 0.0043, "step": 40550 }, { "grad_norm": 0.08110017329454422, "learning_rate": 6.923494802078229e-05, "loss": 0.0042, "step": 40560 }, { "grad_norm": 0.05935639142990112, "learning_rate": 6.921968474685628e-05, "loss": 0.0051, "step": 40570 }, { "grad_norm": 0.08557356148958206, "learning_rate": 6.920441937109445e-05, "loss": 0.005, "step": 40580 }, { "grad_norm": 0.08853638172149658, "learning_rate": 6.91891518951662e-05, "loss": 0.0044, "step": 40590 }, { "grad_norm": 0.0677284300327301, "learning_rate": 6.917388232074114e-05, "loss": 0.0048, "step": 40600 }, { "grad_norm": 0.07196684181690216, "learning_rate": 6.915861064948914e-05, "loss": 0.0042, "step": 40610 }, { "grad_norm": 0.08874333649873734, "learning_rate": 6.914333688308025e-05, "loss": 0.0056, "step": 40620 }, { "grad_norm": 0.06601930409669876, "learning_rate": 6.912806102318487e-05, "loss": 0.0048, "step": 40630 }, { "grad_norm": 0.07260828465223312, "learning_rate": 6.911278307147347e-05, "loss": 0.0041, "step": 40640 }, { "grad_norm": 0.08312512189149857, "learning_rate": 6.909750302961684e-05, "loss": 0.0043, "step": 40650 }, { "grad_norm": 0.09464122354984283, "learning_rate": 6.9082220899286e-05, "loss": 0.0033, "step": 40660 }, { "grad_norm": 0.0810057744383812, "learning_rate": 6.906693668215219e-05, "loss": 0.0045, "step": 40670 }, { "grad_norm": 0.0652090460062027, "learning_rate": 6.905165037988683e-05, "loss": 0.0038, "step": 40680 }, { "grad_norm": 0.07699815928936005, "learning_rate": 6.903636199416164e-05, "loss": 0.004, "step": 40690 }, { "grad_norm": 0.0683513730764389, "learning_rate": 6.902107152664851e-05, "loss": 0.0048, "step": 40700 }, { "grad_norm": 0.0749039575457573, "learning_rate": 6.90057789790196e-05, "loss": 0.0047, "step": 40710 }, { "grad_norm": 0.06258521974086761, "learning_rate": 6.899048435294728e-05, "loss": 0.0043, "step": 40720 }, { "grad_norm": 0.09262017905712128, "learning_rate": 6.897518765010415e-05, "loss": 0.0047, "step": 40730 }, { "grad_norm": 0.05871754139661789, "learning_rate": 6.895988887216302e-05, "loss": 0.0044, "step": 40740 }, { "grad_norm": 0.06468525528907776, "learning_rate": 6.894458802079694e-05, "loss": 0.0046, "step": 40750 }, { "grad_norm": 0.13056500256061554, "learning_rate": 6.892928509767922e-05, "loss": 0.0046, "step": 40760 }, { "grad_norm": 0.10343821346759796, "learning_rate": 6.891398010448333e-05, "loss": 0.0069, "step": 40770 }, { "grad_norm": 0.07717292010784149, "learning_rate": 6.8898673042883e-05, "loss": 0.0055, "step": 40780 }, { "grad_norm": 0.11607875674962997, "learning_rate": 6.888336391455222e-05, "loss": 0.0062, "step": 40790 }, { "grad_norm": 0.10764812678098679, "learning_rate": 6.886805272116513e-05, "loss": 0.0052, "step": 40800 }, { "grad_norm": 0.0856308862566948, "learning_rate": 6.885273946439617e-05, "loss": 0.005, "step": 40810 }, { "grad_norm": 0.11200689524412155, "learning_rate": 6.883742414591998e-05, "loss": 0.0046, "step": 40820 }, { "grad_norm": 0.08375940471887589, "learning_rate": 6.88221067674114e-05, "loss": 0.0049, "step": 40830 }, { "grad_norm": 0.06859447062015533, "learning_rate": 6.88067873305455e-05, "loss": 0.0043, "step": 40840 }, { "grad_norm": 0.06587094813585281, "learning_rate": 6.879146583699765e-05, "loss": 0.0048, "step": 40850 }, { "grad_norm": 0.14108790457248688, "learning_rate": 6.877614228844334e-05, "loss": 0.0057, "step": 40860 }, { "grad_norm": 0.11459188908338547, "learning_rate": 6.876081668655832e-05, "loss": 0.0041, "step": 40870 }, { "grad_norm": 0.11562301218509674, "learning_rate": 6.874548903301861e-05, "loss": 0.0048, "step": 40880 }, { "grad_norm": 0.09509999305009842, "learning_rate": 6.873015932950039e-05, "loss": 0.0051, "step": 40890 }, { "grad_norm": 0.07572507858276367, "learning_rate": 6.871482757768012e-05, "loss": 0.0052, "step": 40900 }, { "grad_norm": 0.06814680248498917, "learning_rate": 6.869949377923442e-05, "loss": 0.0047, "step": 40910 }, { "grad_norm": 0.10773779451847076, "learning_rate": 6.868415793584022e-05, "loss": 0.0047, "step": 40920 }, { "grad_norm": 0.09816776216030121, "learning_rate": 6.86688200491746e-05, "loss": 0.0044, "step": 40930 }, { "grad_norm": 0.07643044739961624, "learning_rate": 6.865348012091488e-05, "loss": 0.0044, "step": 40940 }, { "grad_norm": 0.06398042291402817, "learning_rate": 6.86381381527386e-05, "loss": 0.0045, "step": 40950 }, { "grad_norm": 0.08529924601316452, "learning_rate": 6.862279414632359e-05, "loss": 0.0039, "step": 40960 }, { "grad_norm": 0.07048294693231583, "learning_rate": 6.860744810334778e-05, "loss": 0.0043, "step": 40970 }, { "grad_norm": 0.09283669292926788, "learning_rate": 6.859210002548943e-05, "loss": 0.0036, "step": 40980 }, { "grad_norm": 0.07457323372364044, "learning_rate": 6.8576749914427e-05, "loss": 0.0044, "step": 40990 }, { "grad_norm": 0.08184906840324402, "learning_rate": 6.85613977718391e-05, "loss": 0.0057, "step": 41000 }, { "grad_norm": 0.09244559705257416, "learning_rate": 6.854604359940464e-05, "loss": 0.0052, "step": 41010 }, { "grad_norm": 0.06663653999567032, "learning_rate": 6.853068739880276e-05, "loss": 0.0046, "step": 41020 }, { "grad_norm": 0.06488503515720367, "learning_rate": 6.851532917171276e-05, "loss": 0.0053, "step": 41030 }, { "grad_norm": 0.06891791522502899, "learning_rate": 6.849996891981418e-05, "loss": 0.0044, "step": 41040 }, { "grad_norm": 0.12727737426757812, "learning_rate": 6.848460664478684e-05, "loss": 0.0054, "step": 41050 }, { "grad_norm": 0.15088333189487457, "learning_rate": 6.846924234831069e-05, "loss": 0.0064, "step": 41060 }, { "grad_norm": 0.0620097778737545, "learning_rate": 6.845387603206597e-05, "loss": 0.0047, "step": 41070 }, { "grad_norm": 0.05644894763827324, "learning_rate": 6.843850769773312e-05, "loss": 0.0037, "step": 41080 }, { "grad_norm": 0.05695208162069321, "learning_rate": 6.842313734699278e-05, "loss": 0.0046, "step": 41090 }, { "grad_norm": 0.05716612562537193, "learning_rate": 6.840776498152584e-05, "loss": 0.0044, "step": 41100 }, { "grad_norm": 0.07408593595027924, "learning_rate": 6.83923906030134e-05, "loss": 0.0042, "step": 41110 }, { "grad_norm": 0.08933824300765991, "learning_rate": 6.837701421313677e-05, "loss": 0.0047, "step": 41120 }, { "grad_norm": 0.07940635830163956, "learning_rate": 6.83616358135775e-05, "loss": 0.0055, "step": 41130 }, { "grad_norm": 0.07331209629774094, "learning_rate": 6.834625540601734e-05, "loss": 0.0037, "step": 41140 }, { "grad_norm": 0.09025488793849945, "learning_rate": 6.833087299213829e-05, "loss": 0.0049, "step": 41150 }, { "grad_norm": 0.06690405309200287, "learning_rate": 6.83154885736225e-05, "loss": 0.0048, "step": 41160 }, { "grad_norm": 0.07106394320726395, "learning_rate": 6.830010215215244e-05, "loss": 0.0053, "step": 41170 }, { "grad_norm": 0.07529225200414658, "learning_rate": 6.828471372941072e-05, "loss": 0.0042, "step": 41180 }, { "grad_norm": 0.06441269814968109, "learning_rate": 6.826932330708018e-05, "loss": 0.0055, "step": 41190 }, { "grad_norm": 0.0620594248175621, "learning_rate": 6.825393088684393e-05, "loss": 0.0046, "step": 41200 }, { "grad_norm": 0.07814541459083557, "learning_rate": 6.823853647038524e-05, "loss": 0.0055, "step": 41210 }, { "grad_norm": 0.08797821402549744, "learning_rate": 6.822314005938762e-05, "loss": 0.0041, "step": 41220 }, { "grad_norm": 0.05553248152136803, "learning_rate": 6.820774165553479e-05, "loss": 0.0041, "step": 41230 }, { "grad_norm": 0.12235547602176666, "learning_rate": 6.819234126051072e-05, "loss": 0.0046, "step": 41240 }, { "grad_norm": 0.07137929648160934, "learning_rate": 6.817693887599956e-05, "loss": 0.005, "step": 41250 }, { "grad_norm": 0.07791776955127716, "learning_rate": 6.816153450368568e-05, "loss": 0.0041, "step": 41260 }, { "grad_norm": 0.072683185338974, "learning_rate": 6.814612814525372e-05, "loss": 0.0047, "step": 41270 }, { "grad_norm": 0.05912552401423454, "learning_rate": 6.813071980238846e-05, "loss": 0.004, "step": 41280 }, { "grad_norm": 0.06206320971250534, "learning_rate": 6.811530947677492e-05, "loss": 0.0038, "step": 41290 }, { "grad_norm": 0.07212678343057632, "learning_rate": 6.809989717009839e-05, "loss": 0.0044, "step": 41300 }, { "grad_norm": 0.08110036700963974, "learning_rate": 6.808448288404431e-05, "loss": 0.0054, "step": 41310 }, { "grad_norm": 0.08976566046476364, "learning_rate": 6.806906662029838e-05, "loss": 0.005, "step": 41320 }, { "grad_norm": 0.06665235757827759, "learning_rate": 6.805364838054647e-05, "loss": 0.0043, "step": 41330 }, { "grad_norm": 0.07817550003528595, "learning_rate": 6.803822816647474e-05, "loss": 0.0038, "step": 41340 }, { "grad_norm": 0.09593270719051361, "learning_rate": 6.802280597976949e-05, "loss": 0.0066, "step": 41350 }, { "grad_norm": 0.07746823132038116, "learning_rate": 6.800738182211727e-05, "loss": 0.0037, "step": 41360 }, { "grad_norm": 0.06304735690355301, "learning_rate": 6.799195569520485e-05, "loss": 0.0044, "step": 41370 }, { "grad_norm": 0.06083888188004494, "learning_rate": 6.79765276007192e-05, "loss": 0.0044, "step": 41380 }, { "grad_norm": 0.07286440581083298, "learning_rate": 6.796109754034753e-05, "loss": 0.0038, "step": 41390 }, { "grad_norm": 0.08442754298448563, "learning_rate": 6.794566551577724e-05, "loss": 0.0052, "step": 41400 }, { "grad_norm": 0.09613193571567535, "learning_rate": 6.793023152869594e-05, "loss": 0.0054, "step": 41410 }, { "grad_norm": 0.07983354479074478, "learning_rate": 6.791479558079147e-05, "loss": 0.0054, "step": 41420 }, { "grad_norm": 0.07238148152828217, "learning_rate": 6.789935767375191e-05, "loss": 0.0041, "step": 41430 }, { "grad_norm": 0.07376430928707123, "learning_rate": 6.78839178092655e-05, "loss": 0.0046, "step": 41440 }, { "grad_norm": 0.1052551344037056, "learning_rate": 6.786847598902072e-05, "loss": 0.0055, "step": 41450 }, { "grad_norm": 0.08402921259403229, "learning_rate": 6.785303221470627e-05, "loss": 0.0044, "step": 41460 }, { "grad_norm": 0.07561084628105164, "learning_rate": 6.783758648801108e-05, "loss": 0.005, "step": 41470 }, { "grad_norm": 0.09224306792020798, "learning_rate": 6.782213881062422e-05, "loss": 0.0042, "step": 41480 }, { "grad_norm": 0.07248158752918243, "learning_rate": 6.780668918423508e-05, "loss": 0.0041, "step": 41490 }, { "grad_norm": 0.07998227328062057, "learning_rate": 6.779123761053317e-05, "loss": 0.0047, "step": 41500 }, { "grad_norm": 0.09057130664587021, "learning_rate": 6.777578409120826e-05, "loss": 0.0045, "step": 41510 }, { "grad_norm": 0.08221384882926941, "learning_rate": 6.776032862795033e-05, "loss": 0.0042, "step": 41520 }, { "grad_norm": 0.07536650449037552, "learning_rate": 6.774487122244956e-05, "loss": 0.004, "step": 41530 }, { "grad_norm": 0.09050600230693817, "learning_rate": 6.772941187639637e-05, "loss": 0.0037, "step": 41540 }, { "grad_norm": 0.07936299592256546, "learning_rate": 6.771395059148134e-05, "loss": 0.0045, "step": 41550 }, { "grad_norm": 0.10209564119577408, "learning_rate": 6.76984873693953e-05, "loss": 0.005, "step": 41560 }, { "grad_norm": 0.07224389910697937, "learning_rate": 6.768302221182931e-05, "loss": 0.0042, "step": 41570 }, { "grad_norm": 0.06943856179714203, "learning_rate": 6.766755512047457e-05, "loss": 0.0041, "step": 41580 }, { "grad_norm": 0.07630796730518341, "learning_rate": 6.765208609702259e-05, "loss": 0.0042, "step": 41590 }, { "grad_norm": 0.07643520832061768, "learning_rate": 6.763661514316499e-05, "loss": 0.0037, "step": 41600 }, { "grad_norm": 0.06320050358772278, "learning_rate": 6.76211422605937e-05, "loss": 0.0042, "step": 41610 }, { "grad_norm": 0.07009847462177277, "learning_rate": 6.760566745100076e-05, "loss": 0.0043, "step": 41620 }, { "grad_norm": 0.0636695921421051, "learning_rate": 6.75901907160785e-05, "loss": 0.0038, "step": 41630 }, { "grad_norm": 0.08736097812652588, "learning_rate": 6.757471205751943e-05, "loss": 0.0037, "step": 41640 }, { "grad_norm": 0.07711253315210342, "learning_rate": 6.75592314770163e-05, "loss": 0.0049, "step": 41650 }, { "grad_norm": 0.07921221852302551, "learning_rate": 6.7543748976262e-05, "loss": 0.0061, "step": 41660 }, { "grad_norm": 0.113533154129982, "learning_rate": 6.752826455694968e-05, "loss": 0.0044, "step": 41670 }, { "grad_norm": 0.09609144926071167, "learning_rate": 6.751277822077271e-05, "loss": 0.004, "step": 41680 }, { "grad_norm": 0.06652463227510452, "learning_rate": 6.749728996942465e-05, "loss": 0.0048, "step": 41690 }, { "grad_norm": 0.09259085357189178, "learning_rate": 6.748179980459924e-05, "loss": 0.0042, "step": 41700 }, { "grad_norm": 0.06723839789628983, "learning_rate": 6.746630772799052e-05, "loss": 0.0037, "step": 41710 }, { "grad_norm": 0.0650094747543335, "learning_rate": 6.745081374129262e-05, "loss": 0.0036, "step": 41720 }, { "grad_norm": 0.09261875599622726, "learning_rate": 6.743531784619998e-05, "loss": 0.004, "step": 41730 }, { "grad_norm": 0.07041493058204651, "learning_rate": 6.741982004440719e-05, "loss": 0.0037, "step": 41740 }, { "grad_norm": 0.0691823735833168, "learning_rate": 6.740432033760907e-05, "loss": 0.0037, "step": 41750 }, { "grad_norm": 0.06775230169296265, "learning_rate": 6.738881872750066e-05, "loss": 0.0044, "step": 41760 }, { "grad_norm": 0.05592409521341324, "learning_rate": 6.737331521577715e-05, "loss": 0.004, "step": 41770 }, { "grad_norm": 0.12896278500556946, "learning_rate": 6.735780980413403e-05, "loss": 0.0043, "step": 41780 }, { "grad_norm": 0.08771562576293945, "learning_rate": 6.734230249426692e-05, "loss": 0.0038, "step": 41790 }, { "grad_norm": 0.06691482663154602, "learning_rate": 6.732679328787168e-05, "loss": 0.0043, "step": 41800 }, { "grad_norm": 0.09406394511461258, "learning_rate": 6.731128218664438e-05, "loss": 0.0036, "step": 41810 }, { "grad_norm": 0.07229869067668915, "learning_rate": 6.729576919228129e-05, "loss": 0.0042, "step": 41820 }, { "grad_norm": 0.054244283586740494, "learning_rate": 6.728025430647888e-05, "loss": 0.0039, "step": 41830 }, { "grad_norm": 0.10069834440946579, "learning_rate": 6.726473753093383e-05, "loss": 0.004, "step": 41840 }, { "grad_norm": 0.0981777161359787, "learning_rate": 6.724921886734305e-05, "loss": 0.0042, "step": 41850 }, { "grad_norm": 0.06279345601797104, "learning_rate": 6.723369831740363e-05, "loss": 0.0047, "step": 41860 }, { "grad_norm": 0.08373942971229553, "learning_rate": 6.721817588281288e-05, "loss": 0.0041, "step": 41870 }, { "grad_norm": 0.06954299658536911, "learning_rate": 6.720265156526828e-05, "loss": 0.0041, "step": 41880 }, { "grad_norm": 0.07795540243387222, "learning_rate": 6.718712536646758e-05, "loss": 0.0049, "step": 41890 }, { "grad_norm": 0.0784422904253006, "learning_rate": 6.71715972881087e-05, "loss": 0.004, "step": 41900 }, { "grad_norm": 0.08580942451953888, "learning_rate": 6.715606733188974e-05, "loss": 0.004, "step": 41910 }, { "grad_norm": 0.06167233735322952, "learning_rate": 6.714053549950909e-05, "loss": 0.0038, "step": 41920 }, { "grad_norm": 0.0599786713719368, "learning_rate": 6.71250017926652e-05, "loss": 0.0038, "step": 41930 }, { "grad_norm": 0.10168131440877914, "learning_rate": 6.71094662130569e-05, "loss": 0.0044, "step": 41940 }, { "grad_norm": 0.07022672146558762, "learning_rate": 6.709392876238307e-05, "loss": 0.0045, "step": 41950 }, { "grad_norm": 0.06185537204146385, "learning_rate": 6.707838944234289e-05, "loss": 0.0038, "step": 41960 }, { "grad_norm": 0.062059201300144196, "learning_rate": 6.706284825463574e-05, "loss": 0.0042, "step": 41970 }, { "grad_norm": 0.10867749899625778, "learning_rate": 6.704730520096114e-05, "loss": 0.0048, "step": 41980 }, { "grad_norm": 0.10360252857208252, "learning_rate": 6.703176028301888e-05, "loss": 0.0044, "step": 41990 }, { "grad_norm": 0.08817923814058304, "learning_rate": 6.701621350250892e-05, "loss": 0.0049, "step": 42000 }, { "grad_norm": 0.08940790593624115, "learning_rate": 6.700066486113144e-05, "loss": 0.0047, "step": 42010 }, { "grad_norm": 0.06509345769882202, "learning_rate": 6.69851143605868e-05, "loss": 0.0049, "step": 42020 }, { "grad_norm": 0.12822766602039337, "learning_rate": 6.696956200257561e-05, "loss": 0.0064, "step": 42030 }, { "grad_norm": 0.0635143294930458, "learning_rate": 6.695400778879863e-05, "loss": 0.0033, "step": 42040 }, { "grad_norm": 0.051144134253263474, "learning_rate": 6.693845172095683e-05, "loss": 0.0036, "step": 42050 }, { "grad_norm": 0.09713397920131683, "learning_rate": 6.692289380075142e-05, "loss": 0.0043, "step": 42060 }, { "grad_norm": 0.09570180624723434, "learning_rate": 6.690733402988379e-05, "loss": 0.0043, "step": 42070 }, { "grad_norm": 0.0873761773109436, "learning_rate": 6.689177241005553e-05, "loss": 0.005, "step": 42080 }, { "grad_norm": 0.05938880518078804, "learning_rate": 6.687620894296844e-05, "loss": 0.0044, "step": 42090 }, { "grad_norm": 0.08946742862462997, "learning_rate": 6.686064363032451e-05, "loss": 0.0045, "step": 42100 }, { "grad_norm": 0.08147815614938736, "learning_rate": 6.684507647382596e-05, "loss": 0.004, "step": 42110 }, { "grad_norm": 0.06661177426576614, "learning_rate": 6.682950747517513e-05, "loss": 0.0044, "step": 42120 }, { "grad_norm": 0.08693542331457138, "learning_rate": 6.68139366360747e-05, "loss": 0.0043, "step": 42130 }, { "grad_norm": 0.06304658204317093, "learning_rate": 6.679836395822744e-05, "loss": 0.0035, "step": 42140 }, { "grad_norm": 0.06917764246463776, "learning_rate": 6.678278944333633e-05, "loss": 0.0044, "step": 42150 }, { "grad_norm": 0.09447666257619858, "learning_rate": 6.676721309310462e-05, "loss": 0.0045, "step": 42160 }, { "grad_norm": 0.06395349651575089, "learning_rate": 6.67516349092357e-05, "loss": 0.0039, "step": 42170 }, { "grad_norm": 0.07916457206010818, "learning_rate": 6.673605489343315e-05, "loss": 0.0038, "step": 42180 }, { "grad_norm": 0.07405802607536316, "learning_rate": 6.672047304740081e-05, "loss": 0.0044, "step": 42190 }, { "grad_norm": 0.08742250502109528, "learning_rate": 6.670488937284268e-05, "loss": 0.0036, "step": 42200 }, { "grad_norm": 0.06874753534793854, "learning_rate": 6.668930387146294e-05, "loss": 0.0033, "step": 42210 }, { "grad_norm": 0.07336588948965073, "learning_rate": 6.667371654496605e-05, "loss": 0.0042, "step": 42220 }, { "grad_norm": 0.0670161023736, "learning_rate": 6.665812739505659e-05, "loss": 0.0043, "step": 42230 }, { "grad_norm": 0.0735597088932991, "learning_rate": 6.664253642343935e-05, "loss": 0.0041, "step": 42240 }, { "grad_norm": 0.088657908141613, "learning_rate": 6.662694363181935e-05, "loss": 0.0045, "step": 42250 }, { "grad_norm": 0.07469245791435242, "learning_rate": 6.66113490219018e-05, "loss": 0.0036, "step": 42260 }, { "grad_norm": 0.09726785868406296, "learning_rate": 6.659575259539207e-05, "loss": 0.0049, "step": 42270 }, { "grad_norm": 0.10220000892877579, "learning_rate": 6.65801543539958e-05, "loss": 0.0044, "step": 42280 }, { "grad_norm": 0.07695161551237106, "learning_rate": 6.65645542994188e-05, "loss": 0.0045, "step": 42290 }, { "grad_norm": 0.09444567561149597, "learning_rate": 6.654895243336702e-05, "loss": 0.0043, "step": 42300 }, { "grad_norm": 0.08131881803274155, "learning_rate": 6.65333487575467e-05, "loss": 0.0041, "step": 42310 }, { "grad_norm": 0.06872512400150299, "learning_rate": 6.651774327366422e-05, "loss": 0.004, "step": 42320 }, { "grad_norm": 0.07985895127058029, "learning_rate": 6.650213598342613e-05, "loss": 0.0051, "step": 42330 }, { "grad_norm": 0.07432287186384201, "learning_rate": 6.648652688853931e-05, "loss": 0.0042, "step": 42340 }, { "grad_norm": 0.07174152135848999, "learning_rate": 6.647091599071066e-05, "loss": 0.0045, "step": 42350 }, { "grad_norm": 0.08166220784187317, "learning_rate": 6.645530329164743e-05, "loss": 0.0044, "step": 42360 }, { "grad_norm": 0.07011031359434128, "learning_rate": 6.643968879305694e-05, "loss": 0.0047, "step": 42370 }, { "grad_norm": 0.05949746444821358, "learning_rate": 6.642407249664683e-05, "loss": 0.004, "step": 42380 }, { "grad_norm": 0.07954689115285873, "learning_rate": 6.640845440412483e-05, "loss": 0.0042, "step": 42390 }, { "grad_norm": 0.07933122664690018, "learning_rate": 6.639283451719893e-05, "loss": 0.0044, "step": 42400 }, { "grad_norm": 0.08771912008523941, "learning_rate": 6.637721283757729e-05, "loss": 0.0039, "step": 42410 }, { "grad_norm": 0.08802063018083572, "learning_rate": 6.636158936696828e-05, "loss": 0.0044, "step": 42420 }, { "grad_norm": 0.10210174322128296, "learning_rate": 6.634596410708047e-05, "loss": 0.0042, "step": 42430 }, { "grad_norm": 0.06389381736516953, "learning_rate": 6.633033705962257e-05, "loss": 0.0035, "step": 42440 }, { "grad_norm": 0.08132684975862503, "learning_rate": 6.631470822630359e-05, "loss": 0.0042, "step": 42450 }, { "grad_norm": 0.08739538490772247, "learning_rate": 6.629907760883263e-05, "loss": 0.004, "step": 42460 }, { "grad_norm": 0.07303425669670105, "learning_rate": 6.628344520891905e-05, "loss": 0.0036, "step": 42470 }, { "grad_norm": 0.07535474747419357, "learning_rate": 6.62678110282724e-05, "loss": 0.0038, "step": 42480 }, { "grad_norm": 0.08814558386802673, "learning_rate": 6.625217506860237e-05, "loss": 0.0039, "step": 42490 }, { "grad_norm": 0.06926263123750687, "learning_rate": 6.623653733161892e-05, "loss": 0.0037, "step": 42500 }, { "grad_norm": 0.08911826461553574, "learning_rate": 6.622089781903217e-05, "loss": 0.0047, "step": 42510 }, { "grad_norm": 0.07386519014835358, "learning_rate": 6.620525653255241e-05, "loss": 0.0057, "step": 42520 }, { "grad_norm": 0.0671580582857132, "learning_rate": 6.618961347389015e-05, "loss": 0.0042, "step": 42530 }, { "grad_norm": 0.08993075788021088, "learning_rate": 6.617396864475613e-05, "loss": 0.0044, "step": 42540 }, { "grad_norm": 0.0808316022157669, "learning_rate": 6.61583220468612e-05, "loss": 0.0041, "step": 42550 }, { "grad_norm": 0.08014440536499023, "learning_rate": 6.614267368191645e-05, "loss": 0.0035, "step": 42560 }, { "grad_norm": 0.06403159350156784, "learning_rate": 6.61270235516332e-05, "loss": 0.0045, "step": 42570 }, { "grad_norm": 0.062453996390104294, "learning_rate": 6.611137165772287e-05, "loss": 0.004, "step": 42580 }, { "grad_norm": 0.059707969427108765, "learning_rate": 6.60957180018972e-05, "loss": 0.0036, "step": 42590 }, { "grad_norm": 0.06510977447032928, "learning_rate": 6.608006258586797e-05, "loss": 0.0044, "step": 42600 }, { "grad_norm": 0.05095311626791954, "learning_rate": 6.606440541134731e-05, "loss": 0.0037, "step": 42610 }, { "grad_norm": 0.07099560648202896, "learning_rate": 6.60487464800474e-05, "loss": 0.0034, "step": 42620 }, { "grad_norm": 0.06656499952077866, "learning_rate": 6.603308579368071e-05, "loss": 0.0044, "step": 42630 }, { "grad_norm": 0.07851589471101761, "learning_rate": 6.601742335395987e-05, "loss": 0.0042, "step": 42640 }, { "grad_norm": 0.07632986456155777, "learning_rate": 6.600175916259769e-05, "loss": 0.0041, "step": 42650 }, { "grad_norm": 0.07488664239645004, "learning_rate": 6.598609322130718e-05, "loss": 0.0047, "step": 42660 }, { "grad_norm": 0.10965034365653992, "learning_rate": 6.597042553180155e-05, "loss": 0.0048, "step": 42670 }, { "grad_norm": 0.10175270587205887, "learning_rate": 6.59547560957942e-05, "loss": 0.0044, "step": 42680 }, { "grad_norm": 0.06215304881334305, "learning_rate": 6.59390849149987e-05, "loss": 0.0048, "step": 42690 }, { "grad_norm": 0.08386828750371933, "learning_rate": 6.592341199112886e-05, "loss": 0.0041, "step": 42700 }, { "grad_norm": 0.07584647834300995, "learning_rate": 6.590773732589861e-05, "loss": 0.0039, "step": 42710 }, { "grad_norm": 0.05323515832424164, "learning_rate": 6.589206092102213e-05, "loss": 0.0038, "step": 42720 }, { "grad_norm": 0.05376527085900307, "learning_rate": 6.587638277821375e-05, "loss": 0.0046, "step": 42730 }, { "grad_norm": 0.06026041880249977, "learning_rate": 6.586070289918803e-05, "loss": 0.004, "step": 42740 }, { "grad_norm": 0.06373970955610275, "learning_rate": 6.584502128565968e-05, "loss": 0.0048, "step": 42750 }, { "grad_norm": 0.05956791713833809, "learning_rate": 6.582933793934363e-05, "loss": 0.0041, "step": 42760 }, { "grad_norm": 0.08873078227043152, "learning_rate": 6.581365286195498e-05, "loss": 0.004, "step": 42770 }, { "grad_norm": 0.10285840928554535, "learning_rate": 6.579796605520903e-05, "loss": 0.0042, "step": 42780 }, { "grad_norm": 0.0760187953710556, "learning_rate": 6.578227752082127e-05, "loss": 0.0045, "step": 42790 }, { "grad_norm": 0.04126046970486641, "learning_rate": 6.576658726050735e-05, "loss": 0.0038, "step": 42800 }, { "grad_norm": 0.06034516543149948, "learning_rate": 6.575089527598316e-05, "loss": 0.0048, "step": 42810 }, { "grad_norm": 0.06912381947040558, "learning_rate": 6.573520156896476e-05, "loss": 0.0042, "step": 42820 }, { "grad_norm": 0.09220094233751297, "learning_rate": 6.571950614116835e-05, "loss": 0.0046, "step": 42830 }, { "grad_norm": 0.09071793407201767, "learning_rate": 6.570380899431039e-05, "loss": 0.0038, "step": 42840 }, { "grad_norm": 0.0852905884385109, "learning_rate": 6.568811013010749e-05, "loss": 0.0036, "step": 42850 }, { "grad_norm": 0.07431932538747787, "learning_rate": 6.567240955027645e-05, "loss": 0.0037, "step": 42860 }, { "grad_norm": 0.07248348742723465, "learning_rate": 6.565670725653427e-05, "loss": 0.004, "step": 42870 }, { "grad_norm": 0.10019822418689728, "learning_rate": 6.564100325059812e-05, "loss": 0.004, "step": 42880 }, { "grad_norm": 0.06812435388565063, "learning_rate": 6.562529753418538e-05, "loss": 0.0037, "step": 42890 }, { "grad_norm": 0.074043408036232, "learning_rate": 6.56095901090136e-05, "loss": 0.0047, "step": 42900 }, { "grad_norm": 0.06804030388593674, "learning_rate": 6.55938809768005e-05, "loss": 0.005, "step": 42910 }, { "grad_norm": 0.08303024619817734, "learning_rate": 6.557817013926402e-05, "loss": 0.0048, "step": 42920 }, { "grad_norm": 0.09747615456581116, "learning_rate": 6.556245759812227e-05, "loss": 0.0035, "step": 42930 }, { "grad_norm": 0.07860434055328369, "learning_rate": 6.554674335509357e-05, "loss": 0.0035, "step": 42940 }, { "grad_norm": 0.07405945658683777, "learning_rate": 6.553102741189638e-05, "loss": 0.0039, "step": 42950 }, { "grad_norm": 0.12520018219947815, "learning_rate": 6.55153097702494e-05, "loss": 0.0037, "step": 42960 }, { "grad_norm": 0.07002705335617065, "learning_rate": 6.549959043187146e-05, "loss": 0.0039, "step": 42970 }, { "grad_norm": 0.10002105683088303, "learning_rate": 6.54838693984816e-05, "loss": 0.0046, "step": 42980 }, { "grad_norm": 0.07295511662960052, "learning_rate": 6.546814667179909e-05, "loss": 0.0037, "step": 42990 }, { "grad_norm": 0.08025626093149185, "learning_rate": 6.545242225354328e-05, "loss": 0.0046, "step": 43000 }, { "grad_norm": 0.0850314348936081, "learning_rate": 6.543669614543383e-05, "loss": 0.0038, "step": 43010 }, { "grad_norm": 0.07076384127140045, "learning_rate": 6.542096834919049e-05, "loss": 0.0038, "step": 43020 }, { "grad_norm": 0.055318038910627365, "learning_rate": 6.540523886653324e-05, "loss": 0.0048, "step": 43030 }, { "grad_norm": 0.08218894898891449, "learning_rate": 6.538950769918222e-05, "loss": 0.005, "step": 43040 }, { "grad_norm": 0.08151675015687943, "learning_rate": 6.537377484885779e-05, "loss": 0.0055, "step": 43050 }, { "grad_norm": 0.07969742268323898, "learning_rate": 6.535804031728045e-05, "loss": 0.0044, "step": 43060 }, { "grad_norm": 0.06991486251354218, "learning_rate": 6.534230410617092e-05, "loss": 0.0048, "step": 43070 }, { "grad_norm": 0.09881509840488434, "learning_rate": 6.532656621725007e-05, "loss": 0.0041, "step": 43080 }, { "grad_norm": 0.09661611169576645, "learning_rate": 6.531082665223898e-05, "loss": 0.0041, "step": 43090 }, { "grad_norm": 0.08492619544267654, "learning_rate": 6.529508541285889e-05, "loss": 0.0036, "step": 43100 }, { "grad_norm": 0.06738897413015366, "learning_rate": 6.527934250083128e-05, "loss": 0.004, "step": 43110 }, { "grad_norm": 0.0682816132903099, "learning_rate": 6.526359791787775e-05, "loss": 0.004, "step": 43120 }, { "grad_norm": 0.07929640263319016, "learning_rate": 6.524785166572009e-05, "loss": 0.0045, "step": 43130 }, { "grad_norm": 0.08170577883720398, "learning_rate": 6.523210374608029e-05, "loss": 0.0043, "step": 43140 }, { "grad_norm": 0.08192149549722672, "learning_rate": 6.521635416068054e-05, "loss": 0.0035, "step": 43150 }, { "grad_norm": 0.0621301531791687, "learning_rate": 6.520060291124317e-05, "loss": 0.0041, "step": 43160 }, { "grad_norm": 0.07290739566087723, "learning_rate": 6.518484999949072e-05, "loss": 0.0048, "step": 43170 }, { "grad_norm": 0.06352949142456055, "learning_rate": 6.516909542714591e-05, "loss": 0.0047, "step": 43180 }, { "grad_norm": 0.0670972615480423, "learning_rate": 6.515333919593161e-05, "loss": 0.0039, "step": 43190 }, { "grad_norm": 0.08564222604036331, "learning_rate": 6.513758130757094e-05, "loss": 0.0037, "step": 43200 }, { "grad_norm": 0.06391309946775436, "learning_rate": 6.512182176378713e-05, "loss": 0.0042, "step": 43210 }, { "grad_norm": 0.0740933045744896, "learning_rate": 6.510606056630362e-05, "loss": 0.0042, "step": 43220 }, { "grad_norm": 0.06100131943821907, "learning_rate": 6.509029771684405e-05, "loss": 0.0037, "step": 43230 }, { "grad_norm": 0.11069217324256897, "learning_rate": 6.50745332171322e-05, "loss": 0.0062, "step": 43240 }, { "grad_norm": 0.061559539288282394, "learning_rate": 6.505876706889207e-05, "loss": 0.0039, "step": 43250 }, { "grad_norm": 0.09342855215072632, "learning_rate": 6.504299927384781e-05, "loss": 0.0042, "step": 43260 }, { "grad_norm": 0.07190978527069092, "learning_rate": 6.502722983372378e-05, "loss": 0.005, "step": 43270 }, { "grad_norm": 0.07538823038339615, "learning_rate": 6.501145875024446e-05, "loss": 0.004, "step": 43280 }, { "grad_norm": 0.06859472393989563, "learning_rate": 6.499568602513459e-05, "loss": 0.0043, "step": 43290 }, { "grad_norm": 0.07687506079673767, "learning_rate": 6.497991166011903e-05, "loss": 0.0049, "step": 43300 }, { "grad_norm": 0.09863699227571487, "learning_rate": 6.49641356569229e-05, "loss": 0.0038, "step": 43310 }, { "grad_norm": 0.08843179792165756, "learning_rate": 6.494835801727135e-05, "loss": 0.0041, "step": 43320 }, { "grad_norm": 0.07613302022218704, "learning_rate": 6.493257874288986e-05, "loss": 0.0041, "step": 43330 }, { "grad_norm": 0.08683402836322784, "learning_rate": 6.491679783550399e-05, "loss": 0.0038, "step": 43340 }, { "grad_norm": 0.07924696058034897, "learning_rate": 6.490101529683957e-05, "loss": 0.0046, "step": 43350 }, { "grad_norm": 0.1038660854101181, "learning_rate": 6.488523112862248e-05, "loss": 0.0039, "step": 43360 }, { "grad_norm": 0.06955911964178085, "learning_rate": 6.486944533257891e-05, "loss": 0.0043, "step": 43370 }, { "grad_norm": 0.0615641288459301, "learning_rate": 6.485365791043517e-05, "loss": 0.0041, "step": 43380 }, { "grad_norm": 0.07393352687358856, "learning_rate": 6.483786886391772e-05, "loss": 0.0047, "step": 43390 }, { "grad_norm": 0.08199340850114822, "learning_rate": 6.482207819475323e-05, "loss": 0.0043, "step": 43400 }, { "grad_norm": 0.06529033929109573, "learning_rate": 6.480628590466857e-05, "loss": 0.0052, "step": 43410 }, { "grad_norm": 0.07469425350427628, "learning_rate": 6.479049199539074e-05, "loss": 0.0043, "step": 43420 }, { "grad_norm": 0.06290144473314285, "learning_rate": 6.477469646864693e-05, "loss": 0.0042, "step": 43430 }, { "grad_norm": 0.07126937806606293, "learning_rate": 6.475889932616454e-05, "loss": 0.004, "step": 43440 }, { "grad_norm": 0.06913239508867264, "learning_rate": 6.474310056967111e-05, "loss": 0.0048, "step": 43450 }, { "grad_norm": 0.10686460137367249, "learning_rate": 6.472730020089437e-05, "loss": 0.0043, "step": 43460 }, { "grad_norm": 0.07554525882005692, "learning_rate": 6.471149822156222e-05, "loss": 0.0039, "step": 43470 }, { "grad_norm": 0.0999520942568779, "learning_rate": 6.469569463340275e-05, "loss": 0.0038, "step": 43480 }, { "grad_norm": 0.06070023030042648, "learning_rate": 6.467988943814418e-05, "loss": 0.0038, "step": 43490 }, { "grad_norm": 0.07143709063529968, "learning_rate": 6.4664082637515e-05, "loss": 0.0045, "step": 43500 }, { "grad_norm": 0.08229542523622513, "learning_rate": 6.464827423324377e-05, "loss": 0.0038, "step": 43510 }, { "grad_norm": 0.07938080281019211, "learning_rate": 6.463246422705931e-05, "loss": 0.0047, "step": 43520 }, { "grad_norm": 0.058541540056467056, "learning_rate": 6.461665262069055e-05, "loss": 0.0038, "step": 43530 }, { "grad_norm": 0.08778171986341476, "learning_rate": 6.460083941586665e-05, "loss": 0.0043, "step": 43540 }, { "grad_norm": 0.07182256877422333, "learning_rate": 6.45850246143169e-05, "loss": 0.0042, "step": 43550 }, { "grad_norm": 0.08175814896821976, "learning_rate": 6.456920821777077e-05, "loss": 0.0047, "step": 43560 }, { "grad_norm": 0.07383222132921219, "learning_rate": 6.455339022795796e-05, "loss": 0.0037, "step": 43570 }, { "grad_norm": 0.0894925445318222, "learning_rate": 6.453757064660825e-05, "loss": 0.0043, "step": 43580 }, { "grad_norm": 0.0813092514872551, "learning_rate": 6.452174947545169e-05, "loss": 0.0045, "step": 43590 }, { "grad_norm": 0.09186077862977982, "learning_rate": 6.450592671621842e-05, "loss": 0.0052, "step": 43600 }, { "grad_norm": 0.10348232835531235, "learning_rate": 6.449010237063882e-05, "loss": 0.0041, "step": 43610 }, { "grad_norm": 0.0848819762468338, "learning_rate": 6.447427644044343e-05, "loss": 0.0042, "step": 43620 }, { "grad_norm": 0.06861001998186111, "learning_rate": 6.44584489273629e-05, "loss": 0.0041, "step": 43630 }, { "grad_norm": 0.10247644037008286, "learning_rate": 6.444261983312817e-05, "loss": 0.0049, "step": 43640 }, { "grad_norm": 0.09005314111709595, "learning_rate": 6.442678915947023e-05, "loss": 0.0041, "step": 43650 }, { "grad_norm": 0.09428375959396362, "learning_rate": 6.441095690812032e-05, "loss": 0.0051, "step": 43660 }, { "grad_norm": 0.07646395266056061, "learning_rate": 6.439512308080982e-05, "loss": 0.0044, "step": 43670 }, { "grad_norm": 0.06830459088087082, "learning_rate": 6.43792876792703e-05, "loss": 0.0045, "step": 43680 }, { "grad_norm": 0.07830369472503662, "learning_rate": 6.43634507052335e-05, "loss": 0.0045, "step": 43690 }, { "grad_norm": 0.08408418297767639, "learning_rate": 6.434761216043135e-05, "loss": 0.0049, "step": 43700 }, { "grad_norm": 0.08441822230815887, "learning_rate": 6.433177204659589e-05, "loss": 0.0051, "step": 43710 }, { "grad_norm": 0.07337898015975952, "learning_rate": 6.431593036545938e-05, "loss": 0.0044, "step": 43720 }, { "grad_norm": 0.0843387022614479, "learning_rate": 6.430008711875426e-05, "loss": 0.0051, "step": 43730 }, { "grad_norm": 0.10258673876523972, "learning_rate": 6.42842423082131e-05, "loss": 0.0047, "step": 43740 }, { "grad_norm": 0.08413515239953995, "learning_rate": 6.426839593556868e-05, "loss": 0.0052, "step": 43750 }, { "grad_norm": 0.0914379209280014, "learning_rate": 6.425254800255392e-05, "loss": 0.0061, "step": 43760 }, { "grad_norm": 0.056382305920124054, "learning_rate": 6.423669851090196e-05, "loss": 0.0042, "step": 43770 }, { "grad_norm": 0.09257826954126358, "learning_rate": 6.422084746234604e-05, "loss": 0.0058, "step": 43780 }, { "grad_norm": 0.06204685568809509, "learning_rate": 6.420499485861965e-05, "loss": 0.0044, "step": 43790 }, { "grad_norm": 0.0859215259552002, "learning_rate": 6.418914070145636e-05, "loss": 0.0053, "step": 43800 }, { "grad_norm": 0.09523393213748932, "learning_rate": 6.417328499258996e-05, "loss": 0.0049, "step": 43810 }, { "grad_norm": 0.07625747472047806, "learning_rate": 6.415742773375446e-05, "loss": 0.0049, "step": 43820 }, { "grad_norm": 0.08072372525930405, "learning_rate": 6.414156892668393e-05, "loss": 0.0051, "step": 43830 }, { "grad_norm": 0.06838290393352509, "learning_rate": 6.412570857311267e-05, "loss": 0.004, "step": 43840 }, { "grad_norm": 0.0942726731300354, "learning_rate": 6.410984667477518e-05, "loss": 0.0049, "step": 43850 }, { "grad_norm": 0.07533089816570282, "learning_rate": 6.409398323340607e-05, "loss": 0.0044, "step": 43860 }, { "grad_norm": 0.07623682171106339, "learning_rate": 6.407811825074012e-05, "loss": 0.0045, "step": 43870 }, { "grad_norm": 0.06317724287509918, "learning_rate": 6.406225172851234e-05, "loss": 0.0035, "step": 43880 }, { "grad_norm": 0.06614480167627335, "learning_rate": 6.404638366845786e-05, "loss": 0.004, "step": 43890 }, { "grad_norm": 0.08378256857395172, "learning_rate": 6.403051407231196e-05, "loss": 0.0041, "step": 43900 }, { "grad_norm": 0.051094237715005875, "learning_rate": 6.401464294181016e-05, "loss": 0.005, "step": 43910 }, { "grad_norm": 0.07418079674243927, "learning_rate": 6.399877027868808e-05, "loss": 0.0038, "step": 43920 }, { "grad_norm": 0.0585012212395668, "learning_rate": 6.39828960846815e-05, "loss": 0.0039, "step": 43930 }, { "grad_norm": 0.07488598674535751, "learning_rate": 6.396702036152644e-05, "loss": 0.0039, "step": 43940 }, { "grad_norm": 0.06888324022293091, "learning_rate": 6.395114311095905e-05, "loss": 0.0045, "step": 43950 }, { "grad_norm": 0.08491243422031403, "learning_rate": 6.393526433471559e-05, "loss": 0.0037, "step": 43960 }, { "grad_norm": 0.1136229857802391, "learning_rate": 6.391938403453262e-05, "loss": 0.0039, "step": 43970 }, { "grad_norm": 0.08531636744737625, "learning_rate": 6.390350221214671e-05, "loss": 0.0041, "step": 43980 }, { "grad_norm": 0.09045851975679398, "learning_rate": 6.38876188692947e-05, "loss": 0.0039, "step": 43990 }, { "grad_norm": 0.060841839760541916, "learning_rate": 6.387173400771358e-05, "loss": 0.0035, "step": 44000 }, { "grad_norm": 0.07333169877529144, "learning_rate": 6.385584762914047e-05, "loss": 0.0042, "step": 44010 }, { "grad_norm": 0.08352454006671906, "learning_rate": 6.38399597353127e-05, "loss": 0.0037, "step": 44020 }, { "grad_norm": 0.0843653604388237, "learning_rate": 6.382407032796775e-05, "loss": 0.0038, "step": 44030 }, { "grad_norm": 0.09321712702512741, "learning_rate": 6.380817940884325e-05, "loss": 0.0051, "step": 44040 }, { "grad_norm": 0.08917278051376343, "learning_rate": 6.379228697967702e-05, "loss": 0.0041, "step": 44050 }, { "grad_norm": 0.06497173756361008, "learning_rate": 6.3776393042207e-05, "loss": 0.0043, "step": 44060 }, { "grad_norm": 0.0771455243229866, "learning_rate": 6.376049759817137e-05, "loss": 0.004, "step": 44070 }, { "grad_norm": 0.08287613838911057, "learning_rate": 6.37446006493084e-05, "loss": 0.0039, "step": 44080 }, { "grad_norm": 0.049467794597148895, "learning_rate": 6.372870219735656e-05, "loss": 0.0036, "step": 44090 }, { "grad_norm": 0.05588170513510704, "learning_rate": 6.371280224405452e-05, "loss": 0.004, "step": 44100 }, { "grad_norm": 0.057356491684913635, "learning_rate": 6.369690079114101e-05, "loss": 0.0038, "step": 44110 }, { "grad_norm": 0.07111898809671402, "learning_rate": 6.368099784035504e-05, "loss": 0.0047, "step": 44120 }, { "grad_norm": 0.09259866178035736, "learning_rate": 6.366509339343572e-05, "loss": 0.0044, "step": 44130 }, { "grad_norm": 0.08668159693479538, "learning_rate": 6.364918745212232e-05, "loss": 0.0038, "step": 44140 }, { "grad_norm": 0.07149432599544525, "learning_rate": 6.363328001815431e-05, "loss": 0.0049, "step": 44150 }, { "grad_norm": 0.06487443298101425, "learning_rate": 6.361737109327128e-05, "loss": 0.0038, "step": 44160 }, { "grad_norm": 0.08825235813856125, "learning_rate": 6.360146067921301e-05, "loss": 0.0042, "step": 44170 }, { "grad_norm": 0.08656514436006546, "learning_rate": 6.35855487777195e-05, "loss": 0.0046, "step": 44180 }, { "grad_norm": 0.07430942356586456, "learning_rate": 6.356963539053076e-05, "loss": 0.0042, "step": 44190 }, { "grad_norm": 0.09519527107477188, "learning_rate": 6.355372051938711e-05, "loss": 0.0049, "step": 44200 }, { "grad_norm": 0.06600350886583328, "learning_rate": 6.353780416602894e-05, "loss": 0.0039, "step": 44210 }, { "grad_norm": 0.10742366313934326, "learning_rate": 6.352188633219689e-05, "loss": 0.004, "step": 44220 }, { "grad_norm": 0.06321392208337784, "learning_rate": 6.350596701963166e-05, "loss": 0.004, "step": 44230 }, { "grad_norm": 0.07648911327123642, "learning_rate": 6.349004623007419e-05, "loss": 0.0044, "step": 44240 }, { "grad_norm": 0.07186521589756012, "learning_rate": 6.347412396526555e-05, "loss": 0.0036, "step": 44250 }, { "grad_norm": 0.0755242109298706, "learning_rate": 6.345820022694696e-05, "loss": 0.0044, "step": 44260 }, { "grad_norm": 0.07614386826753616, "learning_rate": 6.344227501685984e-05, "loss": 0.0046, "step": 44270 }, { "grad_norm": 0.07003667950630188, "learning_rate": 6.342634833674572e-05, "loss": 0.0044, "step": 44280 }, { "grad_norm": 0.08190766721963882, "learning_rate": 6.341042018834635e-05, "loss": 0.0041, "step": 44290 }, { "grad_norm": 0.0871797427535057, "learning_rate": 6.339449057340359e-05, "loss": 0.0035, "step": 44300 }, { "grad_norm": 0.10202565044164658, "learning_rate": 6.337855949365945e-05, "loss": 0.0044, "step": 44310 }, { "grad_norm": 0.09826391935348511, "learning_rate": 6.336262695085619e-05, "loss": 0.0042, "step": 44320 }, { "grad_norm": 0.06455304473638535, "learning_rate": 6.334669294673612e-05, "loss": 0.0032, "step": 44330 }, { "grad_norm": 0.06410679966211319, "learning_rate": 6.333075748304179e-05, "loss": 0.0039, "step": 44340 }, { "grad_norm": 0.07728311419487, "learning_rate": 6.331482056151585e-05, "loss": 0.0046, "step": 44350 }, { "grad_norm": 0.08480268716812134, "learning_rate": 6.329888218390117e-05, "loss": 0.0045, "step": 44360 }, { "grad_norm": 0.08310390263795853, "learning_rate": 6.328294235194072e-05, "loss": 0.0043, "step": 44370 }, { "grad_norm": 0.07663335651159286, "learning_rate": 6.326700106737765e-05, "loss": 0.0045, "step": 44380 }, { "grad_norm": 0.08041343092918396, "learning_rate": 6.325105833195531e-05, "loss": 0.004, "step": 44390 }, { "grad_norm": 0.054829709231853485, "learning_rate": 6.323511414741715e-05, "loss": 0.0039, "step": 44400 }, { "grad_norm": 0.09817001968622208, "learning_rate": 6.321916851550678e-05, "loss": 0.0038, "step": 44410 }, { "grad_norm": 0.09168652445077896, "learning_rate": 6.320322143796806e-05, "loss": 0.0036, "step": 44420 }, { "grad_norm": 0.12358789891004562, "learning_rate": 6.318727291654488e-05, "loss": 0.0047, "step": 44430 }, { "grad_norm": 0.07803592085838318, "learning_rate": 6.317132295298134e-05, "loss": 0.0058, "step": 44440 }, { "grad_norm": 0.07205061614513397, "learning_rate": 6.315537154902173e-05, "loss": 0.0052, "step": 44450 }, { "grad_norm": 0.07848362624645233, "learning_rate": 6.313941870641048e-05, "loss": 0.0059, "step": 44460 }, { "grad_norm": 0.07132751494646072, "learning_rate": 6.312346442689214e-05, "loss": 0.0051, "step": 44470 }, { "grad_norm": 0.07258550077676773, "learning_rate": 6.310750871221147e-05, "loss": 0.0057, "step": 44480 }, { "grad_norm": 0.06481947749853134, "learning_rate": 6.309155156411335e-05, "loss": 0.0043, "step": 44490 }, { "grad_norm": 0.05121621862053871, "learning_rate": 6.307559298434284e-05, "loss": 0.0034, "step": 44500 }, { "grad_norm": 0.07302957773208618, "learning_rate": 6.305963297464512e-05, "loss": 0.0041, "step": 44510 }, { "grad_norm": 0.05882759392261505, "learning_rate": 6.304367153676561e-05, "loss": 0.0049, "step": 44520 }, { "grad_norm": 0.06000415235757828, "learning_rate": 6.302770867244978e-05, "loss": 0.0046, "step": 44530 }, { "grad_norm": 0.07230868935585022, "learning_rate": 6.301174438344328e-05, "loss": 0.0052, "step": 44540 }, { "grad_norm": 0.08173223584890366, "learning_rate": 6.299577867149202e-05, "loss": 0.0055, "step": 44550 }, { "grad_norm": 0.11239179968833923, "learning_rate": 6.297981153834192e-05, "loss": 0.0048, "step": 44560 }, { "grad_norm": 0.0782516822218895, "learning_rate": 6.296384298573916e-05, "loss": 0.0043, "step": 44570 }, { "grad_norm": 0.06880445778369904, "learning_rate": 6.294787301543001e-05, "loss": 0.0034, "step": 44580 }, { "grad_norm": 0.08095135539770126, "learning_rate": 6.293190162916095e-05, "loss": 0.0049, "step": 44590 }, { "grad_norm": 0.06196292117238045, "learning_rate": 6.291592882867855e-05, "loss": 0.0039, "step": 44600 }, { "grad_norm": 0.06114300340414047, "learning_rate": 6.28999546157296e-05, "loss": 0.0041, "step": 44610 }, { "grad_norm": 0.055433474481105804, "learning_rate": 6.288397899206102e-05, "loss": 0.0042, "step": 44620 }, { "grad_norm": 0.07717714458703995, "learning_rate": 6.286800195941984e-05, "loss": 0.0049, "step": 44630 }, { "grad_norm": 0.056625016033649445, "learning_rate": 6.285202351955334e-05, "loss": 0.0037, "step": 44640 }, { "grad_norm": 0.06339599192142487, "learning_rate": 6.283604367420887e-05, "loss": 0.0039, "step": 44650 }, { "grad_norm": 0.0816776305437088, "learning_rate": 6.282006242513394e-05, "loss": 0.0034, "step": 44660 }, { "grad_norm": 0.08178457617759705, "learning_rate": 6.280407977407628e-05, "loss": 0.0035, "step": 44670 }, { "grad_norm": 0.060708556324243546, "learning_rate": 6.27880957227837e-05, "loss": 0.0032, "step": 44680 }, { "grad_norm": 0.07832477986812592, "learning_rate": 6.27721102730042e-05, "loss": 0.0041, "step": 44690 }, { "grad_norm": 0.06936945766210556, "learning_rate": 6.275612342648592e-05, "loss": 0.0039, "step": 44700 }, { "grad_norm": 0.08369305729866028, "learning_rate": 6.274013518497716e-05, "loss": 0.0038, "step": 44710 }, { "grad_norm": 0.0790063664317131, "learning_rate": 6.272414555022636e-05, "loss": 0.0043, "step": 44720 }, { "grad_norm": 0.07266976684331894, "learning_rate": 6.270815452398215e-05, "loss": 0.0035, "step": 44730 }, { "grad_norm": 0.058191780000925064, "learning_rate": 6.269216210799326e-05, "loss": 0.0033, "step": 44740 }, { "grad_norm": 0.06824943423271179, "learning_rate": 6.26761683040086e-05, "loss": 0.0039, "step": 44750 }, { "grad_norm": 0.08068878203630447, "learning_rate": 6.266017311377723e-05, "loss": 0.0028, "step": 44760 }, { "grad_norm": 0.08802786469459534, "learning_rate": 6.264417653904839e-05, "loss": 0.0044, "step": 44770 }, { "grad_norm": 0.08583983778953552, "learning_rate": 6.262817858157139e-05, "loss": 0.0036, "step": 44780 }, { "grad_norm": 0.08722619712352753, "learning_rate": 6.261217924309576e-05, "loss": 0.0041, "step": 44790 }, { "grad_norm": 0.09068838506937027, "learning_rate": 6.259617852537118e-05, "loss": 0.0042, "step": 44800 }, { "grad_norm": 0.0789499282836914, "learning_rate": 6.258017643014747e-05, "loss": 0.0041, "step": 44810 }, { "grad_norm": 0.0799926370382309, "learning_rate": 6.256417295917456e-05, "loss": 0.004, "step": 44820 }, { "grad_norm": 0.09541718661785126, "learning_rate": 6.254816811420258e-05, "loss": 0.004, "step": 44830 }, { "grad_norm": 0.0793963223695755, "learning_rate": 6.253216189698183e-05, "loss": 0.0039, "step": 44840 }, { "grad_norm": 0.05925895273685455, "learning_rate": 6.251615430926267e-05, "loss": 0.0031, "step": 44850 }, { "grad_norm": 0.08135078102350235, "learning_rate": 6.25001453527957e-05, "loss": 0.0039, "step": 44860 }, { "grad_norm": 0.060302674770355225, "learning_rate": 6.248413502933164e-05, "loss": 0.0036, "step": 44870 }, { "grad_norm": 0.0696488618850708, "learning_rate": 6.246812334062133e-05, "loss": 0.0032, "step": 44880 }, { "grad_norm": 0.07153579592704773, "learning_rate": 6.245211028841579e-05, "loss": 0.004, "step": 44890 }, { "grad_norm": 0.0791214108467102, "learning_rate": 6.24360958744662e-05, "loss": 0.0038, "step": 44900 }, { "grad_norm": 0.08416208624839783, "learning_rate": 6.242008010052387e-05, "loss": 0.004, "step": 44910 }, { "grad_norm": 0.057030703872442245, "learning_rate": 6.240406296834024e-05, "loss": 0.004, "step": 44920 }, { "grad_norm": 0.06149589270353317, "learning_rate": 6.238804447966694e-05, "loss": 0.0039, "step": 44930 }, { "grad_norm": 0.09727983176708221, "learning_rate": 6.237202463625573e-05, "loss": 0.0049, "step": 44940 }, { "grad_norm": 0.07801500707864761, "learning_rate": 6.235600343985848e-05, "loss": 0.0032, "step": 44950 }, { "grad_norm": 0.08734689652919769, "learning_rate": 6.233998089222729e-05, "loss": 0.0035, "step": 44960 }, { "grad_norm": 0.08546306937932968, "learning_rate": 6.232395699511433e-05, "loss": 0.0041, "step": 44970 }, { "grad_norm": 0.09044583141803741, "learning_rate": 6.230793175027199e-05, "loss": 0.0039, "step": 44980 }, { "grad_norm": 0.056764423847198486, "learning_rate": 6.22919051594527e-05, "loss": 0.0033, "step": 44990 }, { "grad_norm": 0.06265270709991455, "learning_rate": 6.227587722440917e-05, "loss": 0.0037, "step": 45000 }, { "grad_norm": 0.058342333883047104, "learning_rate": 6.225984794689414e-05, "loss": 0.0034, "step": 45010 }, { "grad_norm": 0.06695903092622757, "learning_rate": 6.22438173286606e-05, "loss": 0.0038, "step": 45020 }, { "grad_norm": 0.0785273090004921, "learning_rate": 6.222778537146159e-05, "loss": 0.0036, "step": 45030 }, { "grad_norm": 0.0659889504313469, "learning_rate": 6.221175207705037e-05, "loss": 0.003, "step": 45040 }, { "grad_norm": 0.0607001855969429, "learning_rate": 6.219571744718027e-05, "loss": 0.0037, "step": 45050 }, { "grad_norm": 0.09673601388931274, "learning_rate": 6.217968148360487e-05, "loss": 0.0053, "step": 45060 }, { "grad_norm": 0.043340690433979034, "learning_rate": 6.216364418807782e-05, "loss": 0.0053, "step": 45070 }, { "grad_norm": 0.0896567776799202, "learning_rate": 6.214760556235292e-05, "loss": 0.0051, "step": 45080 }, { "grad_norm": 0.07323874533176422, "learning_rate": 6.213156560818414e-05, "loss": 0.0037, "step": 45090 }, { "grad_norm": 0.05203449726104736, "learning_rate": 6.211552432732559e-05, "loss": 0.0037, "step": 45100 }, { "grad_norm": 0.0716608464717865, "learning_rate": 6.20994817215315e-05, "loss": 0.0039, "step": 45110 }, { "grad_norm": 0.08401075750589371, "learning_rate": 6.208343779255628e-05, "loss": 0.0055, "step": 45120 }, { "grad_norm": 0.0683051273226738, "learning_rate": 6.206739254215449e-05, "loss": 0.0047, "step": 45130 }, { "grad_norm": 0.07208213955163956, "learning_rate": 6.205134597208077e-05, "loss": 0.0038, "step": 45140 }, { "grad_norm": 0.07372859120368958, "learning_rate": 6.203529808409e-05, "loss": 0.0042, "step": 45150 }, { "grad_norm": 0.07236585021018982, "learning_rate": 6.201924887993712e-05, "loss": 0.0042, "step": 45160 }, { "grad_norm": 0.07254685461521149, "learning_rate": 6.200319836137725e-05, "loss": 0.0045, "step": 45170 }, { "grad_norm": 0.07170360535383224, "learning_rate": 6.198714653016565e-05, "loss": 0.0037, "step": 45180 }, { "grad_norm": 0.08846325427293777, "learning_rate": 6.197109338805774e-05, "loss": 0.0051, "step": 45190 }, { "grad_norm": 0.08377646654844284, "learning_rate": 6.195503893680903e-05, "loss": 0.0035, "step": 45200 }, { "grad_norm": 0.089971162378788, "learning_rate": 6.193898317817524e-05, "loss": 0.0053, "step": 45210 }, { "grad_norm": 0.07911268621683121, "learning_rate": 6.192292611391221e-05, "loss": 0.0039, "step": 45220 }, { "grad_norm": 0.09661568701267242, "learning_rate": 6.190686774577591e-05, "loss": 0.0044, "step": 45230 }, { "grad_norm": 0.09325601160526276, "learning_rate": 6.189080807552245e-05, "loss": 0.0057, "step": 45240 }, { "grad_norm": 0.07147861272096634, "learning_rate": 6.187474710490809e-05, "loss": 0.0034, "step": 45250 }, { "grad_norm": 0.08332420885562897, "learning_rate": 6.185868483568926e-05, "loss": 0.0036, "step": 45260 }, { "grad_norm": 0.06542552262544632, "learning_rate": 6.184262126962245e-05, "loss": 0.004, "step": 45270 }, { "grad_norm": 0.08234511315822601, "learning_rate": 6.182655640846442e-05, "loss": 0.0049, "step": 45280 }, { "grad_norm": 0.05371198058128357, "learning_rate": 6.181049025397196e-05, "loss": 0.0035, "step": 45290 }, { "grad_norm": 0.07339148968458176, "learning_rate": 6.179442280790202e-05, "loss": 0.0044, "step": 45300 }, { "grad_norm": 0.11807458102703094, "learning_rate": 6.177835407201174e-05, "loss": 0.0039, "step": 45310 }, { "grad_norm": 0.06606333702802658, "learning_rate": 6.176228404805839e-05, "loss": 0.0038, "step": 45320 }, { "grad_norm": 0.06402727216482162, "learning_rate": 6.174621273779932e-05, "loss": 0.0047, "step": 45330 }, { "grad_norm": 0.09512236714363098, "learning_rate": 6.17301401429921e-05, "loss": 0.0039, "step": 45340 }, { "grad_norm": 0.08579045534133911, "learning_rate": 6.171406626539436e-05, "loss": 0.0048, "step": 45350 }, { "grad_norm": 0.07870178669691086, "learning_rate": 6.169799110676398e-05, "loss": 0.0033, "step": 45360 }, { "grad_norm": 0.061839509755373, "learning_rate": 6.168191466885885e-05, "loss": 0.0045, "step": 45370 }, { "grad_norm": 0.04960300028324127, "learning_rate": 6.166583695343711e-05, "loss": 0.0043, "step": 45380 }, { "grad_norm": 0.058194827288389206, "learning_rate": 6.164975796225698e-05, "loss": 0.0031, "step": 45390 }, { "grad_norm": 0.07552521675825119, "learning_rate": 6.163367769707683e-05, "loss": 0.0036, "step": 45400 }, { "grad_norm": 0.05367990955710411, "learning_rate": 6.161759615965519e-05, "loss": 0.0034, "step": 45410 }, { "grad_norm": 0.0788053572177887, "learning_rate": 6.16015133517507e-05, "loss": 0.0044, "step": 45420 }, { "grad_norm": 0.08263241499662399, "learning_rate": 6.158542927512214e-05, "loss": 0.0034, "step": 45430 }, { "grad_norm": 0.08158045262098312, "learning_rate": 6.156934393152846e-05, "loss": 0.0037, "step": 45440 }, { "grad_norm": 0.07954847067594528, "learning_rate": 6.15532573227287e-05, "loss": 0.0035, "step": 45450 }, { "grad_norm": 0.059452082961797714, "learning_rate": 6.153716945048212e-05, "loss": 0.0037, "step": 45460 }, { "grad_norm": 0.06491929292678833, "learning_rate": 6.152108031654802e-05, "loss": 0.0044, "step": 45470 }, { "grad_norm": 0.10427886247634888, "learning_rate": 6.15049899226859e-05, "loss": 0.0066, "step": 45480 }, { "grad_norm": 0.11541251093149185, "learning_rate": 6.148889827065537e-05, "loss": 0.0045, "step": 45490 }, { "grad_norm": 0.07915934920310974, "learning_rate": 6.147280536221622e-05, "loss": 0.0044, "step": 45500 }, { "grad_norm": 0.07199966907501221, "learning_rate": 6.145671119912832e-05, "loss": 0.0037, "step": 45510 }, { "grad_norm": 0.07757633179426193, "learning_rate": 6.144061578315169e-05, "loss": 0.0047, "step": 45520 }, { "grad_norm": 0.06300689280033112, "learning_rate": 6.142451911604654e-05, "loss": 0.004, "step": 45530 }, { "grad_norm": 0.09589370340108871, "learning_rate": 6.140842119957315e-05, "loss": 0.0059, "step": 45540 }, { "grad_norm": 0.05670624226331711, "learning_rate": 6.139232203549197e-05, "loss": 0.0039, "step": 45550 }, { "grad_norm": 0.05948120355606079, "learning_rate": 6.137622162556357e-05, "loss": 0.0033, "step": 45560 }, { "grad_norm": 0.0603901706635952, "learning_rate": 6.13601199715487e-05, "loss": 0.0037, "step": 45570 }, { "grad_norm": 0.07602547109127045, "learning_rate": 6.134401707520816e-05, "loss": 0.005, "step": 45580 }, { "grad_norm": 0.07666005939245224, "learning_rate": 6.132791293830298e-05, "loss": 0.0037, "step": 45590 }, { "grad_norm": 0.06681861728429794, "learning_rate": 6.131180756259428e-05, "loss": 0.0034, "step": 45600 }, { "grad_norm": 0.04768803343176842, "learning_rate": 6.129570094984331e-05, "loss": 0.0036, "step": 45610 }, { "grad_norm": 0.08611691743135452, "learning_rate": 6.127959310181145e-05, "loss": 0.0039, "step": 45620 }, { "grad_norm": 0.07607194781303406, "learning_rate": 6.126348402026026e-05, "loss": 0.0036, "step": 45630 }, { "grad_norm": 0.10424419492483139, "learning_rate": 6.12473737069514e-05, "loss": 0.0045, "step": 45640 }, { "grad_norm": 0.13786473870277405, "learning_rate": 6.123126216364665e-05, "loss": 0.0051, "step": 45650 }, { "grad_norm": 0.06767603009939194, "learning_rate": 6.121514939210797e-05, "loss": 0.0036, "step": 45660 }, { "grad_norm": 0.09620437771081924, "learning_rate": 6.119903539409741e-05, "loss": 0.0046, "step": 45670 }, { "grad_norm": 0.09709635376930237, "learning_rate": 6.118292017137716e-05, "loss": 0.0041, "step": 45680 }, { "grad_norm": 0.07878581434488297, "learning_rate": 6.116680372570959e-05, "loss": 0.0036, "step": 45690 }, { "grad_norm": 0.08686259388923645, "learning_rate": 6.115068605885713e-05, "loss": 0.004, "step": 45700 }, { "grad_norm": 0.08474677056074142, "learning_rate": 6.113456717258243e-05, "loss": 0.0049, "step": 45710 }, { "grad_norm": 0.08303479105234146, "learning_rate": 6.11184470686482e-05, "loss": 0.0051, "step": 45720 }, { "grad_norm": 0.07940733432769775, "learning_rate": 6.110232574881731e-05, "loss": 0.0039, "step": 45730 }, { "grad_norm": 0.09119252115488052, "learning_rate": 6.108620321485277e-05, "loss": 0.0039, "step": 45740 }, { "grad_norm": 0.08386588841676712, "learning_rate": 6.107007946851773e-05, "loss": 0.0041, "step": 45750 }, { "grad_norm": 0.07626104354858398, "learning_rate": 6.105395451157542e-05, "loss": 0.0044, "step": 45760 }, { "grad_norm": 0.06974992901086807, "learning_rate": 6.103782834578928e-05, "loss": 0.0038, "step": 45770 }, { "grad_norm": 0.08498678356409073, "learning_rate": 6.102170097292281e-05, "loss": 0.0037, "step": 45780 }, { "grad_norm": 0.07426866888999939, "learning_rate": 6.1005572394739716e-05, "loss": 0.0038, "step": 45790 }, { "grad_norm": 0.06697935611009598, "learning_rate": 6.0989442613003765e-05, "loss": 0.0039, "step": 45800 }, { "grad_norm": 0.052268486469984055, "learning_rate": 6.097331162947888e-05, "loss": 0.0034, "step": 45810 }, { "grad_norm": 0.06593772768974304, "learning_rate": 6.095717944592914e-05, "loss": 0.0038, "step": 45820 }, { "grad_norm": 0.07257084548473358, "learning_rate": 6.094104606411873e-05, "loss": 0.0045, "step": 45830 }, { "grad_norm": 0.09404022991657257, "learning_rate": 6.0924911485811966e-05, "loss": 0.0038, "step": 45840 }, { "grad_norm": 0.06211457401514053, "learning_rate": 6.090877571277331e-05, "loss": 0.0039, "step": 45850 }, { "grad_norm": 0.1075909435749054, "learning_rate": 6.089263874676736e-05, "loss": 0.0044, "step": 45860 }, { "grad_norm": 0.0762244462966919, "learning_rate": 6.0876500589558796e-05, "loss": 0.0043, "step": 45870 }, { "grad_norm": 0.07638303935527802, "learning_rate": 6.086036124291248e-05, "loss": 0.0041, "step": 45880 }, { "grad_norm": 0.07302479445934296, "learning_rate": 6.084422070859339e-05, "loss": 0.0037, "step": 45890 }, { "grad_norm": 0.08403456956148148, "learning_rate": 6.082807898836663e-05, "loss": 0.0051, "step": 45900 }, { "grad_norm": 0.07660134881734848, "learning_rate": 6.081193608399742e-05, "loss": 0.0042, "step": 45910 }, { "grad_norm": 0.07673386484384537, "learning_rate": 6.0795791997251164e-05, "loss": 0.0036, "step": 45920 }, { "grad_norm": 0.04471905902028084, "learning_rate": 6.0779646729893294e-05, "loss": 0.0044, "step": 45930 }, { "grad_norm": 0.04931139573454857, "learning_rate": 6.0763500283689476e-05, "loss": 0.0036, "step": 45940 }, { "grad_norm": 0.06310461461544037, "learning_rate": 6.0747352660405455e-05, "loss": 0.0033, "step": 45950 }, { "grad_norm": 0.0901721641421318, "learning_rate": 6.073120386180709e-05, "loss": 0.0041, "step": 45960 }, { "grad_norm": 0.07089871913194656, "learning_rate": 6.0715053889660425e-05, "loss": 0.0039, "step": 45970 }, { "grad_norm": 0.06505977362394333, "learning_rate": 6.069890274573157e-05, "loss": 0.0035, "step": 45980 }, { "grad_norm": 0.08852928131818771, "learning_rate": 6.068275043178679e-05, "loss": 0.0048, "step": 45990 }, { "grad_norm": 0.08416961133480072, "learning_rate": 6.066659694959248e-05, "loss": 0.0047, "step": 46000 }, { "grad_norm": 0.06596454232931137, "learning_rate": 6.0650442300915176e-05, "loss": 0.0052, "step": 46010 }, { "grad_norm": 0.07297929376363754, "learning_rate": 6.063428648752152e-05, "loss": 0.004, "step": 46020 }, { "grad_norm": 0.09559136629104614, "learning_rate": 6.0618129511178266e-05, "loss": 0.0036, "step": 46030 }, { "grad_norm": 0.07620558887720108, "learning_rate": 6.060197137365234e-05, "loss": 0.0043, "step": 46040 }, { "grad_norm": 0.06998409330844879, "learning_rate": 6.058581207671077e-05, "loss": 0.0043, "step": 46050 }, { "grad_norm": 0.08436205983161926, "learning_rate": 6.056965162212072e-05, "loss": 0.0033, "step": 46060 }, { "grad_norm": 0.08034384995698929, "learning_rate": 6.055349001164943e-05, "loss": 0.0039, "step": 46070 }, { "grad_norm": 0.07030408829450607, "learning_rate": 6.0537327247064347e-05, "loss": 0.0048, "step": 46080 }, { "grad_norm": 0.07950497418642044, "learning_rate": 6.0521163330133014e-05, "loss": 0.0043, "step": 46090 }, { "grad_norm": 0.1013571098446846, "learning_rate": 6.050499826262306e-05, "loss": 0.0054, "step": 46100 }, { "grad_norm": 0.09646370261907578, "learning_rate": 6.0488832046302294e-05, "loss": 0.0042, "step": 46110 }, { "grad_norm": 0.08679002523422241, "learning_rate": 6.0472664682938626e-05, "loss": 0.0036, "step": 46120 }, { "grad_norm": 0.06349362432956696, "learning_rate": 6.045649617430009e-05, "loss": 0.0044, "step": 46130 }, { "grad_norm": 0.05035148561000824, "learning_rate": 6.0440326522154866e-05, "loss": 0.0039, "step": 46140 }, { "grad_norm": 0.07231468707323074, "learning_rate": 6.0424155728271224e-05, "loss": 0.0051, "step": 46150 }, { "grad_norm": 0.06749957799911499, "learning_rate": 6.040798379441758e-05, "loss": 0.0037, "step": 46160 }, { "grad_norm": 0.08808626234531403, "learning_rate": 6.0391810722362485e-05, "loss": 0.0033, "step": 46170 }, { "grad_norm": 0.09642906486988068, "learning_rate": 6.037563651387458e-05, "loss": 0.004, "step": 46180 }, { "grad_norm": 0.08194834738969803, "learning_rate": 6.0359461170722666e-05, "loss": 0.0046, "step": 46190 }, { "grad_norm": 0.08298951387405396, "learning_rate": 6.034328469467566e-05, "loss": 0.0036, "step": 46200 }, { "grad_norm": 0.08154194802045822, "learning_rate": 6.0327107087502596e-05, "loss": 0.0046, "step": 46210 }, { "grad_norm": 0.10856829583644867, "learning_rate": 6.031092835097262e-05, "loss": 0.0044, "step": 46220 }, { "grad_norm": 0.0725933164358139, "learning_rate": 6.0294748486855024e-05, "loss": 0.0037, "step": 46230 }, { "grad_norm": 0.0794324204325676, "learning_rate": 6.0278567496919216e-05, "loss": 0.0035, "step": 46240 }, { "grad_norm": 0.06131510064005852, "learning_rate": 6.026238538293472e-05, "loss": 0.0031, "step": 46250 }, { "grad_norm": 0.0730283111333847, "learning_rate": 6.024620214667118e-05, "loss": 0.0034, "step": 46260 }, { "grad_norm": 0.09257549792528152, "learning_rate": 6.0230017789898384e-05, "loss": 0.004, "step": 46270 }, { "grad_norm": 0.07397588342428207, "learning_rate": 6.021383231438622e-05, "loss": 0.0042, "step": 46280 }, { "grad_norm": 0.06644982099533081, "learning_rate": 6.0197645721904704e-05, "loss": 0.0039, "step": 46290 }, { "grad_norm": 0.07289931923151016, "learning_rate": 6.0181458014224e-05, "loss": 0.0048, "step": 46300 }, { "grad_norm": 0.0526733472943306, "learning_rate": 6.016526919311435e-05, "loss": 0.0036, "step": 46310 }, { "grad_norm": 0.06785406172275543, "learning_rate": 6.014907926034613e-05, "loss": 0.0035, "step": 46320 }, { "grad_norm": 0.09522684663534164, "learning_rate": 6.0132888217689875e-05, "loss": 0.0038, "step": 46330 }, { "grad_norm": 0.07681062072515488, "learning_rate": 6.01166960669162e-05, "loss": 0.0049, "step": 46340 }, { "grad_norm": 0.08881966769695282, "learning_rate": 6.0100502809795845e-05, "loss": 0.0035, "step": 46350 }, { "grad_norm": 0.07934550940990448, "learning_rate": 6.00843084480997e-05, "loss": 0.0045, "step": 46360 }, { "grad_norm": 0.069003164768219, "learning_rate": 6.0068112983598736e-05, "loss": 0.0051, "step": 46370 }, { "grad_norm": 0.11645486205816269, "learning_rate": 6.005191641806407e-05, "loss": 0.0043, "step": 46380 }, { "grad_norm": 0.05697302520275116, "learning_rate": 6.003571875326694e-05, "loss": 0.0039, "step": 46390 }, { "grad_norm": 0.06928828358650208, "learning_rate": 6.00195199909787e-05, "loss": 0.0046, "step": 46400 }, { "grad_norm": 0.07863523811101913, "learning_rate": 6.000332013297082e-05, "loss": 0.0036, "step": 46410 }, { "grad_norm": 0.07706917822360992, "learning_rate": 5.998711918101487e-05, "loss": 0.0037, "step": 46420 }, { "grad_norm": 0.06868689507246017, "learning_rate": 5.997091713688261e-05, "loss": 0.0043, "step": 46430 }, { "grad_norm": 0.1005658358335495, "learning_rate": 5.9954714002345836e-05, "loss": 0.0039, "step": 46440 }, { "grad_norm": 0.055541299283504486, "learning_rate": 5.993850977917649e-05, "loss": 0.0042, "step": 46450 }, { "grad_norm": 0.0777706652879715, "learning_rate": 5.992230446914667e-05, "loss": 0.0049, "step": 46460 }, { "grad_norm": 0.05805917829275131, "learning_rate": 5.990609807402855e-05, "loss": 0.004, "step": 46470 }, { "grad_norm": 0.09153320640325546, "learning_rate": 5.988989059559443e-05, "loss": 0.0046, "step": 46480 }, { "grad_norm": 0.06126364693045616, "learning_rate": 5.987368203561675e-05, "loss": 0.0035, "step": 46490 }, { "grad_norm": 0.11456108838319778, "learning_rate": 5.9857472395868055e-05, "loss": 0.0047, "step": 46500 }, { "grad_norm": 0.09093735367059708, "learning_rate": 5.9841261678120983e-05, "loss": 0.0043, "step": 46510 }, { "grad_norm": 0.0735916718840599, "learning_rate": 5.982504988414834e-05, "loss": 0.0051, "step": 46520 }, { "grad_norm": 0.09127077460289001, "learning_rate": 5.9808837015723015e-05, "loss": 0.0041, "step": 46530 }, { "grad_norm": 0.0721723735332489, "learning_rate": 5.9792623074618016e-05, "loss": 0.0041, "step": 46540 }, { "grad_norm": 0.07338955253362656, "learning_rate": 5.977640806260648e-05, "loss": 0.0038, "step": 46550 }, { "grad_norm": 0.10892092436552048, "learning_rate": 5.976019198146166e-05, "loss": 0.005, "step": 46560 }, { "grad_norm": 0.09558948129415512, "learning_rate": 5.974397483295692e-05, "loss": 0.0042, "step": 46570 }, { "grad_norm": 0.08851053565740585, "learning_rate": 5.972775661886575e-05, "loss": 0.0051, "step": 46580 }, { "grad_norm": 0.07779895514249802, "learning_rate": 5.971153734096173e-05, "loss": 0.0045, "step": 46590 }, { "grad_norm": 0.062345974147319794, "learning_rate": 5.969531700101859e-05, "loss": 0.0037, "step": 46600 }, { "grad_norm": 0.053848810493946075, "learning_rate": 5.9679095600810155e-05, "loss": 0.0034, "step": 46610 }, { "grad_norm": 0.051727648824453354, "learning_rate": 5.9662873142110384e-05, "loss": 0.004, "step": 46620 }, { "grad_norm": 0.05593183636665344, "learning_rate": 5.964664962669333e-05, "loss": 0.0042, "step": 46630 }, { "grad_norm": 0.07275684177875519, "learning_rate": 5.9630425056333186e-05, "loss": 0.0049, "step": 46640 }, { "grad_norm": 0.06785522401332855, "learning_rate": 5.961419943280422e-05, "loss": 0.0037, "step": 46650 }, { "grad_norm": 0.10572367161512375, "learning_rate": 5.959797275788087e-05, "loss": 0.0053, "step": 46660 }, { "grad_norm": 0.07764434814453125, "learning_rate": 5.958174503333765e-05, "loss": 0.0045, "step": 46670 }, { "grad_norm": 0.07209793478250504, "learning_rate": 5.9565516260949195e-05, "loss": 0.0049, "step": 46680 }, { "grad_norm": 0.06296129524707794, "learning_rate": 5.9549286442490273e-05, "loss": 0.0033, "step": 46690 }, { "grad_norm": 0.07369379699230194, "learning_rate": 5.953305557973572e-05, "loss": 0.0035, "step": 46700 }, { "grad_norm": 0.10187216848134995, "learning_rate": 5.9516823674460564e-05, "loss": 0.0039, "step": 46710 }, { "grad_norm": 0.10346677899360657, "learning_rate": 5.9500590728439875e-05, "loss": 0.0048, "step": 46720 }, { "grad_norm": 0.07950105518102646, "learning_rate": 5.948435674344886e-05, "loss": 0.0039, "step": 46730 }, { "grad_norm": 0.07972294092178345, "learning_rate": 5.946812172126285e-05, "loss": 0.0047, "step": 46740 }, { "grad_norm": 0.05719251558184624, "learning_rate": 5.945188566365729e-05, "loss": 0.0029, "step": 46750 }, { "grad_norm": 0.12066999077796936, "learning_rate": 5.943564857240773e-05, "loss": 0.0036, "step": 46760 }, { "grad_norm": 0.07307778298854828, "learning_rate": 5.941941044928981e-05, "loss": 0.0034, "step": 46770 }, { "grad_norm": 0.061052847653627396, "learning_rate": 5.940317129607935e-05, "loss": 0.0042, "step": 46780 }, { "grad_norm": 0.056580934673547745, "learning_rate": 5.9386931114552204e-05, "loss": 0.004, "step": 46790 }, { "grad_norm": 0.07861348986625671, "learning_rate": 5.937068990648438e-05, "loss": 0.004, "step": 46800 }, { "grad_norm": 0.08271710574626923, "learning_rate": 5.935444767365199e-05, "loss": 0.0039, "step": 46810 }, { "grad_norm": 0.07703734189271927, "learning_rate": 5.933820441783129e-05, "loss": 0.0045, "step": 46820 }, { "grad_norm": 0.0621643029153347, "learning_rate": 5.932196014079857e-05, "loss": 0.0038, "step": 46830 }, { "grad_norm": 0.08306055516004562, "learning_rate": 5.930571484433032e-05, "loss": 0.0036, "step": 46840 }, { "grad_norm": 0.07810114324092865, "learning_rate": 5.928946853020309e-05, "loss": 0.0045, "step": 46850 }, { "grad_norm": 0.06396647542715073, "learning_rate": 5.927322120019352e-05, "loss": 0.0039, "step": 46860 }, { "grad_norm": 0.06340160220861435, "learning_rate": 5.925697285607844e-05, "loss": 0.0036, "step": 46870 }, { "grad_norm": 0.07789649069309235, "learning_rate": 5.924072349963472e-05, "loss": 0.0042, "step": 46880 }, { "grad_norm": 0.09484162926673889, "learning_rate": 5.922447313263938e-05, "loss": 0.0034, "step": 46890 }, { "grad_norm": 0.07316047698259354, "learning_rate": 5.920822175686952e-05, "loss": 0.0035, "step": 46900 }, { "grad_norm": 0.08416015654802322, "learning_rate": 5.919196937410239e-05, "loss": 0.003, "step": 46910 }, { "grad_norm": 0.05722419545054436, "learning_rate": 5.917571598611529e-05, "loss": 0.004, "step": 46920 }, { "grad_norm": 0.08833671361207962, "learning_rate": 5.9159461594685686e-05, "loss": 0.0041, "step": 46930 }, { "grad_norm": 0.06096859648823738, "learning_rate": 5.914320620159114e-05, "loss": 0.0035, "step": 46940 }, { "grad_norm": 0.056959185749292374, "learning_rate": 5.912694980860931e-05, "loss": 0.0039, "step": 46950 }, { "grad_norm": 0.09445392340421677, "learning_rate": 5.9110692417517964e-05, "loss": 0.0046, "step": 46960 }, { "grad_norm": 0.0807121992111206, "learning_rate": 5.9094434030095e-05, "loss": 0.0046, "step": 46970 }, { "grad_norm": 0.07570084184408188, "learning_rate": 5.907817464811841e-05, "loss": 0.0035, "step": 46980 }, { "grad_norm": 0.05100418999791145, "learning_rate": 5.906191427336627e-05, "loss": 0.0045, "step": 46990 }, { "grad_norm": 0.06299883872270584, "learning_rate": 5.9045652907616835e-05, "loss": 0.0039, "step": 47000 }, { "grad_norm": 0.08639666438102722, "learning_rate": 5.902939055264838e-05, "loss": 0.0036, "step": 47010 }, { "grad_norm": 0.07412246614694595, "learning_rate": 5.901312721023935e-05, "loss": 0.0034, "step": 47020 }, { "grad_norm": 0.08745206892490387, "learning_rate": 5.8996862882168294e-05, "loss": 0.0063, "step": 47030 }, { "grad_norm": 0.06774777919054031, "learning_rate": 5.8980597570213826e-05, "loss": 0.0041, "step": 47040 }, { "grad_norm": 0.09403018653392792, "learning_rate": 5.896433127615471e-05, "loss": 0.0041, "step": 47050 }, { "grad_norm": 0.0692843571305275, "learning_rate": 5.894806400176981e-05, "loss": 0.0031, "step": 47060 }, { "grad_norm": 0.08823847025632858, "learning_rate": 5.893179574883808e-05, "loss": 0.0035, "step": 47070 }, { "grad_norm": 0.0654345229268074, "learning_rate": 5.8915526519138585e-05, "loss": 0.0032, "step": 47080 }, { "grad_norm": 0.08364786952733994, "learning_rate": 5.889925631445053e-05, "loss": 0.0044, "step": 47090 }, { "grad_norm": 0.06336229294538498, "learning_rate": 5.888298513655318e-05, "loss": 0.0035, "step": 47100 }, { "grad_norm": 0.06872545182704926, "learning_rate": 5.886671298722592e-05, "loss": 0.0034, "step": 47110 }, { "grad_norm": 0.07013107091188431, "learning_rate": 5.885043986824828e-05, "loss": 0.0037, "step": 47120 }, { "grad_norm": 0.07266920059919357, "learning_rate": 5.883416578139982e-05, "loss": 0.0043, "step": 47130 }, { "grad_norm": 0.07701953500509262, "learning_rate": 5.881789072846029e-05, "loss": 0.0041, "step": 47140 }, { "grad_norm": 0.07136093825101852, "learning_rate": 5.8801614711209474e-05, "loss": 0.0035, "step": 47150 }, { "grad_norm": 0.08867765218019485, "learning_rate": 5.878533773142732e-05, "loss": 0.0046, "step": 47160 }, { "grad_norm": 0.09230228513479233, "learning_rate": 5.8769059790893834e-05, "loss": 0.004, "step": 47170 }, { "grad_norm": 0.06800482422113419, "learning_rate": 5.8752780891389145e-05, "loss": 0.0037, "step": 47180 }, { "grad_norm": 0.07318036258220673, "learning_rate": 5.873650103469351e-05, "loss": 0.0044, "step": 47190 }, { "grad_norm": 0.05927496775984764, "learning_rate": 5.872022022258725e-05, "loss": 0.0042, "step": 47200 }, { "grad_norm": 0.07058178633451462, "learning_rate": 5.8703938456850814e-05, "loss": 0.003, "step": 47210 }, { "grad_norm": 0.08113192766904831, "learning_rate": 5.868765573926477e-05, "loss": 0.0034, "step": 47220 }, { "grad_norm": 0.07498950511217117, "learning_rate": 5.867137207160974e-05, "loss": 0.0044, "step": 47230 }, { "grad_norm": 0.07881834357976913, "learning_rate": 5.86550874556665e-05, "loss": 0.0047, "step": 47240 }, { "grad_norm": 0.07494540512561798, "learning_rate": 5.863880189321592e-05, "loss": 0.005, "step": 47250 }, { "grad_norm": 0.07566572725772858, "learning_rate": 5.862251538603894e-05, "loss": 0.0037, "step": 47260 }, { "grad_norm": 0.08330239355564117, "learning_rate": 5.8606227935916656e-05, "loss": 0.004, "step": 47270 }, { "grad_norm": 0.07289987057447433, "learning_rate": 5.858993954463021e-05, "loss": 0.004, "step": 47280 }, { "grad_norm": 0.09267823398113251, "learning_rate": 5.8573650213960896e-05, "loss": 0.0037, "step": 47290 }, { "grad_norm": 0.06241631507873535, "learning_rate": 5.8557359945690084e-05, "loss": 0.0037, "step": 47300 }, { "grad_norm": 0.07808317244052887, "learning_rate": 5.8541068741599246e-05, "loss": 0.0039, "step": 47310 }, { "grad_norm": 0.0727616399526596, "learning_rate": 5.8524776603469985e-05, "loss": 0.0042, "step": 47320 }, { "grad_norm": 0.0861850157380104, "learning_rate": 5.850848353308397e-05, "loss": 0.004, "step": 47330 }, { "grad_norm": 0.05840366333723068, "learning_rate": 5.849218953222297e-05, "loss": 0.0037, "step": 47340 }, { "grad_norm": 0.07307076454162598, "learning_rate": 5.847589460266891e-05, "loss": 0.0049, "step": 47350 }, { "grad_norm": 0.09706687182188034, "learning_rate": 5.845959874620375e-05, "loss": 0.0038, "step": 47360 }, { "grad_norm": 0.05745455250144005, "learning_rate": 5.84433019646096e-05, "loss": 0.0036, "step": 47370 }, { "grad_norm": 0.07714205980300903, "learning_rate": 5.842700425966863e-05, "loss": 0.0039, "step": 47380 }, { "grad_norm": 0.09358243644237518, "learning_rate": 5.841070563316315e-05, "loss": 0.0036, "step": 47390 }, { "grad_norm": 0.07686634361743927, "learning_rate": 5.839440608687554e-05, "loss": 0.0035, "step": 47400 }, { "grad_norm": 0.08818287402391434, "learning_rate": 5.837810562258831e-05, "loss": 0.004, "step": 47410 }, { "grad_norm": 0.09152857959270477, "learning_rate": 5.836180424208405e-05, "loss": 0.0031, "step": 47420 }, { "grad_norm": 0.08930201828479767, "learning_rate": 5.834550194714543e-05, "loss": 0.0034, "step": 47430 }, { "grad_norm": 0.07354257255792618, "learning_rate": 5.832919873955528e-05, "loss": 0.0037, "step": 47440 }, { "grad_norm": 0.056344784796237946, "learning_rate": 5.831289462109648e-05, "loss": 0.0037, "step": 47450 }, { "grad_norm": 0.09575742483139038, "learning_rate": 5.829658959355202e-05, "loss": 0.0039, "step": 47460 }, { "grad_norm": 0.081752248108387, "learning_rate": 5.8280283658704995e-05, "loss": 0.0037, "step": 47470 }, { "grad_norm": 0.0936850905418396, "learning_rate": 5.826397681833859e-05, "loss": 0.0036, "step": 47480 }, { "grad_norm": 0.07283613085746765, "learning_rate": 5.824766907423612e-05, "loss": 0.0038, "step": 47490 }, { "grad_norm": 0.08445794135332108, "learning_rate": 5.8231360428180945e-05, "loss": 0.0039, "step": 47500 }, { "grad_norm": 0.08418835699558258, "learning_rate": 5.821505088195658e-05, "loss": 0.0034, "step": 47510 }, { "grad_norm": 0.09377472847700119, "learning_rate": 5.819874043734661e-05, "loss": 0.0041, "step": 47520 }, { "grad_norm": 0.05313543602824211, "learning_rate": 5.8182429096134695e-05, "loss": 0.0036, "step": 47530 }, { "grad_norm": 0.08711869269609451, "learning_rate": 5.816611686010465e-05, "loss": 0.0042, "step": 47540 }, { "grad_norm": 0.06891560554504395, "learning_rate": 5.814980373104033e-05, "loss": 0.0041, "step": 47550 }, { "grad_norm": 0.06791597604751587, "learning_rate": 5.813348971072572e-05, "loss": 0.0034, "step": 47560 }, { "grad_norm": 0.05774970352649689, "learning_rate": 5.811717480094492e-05, "loss": 0.0037, "step": 47570 }, { "grad_norm": 0.06668929755687714, "learning_rate": 5.810085900348209e-05, "loss": 0.0041, "step": 47580 }, { "grad_norm": 0.06606259942054749, "learning_rate": 5.8084542320121483e-05, "loss": 0.0038, "step": 47590 }, { "grad_norm": 0.10296767204999924, "learning_rate": 5.8068224752647497e-05, "loss": 0.0033, "step": 47600 }, { "grad_norm": 0.12205658107995987, "learning_rate": 5.805190630284457e-05, "loss": 0.0041, "step": 47610 }, { "grad_norm": 0.06486785411834717, "learning_rate": 5.803558697249729e-05, "loss": 0.0039, "step": 47620 }, { "grad_norm": 0.07550870627164841, "learning_rate": 5.8019266763390276e-05, "loss": 0.0042, "step": 47630 }, { "grad_norm": 0.0527711883187294, "learning_rate": 5.800294567730833e-05, "loss": 0.004, "step": 47640 }, { "grad_norm": 0.08215492963790894, "learning_rate": 5.798662371603628e-05, "loss": 0.0049, "step": 47650 }, { "grad_norm": 0.07317696511745453, "learning_rate": 5.797030088135904e-05, "loss": 0.0034, "step": 47660 }, { "grad_norm": 0.0756441205739975, "learning_rate": 5.7953977175061705e-05, "loss": 0.0034, "step": 47670 }, { "grad_norm": 0.07031011581420898, "learning_rate": 5.793765259892938e-05, "loss": 0.0038, "step": 47680 }, { "grad_norm": 0.07785550504922867, "learning_rate": 5.792132715474729e-05, "loss": 0.0043, "step": 47690 }, { "grad_norm": 0.0555693656206131, "learning_rate": 5.790500084430078e-05, "loss": 0.0036, "step": 47700 }, { "grad_norm": 0.08463427424430847, "learning_rate": 5.7888673669375274e-05, "loss": 0.0039, "step": 47710 }, { "grad_norm": 0.07201190292835236, "learning_rate": 5.787234563175625e-05, "loss": 0.0033, "step": 47720 }, { "grad_norm": 0.06898626685142517, "learning_rate": 5.7856016733229355e-05, "loss": 0.0037, "step": 47730 }, { "grad_norm": 0.048834629356861115, "learning_rate": 5.7839686975580297e-05, "loss": 0.0035, "step": 47740 }, { "grad_norm": 0.053422510623931885, "learning_rate": 5.782335636059484e-05, "loss": 0.0037, "step": 47750 }, { "grad_norm": 0.07993140071630478, "learning_rate": 5.780702489005889e-05, "loss": 0.0037, "step": 47760 }, { "grad_norm": 0.06092953309416771, "learning_rate": 5.779069256575845e-05, "loss": 0.0045, "step": 47770 }, { "grad_norm": 0.055250078439712524, "learning_rate": 5.7774359389479574e-05, "loss": 0.004, "step": 47780 }, { "grad_norm": 0.053984541445970535, "learning_rate": 5.775802536300845e-05, "loss": 0.0032, "step": 47790 }, { "grad_norm": 0.08773173391819, "learning_rate": 5.774169048813134e-05, "loss": 0.0033, "step": 47800 }, { "grad_norm": 0.06256565451622009, "learning_rate": 5.77253547666346e-05, "loss": 0.0041, "step": 47810 }, { "grad_norm": 0.13198435306549072, "learning_rate": 5.770901820030465e-05, "loss": 0.0035, "step": 47820 }, { "grad_norm": 0.09406384825706482, "learning_rate": 5.769268079092809e-05, "loss": 0.0047, "step": 47830 }, { "grad_norm": 0.07143331319093704, "learning_rate": 5.767634254029151e-05, "loss": 0.0039, "step": 47840 }, { "grad_norm": 0.09517094492912292, "learning_rate": 5.7660003450181655e-05, "loss": 0.0035, "step": 47850 }, { "grad_norm": 0.10112807899713516, "learning_rate": 5.764366352238534e-05, "loss": 0.0039, "step": 47860 }, { "grad_norm": 0.056170035153627396, "learning_rate": 5.7627322758689474e-05, "loss": 0.0034, "step": 47870 }, { "grad_norm": 0.11500968039035797, "learning_rate": 5.761098116088105e-05, "loss": 0.0036, "step": 47880 }, { "grad_norm": 0.06359753012657166, "learning_rate": 5.759463873074717e-05, "loss": 0.004, "step": 47890 }, { "grad_norm": 0.0742361843585968, "learning_rate": 5.757829547007504e-05, "loss": 0.0043, "step": 47900 }, { "grad_norm": 0.06797056645154953, "learning_rate": 5.756195138065189e-05, "loss": 0.0031, "step": 47910 }, { "grad_norm": 0.10225832462310791, "learning_rate": 5.754560646426511e-05, "loss": 0.0042, "step": 47920 }, { "grad_norm": 0.0926908329129219, "learning_rate": 5.752926072270216e-05, "loss": 0.005, "step": 47930 }, { "grad_norm": 0.07131922990083694, "learning_rate": 5.7512914157750563e-05, "loss": 0.0031, "step": 47940 }, { "grad_norm": 0.06336205452680588, "learning_rate": 5.749656677119798e-05, "loss": 0.004, "step": 47950 }, { "grad_norm": 0.04926080256700516, "learning_rate": 5.7480218564832125e-05, "loss": 0.0036, "step": 47960 }, { "grad_norm": 0.09085939079523087, "learning_rate": 5.746386954044082e-05, "loss": 0.0034, "step": 47970 }, { "grad_norm": 0.06883510202169418, "learning_rate": 5.744751969981195e-05, "loss": 0.0048, "step": 47980 }, { "grad_norm": 0.09948855638504028, "learning_rate": 5.7431169044733526e-05, "loss": 0.0039, "step": 47990 }, { "grad_norm": 0.07278603315353394, "learning_rate": 5.741481757699364e-05, "loss": 0.0036, "step": 48000 }, { "grad_norm": 0.11160745471715927, "learning_rate": 5.7398465298380434e-05, "loss": 0.0036, "step": 48010 }, { "grad_norm": 0.06634125113487244, "learning_rate": 5.7382112210682193e-05, "loss": 0.003, "step": 48020 }, { "grad_norm": 0.08563842624425888, "learning_rate": 5.7365758315687266e-05, "loss": 0.0035, "step": 48030 }, { "grad_norm": 0.056989267468452454, "learning_rate": 5.734940361518407e-05, "loss": 0.0031, "step": 48040 }, { "grad_norm": 0.06672725081443787, "learning_rate": 5.733304811096116e-05, "loss": 0.0043, "step": 48050 }, { "grad_norm": 0.06957028061151505, "learning_rate": 5.731669180480713e-05, "loss": 0.0027, "step": 48060 }, { "grad_norm": 0.056139856576919556, "learning_rate": 5.730033469851067e-05, "loss": 0.003, "step": 48070 }, { "grad_norm": 0.06921689957380295, "learning_rate": 5.7283976793860607e-05, "loss": 0.003, "step": 48080 }, { "grad_norm": 0.04505686089396477, "learning_rate": 5.726761809264577e-05, "loss": 0.0037, "step": 48090 }, { "grad_norm": 0.07775874435901642, "learning_rate": 5.7251258596655155e-05, "loss": 0.0034, "step": 48100 }, { "grad_norm": 0.06973040103912354, "learning_rate": 5.72348983076778e-05, "loss": 0.0041, "step": 48110 }, { "grad_norm": 0.07678286731243134, "learning_rate": 5.7218537227502854e-05, "loss": 0.0032, "step": 48120 }, { "grad_norm": 0.07175694406032562, "learning_rate": 5.720217535791951e-05, "loss": 0.0034, "step": 48130 }, { "grad_norm": 0.06886908411979675, "learning_rate": 5.718581270071711e-05, "loss": 0.004, "step": 48140 }, { "grad_norm": 0.07410486042499542, "learning_rate": 5.716944925768505e-05, "loss": 0.0036, "step": 48150 }, { "grad_norm": 0.07808142900466919, "learning_rate": 5.7153085030612786e-05, "loss": 0.0036, "step": 48160 }, { "grad_norm": 0.08384978771209717, "learning_rate": 5.713672002128989e-05, "loss": 0.0037, "step": 48170 }, { "grad_norm": 0.06880206614732742, "learning_rate": 5.712035423150602e-05, "loss": 0.003, "step": 48180 }, { "grad_norm": 0.09165997803211212, "learning_rate": 5.710398766305094e-05, "loss": 0.0039, "step": 48190 }, { "grad_norm": 0.06957393884658813, "learning_rate": 5.708762031771442e-05, "loss": 0.0034, "step": 48200 }, { "grad_norm": 0.11767637729644775, "learning_rate": 5.707125219728642e-05, "loss": 0.0047, "step": 48210 }, { "grad_norm": 0.08818873763084412, "learning_rate": 5.7054883303556905e-05, "loss": 0.0041, "step": 48220 }, { "grad_norm": 0.06633539497852325, "learning_rate": 5.703851363831595e-05, "loss": 0.004, "step": 48230 }, { "grad_norm": 0.12735481560230255, "learning_rate": 5.702214320335374e-05, "loss": 0.0042, "step": 48240 }, { "grad_norm": 0.07602427154779434, "learning_rate": 5.7005772000460514e-05, "loss": 0.0033, "step": 48250 }, { "grad_norm": 0.08488171547651291, "learning_rate": 5.698940003142656e-05, "loss": 0.0037, "step": 48260 }, { "grad_norm": 0.0744636207818985, "learning_rate": 5.697302729804236e-05, "loss": 0.0035, "step": 48270 }, { "grad_norm": 0.06934840232133865, "learning_rate": 5.695665380209837e-05, "loss": 0.004, "step": 48280 }, { "grad_norm": 0.06187822297215462, "learning_rate": 5.6940279545385165e-05, "loss": 0.0038, "step": 48290 }, { "grad_norm": 0.094436876475811, "learning_rate": 5.692390452969344e-05, "loss": 0.0035, "step": 48300 }, { "grad_norm": 0.0938766747713089, "learning_rate": 5.690752875681392e-05, "loss": 0.0045, "step": 48310 }, { "grad_norm": 0.10221950709819794, "learning_rate": 5.6891152228537435e-05, "loss": 0.0046, "step": 48320 }, { "grad_norm": 0.049775347113609314, "learning_rate": 5.687477494665492e-05, "loss": 0.0035, "step": 48330 }, { "grad_norm": 0.12110251188278198, "learning_rate": 5.685839691295733e-05, "loss": 0.0034, "step": 48340 }, { "grad_norm": 0.07145668566226959, "learning_rate": 5.6842018129235786e-05, "loss": 0.004, "step": 48350 }, { "grad_norm": 0.08798589557409286, "learning_rate": 5.6825638597281404e-05, "loss": 0.0045, "step": 48360 }, { "grad_norm": 0.08337345719337463, "learning_rate": 5.680925831888546e-05, "loss": 0.0041, "step": 48370 }, { "grad_norm": 0.10089656710624695, "learning_rate": 5.6792877295839274e-05, "loss": 0.0044, "step": 48380 }, { "grad_norm": 0.07786689698696136, "learning_rate": 5.6776495529934224e-05, "loss": 0.0038, "step": 48390 }, { "grad_norm": 0.05707242712378502, "learning_rate": 5.6760113022961824e-05, "loss": 0.0038, "step": 48400 }, { "grad_norm": 0.09245365858078003, "learning_rate": 5.6743729776713617e-05, "loss": 0.0048, "step": 48410 }, { "grad_norm": 0.09207760542631149, "learning_rate": 5.672734579298126e-05, "loss": 0.0052, "step": 48420 }, { "grad_norm": 0.06878822296857834, "learning_rate": 5.671096107355649e-05, "loss": 0.004, "step": 48430 }, { "grad_norm": 0.09245230257511139, "learning_rate": 5.66945756202311e-05, "loss": 0.0036, "step": 48440 }, { "grad_norm": 0.06667915731668472, "learning_rate": 5.667818943479699e-05, "loss": 0.0045, "step": 48450 }, { "grad_norm": 0.060195982456207275, "learning_rate": 5.666180251904612e-05, "loss": 0.0037, "step": 48460 }, { "grad_norm": 0.06984463334083557, "learning_rate": 5.6645414874770555e-05, "loss": 0.0036, "step": 48470 }, { "grad_norm": 0.07114417850971222, "learning_rate": 5.66290265037624e-05, "loss": 0.0042, "step": 48480 }, { "grad_norm": 0.06837528944015503, "learning_rate": 5.661263740781386e-05, "loss": 0.0042, "step": 48490 }, { "grad_norm": 0.07259264588356018, "learning_rate": 5.6596247588717254e-05, "loss": 0.0042, "step": 48500 }, { "grad_norm": 0.10219115018844604, "learning_rate": 5.6579857048264926e-05, "loss": 0.0035, "step": 48510 }, { "grad_norm": 0.0724710077047348, "learning_rate": 5.6563465788249314e-05, "loss": 0.004, "step": 48520 }, { "grad_norm": 0.05878334492444992, "learning_rate": 5.6547073810462956e-05, "loss": 0.0033, "step": 48530 }, { "grad_norm": 0.06823476403951645, "learning_rate": 5.653068111669846e-05, "loss": 0.0035, "step": 48540 }, { "grad_norm": 0.06889498233795166, "learning_rate": 5.651428770874848e-05, "loss": 0.0039, "step": 48550 }, { "grad_norm": 0.0644097700715065, "learning_rate": 5.64978935884058e-05, "loss": 0.0032, "step": 48560 }, { "grad_norm": 0.06689124554395676, "learning_rate": 5.6481498757463244e-05, "loss": 0.0033, "step": 48570 }, { "grad_norm": 0.08182081580162048, "learning_rate": 5.646510321771373e-05, "loss": 0.0038, "step": 48580 }, { "grad_norm": 0.09005296230316162, "learning_rate": 5.644870697095024e-05, "loss": 0.0044, "step": 48590 }, { "grad_norm": 0.0736415684223175, "learning_rate": 5.643231001896586e-05, "loss": 0.0036, "step": 48600 }, { "grad_norm": 0.06832461804151535, "learning_rate": 5.6415912363553726e-05, "loss": 0.004, "step": 48610 }, { "grad_norm": 0.06748457998037338, "learning_rate": 5.639951400650706e-05, "loss": 0.0038, "step": 48620 }, { "grad_norm": 0.07769463211297989, "learning_rate": 5.6383114949619165e-05, "loss": 0.0045, "step": 48630 }, { "grad_norm": 0.0772714614868164, "learning_rate": 5.636671519468342e-05, "loss": 0.0047, "step": 48640 }, { "grad_norm": 0.06334126740694046, "learning_rate": 5.635031474349327e-05, "loss": 0.0039, "step": 48650 }, { "grad_norm": 0.06001394987106323, "learning_rate": 5.6333913597842246e-05, "loss": 0.004, "step": 48660 }, { "grad_norm": 0.0575326643884182, "learning_rate": 5.6317511759523955e-05, "loss": 0.0041, "step": 48670 }, { "grad_norm": 0.07981747388839722, "learning_rate": 5.630110923033207e-05, "loss": 0.0045, "step": 48680 }, { "grad_norm": 0.0988420769572258, "learning_rate": 5.628470601206036e-05, "loss": 0.0046, "step": 48690 }, { "grad_norm": 0.11974965780973434, "learning_rate": 5.626830210650266e-05, "loss": 0.0039, "step": 48700 }, { "grad_norm": 0.07708338648080826, "learning_rate": 5.625189751545285e-05, "loss": 0.0037, "step": 48710 }, { "grad_norm": 0.047276441007852554, "learning_rate": 5.6235492240704936e-05, "loss": 0.0036, "step": 48720 }, { "grad_norm": 0.07722557336091995, "learning_rate": 5.621908628405296e-05, "loss": 0.0037, "step": 48730 }, { "grad_norm": 0.08200231939554214, "learning_rate": 5.620267964729106e-05, "loss": 0.0034, "step": 48740 }, { "grad_norm": 0.0741569995880127, "learning_rate": 5.618627233221344e-05, "loss": 0.0037, "step": 48750 }, { "grad_norm": 0.08304032683372498, "learning_rate": 5.616986434061438e-05, "loss": 0.0052, "step": 48760 }, { "grad_norm": 0.09837996959686279, "learning_rate": 5.615345567428822e-05, "loss": 0.0046, "step": 48770 }, { "grad_norm": 0.08372166007757187, "learning_rate": 5.613704633502941e-05, "loss": 0.0045, "step": 48780 }, { "grad_norm": 0.067410409450531, "learning_rate": 5.612063632463245e-05, "loss": 0.0043, "step": 48790 }, { "grad_norm": 0.09641712158918381, "learning_rate": 5.610422564489188e-05, "loss": 0.0046, "step": 48800 }, { "grad_norm": 0.07119255512952805, "learning_rate": 5.608781429760239e-05, "loss": 0.0038, "step": 48810 }, { "grad_norm": 0.06449298560619354, "learning_rate": 5.607140228455866e-05, "loss": 0.0031, "step": 48820 }, { "grad_norm": 0.056844789534807205, "learning_rate": 5.605498960755553e-05, "loss": 0.004, "step": 48830 }, { "grad_norm": 0.06448784470558167, "learning_rate": 5.603857626838782e-05, "loss": 0.0036, "step": 48840 }, { "grad_norm": 0.04560238867998123, "learning_rate": 5.60221622688505e-05, "loss": 0.003, "step": 48850 }, { "grad_norm": 0.16936373710632324, "learning_rate": 5.6005747610738565e-05, "loss": 0.0047, "step": 48860 }, { "grad_norm": 0.07889356464147568, "learning_rate": 5.59893322958471e-05, "loss": 0.0032, "step": 48870 }, { "grad_norm": 0.05289025604724884, "learning_rate": 5.5972916325971256e-05, "loss": 0.004, "step": 48880 }, { "grad_norm": 0.08968015015125275, "learning_rate": 5.595649970290628e-05, "loss": 0.0038, "step": 48890 }, { "grad_norm": 0.06241421401500702, "learning_rate": 5.5940082428447426e-05, "loss": 0.0051, "step": 48900 }, { "grad_norm": 0.07007057219743729, "learning_rate": 5.592366450439012e-05, "loss": 0.0036, "step": 48910 }, { "grad_norm": 0.06116761267185211, "learning_rate": 5.590724593252975e-05, "loss": 0.0043, "step": 48920 }, { "grad_norm": 0.0602462962269783, "learning_rate": 5.589082671466184e-05, "loss": 0.0029, "step": 48930 }, { "grad_norm": 0.06060589477419853, "learning_rate": 5.587440685258199e-05, "loss": 0.0037, "step": 48940 }, { "grad_norm": 0.07408536970615387, "learning_rate": 5.585798634808583e-05, "loss": 0.0039, "step": 48950 }, { "grad_norm": 0.08565906435251236, "learning_rate": 5.584156520296909e-05, "loss": 0.0035, "step": 48960 }, { "grad_norm": 0.06628966331481934, "learning_rate": 5.582514341902757e-05, "loss": 0.0041, "step": 48970 }, { "grad_norm": 0.07665491104125977, "learning_rate": 5.580872099805713e-05, "loss": 0.0043, "step": 48980 }, { "grad_norm": 0.09226469695568085, "learning_rate": 5.5792297941853674e-05, "loss": 0.0045, "step": 48990 }, { "grad_norm": 0.07418114691972733, "learning_rate": 5.5775874252213247e-05, "loss": 0.0038, "step": 49000 }, { "grad_norm": 0.0668967068195343, "learning_rate": 5.575944993093189e-05, "loss": 0.0041, "step": 49010 }, { "grad_norm": 0.06524614989757538, "learning_rate": 5.574302497980574e-05, "loss": 0.0042, "step": 49020 }, { "grad_norm": 0.07246392965316772, "learning_rate": 5.5726599400631005e-05, "loss": 0.0037, "step": 49030 }, { "grad_norm": 0.08184044808149338, "learning_rate": 5.571017319520399e-05, "loss": 0.0035, "step": 49040 }, { "grad_norm": 0.07213588058948517, "learning_rate": 5.5693746365320984e-05, "loss": 0.0032, "step": 49050 }, { "grad_norm": 0.045093391090631485, "learning_rate": 5.567731891277848e-05, "loss": 0.0034, "step": 49060 }, { "grad_norm": 0.08792953193187714, "learning_rate": 5.566089083937288e-05, "loss": 0.0031, "step": 49070 }, { "grad_norm": 0.0828268900513649, "learning_rate": 5.564446214690079e-05, "loss": 0.0033, "step": 49080 }, { "grad_norm": 0.05735655501484871, "learning_rate": 5.56280328371588e-05, "loss": 0.0038, "step": 49090 }, { "grad_norm": 0.09256639331579208, "learning_rate": 5.5611602911943596e-05, "loss": 0.0035, "step": 49100 }, { "grad_norm": 0.06744188815355301, "learning_rate": 5.559517237305194e-05, "loss": 0.0034, "step": 49110 }, { "grad_norm": 0.07124193757772446, "learning_rate": 5.557874122228064e-05, "loss": 0.0031, "step": 49120 }, { "grad_norm": 0.09867582470178604, "learning_rate": 5.55623094614266e-05, "loss": 0.0036, "step": 49130 }, { "grad_norm": 0.053525570780038834, "learning_rate": 5.5545877092286743e-05, "loss": 0.0033, "step": 49140 }, { "grad_norm": 0.08371065557003021, "learning_rate": 5.552944411665811e-05, "loss": 0.0041, "step": 49150 }, { "grad_norm": 0.06844134628772736, "learning_rate": 5.5513010536337795e-05, "loss": 0.0033, "step": 49160 }, { "grad_norm": 0.05974738299846649, "learning_rate": 5.5496576353122934e-05, "loss": 0.0033, "step": 49170 }, { "grad_norm": 0.06484295427799225, "learning_rate": 5.548014156881074e-05, "loss": 0.0042, "step": 49180 }, { "grad_norm": 0.11342695355415344, "learning_rate": 5.5463706185198494e-05, "loss": 0.0033, "step": 49190 }, { "grad_norm": 0.068067766726017, "learning_rate": 5.544727020408358e-05, "loss": 0.0046, "step": 49200 }, { "grad_norm": 0.08228372782468796, "learning_rate": 5.5430833627263366e-05, "loss": 0.0046, "step": 49210 }, { "grad_norm": 0.08631081879138947, "learning_rate": 5.5414396456535354e-05, "loss": 0.0039, "step": 49220 }, { "grad_norm": 0.04854148253798485, "learning_rate": 5.539795869369709e-05, "loss": 0.0031, "step": 49230 }, { "grad_norm": 0.05202487111091614, "learning_rate": 5.538152034054618e-05, "loss": 0.0036, "step": 49240 }, { "grad_norm": 0.07482946664094925, "learning_rate": 5.536508139888028e-05, "loss": 0.005, "step": 49250 }, { "grad_norm": 0.06265939772129059, "learning_rate": 5.534864187049716e-05, "loss": 0.004, "step": 49260 }, { "grad_norm": 0.06357700377702713, "learning_rate": 5.5332201757194615e-05, "loss": 0.0046, "step": 49270 }, { "grad_norm": 0.06490597128868103, "learning_rate": 5.531576106077049e-05, "loss": 0.0038, "step": 49280 }, { "grad_norm": 0.08026639372110367, "learning_rate": 5.529931978302272e-05, "loss": 0.0039, "step": 49290 }, { "grad_norm": 0.07832027971744537, "learning_rate": 5.528287792574932e-05, "loss": 0.0037, "step": 49300 }, { "grad_norm": 0.060089871287345886, "learning_rate": 5.5266435490748324e-05, "loss": 0.0043, "step": 49310 }, { "grad_norm": 0.056866854429244995, "learning_rate": 5.524999247981787e-05, "loss": 0.0033, "step": 49320 }, { "grad_norm": 0.05937570333480835, "learning_rate": 5.523354889475613e-05, "loss": 0.0044, "step": 49330 }, { "grad_norm": 0.08340562880039215, "learning_rate": 5.521710473736134e-05, "loss": 0.004, "step": 49340 }, { "grad_norm": 0.04976293817162514, "learning_rate": 5.5200660009431826e-05, "loss": 0.0037, "step": 49350 }, { "grad_norm": 0.06501973420381546, "learning_rate": 5.518421471276596e-05, "loss": 0.0033, "step": 49360 }, { "grad_norm": 0.07492358982563019, "learning_rate": 5.5167768849162174e-05, "loss": 0.003, "step": 49370 }, { "grad_norm": 0.05865941196680069, "learning_rate": 5.515132242041893e-05, "loss": 0.0042, "step": 49380 }, { "grad_norm": 0.05971839278936386, "learning_rate": 5.513487542833483e-05, "loss": 0.0028, "step": 49390 }, { "grad_norm": 0.08208318054676056, "learning_rate": 5.5118427874708466e-05, "loss": 0.0035, "step": 49400 }, { "grad_norm": 0.08270739763975143, "learning_rate": 5.510197976133853e-05, "loss": 0.0035, "step": 49410 }, { "grad_norm": 0.06442738324403763, "learning_rate": 5.508553109002376e-05, "loss": 0.004, "step": 49420 }, { "grad_norm": 0.057379305362701416, "learning_rate": 5.5069081862562957e-05, "loss": 0.0034, "step": 49430 }, { "grad_norm": 0.0646725594997406, "learning_rate": 5.5052632080754965e-05, "loss": 0.0037, "step": 49440 }, { "grad_norm": 0.06433103233575821, "learning_rate": 5.503618174639874e-05, "loss": 0.0034, "step": 49450 }, { "grad_norm": 0.059838876128196716, "learning_rate": 5.501973086129325e-05, "loss": 0.0038, "step": 49460 }, { "grad_norm": 0.0826338455080986, "learning_rate": 5.500327942723753e-05, "loss": 0.0036, "step": 49470 }, { "grad_norm": 0.08660139888525009, "learning_rate": 5.498682744603071e-05, "loss": 0.0039, "step": 49480 }, { "grad_norm": 0.0805753767490387, "learning_rate": 5.497037491947194e-05, "loss": 0.0043, "step": 49490 }, { "grad_norm": 0.07646232098340988, "learning_rate": 5.4953921849360424e-05, "loss": 0.0031, "step": 49500 }, { "grad_norm": 0.0711153894662857, "learning_rate": 5.493746823749547e-05, "loss": 0.0037, "step": 49510 }, { "grad_norm": 0.07588846236467361, "learning_rate": 5.49210140856764e-05, "loss": 0.0036, "step": 49520 }, { "grad_norm": 0.06596280634403229, "learning_rate": 5.4904559395702647e-05, "loss": 0.0031, "step": 49530 }, { "grad_norm": 0.07951243221759796, "learning_rate": 5.488810416937364e-05, "loss": 0.0035, "step": 49540 }, { "grad_norm": 0.08013555407524109, "learning_rate": 5.4871648408488926e-05, "loss": 0.0035, "step": 49550 }, { "grad_norm": 0.06484246999025345, "learning_rate": 5.485519211484807e-05, "loss": 0.0041, "step": 49560 }, { "grad_norm": 0.07089251279830933, "learning_rate": 5.4838735290250676e-05, "loss": 0.0036, "step": 49570 }, { "grad_norm": 0.06283114105463028, "learning_rate": 5.482227793649649e-05, "loss": 0.0037, "step": 49580 }, { "grad_norm": 0.06387211382389069, "learning_rate": 5.480582005538524e-05, "loss": 0.0033, "step": 49590 }, { "grad_norm": 0.060935553163290024, "learning_rate": 5.478936164871671e-05, "loss": 0.0044, "step": 49600 }, { "grad_norm": 0.05937965586781502, "learning_rate": 5.477290271829081e-05, "loss": 0.0034, "step": 49610 }, { "grad_norm": 0.07552070915699005, "learning_rate": 5.475644326590744e-05, "loss": 0.0052, "step": 49620 }, { "grad_norm": 0.10954271256923676, "learning_rate": 5.473998329336658e-05, "loss": 0.0037, "step": 49630 }, { "grad_norm": 0.06721417605876923, "learning_rate": 5.4723522802468286e-05, "loss": 0.0032, "step": 49640 }, { "grad_norm": 0.06466427445411682, "learning_rate": 5.470706179501264e-05, "loss": 0.0039, "step": 49650 }, { "grad_norm": 0.06532682478427887, "learning_rate": 5.4690600272799776e-05, "loss": 0.0038, "step": 49660 }, { "grad_norm": 0.06588335335254669, "learning_rate": 5.467413823762993e-05, "loss": 0.0038, "step": 49670 }, { "grad_norm": 0.06533649563789368, "learning_rate": 5.465767569130335e-05, "loss": 0.0034, "step": 49680 }, { "grad_norm": 0.07811814546585083, "learning_rate": 5.464121263562036e-05, "loss": 0.0031, "step": 49690 }, { "grad_norm": 0.08351100236177444, "learning_rate": 5.4624749072381343e-05, "loss": 0.0037, "step": 49700 }, { "grad_norm": 0.0640532523393631, "learning_rate": 5.460828500338672e-05, "loss": 0.0041, "step": 49710 }, { "grad_norm": 0.0765293687582016, "learning_rate": 5.459182043043698e-05, "loss": 0.0035, "step": 49720 }, { "grad_norm": 0.07618417590856552, "learning_rate": 5.457535535533265e-05, "loss": 0.0045, "step": 49730 }, { "grad_norm": 0.06937538832426071, "learning_rate": 5.4558889779874334e-05, "loss": 0.0031, "step": 49740 }, { "grad_norm": 0.06480446457862854, "learning_rate": 5.454242370586269e-05, "loss": 0.0038, "step": 49750 }, { "grad_norm": 0.057597409933805466, "learning_rate": 5.452595713509843e-05, "loss": 0.0034, "step": 49760 }, { "grad_norm": 0.07384621351957321, "learning_rate": 5.4509490069382274e-05, "loss": 0.0045, "step": 49770 }, { "grad_norm": 0.10680857300758362, "learning_rate": 5.449302251051509e-05, "loss": 0.0035, "step": 49780 }, { "grad_norm": 0.09090961515903473, "learning_rate": 5.44765544602977e-05, "loss": 0.0042, "step": 49790 }, { "grad_norm": 0.06834731996059418, "learning_rate": 5.446008592053107e-05, "loss": 0.0035, "step": 49800 }, { "grad_norm": 0.12508347630500793, "learning_rate": 5.4443616893016135e-05, "loss": 0.0057, "step": 49810 }, { "grad_norm": 0.10095575451850891, "learning_rate": 5.4427147379553934e-05, "loss": 0.0036, "step": 49820 }, { "grad_norm": 0.09656422585248947, "learning_rate": 5.441067738194556e-05, "loss": 0.0033, "step": 49830 }, { "grad_norm": 0.07435194402933121, "learning_rate": 5.439420690199214e-05, "loss": 0.0032, "step": 49840 }, { "grad_norm": 0.05912599340081215, "learning_rate": 5.437773594149484e-05, "loss": 0.0031, "step": 49850 }, { "grad_norm": 0.06885885447263718, "learning_rate": 5.436126450225495e-05, "loss": 0.0032, "step": 49860 }, { "grad_norm": 0.057439323514699936, "learning_rate": 5.434479258607371e-05, "loss": 0.0031, "step": 49870 }, { "grad_norm": 0.09202609211206436, "learning_rate": 5.43283201947525e-05, "loss": 0.0029, "step": 49880 }, { "grad_norm": 0.057380352169275284, "learning_rate": 5.431184733009268e-05, "loss": 0.0037, "step": 49890 }, { "grad_norm": 0.07829853147268295, "learning_rate": 5.4295373993895736e-05, "loss": 0.0035, "step": 49900 }, { "grad_norm": 0.06367641687393188, "learning_rate": 5.4278900187963157e-05, "loss": 0.0042, "step": 49910 }, { "grad_norm": 0.07984942942857742, "learning_rate": 5.426242591409646e-05, "loss": 0.0039, "step": 49920 }, { "grad_norm": 0.06381844729185104, "learning_rate": 5.42459511740973e-05, "loss": 0.0034, "step": 49930 }, { "grad_norm": 0.11364250630140305, "learning_rate": 5.422947596976729e-05, "loss": 0.0031, "step": 49940 }, { "grad_norm": 0.0990348607301712, "learning_rate": 5.4213000302908134e-05, "loss": 0.0033, "step": 49950 }, { "grad_norm": 0.06788288801908493, "learning_rate": 5.419652417532162e-05, "loss": 0.0032, "step": 49960 }, { "grad_norm": 0.06361274421215057, "learning_rate": 5.4180047588809534e-05, "loss": 0.003, "step": 49970 }, { "grad_norm": 0.08557054400444031, "learning_rate": 5.4163570545173704e-05, "loss": 0.0035, "step": 49980 }, { "grad_norm": 0.05853649228811264, "learning_rate": 5.414709304621608e-05, "loss": 0.0032, "step": 49990 }, { "grad_norm": 0.07017309218645096, "learning_rate": 5.413061509373858e-05, "loss": 0.0033, "step": 50000 }, { "grad_norm": 0.054665517061948776, "learning_rate": 5.411413668954324e-05, "loss": 0.0029, "step": 50010 }, { "grad_norm": 0.057707421481609344, "learning_rate": 5.4097657835432083e-05, "loss": 0.0033, "step": 50020 }, { "grad_norm": 0.056612804532051086, "learning_rate": 5.408117853320723e-05, "loss": 0.0034, "step": 50030 }, { "grad_norm": 0.060435034334659576, "learning_rate": 5.406469878467084e-05, "loss": 0.0033, "step": 50040 }, { "grad_norm": 0.07088866829872131, "learning_rate": 5.404821859162509e-05, "loss": 0.0033, "step": 50050 }, { "grad_norm": 0.08826068788766861, "learning_rate": 5.403173795587225e-05, "loss": 0.0033, "step": 50060 }, { "grad_norm": 0.06310148537158966, "learning_rate": 5.4015256879214606e-05, "loss": 0.0029, "step": 50070 }, { "grad_norm": 0.06697944551706314, "learning_rate": 5.39987753634545e-05, "loss": 0.0037, "step": 50080 }, { "grad_norm": 0.06366538256406784, "learning_rate": 5.3982293410394335e-05, "loss": 0.0032, "step": 50090 }, { "grad_norm": 0.07633920758962631, "learning_rate": 5.396581102183655e-05, "loss": 0.0037, "step": 50100 }, { "grad_norm": 0.07721730321645737, "learning_rate": 5.3949328199583634e-05, "loss": 0.0051, "step": 50110 }, { "grad_norm": 0.07052392512559891, "learning_rate": 5.393284494543813e-05, "loss": 0.0036, "step": 50120 }, { "grad_norm": 0.06211366504430771, "learning_rate": 5.391636126120262e-05, "loss": 0.0036, "step": 50130 }, { "grad_norm": 0.08197450637817383, "learning_rate": 5.389987714867971e-05, "loss": 0.0036, "step": 50140 }, { "grad_norm": 0.06796733289957047, "learning_rate": 5.3883392609672114e-05, "loss": 0.0034, "step": 50150 }, { "grad_norm": 0.07603476941585541, "learning_rate": 5.3866907645982545e-05, "loss": 0.0034, "step": 50160 }, { "grad_norm": 0.06292802095413208, "learning_rate": 5.385042225941375e-05, "loss": 0.0029, "step": 50170 }, { "grad_norm": 0.04306085407733917, "learning_rate": 5.3833936451768585e-05, "loss": 0.003, "step": 50180 }, { "grad_norm": 0.05487990751862526, "learning_rate": 5.3817450224849884e-05, "loss": 0.0029, "step": 50190 }, { "grad_norm": 0.08091352880001068, "learning_rate": 5.380096358046056e-05, "loss": 0.0034, "step": 50200 }, { "grad_norm": 0.07248193025588989, "learning_rate": 5.378447652040359e-05, "loss": 0.004, "step": 50210 }, { "grad_norm": 0.060144513845443726, "learning_rate": 5.376798904648195e-05, "loss": 0.0043, "step": 50220 }, { "grad_norm": 0.07755101472139359, "learning_rate": 5.375150116049869e-05, "loss": 0.0041, "step": 50230 }, { "grad_norm": 0.08008889108896255, "learning_rate": 5.373501286425691e-05, "loss": 0.004, "step": 50240 }, { "grad_norm": 0.05646444857120514, "learning_rate": 5.3718524159559726e-05, "loss": 0.0033, "step": 50250 }, { "grad_norm": 0.056898366659879684, "learning_rate": 5.370203504821034e-05, "loss": 0.004, "step": 50260 }, { "grad_norm": 0.06379275023937225, "learning_rate": 5.368554553201196e-05, "loss": 0.0038, "step": 50270 }, { "grad_norm": 0.0846458300948143, "learning_rate": 5.3669055612767874e-05, "loss": 0.0036, "step": 50280 }, { "grad_norm": 0.07813076674938202, "learning_rate": 5.3652565292281376e-05, "loss": 0.0034, "step": 50290 }, { "grad_norm": 0.17035089433193207, "learning_rate": 5.3636074572355824e-05, "loss": 0.0047, "step": 50300 }, { "grad_norm": 0.062361616641283035, "learning_rate": 5.361958345479463e-05, "loss": 0.0036, "step": 50310 }, { "grad_norm": 0.07337586581707001, "learning_rate": 5.360309194140123e-05, "loss": 0.0039, "step": 50320 }, { "grad_norm": 0.05777941644191742, "learning_rate": 5.3586600033979106e-05, "loss": 0.0029, "step": 50330 }, { "grad_norm": 0.058528460562229156, "learning_rate": 5.357010773433181e-05, "loss": 0.003, "step": 50340 }, { "grad_norm": 0.0700049102306366, "learning_rate": 5.3553615044262895e-05, "loss": 0.0037, "step": 50350 }, { "grad_norm": 0.08179289102554321, "learning_rate": 5.3537121965575964e-05, "loss": 0.0036, "step": 50360 }, { "grad_norm": 0.06533723324537277, "learning_rate": 5.352062850007471e-05, "loss": 0.0035, "step": 50370 }, { "grad_norm": 0.07366176694631577, "learning_rate": 5.350413464956282e-05, "loss": 0.0036, "step": 50380 }, { "grad_norm": 0.07962571829557419, "learning_rate": 5.348764041584403e-05, "loss": 0.0041, "step": 50390 }, { "grad_norm": 0.05508190393447876, "learning_rate": 5.347114580072212e-05, "loss": 0.0042, "step": 50400 }, { "grad_norm": 0.056895673274993896, "learning_rate": 5.345465080600093e-05, "loss": 0.0033, "step": 50410 }, { "grad_norm": 0.07272933423519135, "learning_rate": 5.3438155433484326e-05, "loss": 0.0036, "step": 50420 }, { "grad_norm": 0.08296875655651093, "learning_rate": 5.3421659684976197e-05, "loss": 0.0035, "step": 50430 }, { "grad_norm": 0.07785958796739578, "learning_rate": 5.340516356228052e-05, "loss": 0.0039, "step": 50440 }, { "grad_norm": 0.06923791766166687, "learning_rate": 5.338866706720128e-05, "loss": 0.0037, "step": 50450 }, { "grad_norm": 0.08670848608016968, "learning_rate": 5.337217020154249e-05, "loss": 0.004, "step": 50460 }, { "grad_norm": 0.07010604441165924, "learning_rate": 5.335567296710825e-05, "loss": 0.0036, "step": 50470 }, { "grad_norm": 0.07040243595838547, "learning_rate": 5.333917536570265e-05, "loss": 0.0045, "step": 50480 }, { "grad_norm": 0.06922385841608047, "learning_rate": 5.332267739912986e-05, "loss": 0.0036, "step": 50490 }, { "grad_norm": 0.06341268122196198, "learning_rate": 5.330617906919405e-05, "loss": 0.004, "step": 50500 }, { "grad_norm": 0.06444834172725677, "learning_rate": 5.328968037769949e-05, "loss": 0.0032, "step": 50510 }, { "grad_norm": 0.08829896152019501, "learning_rate": 5.327318132645042e-05, "loss": 0.0031, "step": 50520 }, { "grad_norm": 0.09328418225049973, "learning_rate": 5.325668191725116e-05, "loss": 0.0041, "step": 50530 }, { "grad_norm": 0.10202909260988235, "learning_rate": 5.324018215190606e-05, "loss": 0.0072, "step": 50540 }, { "grad_norm": 0.10866667330265045, "learning_rate": 5.3223682032219515e-05, "loss": 0.0056, "step": 50550 }, { "grad_norm": 0.07066420465707779, "learning_rate": 5.320718155999595e-05, "loss": 0.0046, "step": 50560 }, { "grad_norm": 0.07978112995624542, "learning_rate": 5.3190680737039835e-05, "loss": 0.0042, "step": 50570 }, { "grad_norm": 0.05041857808828354, "learning_rate": 5.317417956515567e-05, "loss": 0.004, "step": 50580 }, { "grad_norm": 0.0718233734369278, "learning_rate": 5.315767804614798e-05, "loss": 0.0041, "step": 50590 }, { "grad_norm": 0.10119164735078812, "learning_rate": 5.3141176181821395e-05, "loss": 0.0043, "step": 50600 }, { "grad_norm": 0.07930824905633926, "learning_rate": 5.31246739739805e-05, "loss": 0.0043, "step": 50610 }, { "grad_norm": 0.06169923394918442, "learning_rate": 5.310817142442994e-05, "loss": 0.0036, "step": 50620 }, { "grad_norm": 0.07621653378009796, "learning_rate": 5.309166853497445e-05, "loss": 0.0043, "step": 50630 }, { "grad_norm": 0.07257042080163956, "learning_rate": 5.307516530741873e-05, "loss": 0.0042, "step": 50640 }, { "grad_norm": 0.07987785339355469, "learning_rate": 5.305866174356754e-05, "loss": 0.0046, "step": 50650 }, { "grad_norm": 0.1165158748626709, "learning_rate": 5.304215784522571e-05, "loss": 0.0043, "step": 50660 }, { "grad_norm": 0.07261300086975098, "learning_rate": 5.302565361419808e-05, "loss": 0.0034, "step": 50670 }, { "grad_norm": 0.0809115320444107, "learning_rate": 5.3009149052289507e-05, "loss": 0.0036, "step": 50680 }, { "grad_norm": 0.07205052673816681, "learning_rate": 5.299264416130493e-05, "loss": 0.0048, "step": 50690 }, { "grad_norm": 0.07583373039960861, "learning_rate": 5.297613894304928e-05, "loss": 0.0038, "step": 50700 }, { "grad_norm": 0.0712391808629036, "learning_rate": 5.2959633399327534e-05, "loss": 0.0041, "step": 50710 }, { "grad_norm": 0.0717354416847229, "learning_rate": 5.294312753194476e-05, "loss": 0.0038, "step": 50720 }, { "grad_norm": 0.055430393666028976, "learning_rate": 5.292662134270596e-05, "loss": 0.004, "step": 50730 }, { "grad_norm": 0.06918247789144516, "learning_rate": 5.291011483341626e-05, "loss": 0.0039, "step": 50740 }, { "grad_norm": 0.07605884224176407, "learning_rate": 5.2893608005880767e-05, "loss": 0.0034, "step": 50750 }, { "grad_norm": 0.06544990837574005, "learning_rate": 5.287710086190467e-05, "loss": 0.0033, "step": 50760 }, { "grad_norm": 0.06248186156153679, "learning_rate": 5.286059340329314e-05, "loss": 0.0035, "step": 50770 }, { "grad_norm": 0.07418447732925415, "learning_rate": 5.284408563185141e-05, "loss": 0.0034, "step": 50780 }, { "grad_norm": 0.08605924993753433, "learning_rate": 5.282757754938476e-05, "loss": 0.0035, "step": 50790 }, { "grad_norm": 0.06730370223522186, "learning_rate": 5.281106915769849e-05, "loss": 0.0036, "step": 50800 }, { "grad_norm": 0.08675781637430191, "learning_rate": 5.2794560458597897e-05, "loss": 0.0042, "step": 50810 }, { "grad_norm": 0.06306341290473938, "learning_rate": 5.27780514538884e-05, "loss": 0.003, "step": 50820 }, { "grad_norm": 0.0804428979754448, "learning_rate": 5.2761542145375365e-05, "loss": 0.0046, "step": 50830 }, { "grad_norm": 0.06466881185770035, "learning_rate": 5.274503253486421e-05, "loss": 0.003, "step": 50840 }, { "grad_norm": 0.0683787539601326, "learning_rate": 5.272852262416046e-05, "loss": 0.0041, "step": 50850 }, { "grad_norm": 0.06682602316141129, "learning_rate": 5.2712012415069555e-05, "loss": 0.0052, "step": 50860 }, { "grad_norm": 0.057580702006816864, "learning_rate": 5.269550190939705e-05, "loss": 0.0031, "step": 50870 }, { "grad_norm": 0.06324562430381775, "learning_rate": 5.267899110894852e-05, "loss": 0.0042, "step": 50880 }, { "grad_norm": 0.07374951243400574, "learning_rate": 5.266248001552955e-05, "loss": 0.0037, "step": 50890 }, { "grad_norm": 0.07855379581451416, "learning_rate": 5.264596863094575e-05, "loss": 0.0044, "step": 50900 }, { "grad_norm": 0.07060490548610687, "learning_rate": 5.262945695700282e-05, "loss": 0.0032, "step": 50910 }, { "grad_norm": 0.07795450836420059, "learning_rate": 5.261294499550643e-05, "loss": 0.0041, "step": 50920 }, { "grad_norm": 0.05470183119177818, "learning_rate": 5.25964327482623e-05, "loss": 0.0037, "step": 50930 }, { "grad_norm": 0.07515304535627365, "learning_rate": 5.257992021707617e-05, "loss": 0.004, "step": 50940 }, { "grad_norm": 0.06759148091077805, "learning_rate": 5.256340740375387e-05, "loss": 0.0032, "step": 50950 }, { "grad_norm": 0.11291258037090302, "learning_rate": 5.254689431010117e-05, "loss": 0.0033, "step": 50960 }, { "grad_norm": 0.0795724019408226, "learning_rate": 5.253038093792395e-05, "loss": 0.004, "step": 50970 }, { "grad_norm": 0.08219140768051147, "learning_rate": 5.251386728902806e-05, "loss": 0.0033, "step": 50980 }, { "grad_norm": 0.06149091199040413, "learning_rate": 5.2497353365219446e-05, "loss": 0.0039, "step": 50990 }, { "grad_norm": 0.06310595571994781, "learning_rate": 5.2480839168304e-05, "loss": 0.0039, "step": 51000 }, { "grad_norm": 0.06979726999998093, "learning_rate": 5.246432470008773e-05, "loss": 0.0035, "step": 51010 }, { "grad_norm": 0.09092658013105392, "learning_rate": 5.2447809962376616e-05, "loss": 0.0053, "step": 51020 }, { "grad_norm": 0.08052439242601395, "learning_rate": 5.243129495697666e-05, "loss": 0.0033, "step": 51030 }, { "grad_norm": 0.08320886641740799, "learning_rate": 5.241477968569397e-05, "loss": 0.0037, "step": 51040 }, { "grad_norm": 0.06598082929849625, "learning_rate": 5.23982641503346e-05, "loss": 0.005, "step": 51050 }, { "grad_norm": 0.05494409054517746, "learning_rate": 5.238174835270465e-05, "loss": 0.0045, "step": 51060 }, { "grad_norm": 0.08345332741737366, "learning_rate": 5.23652322946103e-05, "loss": 0.004, "step": 51070 }, { "grad_norm": 0.05516309663653374, "learning_rate": 5.234871597785771e-05, "loss": 0.0035, "step": 51080 }, { "grad_norm": 0.07520344853401184, "learning_rate": 5.233219940425306e-05, "loss": 0.0034, "step": 51090 }, { "grad_norm": 0.07826505601406097, "learning_rate": 5.2315682575602576e-05, "loss": 0.0036, "step": 51100 }, { "grad_norm": 0.08553126454353333, "learning_rate": 5.229916549371254e-05, "loss": 0.0036, "step": 51110 }, { "grad_norm": 0.09469561278820038, "learning_rate": 5.228264816038924e-05, "loss": 0.0037, "step": 51120 }, { "grad_norm": 0.0642663910984993, "learning_rate": 5.2266130577438935e-05, "loss": 0.0041, "step": 51130 }, { "grad_norm": 0.058116599917411804, "learning_rate": 5.224961274666802e-05, "loss": 0.0037, "step": 51140 }, { "grad_norm": 0.08550377935171127, "learning_rate": 5.223309466988283e-05, "loss": 0.0038, "step": 51150 }, { "grad_norm": 0.07917659729719162, "learning_rate": 5.221657634888976e-05, "loss": 0.0036, "step": 51160 }, { "grad_norm": 0.11529430001974106, "learning_rate": 5.2200057785495235e-05, "loss": 0.0039, "step": 51170 }, { "grad_norm": 0.06375934183597565, "learning_rate": 5.2183538981505706e-05, "loss": 0.0032, "step": 51180 }, { "grad_norm": 0.07376255840063095, "learning_rate": 5.216701993872762e-05, "loss": 0.0032, "step": 51190 }, { "grad_norm": 0.07522693276405334, "learning_rate": 5.215050065896752e-05, "loss": 0.0035, "step": 51200 }, { "grad_norm": 0.07620932161808014, "learning_rate": 5.2133981144031875e-05, "loss": 0.0037, "step": 51210 }, { "grad_norm": 0.05983567610383034, "learning_rate": 5.211746139572727e-05, "loss": 0.0032, "step": 51220 }, { "grad_norm": 0.07510463148355484, "learning_rate": 5.210094141586026e-05, "loss": 0.0034, "step": 51230 }, { "grad_norm": 0.07772216200828552, "learning_rate": 5.2084421206237476e-05, "loss": 0.0034, "step": 51240 }, { "grad_norm": 0.09706995636224747, "learning_rate": 5.20679007686655e-05, "loss": 0.0043, "step": 51250 }, { "grad_norm": 0.06595874577760696, "learning_rate": 5.205138010495102e-05, "loss": 0.0038, "step": 51260 }, { "grad_norm": 0.09244339913129807, "learning_rate": 5.20348592169007e-05, "loss": 0.0036, "step": 51270 }, { "grad_norm": 0.06734159588813782, "learning_rate": 5.201833810632123e-05, "loss": 0.0038, "step": 51280 }, { "grad_norm": 0.058210588991642, "learning_rate": 5.200181677501933e-05, "loss": 0.0031, "step": 51290 }, { "grad_norm": 0.06841132044792175, "learning_rate": 5.1985295224801777e-05, "loss": 0.0037, "step": 51300 }, { "grad_norm": 0.06850928068161011, "learning_rate": 5.196877345747531e-05, "loss": 0.0042, "step": 51310 }, { "grad_norm": 0.0950762927532196, "learning_rate": 5.195225147484674e-05, "loss": 0.0036, "step": 51320 }, { "grad_norm": 0.07526274025440216, "learning_rate": 5.193572927872288e-05, "loss": 0.0041, "step": 51330 }, { "grad_norm": 0.06885271519422531, "learning_rate": 5.191920687091059e-05, "loss": 0.003, "step": 51340 }, { "grad_norm": 0.06508543342351913, "learning_rate": 5.1902684253216695e-05, "loss": 0.0035, "step": 51350 }, { "grad_norm": 0.0676957443356514, "learning_rate": 5.1886161427448134e-05, "loss": 0.0026, "step": 51360 }, { "grad_norm": 0.05774585157632828, "learning_rate": 5.186963839541179e-05, "loss": 0.0034, "step": 51370 }, { "grad_norm": 0.06718704849481583, "learning_rate": 5.185311515891459e-05, "loss": 0.0035, "step": 51380 }, { "grad_norm": 0.1073315367102623, "learning_rate": 5.1836591719763516e-05, "loss": 0.004, "step": 51390 }, { "grad_norm": 0.06992622464895248, "learning_rate": 5.182006807976554e-05, "loss": 0.0035, "step": 51400 }, { "grad_norm": 0.06812018901109695, "learning_rate": 5.180354424072763e-05, "loss": 0.0033, "step": 51410 }, { "grad_norm": 0.08330196887254715, "learning_rate": 5.178702020445686e-05, "loss": 0.0029, "step": 51420 }, { "grad_norm": 0.06100519001483917, "learning_rate": 5.177049597276024e-05, "loss": 0.0029, "step": 51430 }, { "grad_norm": 0.0782499834895134, "learning_rate": 5.175397154744482e-05, "loss": 0.0034, "step": 51440 }, { "grad_norm": 0.06621945649385452, "learning_rate": 5.1737446930317736e-05, "loss": 0.0029, "step": 51450 }, { "grad_norm": 0.0804300457239151, "learning_rate": 5.172092212318604e-05, "loss": 0.0036, "step": 51460 }, { "grad_norm": 0.05679449066519737, "learning_rate": 5.170439712785692e-05, "loss": 0.003, "step": 51470 }, { "grad_norm": 0.05708099901676178, "learning_rate": 5.1687871946137467e-05, "loss": 0.0038, "step": 51480 }, { "grad_norm": 0.07336435467004776, "learning_rate": 5.16713465798349e-05, "loss": 0.0035, "step": 51490 }, { "grad_norm": 0.0465201660990715, "learning_rate": 5.165482103075637e-05, "loss": 0.0032, "step": 51500 }, { "grad_norm": 0.07687029987573624, "learning_rate": 5.163829530070909e-05, "loss": 0.0031, "step": 51510 }, { "grad_norm": 0.08335026353597641, "learning_rate": 5.162176939150032e-05, "loss": 0.0034, "step": 51520 }, { "grad_norm": 0.08519884198904037, "learning_rate": 5.160524330493728e-05, "loss": 0.0041, "step": 51530 }, { "grad_norm": 0.1103425920009613, "learning_rate": 5.158871704282724e-05, "loss": 0.0038, "step": 51540 }, { "grad_norm": 0.09899572283029556, "learning_rate": 5.15721906069775e-05, "loss": 0.0035, "step": 51550 }, { "grad_norm": 0.07102768868207932, "learning_rate": 5.155566399919536e-05, "loss": 0.0032, "step": 51560 }, { "grad_norm": 0.06184035912156105, "learning_rate": 5.153913722128813e-05, "loss": 0.0029, "step": 51570 }, { "grad_norm": 0.06636474281549454, "learning_rate": 5.1522610275063176e-05, "loss": 0.0037, "step": 51580 }, { "grad_norm": 0.07207480818033218, "learning_rate": 5.150608316232786e-05, "loss": 0.0032, "step": 51590 }, { "grad_norm": 0.07015467435121536, "learning_rate": 5.148955588488955e-05, "loss": 0.004, "step": 51600 }, { "grad_norm": 0.06934670358896255, "learning_rate": 5.147302844455564e-05, "loss": 0.0031, "step": 51610 }, { "grad_norm": 0.04685693979263306, "learning_rate": 5.145650084313356e-05, "loss": 0.0032, "step": 51620 }, { "grad_norm": 0.06260088086128235, "learning_rate": 5.1439973082430734e-05, "loss": 0.0035, "step": 51630 }, { "grad_norm": 0.08775761723518372, "learning_rate": 5.142344516425461e-05, "loss": 0.0035, "step": 51640 }, { "grad_norm": 0.06390228867530823, "learning_rate": 5.1406917090412673e-05, "loss": 0.0037, "step": 51650 }, { "grad_norm": 0.06216610223054886, "learning_rate": 5.1390388862712404e-05, "loss": 0.0028, "step": 51660 }, { "grad_norm": 0.06857294589281082, "learning_rate": 5.137386048296129e-05, "loss": 0.0041, "step": 51670 }, { "grad_norm": 0.06977370381355286, "learning_rate": 5.135733195296686e-05, "loss": 0.0029, "step": 51680 }, { "grad_norm": 0.0568820983171463, "learning_rate": 5.134080327453664e-05, "loss": 0.0033, "step": 51690 }, { "grad_norm": 0.0577574223279953, "learning_rate": 5.1324274449478204e-05, "loss": 0.0029, "step": 51700 }, { "grad_norm": 0.053509753197431564, "learning_rate": 5.130774547959909e-05, "loss": 0.0035, "step": 51710 }, { "grad_norm": 0.09661770612001419, "learning_rate": 5.129121636670692e-05, "loss": 0.0033, "step": 51720 }, { "grad_norm": 0.07913301140069962, "learning_rate": 5.1274687112609246e-05, "loss": 0.0035, "step": 51730 }, { "grad_norm": 0.07721615582704544, "learning_rate": 5.125815771911372e-05, "loss": 0.0036, "step": 51740 }, { "grad_norm": 0.06709491461515427, "learning_rate": 5.1241628188027956e-05, "loss": 0.0035, "step": 51750 }, { "grad_norm": 0.05949747934937477, "learning_rate": 5.1225098521159585e-05, "loss": 0.0028, "step": 51760 }, { "grad_norm": 0.06442698836326599, "learning_rate": 5.12085687203163e-05, "loss": 0.0034, "step": 51770 }, { "grad_norm": 0.06961604207754135, "learning_rate": 5.119203878730575e-05, "loss": 0.0034, "step": 51780 }, { "grad_norm": 0.06600189954042435, "learning_rate": 5.117550872393564e-05, "loss": 0.0033, "step": 51790 }, { "grad_norm": 0.06089791655540466, "learning_rate": 5.1158978532013645e-05, "loss": 0.0029, "step": 51800 }, { "grad_norm": 0.06437716633081436, "learning_rate": 5.114244821334751e-05, "loss": 0.0032, "step": 51810 }, { "grad_norm": 0.07626593112945557, "learning_rate": 5.112591776974497e-05, "loss": 0.0035, "step": 51820 }, { "grad_norm": 0.08131129294633865, "learning_rate": 5.110938720301374e-05, "loss": 0.0032, "step": 51830 }, { "grad_norm": 0.06866994500160217, "learning_rate": 5.1092856514961595e-05, "loss": 0.0042, "step": 51840 }, { "grad_norm": 0.06092849001288414, "learning_rate": 5.107632570739631e-05, "loss": 0.0031, "step": 51850 }, { "grad_norm": 0.07104381918907166, "learning_rate": 5.1059794782125645e-05, "loss": 0.0038, "step": 51860 }, { "grad_norm": 0.09029772132635117, "learning_rate": 5.104326374095744e-05, "loss": 0.0039, "step": 51870 }, { "grad_norm": 0.06207127496600151, "learning_rate": 5.102673258569947e-05, "loss": 0.0029, "step": 51880 }, { "grad_norm": 0.08369401097297668, "learning_rate": 5.101020131815956e-05, "loss": 0.0036, "step": 51890 }, { "grad_norm": 0.05911509692668915, "learning_rate": 5.099366994014556e-05, "loss": 0.0041, "step": 51900 }, { "grad_norm": 0.07512132078409195, "learning_rate": 5.097713845346532e-05, "loss": 0.0033, "step": 51910 }, { "grad_norm": 0.07579781860113144, "learning_rate": 5.096060685992668e-05, "loss": 0.0034, "step": 51920 }, { "grad_norm": 0.06821195781230927, "learning_rate": 5.0944075161337514e-05, "loss": 0.0038, "step": 51930 }, { "grad_norm": 0.05213622748851776, "learning_rate": 5.09275433595057e-05, "loss": 0.0029, "step": 51940 }, { "grad_norm": 0.048513200134038925, "learning_rate": 5.0911011456239157e-05, "loss": 0.0035, "step": 51950 }, { "grad_norm": 0.07278042286634445, "learning_rate": 5.089447945334574e-05, "loss": 0.0029, "step": 51960 }, { "grad_norm": 0.08944067358970642, "learning_rate": 5.087794735263343e-05, "loss": 0.0034, "step": 51970 }, { "grad_norm": 0.07730905711650848, "learning_rate": 5.086141515591011e-05, "loss": 0.0036, "step": 51980 }, { "grad_norm": 0.051741261035203934, "learning_rate": 5.084488286498371e-05, "loss": 0.0032, "step": 51990 }, { "grad_norm": 0.0749148353934288, "learning_rate": 5.08283504816622e-05, "loss": 0.0032, "step": 52000 }, { "grad_norm": 0.07725045830011368, "learning_rate": 5.081181800775353e-05, "loss": 0.0029, "step": 52010 }, { "grad_norm": 0.07457035034894943, "learning_rate": 5.079528544506565e-05, "loss": 0.0037, "step": 52020 }, { "grad_norm": 0.056497711688280106, "learning_rate": 5.077875279540657e-05, "loss": 0.0029, "step": 52030 }, { "grad_norm": 0.06605591624975204, "learning_rate": 5.076222006058424e-05, "loss": 0.0032, "step": 52040 }, { "grad_norm": 0.05280221626162529, "learning_rate": 5.074568724240667e-05, "loss": 0.0032, "step": 52050 }, { "grad_norm": 0.09395747631788254, "learning_rate": 5.072915434268187e-05, "loss": 0.0031, "step": 52060 }, { "grad_norm": 0.07441899925470352, "learning_rate": 5.071262136321785e-05, "loss": 0.0033, "step": 52070 }, { "grad_norm": 0.059930380433797836, "learning_rate": 5.069608830582262e-05, "loss": 0.0032, "step": 52080 }, { "grad_norm": 0.06414363533258438, "learning_rate": 5.067955517230423e-05, "loss": 0.0036, "step": 52090 }, { "grad_norm": 0.06826767325401306, "learning_rate": 5.06630219644707e-05, "loss": 0.0036, "step": 52100 }, { "grad_norm": 0.0588206946849823, "learning_rate": 5.064648868413008e-05, "loss": 0.0034, "step": 52110 }, { "grad_norm": 0.060773395001888275, "learning_rate": 5.062995533309044e-05, "loss": 0.003, "step": 52120 }, { "grad_norm": 0.07255237549543381, "learning_rate": 5.061342191315983e-05, "loss": 0.003, "step": 52130 }, { "grad_norm": 0.06676558405160904, "learning_rate": 5.0596888426146325e-05, "loss": 0.003, "step": 52140 }, { "grad_norm": 0.06492946296930313, "learning_rate": 5.058035487385798e-05, "loss": 0.0031, "step": 52150 }, { "grad_norm": 0.09347604215145111, "learning_rate": 5.0563821258102915e-05, "loss": 0.0038, "step": 52160 }, { "grad_norm": 0.06492698192596436, "learning_rate": 5.05472875806892e-05, "loss": 0.0035, "step": 52170 }, { "grad_norm": 0.06251589208841324, "learning_rate": 5.053075384342494e-05, "loss": 0.0029, "step": 52180 }, { "grad_norm": 0.06387036293745041, "learning_rate": 5.0514220048118234e-05, "loss": 0.0033, "step": 52190 }, { "grad_norm": 0.05750040337443352, "learning_rate": 5.049768619657721e-05, "loss": 0.003, "step": 52200 }, { "grad_norm": 0.06317470222711563, "learning_rate": 5.048115229060996e-05, "loss": 0.004, "step": 52210 }, { "grad_norm": 0.05720040202140808, "learning_rate": 5.046461833202462e-05, "loss": 0.0031, "step": 52220 }, { "grad_norm": 0.06964945048093796, "learning_rate": 5.044808432262932e-05, "loss": 0.0034, "step": 52230 }, { "grad_norm": 0.06535785645246506, "learning_rate": 5.043155026423219e-05, "loss": 0.0031, "step": 52240 }, { "grad_norm": 0.08881739526987076, "learning_rate": 5.041501615864137e-05, "loss": 0.0034, "step": 52250 }, { "grad_norm": 0.06879284977912903, "learning_rate": 5.039848200766502e-05, "loss": 0.0033, "step": 52260 }, { "grad_norm": 0.071944959461689, "learning_rate": 5.038194781311125e-05, "loss": 0.0031, "step": 52270 }, { "grad_norm": 0.058077968657016754, "learning_rate": 5.0365413576788264e-05, "loss": 0.0042, "step": 52280 }, { "grad_norm": 0.05861680209636688, "learning_rate": 5.0348879300504194e-05, "loss": 0.0031, "step": 52290 }, { "grad_norm": 0.06262992322444916, "learning_rate": 5.033234498606721e-05, "loss": 0.0034, "step": 52300 }, { "grad_norm": 0.06922642886638641, "learning_rate": 5.0315810635285476e-05, "loss": 0.0024, "step": 52310 }, { "grad_norm": 0.0703209862112999, "learning_rate": 5.0299276249967164e-05, "loss": 0.0027, "step": 52320 }, { "grad_norm": 0.07660870999097824, "learning_rate": 5.0282741831920454e-05, "loss": 0.0038, "step": 52330 }, { "grad_norm": 0.0929175615310669, "learning_rate": 5.026620738295351e-05, "loss": 0.003, "step": 52340 }, { "grad_norm": 0.05902264639735222, "learning_rate": 5.024967290487455e-05, "loss": 0.0035, "step": 52350 }, { "grad_norm": 0.07772229611873627, "learning_rate": 5.023313839949172e-05, "loss": 0.0044, "step": 52360 }, { "grad_norm": 0.08850442618131638, "learning_rate": 5.0216603868613224e-05, "loss": 0.0044, "step": 52370 }, { "grad_norm": 0.10531271994113922, "learning_rate": 5.020006931404726e-05, "loss": 0.0033, "step": 52380 }, { "grad_norm": 0.06436370313167572, "learning_rate": 5.0183534737602014e-05, "loss": 0.0036, "step": 52390 }, { "grad_norm": 0.06275825202465057, "learning_rate": 5.0167000141085686e-05, "loss": 0.0036, "step": 52400 }, { "grad_norm": 0.07144682109355927, "learning_rate": 5.015046552630647e-05, "loss": 0.0031, "step": 52410 }, { "grad_norm": 0.04967232793569565, "learning_rate": 5.013393089507258e-05, "loss": 0.0031, "step": 52420 }, { "grad_norm": 0.07455036789178848, "learning_rate": 5.011739624919219e-05, "loss": 0.0037, "step": 52430 }, { "grad_norm": 0.08590510487556458, "learning_rate": 5.0100861590473545e-05, "loss": 0.0044, "step": 52440 }, { "grad_norm": 0.07187063992023468, "learning_rate": 5.008432692072482e-05, "loss": 0.0035, "step": 52450 }, { "grad_norm": 0.07992415875196457, "learning_rate": 5.006779224175423e-05, "loss": 0.0045, "step": 52460 }, { "grad_norm": 0.07203105092048645, "learning_rate": 5.005125755536998e-05, "loss": 0.0041, "step": 52470 }, { "grad_norm": 0.06785236299037933, "learning_rate": 5.003472286338029e-05, "loss": 0.0036, "step": 52480 }, { "grad_norm": 0.09108046442270279, "learning_rate": 5.001818816759335e-05, "loss": 0.0044, "step": 52490 }, { "grad_norm": 0.09610395133495331, "learning_rate": 5.000165346981738e-05, "loss": 0.0049, "step": 52500 }, { "grad_norm": 0.08868717402219772, "learning_rate": 4.9985118771860595e-05, "loss": 0.0034, "step": 52510 }, { "grad_norm": 0.05472894012928009, "learning_rate": 4.996858407553119e-05, "loss": 0.0034, "step": 52520 }, { "grad_norm": 0.06680376082658768, "learning_rate": 4.99520493826374e-05, "loss": 0.0031, "step": 52530 }, { "grad_norm": 0.0807066485285759, "learning_rate": 4.9935514694987384e-05, "loss": 0.0036, "step": 52540 }, { "grad_norm": 0.06293907016515732, "learning_rate": 4.991898001438941e-05, "loss": 0.0042, "step": 52550 }, { "grad_norm": 0.10360202938318253, "learning_rate": 4.990244534265166e-05, "loss": 0.0045, "step": 52560 }, { "grad_norm": 0.09444475173950195, "learning_rate": 4.988591068158234e-05, "loss": 0.0042, "step": 52570 }, { "grad_norm": 0.07077176868915558, "learning_rate": 4.986937603298965e-05, "loss": 0.0031, "step": 52580 }, { "grad_norm": 0.07951189577579498, "learning_rate": 4.985284139868181e-05, "loss": 0.0042, "step": 52590 }, { "grad_norm": 0.06570862233638763, "learning_rate": 4.9836306780467e-05, "loss": 0.0041, "step": 52600 }, { "grad_norm": 0.05499739944934845, "learning_rate": 4.981977218015343e-05, "loss": 0.0034, "step": 52610 }, { "grad_norm": 0.07120577991008759, "learning_rate": 4.980323759954932e-05, "loss": 0.0034, "step": 52620 }, { "grad_norm": 0.0726611390709877, "learning_rate": 4.978670304046284e-05, "loss": 0.0033, "step": 52630 }, { "grad_norm": 0.06706125289201736, "learning_rate": 4.9770168504702195e-05, "loss": 0.0038, "step": 52640 }, { "grad_norm": 0.08310163021087646, "learning_rate": 4.975363399407559e-05, "loss": 0.003, "step": 52650 }, { "grad_norm": 0.08803313225507736, "learning_rate": 4.9737099510391194e-05, "loss": 0.0034, "step": 52660 }, { "grad_norm": 0.09623438119888306, "learning_rate": 4.972056505545719e-05, "loss": 0.0043, "step": 52670 }, { "grad_norm": 0.07002612203359604, "learning_rate": 4.9704030631081775e-05, "loss": 0.0033, "step": 52680 }, { "grad_norm": 0.07270451635122299, "learning_rate": 4.968749623907313e-05, "loss": 0.0027, "step": 52690 }, { "grad_norm": 0.09363117069005966, "learning_rate": 4.967096188123943e-05, "loss": 0.0046, "step": 52700 }, { "grad_norm": 0.06838366389274597, "learning_rate": 4.965442755938884e-05, "loss": 0.0031, "step": 52710 }, { "grad_norm": 0.05612137168645859, "learning_rate": 4.963789327532954e-05, "loss": 0.0031, "step": 52720 }, { "grad_norm": 0.06305637955665588, "learning_rate": 4.962135903086968e-05, "loss": 0.0034, "step": 52730 }, { "grad_norm": 0.07179922610521317, "learning_rate": 4.960482482781741e-05, "loss": 0.0028, "step": 52740 }, { "grad_norm": 0.08037485182285309, "learning_rate": 4.958829066798092e-05, "loss": 0.0038, "step": 52750 }, { "grad_norm": 0.0884278193116188, "learning_rate": 4.957175655316835e-05, "loss": 0.0032, "step": 52760 }, { "grad_norm": 0.05559210106730461, "learning_rate": 4.9555222485187834e-05, "loss": 0.0036, "step": 52770 }, { "grad_norm": 0.06997501105070114, "learning_rate": 4.953868846584752e-05, "loss": 0.0026, "step": 52780 }, { "grad_norm": 0.046659957617521286, "learning_rate": 4.9522154496955544e-05, "loss": 0.0036, "step": 52790 }, { "grad_norm": 0.07799209654331207, "learning_rate": 4.950562058032001e-05, "loss": 0.0033, "step": 52800 }, { "grad_norm": 0.08402960002422333, "learning_rate": 4.948908671774909e-05, "loss": 0.0035, "step": 52810 }, { "grad_norm": 0.06608409434556961, "learning_rate": 4.947255291105087e-05, "loss": 0.0035, "step": 52820 }, { "grad_norm": 0.05214381963014603, "learning_rate": 4.945601916203348e-05, "loss": 0.0034, "step": 52830 }, { "grad_norm": 0.04550303518772125, "learning_rate": 4.9439485472505016e-05, "loss": 0.0027, "step": 52840 }, { "grad_norm": 0.09503559023141861, "learning_rate": 4.942295184427356e-05, "loss": 0.0035, "step": 52850 }, { "grad_norm": 0.09267366677522659, "learning_rate": 4.9406418279147234e-05, "loss": 0.0029, "step": 52860 }, { "grad_norm": 0.10761295258998871, "learning_rate": 4.9389884778934116e-05, "loss": 0.0033, "step": 52870 }, { "grad_norm": 0.08061821758747101, "learning_rate": 4.937335134544228e-05, "loss": 0.0029, "step": 52880 }, { "grad_norm": 0.07780137658119202, "learning_rate": 4.935681798047979e-05, "loss": 0.0034, "step": 52890 }, { "grad_norm": 0.05720636621117592, "learning_rate": 4.934028468585473e-05, "loss": 0.0034, "step": 52900 }, { "grad_norm": 0.05541137233376503, "learning_rate": 4.932375146337514e-05, "loss": 0.0035, "step": 52910 }, { "grad_norm": 0.06180834025144577, "learning_rate": 4.930721831484908e-05, "loss": 0.0024, "step": 52920 }, { "grad_norm": 0.054075852036476135, "learning_rate": 4.929068524208456e-05, "loss": 0.0036, "step": 52930 }, { "grad_norm": 0.0658869668841362, "learning_rate": 4.927415224688965e-05, "loss": 0.0035, "step": 52940 }, { "grad_norm": 0.060975633561611176, "learning_rate": 4.9257619331072375e-05, "loss": 0.0033, "step": 52950 }, { "grad_norm": 0.0661262795329094, "learning_rate": 4.924108649644072e-05, "loss": 0.0037, "step": 52960 }, { "grad_norm": 0.05340520292520523, "learning_rate": 4.922455374480271e-05, "loss": 0.003, "step": 52970 }, { "grad_norm": 0.06967362016439438, "learning_rate": 4.920802107796634e-05, "loss": 0.0029, "step": 52980 }, { "grad_norm": 0.06990180909633636, "learning_rate": 4.9191488497739564e-05, "loss": 0.0033, "step": 52990 }, { "grad_norm": 0.07161986082792282, "learning_rate": 4.917495600593042e-05, "loss": 0.0026, "step": 53000 }, { "grad_norm": 0.06393011659383774, "learning_rate": 4.915842360434684e-05, "loss": 0.0039, "step": 53010 }, { "grad_norm": 0.05784067139029503, "learning_rate": 4.914189129479679e-05, "loss": 0.0033, "step": 53020 }, { "grad_norm": 0.08438920974731445, "learning_rate": 4.912535907908822e-05, "loss": 0.0037, "step": 53030 }, { "grad_norm": 0.05828727036714554, "learning_rate": 4.910882695902907e-05, "loss": 0.0036, "step": 53040 }, { "grad_norm": 0.08790067583322525, "learning_rate": 4.9092294936427234e-05, "loss": 0.0041, "step": 53050 }, { "grad_norm": 0.07625661045312881, "learning_rate": 4.9075763013090685e-05, "loss": 0.004, "step": 53060 }, { "grad_norm": 0.06796441972255707, "learning_rate": 4.9059231190827305e-05, "loss": 0.0032, "step": 53070 }, { "grad_norm": 0.053501542657613754, "learning_rate": 4.904269947144499e-05, "loss": 0.0032, "step": 53080 }, { "grad_norm": 0.05858913064002991, "learning_rate": 4.902616785675161e-05, "loss": 0.0034, "step": 53090 }, { "grad_norm": 0.0663311704993248, "learning_rate": 4.900963634855505e-05, "loss": 0.0037, "step": 53100 }, { "grad_norm": 0.0669616088271141, "learning_rate": 4.8993104948663173e-05, "loss": 0.0031, "step": 53110 }, { "grad_norm": 0.0933196172118187, "learning_rate": 4.8976573658883823e-05, "loss": 0.0027, "step": 53120 }, { "grad_norm": 0.07158204168081284, "learning_rate": 4.896004248102484e-05, "loss": 0.0031, "step": 53130 }, { "grad_norm": 0.051149044185876846, "learning_rate": 4.8943511416894045e-05, "loss": 0.0031, "step": 53140 }, { "grad_norm": 0.07499907910823822, "learning_rate": 4.892698046829926e-05, "loss": 0.0035, "step": 53150 }, { "grad_norm": 0.061564598232507706, "learning_rate": 4.8910449637048275e-05, "loss": 0.0031, "step": 53160 }, { "grad_norm": 0.05602720007300377, "learning_rate": 4.8893918924948884e-05, "loss": 0.0038, "step": 53170 }, { "grad_norm": 0.08323761075735092, "learning_rate": 4.8877388333808835e-05, "loss": 0.0027, "step": 53180 }, { "grad_norm": 0.06559024751186371, "learning_rate": 4.8860857865435925e-05, "loss": 0.0048, "step": 53190 }, { "grad_norm": 0.06978875398635864, "learning_rate": 4.884432752163789e-05, "loss": 0.0028, "step": 53200 }, { "grad_norm": 0.051463719457387924, "learning_rate": 4.8827797304222466e-05, "loss": 0.0032, "step": 53210 }, { "grad_norm": 0.05410291254520416, "learning_rate": 4.8811267214997366e-05, "loss": 0.0039, "step": 53220 }, { "grad_norm": 0.053536731749773026, "learning_rate": 4.879473725577029e-05, "loss": 0.0031, "step": 53230 }, { "grad_norm": 0.07605814933776855, "learning_rate": 4.877820742834891e-05, "loss": 0.0033, "step": 53240 }, { "grad_norm": 0.08021663129329681, "learning_rate": 4.8761677734540956e-05, "loss": 0.0044, "step": 53250 }, { "grad_norm": 0.09806166589260101, "learning_rate": 4.874514817615407e-05, "loss": 0.0052, "step": 53260 }, { "grad_norm": 0.09026550501585007, "learning_rate": 4.872861875499588e-05, "loss": 0.0036, "step": 53270 }, { "grad_norm": 0.07312525808811188, "learning_rate": 4.871208947287404e-05, "loss": 0.0036, "step": 53280 }, { "grad_norm": 0.0794229656457901, "learning_rate": 4.869556033159615e-05, "loss": 0.0035, "step": 53290 }, { "grad_norm": 0.06955568492412567, "learning_rate": 4.8679031332969835e-05, "loss": 0.0025, "step": 53300 }, { "grad_norm": 0.08062949031591415, "learning_rate": 4.8662502478802635e-05, "loss": 0.0034, "step": 53310 }, { "grad_norm": 0.08541062474250793, "learning_rate": 4.8645973770902176e-05, "loss": 0.0039, "step": 53320 }, { "grad_norm": 0.07068775594234467, "learning_rate": 4.862944521107599e-05, "loss": 0.0035, "step": 53330 }, { "grad_norm": 0.10678276419639587, "learning_rate": 4.861291680113162e-05, "loss": 0.0041, "step": 53340 }, { "grad_norm": 0.08631134778261185, "learning_rate": 4.8596388542876563e-05, "loss": 0.0035, "step": 53350 }, { "grad_norm": 0.057702623307704926, "learning_rate": 4.857986043811836e-05, "loss": 0.0034, "step": 53360 }, { "grad_norm": 0.09221604466438293, "learning_rate": 4.856333248866448e-05, "loss": 0.0038, "step": 53370 }, { "grad_norm": 0.06475306302309036, "learning_rate": 4.85468046963224e-05, "loss": 0.0032, "step": 53380 }, { "grad_norm": 0.09266264736652374, "learning_rate": 4.853027706289957e-05, "loss": 0.0037, "step": 53390 }, { "grad_norm": 0.05327007547020912, "learning_rate": 4.851374959020344e-05, "loss": 0.0028, "step": 53400 }, { "grad_norm": 0.06559227406978607, "learning_rate": 4.849722228004141e-05, "loss": 0.003, "step": 53410 }, { "grad_norm": 0.0719248354434967, "learning_rate": 4.848069513422091e-05, "loss": 0.0031, "step": 53420 }, { "grad_norm": 0.098762646317482, "learning_rate": 4.8464168154549294e-05, "loss": 0.0046, "step": 53430 }, { "grad_norm": 0.08139393478631973, "learning_rate": 4.844764134283392e-05, "loss": 0.0046, "step": 53440 }, { "grad_norm": 0.07409346103668213, "learning_rate": 4.843111470088218e-05, "loss": 0.0042, "step": 53450 }, { "grad_norm": 0.046192266047000885, "learning_rate": 4.841458823050137e-05, "loss": 0.0037, "step": 53460 }, { "grad_norm": 0.07586260139942169, "learning_rate": 4.8398061933498816e-05, "loss": 0.0042, "step": 53470 }, { "grad_norm": 0.07091003656387329, "learning_rate": 4.83815358116818e-05, "loss": 0.0033, "step": 53480 }, { "grad_norm": 0.06430994719266891, "learning_rate": 4.8365009866857584e-05, "loss": 0.003, "step": 53490 }, { "grad_norm": 0.05793431028723717, "learning_rate": 4.834848410083342e-05, "loss": 0.0031, "step": 53500 }, { "grad_norm": 0.05746601149439812, "learning_rate": 4.833195851541657e-05, "loss": 0.0046, "step": 53510 }, { "grad_norm": 0.0583811029791832, "learning_rate": 4.831543311241423e-05, "loss": 0.004, "step": 53520 }, { "grad_norm": 0.08482252061367035, "learning_rate": 4.829890789363359e-05, "loss": 0.0034, "step": 53530 }, { "grad_norm": 0.047158971428871155, "learning_rate": 4.828238286088182e-05, "loss": 0.0031, "step": 53540 }, { "grad_norm": 0.05724490061402321, "learning_rate": 4.826585801596609e-05, "loss": 0.0039, "step": 53550 }, { "grad_norm": 0.056953974068164825, "learning_rate": 4.8249333360693486e-05, "loss": 0.0027, "step": 53560 }, { "grad_norm": 0.054950617253780365, "learning_rate": 4.8232808896871185e-05, "loss": 0.0033, "step": 53570 }, { "grad_norm": 0.07672598958015442, "learning_rate": 4.821628462630624e-05, "loss": 0.0033, "step": 53580 }, { "grad_norm": 0.08971908688545227, "learning_rate": 4.8199760550805735e-05, "loss": 0.003, "step": 53590 }, { "grad_norm": 0.04559826850891113, "learning_rate": 4.818323667217669e-05, "loss": 0.003, "step": 53600 }, { "grad_norm": 0.07949434220790863, "learning_rate": 4.8166712992226165e-05, "loss": 0.0038, "step": 53610 }, { "grad_norm": 0.051001064479351044, "learning_rate": 4.8150189512761154e-05, "loss": 0.0029, "step": 53620 }, { "grad_norm": 0.057055335491895676, "learning_rate": 4.813366623558862e-05, "loss": 0.0028, "step": 53630 }, { "grad_norm": 0.06829923391342163, "learning_rate": 4.811714316251554e-05, "loss": 0.0034, "step": 53640 }, { "grad_norm": 0.06378162652254105, "learning_rate": 4.810062029534886e-05, "loss": 0.003, "step": 53650 }, { "grad_norm": 0.06634204834699631, "learning_rate": 4.8084097635895505e-05, "loss": 0.0037, "step": 53660 }, { "grad_norm": 0.06324523687362671, "learning_rate": 4.8067575185962335e-05, "loss": 0.0036, "step": 53670 }, { "grad_norm": 0.08624235540628433, "learning_rate": 4.805105294735625e-05, "loss": 0.004, "step": 53680 }, { "grad_norm": 0.07251930236816406, "learning_rate": 4.803453092188406e-05, "loss": 0.0027, "step": 53690 }, { "grad_norm": 0.05246073752641678, "learning_rate": 4.801800911135263e-05, "loss": 0.0031, "step": 53700 }, { "grad_norm": 0.06072588264942169, "learning_rate": 4.8001487517568744e-05, "loss": 0.0038, "step": 53710 }, { "grad_norm": 0.0860501229763031, "learning_rate": 4.798496614233918e-05, "loss": 0.0037, "step": 53720 }, { "grad_norm": 0.05935533344745636, "learning_rate": 4.796844498747069e-05, "loss": 0.0028, "step": 53730 }, { "grad_norm": 0.08770627528429031, "learning_rate": 4.795192405477e-05, "loss": 0.0039, "step": 53740 }, { "grad_norm": 0.08408554643392563, "learning_rate": 4.79354033460438e-05, "loss": 0.0037, "step": 53750 }, { "grad_norm": 0.06511930376291275, "learning_rate": 4.79188828630988e-05, "loss": 0.0036, "step": 53760 }, { "grad_norm": 0.07530150562524796, "learning_rate": 4.790236260774165e-05, "loss": 0.0033, "step": 53770 }, { "grad_norm": 0.05505754053592682, "learning_rate": 4.788584258177896e-05, "loss": 0.0036, "step": 53780 }, { "grad_norm": 0.08236242085695267, "learning_rate": 4.7869322787017356e-05, "loss": 0.0049, "step": 53790 }, { "grad_norm": 0.07911469787359238, "learning_rate": 4.78528032252634e-05, "loss": 0.0035, "step": 53800 }, { "grad_norm": 0.0870356410741806, "learning_rate": 4.783628389832366e-05, "loss": 0.0039, "step": 53810 }, { "grad_norm": 0.059382062405347824, "learning_rate": 4.7819764808004644e-05, "loss": 0.0033, "step": 53820 }, { "grad_norm": 0.07497111707925797, "learning_rate": 4.780324595611289e-05, "loss": 0.0033, "step": 53830 }, { "grad_norm": 0.06290271878242493, "learning_rate": 4.7786727344454854e-05, "loss": 0.0029, "step": 53840 }, { "grad_norm": 0.06010046601295471, "learning_rate": 4.777020897483697e-05, "loss": 0.0024, "step": 53850 }, { "grad_norm": 0.07845716923475266, "learning_rate": 4.775369084906569e-05, "loss": 0.0048, "step": 53860 }, { "grad_norm": 0.04388466849923134, "learning_rate": 4.773717296894742e-05, "loss": 0.0027, "step": 53870 }, { "grad_norm": 0.08402492105960846, "learning_rate": 4.772065533628848e-05, "loss": 0.004, "step": 53880 }, { "grad_norm": 0.06651535630226135, "learning_rate": 4.7704137952895265e-05, "loss": 0.0027, "step": 53890 }, { "grad_norm": 0.05736039951443672, "learning_rate": 4.768762082057407e-05, "loss": 0.0033, "step": 53900 }, { "grad_norm": 0.051002781838178635, "learning_rate": 4.7671103941131186e-05, "loss": 0.0028, "step": 53910 }, { "grad_norm": 0.09114585071802139, "learning_rate": 4.765458731637287e-05, "loss": 0.0032, "step": 53920 }, { "grad_norm": 0.0711713433265686, "learning_rate": 4.763807094810537e-05, "loss": 0.0033, "step": 53930 }, { "grad_norm": 0.07345262169837952, "learning_rate": 4.762155483813487e-05, "loss": 0.0041, "step": 53940 }, { "grad_norm": 0.11144356429576874, "learning_rate": 4.760503898826754e-05, "loss": 0.0039, "step": 53950 }, { "grad_norm": 0.266823947429657, "learning_rate": 4.758852340030957e-05, "loss": 0.0045, "step": 53960 }, { "grad_norm": 0.11386942118406296, "learning_rate": 4.757200807606705e-05, "loss": 0.0035, "step": 53970 }, { "grad_norm": 0.06266681104898453, "learning_rate": 4.755549301734607e-05, "loss": 0.0039, "step": 53980 }, { "grad_norm": 0.09455517679452896, "learning_rate": 4.753897822595271e-05, "loss": 0.0036, "step": 53990 }, { "grad_norm": 0.0813014879822731, "learning_rate": 4.7522463703692994e-05, "loss": 0.0044, "step": 54000 }, { "grad_norm": 0.06394968926906586, "learning_rate": 4.75059494523729e-05, "loss": 0.0033, "step": 54010 }, { "grad_norm": 0.060763467103242874, "learning_rate": 4.748943547379843e-05, "loss": 0.003, "step": 54020 }, { "grad_norm": 0.0741802230477333, "learning_rate": 4.7472921769775535e-05, "loss": 0.0046, "step": 54030 }, { "grad_norm": 0.06373326480388641, "learning_rate": 4.745640834211011e-05, "loss": 0.0036, "step": 54040 }, { "grad_norm": 0.07080848515033722, "learning_rate": 4.7439895192608054e-05, "loss": 0.0039, "step": 54050 }, { "grad_norm": 0.06788022816181183, "learning_rate": 4.7423382323075186e-05, "loss": 0.0027, "step": 54060 }, { "grad_norm": 0.046303730458021164, "learning_rate": 4.740686973531737e-05, "loss": 0.0026, "step": 54070 }, { "grad_norm": 0.09974269568920135, "learning_rate": 4.7390357431140375e-05, "loss": 0.0032, "step": 54080 }, { "grad_norm": 0.08722911775112152, "learning_rate": 4.737384541234998e-05, "loss": 0.0034, "step": 54090 }, { "grad_norm": 0.05962372198700905, "learning_rate": 4.735733368075189e-05, "loss": 0.0033, "step": 54100 }, { "grad_norm": 0.10383760184049606, "learning_rate": 4.734082223815182e-05, "loss": 0.0034, "step": 54110 }, { "grad_norm": 0.0743011012673378, "learning_rate": 4.7324311086355436e-05, "loss": 0.0029, "step": 54120 }, { "grad_norm": 0.061928413808345795, "learning_rate": 4.730780022716836e-05, "loss": 0.0033, "step": 54130 }, { "grad_norm": 0.08448325097560883, "learning_rate": 4.729128966239619e-05, "loss": 0.0027, "step": 54140 }, { "grad_norm": 0.07526780664920807, "learning_rate": 4.7274779393844536e-05, "loss": 0.003, "step": 54150 }, { "grad_norm": 0.05872824043035507, "learning_rate": 4.725826942331891e-05, "loss": 0.0034, "step": 54160 }, { "grad_norm": 0.06189222261309624, "learning_rate": 4.724175975262481e-05, "loss": 0.003, "step": 54170 }, { "grad_norm": 0.09532486647367477, "learning_rate": 4.722525038356772e-05, "loss": 0.0031, "step": 54180 }, { "grad_norm": 0.0642583817243576, "learning_rate": 4.7208741317953075e-05, "loss": 0.0034, "step": 54190 }, { "grad_norm": 0.07169972360134125, "learning_rate": 4.7192232557586264e-05, "loss": 0.0032, "step": 54200 }, { "grad_norm": 0.05971897393465042, "learning_rate": 4.717572410427269e-05, "loss": 0.0033, "step": 54210 }, { "grad_norm": 0.0644720271229744, "learning_rate": 4.71592159598177e-05, "loss": 0.0028, "step": 54220 }, { "grad_norm": 0.09099935740232468, "learning_rate": 4.714270812602657e-05, "loss": 0.0032, "step": 54230 }, { "grad_norm": 0.07905084639787674, "learning_rate": 4.712620060470458e-05, "loss": 0.0034, "step": 54240 }, { "grad_norm": 0.08909197151660919, "learning_rate": 4.7109693397656986e-05, "loss": 0.0041, "step": 54250 }, { "grad_norm": 0.06565943360328674, "learning_rate": 4.709318650668894e-05, "loss": 0.0035, "step": 54260 }, { "grad_norm": 0.12144055962562561, "learning_rate": 4.7076679933605676e-05, "loss": 0.0041, "step": 54270 }, { "grad_norm": 0.06612330675125122, "learning_rate": 4.70601736802123e-05, "loss": 0.0032, "step": 54280 }, { "grad_norm": 0.08452392369508743, "learning_rate": 4.704366774831391e-05, "loss": 0.0036, "step": 54290 }, { "grad_norm": 0.06264373660087585, "learning_rate": 4.702716213971557e-05, "loss": 0.0033, "step": 54300 }, { "grad_norm": 0.06472636759281158, "learning_rate": 4.70106568562223e-05, "loss": 0.0032, "step": 54310 }, { "grad_norm": 0.0748477652668953, "learning_rate": 4.6994151899639105e-05, "loss": 0.0031, "step": 54320 }, { "grad_norm": 0.056615378707647324, "learning_rate": 4.697764727177093e-05, "loss": 0.0028, "step": 54330 }, { "grad_norm": 0.04544685408473015, "learning_rate": 4.696114297442272e-05, "loss": 0.0031, "step": 54340 }, { "grad_norm": 0.07466883212327957, "learning_rate": 4.694463900939933e-05, "loss": 0.0026, "step": 54350 }, { "grad_norm": 0.05854646861553192, "learning_rate": 4.692813537850564e-05, "loss": 0.0026, "step": 54360 }, { "grad_norm": 0.08927491307258606, "learning_rate": 4.6911632083546445e-05, "loss": 0.0039, "step": 54370 }, { "grad_norm": 0.08239959180355072, "learning_rate": 4.6895129126326525e-05, "loss": 0.0028, "step": 54380 }, { "grad_norm": 0.05876016989350319, "learning_rate": 4.687862650865059e-05, "loss": 0.0033, "step": 54390 }, { "grad_norm": 0.038624707609415054, "learning_rate": 4.686212423232339e-05, "loss": 0.003, "step": 54400 }, { "grad_norm": 0.045781202614307404, "learning_rate": 4.6845622299149565e-05, "loss": 0.0036, "step": 54410 }, { "grad_norm": 0.050153546035289764, "learning_rate": 4.682912071093374e-05, "loss": 0.003, "step": 54420 }, { "grad_norm": 0.07014387845993042, "learning_rate": 4.681261946948052e-05, "loss": 0.0031, "step": 54430 }, { "grad_norm": 0.07860150933265686, "learning_rate": 4.679611857659443e-05, "loss": 0.003, "step": 54440 }, { "grad_norm": 0.050949208438396454, "learning_rate": 4.6779618034079975e-05, "loss": 0.0029, "step": 54450 }, { "grad_norm": 0.05407052859663963, "learning_rate": 4.6763117843741675e-05, "loss": 0.0028, "step": 54460 }, { "grad_norm": 0.06619547307491302, "learning_rate": 4.674661800738394e-05, "loss": 0.0037, "step": 54470 }, { "grad_norm": 0.05227390676736832, "learning_rate": 4.673011852681117e-05, "loss": 0.0028, "step": 54480 }, { "grad_norm": 0.05063511058688164, "learning_rate": 4.671361940382772e-05, "loss": 0.003, "step": 54490 }, { "grad_norm": 0.07155270129442215, "learning_rate": 4.669712064023791e-05, "loss": 0.0035, "step": 54500 }, { "grad_norm": 0.07092316448688507, "learning_rate": 4.668062223784602e-05, "loss": 0.0027, "step": 54510 }, { "grad_norm": 0.07757535576820374, "learning_rate": 4.6664124198456274e-05, "loss": 0.0037, "step": 54520 }, { "grad_norm": 0.07443279772996902, "learning_rate": 4.6647626523872905e-05, "loss": 0.0035, "step": 54530 }, { "grad_norm": 0.06141067296266556, "learning_rate": 4.6631129215900066e-05, "loss": 0.0034, "step": 54540 }, { "grad_norm": 0.07277939468622208, "learning_rate": 4.661463227634187e-05, "loss": 0.0032, "step": 54550 }, { "grad_norm": 0.04173189401626587, "learning_rate": 4.659813570700238e-05, "loss": 0.0029, "step": 54560 }, { "grad_norm": 0.053435735404491425, "learning_rate": 4.658163950968568e-05, "loss": 0.0029, "step": 54570 }, { "grad_norm": 0.0534503199160099, "learning_rate": 4.6565143686195726e-05, "loss": 0.003, "step": 54580 }, { "grad_norm": 0.08699408918619156, "learning_rate": 4.654864823833649e-05, "loss": 0.0033, "step": 54590 }, { "grad_norm": 0.06105054169893265, "learning_rate": 4.65321531679119e-05, "loss": 0.0032, "step": 54600 }, { "grad_norm": 0.09150438010692596, "learning_rate": 4.6515658476725834e-05, "loss": 0.0045, "step": 54610 }, { "grad_norm": 0.05437855049967766, "learning_rate": 4.649916416658212e-05, "loss": 0.0029, "step": 54620 }, { "grad_norm": 0.05996660515666008, "learning_rate": 4.648267023928455e-05, "loss": 0.0027, "step": 54630 }, { "grad_norm": 0.06958317011594772, "learning_rate": 4.646617669663687e-05, "loss": 0.0033, "step": 54640 }, { "grad_norm": 0.04455788433551788, "learning_rate": 4.644968354044277e-05, "loss": 0.0022, "step": 54650 }, { "grad_norm": 0.06456759572029114, "learning_rate": 4.643319077250596e-05, "loss": 0.0042, "step": 54660 }, { "grad_norm": 0.05212904512882233, "learning_rate": 4.6416698394630055e-05, "loss": 0.0031, "step": 54670 }, { "grad_norm": 0.06856989115476608, "learning_rate": 4.640020640861862e-05, "loss": 0.0033, "step": 54680 }, { "grad_norm": 0.05334915965795517, "learning_rate": 4.6383714816275195e-05, "loss": 0.0037, "step": 54690 }, { "grad_norm": 0.05099748447537422, "learning_rate": 4.636722361940329e-05, "loss": 0.0032, "step": 54700 }, { "grad_norm": 0.055312737822532654, "learning_rate": 4.635073281980632e-05, "loss": 0.0029, "step": 54710 }, { "grad_norm": 0.07062935829162598, "learning_rate": 4.633424241928775e-05, "loss": 0.0035, "step": 54720 }, { "grad_norm": 0.06948681175708771, "learning_rate": 4.6317752419650914e-05, "loss": 0.0025, "step": 54730 }, { "grad_norm": 0.05917995050549507, "learning_rate": 4.630126282269914e-05, "loss": 0.003, "step": 54740 }, { "grad_norm": 0.05298908054828644, "learning_rate": 4.628477363023569e-05, "loss": 0.0028, "step": 54750 }, { "grad_norm": 0.05773909389972687, "learning_rate": 4.626828484406383e-05, "loss": 0.0033, "step": 54760 }, { "grad_norm": 0.052411019802093506, "learning_rate": 4.625179646598669e-05, "loss": 0.003, "step": 54770 }, { "grad_norm": 0.07535979896783829, "learning_rate": 4.6235308497807487e-05, "loss": 0.0026, "step": 54780 }, { "grad_norm": 0.050656501203775406, "learning_rate": 4.6218820941329276e-05, "loss": 0.0033, "step": 54790 }, { "grad_norm": 0.07911471277475357, "learning_rate": 4.620233379835513e-05, "loss": 0.0029, "step": 54800 }, { "grad_norm": 0.04967194423079491, "learning_rate": 4.618584707068804e-05, "loss": 0.0027, "step": 54810 }, { "grad_norm": 0.059675440192222595, "learning_rate": 4.616936076013099e-05, "loss": 0.0029, "step": 54820 }, { "grad_norm": 0.05767707899212837, "learning_rate": 4.615287486848688e-05, "loss": 0.0041, "step": 54830 }, { "grad_norm": 0.08728477358818054, "learning_rate": 4.6136389397558584e-05, "loss": 0.0031, "step": 54840 }, { "grad_norm": 0.051533427089452744, "learning_rate": 4.6119904349148935e-05, "loss": 0.0036, "step": 54850 }, { "grad_norm": 0.07269230484962463, "learning_rate": 4.6103419725060726e-05, "loss": 0.0027, "step": 54860 }, { "grad_norm": 0.10558871179819107, "learning_rate": 4.608693552709667e-05, "loss": 0.003, "step": 54870 }, { "grad_norm": 0.06860004365444183, "learning_rate": 4.607045175705947e-05, "loss": 0.003, "step": 54880 }, { "grad_norm": 0.0808604285120964, "learning_rate": 4.605396841675176e-05, "loss": 0.0027, "step": 54890 }, { "grad_norm": 0.04420478269457817, "learning_rate": 4.6037485507976094e-05, "loss": 0.0027, "step": 54900 }, { "grad_norm": 0.07897201925516129, "learning_rate": 4.602100303253509e-05, "loss": 0.0025, "step": 54910 }, { "grad_norm": 0.08375813812017441, "learning_rate": 4.600452099223122e-05, "loss": 0.0032, "step": 54920 }, { "grad_norm": 0.08023210614919662, "learning_rate": 4.598803938886691e-05, "loss": 0.0032, "step": 54930 }, { "grad_norm": 0.07941943407058716, "learning_rate": 4.597155822424459e-05, "loss": 0.003, "step": 54940 }, { "grad_norm": 0.05325678363442421, "learning_rate": 4.59550775001666e-05, "loss": 0.0032, "step": 54950 }, { "grad_norm": 0.06985723972320557, "learning_rate": 4.5938597218435234e-05, "loss": 0.0041, "step": 54960 }, { "grad_norm": 0.07813183218240738, "learning_rate": 4.592211738085279e-05, "loss": 0.0032, "step": 54970 }, { "grad_norm": 0.06447471678256989, "learning_rate": 4.590563798922146e-05, "loss": 0.003, "step": 54980 }, { "grad_norm": 0.06318395584821701, "learning_rate": 4.5889159045343404e-05, "loss": 0.0035, "step": 54990 }, { "grad_norm": 0.07236973941326141, "learning_rate": 4.5872680551020734e-05, "loss": 0.0033, "step": 55000 }, { "grad_norm": 0.06600137054920197, "learning_rate": 4.585620250805552e-05, "loss": 0.0032, "step": 55010 }, { "grad_norm": 0.06729906052350998, "learning_rate": 4.583972491824975e-05, "loss": 0.0035, "step": 55020 }, { "grad_norm": 0.0588449165225029, "learning_rate": 4.582324778340541e-05, "loss": 0.0028, "step": 55030 }, { "grad_norm": 0.05812633037567139, "learning_rate": 4.5806771105324424e-05, "loss": 0.0037, "step": 55040 }, { "grad_norm": 0.0962103009223938, "learning_rate": 4.579029488580865e-05, "loss": 0.0035, "step": 55050 }, { "grad_norm": 0.06211818382143974, "learning_rate": 4.5773819126659876e-05, "loss": 0.0034, "step": 55060 }, { "grad_norm": 0.04616987705230713, "learning_rate": 4.575734382967991e-05, "loss": 0.0039, "step": 55070 }, { "grad_norm": 0.07016227394342422, "learning_rate": 4.574086899667044e-05, "loss": 0.0032, "step": 55080 }, { "grad_norm": 0.07663194835186005, "learning_rate": 4.5724394629433106e-05, "loss": 0.0031, "step": 55090 }, { "grad_norm": 0.08117873221635818, "learning_rate": 4.570792072976958e-05, "loss": 0.0032, "step": 55100 }, { "grad_norm": 0.045369986444711685, "learning_rate": 4.569144729948138e-05, "loss": 0.0028, "step": 55110 }, { "grad_norm": 0.07596664130687714, "learning_rate": 4.5674974340370033e-05, "loss": 0.0034, "step": 55120 }, { "grad_norm": 0.07435238361358643, "learning_rate": 4.565850185423699e-05, "loss": 0.003, "step": 55130 }, { "grad_norm": 0.0863046944141388, "learning_rate": 4.564202984288365e-05, "loss": 0.0032, "step": 55140 }, { "grad_norm": 0.07870718091726303, "learning_rate": 4.5625558308111356e-05, "loss": 0.0029, "step": 55150 }, { "grad_norm": 0.07779323309659958, "learning_rate": 4.560908725172145e-05, "loss": 0.0033, "step": 55160 }, { "grad_norm": 0.05157947167754173, "learning_rate": 4.559261667551516e-05, "loss": 0.0029, "step": 55170 }, { "grad_norm": 0.06915362179279327, "learning_rate": 4.5576146581293685e-05, "loss": 0.0027, "step": 55180 }, { "grad_norm": 0.07046613097190857, "learning_rate": 4.5559676970858164e-05, "loss": 0.0028, "step": 55190 }, { "grad_norm": 0.07107347249984741, "learning_rate": 4.5543207846009705e-05, "loss": 0.0037, "step": 55200 }, { "grad_norm": 0.049044668674468994, "learning_rate": 4.552673920854933e-05, "loss": 0.0026, "step": 55210 }, { "grad_norm": 0.07325431704521179, "learning_rate": 4.5510271060278006e-05, "loss": 0.0031, "step": 55220 }, { "grad_norm": 0.08526793122291565, "learning_rate": 4.5493803402996704e-05, "loss": 0.004, "step": 55230 }, { "grad_norm": 0.05479258671402931, "learning_rate": 4.5477336238506295e-05, "loss": 0.0033, "step": 55240 }, { "grad_norm": 0.07459647208452225, "learning_rate": 4.546086956860759e-05, "loss": 0.0031, "step": 55250 }, { "grad_norm": 0.051242392510175705, "learning_rate": 4.5444403395101374e-05, "loss": 0.003, "step": 55260 }, { "grad_norm": 0.06607018411159515, "learning_rate": 4.542793771978834e-05, "loss": 0.0046, "step": 55270 }, { "grad_norm": 0.08295796811580658, "learning_rate": 4.541147254446916e-05, "loss": 0.0037, "step": 55280 }, { "grad_norm": 0.06270917505025864, "learning_rate": 4.539500787094445e-05, "loss": 0.0033, "step": 55290 }, { "grad_norm": 0.06443853676319122, "learning_rate": 4.537854370101475e-05, "loss": 0.0032, "step": 55300 }, { "grad_norm": 0.060362886637449265, "learning_rate": 4.5362080036480566e-05, "loss": 0.0026, "step": 55310 }, { "grad_norm": 0.06391627341508865, "learning_rate": 4.534561687914235e-05, "loss": 0.0033, "step": 55320 }, { "grad_norm": 0.06386047601699829, "learning_rate": 4.5329154230800466e-05, "loss": 0.0033, "step": 55330 }, { "grad_norm": 0.07683175802230835, "learning_rate": 4.531269209325525e-05, "loss": 0.0029, "step": 55340 }, { "grad_norm": 0.08171523362398148, "learning_rate": 4.529623046830696e-05, "loss": 0.0042, "step": 55350 }, { "grad_norm": 0.08696630597114563, "learning_rate": 4.5279769357755866e-05, "loss": 0.0031, "step": 55360 }, { "grad_norm": 0.0643158033490181, "learning_rate": 4.5263308763402084e-05, "loss": 0.0027, "step": 55370 }, { "grad_norm": 0.06718318164348602, "learning_rate": 4.524684868704574e-05, "loss": 0.0031, "step": 55380 }, { "grad_norm": 0.0540274903178215, "learning_rate": 4.523038913048688e-05, "loss": 0.0031, "step": 55390 }, { "grad_norm": 0.07422367483377457, "learning_rate": 4.521393009552549e-05, "loss": 0.0036, "step": 55400 }, { "grad_norm": 0.0690774917602539, "learning_rate": 4.5197471583961484e-05, "loss": 0.0039, "step": 55410 }, { "grad_norm": 0.060594625771045685, "learning_rate": 4.51810135975948e-05, "loss": 0.0043, "step": 55420 }, { "grad_norm": 0.07700345665216446, "learning_rate": 4.516455613822521e-05, "loss": 0.0031, "step": 55430 }, { "grad_norm": 0.05228946730494499, "learning_rate": 4.514809920765249e-05, "loss": 0.0045, "step": 55440 }, { "grad_norm": 0.06540871411561966, "learning_rate": 4.5131642807676356e-05, "loss": 0.0034, "step": 55450 }, { "grad_norm": 0.061475519090890884, "learning_rate": 4.511518694009644e-05, "loss": 0.0024, "step": 55460 }, { "grad_norm": 0.06038155034184456, "learning_rate": 4.509873160671231e-05, "loss": 0.0028, "step": 55470 }, { "grad_norm": 0.06515348702669144, "learning_rate": 4.508227680932356e-05, "loss": 0.0029, "step": 55480 }, { "grad_norm": 0.06070299074053764, "learning_rate": 4.5065822549729605e-05, "loss": 0.003, "step": 55490 }, { "grad_norm": 0.047452423721551895, "learning_rate": 4.504936882972989e-05, "loss": 0.004, "step": 55500 }, { "grad_norm": 0.0469944141805172, "learning_rate": 4.503291565112374e-05, "loss": 0.0027, "step": 55510 }, { "grad_norm": 0.0931006669998169, "learning_rate": 4.501646301571048e-05, "loss": 0.0047, "step": 55520 }, { "grad_norm": 0.07478520274162292, "learning_rate": 4.500001092528933e-05, "loss": 0.0033, "step": 55530 }, { "grad_norm": 0.070989228785038, "learning_rate": 4.4983559381659455e-05, "loss": 0.0036, "step": 55540 }, { "grad_norm": 0.07212956249713898, "learning_rate": 4.496710838661999e-05, "loss": 0.0029, "step": 55550 }, { "grad_norm": 0.04727616906166077, "learning_rate": 4.495065794196999e-05, "loss": 0.0032, "step": 55560 }, { "grad_norm": 0.051315587013959885, "learning_rate": 4.4934208049508455e-05, "loss": 0.003, "step": 55570 }, { "grad_norm": 0.0775102823972702, "learning_rate": 4.4917758711034303e-05, "loss": 0.0028, "step": 55580 }, { "grad_norm": 0.06118699163198471, "learning_rate": 4.490130992834642e-05, "loss": 0.0026, "step": 55590 }, { "grad_norm": 0.05661783367395401, "learning_rate": 4.488486170324359e-05, "loss": 0.0031, "step": 55600 }, { "grad_norm": 0.06378072500228882, "learning_rate": 4.486841403752463e-05, "loss": 0.0028, "step": 55610 }, { "grad_norm": 0.08293759077787399, "learning_rate": 4.4851966932988174e-05, "loss": 0.0026, "step": 55620 }, { "grad_norm": 0.06410714238882065, "learning_rate": 4.4835520391432894e-05, "loss": 0.0032, "step": 55630 }, { "grad_norm": 0.04945884272456169, "learning_rate": 4.481907441465733e-05, "loss": 0.0031, "step": 55640 }, { "grad_norm": 0.04525325074791908, "learning_rate": 4.480262900446e-05, "loss": 0.0028, "step": 55650 }, { "grad_norm": 0.040576111525297165, "learning_rate": 4.478618416263933e-05, "loss": 0.0044, "step": 55660 }, { "grad_norm": 0.0699702799320221, "learning_rate": 4.476973989099373e-05, "loss": 0.0028, "step": 55670 }, { "grad_norm": 0.09259723126888275, "learning_rate": 4.475329619132153e-05, "loss": 0.0032, "step": 55680 }, { "grad_norm": 0.0569688081741333, "learning_rate": 4.4736853065420956e-05, "loss": 0.0026, "step": 55690 }, { "grad_norm": 0.04823823645710945, "learning_rate": 4.472041051509023e-05, "loss": 0.0032, "step": 55700 }, { "grad_norm": 0.07045289129018784, "learning_rate": 4.470396854212746e-05, "loss": 0.0047, "step": 55710 }, { "grad_norm": 0.06651508063077927, "learning_rate": 4.468752714833072e-05, "loss": 0.0032, "step": 55720 }, { "grad_norm": 0.06885331869125366, "learning_rate": 4.467108633549803e-05, "loss": 0.0028, "step": 55730 }, { "grad_norm": 0.07498679310083389, "learning_rate": 4.4654646105427335e-05, "loss": 0.0032, "step": 55740 }, { "grad_norm": 0.06444749981164932, "learning_rate": 4.463820645991651e-05, "loss": 0.0035, "step": 55750 }, { "grad_norm": 0.05356481671333313, "learning_rate": 4.462176740076334e-05, "loss": 0.0035, "step": 55760 }, { "grad_norm": 0.07050257176160812, "learning_rate": 4.4605328929765624e-05, "loss": 0.0033, "step": 55770 }, { "grad_norm": 0.07976111024618149, "learning_rate": 4.458889104872102e-05, "loss": 0.0044, "step": 55780 }, { "grad_norm": 0.06371920555830002, "learning_rate": 4.457245375942715e-05, "loss": 0.0032, "step": 55790 }, { "grad_norm": 0.095230333507061, "learning_rate": 4.455601706368157e-05, "loss": 0.0042, "step": 55800 }, { "grad_norm": 0.08352255076169968, "learning_rate": 4.453958096328179e-05, "loss": 0.0027, "step": 55810 }, { "grad_norm": 0.06356534361839294, "learning_rate": 4.452314546002523e-05, "loss": 0.0029, "step": 55820 }, { "grad_norm": 0.06007500737905502, "learning_rate": 4.4506710555709246e-05, "loss": 0.0025, "step": 55830 }, { "grad_norm": 0.06841105967760086, "learning_rate": 4.449027625213114e-05, "loss": 0.0033, "step": 55840 }, { "grad_norm": 0.05914853513240814, "learning_rate": 4.4473842551088135e-05, "loss": 0.0026, "step": 55850 }, { "grad_norm": 0.07606426626443863, "learning_rate": 4.445740945437738e-05, "loss": 0.0037, "step": 55860 }, { "grad_norm": 0.08354976773262024, "learning_rate": 4.444097696379602e-05, "loss": 0.0034, "step": 55870 }, { "grad_norm": 0.05964629352092743, "learning_rate": 4.442454508114106e-05, "loss": 0.0028, "step": 55880 }, { "grad_norm": 0.080839604139328, "learning_rate": 4.4408113808209465e-05, "loss": 0.0031, "step": 55890 }, { "grad_norm": 0.061718255281448364, "learning_rate": 4.439168314679813e-05, "loss": 0.0029, "step": 55900 }, { "grad_norm": 0.051144879311323166, "learning_rate": 4.43752530987039e-05, "loss": 0.003, "step": 55910 }, { "grad_norm": 0.06182872876524925, "learning_rate": 4.4358823665723515e-05, "loss": 0.0034, "step": 55920 }, { "grad_norm": 0.05098898336291313, "learning_rate": 4.434239484965371e-05, "loss": 0.003, "step": 55930 }, { "grad_norm": 0.054156169295310974, "learning_rate": 4.4325966652291103e-05, "loss": 0.0025, "step": 55940 }, { "grad_norm": 0.07329972833395004, "learning_rate": 4.430953907543225e-05, "loss": 0.0027, "step": 55950 }, { "grad_norm": 0.06405861675739288, "learning_rate": 4.4293112120873645e-05, "loss": 0.0033, "step": 55960 }, { "grad_norm": 0.07071537524461746, "learning_rate": 4.427668579041172e-05, "loss": 0.0034, "step": 55970 }, { "grad_norm": 0.06181332841515541, "learning_rate": 4.4260260085842816e-05, "loss": 0.0029, "step": 55980 }, { "grad_norm": 0.06752520054578781, "learning_rate": 4.4243835008963256e-05, "loss": 0.0027, "step": 55990 }, { "grad_norm": 0.08345477283000946, "learning_rate": 4.422741056156925e-05, "loss": 0.0037, "step": 56000 }, { "grad_norm": 0.07267504185438156, "learning_rate": 4.421098674545693e-05, "loss": 0.0031, "step": 56010 }, { "grad_norm": 0.06522353738546371, "learning_rate": 4.419456356242241e-05, "loss": 0.0036, "step": 56020 }, { "grad_norm": 0.07751329988241196, "learning_rate": 4.4178141014261685e-05, "loss": 0.003, "step": 56030 }, { "grad_norm": 0.08040901273488998, "learning_rate": 4.416171910277071e-05, "loss": 0.0035, "step": 56040 }, { "grad_norm": 0.061858147382736206, "learning_rate": 4.414529782974533e-05, "loss": 0.0039, "step": 56050 }, { "grad_norm": 0.06350447237491608, "learning_rate": 4.41288771969814e-05, "loss": 0.0035, "step": 56060 }, { "grad_norm": 0.07953305542469025, "learning_rate": 4.411245720627464e-05, "loss": 0.0037, "step": 56070 }, { "grad_norm": 0.05980215594172478, "learning_rate": 4.40960378594207e-05, "loss": 0.0032, "step": 56080 }, { "grad_norm": 0.06408168375492096, "learning_rate": 4.407961915821519e-05, "loss": 0.003, "step": 56090 }, { "grad_norm": 0.059422463178634644, "learning_rate": 4.406320110445362e-05, "loss": 0.0033, "step": 56100 }, { "grad_norm": 0.06347392499446869, "learning_rate": 4.404678369993144e-05, "loss": 0.0031, "step": 56110 }, { "grad_norm": 0.05351708456873894, "learning_rate": 4.403036694644406e-05, "loss": 0.003, "step": 56120 }, { "grad_norm": 0.06046707555651665, "learning_rate": 4.4013950845786764e-05, "loss": 0.0033, "step": 56130 }, { "grad_norm": 0.06878339499235153, "learning_rate": 4.399753539975482e-05, "loss": 0.0031, "step": 56140 }, { "grad_norm": 0.06894581764936447, "learning_rate": 4.398112061014337e-05, "loss": 0.0031, "step": 56150 }, { "grad_norm": 0.07712811976671219, "learning_rate": 4.396470647874753e-05, "loss": 0.0028, "step": 56160 }, { "grad_norm": 0.049274321645498276, "learning_rate": 4.394829300736229e-05, "loss": 0.0028, "step": 56170 }, { "grad_norm": 0.05235174298286438, "learning_rate": 4.393188019778265e-05, "loss": 0.003, "step": 56180 }, { "grad_norm": 0.04556494578719139, "learning_rate": 4.391546805180347e-05, "loss": 0.0025, "step": 56190 }, { "grad_norm": 0.06305812299251556, "learning_rate": 4.389905657121955e-05, "loss": 0.003, "step": 56200 }, { "grad_norm": 0.04542577266693115, "learning_rate": 4.3882645757825635e-05, "loss": 0.0027, "step": 56210 }, { "grad_norm": 0.06571152061223984, "learning_rate": 4.386623561341637e-05, "loss": 0.0028, "step": 56220 }, { "grad_norm": 0.07653562724590302, "learning_rate": 4.384982613978637e-05, "loss": 0.0032, "step": 56230 }, { "grad_norm": 0.07048452645540237, "learning_rate": 4.383341733873012e-05, "loss": 0.0027, "step": 56240 }, { "grad_norm": 0.06764551252126694, "learning_rate": 4.381700921204209e-05, "loss": 0.0032, "step": 56250 }, { "grad_norm": 0.0628749206662178, "learning_rate": 4.380060176151663e-05, "loss": 0.0027, "step": 56260 }, { "grad_norm": 0.05350765958428383, "learning_rate": 4.378419498894805e-05, "loss": 0.0028, "step": 56270 }, { "grad_norm": 0.06720013171434402, "learning_rate": 4.376778889613056e-05, "loss": 0.0034, "step": 56280 }, { "grad_norm": 0.07534290850162506, "learning_rate": 4.375138348485831e-05, "loss": 0.0029, "step": 56290 }, { "grad_norm": 0.0558319017291069, "learning_rate": 4.3734978756925346e-05, "loss": 0.0023, "step": 56300 }, { "grad_norm": 0.05489353463053703, "learning_rate": 4.3718574714125706e-05, "loss": 0.0027, "step": 56310 }, { "grad_norm": 0.03918996825814247, "learning_rate": 4.370217135825329e-05, "loss": 0.0019, "step": 56320 }, { "grad_norm": 0.04424019157886505, "learning_rate": 4.368576869110194e-05, "loss": 0.0028, "step": 56330 }, { "grad_norm": 0.04540124535560608, "learning_rate": 4.366936671446544e-05, "loss": 0.0028, "step": 56340 }, { "grad_norm": 0.06872114539146423, "learning_rate": 4.3652965430137474e-05, "loss": 0.0027, "step": 56350 }, { "grad_norm": 0.051628097891807556, "learning_rate": 4.3636564839911646e-05, "loss": 0.0033, "step": 56360 }, { "grad_norm": 0.08118006587028503, "learning_rate": 4.3620164945581545e-05, "loss": 0.0033, "step": 56370 }, { "grad_norm": 0.08205832540988922, "learning_rate": 4.360376574894061e-05, "loss": 0.0036, "step": 56380 }, { "grad_norm": 0.058015674352645874, "learning_rate": 4.358736725178224e-05, "loss": 0.0031, "step": 56390 }, { "grad_norm": 0.08266235142946243, "learning_rate": 4.357096945589974e-05, "loss": 0.0029, "step": 56400 }, { "grad_norm": 0.06693423539400101, "learning_rate": 4.355457236308636e-05, "loss": 0.0029, "step": 56410 }, { "grad_norm": 0.072325199842453, "learning_rate": 4.353817597513526e-05, "loss": 0.0028, "step": 56420 }, { "grad_norm": 0.06706110388040543, "learning_rate": 4.352178029383948e-05, "loss": 0.0031, "step": 56430 }, { "grad_norm": 0.05144727602601051, "learning_rate": 4.3505385320992105e-05, "loss": 0.0039, "step": 56440 }, { "grad_norm": 0.06985780596733093, "learning_rate": 4.348899105838602e-05, "loss": 0.0026, "step": 56450 }, { "grad_norm": 0.07966051995754242, "learning_rate": 4.3472597507814087e-05, "loss": 0.0033, "step": 56460 }, { "grad_norm": 0.07085063308477402, "learning_rate": 4.3456204671069066e-05, "loss": 0.0028, "step": 56470 }, { "grad_norm": 0.0629429742693901, "learning_rate": 4.343981254994367e-05, "loss": 0.0029, "step": 56480 }, { "grad_norm": 0.05527304485440254, "learning_rate": 4.34234211462305e-05, "loss": 0.0036, "step": 56490 }, { "grad_norm": 0.06489303708076477, "learning_rate": 4.340703046172213e-05, "loss": 0.0027, "step": 56500 }, { "grad_norm": 0.0852188691496849, "learning_rate": 4.339064049821097e-05, "loss": 0.0028, "step": 56510 }, { "grad_norm": 0.06024029850959778, "learning_rate": 4.3374251257489446e-05, "loss": 0.0029, "step": 56520 }, { "grad_norm": 0.06638191640377045, "learning_rate": 4.335786274134984e-05, "loss": 0.0027, "step": 56530 }, { "grad_norm": 0.06171778589487076, "learning_rate": 4.334147495158438e-05, "loss": 0.0031, "step": 56540 }, { "grad_norm": 0.07321732491254807, "learning_rate": 4.33250878899852e-05, "loss": 0.003, "step": 56550 }, { "grad_norm": 0.057023871690034866, "learning_rate": 4.330870155834437e-05, "loss": 0.0033, "step": 56560 }, { "grad_norm": 0.10232871025800705, "learning_rate": 4.3292315958453884e-05, "loss": 0.0042, "step": 56570 }, { "grad_norm": 0.07455004751682281, "learning_rate": 4.3275931092105646e-05, "loss": 0.0032, "step": 56580 }, { "grad_norm": 0.07831097394227982, "learning_rate": 4.325954696109147e-05, "loss": 0.0034, "step": 56590 }, { "grad_norm": 0.06479920446872711, "learning_rate": 4.32431635672031e-05, "loss": 0.0037, "step": 56600 }, { "grad_norm": 0.08474186062812805, "learning_rate": 4.3226780912232214e-05, "loss": 0.0036, "step": 56610 }, { "grad_norm": 0.07694212347269058, "learning_rate": 4.321039899797036e-05, "loss": 0.0031, "step": 56620 }, { "grad_norm": 0.07284723222255707, "learning_rate": 4.319401782620908e-05, "loss": 0.0029, "step": 56630 }, { "grad_norm": 0.1034780889749527, "learning_rate": 4.317763739873978e-05, "loss": 0.0036, "step": 56640 }, { "grad_norm": 0.08436909317970276, "learning_rate": 4.31612577173538e-05, "loss": 0.0032, "step": 56650 }, { "grad_norm": 0.08441431820392609, "learning_rate": 4.314487878384239e-05, "loss": 0.0029, "step": 56660 }, { "grad_norm": 0.08239561319351196, "learning_rate": 4.3128500599996736e-05, "loss": 0.0033, "step": 56670 }, { "grad_norm": 0.07450125366449356, "learning_rate": 4.31121231676079e-05, "loss": 0.0031, "step": 56680 }, { "grad_norm": 0.04996303841471672, "learning_rate": 4.309574648846694e-05, "loss": 0.0029, "step": 56690 }, { "grad_norm": 0.08704829216003418, "learning_rate": 4.3079370564364755e-05, "loss": 0.0035, "step": 56700 }, { "grad_norm": 0.07394249737262726, "learning_rate": 4.306299539709221e-05, "loss": 0.0044, "step": 56710 }, { "grad_norm": 0.09265124052762985, "learning_rate": 4.3046620988440046e-05, "loss": 0.0042, "step": 56720 }, { "grad_norm": 0.05605536326766014, "learning_rate": 4.303024734019897e-05, "loss": 0.0028, "step": 56730 }, { "grad_norm": 0.054041218012571335, "learning_rate": 4.3013874454159565e-05, "loss": 0.0025, "step": 56740 }, { "grad_norm": 0.06564813107252121, "learning_rate": 4.299750233211233e-05, "loss": 0.0035, "step": 56750 }, { "grad_norm": 0.07537639886140823, "learning_rate": 4.2981130975847715e-05, "loss": 0.0033, "step": 56760 }, { "grad_norm": 0.06897696107625961, "learning_rate": 4.2964760387156076e-05, "loss": 0.0034, "step": 56770 }, { "grad_norm": 0.06993842124938965, "learning_rate": 4.294839056782767e-05, "loss": 0.0034, "step": 56780 }, { "grad_norm": 0.06506302952766418, "learning_rate": 4.293202151965266e-05, "loss": 0.0029, "step": 56790 }, { "grad_norm": 0.06629980355501175, "learning_rate": 4.291565324442115e-05, "loss": 0.0032, "step": 56800 }, { "grad_norm": 0.08566152304410934, "learning_rate": 4.2899285743923135e-05, "loss": 0.0036, "step": 56810 }, { "grad_norm": 0.0644504725933075, "learning_rate": 4.288291901994857e-05, "loss": 0.0031, "step": 56820 }, { "grad_norm": 0.07169291377067566, "learning_rate": 4.286655307428729e-05, "loss": 0.0029, "step": 56830 }, { "grad_norm": 0.062115151435136795, "learning_rate": 4.285018790872903e-05, "loss": 0.0041, "step": 56840 }, { "grad_norm": 0.050516482442617416, "learning_rate": 4.283382352506347e-05, "loss": 0.0034, "step": 56850 }, { "grad_norm": 0.0637320950627327, "learning_rate": 4.2817459925080205e-05, "loss": 0.0026, "step": 56860 }, { "grad_norm": 0.06419476121664047, "learning_rate": 4.28010971105687e-05, "loss": 0.0026, "step": 56870 }, { "grad_norm": 0.07619834691286087, "learning_rate": 4.278473508331841e-05, "loss": 0.0025, "step": 56880 }, { "grad_norm": 0.08171170949935913, "learning_rate": 4.276837384511864e-05, "loss": 0.0034, "step": 56890 }, { "grad_norm": 0.05218949913978577, "learning_rate": 4.275201339775864e-05, "loss": 0.0035, "step": 56900 }, { "grad_norm": 0.0644574835896492, "learning_rate": 4.273565374302756e-05, "loss": 0.0031, "step": 56910 }, { "grad_norm": 0.04919494315981865, "learning_rate": 4.2719294882714454e-05, "loss": 0.0026, "step": 56920 }, { "grad_norm": 0.04526636749505997, "learning_rate": 4.270293681860831e-05, "loss": 0.0024, "step": 56930 }, { "grad_norm": 0.06949150562286377, "learning_rate": 4.268657955249802e-05, "loss": 0.0025, "step": 56940 }, { "grad_norm": 0.05576414614915848, "learning_rate": 4.267022308617241e-05, "loss": 0.0028, "step": 56950 }, { "grad_norm": 0.0619007907807827, "learning_rate": 4.265386742142018e-05, "loss": 0.0034, "step": 56960 }, { "grad_norm": 0.08212394267320633, "learning_rate": 4.263751256002996e-05, "loss": 0.0032, "step": 56970 }, { "grad_norm": 0.04018263891339302, "learning_rate": 4.26211585037903e-05, "loss": 0.0031, "step": 56980 }, { "grad_norm": 0.05887237936258316, "learning_rate": 4.2604805254489656e-05, "loss": 0.0036, "step": 56990 }, { "grad_norm": 0.06460556387901306, "learning_rate": 4.258845281391638e-05, "loss": 0.003, "step": 57000 }, { "grad_norm": 0.07410982996225357, "learning_rate": 4.257210118385876e-05, "loss": 0.0031, "step": 57010 }, { "grad_norm": 0.060336824506521225, "learning_rate": 4.2555750366104994e-05, "loss": 0.0034, "step": 57020 }, { "grad_norm": 0.07465240359306335, "learning_rate": 4.2539400362443184e-05, "loss": 0.0043, "step": 57030 }, { "grad_norm": 0.06435434520244598, "learning_rate": 4.252305117466133e-05, "loss": 0.0031, "step": 57040 }, { "grad_norm": 0.06869972497224808, "learning_rate": 4.2506702804547356e-05, "loss": 0.0034, "step": 57050 }, { "grad_norm": 0.06115391477942467, "learning_rate": 4.249035525388907e-05, "loss": 0.0033, "step": 57060 }, { "grad_norm": 0.05451355502009392, "learning_rate": 4.247400852447427e-05, "loss": 0.0033, "step": 57070 }, { "grad_norm": 0.058689143508672714, "learning_rate": 4.245766261809059e-05, "loss": 0.0031, "step": 57080 }, { "grad_norm": 0.06528064608573914, "learning_rate": 4.244131753652558e-05, "loss": 0.0033, "step": 57090 }, { "grad_norm": 0.07283536344766617, "learning_rate": 4.242497328156672e-05, "loss": 0.0027, "step": 57100 }, { "grad_norm": 0.05912497267127037, "learning_rate": 4.240862985500139e-05, "loss": 0.0031, "step": 57110 }, { "grad_norm": 0.08165616542100906, "learning_rate": 4.2392287258616894e-05, "loss": 0.0035, "step": 57120 }, { "grad_norm": 0.05271412804722786, "learning_rate": 4.2375945494200396e-05, "loss": 0.0025, "step": 57130 }, { "grad_norm": 0.05970517545938492, "learning_rate": 4.2359604563539054e-05, "loss": 0.003, "step": 57140 }, { "grad_norm": 0.06884311884641647, "learning_rate": 4.2343264468419885e-05, "loss": 0.0028, "step": 57150 }, { "grad_norm": 0.09161267429590225, "learning_rate": 4.232692521062979e-05, "loss": 0.0031, "step": 57160 }, { "grad_norm": 0.07455874234437943, "learning_rate": 4.231058679195562e-05, "loss": 0.0028, "step": 57170 }, { "grad_norm": 0.0851885974407196, "learning_rate": 4.229424921418411e-05, "loss": 0.0028, "step": 57180 }, { "grad_norm": 0.06484757363796234, "learning_rate": 4.2277912479101916e-05, "loss": 0.0032, "step": 57190 }, { "grad_norm": 0.06275199353694916, "learning_rate": 4.226157658849562e-05, "loss": 0.0031, "step": 57200 }, { "grad_norm": 0.06741860508918762, "learning_rate": 4.2245241544151664e-05, "loss": 0.0048, "step": 57210 }, { "grad_norm": 0.09024395793676376, "learning_rate": 4.2228907347856425e-05, "loss": 0.0032, "step": 57220 }, { "grad_norm": 0.06400419771671295, "learning_rate": 4.2212574001396214e-05, "loss": 0.004, "step": 57230 }, { "grad_norm": 0.06327425688505173, "learning_rate": 4.21962415065572e-05, "loss": 0.0033, "step": 57240 }, { "grad_norm": 0.08759024739265442, "learning_rate": 4.217990986512548e-05, "loss": 0.0036, "step": 57250 }, { "grad_norm": 0.09157595783472061, "learning_rate": 4.216357907888704e-05, "loss": 0.0038, "step": 57260 }, { "grad_norm": 0.04855271056294441, "learning_rate": 4.2147249149627824e-05, "loss": 0.0031, "step": 57270 }, { "grad_norm": 0.06506556272506714, "learning_rate": 4.213092007913364e-05, "loss": 0.0029, "step": 57280 }, { "grad_norm": 0.08366706967353821, "learning_rate": 4.2114591869190205e-05, "loss": 0.0035, "step": 57290 }, { "grad_norm": 0.07450196892023087, "learning_rate": 4.209826452158315e-05, "loss": 0.0029, "step": 57300 }, { "grad_norm": 0.0340396948158741, "learning_rate": 4.2081938038098e-05, "loss": 0.0023, "step": 57310 }, { "grad_norm": 0.05458392575383186, "learning_rate": 4.2065612420520185e-05, "loss": 0.0039, "step": 57320 }, { "grad_norm": 0.060650452971458435, "learning_rate": 4.204928767063509e-05, "loss": 0.0026, "step": 57330 }, { "grad_norm": 0.07418211549520493, "learning_rate": 4.2032963790227944e-05, "loss": 0.0029, "step": 57340 }, { "grad_norm": 0.07461494952440262, "learning_rate": 4.2016640781083887e-05, "loss": 0.0038, "step": 57350 }, { "grad_norm": 0.05667170509696007, "learning_rate": 4.2000318644988004e-05, "loss": 0.0025, "step": 57360 }, { "grad_norm": 0.052400052547454834, "learning_rate": 4.198399738372524e-05, "loss": 0.0026, "step": 57370 }, { "grad_norm": 0.06897326558828354, "learning_rate": 4.196767699908045e-05, "loss": 0.0031, "step": 57380 }, { "grad_norm": 0.05723177269101143, "learning_rate": 4.195135749283845e-05, "loss": 0.0036, "step": 57390 }, { "grad_norm": 0.06220942735671997, "learning_rate": 4.1935038866783894e-05, "loss": 0.0035, "step": 57400 }, { "grad_norm": 0.07384882867336273, "learning_rate": 4.1918721122701355e-05, "loss": 0.003, "step": 57410 }, { "grad_norm": 0.07880334556102753, "learning_rate": 4.190240426237533e-05, "loss": 0.0028, "step": 57420 }, { "grad_norm": 0.05569564923644066, "learning_rate": 4.188608828759019e-05, "loss": 0.0025, "step": 57430 }, { "grad_norm": 0.06011110171675682, "learning_rate": 4.1869773200130246e-05, "loss": 0.0034, "step": 57440 }, { "grad_norm": 0.06139937788248062, "learning_rate": 4.185345900177967e-05, "loss": 0.0029, "step": 57450 }, { "grad_norm": 0.060030315071344376, "learning_rate": 4.183714569432258e-05, "loss": 0.0026, "step": 57460 }, { "grad_norm": 0.061602089554071426, "learning_rate": 4.182083327954295e-05, "loss": 0.0036, "step": 57470 }, { "grad_norm": 0.05596744269132614, "learning_rate": 4.1804521759224714e-05, "loss": 0.003, "step": 57480 }, { "grad_norm": 0.06948375701904297, "learning_rate": 4.178821113515165e-05, "loss": 0.0024, "step": 57490 }, { "grad_norm": 0.07729671150445938, "learning_rate": 4.1771901409107476e-05, "loss": 0.0033, "step": 57500 }, { "grad_norm": 0.05550909787416458, "learning_rate": 4.175559258287577e-05, "loss": 0.0025, "step": 57510 }, { "grad_norm": 0.06842321902513504, "learning_rate": 4.173928465824008e-05, "loss": 0.0031, "step": 57520 }, { "grad_norm": 0.05983792990446091, "learning_rate": 4.17229776369838e-05, "loss": 0.004, "step": 57530 }, { "grad_norm": 0.07546954602003098, "learning_rate": 4.170667152089025e-05, "loss": 0.0041, "step": 57540 }, { "grad_norm": 0.0549384243786335, "learning_rate": 4.1690366311742626e-05, "loss": 0.0042, "step": 57550 }, { "grad_norm": 0.058468904346227646, "learning_rate": 4.1674062011324055e-05, "loss": 0.0029, "step": 57560 }, { "grad_norm": 0.07872319966554642, "learning_rate": 4.165775862141752e-05, "loss": 0.003, "step": 57570 }, { "grad_norm": 0.056899040937423706, "learning_rate": 4.164145614380598e-05, "loss": 0.0032, "step": 57580 }, { "grad_norm": 0.08156264573335648, "learning_rate": 4.162515458027223e-05, "loss": 0.0038, "step": 57590 }, { "grad_norm": 0.09631458669900894, "learning_rate": 4.160885393259899e-05, "loss": 0.0034, "step": 57600 }, { "grad_norm": 0.0632631704211235, "learning_rate": 4.159255420256886e-05, "loss": 0.0024, "step": 57610 }, { "grad_norm": 0.06275101006031036, "learning_rate": 4.157625539196436e-05, "loss": 0.0028, "step": 57620 }, { "grad_norm": 0.08168555796146393, "learning_rate": 4.155995750256791e-05, "loss": 0.0029, "step": 57630 }, { "grad_norm": 0.07394898682832718, "learning_rate": 4.15436605361618e-05, "loss": 0.0034, "step": 57640 }, { "grad_norm": 0.1153346374630928, "learning_rate": 4.152736449452827e-05, "loss": 0.0029, "step": 57650 }, { "grad_norm": 0.04123327508568764, "learning_rate": 4.151106937944943e-05, "loss": 0.0025, "step": 57660 }, { "grad_norm": 0.043483756482601166, "learning_rate": 4.149477519270727e-05, "loss": 0.0037, "step": 57670 }, { "grad_norm": 0.06320126354694366, "learning_rate": 4.147848193608371e-05, "loss": 0.0044, "step": 57680 }, { "grad_norm": 0.060807958245277405, "learning_rate": 4.146218961136056e-05, "loss": 0.0031, "step": 57690 }, { "grad_norm": 0.0890812799334526, "learning_rate": 4.14458982203195e-05, "loss": 0.0037, "step": 57700 }, { "grad_norm": 0.06820295006036758, "learning_rate": 4.1429607764742165e-05, "loss": 0.0037, "step": 57710 }, { "grad_norm": 0.05950560420751572, "learning_rate": 4.141331824641003e-05, "loss": 0.0037, "step": 57720 }, { "grad_norm": 0.07680241018533707, "learning_rate": 4.139702966710452e-05, "loss": 0.0026, "step": 57730 }, { "grad_norm": 0.07964510470628738, "learning_rate": 4.13807420286069e-05, "loss": 0.003, "step": 57740 }, { "grad_norm": 0.06580732017755508, "learning_rate": 4.136445533269837e-05, "loss": 0.0027, "step": 57750 }, { "grad_norm": 0.08175873756408691, "learning_rate": 4.134816958116003e-05, "loss": 0.0036, "step": 57760 }, { "grad_norm": 0.06021289899945259, "learning_rate": 4.133188477577283e-05, "loss": 0.0026, "step": 57770 }, { "grad_norm": 0.06739094108343124, "learning_rate": 4.131560091831771e-05, "loss": 0.0026, "step": 57780 }, { "grad_norm": 0.07387366145849228, "learning_rate": 4.1299318010575405e-05, "loss": 0.0033, "step": 57790 }, { "grad_norm": 0.08685587346553802, "learning_rate": 4.128303605432661e-05, "loss": 0.0037, "step": 57800 }, { "grad_norm": 0.07855407893657684, "learning_rate": 4.126675505135188e-05, "loss": 0.0028, "step": 57810 }, { "grad_norm": 0.06603267788887024, "learning_rate": 4.125047500343169e-05, "loss": 0.0026, "step": 57820 }, { "grad_norm": 0.04843095690011978, "learning_rate": 4.1234195912346376e-05, "loss": 0.0031, "step": 57830 }, { "grad_norm": 0.0584278367459774, "learning_rate": 4.1217917779876235e-05, "loss": 0.0029, "step": 57840 }, { "grad_norm": 0.0568060539662838, "learning_rate": 4.120164060780141e-05, "loss": 0.0025, "step": 57850 }, { "grad_norm": 0.04594050347805023, "learning_rate": 4.1185364397901935e-05, "loss": 0.0024, "step": 57860 }, { "grad_norm": 0.06295958161354065, "learning_rate": 4.116908915195776e-05, "loss": 0.0028, "step": 57870 }, { "grad_norm": 0.06555863469839096, "learning_rate": 4.115281487174872e-05, "loss": 0.0035, "step": 57880 }, { "grad_norm": 0.062072958797216415, "learning_rate": 4.113654155905452e-05, "loss": 0.0029, "step": 57890 }, { "grad_norm": 0.07132801413536072, "learning_rate": 4.112026921565483e-05, "loss": 0.0041, "step": 57900 }, { "grad_norm": 0.07460293918848038, "learning_rate": 4.1103997843329147e-05, "loss": 0.003, "step": 57910 }, { "grad_norm": 0.07006032764911652, "learning_rate": 4.108772744385689e-05, "loss": 0.0025, "step": 57920 }, { "grad_norm": 0.06860900670289993, "learning_rate": 4.1071458019017346e-05, "loss": 0.003, "step": 57930 }, { "grad_norm": 0.07321260869503021, "learning_rate": 4.1055189570589734e-05, "loss": 0.0036, "step": 57940 }, { "grad_norm": 0.07093542814254761, "learning_rate": 4.103892210035315e-05, "loss": 0.0035, "step": 57950 }, { "grad_norm": 0.0640445351600647, "learning_rate": 4.102265561008656e-05, "loss": 0.003, "step": 57960 }, { "grad_norm": 0.06567678600549698, "learning_rate": 4.100639010156885e-05, "loss": 0.0023, "step": 57970 }, { "grad_norm": 0.05986884981393814, "learning_rate": 4.099012557657882e-05, "loss": 0.0027, "step": 57980 }, { "grad_norm": 0.06292091310024261, "learning_rate": 4.09738620368951e-05, "loss": 0.0029, "step": 57990 }, { "grad_norm": 0.06833485513925552, "learning_rate": 4.095759948429626e-05, "loss": 0.0029, "step": 58000 }, { "grad_norm": 0.06261065602302551, "learning_rate": 4.094133792056075e-05, "loss": 0.0039, "step": 58010 }, { "grad_norm": 0.08527731150388718, "learning_rate": 4.092507734746688e-05, "loss": 0.0028, "step": 58020 }, { "grad_norm": 0.07357887923717499, "learning_rate": 4.090881776679293e-05, "loss": 0.0027, "step": 58030 }, { "grad_norm": 0.05840827897191048, "learning_rate": 4.0892559180316994e-05, "loss": 0.0038, "step": 58040 }, { "grad_norm": 0.06901530921459198, "learning_rate": 4.08763015898171e-05, "loss": 0.003, "step": 58050 }, { "grad_norm": 0.1005757600069046, "learning_rate": 4.0860044997071145e-05, "loss": 0.0032, "step": 58060 }, { "grad_norm": 0.07116129249334335, "learning_rate": 4.084378940385693e-05, "loss": 0.0036, "step": 58070 }, { "grad_norm": 0.06992626935243607, "learning_rate": 4.082753481195211e-05, "loss": 0.0035, "step": 58080 }, { "grad_norm": 0.06374181807041168, "learning_rate": 4.0811281223134324e-05, "loss": 0.0037, "step": 58090 }, { "grad_norm": 0.0764954537153244, "learning_rate": 4.0795028639181e-05, "loss": 0.0034, "step": 58100 }, { "grad_norm": 0.11529205739498138, "learning_rate": 4.077877706186951e-05, "loss": 0.0034, "step": 58110 }, { "grad_norm": 0.06531000882387161, "learning_rate": 4.076252649297709e-05, "loss": 0.0031, "step": 58120 }, { "grad_norm": 0.0585605725646019, "learning_rate": 4.07462769342809e-05, "loss": 0.0021, "step": 58130 }, { "grad_norm": 0.05106666311621666, "learning_rate": 4.073002838755793e-05, "loss": 0.0031, "step": 58140 }, { "grad_norm": 0.05393381416797638, "learning_rate": 4.0713780854585115e-05, "loss": 0.0024, "step": 58150 }, { "grad_norm": 0.05027082934975624, "learning_rate": 4.069753433713927e-05, "loss": 0.0023, "step": 58160 }, { "grad_norm": 0.052522387355566025, "learning_rate": 4.068128883699709e-05, "loss": 0.0034, "step": 58170 }, { "grad_norm": 0.05906074494123459, "learning_rate": 4.0665044355935155e-05, "loss": 0.0031, "step": 58180 }, { "grad_norm": 0.06698473542928696, "learning_rate": 4.064880089572993e-05, "loss": 0.0036, "step": 58190 }, { "grad_norm": 0.07018470764160156, "learning_rate": 4.063255845815779e-05, "loss": 0.0026, "step": 58200 }, { "grad_norm": 0.04003674536943436, "learning_rate": 4.0616317044994945e-05, "loss": 0.0025, "step": 58210 }, { "grad_norm": 0.06907110661268234, "learning_rate": 4.0600076658017585e-05, "loss": 0.0028, "step": 58220 }, { "grad_norm": 0.0904969871044159, "learning_rate": 4.0583837299001706e-05, "loss": 0.0028, "step": 58230 }, { "grad_norm": 0.07979042828083038, "learning_rate": 4.056759896972322e-05, "loss": 0.003, "step": 58240 }, { "grad_norm": 0.09241118282079697, "learning_rate": 4.055136167195793e-05, "loss": 0.0033, "step": 58250 }, { "grad_norm": 0.05414407700300217, "learning_rate": 4.053512540748152e-05, "loss": 0.003, "step": 58260 }, { "grad_norm": 0.053112905472517014, "learning_rate": 4.051889017806955e-05, "loss": 0.0033, "step": 58270 }, { "grad_norm": 0.07500405609607697, "learning_rate": 4.0502655985497516e-05, "loss": 0.0031, "step": 58280 }, { "grad_norm": 0.07781071960926056, "learning_rate": 4.048642283154074e-05, "loss": 0.0035, "step": 58290 }, { "grad_norm": 0.06022065877914429, "learning_rate": 4.0470190717974464e-05, "loss": 0.0039, "step": 58300 }, { "grad_norm": 0.04311254620552063, "learning_rate": 4.045395964657381e-05, "loss": 0.0025, "step": 58310 }, { "grad_norm": 0.055600572377443314, "learning_rate": 4.043772961911376e-05, "loss": 0.0025, "step": 58320 }, { "grad_norm": 0.056040164083242416, "learning_rate": 4.042150063736924e-05, "loss": 0.0027, "step": 58330 }, { "grad_norm": 0.06222902610898018, "learning_rate": 4.0405272703114976e-05, "loss": 0.0029, "step": 58340 }, { "grad_norm": 0.07238715887069702, "learning_rate": 4.038904581812569e-05, "loss": 0.0041, "step": 58350 }, { "grad_norm": 0.06539846956729889, "learning_rate": 4.03728199841759e-05, "loss": 0.0023, "step": 58360 }, { "grad_norm": 0.05476079881191254, "learning_rate": 4.0356595203040044e-05, "loss": 0.0029, "step": 58370 }, { "grad_norm": 0.04276774451136589, "learning_rate": 4.0340371476492426e-05, "loss": 0.0027, "step": 58380 }, { "grad_norm": 0.06386934965848923, "learning_rate": 4.032414880630727e-05, "loss": 0.0023, "step": 58390 }, { "grad_norm": 0.05533034726977348, "learning_rate": 4.0307927194258646e-05, "loss": 0.0036, "step": 58400 }, { "grad_norm": 0.056875020265579224, "learning_rate": 4.029170664212054e-05, "loss": 0.0026, "step": 58410 }, { "grad_norm": 0.04987052083015442, "learning_rate": 4.0275487151666794e-05, "loss": 0.0022, "step": 58420 }, { "grad_norm": 0.059306200593709946, "learning_rate": 4.025926872467116e-05, "loss": 0.0033, "step": 58430 }, { "grad_norm": 0.049694109708070755, "learning_rate": 4.024305136290725e-05, "loss": 0.0035, "step": 58440 }, { "grad_norm": 0.07142554223537445, "learning_rate": 4.022683506814858e-05, "loss": 0.0025, "step": 58450 }, { "grad_norm": 0.07530716061592102, "learning_rate": 4.021061984216853e-05, "loss": 0.0031, "step": 58460 }, { "grad_norm": 0.06718266010284424, "learning_rate": 4.019440568674035e-05, "loss": 0.0025, "step": 58470 }, { "grad_norm": 0.06511162221431732, "learning_rate": 4.017819260363724e-05, "loss": 0.0031, "step": 58480 }, { "grad_norm": 0.04934697970747948, "learning_rate": 4.016198059463223e-05, "loss": 0.0032, "step": 58490 }, { "grad_norm": 0.0664902999997139, "learning_rate": 4.0145769661498233e-05, "loss": 0.0025, "step": 58500 }, { "grad_norm": 0.04582475870847702, "learning_rate": 4.012955980600804e-05, "loss": 0.0031, "step": 58510 }, { "grad_norm": 0.07205628603696823, "learning_rate": 4.0113351029934355e-05, "loss": 0.0032, "step": 58520 }, { "grad_norm": 0.06295671314001083, "learning_rate": 4.0097143335049714e-05, "loss": 0.0034, "step": 58530 }, { "grad_norm": 0.058777764439582825, "learning_rate": 4.0080936723126606e-05, "loss": 0.0025, "step": 58540 }, { "grad_norm": 0.0861019641160965, "learning_rate": 4.006473119593735e-05, "loss": 0.0024, "step": 58550 }, { "grad_norm": 0.05551362410187721, "learning_rate": 4.004852675525415e-05, "loss": 0.003, "step": 58560 }, { "grad_norm": 0.0679679811000824, "learning_rate": 4.00323234028491e-05, "loss": 0.0036, "step": 58570 }, { "grad_norm": 0.06979599595069885, "learning_rate": 4.0016121140494186e-05, "loss": 0.0031, "step": 58580 }, { "grad_norm": 0.06929274648427963, "learning_rate": 3.999991996996122e-05, "loss": 0.0029, "step": 58590 }, { "grad_norm": 0.0672278180718422, "learning_rate": 3.9983719893022e-05, "loss": 0.0031, "step": 58600 }, { "grad_norm": 0.07860026508569717, "learning_rate": 3.996752091144812e-05, "loss": 0.0028, "step": 58610 }, { "grad_norm": 0.0979527086019516, "learning_rate": 3.9951323027011055e-05, "loss": 0.0027, "step": 58620 }, { "grad_norm": 0.04652149975299835, "learning_rate": 3.993512624148219e-05, "loss": 0.0031, "step": 58630 }, { "grad_norm": 0.07859644293785095, "learning_rate": 3.99189305566328e-05, "loss": 0.0027, "step": 58640 }, { "grad_norm": 0.05165974795818329, "learning_rate": 3.990273597423401e-05, "loss": 0.0034, "step": 58650 }, { "grad_norm": 0.0728430524468422, "learning_rate": 3.988654249605681e-05, "loss": 0.0038, "step": 58660 }, { "grad_norm": 0.05867123603820801, "learning_rate": 3.987035012387213e-05, "loss": 0.0027, "step": 58670 }, { "grad_norm": 0.10106132179498672, "learning_rate": 3.9854158859450735e-05, "loss": 0.0029, "step": 58680 }, { "grad_norm": 0.0684896931052208, "learning_rate": 3.983796870456328e-05, "loss": 0.0031, "step": 58690 }, { "grad_norm": 0.04890608787536621, "learning_rate": 3.982177966098028e-05, "loss": 0.0035, "step": 58700 }, { "grad_norm": 0.0516398660838604, "learning_rate": 3.980559173047216e-05, "loss": 0.0026, "step": 58710 }, { "grad_norm": 0.057605136185884476, "learning_rate": 3.9789404914809176e-05, "loss": 0.0032, "step": 58720 }, { "grad_norm": 0.05657421052455902, "learning_rate": 3.9773219215761544e-05, "loss": 0.0032, "step": 58730 }, { "grad_norm": 0.06170077994465828, "learning_rate": 3.975703463509928e-05, "loss": 0.0029, "step": 58740 }, { "grad_norm": 0.06443402171134949, "learning_rate": 3.974085117459231e-05, "loss": 0.0025, "step": 58750 }, { "grad_norm": 0.07315517961978912, "learning_rate": 3.972466883601043e-05, "loss": 0.0034, "step": 58760 }, { "grad_norm": 0.08576364815235138, "learning_rate": 3.9708487621123316e-05, "loss": 0.0031, "step": 58770 }, { "grad_norm": 0.08412965387105942, "learning_rate": 3.969230753170051e-05, "loss": 0.0026, "step": 58780 }, { "grad_norm": 0.08222927898168564, "learning_rate": 3.967612856951146e-05, "loss": 0.0038, "step": 58790 }, { "grad_norm": 0.04882480949163437, "learning_rate": 3.9659950736325476e-05, "loss": 0.0035, "step": 58800 }, { "grad_norm": 0.06703003495931625, "learning_rate": 3.964377403391174e-05, "loss": 0.0027, "step": 58810 }, { "grad_norm": 0.05926154553890228, "learning_rate": 3.9627598464039303e-05, "loss": 0.003, "step": 58820 }, { "grad_norm": 0.07973229885101318, "learning_rate": 3.9611424028477096e-05, "loss": 0.003, "step": 58830 }, { "grad_norm": 0.0673028826713562, "learning_rate": 3.959525072899394e-05, "loss": 0.0037, "step": 58840 }, { "grad_norm": 0.055879924446344376, "learning_rate": 3.9579078567358506e-05, "loss": 0.0025, "step": 58850 }, { "grad_norm": 0.052396759390830994, "learning_rate": 3.956290754533939e-05, "loss": 0.0027, "step": 58860 }, { "grad_norm": 0.06555281579494476, "learning_rate": 3.954673766470501e-05, "loss": 0.0028, "step": 58870 }, { "grad_norm": 0.05011405050754547, "learning_rate": 3.9530568927223674e-05, "loss": 0.003, "step": 58880 }, { "grad_norm": 0.07482147216796875, "learning_rate": 3.95144013346636e-05, "loss": 0.003, "step": 58890 }, { "grad_norm": 0.05710356682538986, "learning_rate": 3.949823488879282e-05, "loss": 0.003, "step": 58900 }, { "grad_norm": 0.0682315081357956, "learning_rate": 3.9482069591379276e-05, "loss": 0.0026, "step": 58910 }, { "grad_norm": 0.05482511222362518, "learning_rate": 3.9465905444190784e-05, "loss": 0.0024, "step": 58920 }, { "grad_norm": 0.108017697930336, "learning_rate": 3.944974244899505e-05, "loss": 0.0031, "step": 58930 }, { "grad_norm": 0.06765371561050415, "learning_rate": 3.9433580607559616e-05, "loss": 0.0028, "step": 58940 }, { "grad_norm": 0.0529576875269413, "learning_rate": 3.941741992165193e-05, "loss": 0.0029, "step": 58950 }, { "grad_norm": 0.0428047850728035, "learning_rate": 3.940126039303928e-05, "loss": 0.0028, "step": 58960 }, { "grad_norm": 0.047024015337228775, "learning_rate": 3.938510202348886e-05, "loss": 0.003, "step": 58970 }, { "grad_norm": 0.07191241532564163, "learning_rate": 3.93689448147677e-05, "loss": 0.0024, "step": 58980 }, { "grad_norm": 0.06497794389724731, "learning_rate": 3.935278876864278e-05, "loss": 0.0025, "step": 58990 }, { "grad_norm": 0.07055133581161499, "learning_rate": 3.933663388688087e-05, "loss": 0.0034, "step": 59000 }, { "grad_norm": 0.05293697118759155, "learning_rate": 3.9320480171248645e-05, "loss": 0.0029, "step": 59010 }, { "grad_norm": 0.095303975045681, "learning_rate": 3.930432762351265e-05, "loss": 0.0024, "step": 59020 }, { "grad_norm": 0.05874602496623993, "learning_rate": 3.928817624543931e-05, "loss": 0.0033, "step": 59030 }, { "grad_norm": 0.058222513645887375, "learning_rate": 3.9272026038794885e-05, "loss": 0.0028, "step": 59040 }, { "grad_norm": 0.07444901019334793, "learning_rate": 3.9255877005345586e-05, "loss": 0.0024, "step": 59050 }, { "grad_norm": 0.08076885342597961, "learning_rate": 3.923972914685743e-05, "loss": 0.0027, "step": 59060 }, { "grad_norm": 0.06723776459693909, "learning_rate": 3.922358246509631e-05, "loss": 0.0026, "step": 59070 }, { "grad_norm": 0.047618068754673004, "learning_rate": 3.920743696182801e-05, "loss": 0.0023, "step": 59080 }, { "grad_norm": 0.07044854760169983, "learning_rate": 3.9191292638818164e-05, "loss": 0.0029, "step": 59090 }, { "grad_norm": 0.09019937366247177, "learning_rate": 3.9175149497832316e-05, "loss": 0.0045, "step": 59100 }, { "grad_norm": 0.10152985155582428, "learning_rate": 3.915900754063584e-05, "loss": 0.0039, "step": 59110 }, { "grad_norm": 0.07896628975868225, "learning_rate": 3.9142866768994014e-05, "loss": 0.0026, "step": 59120 }, { "grad_norm": 0.08491577953100204, "learning_rate": 3.912672718467193e-05, "loss": 0.0037, "step": 59130 }, { "grad_norm": 0.06838691234588623, "learning_rate": 3.911058878943463e-05, "loss": 0.0025, "step": 59140 }, { "grad_norm": 0.05733520910143852, "learning_rate": 3.9094451585046974e-05, "loss": 0.0024, "step": 59150 }, { "grad_norm": 0.07726708799600601, "learning_rate": 3.90783155732737e-05, "loss": 0.0025, "step": 59160 }, { "grad_norm": 0.044344592839479446, "learning_rate": 3.9062180755879373e-05, "loss": 0.0024, "step": 59170 }, { "grad_norm": 0.05474383383989334, "learning_rate": 3.904604713462855e-05, "loss": 0.0043, "step": 59180 }, { "grad_norm": 0.08159075677394867, "learning_rate": 3.902991471128554e-05, "loss": 0.0029, "step": 59190 }, { "grad_norm": 0.058816224336624146, "learning_rate": 3.901378348761456e-05, "loss": 0.0029, "step": 59200 }, { "grad_norm": 0.06179400533437729, "learning_rate": 3.899765346537971e-05, "loss": 0.0027, "step": 59210 }, { "grad_norm": 0.04908617585897446, "learning_rate": 3.898152464634493e-05, "loss": 0.0028, "step": 59220 }, { "grad_norm": 0.06734712421894073, "learning_rate": 3.896539703227402e-05, "loss": 0.0042, "step": 59230 }, { "grad_norm": 0.07227928936481476, "learning_rate": 3.894927062493073e-05, "loss": 0.0031, "step": 59240 }, { "grad_norm": 0.056654851883649826, "learning_rate": 3.893314542607859e-05, "loss": 0.0035, "step": 59250 }, { "grad_norm": 0.06768136471509933, "learning_rate": 3.891702143748103e-05, "loss": 0.0026, "step": 59260 }, { "grad_norm": 0.06590335816144943, "learning_rate": 3.890089866090133e-05, "loss": 0.0033, "step": 59270 }, { "grad_norm": 0.03810388594865799, "learning_rate": 3.888477709810267e-05, "loss": 0.0027, "step": 59280 }, { "grad_norm": 0.054734304547309875, "learning_rate": 3.886865675084806e-05, "loss": 0.0039, "step": 59290 }, { "grad_norm": 0.08558259159326553, "learning_rate": 3.885253762090043e-05, "loss": 0.003, "step": 59300 }, { "grad_norm": 0.08196742087602615, "learning_rate": 3.8836419710022534e-05, "loss": 0.0035, "step": 59310 }, { "grad_norm": 0.05241088569164276, "learning_rate": 3.882030301997698e-05, "loss": 0.0027, "step": 59320 }, { "grad_norm": 0.06160182133316994, "learning_rate": 3.880418755252629e-05, "loss": 0.003, "step": 59330 }, { "grad_norm": 0.053849659860134125, "learning_rate": 3.87880733094328e-05, "loss": 0.0023, "step": 59340 }, { "grad_norm": 0.060595881193876266, "learning_rate": 3.877196029245877e-05, "loss": 0.0026, "step": 59350 }, { "grad_norm": 0.05005432665348053, "learning_rate": 3.875584850336627e-05, "loss": 0.0037, "step": 59360 }, { "grad_norm": 0.06167981028556824, "learning_rate": 3.873973794391728e-05, "loss": 0.0032, "step": 59370 }, { "grad_norm": 0.06534036248922348, "learning_rate": 3.872362861587361e-05, "loss": 0.0027, "step": 59380 }, { "grad_norm": 0.06858213990926743, "learning_rate": 3.870752052099698e-05, "loss": 0.0029, "step": 59390 }, { "grad_norm": 0.07312510907649994, "learning_rate": 3.8691413661048916e-05, "loss": 0.0028, "step": 59400 }, { "grad_norm": 0.05542868748307228, "learning_rate": 3.867530803779085e-05, "loss": 0.0029, "step": 59410 }, { "grad_norm": 0.07244444638490677, "learning_rate": 3.8659203652984054e-05, "loss": 0.0032, "step": 59420 }, { "grad_norm": 0.07752229273319244, "learning_rate": 3.864310050838972e-05, "loss": 0.0024, "step": 59430 }, { "grad_norm": 0.05460160970687866, "learning_rate": 3.8626998605768835e-05, "loss": 0.0029, "step": 59440 }, { "grad_norm": 0.05395353212952614, "learning_rate": 3.8610897946882294e-05, "loss": 0.0024, "step": 59450 }, { "grad_norm": 0.07767852395772934, "learning_rate": 3.859479853349082e-05, "loss": 0.0034, "step": 59460 }, { "grad_norm": 0.07430329918861389, "learning_rate": 3.857870036735503e-05, "loss": 0.0027, "step": 59470 }, { "grad_norm": 0.03968563303351402, "learning_rate": 3.856260345023538e-05, "loss": 0.0032, "step": 59480 }, { "grad_norm": 0.053960442543029785, "learning_rate": 3.854650778389224e-05, "loss": 0.0034, "step": 59490 }, { "grad_norm": 0.0645170733332634, "learning_rate": 3.8530413370085786e-05, "loss": 0.0023, "step": 59500 }, { "grad_norm": 0.059934716671705246, "learning_rate": 3.851432021057608e-05, "loss": 0.0028, "step": 59510 }, { "grad_norm": 0.06188683956861496, "learning_rate": 3.849822830712306e-05, "loss": 0.0025, "step": 59520 }, { "grad_norm": 0.06047441437840462, "learning_rate": 3.848213766148649e-05, "loss": 0.0025, "step": 59530 }, { "grad_norm": 0.05819740518927574, "learning_rate": 3.846604827542602e-05, "loss": 0.0032, "step": 59540 }, { "grad_norm": 0.06585265696048737, "learning_rate": 3.844996015070117e-05, "loss": 0.0031, "step": 59550 }, { "grad_norm": 0.07618711143732071, "learning_rate": 3.843387328907132e-05, "loss": 0.0032, "step": 59560 }, { "grad_norm": 0.059164028614759445, "learning_rate": 3.8417787692295696e-05, "loss": 0.005, "step": 59570 }, { "grad_norm": 0.08624501526355743, "learning_rate": 3.84017033621334e-05, "loss": 0.0026, "step": 59580 }, { "grad_norm": 0.06149924546480179, "learning_rate": 3.838562030034338e-05, "loss": 0.0027, "step": 59590 }, { "grad_norm": 0.06801342219114304, "learning_rate": 3.836953850868447e-05, "loss": 0.003, "step": 59600 }, { "grad_norm": 0.07893604785203934, "learning_rate": 3.835345798891533e-05, "loss": 0.0028, "step": 59610 }, { "grad_norm": 0.06020364910364151, "learning_rate": 3.833737874279452e-05, "loss": 0.0031, "step": 59620 }, { "grad_norm": 0.054131340235471725, "learning_rate": 3.8321300772080433e-05, "loss": 0.0029, "step": 59630 }, { "grad_norm": 0.07033047080039978, "learning_rate": 3.8305224078531345e-05, "loss": 0.0024, "step": 59640 }, { "grad_norm": 0.06798569858074188, "learning_rate": 3.828914866390537e-05, "loss": 0.0029, "step": 59650 }, { "grad_norm": 0.05462034046649933, "learning_rate": 3.827307452996048e-05, "loss": 0.0026, "step": 59660 }, { "grad_norm": 0.07081786543130875, "learning_rate": 3.825700167845454e-05, "loss": 0.0031, "step": 59670 }, { "grad_norm": 0.04663966968655586, "learning_rate": 3.8240930111145214e-05, "loss": 0.0024, "step": 59680 }, { "grad_norm": 0.05785128101706505, "learning_rate": 3.8224859829790116e-05, "loss": 0.0034, "step": 59690 }, { "grad_norm": 0.09340523183345795, "learning_rate": 3.820879083614664e-05, "loss": 0.0026, "step": 59700 }, { "grad_norm": 0.06455804407596588, "learning_rate": 3.819272313197208e-05, "loss": 0.0031, "step": 59710 }, { "grad_norm": 0.05785466730594635, "learning_rate": 3.817665671902355e-05, "loss": 0.0024, "step": 59720 }, { "grad_norm": 0.07104582339525223, "learning_rate": 3.816059159905808e-05, "loss": 0.0036, "step": 59730 }, { "grad_norm": 0.08414129912853241, "learning_rate": 3.8144527773832476e-05, "loss": 0.0032, "step": 59740 }, { "grad_norm": 0.04911267012357712, "learning_rate": 3.812846524510352e-05, "loss": 0.0024, "step": 59750 }, { "grad_norm": 0.06540028005838394, "learning_rate": 3.811240401462775e-05, "loss": 0.003, "step": 59760 }, { "grad_norm": 0.09111922979354858, "learning_rate": 3.809634408416162e-05, "loss": 0.0029, "step": 59770 }, { "grad_norm": 0.07280737906694412, "learning_rate": 3.8080285455461394e-05, "loss": 0.0026, "step": 59780 }, { "grad_norm": 0.06384497880935669, "learning_rate": 3.806422813028323e-05, "loss": 0.0029, "step": 59790 }, { "grad_norm": 0.04671746864914894, "learning_rate": 3.8048172110383114e-05, "loss": 0.0021, "step": 59800 }, { "grad_norm": 0.07479143887758255, "learning_rate": 3.8032117397516944e-05, "loss": 0.0035, "step": 59810 }, { "grad_norm": 0.06435389071702957, "learning_rate": 3.801606399344042e-05, "loss": 0.0027, "step": 59820 }, { "grad_norm": 0.06457728892564774, "learning_rate": 3.800001189990913e-05, "loss": 0.0028, "step": 59830 }, { "grad_norm": 0.061806291341781616, "learning_rate": 3.798396111867847e-05, "loss": 0.0037, "step": 59840 }, { "grad_norm": 0.06904549151659012, "learning_rate": 3.796791165150378e-05, "loss": 0.0023, "step": 59850 }, { "grad_norm": 0.061663445085287094, "learning_rate": 3.795186350014018e-05, "loss": 0.0036, "step": 59860 }, { "grad_norm": 0.06710264831781387, "learning_rate": 3.793581666634266e-05, "loss": 0.0037, "step": 59870 }, { "grad_norm": 0.06324759125709534, "learning_rate": 3.791977115186609e-05, "loss": 0.0027, "step": 59880 }, { "grad_norm": 0.07103507965803146, "learning_rate": 3.79037269584652e-05, "loss": 0.0023, "step": 59890 }, { "grad_norm": 0.06789713352918625, "learning_rate": 3.7887684087894545e-05, "loss": 0.0024, "step": 59900 }, { "grad_norm": 0.06157982721924782, "learning_rate": 3.787164254190854e-05, "loss": 0.0026, "step": 59910 }, { "grad_norm": 0.07791022211313248, "learning_rate": 3.7855602322261473e-05, "loss": 0.0026, "step": 59920 }, { "grad_norm": 0.05927331745624542, "learning_rate": 3.783956343070746e-05, "loss": 0.0031, "step": 59930 }, { "grad_norm": 0.0628805011510849, "learning_rate": 3.7823525869000526e-05, "loss": 0.003, "step": 59940 }, { "grad_norm": 0.06882818788290024, "learning_rate": 3.78074896388945e-05, "loss": 0.0026, "step": 59950 }, { "grad_norm": 0.06187071278691292, "learning_rate": 3.7791454742143075e-05, "loss": 0.0032, "step": 59960 }, { "grad_norm": 0.0716506838798523, "learning_rate": 3.7775421180499795e-05, "loss": 0.0029, "step": 59970 }, { "grad_norm": 0.06420423835515976, "learning_rate": 3.7759388955718086e-05, "loss": 0.0029, "step": 59980 }, { "grad_norm": 0.06839970499277115, "learning_rate": 3.774335806955117e-05, "loss": 0.0027, "step": 59990 }, { "grad_norm": 0.07429157942533493, "learning_rate": 3.7727328523752215e-05, "loss": 0.0028, "step": 60000 }, { "grad_norm": 0.05932061746716499, "learning_rate": 3.771130032007415e-05, "loss": 0.0032, "step": 60010 }, { "grad_norm": 0.053683578968048096, "learning_rate": 3.769527346026982e-05, "loss": 0.0022, "step": 60020 }, { "grad_norm": 0.07257682830095291, "learning_rate": 3.767924794609188e-05, "loss": 0.003, "step": 60030 }, { "grad_norm": 0.10665090382099152, "learning_rate": 3.7663223779292854e-05, "loss": 0.0028, "step": 60040 }, { "grad_norm": 0.05953869968652725, "learning_rate": 3.764720096162512e-05, "loss": 0.0032, "step": 60050 }, { "grad_norm": 0.08398914337158203, "learning_rate": 3.763117949484092e-05, "loss": 0.0029, "step": 60060 }, { "grad_norm": 0.06465239822864532, "learning_rate": 3.7615159380692344e-05, "loss": 0.0043, "step": 60070 }, { "grad_norm": 0.06772226840257645, "learning_rate": 3.759914062093132e-05, "loss": 0.0029, "step": 60080 }, { "grad_norm": 0.06908568739891052, "learning_rate": 3.7583123217309615e-05, "loss": 0.0037, "step": 60090 }, { "grad_norm": 0.05654367059469223, "learning_rate": 3.7567107171578904e-05, "loss": 0.0025, "step": 60100 }, { "grad_norm": 0.07058410346508026, "learning_rate": 3.755109248549066e-05, "loss": 0.003, "step": 60110 }, { "grad_norm": 0.06432691961526871, "learning_rate": 3.75350791607962e-05, "loss": 0.0032, "step": 60120 }, { "grad_norm": 0.06279220432043076, "learning_rate": 3.751906719924676e-05, "loss": 0.0025, "step": 60130 }, { "grad_norm": 0.04497166723012924, "learning_rate": 3.750305660259337e-05, "loss": 0.0026, "step": 60140 }, { "grad_norm": 0.061053626239299774, "learning_rate": 3.748704737258693e-05, "loss": 0.0037, "step": 60150 }, { "grad_norm": 0.0742373913526535, "learning_rate": 3.747103951097816e-05, "loss": 0.0035, "step": 60160 }, { "grad_norm": 0.0562576986849308, "learning_rate": 3.745503301951768e-05, "loss": 0.0026, "step": 60170 }, { "grad_norm": 0.048887982964515686, "learning_rate": 3.743902789995592e-05, "loss": 0.0022, "step": 60180 }, { "grad_norm": 0.06789197772741318, "learning_rate": 3.7423024154043183e-05, "loss": 0.0023, "step": 60190 }, { "grad_norm": 0.06698975712060928, "learning_rate": 3.740702178352963e-05, "loss": 0.0026, "step": 60200 }, { "grad_norm": 0.058179695159196854, "learning_rate": 3.739102079016523e-05, "loss": 0.0029, "step": 60210 }, { "grad_norm": 0.04531664773821831, "learning_rate": 3.7375021175699846e-05, "loss": 0.003, "step": 60220 }, { "grad_norm": 0.06891314685344696, "learning_rate": 3.735902294188317e-05, "loss": 0.0036, "step": 60230 }, { "grad_norm": 0.1067381277680397, "learning_rate": 3.734302609046473e-05, "loss": 0.0023, "step": 60240 }, { "grad_norm": 0.056214481592178345, "learning_rate": 3.732703062319391e-05, "loss": 0.003, "step": 60250 }, { "grad_norm": 0.0626508817076683, "learning_rate": 3.731103654181999e-05, "loss": 0.0024, "step": 60260 }, { "grad_norm": 0.04153347760438919, "learning_rate": 3.729504384809203e-05, "loss": 0.0019, "step": 60270 }, { "grad_norm": 0.06842529028654099, "learning_rate": 3.727905254375898e-05, "loss": 0.0028, "step": 60280 }, { "grad_norm": 0.062158867716789246, "learning_rate": 3.7263062630569614e-05, "loss": 0.0029, "step": 60290 }, { "grad_norm": 0.0726267620921135, "learning_rate": 3.724707411027256e-05, "loss": 0.0038, "step": 60300 }, { "grad_norm": 0.07329197227954865, "learning_rate": 3.723108698461631e-05, "loss": 0.0031, "step": 60310 }, { "grad_norm": 0.046974360942840576, "learning_rate": 3.72151012553492e-05, "loss": 0.003, "step": 60320 }, { "grad_norm": 0.04897432029247284, "learning_rate": 3.719911692421939e-05, "loss": 0.0022, "step": 60330 }, { "grad_norm": 0.052276525646448135, "learning_rate": 3.71831339929749e-05, "loss": 0.0024, "step": 60340 }, { "grad_norm": 0.07405878603458405, "learning_rate": 3.716715246336362e-05, "loss": 0.003, "step": 60350 }, { "grad_norm": 0.058376215398311615, "learning_rate": 3.715117233713324e-05, "loss": 0.0032, "step": 60360 }, { "grad_norm": 0.05792124941945076, "learning_rate": 3.713519361603135e-05, "loss": 0.0023, "step": 60370 }, { "grad_norm": 0.08567063510417938, "learning_rate": 3.711921630180532e-05, "loss": 0.0026, "step": 60380 }, { "grad_norm": 0.07500679045915604, "learning_rate": 3.710324039620245e-05, "loss": 0.0027, "step": 60390 }, { "grad_norm": 0.0631362572312355, "learning_rate": 3.708726590096982e-05, "loss": 0.0027, "step": 60400 }, { "grad_norm": 0.06065942347049713, "learning_rate": 3.707129281785437e-05, "loss": 0.0031, "step": 60410 }, { "grad_norm": 0.06601180881261826, "learning_rate": 3.705532114860291e-05, "loss": 0.0055, "step": 60420 }, { "grad_norm": 0.06406260281801224, "learning_rate": 3.703935089496207e-05, "loss": 0.0026, "step": 60430 }, { "grad_norm": 0.06580217182636261, "learning_rate": 3.702338205867829e-05, "loss": 0.0023, "step": 60440 }, { "grad_norm": 0.06226259469985962, "learning_rate": 3.7007414641497976e-05, "loss": 0.0035, "step": 60450 }, { "grad_norm": 0.08074760437011719, "learning_rate": 3.699144864516726e-05, "loss": 0.0028, "step": 60460 }, { "grad_norm": 0.06574268639087677, "learning_rate": 3.6975484071432153e-05, "loss": 0.0037, "step": 60470 }, { "grad_norm": 0.06583687663078308, "learning_rate": 3.695952092203854e-05, "loss": 0.0025, "step": 60480 }, { "grad_norm": 0.07192825525999069, "learning_rate": 3.69435591987321e-05, "loss": 0.0026, "step": 60490 }, { "grad_norm": 0.05789169669151306, "learning_rate": 3.6927598903258374e-05, "loss": 0.0025, "step": 60500 }, { "grad_norm": 0.0674050822854042, "learning_rate": 3.691164003736281e-05, "loss": 0.0033, "step": 60510 }, { "grad_norm": 0.05130438879132271, "learning_rate": 3.689568260279061e-05, "loss": 0.0023, "step": 60520 }, { "grad_norm": 0.04103221371769905, "learning_rate": 3.687972660128686e-05, "loss": 0.0026, "step": 60530 }, { "grad_norm": 0.0777270495891571, "learning_rate": 3.686377203459648e-05, "loss": 0.0034, "step": 60540 }, { "grad_norm": 0.04939550533890724, "learning_rate": 3.6847818904464226e-05, "loss": 0.0022, "step": 60550 }, { "grad_norm": 0.05491361394524574, "learning_rate": 3.683186721263474e-05, "loss": 0.0023, "step": 60560 }, { "grad_norm": 0.06933973729610443, "learning_rate": 3.6815916960852435e-05, "loss": 0.0027, "step": 60570 }, { "grad_norm": 0.05782342702150345, "learning_rate": 3.679996815086165e-05, "loss": 0.0029, "step": 60580 }, { "grad_norm": 0.09309221804141998, "learning_rate": 3.678402078440649e-05, "loss": 0.0025, "step": 60590 }, { "grad_norm": 0.06247403472661972, "learning_rate": 3.676807486323096e-05, "loss": 0.0033, "step": 60600 }, { "grad_norm": 0.04932713136076927, "learning_rate": 3.6752130389078864e-05, "loss": 0.0028, "step": 60610 }, { "grad_norm": 0.06850656121969223, "learning_rate": 3.673618736369388e-05, "loss": 0.0032, "step": 60620 }, { "grad_norm": 0.06058381497859955, "learning_rate": 3.6720245788819474e-05, "loss": 0.0023, "step": 60630 }, { "grad_norm": 0.09097915142774582, "learning_rate": 3.6704305666199044e-05, "loss": 0.0024, "step": 60640 }, { "grad_norm": 0.048255305737257004, "learning_rate": 3.668836699757576e-05, "loss": 0.0035, "step": 60650 }, { "grad_norm": 0.05024537071585655, "learning_rate": 3.667242978469265e-05, "loss": 0.0029, "step": 60660 }, { "grad_norm": 0.085719995200634, "learning_rate": 3.6656494029292575e-05, "loss": 0.0034, "step": 60670 }, { "grad_norm": 0.06346305459737778, "learning_rate": 3.664055973311825e-05, "loss": 0.0034, "step": 60680 }, { "grad_norm": 0.055215492844581604, "learning_rate": 3.662462689791221e-05, "loss": 0.0024, "step": 60690 }, { "grad_norm": 0.06537245959043503, "learning_rate": 3.660869552541689e-05, "loss": 0.003, "step": 60700 }, { "grad_norm": 0.056704722344875336, "learning_rate": 3.659276561737448e-05, "loss": 0.0041, "step": 60710 }, { "grad_norm": 0.06770981103181839, "learning_rate": 3.657683717552708e-05, "loss": 0.0033, "step": 60720 }, { "grad_norm": 0.06586288660764694, "learning_rate": 3.6560910201616574e-05, "loss": 0.003, "step": 60730 }, { "grad_norm": 0.07900906354188919, "learning_rate": 3.654498469738472e-05, "loss": 0.0024, "step": 60740 }, { "grad_norm": 0.04356486722826958, "learning_rate": 3.6529060664573126e-05, "loss": 0.0028, "step": 60750 }, { "grad_norm": 0.05263645574450493, "learning_rate": 3.6513138104923176e-05, "loss": 0.0031, "step": 60760 }, { "grad_norm": 0.06350512057542801, "learning_rate": 3.6497217020176176e-05, "loss": 0.0031, "step": 60770 }, { "grad_norm": 0.06535197794437408, "learning_rate": 3.648129741207323e-05, "loss": 0.0034, "step": 60780 }, { "grad_norm": 0.06922885030508041, "learning_rate": 3.646537928235527e-05, "loss": 0.0039, "step": 60790 }, { "grad_norm": 0.07905121892690659, "learning_rate": 3.6449462632763075e-05, "loss": 0.0027, "step": 60800 }, { "grad_norm": 0.06500363349914551, "learning_rate": 3.643354746503729e-05, "loss": 0.0029, "step": 60810 }, { "grad_norm": 0.07421945780515671, "learning_rate": 3.6417633780918345e-05, "loss": 0.0031, "step": 60820 }, { "grad_norm": 0.05844279006123543, "learning_rate": 3.640172158214655e-05, "loss": 0.0024, "step": 60830 }, { "grad_norm": 0.07230200618505478, "learning_rate": 3.6385810870462035e-05, "loss": 0.0026, "step": 60840 }, { "grad_norm": 0.045148223638534546, "learning_rate": 3.636990164760479e-05, "loss": 0.0022, "step": 60850 }, { "grad_norm": 0.03665407374501228, "learning_rate": 3.6353993915314596e-05, "loss": 0.0023, "step": 60860 }, { "grad_norm": 0.07236554473638535, "learning_rate": 3.6338087675331124e-05, "loss": 0.0026, "step": 60870 }, { "grad_norm": 0.061996523290872574, "learning_rate": 3.632218292939383e-05, "loss": 0.0024, "step": 60880 }, { "grad_norm": 0.06845278292894363, "learning_rate": 3.6306279679242033e-05, "loss": 0.0023, "step": 60890 }, { "grad_norm": 0.04932113736867905, "learning_rate": 3.629037792661492e-05, "loss": 0.0025, "step": 60900 }, { "grad_norm": 0.05836385115981102, "learning_rate": 3.627447767325147e-05, "loss": 0.0029, "step": 60910 }, { "grad_norm": 0.05501779913902283, "learning_rate": 3.62585789208905e-05, "loss": 0.0031, "step": 60920 }, { "grad_norm": 0.057774074375629425, "learning_rate": 3.624268167127068e-05, "loss": 0.0025, "step": 60930 }, { "grad_norm": 0.09213389456272125, "learning_rate": 3.622678592613051e-05, "loss": 0.0032, "step": 60940 }, { "grad_norm": 0.06652417033910751, "learning_rate": 3.6210891687208304e-05, "loss": 0.0026, "step": 60950 }, { "grad_norm": 0.06499955803155899, "learning_rate": 3.619499895624228e-05, "loss": 0.0028, "step": 60960 }, { "grad_norm": 0.0697583481669426, "learning_rate": 3.61791077349704e-05, "loss": 0.0027, "step": 60970 }, { "grad_norm": 0.08403593301773071, "learning_rate": 3.616321802513053e-05, "loss": 0.0032, "step": 60980 }, { "grad_norm": 0.06482376158237457, "learning_rate": 3.6147329828460325e-05, "loss": 0.0036, "step": 60990 }, { "grad_norm": 0.06562898308038712, "learning_rate": 3.6131443146697306e-05, "loss": 0.0029, "step": 61000 }, { "grad_norm": 0.06019136682152748, "learning_rate": 3.611555798157879e-05, "loss": 0.0027, "step": 61010 }, { "grad_norm": 0.08127125352621078, "learning_rate": 3.6099674334842e-05, "loss": 0.0022, "step": 61020 }, { "grad_norm": 0.047182075679302216, "learning_rate": 3.6083792208223934e-05, "loss": 0.0024, "step": 61030 }, { "grad_norm": 0.04736270755529404, "learning_rate": 3.606791160346142e-05, "loss": 0.0027, "step": 61040 }, { "grad_norm": 0.04601503163576126, "learning_rate": 3.6052032522291134e-05, "loss": 0.0027, "step": 61050 }, { "grad_norm": 0.06473129242658615, "learning_rate": 3.603615496644962e-05, "loss": 0.0027, "step": 61060 }, { "grad_norm": 0.06808187812566757, "learning_rate": 3.60202789376732e-05, "loss": 0.0028, "step": 61070 }, { "grad_norm": 0.07960399240255356, "learning_rate": 3.6004404437698036e-05, "loss": 0.0029, "step": 61080 }, { "grad_norm": 0.06727301329374313, "learning_rate": 3.5988531468260184e-05, "loss": 0.003, "step": 61090 }, { "grad_norm": 0.05485169589519501, "learning_rate": 3.597266003109546e-05, "loss": 0.0021, "step": 61100 }, { "grad_norm": 0.07059650123119354, "learning_rate": 3.5956790127939557e-05, "loss": 0.0029, "step": 61110 }, { "grad_norm": 0.07879268378019333, "learning_rate": 3.594092176052796e-05, "loss": 0.0031, "step": 61120 }, { "grad_norm": 0.06852957606315613, "learning_rate": 3.592505493059603e-05, "loss": 0.003, "step": 61130 }, { "grad_norm": 0.0654948428273201, "learning_rate": 3.5909189639878905e-05, "loss": 0.0026, "step": 61140 }, { "grad_norm": 0.06479013711214066, "learning_rate": 3.5893325890111626e-05, "loss": 0.0023, "step": 61150 }, { "grad_norm": 0.04894367977976799, "learning_rate": 3.5877463683029034e-05, "loss": 0.0025, "step": 61160 }, { "grad_norm": 0.06701627373695374, "learning_rate": 3.586160302036578e-05, "loss": 0.003, "step": 61170 }, { "grad_norm": 0.04606137424707413, "learning_rate": 3.584574390385636e-05, "loss": 0.0031, "step": 61180 }, { "grad_norm": 0.068794846534729, "learning_rate": 3.582988633523511e-05, "loss": 0.0027, "step": 61190 }, { "grad_norm": 0.10908526182174683, "learning_rate": 3.581403031623616e-05, "loss": 0.0029, "step": 61200 }, { "grad_norm": 0.0486188679933548, "learning_rate": 3.579817584859354e-05, "loss": 0.003, "step": 61210 }, { "grad_norm": 0.046805936843156815, "learning_rate": 3.578232293404107e-05, "loss": 0.0027, "step": 61220 }, { "grad_norm": 0.06750554591417313, "learning_rate": 3.5766471574312385e-05, "loss": 0.0028, "step": 61230 }, { "grad_norm": 0.061017319560050964, "learning_rate": 3.5750621771140964e-05, "loss": 0.0024, "step": 61240 }, { "grad_norm": 0.051795586943626404, "learning_rate": 3.5734773526260115e-05, "loss": 0.0031, "step": 61250 }, { "grad_norm": 0.08801701664924622, "learning_rate": 3.5718926841402996e-05, "loss": 0.0025, "step": 61260 }, { "grad_norm": 0.05742872506380081, "learning_rate": 3.5703081718302555e-05, "loss": 0.0029, "step": 61270 }, { "grad_norm": 0.05691434442996979, "learning_rate": 3.568723815869162e-05, "loss": 0.0029, "step": 61280 }, { "grad_norm": 0.07291585206985474, "learning_rate": 3.5671396164302777e-05, "loss": 0.0026, "step": 61290 }, { "grad_norm": 0.06683072447776794, "learning_rate": 3.565555573686853e-05, "loss": 0.0025, "step": 61300 }, { "grad_norm": 0.06187232956290245, "learning_rate": 3.563971687812113e-05, "loss": 0.0039, "step": 61310 }, { "grad_norm": 0.055409956723451614, "learning_rate": 3.5623879589792706e-05, "loss": 0.0024, "step": 61320 }, { "grad_norm": 0.07287828624248505, "learning_rate": 3.5608043873615185e-05, "loss": 0.0036, "step": 61330 }, { "grad_norm": 0.05955563485622406, "learning_rate": 3.559220973132035e-05, "loss": 0.0025, "step": 61340 }, { "grad_norm": 0.059489380568265915, "learning_rate": 3.557637716463981e-05, "loss": 0.0032, "step": 61350 }, { "grad_norm": 0.062475886195898056, "learning_rate": 3.556054617530497e-05, "loss": 0.003, "step": 61360 }, { "grad_norm": 0.06575299799442291, "learning_rate": 3.5544716765047106e-05, "loss": 0.0027, "step": 61370 }, { "grad_norm": 0.08411896228790283, "learning_rate": 3.5528888935597274e-05, "loss": 0.003, "step": 61380 }, { "grad_norm": 0.0590788759291172, "learning_rate": 3.551306268868636e-05, "loss": 0.0023, "step": 61390 }, { "grad_norm": 0.05953570082783699, "learning_rate": 3.5497238026045166e-05, "loss": 0.0028, "step": 61400 }, { "grad_norm": 0.06419955939054489, "learning_rate": 3.548141494940421e-05, "loss": 0.003, "step": 61410 }, { "grad_norm": 0.06103815883398056, "learning_rate": 3.546559346049388e-05, "loss": 0.0029, "step": 61420 }, { "grad_norm": 0.07278071343898773, "learning_rate": 3.544977356104441e-05, "loss": 0.0026, "step": 61430 }, { "grad_norm": 0.04202991724014282, "learning_rate": 3.543395525278582e-05, "loss": 0.0027, "step": 61440 }, { "grad_norm": 0.038012079894542694, "learning_rate": 3.5418138537447985e-05, "loss": 0.002, "step": 61450 }, { "grad_norm": 0.06885070353746414, "learning_rate": 3.5402323416760574e-05, "loss": 0.0036, "step": 61460 }, { "grad_norm": 0.05608708783984184, "learning_rate": 3.538650989245316e-05, "loss": 0.0026, "step": 61470 }, { "grad_norm": 0.07885339856147766, "learning_rate": 3.537069796625504e-05, "loss": 0.0039, "step": 61480 }, { "grad_norm": 0.0772753432393074, "learning_rate": 3.53548876398954e-05, "loss": 0.003, "step": 61490 }, { "grad_norm": 0.08204096555709839, "learning_rate": 3.5339078915103214e-05, "loss": 0.0026, "step": 61500 }, { "grad_norm": 0.08302842080593109, "learning_rate": 3.532327179360733e-05, "loss": 0.0041, "step": 61510 }, { "grad_norm": 0.06300140917301178, "learning_rate": 3.530746627713636e-05, "loss": 0.0028, "step": 61520 }, { "grad_norm": 0.049282096326351166, "learning_rate": 3.5291662367418805e-05, "loss": 0.0023, "step": 61530 }, { "grad_norm": 0.05756557360291481, "learning_rate": 3.527586006618293e-05, "loss": 0.0034, "step": 61540 }, { "grad_norm": 0.07170800119638443, "learning_rate": 3.526005937515687e-05, "loss": 0.0028, "step": 61550 }, { "grad_norm": 0.06108126416802406, "learning_rate": 3.524426029606856e-05, "loss": 0.0034, "step": 61560 }, { "grad_norm": 0.057059597223997116, "learning_rate": 3.5228462830645746e-05, "loss": 0.0035, "step": 61570 }, { "grad_norm": 0.09498163312673569, "learning_rate": 3.5212666980616036e-05, "loss": 0.0034, "step": 61580 }, { "grad_norm": 0.08565530180931091, "learning_rate": 3.519687274770681e-05, "loss": 0.0023, "step": 61590 }, { "grad_norm": 0.06916496157646179, "learning_rate": 3.518108013364535e-05, "loss": 0.0025, "step": 61600 }, { "grad_norm": 0.09328217804431915, "learning_rate": 3.516528914015868e-05, "loss": 0.0029, "step": 61610 }, { "grad_norm": 0.07850050181150436, "learning_rate": 3.5149499768973694e-05, "loss": 0.0026, "step": 61620 }, { "grad_norm": 0.058690477162599564, "learning_rate": 3.513371202181708e-05, "loss": 0.0023, "step": 61630 }, { "grad_norm": 0.04957148805260658, "learning_rate": 3.511792590041537e-05, "loss": 0.0029, "step": 61640 }, { "grad_norm": 0.04779615253210068, "learning_rate": 3.510214140649488e-05, "loss": 0.0022, "step": 61650 }, { "grad_norm": 0.055061694234609604, "learning_rate": 3.508635854178184e-05, "loss": 0.0025, "step": 61660 }, { "grad_norm": 0.04784173518419266, "learning_rate": 3.50705773080022e-05, "loss": 0.0035, "step": 61670 }, { "grad_norm": 0.07117410749197006, "learning_rate": 3.505479770688178e-05, "loss": 0.0027, "step": 61680 }, { "grad_norm": 0.07428506016731262, "learning_rate": 3.503901974014622e-05, "loss": 0.003, "step": 61690 }, { "grad_norm": 0.06020336225628853, "learning_rate": 3.502324340952095e-05, "loss": 0.0031, "step": 61700 }, { "grad_norm": 0.0658116489648819, "learning_rate": 3.500746871673126e-05, "loss": 0.0034, "step": 61710 }, { "grad_norm": 0.05961638689041138, "learning_rate": 3.499169566350227e-05, "loss": 0.0026, "step": 61720 }, { "grad_norm": 0.08942313492298126, "learning_rate": 3.4975924251558876e-05, "loss": 0.0029, "step": 61730 }, { "grad_norm": 0.059693802148103714, "learning_rate": 3.496015448262582e-05, "loss": 0.0024, "step": 61740 }, { "grad_norm": 0.038471367210149765, "learning_rate": 3.494438635842764e-05, "loss": 0.0019, "step": 61750 }, { "grad_norm": 0.06588133424520493, "learning_rate": 3.492861988068874e-05, "loss": 0.0032, "step": 61760 }, { "grad_norm": 0.03870581462979317, "learning_rate": 3.491285505113332e-05, "loss": 0.0023, "step": 61770 }, { "grad_norm": 0.055379267781972885, "learning_rate": 3.4897091871485376e-05, "loss": 0.0026, "step": 61780 }, { "grad_norm": 0.07831909507513046, "learning_rate": 3.488133034346875e-05, "loss": 0.0025, "step": 61790 }, { "grad_norm": 0.058316413313150406, "learning_rate": 3.486557046880713e-05, "loss": 0.0026, "step": 61800 }, { "grad_norm": 0.05969758331775665, "learning_rate": 3.484981224922395e-05, "loss": 0.0032, "step": 61810 }, { "grad_norm": 0.06287802010774612, "learning_rate": 3.4834055686442544e-05, "loss": 0.0032, "step": 61820 }, { "grad_norm": 0.07141484320163727, "learning_rate": 3.4818300782186e-05, "loss": 0.0031, "step": 61830 }, { "grad_norm": 0.07814483344554901, "learning_rate": 3.4802547538177236e-05, "loss": 0.0024, "step": 61840 }, { "grad_norm": 0.07189463078975677, "learning_rate": 3.478679595613905e-05, "loss": 0.0037, "step": 61850 }, { "grad_norm": 0.059436310082674026, "learning_rate": 3.4771046037793975e-05, "loss": 0.0025, "step": 61860 }, { "grad_norm": 0.06778975576162338, "learning_rate": 3.4755297784864416e-05, "loss": 0.0028, "step": 61870 }, { "grad_norm": 0.05674654245376587, "learning_rate": 3.473955119907257e-05, "loss": 0.0032, "step": 61880 }, { "grad_norm": 0.06269749999046326, "learning_rate": 3.472380628214046e-05, "loss": 0.0031, "step": 61890 }, { "grad_norm": 0.08037155866622925, "learning_rate": 3.4708063035789906e-05, "loss": 0.0027, "step": 61900 }, { "grad_norm": 0.04611935466527939, "learning_rate": 3.4692321461742614e-05, "loss": 0.0027, "step": 61910 }, { "grad_norm": 0.05928722769021988, "learning_rate": 3.4676581561720023e-05, "loss": 0.004, "step": 61920 }, { "grad_norm": 0.10911307483911514, "learning_rate": 3.466084333744344e-05, "loss": 0.003, "step": 61930 }, { "grad_norm": 0.05526861175894737, "learning_rate": 3.4645106790633964e-05, "loss": 0.0033, "step": 61940 }, { "grad_norm": 0.06370711326599121, "learning_rate": 3.4629371923012526e-05, "loss": 0.0034, "step": 61950 }, { "grad_norm": 0.04680819436907768, "learning_rate": 3.4613638736299856e-05, "loss": 0.003, "step": 61960 }, { "grad_norm": 0.06699759513139725, "learning_rate": 3.459790723221652e-05, "loss": 0.003, "step": 61970 }, { "grad_norm": 0.06371438503265381, "learning_rate": 3.45821774124829e-05, "loss": 0.0032, "step": 61980 }, { "grad_norm": 0.0563972145318985, "learning_rate": 3.456644927881918e-05, "loss": 0.0023, "step": 61990 }, { "grad_norm": 0.07267359644174576, "learning_rate": 3.4550722832945354e-05, "loss": 0.0032, "step": 62000 }, { "grad_norm": 0.05760161206126213, "learning_rate": 3.453499807658126e-05, "loss": 0.0031, "step": 62010 }, { "grad_norm": 0.056310150772333145, "learning_rate": 3.451927501144653e-05, "loss": 0.003, "step": 62020 }, { "grad_norm": 0.05307890474796295, "learning_rate": 3.450355363926059e-05, "loss": 0.0028, "step": 62030 }, { "grad_norm": 0.056326042860746384, "learning_rate": 3.448783396174273e-05, "loss": 0.0033, "step": 62040 }, { "grad_norm": 0.06373751163482666, "learning_rate": 3.4472115980612034e-05, "loss": 0.0029, "step": 62050 }, { "grad_norm": 0.060213129967451096, "learning_rate": 3.4456399697587396e-05, "loss": 0.0031, "step": 62060 }, { "grad_norm": 0.0546426996588707, "learning_rate": 3.4440685114387506e-05, "loss": 0.0037, "step": 62070 }, { "grad_norm": 0.05756569653749466, "learning_rate": 3.442497223273091e-05, "loss": 0.0029, "step": 62080 }, { "grad_norm": 0.058665841817855835, "learning_rate": 3.4409261054335926e-05, "loss": 0.0025, "step": 62090 }, { "grad_norm": 0.05650266632437706, "learning_rate": 3.439355158092069e-05, "loss": 0.0028, "step": 62100 }, { "grad_norm": 0.08932528644800186, "learning_rate": 3.437784381420322e-05, "loss": 0.0033, "step": 62110 }, { "grad_norm": 0.05936693772673607, "learning_rate": 3.436213775590126e-05, "loss": 0.0024, "step": 62120 }, { "grad_norm": 0.06796572357416153, "learning_rate": 3.4346433407732405e-05, "loss": 0.0032, "step": 62130 }, { "grad_norm": 0.04775839298963547, "learning_rate": 3.433073077141405e-05, "loss": 0.0034, "step": 62140 }, { "grad_norm": 0.057533398270606995, "learning_rate": 3.431502984866343e-05, "loss": 0.0026, "step": 62150 }, { "grad_norm": 0.051819637417793274, "learning_rate": 3.429933064119753e-05, "loss": 0.0032, "step": 62160 }, { "grad_norm": 0.05158036947250366, "learning_rate": 3.4283633150733255e-05, "loss": 0.0023, "step": 62170 }, { "grad_norm": 0.04971320927143097, "learning_rate": 3.4267937378987225e-05, "loss": 0.0023, "step": 62180 }, { "grad_norm": 0.058757711201906204, "learning_rate": 3.425224332767592e-05, "loss": 0.0024, "step": 62190 }, { "grad_norm": 0.06666769087314606, "learning_rate": 3.4236550998515595e-05, "loss": 0.0027, "step": 62200 }, { "grad_norm": 0.0958896279335022, "learning_rate": 3.422086039322235e-05, "loss": 0.0025, "step": 62210 }, { "grad_norm": 0.07929050177335739, "learning_rate": 3.420517151351209e-05, "loss": 0.0026, "step": 62220 }, { "grad_norm": 0.07829775661230087, "learning_rate": 3.418948436110053e-05, "loss": 0.0026, "step": 62230 }, { "grad_norm": 0.06883122026920319, "learning_rate": 3.417379893770319e-05, "loss": 0.0028, "step": 62240 }, { "grad_norm": 0.05308709666132927, "learning_rate": 3.41581152450354e-05, "loss": 0.0033, "step": 62250 }, { "grad_norm": 0.057875439524650574, "learning_rate": 3.414243328481231e-05, "loss": 0.0026, "step": 62260 }, { "grad_norm": 0.055482249706983566, "learning_rate": 3.412675305874888e-05, "loss": 0.0029, "step": 62270 }, { "grad_norm": 0.040595415979623795, "learning_rate": 3.4111074568559876e-05, "loss": 0.0031, "step": 62280 }, { "grad_norm": 0.06319733709096909, "learning_rate": 3.4095397815959846e-05, "loss": 0.0028, "step": 62290 }, { "grad_norm": 0.06984692066907883, "learning_rate": 3.407972280266322e-05, "loss": 0.0024, "step": 62300 }, { "grad_norm": 0.05566209554672241, "learning_rate": 3.4064049530384166e-05, "loss": 0.003, "step": 62310 }, { "grad_norm": 0.07253218442201614, "learning_rate": 3.40483780008367e-05, "loss": 0.0026, "step": 62320 }, { "grad_norm": 0.0831046774983406, "learning_rate": 3.4032708215734635e-05, "loss": 0.0034, "step": 62330 }, { "grad_norm": 0.0634043738245964, "learning_rate": 3.4017040176791584e-05, "loss": 0.0022, "step": 62340 }, { "grad_norm": 0.050769075751304626, "learning_rate": 3.4001373885720976e-05, "loss": 0.0042, "step": 62350 }, { "grad_norm": 0.04208439961075783, "learning_rate": 3.398570934423608e-05, "loss": 0.0021, "step": 62360 }, { "grad_norm": 0.07715088874101639, "learning_rate": 3.397004655404994e-05, "loss": 0.0027, "step": 62370 }, { "grad_norm": 0.0682484358549118, "learning_rate": 3.395438551687541e-05, "loss": 0.0029, "step": 62380 }, { "grad_norm": 0.0948072150349617, "learning_rate": 3.393872623442515e-05, "loss": 0.003, "step": 62390 }, { "grad_norm": 0.08039486408233643, "learning_rate": 3.392306870841164e-05, "loss": 0.0028, "step": 62400 }, { "grad_norm": 0.061738062649965286, "learning_rate": 3.390741294054715e-05, "loss": 0.0022, "step": 62410 }, { "grad_norm": 0.08411921560764313, "learning_rate": 3.38917589325438e-05, "loss": 0.0032, "step": 62420 }, { "grad_norm": 0.06391341984272003, "learning_rate": 3.387610668611349e-05, "loss": 0.0027, "step": 62430 }, { "grad_norm": 0.06659269332885742, "learning_rate": 3.38604562029679e-05, "loss": 0.0021, "step": 62440 }, { "grad_norm": 0.06922896951436996, "learning_rate": 3.384480748481855e-05, "loss": 0.0026, "step": 62450 }, { "grad_norm": 0.06990694254636765, "learning_rate": 3.382916053337676e-05, "loss": 0.0025, "step": 62460 }, { "grad_norm": 0.05860087648034096, "learning_rate": 3.381351535035366e-05, "loss": 0.0028, "step": 62470 }, { "grad_norm": 0.04511444270610809, "learning_rate": 3.379787193746019e-05, "loss": 0.0027, "step": 62480 }, { "grad_norm": 0.07661283761262894, "learning_rate": 3.378223029640709e-05, "loss": 0.0026, "step": 62490 }, { "grad_norm": 0.05961618199944496, "learning_rate": 3.376659042890489e-05, "loss": 0.0025, "step": 62500 }, { "grad_norm": 0.08485548198223114, "learning_rate": 3.375095233666397e-05, "loss": 0.0024, "step": 62510 }, { "grad_norm": 0.04768521711230278, "learning_rate": 3.373531602139447e-05, "loss": 0.0022, "step": 62520 }, { "grad_norm": 0.04080211743712425, "learning_rate": 3.371968148480635e-05, "loss": 0.0022, "step": 62530 }, { "grad_norm": 0.06338028609752655, "learning_rate": 3.370404872860937e-05, "loss": 0.0025, "step": 62540 }, { "grad_norm": 0.06857523322105408, "learning_rate": 3.3688417754513145e-05, "loss": 0.003, "step": 62550 }, { "grad_norm": 0.04690919443964958, "learning_rate": 3.367278856422703e-05, "loss": 0.0025, "step": 62560 }, { "grad_norm": 0.06083961948752403, "learning_rate": 3.365716115946021e-05, "loss": 0.0022, "step": 62570 }, { "grad_norm": 0.05077467858791351, "learning_rate": 3.3641535541921666e-05, "loss": 0.0021, "step": 62580 }, { "grad_norm": 0.055088553577661514, "learning_rate": 3.36259117133202e-05, "loss": 0.0023, "step": 62590 }, { "grad_norm": 0.047382939606904984, "learning_rate": 3.36102896753644e-05, "loss": 0.0029, "step": 62600 }, { "grad_norm": 0.06309669464826584, "learning_rate": 3.35946694297627e-05, "loss": 0.0026, "step": 62610 }, { "grad_norm": 0.07421211898326874, "learning_rate": 3.357905097822328e-05, "loss": 0.0031, "step": 62620 }, { "grad_norm": 0.07058247178792953, "learning_rate": 3.356343432245416e-05, "loss": 0.0024, "step": 62630 }, { "grad_norm": 0.06363221257925034, "learning_rate": 3.354781946416315e-05, "loss": 0.0023, "step": 62640 }, { "grad_norm": 0.06010299175977707, "learning_rate": 3.3532206405057864e-05, "loss": 0.0023, "step": 62650 }, { "grad_norm": 0.08022269606590271, "learning_rate": 3.351659514684573e-05, "loss": 0.0025, "step": 62660 }, { "grad_norm": 0.055719953030347824, "learning_rate": 3.350098569123395e-05, "loss": 0.003, "step": 62670 }, { "grad_norm": 0.06023884192109108, "learning_rate": 3.348537803992958e-05, "loss": 0.0025, "step": 62680 }, { "grad_norm": 0.074802465736866, "learning_rate": 3.346977219463945e-05, "loss": 0.0023, "step": 62690 }, { "grad_norm": 0.05325459688901901, "learning_rate": 3.345416815707017e-05, "loss": 0.0028, "step": 62700 }, { "grad_norm": 0.06026231497526169, "learning_rate": 3.3438565928928185e-05, "loss": 0.0024, "step": 62710 }, { "grad_norm": 0.06307011097669601, "learning_rate": 3.3422965511919736e-05, "loss": 0.0025, "step": 62720 }, { "grad_norm": 0.057777758687734604, "learning_rate": 3.3407366907750846e-05, "loss": 0.0024, "step": 62730 }, { "grad_norm": 0.06303852051496506, "learning_rate": 3.339177011812737e-05, "loss": 0.0026, "step": 62740 }, { "grad_norm": 0.05085808038711548, "learning_rate": 3.3376175144754935e-05, "loss": 0.0024, "step": 62750 }, { "grad_norm": 0.07408450543880463, "learning_rate": 3.3360581989339004e-05, "loss": 0.0026, "step": 62760 }, { "grad_norm": 0.07377558946609497, "learning_rate": 3.3344990653584804e-05, "loss": 0.0034, "step": 62770 }, { "grad_norm": 0.06866656988859177, "learning_rate": 3.332940113919739e-05, "loss": 0.0031, "step": 62780 }, { "grad_norm": 0.06839365512132645, "learning_rate": 3.331381344788161e-05, "loss": 0.0035, "step": 62790 }, { "grad_norm": 0.06947232037782669, "learning_rate": 3.3298227581342076e-05, "loss": 0.0022, "step": 62800 }, { "grad_norm": 0.07522723823785782, "learning_rate": 3.328264354128329e-05, "loss": 0.0026, "step": 62810 }, { "grad_norm": 0.05825721099972725, "learning_rate": 3.3267061329409465e-05, "loss": 0.0023, "step": 62820 }, { "grad_norm": 0.05416441336274147, "learning_rate": 3.325148094742466e-05, "loss": 0.0028, "step": 62830 }, { "grad_norm": 0.051908764988183975, "learning_rate": 3.323590239703271e-05, "loss": 0.0024, "step": 62840 }, { "grad_norm": 0.08619175851345062, "learning_rate": 3.322032567993728e-05, "loss": 0.0024, "step": 62850 }, { "grad_norm": 0.07285597920417786, "learning_rate": 3.3204750797841774e-05, "loss": 0.0031, "step": 62860 }, { "grad_norm": 0.05035511776804924, "learning_rate": 3.31891777524495e-05, "loss": 0.0023, "step": 62870 }, { "grad_norm": 0.052598610520362854, "learning_rate": 3.317360654546347e-05, "loss": 0.0025, "step": 62880 }, { "grad_norm": 0.07448230683803558, "learning_rate": 3.3158037178586534e-05, "loss": 0.0024, "step": 62890 }, { "grad_norm": 0.04765021800994873, "learning_rate": 3.3142469653521316e-05, "loss": 0.0025, "step": 62900 }, { "grad_norm": 0.05265149101614952, "learning_rate": 3.3126903971970284e-05, "loss": 0.0023, "step": 62910 }, { "grad_norm": 0.09823568910360336, "learning_rate": 3.311134013563564e-05, "loss": 0.0022, "step": 62920 }, { "grad_norm": 0.0637914165854454, "learning_rate": 3.309577814621947e-05, "loss": 0.0044, "step": 62930 }, { "grad_norm": 0.050768639892339706, "learning_rate": 3.308021800542359e-05, "loss": 0.0026, "step": 62940 }, { "grad_norm": 0.06547202169895172, "learning_rate": 3.306465971494963e-05, "loss": 0.0025, "step": 62950 }, { "grad_norm": 0.05973408743739128, "learning_rate": 3.304910327649901e-05, "loss": 0.0029, "step": 62960 }, { "grad_norm": 0.05996359884738922, "learning_rate": 3.303354869177297e-05, "loss": 0.0021, "step": 62970 }, { "grad_norm": 0.06436368077993393, "learning_rate": 3.301799596247255e-05, "loss": 0.0025, "step": 62980 }, { "grad_norm": 0.0535057932138443, "learning_rate": 3.300244509029854e-05, "loss": 0.0029, "step": 62990 }, { "grad_norm": 0.06446542590856552, "learning_rate": 3.2986896076951585e-05, "loss": 0.0033, "step": 63000 }, { "grad_norm": 0.06753893196582794, "learning_rate": 3.297134892413211e-05, "loss": 0.0029, "step": 63010 }, { "grad_norm": 0.06885191798210144, "learning_rate": 3.295580363354031e-05, "loss": 0.0026, "step": 63020 }, { "grad_norm": 0.06288942694664001, "learning_rate": 3.2940260206876194e-05, "loss": 0.0026, "step": 63030 }, { "grad_norm": 0.05496155843138695, "learning_rate": 3.292471864583958e-05, "loss": 0.0023, "step": 63040 }, { "grad_norm": 0.04344223067164421, "learning_rate": 3.290917895213004e-05, "loss": 0.0028, "step": 63050 }, { "grad_norm": 0.051843419671058655, "learning_rate": 3.289364112744701e-05, "loss": 0.0026, "step": 63060 }, { "grad_norm": 0.050235114991664886, "learning_rate": 3.287810517348967e-05, "loss": 0.0025, "step": 63070 }, { "grad_norm": 0.06437966972589493, "learning_rate": 3.2862571091957014e-05, "loss": 0.0026, "step": 63080 }, { "grad_norm": 0.06561190634965897, "learning_rate": 3.284703888454781e-05, "loss": 0.002, "step": 63090 }, { "grad_norm": 0.08078060299158096, "learning_rate": 3.2831508552960646e-05, "loss": 0.0034, "step": 63100 }, { "grad_norm": 0.05638226866722107, "learning_rate": 3.281598009889387e-05, "loss": 0.0027, "step": 63110 }, { "grad_norm": 0.07786911725997925, "learning_rate": 3.28004535240457e-05, "loss": 0.0025, "step": 63120 }, { "grad_norm": 0.06067291274666786, "learning_rate": 3.278492883011407e-05, "loss": 0.0023, "step": 63130 }, { "grad_norm": 0.05528278276324272, "learning_rate": 3.276940601879675e-05, "loss": 0.0034, "step": 63140 }, { "grad_norm": 0.05822327733039856, "learning_rate": 3.275388509179127e-05, "loss": 0.0026, "step": 63150 }, { "grad_norm": 0.06571176648139954, "learning_rate": 3.2738366050794985e-05, "loss": 0.0025, "step": 63160 }, { "grad_norm": 0.04650061950087547, "learning_rate": 3.2722848897505034e-05, "loss": 0.0025, "step": 63170 }, { "grad_norm": 0.05959297716617584, "learning_rate": 3.2707333633618356e-05, "loss": 0.0023, "step": 63180 }, { "grad_norm": 0.06807810068130493, "learning_rate": 3.269182026083167e-05, "loss": 0.0035, "step": 63190 }, { "grad_norm": 0.05914948135614395, "learning_rate": 3.2676308780841516e-05, "loss": 0.003, "step": 63200 }, { "grad_norm": 0.056059420108795166, "learning_rate": 3.266079919534416e-05, "loss": 0.0028, "step": 63210 }, { "grad_norm": 0.056043654680252075, "learning_rate": 3.264529150603575e-05, "loss": 0.0022, "step": 63220 }, { "grad_norm": 0.07084524631500244, "learning_rate": 3.262978571461217e-05, "loss": 0.0031, "step": 63230 }, { "grad_norm": 0.0885779857635498, "learning_rate": 3.26142818227691e-05, "loss": 0.0032, "step": 63240 }, { "grad_norm": 0.06426840275526047, "learning_rate": 3.259877983220203e-05, "loss": 0.0028, "step": 63250 }, { "grad_norm": 0.04936695471405983, "learning_rate": 3.2583279744606246e-05, "loss": 0.0025, "step": 63260 }, { "grad_norm": 0.08332053571939468, "learning_rate": 3.25677815616768e-05, "loss": 0.0026, "step": 63270 }, { "grad_norm": 0.09027498215436935, "learning_rate": 3.255228528510856e-05, "loss": 0.0029, "step": 63280 }, { "grad_norm": 0.05725869536399841, "learning_rate": 3.253679091659617e-05, "loss": 0.0033, "step": 63290 }, { "grad_norm": 0.06227702274918556, "learning_rate": 3.252129845783404e-05, "loss": 0.0028, "step": 63300 }, { "grad_norm": 0.07623329758644104, "learning_rate": 3.2505807910516456e-05, "loss": 0.0042, "step": 63310 }, { "grad_norm": 0.050341080874204636, "learning_rate": 3.249031927633742e-05, "loss": 0.0028, "step": 63320 }, { "grad_norm": 0.05137835815548897, "learning_rate": 3.2474832556990737e-05, "loss": 0.0024, "step": 63330 }, { "grad_norm": 0.040860362350940704, "learning_rate": 3.245934775417002e-05, "loss": 0.0023, "step": 63340 }, { "grad_norm": 0.04767870530486107, "learning_rate": 3.244386486956866e-05, "loss": 0.0031, "step": 63350 }, { "grad_norm": 0.07070446014404297, "learning_rate": 3.2428383904879846e-05, "loss": 0.003, "step": 63360 }, { "grad_norm": 0.055980831384658813, "learning_rate": 3.241290486179652e-05, "loss": 0.0026, "step": 63370 }, { "grad_norm": 0.08420950174331665, "learning_rate": 3.239742774201151e-05, "loss": 0.0026, "step": 63380 }, { "grad_norm": 0.06299411505460739, "learning_rate": 3.238195254721733e-05, "loss": 0.0024, "step": 63390 }, { "grad_norm": 0.07729804515838623, "learning_rate": 3.236647927910633e-05, "loss": 0.0027, "step": 63400 }, { "grad_norm": 0.08500229567289352, "learning_rate": 3.2351007939370646e-05, "loss": 0.0032, "step": 63410 }, { "grad_norm": 0.049292486160993576, "learning_rate": 3.2335538529702184e-05, "loss": 0.0025, "step": 63420 }, { "grad_norm": 0.06338073313236237, "learning_rate": 3.232007105179268e-05, "loss": 0.0033, "step": 63430 }, { "grad_norm": 0.07876711338758469, "learning_rate": 3.230460550733363e-05, "loss": 0.0027, "step": 63440 }, { "grad_norm": 0.06744711846113205, "learning_rate": 3.2289141898016315e-05, "loss": 0.003, "step": 63450 }, { "grad_norm": 0.04894760251045227, "learning_rate": 3.227368022553181e-05, "loss": 0.0029, "step": 63460 }, { "grad_norm": 0.0746828243136406, "learning_rate": 3.225822049157099e-05, "loss": 0.0026, "step": 63470 }, { "grad_norm": 0.06580240279436111, "learning_rate": 3.2242762697824494e-05, "loss": 0.0028, "step": 63480 }, { "grad_norm": 0.06041066721081734, "learning_rate": 3.222730684598279e-05, "loss": 0.0025, "step": 63490 }, { "grad_norm": 0.061480868607759476, "learning_rate": 3.221185293773605e-05, "loss": 0.0035, "step": 63500 }, { "grad_norm": 0.05879516899585724, "learning_rate": 3.2196400974774354e-05, "loss": 0.0026, "step": 63510 }, { "grad_norm": 0.07393255829811096, "learning_rate": 3.2180950958787485e-05, "loss": 0.0028, "step": 63520 }, { "grad_norm": 0.05257195606827736, "learning_rate": 3.2165502891465024e-05, "loss": 0.0025, "step": 63530 }, { "grad_norm": 0.07401112467050552, "learning_rate": 3.215005677449636e-05, "loss": 0.0025, "step": 63540 }, { "grad_norm": 0.04865945875644684, "learning_rate": 3.2134612609570644e-05, "loss": 0.0022, "step": 63550 }, { "grad_norm": 0.0592915415763855, "learning_rate": 3.211917039837682e-05, "loss": 0.0029, "step": 63560 }, { "grad_norm": 0.05674946308135986, "learning_rate": 3.210373014260365e-05, "loss": 0.0028, "step": 63570 }, { "grad_norm": 0.04843844473361969, "learning_rate": 3.208829184393965e-05, "loss": 0.002, "step": 63580 }, { "grad_norm": 0.07972445338964462, "learning_rate": 3.207285550407312e-05, "loss": 0.0022, "step": 63590 }, { "grad_norm": 0.0613914430141449, "learning_rate": 3.205742112469217e-05, "loss": 0.0023, "step": 63600 }, { "grad_norm": 0.03929472714662552, "learning_rate": 3.204198870748465e-05, "loss": 0.0026, "step": 63610 }, { "grad_norm": 0.05456145480275154, "learning_rate": 3.202655825413824e-05, "loss": 0.0023, "step": 63620 }, { "grad_norm": 0.06764529645442963, "learning_rate": 3.2011129766340416e-05, "loss": 0.0022, "step": 63630 }, { "grad_norm": 0.0611766092479229, "learning_rate": 3.1995703245778384e-05, "loss": 0.003, "step": 63640 }, { "grad_norm": 0.04856373742222786, "learning_rate": 3.1980278694139186e-05, "loss": 0.0025, "step": 63650 }, { "grad_norm": 0.07522802799940109, "learning_rate": 3.196485611310962e-05, "loss": 0.0041, "step": 63660 }, { "grad_norm": 0.062493208795785904, "learning_rate": 3.1949435504376257e-05, "loss": 0.0022, "step": 63670 }, { "grad_norm": 0.060006678104400635, "learning_rate": 3.1934016869625505e-05, "loss": 0.0022, "step": 63680 }, { "grad_norm": 0.06323785334825516, "learning_rate": 3.191860021054349e-05, "loss": 0.003, "step": 63690 }, { "grad_norm": 0.06832127273082733, "learning_rate": 3.190318552881618e-05, "loss": 0.0033, "step": 63700 }, { "grad_norm": 0.06945522874593735, "learning_rate": 3.1887772826129285e-05, "loss": 0.0026, "step": 63710 }, { "grad_norm": 0.07958973199129105, "learning_rate": 3.1872362104168335e-05, "loss": 0.0023, "step": 63720 }, { "grad_norm": 0.05753497779369354, "learning_rate": 3.185695336461861e-05, "loss": 0.0034, "step": 63730 }, { "grad_norm": 0.05108095332980156, "learning_rate": 3.184154660916519e-05, "loss": 0.0026, "step": 63740 }, { "grad_norm": 0.05048516392707825, "learning_rate": 3.1826141839492906e-05, "loss": 0.0024, "step": 63750 }, { "grad_norm": 0.08219639211893082, "learning_rate": 3.181073905728644e-05, "loss": 0.0026, "step": 63760 }, { "grad_norm": 0.05933394655585289, "learning_rate": 3.1795338264230214e-05, "loss": 0.0023, "step": 63770 }, { "grad_norm": 0.0601360984146595, "learning_rate": 3.1779939462008426e-05, "loss": 0.0027, "step": 63780 }, { "grad_norm": 0.05310734733939171, "learning_rate": 3.1764542652305066e-05, "loss": 0.0021, "step": 63790 }, { "grad_norm": 0.05270487070083618, "learning_rate": 3.17491478368039e-05, "loss": 0.0035, "step": 63800 }, { "grad_norm": 0.052894558757543564, "learning_rate": 3.173375501718847e-05, "loss": 0.0028, "step": 63810 }, { "grad_norm": 0.061275582760572433, "learning_rate": 3.171836419514215e-05, "loss": 0.004, "step": 63820 }, { "grad_norm": 0.061289843171834946, "learning_rate": 3.170297537234803e-05, "loss": 0.0036, "step": 63830 }, { "grad_norm": 0.06602523475885391, "learning_rate": 3.168758855048901e-05, "loss": 0.0025, "step": 63840 }, { "grad_norm": 0.0571507029235363, "learning_rate": 3.1672203731247785e-05, "loss": 0.0021, "step": 63850 }, { "grad_norm": 0.1010730117559433, "learning_rate": 3.16568209163068e-05, "loss": 0.0029, "step": 63860 }, { "grad_norm": 0.04313153773546219, "learning_rate": 3.1641440107348295e-05, "loss": 0.0025, "step": 63870 }, { "grad_norm": 0.07487238198518753, "learning_rate": 3.162606130605429e-05, "loss": 0.0024, "step": 63880 }, { "grad_norm": 0.05132739618420601, "learning_rate": 3.16106845141066e-05, "loss": 0.0029, "step": 63890 }, { "grad_norm": 0.055210474878549576, "learning_rate": 3.159530973318682e-05, "loss": 0.0028, "step": 63900 }, { "grad_norm": 0.05267881974577904, "learning_rate": 3.1579936964976286e-05, "loss": 0.0028, "step": 63910 }, { "grad_norm": 0.06470435112714767, "learning_rate": 3.156456621115615e-05, "loss": 0.0022, "step": 63920 }, { "grad_norm": 0.045241955667734146, "learning_rate": 3.154919747340734e-05, "loss": 0.0027, "step": 63930 }, { "grad_norm": 0.060275956988334656, "learning_rate": 3.153383075341054e-05, "loss": 0.0035, "step": 63940 }, { "grad_norm": 0.05586893483996391, "learning_rate": 3.1518466052846264e-05, "loss": 0.0022, "step": 63950 }, { "grad_norm": 0.06394504010677338, "learning_rate": 3.150310337339474e-05, "loss": 0.0025, "step": 63960 }, { "grad_norm": 0.05006295070052147, "learning_rate": 3.148774271673603e-05, "loss": 0.0028, "step": 63970 }, { "grad_norm": 0.04588710516691208, "learning_rate": 3.147238408454995e-05, "loss": 0.0033, "step": 63980 }, { "grad_norm": 0.04324697330594063, "learning_rate": 3.1457027478516086e-05, "loss": 0.0021, "step": 63990 }, { "grad_norm": 0.06190486624836922, "learning_rate": 3.144167290031382e-05, "loss": 0.0025, "step": 64000 }, { "grad_norm": 0.043557800352573395, "learning_rate": 3.142632035162227e-05, "loss": 0.0019, "step": 64010 }, { "grad_norm": 0.07004143297672272, "learning_rate": 3.141096983412044e-05, "loss": 0.0024, "step": 64020 }, { "grad_norm": 0.09566680341959, "learning_rate": 3.139562134948699e-05, "loss": 0.0026, "step": 64030 }, { "grad_norm": 0.06771042943000793, "learning_rate": 3.1380274899400424e-05, "loss": 0.003, "step": 64040 }, { "grad_norm": 0.06147133558988571, "learning_rate": 3.1364930485539e-05, "loss": 0.0027, "step": 64050 }, { "grad_norm": 0.04915378615260124, "learning_rate": 3.134958810958076e-05, "loss": 0.0023, "step": 64060 }, { "grad_norm": 0.04129202291369438, "learning_rate": 3.13342477732035e-05, "loss": 0.0022, "step": 64070 }, { "grad_norm": 0.052996572107076645, "learning_rate": 3.131890947808487e-05, "loss": 0.0027, "step": 64080 }, { "grad_norm": 0.053971678018569946, "learning_rate": 3.13035732259022e-05, "loss": 0.0019, "step": 64090 }, { "grad_norm": 0.050602447241544724, "learning_rate": 3.1288239018332656e-05, "loss": 0.0025, "step": 64100 }, { "grad_norm": 0.040925536304712296, "learning_rate": 3.1272906857053164e-05, "loss": 0.003, "step": 64110 }, { "grad_norm": 0.09308667480945587, "learning_rate": 3.125757674374042e-05, "loss": 0.0026, "step": 64120 }, { "grad_norm": 0.05898602306842804, "learning_rate": 3.124224868007088e-05, "loss": 0.0024, "step": 64130 }, { "grad_norm": 0.06412754952907562, "learning_rate": 3.1226922667720854e-05, "loss": 0.0026, "step": 64140 }, { "grad_norm": 0.053713761270046234, "learning_rate": 3.121159870836633e-05, "loss": 0.0025, "step": 64150 }, { "grad_norm": 0.05374186113476753, "learning_rate": 3.119627680368312e-05, "loss": 0.0028, "step": 64160 }, { "grad_norm": 0.052399296313524246, "learning_rate": 3.1180956955346816e-05, "loss": 0.0023, "step": 64170 }, { "grad_norm": 0.0508464053273201, "learning_rate": 3.116563916503276e-05, "loss": 0.0027, "step": 64180 }, { "grad_norm": 0.07264117896556854, "learning_rate": 3.1150323434416095e-05, "loss": 0.0026, "step": 64190 }, { "grad_norm": 0.06400381028652191, "learning_rate": 3.1135009765171696e-05, "loss": 0.0024, "step": 64200 }, { "grad_norm": 0.059792883694171906, "learning_rate": 3.111969815897429e-05, "loss": 0.0024, "step": 64210 }, { "grad_norm": 0.06460274755954742, "learning_rate": 3.1104388617498295e-05, "loss": 0.0021, "step": 64220 }, { "grad_norm": 0.061943355947732925, "learning_rate": 3.1089081142417966e-05, "loss": 0.0022, "step": 64230 }, { "grad_norm": 0.08654847741127014, "learning_rate": 3.1073775735407285e-05, "loss": 0.0023, "step": 64240 }, { "grad_norm": 0.04024116322398186, "learning_rate": 3.1058472398140035e-05, "loss": 0.0028, "step": 64250 }, { "grad_norm": 0.07796408981084824, "learning_rate": 3.1043171132289736e-05, "loss": 0.0027, "step": 64260 }, { "grad_norm": 0.05126258358359337, "learning_rate": 3.1027871939529774e-05, "loss": 0.002, "step": 64270 }, { "grad_norm": 0.054198864847421646, "learning_rate": 3.1012574821533214e-05, "loss": 0.0026, "step": 64280 }, { "grad_norm": 0.06250904500484467, "learning_rate": 3.099727977997291e-05, "loss": 0.0039, "step": 64290 }, { "grad_norm": 0.06580889225006104, "learning_rate": 3.098198681652154e-05, "loss": 0.003, "step": 64300 }, { "grad_norm": 0.06333482265472412, "learning_rate": 3.0966695932851475e-05, "loss": 0.0027, "step": 64310 }, { "grad_norm": 0.049889061599969864, "learning_rate": 3.0951407130634926e-05, "loss": 0.0025, "step": 64320 }, { "grad_norm": 0.06622134894132614, "learning_rate": 3.0936120411543864e-05, "loss": 0.0026, "step": 64330 }, { "grad_norm": 0.07114749401807785, "learning_rate": 3.0920835777250015e-05, "loss": 0.0025, "step": 64340 }, { "grad_norm": 0.060414206236600876, "learning_rate": 3.090555322942488e-05, "loss": 0.0028, "step": 64350 }, { "grad_norm": 0.04588286951184273, "learning_rate": 3.089027276973973e-05, "loss": 0.0023, "step": 64360 }, { "grad_norm": 0.06045074388384819, "learning_rate": 3.0874994399865615e-05, "loss": 0.0022, "step": 64370 }, { "grad_norm": 0.0380813330411911, "learning_rate": 3.085971812147337e-05, "loss": 0.0022, "step": 64380 }, { "grad_norm": 0.07590338587760925, "learning_rate": 3.0844443936233557e-05, "loss": 0.0029, "step": 64390 }, { "grad_norm": 0.05451744422316551, "learning_rate": 3.082917184581657e-05, "loss": 0.0027, "step": 64400 }, { "grad_norm": 0.051806896924972534, "learning_rate": 3.0813901851892513e-05, "loss": 0.0025, "step": 64410 }, { "grad_norm": 0.07100978493690491, "learning_rate": 3.0798633956131316e-05, "loss": 0.0026, "step": 64420 }, { "grad_norm": 0.04470217227935791, "learning_rate": 3.078336816020264e-05, "loss": 0.0022, "step": 64430 }, { "grad_norm": 0.05370499566197395, "learning_rate": 3.0768104465775935e-05, "loss": 0.0025, "step": 64440 }, { "grad_norm": 0.054073840379714966, "learning_rate": 3.0752842874520396e-05, "loss": 0.0021, "step": 64450 }, { "grad_norm": 0.05926162749528885, "learning_rate": 3.073758338810503e-05, "loss": 0.0025, "step": 64460 }, { "grad_norm": 0.061266280710697174, "learning_rate": 3.072232600819859e-05, "loss": 0.0026, "step": 64470 }, { "grad_norm": 0.06859979778528214, "learning_rate": 3.070707073646959e-05, "loss": 0.0025, "step": 64480 }, { "grad_norm": 0.06704802811145782, "learning_rate": 3.069181757458633e-05, "loss": 0.0022, "step": 64490 }, { "grad_norm": 0.05729061737656593, "learning_rate": 3.067656652421687e-05, "loss": 0.0023, "step": 64500 }, { "grad_norm": 0.06882583349943161, "learning_rate": 3.0661317587029025e-05, "loss": 0.0026, "step": 64510 }, { "grad_norm": 0.05844811350107193, "learning_rate": 3.064607076469043e-05, "loss": 0.0023, "step": 64520 }, { "grad_norm": 0.04724261164665222, "learning_rate": 3.063082605886845e-05, "loss": 0.0031, "step": 64530 }, { "grad_norm": 0.05560118332505226, "learning_rate": 3.061558347123019e-05, "loss": 0.0022, "step": 64540 }, { "grad_norm": 0.07079322636127472, "learning_rate": 3.060034300344259e-05, "loss": 0.0027, "step": 64550 }, { "grad_norm": 0.0669875219464302, "learning_rate": 3.058510465717231e-05, "loss": 0.0026, "step": 64560 }, { "grad_norm": 0.0661267563700676, "learning_rate": 3.0569868434085804e-05, "loss": 0.0031, "step": 64570 }, { "grad_norm": 0.0678272545337677, "learning_rate": 3.055463433584924e-05, "loss": 0.0028, "step": 64580 }, { "grad_norm": 0.06155514344573021, "learning_rate": 3.0539402364128656e-05, "loss": 0.0026, "step": 64590 }, { "grad_norm": 0.04686637222766876, "learning_rate": 3.052417252058977e-05, "loss": 0.0021, "step": 64600 }, { "grad_norm": 0.06504065543413162, "learning_rate": 3.0508944806898092e-05, "loss": 0.0031, "step": 64610 }, { "grad_norm": 0.07589630037546158, "learning_rate": 3.04937192247189e-05, "loss": 0.0037, "step": 64620 }, { "grad_norm": 0.06875956058502197, "learning_rate": 3.047849577571725e-05, "loss": 0.0033, "step": 64630 }, { "grad_norm": 0.064798504114151, "learning_rate": 3.046327446155795e-05, "loss": 0.0025, "step": 64640 }, { "grad_norm": 0.047990743070840836, "learning_rate": 3.0448055283905587e-05, "loss": 0.0024, "step": 64650 }, { "grad_norm": 0.06754280626773834, "learning_rate": 3.0432838244424494e-05, "loss": 0.0031, "step": 64660 }, { "grad_norm": 0.058073993772268295, "learning_rate": 3.04176233447788e-05, "loss": 0.0037, "step": 64670 }, { "grad_norm": 0.0520002655684948, "learning_rate": 3.040241058663238e-05, "loss": 0.0024, "step": 64680 }, { "grad_norm": 0.052769046276807785, "learning_rate": 3.038719997164887e-05, "loss": 0.0023, "step": 64690 }, { "grad_norm": 0.05690388381481171, "learning_rate": 3.0371991501491683e-05, "loss": 0.0034, "step": 64700 }, { "grad_norm": 0.058898039162158966, "learning_rate": 3.0356785177823975e-05, "loss": 0.003, "step": 64710 }, { "grad_norm": 0.052228718996047974, "learning_rate": 3.0341581002308726e-05, "loss": 0.0025, "step": 64720 }, { "grad_norm": 0.05871758237481117, "learning_rate": 3.0326378976608627e-05, "loss": 0.0018, "step": 64730 }, { "grad_norm": 0.06969272345304489, "learning_rate": 3.0311179102386146e-05, "loss": 0.0028, "step": 64740 }, { "grad_norm": 0.051897287368774414, "learning_rate": 3.0295981381303513e-05, "loss": 0.0027, "step": 64750 }, { "grad_norm": 0.09652256965637207, "learning_rate": 3.0280785815022727e-05, "loss": 0.0031, "step": 64760 }, { "grad_norm": 0.07049515843391418, "learning_rate": 3.0265592405205544e-05, "loss": 0.0023, "step": 64770 }, { "grad_norm": 0.08761894702911377, "learning_rate": 3.0250401153513517e-05, "loss": 0.0027, "step": 64780 }, { "grad_norm": 0.05383254587650299, "learning_rate": 3.0235212061607936e-05, "loss": 0.0021, "step": 64790 }, { "grad_norm": 0.053300466388463974, "learning_rate": 3.0220025131149842e-05, "loss": 0.0031, "step": 64800 }, { "grad_norm": 0.042281895875930786, "learning_rate": 3.020484036380007e-05, "loss": 0.0029, "step": 64810 }, { "grad_norm": 0.05222807824611664, "learning_rate": 3.0189657761219188e-05, "loss": 0.0024, "step": 64820 }, { "grad_norm": 0.061951443552970886, "learning_rate": 3.017447732506753e-05, "loss": 0.0029, "step": 64830 }, { "grad_norm": 0.05633161589503288, "learning_rate": 3.015929905700524e-05, "loss": 0.0025, "step": 64840 }, { "grad_norm": 0.055040143430233, "learning_rate": 3.014412295869218e-05, "loss": 0.0025, "step": 64850 }, { "grad_norm": 0.059883590787649155, "learning_rate": 3.0128949031787977e-05, "loss": 0.0027, "step": 64860 }, { "grad_norm": 0.06020614877343178, "learning_rate": 3.0113777277952022e-05, "loss": 0.0028, "step": 64870 }, { "grad_norm": 0.09634587168693542, "learning_rate": 3.0098607698843496e-05, "loss": 0.0021, "step": 64880 }, { "grad_norm": 0.0907045379281044, "learning_rate": 3.0083440296121312e-05, "loss": 0.0022, "step": 64890 }, { "grad_norm": 0.04664657264947891, "learning_rate": 3.0068275071444142e-05, "loss": 0.0033, "step": 64900 }, { "grad_norm": 0.06772950291633606, "learning_rate": 3.0053112026470442e-05, "loss": 0.0029, "step": 64910 }, { "grad_norm": 0.07647258788347244, "learning_rate": 3.0037951162858434e-05, "loss": 0.0026, "step": 64920 }, { "grad_norm": 0.06740279495716095, "learning_rate": 3.002279248226607e-05, "loss": 0.0027, "step": 64930 }, { "grad_norm": 0.05765146389603615, "learning_rate": 3.000763598635109e-05, "loss": 0.0022, "step": 64940 }, { "grad_norm": 0.04483206570148468, "learning_rate": 2.999248167677098e-05, "loss": 0.002, "step": 64950 }, { "grad_norm": 0.05421774834394455, "learning_rate": 2.9977329555182976e-05, "loss": 0.0024, "step": 64960 }, { "grad_norm": 0.05892917886376381, "learning_rate": 2.9962179623244124e-05, "loss": 0.0029, "step": 64970 }, { "grad_norm": 0.05318215489387512, "learning_rate": 2.9947031882611175e-05, "loss": 0.0036, "step": 64980 }, { "grad_norm": 0.08062678575515747, "learning_rate": 2.9931886334940685e-05, "loss": 0.0025, "step": 64990 }, { "grad_norm": 0.04983380809426308, "learning_rate": 2.9916742981888923e-05, "loss": 0.0025, "step": 65000 }, { "grad_norm": 0.05046211555600166, "learning_rate": 2.990160182511196e-05, "loss": 0.0023, "step": 65010 }, { "grad_norm": 0.04421152174472809, "learning_rate": 2.9886462866265574e-05, "loss": 0.0025, "step": 65020 }, { "grad_norm": 0.06328349560499191, "learning_rate": 2.9871326107005392e-05, "loss": 0.0031, "step": 65030 }, { "grad_norm": 0.057909052819013596, "learning_rate": 2.9856191548986735e-05, "loss": 0.0023, "step": 65040 }, { "grad_norm": 0.08039800077676773, "learning_rate": 2.9841059193864672e-05, "loss": 0.0026, "step": 65050 }, { "grad_norm": 0.08670505881309509, "learning_rate": 2.982592904329407e-05, "loss": 0.0026, "step": 65060 }, { "grad_norm": 0.05275846645236015, "learning_rate": 2.981080109892954e-05, "loss": 0.0034, "step": 65070 }, { "grad_norm": 0.1778343915939331, "learning_rate": 2.9795675362425434e-05, "loss": 0.0025, "step": 65080 }, { "grad_norm": 0.06413696706295013, "learning_rate": 2.97805518354359e-05, "loss": 0.0033, "step": 65090 }, { "grad_norm": 0.05683925375342369, "learning_rate": 2.9765430519614825e-05, "loss": 0.0024, "step": 65100 }, { "grad_norm": 0.04838646575808525, "learning_rate": 2.975031141661584e-05, "loss": 0.003, "step": 65110 }, { "grad_norm": 0.04099755734205246, "learning_rate": 2.9735194528092347e-05, "loss": 0.003, "step": 65120 }, { "grad_norm": 0.04171539843082428, "learning_rate": 2.9720079855697525e-05, "loss": 0.0023, "step": 65130 }, { "grad_norm": 0.06200290471315384, "learning_rate": 2.970496740108427e-05, "loss": 0.0023, "step": 65140 }, { "grad_norm": 0.0732104554772377, "learning_rate": 2.968985716590527e-05, "loss": 0.0026, "step": 65150 }, { "grad_norm": 0.04634767398238182, "learning_rate": 2.9674749151812953e-05, "loss": 0.0022, "step": 65160 }, { "grad_norm": 0.062239743769168854, "learning_rate": 2.965964336045951e-05, "loss": 0.003, "step": 65170 }, { "grad_norm": 0.06675580888986588, "learning_rate": 2.9644539793496894e-05, "loss": 0.0027, "step": 65180 }, { "grad_norm": 0.06525734812021255, "learning_rate": 2.9629438452576807e-05, "loss": 0.0023, "step": 65190 }, { "grad_norm": 0.05632927268743515, "learning_rate": 2.9614339339350698e-05, "loss": 0.0027, "step": 65200 }, { "grad_norm": 0.060797128826379776, "learning_rate": 2.959924245546979e-05, "loss": 0.0027, "step": 65210 }, { "grad_norm": 0.07352002710103989, "learning_rate": 2.9584147802585038e-05, "loss": 0.0027, "step": 65220 }, { "grad_norm": 0.07318584620952606, "learning_rate": 2.956905538234721e-05, "loss": 0.0027, "step": 65230 }, { "grad_norm": 0.039394304156303406, "learning_rate": 2.9553965196406764e-05, "loss": 0.0022, "step": 65240 }, { "grad_norm": 0.04490797966718674, "learning_rate": 2.9538877246413943e-05, "loss": 0.0019, "step": 65250 }, { "grad_norm": 0.03338265419006348, "learning_rate": 2.952379153401874e-05, "loss": 0.0017, "step": 65260 }, { "grad_norm": 0.048198726028203964, "learning_rate": 2.9508708060870914e-05, "loss": 0.0023, "step": 65270 }, { "grad_norm": 0.04983265697956085, "learning_rate": 2.9493626828619948e-05, "loss": 0.0024, "step": 65280 }, { "grad_norm": 0.05642365664243698, "learning_rate": 2.9478547838915137e-05, "loss": 0.0025, "step": 65290 }, { "grad_norm": 0.07141723483800888, "learning_rate": 2.946347109340548e-05, "loss": 0.0037, "step": 65300 }, { "grad_norm": 0.047724563628435135, "learning_rate": 2.944839659373976e-05, "loss": 0.0023, "step": 65310 }, { "grad_norm": 0.07485931366682053, "learning_rate": 2.943332434156647e-05, "loss": 0.0025, "step": 65320 }, { "grad_norm": 0.05600113049149513, "learning_rate": 2.9418254338533912e-05, "loss": 0.0026, "step": 65330 }, { "grad_norm": 0.06198926642537117, "learning_rate": 2.940318658629011e-05, "loss": 0.0028, "step": 65340 }, { "grad_norm": 0.0544840507209301, "learning_rate": 2.938812108648287e-05, "loss": 0.0023, "step": 65350 }, { "grad_norm": 0.04522804543375969, "learning_rate": 2.937305784075972e-05, "loss": 0.0023, "step": 65360 }, { "grad_norm": 0.05295614153146744, "learning_rate": 2.9357996850767937e-05, "loss": 0.0026, "step": 65370 }, { "grad_norm": 0.06648129969835281, "learning_rate": 2.9342938118154595e-05, "loss": 0.0024, "step": 65380 }, { "grad_norm": 0.04258011654019356, "learning_rate": 2.9327881644566485e-05, "loss": 0.0026, "step": 65390 }, { "grad_norm": 0.044639404863119125, "learning_rate": 2.9312827431650157e-05, "loss": 0.0034, "step": 65400 }, { "grad_norm": 0.06254185736179352, "learning_rate": 2.92977754810519e-05, "loss": 0.0027, "step": 65410 }, { "grad_norm": 0.05553703010082245, "learning_rate": 2.928272579441781e-05, "loss": 0.0024, "step": 65420 }, { "grad_norm": 0.06706558167934418, "learning_rate": 2.9267678373393676e-05, "loss": 0.0026, "step": 65430 }, { "grad_norm": 0.062150295823812485, "learning_rate": 2.925263321962507e-05, "loss": 0.0021, "step": 65440 }, { "grad_norm": 0.08587118983268738, "learning_rate": 2.9237590334757292e-05, "loss": 0.0028, "step": 65450 }, { "grad_norm": 0.05476953461766243, "learning_rate": 2.922254972043543e-05, "loss": 0.0023, "step": 65460 }, { "grad_norm": 0.07384809106588364, "learning_rate": 2.9207511378304265e-05, "loss": 0.0021, "step": 65470 }, { "grad_norm": 0.050116539001464844, "learning_rate": 2.9192475310008428e-05, "loss": 0.0033, "step": 65480 }, { "grad_norm": 0.06777410954236984, "learning_rate": 2.917744151719219e-05, "loss": 0.0023, "step": 65490 }, { "grad_norm": 0.047722719609737396, "learning_rate": 2.916241000149964e-05, "loss": 0.0024, "step": 65500 }, { "grad_norm": 0.07159249484539032, "learning_rate": 2.9147380764574622e-05, "loss": 0.0034, "step": 65510 }, { "grad_norm": 0.06787446141242981, "learning_rate": 2.9132353808060674e-05, "loss": 0.0025, "step": 65520 }, { "grad_norm": 0.06461642682552338, "learning_rate": 2.9117329133601134e-05, "loss": 0.0042, "step": 65530 }, { "grad_norm": 0.04617998003959656, "learning_rate": 2.9102306742839093e-05, "loss": 0.0024, "step": 65540 }, { "grad_norm": 0.05442268028855324, "learning_rate": 2.9087286637417375e-05, "loss": 0.0024, "step": 65550 }, { "grad_norm": 0.05573124811053276, "learning_rate": 2.9072268818978537e-05, "loss": 0.0025, "step": 65560 }, { "grad_norm": 0.058615922927856445, "learning_rate": 2.905725328916493e-05, "loss": 0.0026, "step": 65570 }, { "grad_norm": 0.04926380515098572, "learning_rate": 2.9042240049618608e-05, "loss": 0.0017, "step": 65580 }, { "grad_norm": 0.07683757692575455, "learning_rate": 2.9027229101981413e-05, "loss": 0.0023, "step": 65590 }, { "grad_norm": 0.07208523899316788, "learning_rate": 2.901222044789488e-05, "loss": 0.0022, "step": 65600 }, { "grad_norm": 0.05055234581232071, "learning_rate": 2.89972140890004e-05, "loss": 0.0024, "step": 65610 }, { "grad_norm": 0.056392934173345566, "learning_rate": 2.8982210026938988e-05, "loss": 0.0021, "step": 65620 }, { "grad_norm": 0.057457685470581055, "learning_rate": 2.896720826335151e-05, "loss": 0.0028, "step": 65630 }, { "grad_norm": 0.05212629586458206, "learning_rate": 2.8952208799878493e-05, "loss": 0.0021, "step": 65640 }, { "grad_norm": 0.04726032167673111, "learning_rate": 2.89372116381603e-05, "loss": 0.0019, "step": 65650 }, { "grad_norm": 0.04799756035208702, "learning_rate": 2.8922216779836932e-05, "loss": 0.0022, "step": 65660 }, { "grad_norm": 0.04497922584414482, "learning_rate": 2.8907224226548285e-05, "loss": 0.002, "step": 65670 }, { "grad_norm": 0.038699910044670105, "learning_rate": 2.8892233979933868e-05, "loss": 0.0019, "step": 65680 }, { "grad_norm": 0.07195287942886353, "learning_rate": 2.8877246041633026e-05, "loss": 0.0028, "step": 65690 }, { "grad_norm": 0.06345163285732269, "learning_rate": 2.886226041328478e-05, "loss": 0.0022, "step": 65700 }, { "grad_norm": 0.0318060964345932, "learning_rate": 2.8847277096527957e-05, "loss": 0.0038, "step": 65710 }, { "grad_norm": 0.06278829276561737, "learning_rate": 2.8832296093001105e-05, "loss": 0.0023, "step": 65720 }, { "grad_norm": 0.05093918740749359, "learning_rate": 2.8817317404342547e-05, "loss": 0.0023, "step": 65730 }, { "grad_norm": 0.045067425817251205, "learning_rate": 2.880234103219029e-05, "loss": 0.0024, "step": 65740 }, { "grad_norm": 0.055026065558195114, "learning_rate": 2.878736697818215e-05, "loss": 0.0027, "step": 65750 }, { "grad_norm": 0.06564971804618835, "learning_rate": 2.877239524395568e-05, "loss": 0.0029, "step": 65760 }, { "grad_norm": 0.056583140045404434, "learning_rate": 2.8757425831148134e-05, "loss": 0.0025, "step": 65770 }, { "grad_norm": 0.05349380150437355, "learning_rate": 2.874245874139658e-05, "loss": 0.0024, "step": 65780 }, { "grad_norm": 0.08960241824388504, "learning_rate": 2.8727493976337737e-05, "loss": 0.0027, "step": 65790 }, { "grad_norm": 0.09235375374555588, "learning_rate": 2.8712531537608218e-05, "loss": 0.0028, "step": 65800 }, { "grad_norm": 0.0573439821600914, "learning_rate": 2.8697571426844217e-05, "loss": 0.0023, "step": 65810 }, { "grad_norm": 0.04315672069787979, "learning_rate": 2.86826136456818e-05, "loss": 0.0025, "step": 65820 }, { "grad_norm": 0.06499294936656952, "learning_rate": 2.8667658195756685e-05, "loss": 0.0032, "step": 65830 }, { "grad_norm": 0.06185344606637955, "learning_rate": 2.8652705078704412e-05, "loss": 0.0036, "step": 65840 }, { "grad_norm": 0.05943341925740242, "learning_rate": 2.863775429616018e-05, "loss": 0.0026, "step": 65850 }, { "grad_norm": 0.06146295368671417, "learning_rate": 2.862280584975906e-05, "loss": 0.0028, "step": 65860 }, { "grad_norm": 0.07706128060817719, "learning_rate": 2.8607859741135735e-05, "loss": 0.0036, "step": 65870 }, { "grad_norm": 0.048502445220947266, "learning_rate": 2.8592915971924723e-05, "loss": 0.0029, "step": 65880 }, { "grad_norm": 0.046849872916936874, "learning_rate": 2.8577974543760212e-05, "loss": 0.0021, "step": 65890 }, { "grad_norm": 0.05714605376124382, "learning_rate": 2.8563035458276222e-05, "loss": 0.0023, "step": 65900 }, { "grad_norm": 0.04075569659471512, "learning_rate": 2.8548098717106424e-05, "loss": 0.002, "step": 65910 }, { "grad_norm": 0.0847310945391655, "learning_rate": 2.85331643218843e-05, "loss": 0.0022, "step": 65920 }, { "grad_norm": 0.05417923256754875, "learning_rate": 2.8518232274243046e-05, "loss": 0.0028, "step": 65930 }, { "grad_norm": 0.05981681868433952, "learning_rate": 2.8503302575815638e-05, "loss": 0.0031, "step": 65940 }, { "grad_norm": 0.07188164442777634, "learning_rate": 2.8488375228234716e-05, "loss": 0.0032, "step": 65950 }, { "grad_norm": 0.06273775547742844, "learning_rate": 2.8473450233132738e-05, "loss": 0.0029, "step": 65960 }, { "grad_norm": 0.049521394073963165, "learning_rate": 2.8458527592141898e-05, "loss": 0.0029, "step": 65970 }, { "grad_norm": 0.05967878922820091, "learning_rate": 2.8443607306894073e-05, "loss": 0.0021, "step": 65980 }, { "grad_norm": 0.10376645624637604, "learning_rate": 2.8428689379020946e-05, "loss": 0.0025, "step": 65990 }, { "grad_norm": 0.063194639980793, "learning_rate": 2.841377381015391e-05, "loss": 0.003, "step": 66000 }, { "grad_norm": 0.06270228326320648, "learning_rate": 2.8398860601924144e-05, "loss": 0.0031, "step": 66010 }, { "grad_norm": 0.05850830301642418, "learning_rate": 2.838394975596248e-05, "loss": 0.0023, "step": 66020 }, { "grad_norm": 0.04813828319311142, "learning_rate": 2.83690412738996e-05, "loss": 0.002, "step": 66030 }, { "grad_norm": 0.049466755241155624, "learning_rate": 2.835413515736579e-05, "loss": 0.0023, "step": 66040 }, { "grad_norm": 0.07804924249649048, "learning_rate": 2.8339231407991267e-05, "loss": 0.0036, "step": 66050 }, { "grad_norm": 0.061738330870866776, "learning_rate": 2.832433002740581e-05, "loss": 0.0024, "step": 66060 }, { "grad_norm": 0.03196750208735466, "learning_rate": 2.8309431017239047e-05, "loss": 0.0021, "step": 66070 }, { "grad_norm": 0.055952172726392746, "learning_rate": 2.829453437912028e-05, "loss": 0.002, "step": 66080 }, { "grad_norm": 0.06103792041540146, "learning_rate": 2.8279640114678617e-05, "loss": 0.0025, "step": 66090 }, { "grad_norm": 0.12263219058513641, "learning_rate": 2.8264748225542836e-05, "loss": 0.0023, "step": 66100 }, { "grad_norm": 0.05688204616308212, "learning_rate": 2.8249858713341516e-05, "loss": 0.0024, "step": 66110 }, { "grad_norm": 0.05982901155948639, "learning_rate": 2.823497157970294e-05, "loss": 0.0024, "step": 66120 }, { "grad_norm": 0.04785102978348732, "learning_rate": 2.8220086826255164e-05, "loss": 0.0021, "step": 66130 }, { "grad_norm": 0.05304838716983795, "learning_rate": 2.8205204454625934e-05, "loss": 0.0023, "step": 66140 }, { "grad_norm": 0.05498427525162697, "learning_rate": 2.8190324466442798e-05, "loss": 0.0022, "step": 66150 }, { "grad_norm": 0.04385814070701599, "learning_rate": 2.8175446863332965e-05, "loss": 0.0026, "step": 66160 }, { "grad_norm": 0.055175457149744034, "learning_rate": 2.816057164692345e-05, "loss": 0.0022, "step": 66170 }, { "grad_norm": 0.04811447486281395, "learning_rate": 2.8145698818840992e-05, "loss": 0.0019, "step": 66180 }, { "grad_norm": 0.0649375170469284, "learning_rate": 2.8130828380712065e-05, "loss": 0.0032, "step": 66190 }, { "grad_norm": 0.05165204778313637, "learning_rate": 2.811596033416285e-05, "loss": 0.002, "step": 66200 }, { "grad_norm": 0.04586106538772583, "learning_rate": 2.8101094680819307e-05, "loss": 0.0029, "step": 66210 }, { "grad_norm": 0.06483720242977142, "learning_rate": 2.808623142230715e-05, "loss": 0.0023, "step": 66220 }, { "grad_norm": 0.11174393445253372, "learning_rate": 2.8071370560251757e-05, "loss": 0.0024, "step": 66230 }, { "grad_norm": 0.05728275328874588, "learning_rate": 2.805651209627831e-05, "loss": 0.002, "step": 66240 }, { "grad_norm": 0.06411654502153397, "learning_rate": 2.804165603201171e-05, "loss": 0.0027, "step": 66250 }, { "grad_norm": 0.07242955267429352, "learning_rate": 2.802680236907661e-05, "loss": 0.0025, "step": 66260 }, { "grad_norm": 0.03765071928501129, "learning_rate": 2.8011951109097344e-05, "loss": 0.0028, "step": 66270 }, { "grad_norm": 0.10221756249666214, "learning_rate": 2.7997102253698072e-05, "loss": 0.0024, "step": 66280 }, { "grad_norm": 0.0637354776263237, "learning_rate": 2.7982255804502588e-05, "loss": 0.0023, "step": 66290 }, { "grad_norm": 0.0463545098900795, "learning_rate": 2.7967411763134516e-05, "loss": 0.0022, "step": 66300 }, { "grad_norm": 0.04987552389502525, "learning_rate": 2.7952570131217166e-05, "loss": 0.0027, "step": 66310 }, { "grad_norm": 0.051038723438978195, "learning_rate": 2.7937730910373618e-05, "loss": 0.0034, "step": 66320 }, { "grad_norm": 0.05974247679114342, "learning_rate": 2.7922894102226628e-05, "loss": 0.0019, "step": 66330 }, { "grad_norm": 0.05643784627318382, "learning_rate": 2.790805970839876e-05, "loss": 0.0022, "step": 66340 }, { "grad_norm": 0.062211886048316956, "learning_rate": 2.789322773051225e-05, "loss": 0.0026, "step": 66350 }, { "grad_norm": 0.05716157704591751, "learning_rate": 2.7878398170189124e-05, "loss": 0.0032, "step": 66360 }, { "grad_norm": 0.06087365746498108, "learning_rate": 2.7863571029051112e-05, "loss": 0.0024, "step": 66370 }, { "grad_norm": 0.04881703853607178, "learning_rate": 2.784874630871971e-05, "loss": 0.0023, "step": 66380 }, { "grad_norm": 0.07086341083049774, "learning_rate": 2.7833924010816086e-05, "loss": 0.0022, "step": 66390 }, { "grad_norm": 0.07755827158689499, "learning_rate": 2.7819104136961226e-05, "loss": 0.0026, "step": 66400 }, { "grad_norm": 0.06584462523460388, "learning_rate": 2.780428668877577e-05, "loss": 0.0025, "step": 66410 }, { "grad_norm": 0.0829475149512291, "learning_rate": 2.7789471667880152e-05, "loss": 0.0027, "step": 66420 }, { "grad_norm": 0.0625976175069809, "learning_rate": 2.7774659075894523e-05, "loss": 0.0022, "step": 66430 }, { "grad_norm": 0.06119221821427345, "learning_rate": 2.7759848914438778e-05, "loss": 0.0022, "step": 66440 }, { "grad_norm": 0.05322447046637535, "learning_rate": 2.7745041185132496e-05, "loss": 0.002, "step": 66450 }, { "grad_norm": 0.05272011458873749, "learning_rate": 2.773023588959506e-05, "loss": 0.0025, "step": 66460 }, { "grad_norm": 0.06845401972532272, "learning_rate": 2.7715433029445555e-05, "loss": 0.0033, "step": 66470 }, { "grad_norm": 0.05740690231323242, "learning_rate": 2.7700632606302766e-05, "loss": 0.0027, "step": 66480 }, { "grad_norm": 0.07307030260562897, "learning_rate": 2.768583462178528e-05, "loss": 0.0029, "step": 66490 }, { "grad_norm": 0.05968242138624191, "learning_rate": 2.767103907751136e-05, "loss": 0.0036, "step": 66500 }, { "grad_norm": 0.07207556813955307, "learning_rate": 2.7656245975099065e-05, "loss": 0.0027, "step": 66510 }, { "grad_norm": 0.04668368026614189, "learning_rate": 2.7641455316166093e-05, "loss": 0.0017, "step": 66520 }, { "grad_norm": 0.053371407091617584, "learning_rate": 2.7626667102329973e-05, "loss": 0.002, "step": 66530 }, { "grad_norm": 0.05535179376602173, "learning_rate": 2.761188133520788e-05, "loss": 0.0024, "step": 66540 }, { "grad_norm": 0.05582769215106964, "learning_rate": 2.7597098016416784e-05, "loss": 0.0026, "step": 66550 }, { "grad_norm": 0.06325406581163406, "learning_rate": 2.7582317147573368e-05, "loss": 0.0019, "step": 66560 }, { "grad_norm": 0.04086614027619362, "learning_rate": 2.7567538730294064e-05, "loss": 0.002, "step": 66570 }, { "grad_norm": 0.0718701109290123, "learning_rate": 2.7552762766194973e-05, "loss": 0.0026, "step": 66580 }, { "grad_norm": 0.05967329815030098, "learning_rate": 2.753798925689202e-05, "loss": 0.0025, "step": 66590 }, { "grad_norm": 0.07507141679525375, "learning_rate": 2.7523218204000774e-05, "loss": 0.0031, "step": 66600 }, { "grad_norm": 0.041047126054763794, "learning_rate": 2.7508449609136612e-05, "loss": 0.0019, "step": 66610 }, { "grad_norm": 0.05647389218211174, "learning_rate": 2.749368347391454e-05, "loss": 0.0026, "step": 66620 }, { "grad_norm": 0.05461641773581505, "learning_rate": 2.7478919799949443e-05, "loss": 0.0025, "step": 66630 }, { "grad_norm": 0.06411527097225189, "learning_rate": 2.7464158588855803e-05, "loss": 0.0018, "step": 66640 }, { "grad_norm": 0.0688675194978714, "learning_rate": 2.7449399842247918e-05, "loss": 0.0033, "step": 66650 }, { "grad_norm": 0.06584621965885162, "learning_rate": 2.7434643561739736e-05, "loss": 0.0028, "step": 66660 }, { "grad_norm": 0.057245902717113495, "learning_rate": 2.7419889748945016e-05, "loss": 0.0029, "step": 66670 }, { "grad_norm": 0.07076963037252426, "learning_rate": 2.7405138405477203e-05, "loss": 0.0038, "step": 66680 }, { "grad_norm": 0.07901029288768768, "learning_rate": 2.73903895329495e-05, "loss": 0.0025, "step": 66690 }, { "grad_norm": 0.07327196002006531, "learning_rate": 2.737564313297478e-05, "loss": 0.0021, "step": 66700 }, { "grad_norm": 0.07155632227659225, "learning_rate": 2.736089920716572e-05, "loss": 0.0029, "step": 66710 }, { "grad_norm": 0.051145076751708984, "learning_rate": 2.7346157757134706e-05, "loss": 0.0021, "step": 66720 }, { "grad_norm": 0.043743714690208435, "learning_rate": 2.733141878449379e-05, "loss": 0.0026, "step": 66730 }, { "grad_norm": 0.07684037834405899, "learning_rate": 2.7316682290854844e-05, "loss": 0.0034, "step": 66740 }, { "grad_norm": 0.06477805227041245, "learning_rate": 2.7301948277829414e-05, "loss": 0.0021, "step": 66750 }, { "grad_norm": 0.05960789695382118, "learning_rate": 2.7287216747028814e-05, "loss": 0.0024, "step": 66760 }, { "grad_norm": 0.05896301195025444, "learning_rate": 2.7272487700064025e-05, "loss": 0.0029, "step": 66770 }, { "grad_norm": 0.06300297379493713, "learning_rate": 2.725776113854583e-05, "loss": 0.0023, "step": 66780 }, { "grad_norm": 0.050842639058828354, "learning_rate": 2.7243037064084664e-05, "loss": 0.0025, "step": 66790 }, { "grad_norm": 0.06940463930368423, "learning_rate": 2.7228315478290766e-05, "loss": 0.0025, "step": 66800 }, { "grad_norm": 0.05564763396978378, "learning_rate": 2.721359638277402e-05, "loss": 0.002, "step": 66810 }, { "grad_norm": 0.04633207619190216, "learning_rate": 2.7198879779144148e-05, "loss": 0.002, "step": 66820 }, { "grad_norm": 0.07564835250377655, "learning_rate": 2.7184165669010485e-05, "loss": 0.0025, "step": 66830 }, { "grad_norm": 0.03854825720191002, "learning_rate": 2.716945405398218e-05, "loss": 0.002, "step": 66840 }, { "grad_norm": 0.06861714273691177, "learning_rate": 2.715474493566803e-05, "loss": 0.0036, "step": 66850 }, { "grad_norm": 0.07817194610834122, "learning_rate": 2.7140038315676652e-05, "loss": 0.0028, "step": 66860 }, { "grad_norm": 0.08550798892974854, "learning_rate": 2.7125334195616275e-05, "loss": 0.0026, "step": 66870 }, { "grad_norm": 0.05102824047207832, "learning_rate": 2.7110632577094997e-05, "loss": 0.0028, "step": 66880 }, { "grad_norm": 0.0688033401966095, "learning_rate": 2.709593346172051e-05, "loss": 0.0019, "step": 66890 }, { "grad_norm": 0.05239032581448555, "learning_rate": 2.708123685110032e-05, "loss": 0.0018, "step": 66900 }, { "grad_norm": 0.07459867000579834, "learning_rate": 2.7066542746841588e-05, "loss": 0.0022, "step": 66910 }, { "grad_norm": 0.06034648045897484, "learning_rate": 2.7051851150551267e-05, "loss": 0.0021, "step": 66920 }, { "grad_norm": 0.061899423599243164, "learning_rate": 2.7037162063835996e-05, "loss": 0.0021, "step": 66930 }, { "grad_norm": 0.060022395104169846, "learning_rate": 2.702247548830218e-05, "loss": 0.0026, "step": 66940 }, { "grad_norm": 0.06857238709926605, "learning_rate": 2.7007791425555883e-05, "loss": 0.0029, "step": 66950 }, { "grad_norm": 0.06388877332210541, "learning_rate": 2.6993109877202945e-05, "loss": 0.0036, "step": 66960 }, { "grad_norm": 0.06684061139822006, "learning_rate": 2.6978430844848935e-05, "loss": 0.002, "step": 66970 }, { "grad_norm": 0.0659150630235672, "learning_rate": 2.69637543300991e-05, "loss": 0.0027, "step": 66980 }, { "grad_norm": 0.042193543165922165, "learning_rate": 2.694908033455848e-05, "loss": 0.0021, "step": 66990 }, { "grad_norm": 0.08498218655586243, "learning_rate": 2.6934408859831733e-05, "loss": 0.0022, "step": 67000 }, { "grad_norm": 0.04271075129508972, "learning_rate": 2.69197399075234e-05, "loss": 0.0023, "step": 67010 }, { "grad_norm": 0.0907776728272438, "learning_rate": 2.6905073479237584e-05, "loss": 0.0028, "step": 67020 }, { "grad_norm": 0.05650918930768967, "learning_rate": 2.6890409576578235e-05, "loss": 0.0024, "step": 67030 }, { "grad_norm": 0.0726732686161995, "learning_rate": 2.687574820114892e-05, "loss": 0.0026, "step": 67040 }, { "grad_norm": 0.053562674671411514, "learning_rate": 2.686108935455305e-05, "loss": 0.0027, "step": 67050 }, { "grad_norm": 0.06287633627653122, "learning_rate": 2.6846433038393603e-05, "loss": 0.0026, "step": 67060 }, { "grad_norm": 0.0533057302236557, "learning_rate": 2.6831779254273482e-05, "loss": 0.0018, "step": 67070 }, { "grad_norm": 0.06550297141075134, "learning_rate": 2.6817128003795122e-05, "loss": 0.0032, "step": 67080 }, { "grad_norm": 0.08799513429403305, "learning_rate": 2.680247928856081e-05, "loss": 0.0027, "step": 67090 }, { "grad_norm": 0.07383903115987778, "learning_rate": 2.6787833110172466e-05, "loss": 0.0034, "step": 67100 }, { "grad_norm": 0.044794630259275436, "learning_rate": 2.6773189470231807e-05, "loss": 0.0033, "step": 67110 }, { "grad_norm": 0.0506121963262558, "learning_rate": 2.675854837034022e-05, "loss": 0.0028, "step": 67120 }, { "grad_norm": 0.060744449496269226, "learning_rate": 2.6743909812098833e-05, "loss": 0.0019, "step": 67130 }, { "grad_norm": 0.04676973819732666, "learning_rate": 2.67292737971085e-05, "loss": 0.0026, "step": 67140 }, { "grad_norm": 0.04760213568806648, "learning_rate": 2.671464032696982e-05, "loss": 0.0018, "step": 67150 }, { "grad_norm": 0.054836299270391464, "learning_rate": 2.6700009403283055e-05, "loss": 0.0027, "step": 67160 }, { "grad_norm": 0.07166089862585068, "learning_rate": 2.6685381027648214e-05, "loss": 0.0021, "step": 67170 }, { "grad_norm": 0.07796652615070343, "learning_rate": 2.6670755201665086e-05, "loss": 0.0021, "step": 67180 }, { "grad_norm": 0.09194142371416092, "learning_rate": 2.6656131926933066e-05, "loss": 0.0025, "step": 67190 }, { "grad_norm": 0.06537343561649323, "learning_rate": 2.664151120505136e-05, "loss": 0.0024, "step": 67200 }, { "grad_norm": 0.06887804716825485, "learning_rate": 2.6626893037618873e-05, "loss": 0.0027, "step": 67210 }, { "grad_norm": 0.04534941539168358, "learning_rate": 2.6612277426234243e-05, "loss": 0.0022, "step": 67220 }, { "grad_norm": 0.06949389725923538, "learning_rate": 2.659766437249577e-05, "loss": 0.0021, "step": 67230 }, { "grad_norm": 0.05010828748345375, "learning_rate": 2.6583053878001557e-05, "loss": 0.0019, "step": 67240 }, { "grad_norm": 0.046339251101017, "learning_rate": 2.6568445944349323e-05, "loss": 0.002, "step": 67250 }, { "grad_norm": 0.06233410909771919, "learning_rate": 2.6553840573136657e-05, "loss": 0.0024, "step": 67260 }, { "grad_norm": 0.059478338807821274, "learning_rate": 2.6539237765960706e-05, "loss": 0.0023, "step": 67270 }, { "grad_norm": 0.06524724513292313, "learning_rate": 2.6524637524418465e-05, "loss": 0.0023, "step": 67280 }, { "grad_norm": 0.07559829950332642, "learning_rate": 2.651003985010655e-05, "loss": 0.0023, "step": 67290 }, { "grad_norm": 0.05770433321595192, "learning_rate": 2.649544474462138e-05, "loss": 0.002, "step": 67300 }, { "grad_norm": 0.0682959035038948, "learning_rate": 2.6480852209559015e-05, "loss": 0.0028, "step": 67310 }, { "grad_norm": 0.06128743290901184, "learning_rate": 2.6466262246515296e-05, "loss": 0.0018, "step": 67320 }, { "grad_norm": 0.0615132674574852, "learning_rate": 2.6451674857085746e-05, "loss": 0.0024, "step": 67330 }, { "grad_norm": 0.06199907511472702, "learning_rate": 2.6437090042865655e-05, "loss": 0.0027, "step": 67340 }, { "grad_norm": 0.04851154237985611, "learning_rate": 2.6422507805449947e-05, "loss": 0.0021, "step": 67350 }, { "grad_norm": 0.044512487947940826, "learning_rate": 2.6407928146433357e-05, "loss": 0.0025, "step": 67360 }, { "grad_norm": 0.05058961361646652, "learning_rate": 2.639335106741026e-05, "loss": 0.0023, "step": 67370 }, { "grad_norm": 0.05527103319764137, "learning_rate": 2.6378776569974796e-05, "loss": 0.0021, "step": 67380 }, { "grad_norm": 0.05640365183353424, "learning_rate": 2.6364204655720814e-05, "loss": 0.0025, "step": 67390 }, { "grad_norm": 0.06439373642206192, "learning_rate": 2.6349635326241894e-05, "loss": 0.0035, "step": 67400 }, { "grad_norm": 0.06362926959991455, "learning_rate": 2.633506858313127e-05, "loss": 0.003, "step": 67410 }, { "grad_norm": 0.059032391756772995, "learning_rate": 2.6320504427981975e-05, "loss": 0.0031, "step": 67420 }, { "grad_norm": 0.07403331995010376, "learning_rate": 2.6305942862386734e-05, "loss": 0.0024, "step": 67430 }, { "grad_norm": 0.04156756401062012, "learning_rate": 2.629138388793793e-05, "loss": 0.0025, "step": 67440 }, { "grad_norm": 0.048761118203401566, "learning_rate": 2.6276827506227743e-05, "loss": 0.0033, "step": 67450 }, { "grad_norm": 0.06567883491516113, "learning_rate": 2.6262273718848037e-05, "loss": 0.0027, "step": 67460 }, { "grad_norm": 0.062177106738090515, "learning_rate": 2.6247722527390407e-05, "loss": 0.0023, "step": 67470 }, { "grad_norm": 0.06637828797101974, "learning_rate": 2.6233173933446104e-05, "loss": 0.0021, "step": 67480 }, { "grad_norm": 0.07752613723278046, "learning_rate": 2.6218627938606188e-05, "loss": 0.0022, "step": 67490 }, { "grad_norm": 0.0672261044383049, "learning_rate": 2.620408454446135e-05, "loss": 0.0023, "step": 67500 }, { "grad_norm": 0.046736110001802444, "learning_rate": 2.6189543752602046e-05, "loss": 0.002, "step": 67510 }, { "grad_norm": 0.04847179353237152, "learning_rate": 2.6175005564618444e-05, "loss": 0.0019, "step": 67520 }, { "grad_norm": 0.05162876099348068, "learning_rate": 2.6160469982100428e-05, "loss": 0.0021, "step": 67530 }, { "grad_norm": 0.043191585689783096, "learning_rate": 2.6145937006637555e-05, "loss": 0.002, "step": 67540 }, { "grad_norm": 0.05241096019744873, "learning_rate": 2.6131406639819166e-05, "loss": 0.0019, "step": 67550 }, { "grad_norm": 0.0647682175040245, "learning_rate": 2.611687888323424e-05, "loss": 0.003, "step": 67560 }, { "grad_norm": 0.05246301367878914, "learning_rate": 2.610235373847153e-05, "loss": 0.0022, "step": 67570 }, { "grad_norm": 0.048372942954301834, "learning_rate": 2.6087831207119494e-05, "loss": 0.0022, "step": 67580 }, { "grad_norm": 0.03927331045269966, "learning_rate": 2.60733112907663e-05, "loss": 0.0023, "step": 67590 }, { "grad_norm": 0.09289488941431046, "learning_rate": 2.6058793990999786e-05, "loss": 0.0026, "step": 67600 }, { "grad_norm": 0.1444113403558731, "learning_rate": 2.6044279309407593e-05, "loss": 0.002, "step": 67610 }, { "grad_norm": 0.05259755998849869, "learning_rate": 2.6029767247576976e-05, "loss": 0.0023, "step": 67620 }, { "grad_norm": 0.049031395465135574, "learning_rate": 2.601525780709497e-05, "loss": 0.0032, "step": 67630 }, { "grad_norm": 0.05788969248533249, "learning_rate": 2.6000750989548318e-05, "loss": 0.002, "step": 67640 }, { "grad_norm": 0.04554787278175354, "learning_rate": 2.5986246796523462e-05, "loss": 0.0026, "step": 67650 }, { "grad_norm": 0.06876665353775024, "learning_rate": 2.5971745229606546e-05, "loss": 0.0022, "step": 67660 }, { "grad_norm": 0.07278107106685638, "learning_rate": 2.5957246290383442e-05, "loss": 0.003, "step": 67670 }, { "grad_norm": 0.05225350335240364, "learning_rate": 2.594274998043975e-05, "loss": 0.0023, "step": 67680 }, { "grad_norm": 0.07003678381443024, "learning_rate": 2.5928256301360733e-05, "loss": 0.0027, "step": 67690 }, { "grad_norm": 0.061015188694000244, "learning_rate": 2.591376525473142e-05, "loss": 0.0029, "step": 67700 }, { "grad_norm": 0.06814375519752502, "learning_rate": 2.5899276842136523e-05, "loss": 0.0021, "step": 67710 }, { "grad_norm": 0.05301973223686218, "learning_rate": 2.5884791065160495e-05, "loss": 0.0025, "step": 67720 }, { "grad_norm": 0.04914341866970062, "learning_rate": 2.5870307925387436e-05, "loss": 0.0021, "step": 67730 }, { "grad_norm": 0.07069030404090881, "learning_rate": 2.585582742440125e-05, "loss": 0.0025, "step": 67740 }, { "grad_norm": 0.06266040354967117, "learning_rate": 2.5841349563785455e-05, "loss": 0.0031, "step": 67750 }, { "grad_norm": 0.07661467045545578, "learning_rate": 2.5826874345123355e-05, "loss": 0.0024, "step": 67760 }, { "grad_norm": 0.05528261885046959, "learning_rate": 2.581240176999792e-05, "loss": 0.0022, "step": 67770 }, { "grad_norm": 0.07003599405288696, "learning_rate": 2.579793183999189e-05, "loss": 0.0028, "step": 67780 }, { "grad_norm": 0.05788566917181015, "learning_rate": 2.5783464556687627e-05, "loss": 0.0023, "step": 67790 }, { "grad_norm": 0.036370184272527695, "learning_rate": 2.576899992166729e-05, "loss": 0.0019, "step": 67800 }, { "grad_norm": 0.07970186322927475, "learning_rate": 2.5754537936512668e-05, "loss": 0.0024, "step": 67810 }, { "grad_norm": 0.05115431174635887, "learning_rate": 2.574007860280535e-05, "loss": 0.0028, "step": 67820 }, { "grad_norm": 0.05926977097988129, "learning_rate": 2.5725621922126518e-05, "loss": 0.0034, "step": 67830 }, { "grad_norm": 0.044082872569561005, "learning_rate": 2.571116789605722e-05, "loss": 0.002, "step": 67840 }, { "grad_norm": 0.059051308780908585, "learning_rate": 2.5696716526178054e-05, "loss": 0.0023, "step": 67850 }, { "grad_norm": 0.06863421946763992, "learning_rate": 2.568226781406946e-05, "loss": 0.0021, "step": 67860 }, { "grad_norm": 0.05075230076909065, "learning_rate": 2.5667821761311474e-05, "loss": 0.0026, "step": 67870 }, { "grad_norm": 0.0724424198269844, "learning_rate": 2.5653378369483917e-05, "loss": 0.0025, "step": 67880 }, { "grad_norm": 0.051891833543777466, "learning_rate": 2.5638937640166292e-05, "loss": 0.0024, "step": 67890 }, { "grad_norm": 0.05006951466202736, "learning_rate": 2.5624499574937842e-05, "loss": 0.003, "step": 67900 }, { "grad_norm": 0.060434505343437195, "learning_rate": 2.5610064175377456e-05, "loss": 0.0026, "step": 67910 }, { "grad_norm": 0.04808475077152252, "learning_rate": 2.5595631443063782e-05, "loss": 0.0021, "step": 67920 }, { "grad_norm": 0.04556798189878464, "learning_rate": 2.5581201379575177e-05, "loss": 0.0021, "step": 67930 }, { "grad_norm": 0.050820641219615936, "learning_rate": 2.5566773986489655e-05, "loss": 0.0022, "step": 67940 }, { "grad_norm": 0.027577683329582214, "learning_rate": 2.5552349265385e-05, "loss": 0.0018, "step": 67950 }, { "grad_norm": 0.06993094086647034, "learning_rate": 2.5537927217838675e-05, "loss": 0.0025, "step": 67960 }, { "grad_norm": 0.10473601520061493, "learning_rate": 2.5523507845427864e-05, "loss": 0.0026, "step": 67970 }, { "grad_norm": 0.06455383449792862, "learning_rate": 2.5509091149729413e-05, "loss": 0.0033, "step": 67980 }, { "grad_norm": 0.04773185774683952, "learning_rate": 2.5494677132319962e-05, "loss": 0.0027, "step": 67990 }, { "grad_norm": 0.05790354311466217, "learning_rate": 2.548026579477575e-05, "loss": 0.0026, "step": 68000 }, { "grad_norm": 0.062263526022434235, "learning_rate": 2.5465857138672834e-05, "loss": 0.0022, "step": 68010 }, { "grad_norm": 0.05436711013317108, "learning_rate": 2.5451451165586848e-05, "loss": 0.0027, "step": 68020 }, { "grad_norm": 0.05929498001933098, "learning_rate": 2.543704787709329e-05, "loss": 0.0022, "step": 68030 }, { "grad_norm": 0.055329736322164536, "learning_rate": 2.542264727476723e-05, "loss": 0.0021, "step": 68040 }, { "grad_norm": 0.04530123621225357, "learning_rate": 2.5408249360183535e-05, "loss": 0.0022, "step": 68050 }, { "grad_norm": 0.059242356568574905, "learning_rate": 2.5393854134916694e-05, "loss": 0.0027, "step": 68060 }, { "grad_norm": 0.047400787472724915, "learning_rate": 2.5379461600540993e-05, "loss": 0.0019, "step": 68070 }, { "grad_norm": 0.048951636999845505, "learning_rate": 2.5365071758630317e-05, "loss": 0.0021, "step": 68080 }, { "grad_norm": 0.056076135486364365, "learning_rate": 2.535068461075839e-05, "loss": 0.0025, "step": 68090 }, { "grad_norm": 0.057655103504657745, "learning_rate": 2.5336300158498516e-05, "loss": 0.0024, "step": 68100 }, { "grad_norm": 0.053324513137340546, "learning_rate": 2.5321918403423793e-05, "loss": 0.0029, "step": 68110 }, { "grad_norm": 0.046414852142333984, "learning_rate": 2.5307539347106957e-05, "loss": 0.0026, "step": 68120 }, { "grad_norm": 0.06402938067913055, "learning_rate": 2.529316299112049e-05, "loss": 0.0023, "step": 68130 }, { "grad_norm": 0.05467269942164421, "learning_rate": 2.5278789337036567e-05, "loss": 0.0022, "step": 68140 }, { "grad_norm": 0.06203664839267731, "learning_rate": 2.52644183864271e-05, "loss": 0.002, "step": 68150 }, { "grad_norm": 0.0668409913778305, "learning_rate": 2.525005014086363e-05, "loss": 0.0023, "step": 68160 }, { "grad_norm": 0.05893171206116676, "learning_rate": 2.5235684601917465e-05, "loss": 0.0022, "step": 68170 }, { "grad_norm": 0.060814838856458664, "learning_rate": 2.5221321771159618e-05, "loss": 0.0027, "step": 68180 }, { "grad_norm": 0.05651212856173515, "learning_rate": 2.5206961650160743e-05, "loss": 0.0026, "step": 68190 }, { "grad_norm": 0.054333921521902084, "learning_rate": 2.519260424049129e-05, "loss": 0.0035, "step": 68200 }, { "grad_norm": 0.06114961579442024, "learning_rate": 2.5178249543721304e-05, "loss": 0.0025, "step": 68210 }, { "grad_norm": 0.07479935884475708, "learning_rate": 2.5163897561420668e-05, "loss": 0.0018, "step": 68220 }, { "grad_norm": 0.05339023470878601, "learning_rate": 2.514954829515883e-05, "loss": 0.002, "step": 68230 }, { "grad_norm": 0.061875682324171066, "learning_rate": 2.513520174650505e-05, "loss": 0.0022, "step": 68240 }, { "grad_norm": 0.041639283299446106, "learning_rate": 2.5120857917028207e-05, "loss": 0.0018, "step": 68250 }, { "grad_norm": 0.051368992775678635, "learning_rate": 2.5106516808296953e-05, "loss": 0.0023, "step": 68260 }, { "grad_norm": 0.057670708745718, "learning_rate": 2.5092178421879552e-05, "loss": 0.0021, "step": 68270 }, { "grad_norm": 0.08895961195230484, "learning_rate": 2.5077842759344116e-05, "loss": 0.0017, "step": 68280 }, { "grad_norm": 0.0476599857211113, "learning_rate": 2.5063509822258304e-05, "loss": 0.003, "step": 68290 }, { "grad_norm": 0.053101345896720886, "learning_rate": 2.5049179612189576e-05, "loss": 0.0027, "step": 68300 }, { "grad_norm": 0.04298372566699982, "learning_rate": 2.5034852130705043e-05, "loss": 0.002, "step": 68310 }, { "grad_norm": 0.07811571657657623, "learning_rate": 2.5020527379371572e-05, "loss": 0.0023, "step": 68320 }, { "grad_norm": 0.07011055946350098, "learning_rate": 2.500620535975564e-05, "loss": 0.0027, "step": 68330 }, { "grad_norm": 0.08051230013370514, "learning_rate": 2.4991886073423515e-05, "loss": 0.0023, "step": 68340 }, { "grad_norm": 0.06805826723575592, "learning_rate": 2.4977569521941134e-05, "loss": 0.002, "step": 68350 }, { "grad_norm": 0.04853127524256706, "learning_rate": 2.4963255706874143e-05, "loss": 0.0024, "step": 68360 }, { "grad_norm": 0.04187648370862007, "learning_rate": 2.4948944629787853e-05, "loss": 0.0021, "step": 68370 }, { "grad_norm": 0.058623477816581726, "learning_rate": 2.493463629224731e-05, "loss": 0.003, "step": 68380 }, { "grad_norm": 0.047732532024383545, "learning_rate": 2.492033069581728e-05, "loss": 0.0023, "step": 68390 }, { "grad_norm": 0.062003377825021744, "learning_rate": 2.4906027842062137e-05, "loss": 0.0023, "step": 68400 }, { "grad_norm": 0.05105701461434364, "learning_rate": 2.48917277325461e-05, "loss": 0.0021, "step": 68410 }, { "grad_norm": 0.0629589781165123, "learning_rate": 2.487743036883295e-05, "loss": 0.002, "step": 68420 }, { "grad_norm": 0.062167130410671234, "learning_rate": 2.486313575248626e-05, "loss": 0.0021, "step": 68430 }, { "grad_norm": 0.07910401374101639, "learning_rate": 2.4848843885069235e-05, "loss": 0.002, "step": 68440 }, { "grad_norm": 0.06255442649126053, "learning_rate": 2.483455476814484e-05, "loss": 0.0023, "step": 68450 }, { "grad_norm": 0.05333688110113144, "learning_rate": 2.4820268403275672e-05, "loss": 0.0023, "step": 68460 }, { "grad_norm": 0.0641593262553215, "learning_rate": 2.480598479202413e-05, "loss": 0.0025, "step": 68470 }, { "grad_norm": 0.07824418693780899, "learning_rate": 2.4791703935952193e-05, "loss": 0.0026, "step": 68480 }, { "grad_norm": 0.06939341127872467, "learning_rate": 2.477742583662163e-05, "loss": 0.0024, "step": 68490 }, { "grad_norm": 0.07141284644603729, "learning_rate": 2.476315049559384e-05, "loss": 0.0029, "step": 68500 }, { "grad_norm": 0.059316277503967285, "learning_rate": 2.4748877914429987e-05, "loss": 0.0026, "step": 68510 }, { "grad_norm": 0.03893086314201355, "learning_rate": 2.4734608094690857e-05, "loss": 0.0019, "step": 68520 }, { "grad_norm": 0.031027525663375854, "learning_rate": 2.4720341037937006e-05, "loss": 0.0016, "step": 68530 }, { "grad_norm": 0.12440204620361328, "learning_rate": 2.4706076745728645e-05, "loss": 0.0024, "step": 68540 }, { "grad_norm": 0.06208639219403267, "learning_rate": 2.469181521962573e-05, "loss": 0.0022, "step": 68550 }, { "grad_norm": 0.0511569119989872, "learning_rate": 2.467755646118783e-05, "loss": 0.002, "step": 68560 }, { "grad_norm": 0.054260753095149994, "learning_rate": 2.46633004719743e-05, "loss": 0.0027, "step": 68570 }, { "grad_norm": 0.13617298007011414, "learning_rate": 2.4649047253544123e-05, "loss": 0.0027, "step": 68580 }, { "grad_norm": 0.06508111208677292, "learning_rate": 2.4634796807456016e-05, "loss": 0.0024, "step": 68590 }, { "grad_norm": 0.05221117287874222, "learning_rate": 2.46205491352684e-05, "loss": 0.0023, "step": 68600 }, { "grad_norm": 0.03946975618600845, "learning_rate": 2.4606304238539375e-05, "loss": 0.0021, "step": 68610 }, { "grad_norm": 0.04497429355978966, "learning_rate": 2.4592062118826753e-05, "loss": 0.0018, "step": 68620 }, { "grad_norm": 0.049214866012334824, "learning_rate": 2.4577822777688004e-05, "loss": 0.0022, "step": 68630 }, { "grad_norm": 0.08848001807928085, "learning_rate": 2.4563586216680347e-05, "loss": 0.0023, "step": 68640 }, { "grad_norm": 0.052797865122556686, "learning_rate": 2.4549352437360622e-05, "loss": 0.0023, "step": 68650 }, { "grad_norm": 0.04884320870041847, "learning_rate": 2.4535121441285493e-05, "loss": 0.0024, "step": 68660 }, { "grad_norm": 0.037950415164232254, "learning_rate": 2.4520893230011172e-05, "loss": 0.0025, "step": 68670 }, { "grad_norm": 0.043771643191576004, "learning_rate": 2.450666780509368e-05, "loss": 0.0021, "step": 68680 }, { "grad_norm": 0.04116211086511612, "learning_rate": 2.4492445168088656e-05, "loss": 0.0025, "step": 68690 }, { "grad_norm": 0.06019086390733719, "learning_rate": 2.447822532055149e-05, "loss": 0.0027, "step": 68700 }, { "grad_norm": 0.0495084747672081, "learning_rate": 2.4464008264037212e-05, "loss": 0.0031, "step": 68710 }, { "grad_norm": 0.05687027424573898, "learning_rate": 2.4449794000100605e-05, "loss": 0.002, "step": 68720 }, { "grad_norm": 0.043411917984485626, "learning_rate": 2.4435582530296108e-05, "loss": 0.0023, "step": 68730 }, { "grad_norm": 0.04710351303219795, "learning_rate": 2.4421373856177887e-05, "loss": 0.0024, "step": 68740 }, { "grad_norm": 0.05289583280682564, "learning_rate": 2.440716797929975e-05, "loss": 0.0021, "step": 68750 }, { "grad_norm": 0.05589192733168602, "learning_rate": 2.439296490121526e-05, "loss": 0.002, "step": 68760 }, { "grad_norm": 0.06238916888833046, "learning_rate": 2.4378764623477608e-05, "loss": 0.0023, "step": 68770 }, { "grad_norm": 0.03961126133799553, "learning_rate": 2.436456714763974e-05, "loss": 0.0029, "step": 68780 }, { "grad_norm": 0.045784685760736465, "learning_rate": 2.4350372475254262e-05, "loss": 0.0021, "step": 68790 }, { "grad_norm": 0.04545567184686661, "learning_rate": 2.433618060787351e-05, "loss": 0.0019, "step": 68800 }, { "grad_norm": 0.04062729701399803, "learning_rate": 2.432199154704944e-05, "loss": 0.002, "step": 68810 }, { "grad_norm": 0.04356148838996887, "learning_rate": 2.430780529433377e-05, "loss": 0.0025, "step": 68820 }, { "grad_norm": 0.05941466987133026, "learning_rate": 2.42936218512779e-05, "loss": 0.0028, "step": 68830 }, { "grad_norm": 0.07314388453960419, "learning_rate": 2.427944121943288e-05, "loss": 0.0026, "step": 68840 }, { "grad_norm": 0.045986201614141464, "learning_rate": 2.4265263400349497e-05, "loss": 0.0021, "step": 68850 }, { "grad_norm": 0.0879591852426529, "learning_rate": 2.4251088395578213e-05, "loss": 0.0021, "step": 68860 }, { "grad_norm": 0.04188130795955658, "learning_rate": 2.4236916206669214e-05, "loss": 0.0017, "step": 68870 }, { "grad_norm": 0.07415197044610977, "learning_rate": 2.4222746835172304e-05, "loss": 0.0021, "step": 68880 }, { "grad_norm": 0.051582496613264084, "learning_rate": 2.4208580282637065e-05, "loss": 0.0025, "step": 68890 }, { "grad_norm": 0.06299480050802231, "learning_rate": 2.4194416550612687e-05, "loss": 0.0024, "step": 68900 }, { "grad_norm": 0.07515306025743484, "learning_rate": 2.418025564064812e-05, "loss": 0.0025, "step": 68910 }, { "grad_norm": 0.05552281066775322, "learning_rate": 2.4166097554291978e-05, "loss": 0.0021, "step": 68920 }, { "grad_norm": 0.052582915872335434, "learning_rate": 2.4151942293092583e-05, "loss": 0.0018, "step": 68930 }, { "grad_norm": 0.04967055842280388, "learning_rate": 2.41377898585979e-05, "loss": 0.0023, "step": 68940 }, { "grad_norm": 0.0597202405333519, "learning_rate": 2.4123640252355652e-05, "loss": 0.0027, "step": 68950 }, { "grad_norm": 0.07643215358257294, "learning_rate": 2.410949347591318e-05, "loss": 0.002, "step": 68960 }, { "grad_norm": 0.05602199584245682, "learning_rate": 2.4095349530817578e-05, "loss": 0.0023, "step": 68970 }, { "grad_norm": 0.037467170506715775, "learning_rate": 2.4081208418615603e-05, "loss": 0.0021, "step": 68980 }, { "grad_norm": 0.04189116880297661, "learning_rate": 2.4067070140853725e-05, "loss": 0.0022, "step": 68990 }, { "grad_norm": 0.06165102496743202, "learning_rate": 2.405293469907805e-05, "loss": 0.0023, "step": 69000 }, { "grad_norm": 0.07245092839002609, "learning_rate": 2.4038802094834446e-05, "loss": 0.0023, "step": 69010 }, { "grad_norm": 0.051576677709817886, "learning_rate": 2.4024672329668395e-05, "loss": 0.0025, "step": 69020 }, { "grad_norm": 0.04607393220067024, "learning_rate": 2.4010545405125125e-05, "loss": 0.0023, "step": 69030 }, { "grad_norm": 0.057711441069841385, "learning_rate": 2.3996421322749528e-05, "loss": 0.0029, "step": 69040 }, { "grad_norm": 0.05955936759710312, "learning_rate": 2.3982300084086224e-05, "loss": 0.0021, "step": 69050 }, { "grad_norm": 0.039203375577926636, "learning_rate": 2.396818169067944e-05, "loss": 0.0017, "step": 69060 }, { "grad_norm": 0.0523466020822525, "learning_rate": 2.3954066144073178e-05, "loss": 0.0021, "step": 69070 }, { "grad_norm": 0.04415791109204292, "learning_rate": 2.39399534458111e-05, "loss": 0.002, "step": 69080 }, { "grad_norm": 0.06917636841535568, "learning_rate": 2.392584359743651e-05, "loss": 0.0025, "step": 69090 }, { "grad_norm": 0.06026550382375717, "learning_rate": 2.3911736600492463e-05, "loss": 0.0025, "step": 69100 }, { "grad_norm": 0.05153242126107216, "learning_rate": 2.3897632456521672e-05, "loss": 0.0024, "step": 69110 }, { "grad_norm": 0.06206776574254036, "learning_rate": 2.388353116706658e-05, "loss": 0.0026, "step": 69120 }, { "grad_norm": 0.06091291457414627, "learning_rate": 2.3869432733669232e-05, "loss": 0.0017, "step": 69130 }, { "grad_norm": 0.04137445241212845, "learning_rate": 2.385533715787145e-05, "loss": 0.0019, "step": 69140 }, { "grad_norm": 0.05680107697844505, "learning_rate": 2.3841244441214673e-05, "loss": 0.002, "step": 69150 }, { "grad_norm": 0.053052760660648346, "learning_rate": 2.3827154585240073e-05, "loss": 0.0027, "step": 69160 }, { "grad_norm": 0.06318061053752899, "learning_rate": 2.3813067591488498e-05, "loss": 0.0019, "step": 69170 }, { "grad_norm": 0.05332854762673378, "learning_rate": 2.3798983461500494e-05, "loss": 0.0029, "step": 69180 }, { "grad_norm": 0.04376162216067314, "learning_rate": 2.3784902196816256e-05, "loss": 0.0021, "step": 69190 }, { "grad_norm": 0.05623023957014084, "learning_rate": 2.3770823798975716e-05, "loss": 0.002, "step": 69200 }, { "grad_norm": 0.030510850250720978, "learning_rate": 2.3756748269518424e-05, "loss": 0.0019, "step": 69210 }, { "grad_norm": 0.06048944965004921, "learning_rate": 2.374267560998371e-05, "loss": 0.0022, "step": 69220 }, { "grad_norm": 0.04080214351415634, "learning_rate": 2.3728605821910477e-05, "loss": 0.0023, "step": 69230 }, { "grad_norm": 0.07840926945209503, "learning_rate": 2.3714538906837453e-05, "loss": 0.0028, "step": 69240 }, { "grad_norm": 0.05395852401852608, "learning_rate": 2.3700474866302913e-05, "loss": 0.0023, "step": 69250 }, { "grad_norm": 0.06114889308810234, "learning_rate": 2.3686413701844917e-05, "loss": 0.0027, "step": 69260 }, { "grad_norm": 0.05705031380057335, "learning_rate": 2.367235541500114e-05, "loss": 0.0021, "step": 69270 }, { "grad_norm": 0.06112752854824066, "learning_rate": 2.3658300007308993e-05, "loss": 0.0028, "step": 69280 }, { "grad_norm": 0.049115631729364395, "learning_rate": 2.364424748030555e-05, "loss": 0.0022, "step": 69290 }, { "grad_norm": 0.041185129433870316, "learning_rate": 2.3630197835527595e-05, "loss": 0.0018, "step": 69300 }, { "grad_norm": 0.06008776277303696, "learning_rate": 2.3616151074511546e-05, "loss": 0.0024, "step": 69310 }, { "grad_norm": 0.047694556415081024, "learning_rate": 2.3602107198793545e-05, "loss": 0.0022, "step": 69320 }, { "grad_norm": 0.053509172052145004, "learning_rate": 2.3588066209909427e-05, "loss": 0.002, "step": 69330 }, { "grad_norm": 0.06753462553024292, "learning_rate": 2.3574028109394662e-05, "loss": 0.0031, "step": 69340 }, { "grad_norm": 0.0778406485915184, "learning_rate": 2.3559992898784454e-05, "loss": 0.0026, "step": 69350 }, { "grad_norm": 0.05205363780260086, "learning_rate": 2.354596057961367e-05, "loss": 0.0024, "step": 69360 }, { "grad_norm": 0.051921240985393524, "learning_rate": 2.353193115341688e-05, "loss": 0.0023, "step": 69370 }, { "grad_norm": 0.04081685096025467, "learning_rate": 2.3517904621728294e-05, "loss": 0.0021, "step": 69380 }, { "grad_norm": 0.05035002902150154, "learning_rate": 2.3503880986081855e-05, "loss": 0.002, "step": 69390 }, { "grad_norm": 0.06346365809440613, "learning_rate": 2.3489860248011136e-05, "loss": 0.0027, "step": 69400 }, { "grad_norm": 0.047542739659547806, "learning_rate": 2.3475842409049464e-05, "loss": 0.002, "step": 69410 }, { "grad_norm": 0.04461589455604553, "learning_rate": 2.346182747072975e-05, "loss": 0.002, "step": 69420 }, { "grad_norm": 0.06902729719877243, "learning_rate": 2.3447815434584718e-05, "loss": 0.0021, "step": 69430 }, { "grad_norm": 0.060280993580818176, "learning_rate": 2.3433806302146656e-05, "loss": 0.002, "step": 69440 }, { "grad_norm": 0.07867702096700668, "learning_rate": 2.3419800074947612e-05, "loss": 0.0023, "step": 69450 }, { "grad_norm": 0.08142132312059402, "learning_rate": 2.3405796754519248e-05, "loss": 0.0035, "step": 69460 }, { "grad_norm": 0.05724227428436279, "learning_rate": 2.3391796342392992e-05, "loss": 0.003, "step": 69470 }, { "grad_norm": 0.10644937306642532, "learning_rate": 2.337779884009984e-05, "loss": 0.0023, "step": 69480 }, { "grad_norm": 0.0881405919790268, "learning_rate": 2.3363804249170624e-05, "loss": 0.0025, "step": 69490 }, { "grad_norm": 0.05688821151852608, "learning_rate": 2.3349812571135703e-05, "loss": 0.0022, "step": 69500 }, { "grad_norm": 0.08145352452993393, "learning_rate": 2.333582380752523e-05, "loss": 0.0021, "step": 69510 }, { "grad_norm": 0.0496055893599987, "learning_rate": 2.3321837959868963e-05, "loss": 0.0019, "step": 69520 }, { "grad_norm": 0.0620763897895813, "learning_rate": 2.3307855029696383e-05, "loss": 0.0021, "step": 69530 }, { "grad_norm": 0.05411262810230255, "learning_rate": 2.3293875018536648e-05, "loss": 0.0029, "step": 69540 }, { "grad_norm": 0.050127286463975906, "learning_rate": 2.327989792791861e-05, "loss": 0.0023, "step": 69550 }, { "grad_norm": 0.053168073296546936, "learning_rate": 2.326592375937074e-05, "loss": 0.0025, "step": 69560 }, { "grad_norm": 0.06362290680408478, "learning_rate": 2.3251952514421248e-05, "loss": 0.0021, "step": 69570 }, { "grad_norm": 0.05125607177615166, "learning_rate": 2.323798419459804e-05, "loss": 0.0019, "step": 69580 }, { "grad_norm": 0.04186088964343071, "learning_rate": 2.3224018801428625e-05, "loss": 0.002, "step": 69590 }, { "grad_norm": 0.06224272772669792, "learning_rate": 2.3210056336440273e-05, "loss": 0.0027, "step": 69600 }, { "grad_norm": 0.05517403781414032, "learning_rate": 2.3196096801159844e-05, "loss": 0.0024, "step": 69610 }, { "grad_norm": 0.07640339434146881, "learning_rate": 2.3182140197114012e-05, "loss": 0.0021, "step": 69620 }, { "grad_norm": 0.0481409914791584, "learning_rate": 2.3168186525828985e-05, "loss": 0.0027, "step": 69630 }, { "grad_norm": 0.06747792661190033, "learning_rate": 2.3154235788830757e-05, "loss": 0.0029, "step": 69640 }, { "grad_norm": 0.06749806553125381, "learning_rate": 2.3140287987644927e-05, "loss": 0.0022, "step": 69650 }, { "grad_norm": 0.09017924964427948, "learning_rate": 2.3126343123796833e-05, "loss": 0.0021, "step": 69660 }, { "grad_norm": 0.08944129198789597, "learning_rate": 2.3112401198811418e-05, "loss": 0.0019, "step": 69670 }, { "grad_norm": 0.048711784183979034, "learning_rate": 2.3098462214213418e-05, "loss": 0.0031, "step": 69680 }, { "grad_norm": 0.05952563136816025, "learning_rate": 2.308452617152712e-05, "loss": 0.0031, "step": 69690 }, { "grad_norm": 0.055745769292116165, "learning_rate": 2.3070593072276603e-05, "loss": 0.0028, "step": 69700 }, { "grad_norm": 0.07910425215959549, "learning_rate": 2.3056662917985518e-05, "loss": 0.0024, "step": 69710 }, { "grad_norm": 0.0856800526380539, "learning_rate": 2.3042735710177283e-05, "loss": 0.0027, "step": 69720 }, { "grad_norm": 0.0628991350531578, "learning_rate": 2.302881145037492e-05, "loss": 0.0019, "step": 69730 }, { "grad_norm": 0.046999696642160416, "learning_rate": 2.3014890140101192e-05, "loss": 0.0023, "step": 69740 }, { "grad_norm": 0.0710316151380539, "learning_rate": 2.30009717808785e-05, "loss": 0.002, "step": 69750 }, { "grad_norm": 0.05182887986302376, "learning_rate": 2.2987056374228967e-05, "loss": 0.0021, "step": 69760 }, { "grad_norm": 0.04526283219456673, "learning_rate": 2.297314392167431e-05, "loss": 0.0018, "step": 69770 }, { "grad_norm": 0.048687875270843506, "learning_rate": 2.295923442473601e-05, "loss": 0.002, "step": 69780 }, { "grad_norm": 0.048406682908535004, "learning_rate": 2.294532788493519e-05, "loss": 0.0019, "step": 69790 }, { "grad_norm": 0.0391528457403183, "learning_rate": 2.293142430379262e-05, "loss": 0.002, "step": 69800 }, { "grad_norm": 0.05135766416788101, "learning_rate": 2.291752368282879e-05, "loss": 0.0019, "step": 69810 }, { "grad_norm": 0.05380572751164436, "learning_rate": 2.290362602356385e-05, "loss": 0.0019, "step": 69820 }, { "grad_norm": 0.06833873689174652, "learning_rate": 2.288973132751765e-05, "loss": 0.002, "step": 69830 }, { "grad_norm": 0.06072455272078514, "learning_rate": 2.287583959620965e-05, "loss": 0.0021, "step": 69840 }, { "grad_norm": 0.04976826161146164, "learning_rate": 2.2861950831159074e-05, "loss": 0.0024, "step": 69850 }, { "grad_norm": 0.053033411502838135, "learning_rate": 2.284806503388472e-05, "loss": 0.0023, "step": 69860 }, { "grad_norm": 0.059203967452049255, "learning_rate": 2.283418220590518e-05, "loss": 0.0016, "step": 69870 }, { "grad_norm": 0.04883643239736557, "learning_rate": 2.282030234873862e-05, "loss": 0.0025, "step": 69880 }, { "grad_norm": 0.05825231969356537, "learning_rate": 2.2806425463902948e-05, "loss": 0.0024, "step": 69890 }, { "grad_norm": 0.11678364127874374, "learning_rate": 2.279255155291568e-05, "loss": 0.0022, "step": 69900 }, { "grad_norm": 0.051528461277484894, "learning_rate": 2.277868061729409e-05, "loss": 0.0016, "step": 69910 }, { "grad_norm": 0.048129916191101074, "learning_rate": 2.276481265855504e-05, "loss": 0.002, "step": 69920 }, { "grad_norm": 0.05998639762401581, "learning_rate": 2.2750947678215128e-05, "loss": 0.0018, "step": 69930 }, { "grad_norm": 0.06505908817052841, "learning_rate": 2.2737085677790616e-05, "loss": 0.0028, "step": 69940 }, { "grad_norm": 0.08012817054986954, "learning_rate": 2.2723226658797443e-05, "loss": 0.0022, "step": 69950 }, { "grad_norm": 0.04793289303779602, "learning_rate": 2.270937062275117e-05, "loss": 0.0023, "step": 69960 }, { "grad_norm": 0.05240757018327713, "learning_rate": 2.2695517571167113e-05, "loss": 0.002, "step": 69970 }, { "grad_norm": 0.055111050605773926, "learning_rate": 2.268166750556019e-05, "loss": 0.0024, "step": 69980 }, { "grad_norm": 0.03870537504553795, "learning_rate": 2.266782042744503e-05, "loss": 0.0019, "step": 69990 }, { "grad_norm": 0.05967156961560249, "learning_rate": 2.2653976338335935e-05, "loss": 0.002, "step": 70000 } ], "logging_steps": 10, "max_steps": 100000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 10000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 128, "trial_name": null, "trial_params": null }